Annual Review of Immunology Volume 24 2006

Contents ARI 7 February 2006 11:41 Contents Annual Review of Immunology Volume 24, 2006 Annu. Rev. Immunol. 2006.2...

Author: Annual Reviews

36 downloads 1845 Views 10MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ARI

7 February 2006

11:47

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Boardpage

The author in his ofﬁce holding an enlargement of the structure of the MHC class II/HA 306-318 complex, from Stern et al. (71).

ANRV270-IY24-01

ARI

13 February 2006

11:14

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger Department of Molecular and Cellular Biology, Harvard University, Cambridge, Massachusetts 02138; email: [email protected]

Annu. Rev. Immunol. 2006. 24:1–31 First published online as a Review in Advance on November 28, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090703 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0001$20.00

Key Words bacterial cell walls, penicillin, histocompatibility proteins, MHC, HLA, NK cells

Abstract After starting out to become a physician, by a series of accidents I found myself at NIH in 1951 during its most productive growth phase. At age 26, I had a fully funded, independent laboratory and did not know what to work on. With advice from colleagues, I initiated a study of how penicillin kills bacteria. Twenty years later, my lab had outlined the structure and biosynthesis of the peptidoglycan of bacterial cell walls and had discovered that penicillin inhibited the terminal step in its biosynthesis catalyzed by transpeptidases. I then switched ﬁelds, moving to Harvard in 1968 and beginning the study of human HLA proteins. Twenty-ﬁve years later, the last half of which was spent in a stimulating collaboration with the late Don Wiley, our labs had isolated, crystallized, and elucidated the three-dimensional structures of these molecules and shown that their principal function was to present peptides to the immune system in initiating an immune response. More recently, the laboratory has focused on natural killer cells and their roles in peripheral blood and in the pregnant uterine decidua. It has been a wonderful scientiﬁc journey.

1

ANRV270-IY24-01

ARI

13 February 2006

11:14

EARLIEST MEMORIES 1925–1942

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

I was born in 1925, the ﬁrst of three brothers whose grandparents had all emigrated in the 1890s from Eastern Europe: Belarus, Ukraine, and Lithuania. My grandparents met in the great melting pot of immigrants, New York City. I am not absolutely certain of the origin of the paternal grandfather who gave me my surname. The best memories in our family say it was Czernowitz, which is now in the southwest corner of Ukraine but was then part of the Austro-Hungarian Empire; at the time of his naturalization, he renounced allegiance to Franz Josef I, Emperor of Austria and Pontiﬁcal King of Hungary. He was ﬁrst employed as a carpenter in the growing city, but by the time I was born he was the relatively prosperous owner of a construction company. He built the building at 3320 Ditmars Blvd. in Astoria, Queens, a poor working-class area, as a graduation gift to his only son, my father, who was a dentist. My father never sent a patient a bill or asked for payment, maybe in gratitude for his family’s good fortune. I can remember patients returning years later to pay their debts. My earliest memory is of my father’s ofﬁce and waiting room on the second ﬂoor, which are still there. Behind the waiting room was our family apartment. On the ground ﬂoor were a drug store and tailor shop, both sources of rental income for our family. On the third ﬂoor was a small apartment for a couple who cared for the property and cleaned my father’s ofﬁce. By the time I started school, we had moved to an apartment in a row of two-story buildings across the street from the school I ﬁrst attended, quite close to the Steinway piano factory. We soon moved to another rental in a two-family house in Flushing. Then my grandfather built two adjacent homes for himself and my parents and their growing family. On many evenings, my father, my grandfather, and his two sons-in-law gathered to play pinochle and smoke cigars. There were no television sets then; I suppose that today families gather similarly to watch TV. I attended

2

Strominger

P.S. 32 and then Bayside High School. In sixth grade I was moved ahead a year, and in high school I graduated second in my class by a slim margin. I was surprised but extremely pleased when I was admitted to Harvard, despite its known quota on admission of Jewish students. I have no idea what made me stand out. My two brothers went to Yale and to the University of Chicago. Both became university professors, one of pediatrics at Washington University in St. Louis and the other of neuroscience at Albany Medical College.

HARVARD AND CHELSEA NAVAL HOSPITAL, 1942–1944 I began college in the summer of 1942, only a few months after Pearl Harbor and the engagement of the United States in World War II. I was not yet 17 and was too young to be drafted or to enlist in the armed service training program at Harvard or other universities. Fifteen months later, however, I enlisted in the Navy V-12 program at Harvard College. Kirkland House, where I lived, had been commandeered by the V-12 program. I have vivid memories of that period—up at dawn to run a three-mile loop around the bridges that cross the Charles River. Lights were out at dark because of the fear of attack, presumably from enemy submarines. That did not give much time for studying, so after dark Alex Rich (who lived in the same entryway and remains a close friend) and I would sneak under the stairs at the basement level with a blanket over our heads and read by ﬂashlight. Harvard had three semesters per year and no summer vacation. I was only at Harvard for about 20 months, but in that time I completed 29 of the 32 courses required for graduation. I majored in psychology while taking the minimum number of courses that would satisfy the requirements for entrance to medical school. I was not much interested in biology or the other natural sciences, and I thought that psychology would be an excellent preparation for

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

medical school (or, as a I later realized, for running a laboratory). I worked as a volunteer at Massachusetts General Hospital and was on duty the night of the Coconut Grove ﬁre that claimed nearly 500 lives. I can still see the charred bodies lining a main corridor of the hospital. I was at Harvard for only ﬁve and a half terms. I do not know how I managed to take so many courses in that period, but the last ﬁve were easy. Students were continuously being sent away for military service, so the university had a policy that if you completed the midterm of a course, you got full credit. As soon as the Navy realized in March 1944 that I had completed my premedical requirements, they assigned me as a corpsman to Chelsea Naval Hospital across the Mystic River. It still stands as a condominium complex today. At the naval hospital, I had the top deck of a triple-decker bunk; one morning I broke my big toe leaping out of bed when I realized that the commandant was coming up the stairs for an inspection. I was assigned to the plaster room, where my job was to roll plaster after various kinds of broken bones had been set by the naval doctor. The plaster room was run by a marine who had returned from the South Paciﬁc and was often sufﬁciently inebriated so that the running of the plaster room was left to me and another V-12 corpsman (Sig Gunderson). I was the oldest grandchild on both sides of the family and the family’s ambition was that I become a doctor. It is a familiar story of upwardly mobile immigrant families that I see repeated today in the children of recent Asian immigrants in their quest for both economic stability and a useful humanitarian life. I had no interest in science and no exposure to it. U.S. Navy policy in those days was that a V-12 corpsman could apply to medical school; those who were admitted were sent to school as naval trainees. I was admitted to Yale Medical School in September 1944, after only ﬁve months at Chelsea Naval Hospital.

YALE MEDICAL SCHOOL, 1944–1948 Yale Medical School was then and is now an unusual place. My class had only 44 students, so I knew all of them; a third became close friends. There were no examinations and no forced attendance at any class. The only requirements were to pass Part I of the Examination of the National Board of Medical Examiners (Basic Science) after two years and Part II (Clinical Science) after four years. In addition, an original thesis was required at graduation. We took the usual courses in the ﬁrst two years. During the ﬁrst year, while taking physiology, I met John Brobeck, then an assistant professor in physiology. He was engaged in the study of hypothalamic hyperphagia induced by lesions in the ventromedial nucleus of the hypothalamus placed in rats with electrodes attached to the stereotactic Horsley-Clarke instrument. He needed someone to help him by measuring the food intake of these rats and their controls on various diets and with various activities, to weigh them, and later to help in placing lesions. Medical school seemed a little cut and dry, and, although I am not sure how, I got the job. It was simple, but I liked it. It allowed me to escape the routine of medical school, and John was a stimulating mentor. The job lasted through medical school, and I wrote my thesis on the topic. I was the ﬁrst recipient of the Borden Undergraduate Research Award at Yale Medical School. I wrote four papers on this topic, none of them startling, but I was ﬁrst author of three and sole author of one, the ﬁrst entry in my publication list. I was also the ﬁrst author of an abstract, presented at the Federation meetings, but I do not remember whether I presented it or John did. I was hooked. Research was fun, and there was a small conference room where I played bridge with my friends and any nursing students we could persuade to join us. Female medical students were very rare. A seventh paper also early appears in my publication list, entitled “A National Science

www.annualreviews.org • My Scientific Journey

3

ARI

13 February 2006

11:14

Foundation.” It is a summary of the congressional debate regarding whether to establish NSF. It was written in connection with my membership in the Association of Interns and Medical Students (AIMS) in which the Yale chapter was very active, discussed more below. The paper was republished recently as a “classic” by the Yale Journal of Biology and Medicine in which it originally appeared (1). I am not sure whether it is considered a classic because of what it said or because its author had by then become a prominent scientist. It was wartime when I began medical school. I was still in the Navy, but our uniforms had changed from those of seamen to those of ofﬁcers in training. Medical school was on an accelerated basis, i.e., three semesters per year, so that eight semesters, or four years, could be completed in slightly less than three years. Suddenly, in 1945, the war ended, ﬁrst in Europe and then in Asia. I completed my second year 16 months after I had started, in March 1946, and I was discharged from the Navy. The third year would not begin again on the regular schedule until September of that year. I had six months with nothing to do. I decided on an adventure. Another assistant professor in physiology, Bob Livingston, was from Seattle, and his family knew the owner of a logging company based in Seattle but operating in British Columbia, the O’Brien Logging Company. He got me a job with the company. I thought it would be great fun, but John Brobeck did not agree. He made me a deal: If I worked for his friend, Leo Samuels, the chairman of the Biochemistry Department at the relatively new Medical School of the University of Utah in Salt Lake City, for three of the six months, John would help with my transportation expenses. Salt Lake City also seemed like a great adventure to someone who had never been off the East Coast, so I took him up on the offer. My job was to measure the body fat of rats. This involved macerating the whole rat in a meat grinder, extracting the lipids, and then titrat-

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

4

Strominger

ing the fatty acids with sodium hydroxide, my ﬁrst experience in biochemistry. For the remaining three months, I went to Seattle and then on to the coast of British Columbia, somewhere between Stillwater and the Queen Charlotte Islands, where the O’Brien Company was operating. Its owner was extremely concerned that a medical student could hurt his hands in the logging operation, so to my extreme disappointment they put me to work washing dishes for the crew. It took me several weeks to talk my way out of that. I was then promoted to pounding spikes on a railroad being built from the site of the logging camp to the coast, so that the logs could be taken down to the sea and ﬂoated. Finally, after another few weeks, I managed to get where I wanted to be, working in the crew that was harvesting gigantic cedar trees, 6 to 10 ft in diameter. My job was to “choke” the log that had been topped by the “toppers” and other crews that had then felled the gigantic trees and cut them into sections of about 20 ft. Choking the log meant hauling a small steel cable wire from the top of a crane that had been erected at a collecting site, to a log, then around it and back to the crane. A very heavy chain could then be pulled out by the wire to “choke” the log, i.e., to fasten the heavy chain around the log. The log would then be pulled back to a pile at the site of the crane and eventually loaded on a railroad car bound for the sea. It was a solitary, isolated life among men of a totally different background than I had experienced. I had a lot of time to think. Medical school began again in September with the clinical courses. I remember only a few things from those two years. Francis G. Blake was professor of medicine and a prominent American physician. While I was taking my medicine clerkship in 1946, he received a new drug to administer to a patient at the Yale–New Haven Hospital. It was called penicillin, and it was said to cure infectious disease. This was given to a patient at the hospital with miraculous results; I believe it was the ﬁrst patient treated with penicillin in the

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

United States. It made a deep and lasting impression on me. I also delivered several babies, or more precisely, I caught them as they were given birth by multiparous women. During my surgery clerkship, I rebuilt a sailboat from a hull that had been washed up on Lighthouse Beach during the famous hurricane of 1938; the rigging and sails were intact in a nearby barn. I continued to work with John Brobeck. I passed Part II of the National Board of Medical Examiners, ﬁnishing second or third in my class, to the surprise of the faculty, my fellow classmates, and myself.

INTERNSHIP AND POSTDOCTORAL FELLOWSHIP AT WASHINGTON UNIVERSITY SCHOOL OF MEDICINE, 1948–1951 I had obtained my internship in internal medicine on the Ward Medical Service at Barnes Hospital in St. Louis afﬁliated with Washington University School of Medicine. The professor of medicine, W. Barry Wood, had been appointed in his early thirties and was one of the young stars of American medicine. Earlier, he had been an allAmerican football player at Harvard College (and liked to demonstrate his football prowess by inviting all the interns and residents on Sundays for a pickup game of touch football at his beautiful suburban residence). He was also extremely interested in research. I did not like being an intern. It seemed very routine. I decided not to pursue a residency but instead to become a postdoctoral fellow. On the advice of the dean of Yale Medical School, C.N.H. Long, a biochemist whom I knew well enough to ask for guidance, I applied to Oliver H. Lowry, who had been appointed chairman of the Department of Pharmacology at Washington University and was in his mid- to late-thirties. Very few fellowships were available at that time, but ultimately I received one under circumstances that were so accidental and unusual that I cannot relate them

here. Science at Washington University was famous worldwide. Joseph Erlanger (physiology) and Herbert Gasser (pharmacology) were early neurophysiologists who had uncovered the relationship between neuroﬁber size and velocity of nerve conduction. Importantly, Carl and Gerty Cori were widely known for their studies of the mechanism of glycogen synthesis and of glycogen storage diseases. Carl had recently vacated the chairmanship of pharmacology to become the head of biochemistry one ﬂoor below (in terms of contemporary issues, Gerty was a research associate; she became a professor after they shared a Nobel Prize). Research in physiology, pharmacology, biochemistry, and microbiology, where Jacques Bronfenbrenner had been involved in the discovery of bacteriophage and Al Hershey had initiated his famous studies of these bacterial viruses, was very active. Martin Kamen, who had used C11 (a very short-lived carbon isotope, the only radiocarbon known at the time) as a tracer in biology and had then created C14 as a very long-lived tracer, was carrying out research in the Radiology Department. And Ollie Lowry had undertaken a difﬁcult problem, the dissection of single cells from the nervous system and the measurement of their enzyme activities. What an atmosphere in which to be introduced to research! I stayed for two years, although I did not do extremely well. Somewhat discouraged, I opted to go back to ﬁnish my residency in medicine at the University of Chicago, where the young woman I later married was a student. But then the drama began. The Korean War had started in 1951, and those of us who had been trained by the Navy but had not served were called back into service in the spring. I learned that Washington University School of Medicine had been asked to help vitalize medical education in Bangkok, Thailand, by providing staff for the medical school at Siriraj Hospital and that I could join this program as an ofﬁcer in the U.S. Public Health Service as an alternative to the U.S. Navy. I enlisted. Ann and I were www.annualreviews.org • My Scientific Journey

5

ANRV270-IY24-01

ARI

13 February 2006

11:14

married in May and went to Bangkok a month later, we thought for two years. By the time we left for Thailand, I had published two abstracts and completed the work for one paper.

BANGKOK AND A SECURITY HEARING, 1951: HOW I BECAME A BIOCHEMIST

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Bangkok was extraordinarily exotic then, a relatively small city of 500,000 people. Many of the streets were canals (klongs). Each morning, I took a small boat to Siriraj Hospital across the Menam Chao Praya, a large, dirty river that drained much of the country. I carefully sat in the middle of the boat, not wishing to be splashed with the unsanitary water. My task was to help Professor Ouay Ketusinh, the chairman of pharmacology, modernize teaching. I met the members of his department and we talked and consulted every day. I also had many tasks to establish a life in Bangkok: renting a house from a prince (of which there were many because the kings had formerly been polygamous), purchasing a car, and ﬁnding furniture. It did not last long. After only two or three months, I was abruptly called to the ofﬁce of the chief of the Public Health Service mission to whom I was responsible and then to the American ambassador. A cable had been received that I was to leave Bangkok within 24 hours and return to Washington. Why? I asked. The cable message contained a coded reference; the ambassador opened a reference book that showed that I was being recalled under a code that deﬁned me as a security risk. I had no idea why. In protesting the haste, because I had a car to sell, furniture to deal with, and a rental contract from which to disengage myself, I was given an additional 48 hours. I wanted to stop in Europe on the return trip, but that was not permitted. However, the travel ofﬁce carefully arranged for my plane from Bangkok to arrive in Amsterdam after the last (possibly the only) departure for the United States. So in fact, I had a 20-hour stopover. 6

Strominger

Soon after we arrived in Washington, I called a friend, Arthur Murphy, who had married a young woman I had known at Wellesley when I was an undergraduate. She had recently given birth to their second child and had been taken to the hospital that day with a high fever. She died four days later of bulbar polio; post-partum women were unusually susceptible. It was only a few years before the introduction of polio vaccination. Why do I mention this personal episode? It gave me enormous perspective during the not-soserious events that followed. The next day, I reported to the chief of the Division of Commissioned Ofﬁcers in the Public Health Service, who asked me to resign my commission or the president would revoke it. Why? I asked again. I was told that I was a security risk, but they would not provide any details. “Please resign your commission.” Fortunately for me, my brother, who was a medical student in St. Louis then, managed to notify Oliver Lowry and Robert Moore, the dean of the Medical School, about what was happening to me. After all, I had originally gone to Thailand under the auspices of Washington University. They protested mightily, and I was told that I would receive a bill of charges and ultimately a security hearing, if I wanted one. My friend Arthur Murphy was a Justice Department lawyer just out of law school. He could not help me in a case that involved the government, but he told me of several law ﬁrms in Washington that were taking cases such as mine pro bono. First, I went to see Mr. Abraham Fortas, a member of the ﬁrm of Fortas and Porter at that time (and later a Supreme Court Justice). He listened sympathetically but told me that their ﬁrm already had too many pro bono cases. He and Arthur Murphy both suggested another Washington ﬁrm, Covington and Burling. There, another young lawyer, Paul Warnke, Arthur’s classmate and a former law clerk with Dean Acheson, agreed to be my lawyer (Dean Acheson had, by that time, taken leave from the ﬁrm to be secretary of state). Paul later became assistant secretary of defense for international

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

security affairs and still later was appointed as ambassador to negotiate the Strategic Arms Limitations Treaty (SALT) with the Soviet Union. I met many young lawyers in Washington during this period, and, aside from “the problem,” it was a very interesting experience. Paul immediately asked for the bill of charges, which the Loyalty Review Board reluctantly agreed to provide after several months. The bill listed three charges: (a) I had had friends in medical school who were allegedly members of the Communist Party; (b) I had a library of communist books; and (c) I had subscribed to the Daily Worker, the publication of the American Communist Party. None of these was true to my knowledge. What had happened? At the time I left, the FBI had not completed its investigation of me, routine for those assigned to an overseas mission. When they did, they decided I was a security risk. As mentioned above, I had been an active member of the Association of Interns and Medical Students (AIMS). AIMS had campaigned for the introduction of health insurance in the United States, which, at that time, did not exist in any form, private or public. As a result, the powerful American Medical Association branded the organization communist. All the individuals listed in my bill of charges had been AIMS members. The politics of some were to the right and others to the left of mine, but I had never asked any of my friends about their formal political afﬁliations, nor did I think it was appropriate for the Loyalty Review Board to ask. I had obtained my so-called “library” of communist books in a political science course at Yale while in medical school, one of three that I took in late afternoons to complete the requirements for my AB degree from Harvard. There were texts by Engel, Marx, Lenin, and others. My alleged Daily Worker subscription was actually the Progressive, the publication of the party by the same name led by Henry Wallace, of which my landlady, through whose mailbox my own mail came, told the FBI when they interviewed her about me.

The security hearing was short, and I was cleared. However, the Public Health Service told me that I could not return to Bangkok because that would require an additional clearance by the State Department, which had a much tougher Loyalty Review Board, and an additional six months to schedule a hearing. Instead, they suggested an assignment to the National Institutes of Health for the remainder of my appointment as a commissioned ofﬁcer. NIH? What was that? I had never heard of it. At that time, NIH was a small operation in Bethesda, Maryland, composed of only ﬁve or six buildings. But it was a stroke of good fortune. It was really the beginning of my scientiﬁc career. I owe it to Senator Joseph McCarthy!

NATIONAL INSTITUTES OF HEALTH, 1951–1954 To join the NIH, I had to ﬁnd a lab that would accept me. I soon discovered that there were many interesting young scientists at NIH, and I spoke with many of them. I was interested in the National Heart Institute, but the director told his laboratory chiefs that he would not agree to my placement; he did not want his institute associated with someone who would have a security hearing. In the end, only two individuals offered me placements within their laboratories: Sanford Rosenthal, chief of the Laboratory of Pharmacology in the National Institute of Arthritis and Metabolic Diseases, and Seymour Kety, the newly appointed director of the newly established National Institute of Mental Health. The latter, however, was so new that it did not yet have laboratory space, so I gratefully accepted Sanford Rosenthal’s offer. Among the many honors I have received, I treasure the invitation to give the lecture at NIH in honor of his eightieth birthday. The Laboratory of Pharmacology was a fortunate choice. On my ﬁrst day at work, I walked into Sanford’s ofﬁce in Building 4 and asked what he would like me to work on. He said that my one-unit laboratory was www.annualreviews.org • My Scientific Journey

7

ARI

13 February 2006

11:14

adjacent to his ofﬁce and that I could work on anything I wanted. I was stunned. I was 26 years old and had little research experience, but I had a fully funded independent laboratory. What should I work on? Before leaving Oliver Lowry’s neurochemistry laboratory in St. Louis, I had been thinking about focusing my effort on how drugs work in the nervous system, e.g., why does morphine obtund pain without loss of consciousness, whereas barbiturates put you to sleep? These drugs must have differential effects within the complex cellular components of the nervous system. Could the effects be studied with the single-cell dissection techniques pioneered by Lowry? I decided it was much too difﬁcult a problem for 1951 and that, until we could understand how a drug like penicillin worked on a homogenous cell population like bacteria, unraveling the complexities of the nervous system would be too difﬁcult. Thus, penicillin’s mechanism of action was on my mind when Alan Mehler, a member of the Rosenthal laboratory and earlier Severo Ochoa’s only graduate student, pointed out an interesting abstract that was published in the most recent Federation Proceedings. It was based on the PhD thesis of James T. (Ted) Park at the University of Wisconsin, who had isolated from penicillin-inhibited staphylococci a very unusual nucleotide. It contained uridine diphosphate, linked to a previously undescribed sugar whose structure had not yet been elucidated, to which was linked a peptide. Could this nucleotide be involved in the then mysterious relationship between nucleic acid biosynthesis and protein biosynthesis? Alan also told me that Ted Park was in the armed services and had been assigned to the Army Bacteriological Warfare Laboratory at Fort Detrick, Maryland, very close to Bethesda. Park was not able to continue to work on the nucleotide he had discovered. I phoned him and we met for a picnic lunch at a brook outside the gates of Fort Detrick because I was not allowed in. He was very generous in providing me with advice and helping me get started. I had found a problem that really interested me

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

8

Strominger

and a way to get into it. Was this nucleotide that accumulated in Staphylococcus aureus related to the mechanism of action of penicillin? Did it have anything to do with protein and nucleic acid synthesis? Alan Mehler was a well of knowledge about biochemistry as was Herb Tabor, another member of the Rosenthal lab and now the editor of the Journal of Biological Chemistry. Through them I was introduced to the biochemical ferment then being generated at NIH by young scientists in Building 3 (several of whom were close friends with Herb Tabor and Alan Mehler), such as Arthur Kornberg (now known for his studies of DNA synthesis), Bernard Horecker (mechanism of carbon dioxide ﬁxation for sugar synthesis in plants), Jesse Rabinowitz (1-carbon transfer involving folic acid derivatives), Bruce Ames (environmental carcinogens), Leon Heppel and Herman Kalckar (nucleotide metabolism), Osamu Hayaishi (oxygenases in mammals), Earl and Terry Stadtman (microbial biochemistry), Christian Anﬁnsen (protein structure), Bernard Brodie and Julius Axlerod (drug metabolism), Herb Tabor (spermine and related polyamines), Dan Steinberg (phytanic acid storage disease), and many, many others. Three of these individuals later received Nobel prizes, and all have ultimately been elected to the National Academy of Science. My work evolved in two directions: the work I did independently on the mechanism of penicillin action, and the work I did in collaborations (principally with Herman Kalckar and Leon Heppel, which allowed me to gain expertise in enzymology). First, I learned how to use the Park method to isolate the unusual uridine nucleotides from penicillin-inhibited S. aureus (2). It involved fractional precipitation of mercury salts by ethanol and ultimate measurement of the amount of nucleotide by the UV absorption of uridine. It was clear that I would not make very much progress using this time-consuming and inexact procedure. Then I came across an obscure paper that described a method for measuring glucosamine. Could it be adapted

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

to measure the Park nucleotide after mild acid hydrolysis to free the acetylamino sugar? The molar extinction coefﬁcient of the chromagen formed in the published procedure was exceedingly low, but Oliver Lowry was a methodologist, and I had learned that playing with conditions of colorimetric reactions could be rewarding. Late one afternoon, I decided to try different buffers at different pHs. The last one I used was sodium tetraborate, an internal salt that was used in the Lowry laboratory as a pH standard, pH 9.15, and that I had on my reagent shelf for that purpose. I added it, went home for supper, allowing the various trials to incubate at 37◦ . When I came back after supper, my eyes bulged. A tube to which I had added borate was an intense dark purple color. This small thing was the break I needed, and it provided me with a quantitative method of measuring the accumulation of the Park nucleotides. I only published the details of the method a few years later when I encountered Luis Leloir in Copenhagen and learned that a student in his laboratory had developed a similar method. We published it together, four years later in 1955 (3). Two years after that, I published the use of the method to measure the kinetics of accumulation of the nucleotide and many of its features, using 50– 100 ml cultures (4). Almost nothing had to be rushed into publication in the late 1950s. During the same period, my work with Leon Heppel taught me how to purify enzymes and to keep careful track of units of activity, yields, and speciﬁc activity during the puriﬁcation procedure (5), knowledge that proved extremely important later. With Herman Kalckar, who had recently come to NIH from his native Denmark, we isolated another uridine diphospho-sugar compound (UDP-glucose) that had been described by Leloir. Again, we began with precipitation of mercury salts from yeast, but now used the recently introduced anion exchange chromatography to purify the nucleotide. We built our own fraction collector (they could not be purchased then). We knew that UDPglucose was important, but we did not have

a good idea what to do with it. Then one afternoon, in the men’s room, I encountered Julius Axlerod and, while standing side by side, asked him what he was doing. He was studying the synthesis of morphine glucuronide in liver slices. Soon after, we designed an experiment in which UDP-glucose was added to minced liver slices together with the oxidizing agent, NAD (nicotinamide adenine dinucleotide) [then called DPN (diphosphopyridine nucleotide)]. UDP-glucuronic acid was formed and used for the synthesis of the glucoronide. In one afternoon, we had carried out the ﬁrst metabolic transformation of a sugar while attached to a nucleotide (6). This technique has had many extensions for other nucleotides of this type and for other sugars. For example, Herman Kalckar went on to use it to study the utilization of galactose in mammals and to discover that mutation in a uridyl transferase in which galactose 1-phosphate is exchanged for glucose 1-phosphate in UDPglucose is the defect in the metabolic disease congenital galactosemia (7).

SABBATICAL LEAVE IN COPENHAGEN, DENMARK, AND CAMBRIDGE, UK, 1955 After two years at NIH, I knew I wanted to leave. Despite the scientiﬁc excitement, the security hearing had left a bad taste in my mouth for government service. Oliver Lowry invited me to return to his department in St. Louis as an assistant professor. I said that I would like to take the job but that I wanted to see how science is conducted in Europe ﬁrst. He gave me the ﬁrst year of my new appointment as a sabbatical leave, and I obtained support from the Commonwealth Fund for the year in Europe. I wanted to go to Cambridge University, but Ernest Gale, a microbiologist in the Biochemistry Department who was studying penicillin, had no room in his lab that year, and Roy Markham could only take me for the second half. Lowry had worked in Copenhagen with the prominent Danish protein chemist, Kaj Linderstrom-Lang at the www.annualreviews.org • My Scientific Journey

9

ARI

13 February 2006

11:14

Carlsberg Laboratory, and suggested I spend the ﬁrst six months there. It turned out to be an amazingly fortuitous experience. The Carlsberg Laboratory on the grounds of and funded by the Carlsberg Brewery was at that time one of the principal research institutions in biochemistry in Europe. Linderstrom-Lang was a joyful, gregarious laboratory head whose laboratory had previously seen development of the ﬁrst method for nitrogen determination by Kjeldahl and then the introduction of the concept of pH by Sorenson. A very important feature of the laboratory has been often noted by others, but I will say it again. It had a huge two-way refrigerator. One side opened to the laboratory and the other side to the brewery from which it was kept stocked. I needed the cap of a bottle of Carlsberg Beer to illustrate the match shooting game that took place every day at lunch time. The other postdoctoral fellows and sabbatical visitors there at the same time included Chris Anﬁnsen (NIH and Johns Hopkins), Bill Harrington (Johns Hopkins), Fred Richards (Yale), Ieuan Harris (Cambridge, UK), John and Charlotte Schellman (Oregon), and Don Wetlaufer (Minnesota). Chris and Bill lived only a few blocks from us in Hellerup, a Copenhagen suburb. Bill was the only one with a car, so the three of us commuted across town to the Carlsberg Laboratory, 30 minutes each way, more time to talk about science. My bench space, about four feet, had a cupboard full of copper pots that I wanted to clear out. Nothing doing; they were Kjeldahl’s pots. Everyone was working on some aspect of protein or peptide chemistry, so I thought I might as well do the same. I worked out the sequence of the pentapeptide in UDP-acetylmuramylpentapeptide, the Park nucleotide: l-Ala-γ-dGlu-l-Lys-d-Ala-d-Ala, using simple partial acid hydrolysis and paper chromatography, coupled with the stereospeciﬁcity of several enzymes to identify the stereoisomer (8). There was a steady ﬂow of important scientiﬁc visitors; three of them were critical to me. Milton Salton from Cambridge said he

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

10

Strominger

had described a new morphological structure called the bacterial cell wall and that it contained a very unusual sugar. Later I showed that this sugar and the one in the uridine nucleotide were the same. Luis Leloir from Buenos Aires came, and during discussion we discovered that we each had invented a new method for acetylglucosamine determination, which I have already described above. Esmond Snell from Berkeley arrived and talked about a peculiar dipeptide that was present in large amount in lactobacilli and had been identiﬁed as d-alanyl-d-alanine. That was it, I was on the track. But the six-month stay in Copenhagen came to an end. By then, Fred Richards had moved to Cambridge and, because Ann was by this time eight months pregnant, he had found a ﬂat for us on Brunswick Walk facing Midsummer Common. Any readers who have been to Cambridge will know that beautiful spot. The ﬂat, however, was a little squalid. As a result, Ann was allowed to have a hospital delivery at Mill Road Maternity Hospital, which is fortunate because it was a difﬁcult delivery and our son Andy had to be extracted quickly. A month later, Alex and Jane Rich moved in for a week’s stay with our two-year old son, Paul, allowing Ann, me, and Andy to have a driving holiday in Scotland all the way to Skye. I was at the Molteno Institute for Parasitology, where Roy Markham had his laboratory. I learned a great deal about separation of nucleotides by paper electrophoresis and worked with Leslie Mapson across the road on UDP-glucose dehydrogenase from plants. Roy’s laboratory was on the top ﬂoor. A wonderful older scientist, David Keilin, was on the ﬂoor below. He had discovered cytochromes while looking at translucent worms with a hand spectroscope and later found them in many other organisms, including translucent Antarctic ﬁsh that have no hemoglobin and swim slowly under the ice. His museum was on the top ﬂoor with Roy’s lab. It doubled as a small lecture theater. Then one day, Max Perutz began his lecture course on protein

ANRV270-IY24-01

ARI

13 February 2006

11:14

crystallography in that lecture space, the ﬁrst such course anywhere in the world, for perhaps a dozen students. I had no choice but to listen, although I understood very little, because Max had asked Roy to request that that “young American” not run the centrifuge while he was lecturing.

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

WASHINGTON UNIVERSITY SCHOOL OF MEDICINE, ST. LOUIS, 1956–1963 Time passed quickly, and I was soon back in St. Louis as an assistant professor of pharmacology. I shared a lab initially with Bob Furchgott who had just discovered endothelial relaxing factor, later identiﬁed as nitric oxide, using a Rube-Goldberg-type apparatus common to pharmacologists at that time. Bob left six months later to become head of pharmacology at Downstate Medical School in New York, and much later he was widely recognized for the work he had done in St. Louis and New York. I had my own small lab again. On the way back from Europe, I had met Ted Park again. He was working on the same phenomenon that I was, namely, that the uridine nucleotide that accumulated in penicillin-treated staphylococci was a precursor of the bacterial cell wall. We published two papers together rather than compete with separate publications (9, 10). Isotope studies strongly supported the conjecture (11). I spent the next eight years showing that the nucleotide was in fact a biosynthetic precursor of the cell wall, isolating cell walls, determining their structure with Jean-Marie Ghuysen using the speciﬁcities of bacteriolytic enzymes to dissect the structure (12), and working out the mode of biosynthesis of UDPacetylmuramyl-pentapeptide (13). Finally, I got my foot in the door and we developed a cell-free system to study the utilization of two nucleotides, UDP-acetylglucosamine and UDP-acetylmuramyl-perntapeptide, for cell wall peptidoglycan biosynthesis (14), and we also extended the study of transformations of sugars while attached to nucleotides (15). I had

my ﬁrst excursion into immunology [although in fact the components of the l-lysine and meso-diaminopimelic acid–containing peptidoglycans, whose structures were elucidated, have in the past year turned out to be key molecules in understanding a component of innate immunity, the intracellular nucleotidebinding oligomerization domain (NOD) receptor for these components of the peptidoglycans] (16). Teichoic acid, a speciesspeciﬁc “ornament” attached to the cell wall of S. aureus, has an acetylglucosamine residue as a substitution on polyribitol phosphate. Rabbit antisera prepared against S. aureus cells agglutinated cell walls of this organism and were speciﬁc for α-acetylglucosaminyl ribitol, which accounted for only 15% of the total acetlyglucosamine, whereas the remaining 85% of β-acetylglucosaminyl ribitol was immunologically inert (17). Another excursion into immunobiology was a minisabbatical at the Wanderforschungs Institute (now the Max Planck Institute for Immunology) in Freiburg, Germany. The work there was focused on the chemical structure of lipopolysaccharides (LPS) of Gram-negative bacteria (18). A particular focus was the relation between the serological speciﬁcities of the many strains of Salmonella and Escherichia coli and the structures and linkages of the sugar components of LPS, including several unusual sugars, 3,6-dideoxyhexoses such as colitose and tyvelose, in a collaboration with Fritz Kaufmann, a prominent bacterial serologist at the Statens Serologische Institut in Copenhagen. More about how this work got me into HLA below.

UNIVERSITY OF WISCONSIN, 1964–1968 I had found myself increasingly frustrated by the opportunities available at Washington University and about the lack of scientiﬁc stimulation. The school was in temporary doldrums. So, when an offer came to be chairman of the Pharmacology Department at the University of Wisconsin, I took it. The university www.annualreviews.org • My Scientific Journey

11

ARI

13 February 2006

11:14

was extremely generous to a department chairman. In addition to secretarial help, they provided positions for a senior research associate, a technical assistant, and a dishwasher and substantial funds for supplies. Those four years at the University of Wisconsin were extremely productive. The foundation for my work had already been laid at NIH and at Washington University. The structure of cell walls (of which I am especially proud) had been outlined, particularly the presence in them of peptide cross bridges between acetylmuramyl-peptide subunits. The structural analysis led Donald Tipper to the structural analog hypothesis of penicillin action (see below). A particulate enzyme system that catalyzed cell wall biosynthesis from two uridine nucleotide precursors had been established (14), or, at least, we thought it had been established. Initially, in the radioactive incorporation assay, we found 50 cpm of labeled precursor incorporated above a background of 50 cpm. As we continued to work with the system, it soon went up to 500–1000 cpm above the background. I do not know why. We did not change anything. Years later, the same story was repeated in the yields of HLA proteins. You have to have intuition and faith sometimes. A long series of papers have described the steps in the biosynthesis of the peptidoglycan of bacterial cell walls, including the species speciﬁcity in the nature of the peptide cross bridges and the utilization of diaminopimelic acid–containing nucleotides (rather than those containing lysine) in E. coli (19). With each new step in the biosynthesis that we discovered, we tested its sensitivity to penicillin. None were sensitive until the very last step. Finally, we could show that the formation of cross bridges in the particulate system was inhibited by penicillins, and we postulated that the LD-cyclic dipeptide, penicillin, was an analog of the C-terminal end of acetylmuramyl-pentapeptide, i.e., dalanyl-d-alanine, that released d-alanine in the trans-peptidation reaction to form the cross bridge (20). Only recently has it been shown that the hypothesis of the structural

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

12

Strominger

analogy was correct, i.e., that substrate and penicillin lay in the same space in the enzyme (21). Michio Matsuhashi and Kazuo Izaki identiﬁed the double-headed enzymes that catalyzed both trans-glycosylation and trans-peptidation (19). “Enzymes” is plural because multiple penicillin-binding proteins were described that have different roles in cell wall synthesis and that catalyzed this reaction. Along the way, lipid intermediates in peptidoglycan synthesis were also described by John Anderson and later identiﬁed by Yasuo Higashi as C55 -isoprenyl phosphate, the lipid that is involved in the transport of these molecules across the cell membrane (22, 23). There was another research vein also, the description of unusual bacterial sugars and of the mechanism of their biosynthesis. These included 3,6-dideoxyhexoses and two new ones, 4,6-dideoxy-4-acetamidohexoses, that are involved in speciﬁcity determination among different bacterial strains (24). These studies had important consequences later. What has this to do with immunology? Surprisingly, a lot. I had often wondered why no immunological system involving the peptidoglycan or its various components had been described—no antibodies and no T cell recognition phenomena. They are unique bacterial components, and they seemed like logical targets for the immune system of mammals. It took 30–40 years longer for their importance in immunology to be outlined. JeanFrancois Petit, who had worked on peptidoglycan structure and synthesis in my lab (25), returned to Paris and was involved in the discovery that muramyl dipeptide [really Nacetylmuramyl dipeptide (26)] and its derivatives are powerful immunoadjuvants (27), but he had no explanation of the mechanism. Only in the past 10 years, since the innate immune system has become a center-stage player in immunology, has it become clear that these substances are ligands for receptors in the innate immune system helping to initiate immune system responses (28). And, in the past year, a second innate immune system to deal with intracellular bacteria has been uncovered

ANRV270-IY24-01

ARI

13 February 2006

11:14

(16). Acetylmuramyl-l-Ala-γ-d-Glu-mesoDap, or more simply γ-d-Glu-meso-Dap, called iE-DAP, was shown to be a ligand that initiates immune responses as a ligand for the NOD1 receptor (approximately 50 years after its discovery as a component of the cell wall peptidoglycan of some Gram-negative bacteria!). NOD2 recognizes N-acetylmuramyl-lAla-γ-d-Glu, “muramyl dipeptide.”

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

HARVARD UNIVERSITY 1968–PRESENT In the midst of this exciting work, a phone call came from Konrad Bloch and John Edsall, who invited me to join a committee on biochemistry and molecular biology at Harvard as professor of biology, with the expectation that the committee would become a department within a few years. It was an irresistible opportunity to join a small group of outstanding scientists to form a new department in a major university. The other members were Paul Doty, Matt Meselson, and Jim Watson. I was invited to ﬁll a niche in research and instruction in microbial biochemistry, a central theme in biology at that time. Neither they, nor I, knew that I would soon turn my attention to molecular immunology. I had done research in pharmacology for the past 17 years and had been associated with a medical institution for more than 20 years. I welcomed the opportunity to join the faculty of arts and sciences with the stimulation and challenges that a new environment was certain to bring. I had thought for about a year that the penicillin problem was almost solved and that I should look for new pastures. Of course, 40 years later, it still has not been completely solved, and much interesting current work that builds on our old work has been stimulated by the emergence of drug-resistant bacteria. I had only recently passed my fortieth birthday, and I thought that if I were to have an exciting scientiﬁc future, I should ﬁnd another problem. Transplantation biology was in the air, brought there by the advent of kidney grafts and then, in the early 1960s, by heart grafts.

Little was known about the mechanism of allograft rejection, and there was no knowledge of the transplantation antigens that had been named by Peter Gorer (29) during his studies of graft acceptance or rejection in outbred mice. Gorer’s work stimulated the development of inbred strains of mice by George Snell (30). When I moved to Harvard, I decided that it was now or never to change ﬁelds. At ﬁrst, only two brave postdoctoral fellows (Wolf Droege and Dieter Malchow) were interested in working in this new area. The rest of the lab continued to work on bacterial cell walls, the mechanism of their biosynthesis, and how penicillin worked. Over a period of 15 years, however, the number of people working in the former slowly increased and those working in the latter decreased. By 1985, no one who came to the lab was interested in microbial biochemistry. I knew there was interesting work still to do in that area, but I had no choice but to stop working on the penicillin and bacterial cell wall problem. I had been continuously supported by an NIH grant in this area for 29 years, from 1956 when I moved from NIH to Washington University and obtained my ﬁrst research grant, until 1985 when I did not renew it. One notable event inﬂuenced my decision to move toward the study of histocompatibility proteins. In the mid-1960s, Salvador Luria from MIT obtained a grant from NSF to organize a yearly conference that he called “Microdermatology”—the study of the surfaces of microorganisms. The meetings were always held in some interesting place, and one year it was held at the Abbay´e-Royaumont near Paris. Salva always invited an additional speaker working in some distantly related area. That year, the speaker was Allan Davies from the Microbial Research Establishment at Porton, UK (Britain’s equivalent of Fort Detrick). Allan had discovered a number of the 3,6-dideoxyhexoses that give speciﬁcity to bacterial surfaces. He described this work and ended with the speculation that the speciﬁcity of transplantation antigen might also be www.annualreviews.org • My Scientific Journey

13

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

Human leukocyte antigen (HLA): solving the structure of HLA proteins was especially important in the 1970s because it had become clear from the work of Jean Dausset, Jon van Rood, and many others that these substances were critical in transplantation rejection.

14

11:14

determined by unusual sugars. I traveled back with him to London, and he offered to be my guide on a sightseeing trip. First we went to Westminster Abbey, just across the street from the Houses of Parliament, and sat down on a bench in front of the statue of Abraham Lincoln at one side of Parliament Square. We never did any more sightseeing. He spent the next few hours introducing me to transplantation biology. I was enthralled, and it moved this ﬁeld to the top of my ideas for future projects. Whether or not sugars were involved, it was certainly a fascinating area for research. After an abortive start trying to study the homing of lymphocytes to spleen and lymph nodes and its inhibition by pertussis toxin (as had been described by Jim Gowans), two important contributions from former postdoctoral fellows who had worked on the bacterial cell wall/penicillin problem ﬁnally got me started working on HLA antigens. When Stan Nathenson left my lab, he went to work with Allan Davies on murine transplantation antigens and discovered that they could be solubilized from the surfaces of cells by papain (31). At that time, one could not purify a membrane-bound substance, so this contribution was critical. But how could one purify a substance without any idea about its chemical nature or any way to measure it? Arnold Sanderson, who had worked on teichoic acid (32), came to Boston Children’s Hospital on a sabbatical leave from the Blond Institute at Queen Elizabeth Hospital in East Grinstead, UK. There, a Center for Transplantation Biology had been established under the direction of the Danish immunologist, Morten Simonsen. Arnold had been attempting to study human histocompatibility proteins and had even solubilized them with papain and separated the HLA-A and HLA-B speciﬁcities by ion-exchange chromatography (33). I asked if he would show me the assays. First, he showed me the trypan blue exclusion assay that was then used in tissue typing. The dye is excluded from lymphocytes, but, in the presence of pregnancy alloantiserum of Strominger

the right speciﬁcity and complement, holes are punched in the membrane and the dye enters. Thus, the percentage of blue cells and the inhibition of their appearance by soluble HLA antigen could be scored microscopically. Then he showed me the chromium51 release assay that Hans Wigzell had recently developed and that he had used in a more quantitative way. I had learned much about methods from Ollie Lowry, and with a little ﬁddling I could easily make the assay for HLA antigens strictly quantitative. I had had a great experience learning how to purify enzymes with Leon Heppel (a master of the trade), and so I set about purifying HLA proteins as though they were an enzyme, keeping very careful track of speciﬁc activity, fold puriﬁcation, and yield at each step. I quickly realized that to obtain any substantial amount of the protein, very large amounts of starting material would be needed. Fortunately, at this juncture, several generous colleagues appeared who realized how interesting and potentially important the problem was and offered to grow B cell lymphoblastoid cell lines (LCLs) in 100-l batches, ∼100 gm of packed cell (particularly Dean Mann at NIH and Jim Woody and Mike Strong at the U.S. Naval Medical Research Laboratory across the street, together with Navy Corpsman Tommy Williams, who was fantastically devoted to the project). LCLs were only a few years old at the time. Epstein Barr virus (EBV) had only very recently been discovered and been shown to be able to transform and establish LCL in vitro. Thus, one could establish LCL from the B cells of any individual. But these cell lines all had the difﬁculty that, because the products of all three HLA loci that were then known, HLA-A, HLA-B, and HLA-C, were expressed codominantly, a mixture of six speciﬁcities resulted. Then I thought that using homozygous cell lines obtained from the offspring of consanguineous marriages would simplify the problem. There were very good church records of ﬁrst-cousin marriages in the Netherlands, but still another source might be inbred religious

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

populations in the United States. Dean Mann had come from such a community, the Indiana Amish, so he established a line from an individual in that community, JY. The JY cell line was HLA-A2/HLA-B7 homozygous, or so we thought. Years later, we learned that JY expresses two closely related subtypes of HLA-B7, and when class II MHC loci were discovered later, JY was heterozygous at those loci (HLA-DR4, HLA-DR6)! However, other homozygous cell lines were extremely useful in doubling the yield and simplifying the problem of separating speciﬁcities, which was essential to deﬁne the polymorphism. A further problem was funding the research. The NIH Study Section had given my application a low priority score for several reasons: (a) The amount of material was too small to be able to obtain enough to work with; (b) we were using the wrong solubilization method—we should have been using LiCl solubilization that had been reported by another group; and (c) I had no prior experience in immunology (despite my success in microbial biochemistry). Fortunately, I had sufﬁcient funds in the bacterial cell wall/penicillin problem that I could divert a small amount to support one or two people working on HLA antigens. I rewrote the grant application and paid special attention to emphasizing the LiCl extraction method. However, we continued to use the papain solubilization procedure, which we knew was far superior in yield and reproducibility. The application was eventually funded, and, although it supported important continuing work, by that time Mervyn Turner and Peter Creswell had nearly completed the ﬁrst puriﬁcation to homogeneity of human histocompatibility proteins (HLA-A2 and HLA-B7) (see Table 1) (34). Pete Creswell had already shown that papain-solubilized radioactive material seemed to contain two peptide fragments (35). In a collaboration with Howard Grey, they showed that the small subunit of the HLA antigen was β2microglobulin, a small urinary protein with an immunoglobulin-like sequence that had

been isolated form nephrotic human urine and was thought to be a free immunoglobulin domain (36, 37). Tim Springer then demonstrated that detergent-solubilized HLA antigen contains an additional transmembrane region and an intracytoplasmic region and was also composed of two subunits (38), thus ruling out the hypothesis that the two subunits were the products of papain proteolysis. Harry Orr and Jose Lopez de Castro obtained amino acid sequence information using an inherited sequenator left behind by Klaus Weber when he moved with Jim Watson to Cold Spring Harbor (39, 40). Sequencing was a struggle learned by one postdoc after another, until ﬁnally in 1982 I established the protein-sequencing facility, now a world-class and widely used facility called the Harvard Microchemistry Facility run by Bill Lane. The sequences of the heavy (α) chain allowed us to determine that it was composed of three domains and that the third of these (α3) was an immunoglobulin-like domain (39, 40). That marked the beginning of the immunoglobulin superfamily of proteins. Later, DNA sequence information showed that the three domains were encoded in separate exons, conﬁrming the deduction from the protein sequences and also later conﬁrmed by the X-ray determined three-dimensional structure, although α1 and α2 folded together to form a superdomain. Many of the sites of polymorphism could also be deduced from the ﬁrst several sequences.

CLASS II MHC PROTEINS Soon after puriﬁcation of HLA antigens (in reality what we now refer to as class I MHC proteins, or MHCI), a new postdoctoral fellow, Bob Humphreys, observed that during the last step of a puriﬁcation of the papain-solubilized protein by ion-exchange chromatography, a second protein peak of approximately equal size was eluted just before the HLA antigen (41). He ran an SDS gel and found that this material also contained two polypeptide chains of 23 kDa and 30 kDa. www.annualreviews.org • My Scientific Journey

15

ANRV270-IY24-01

Table 1

ARI

13 February 2006

Purification of HLA antigensa

Step HL-A2b 1. Crude membranes 2. Membranes after zonal centrifugation (Peak II) 3. a. Membranes after papain b. Papain digest 4. CM52 chromatography 5. Sephadex G-150 chromatography 6. DE52 chromatography

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

11:14

HL-A7b 1. Crude membranes 2. Membranes after zonal centrifugation (Peak II) 3. a. Membranes after papain b. Papain digest 4. CM52 chromatography 5. Sephadex G-150 chromatography 6. DE52 chromatography

Titer

Volume (ml)

14,600 30,900

53 23

290 25,000 50,000 15,500

Total units

Total protein (mg)

Specific activity

Purification

Yield %

776,000 710,000

986 225

787 3,150

(1) 4

(100) 91

14 9 1.4 10

4,060 225,000 70,000 155,000

145 45 ND 6

28 5,000

6.3

25,800

33

0.5 29 9 20

35,400

3.5

124,000

0.49

253,000

320

16

1,500 3,100

53 23

79,500 72,300

986 225

80.6 321

1 4

100 91

115 3,300 13,000 2,000

14 9 1.4 10

1,610 29,700 18,200 20,000

145 45 ND 6

11.1 660

8.2

3,330

41

2 37 25 25

2,500

4.4

11,000

0.56

20,000

248

14

a

In this preparation, 53 ml of packed frozen cells (RPMI 4265) were employed. First published in Reference 34. MHC class I HL-A2 is now called HLA-A2. HL-A7 is now called HLA-B7. They are products of different loci, not fully known at the time of this puriﬁcation. ND, not determined.

b

We had to study it. Rabbits were immunized, and the immune serum was found to react with a protein that was found on human peripheral B cells but not on other cells among peripheral blood lymphocytes. Just at that point, I was invited to a meeting in Aarhus, Denmark, the 1975 Histocompatibility Testing Workshop, at which the speciﬁcities of pregnancy alloantisera were deﬁned. In addition to the workshop, a symposium on recent developments was held to which I had been invited. I was very surprised to learn that Jon van Rood and his colleagues in Leiden had found some pregnancy alloantisera that also detected a B lymphocyte–speciﬁc protein. The HLA-D speciﬁcity had been previously deﬁned by the people that participated in this workshop as proteins involved in mixed-lymphocyte culture (MLC) proliferation that could not be classiﬁed as HLA-A, B, or C. Van Rood and colleagues referred to 16

Strominger

the new antisera as HLA-DR (d-Related) because these antisera also seemed to block the MLC. These speciﬁcities were found to be genetically linked to the HLA antigens (class I) in a region called the major histocompatibility complex (MHC). They were the human analog of the immune response genes that had been characterized in rodents. Structural studies taken up by Jim Kaufman and Debbie Shackelford ended with an illustration of the striking structural similarities of class I and class II MHC proteins (Figure 1) (42), differing at that stage in our knowledge mainly in the linkage of the four extracellular domains and in the demonstration using alloantisera provided by the Leiden group that a second class II MHC heterodimer, distinct from HLA-DR, could be immunoprecipitated (43). This material was called DC1. DC (now called HLA-DQ) was Dora Centis, the research assistant in Roberto Tosi’s

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

laboratory in Rome who had ﬁrst noticed this type of unusual pregnancy alloantiserum. The third class II protein isotype, originally called SB but renamed HLA-DP, was discovered by Stephen Shaw at NIH. It had originally been thought that humans expressed only one class II MHC protein, whereas mice expressed two! Thus, by 1979 we had gotten as far as we could with protein chemistry of class I and class II MHC proteins, but we still had not answered the fundamental questions that had driven the initiation of this work. Where exactly was the polymorphism located within the molecule? What was the evolutionary force that had driven it? It could not have been the exchange of surgical grafts that originally deﬁned transplantation proteins. And what was the function of the polymorphism? The next steps in answering some of these questions were cDNA and genomic cloning, which had recently been introduced into biochemistry and molecular biology, and three-dimensional structure determination by X-ray crystallography. Both were in their infancy, but advances in techniques soon brought them to the center stage of modern biology.

cDNA AND GENOMIC CLONING OF HLA ANTIGENS Hidde Ploegh ﬁrst came to the lab to work on the penicillin problem as an undergraduate from Groningen, the Netherlands. He later returned as a PhD student from Leiden and began to study the biosynthesis of the HLA proteins. Members of the lab knew very little about the emerging ﬁeld of molecular biology, but Wally Gilbert was just down the hall working out the Maxam-Gilbert DNA sequencing method. When Hidde said he would like to try to clone a class I MHC cDNA, I urged him to give it a try but not to let go of his biosynthetic work until he was sure the cloning would work. The cloning did work, and he reported the ﬁrst cDNA clone for an MHC protein (44). It was not a very big clone, but Harry Orr used it to show that there were at least six class I–like genes in the human genome

(45). He later identiﬁed the additional three as the molecules now called HLA-E, HLAF, and HLA-G (46). At this point, everybody jumped in, and additional clones were rapidly reported. In our lab, the HLA-DR α cDNA was obtained by a very unusual method. A monoclonal antibody that recognized the heavy chain of HLA-DR was used to immunopurify polysomes containing only this mRNA in a single step (47)! cDNA and genomic cloning and sequencing led to the sequence of the HLA-DR α chain (48), and continued analysis revealed the complexity of the MHC class II α chain loci (49). Many labs cloned HLA-DR β chain genes; our contribution was a cDNA clone for HLA-DP (then called SB) (50). Then genomic cloning and cosmid cloning came to the fore. Michael Steinmetz in Lee Hood’s lab had been unusually successful in constructing a complete map of the MHC class II region in the mouse using cosmids (51), and we wanted to do the same for the human MHC class II region. It was not that easy. Eventually, however, Kiyo Okada made a good map from cosmid clones of subregions of the class II region of the human www.annualreviews.org • My Scientific Journey

Figure 1 An illustration of the structures of class I and class II MHC proteins as visualized in 1984 (42, 105).

17

ARI

13 February 2006

11:14

MHC (52, 53), and Thomas Spies made an overlapping map of the entire MHC class I, class II, and class III regions using the discovery that TNF genes were encoded near the center to link the two halves (54, 55). Many new genes were discovered in the class III subregion (56), although the functions of most are still to be identiﬁed. Notably, however, soon after he established an independent lab. Thomas identiﬁed two new genes near the HLA-B locus called MICA and MICB, which have recently assumed importance as ligands on target cells for natural killer (NK) cells (57).

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

X-RAY CRYSTALLOGRAPHY AND THE THREE-DIMENSIONAL STRUCTURE OF CLASS I AND CLASS II HLA PROTEINS As I have said, we had gone about as far as we could in analyzing the structure of HLA proteins by 1979. Protein sequence analysis had shown that the polymorphic residues were spread throughout the linear sequence of the ﬁrst two domains of the structure (40, 58). Attempting to determine the three-dimensional structure was the next logical step. During the period that I was serving my three-year term as chairman of the department (1970–1973), a young crystallographer, Don Wiley, a student in the Biophysics Program at Harvard who did his PhD thesis on aspartic transcarbamylase with Bill Lipscomb as his advisor, was appointed as an assistant professor. He was sufﬁciently promising so that we appointed him directly after he received his PhD degree without having done a postdoctoral fellowship. Don was a generation younger than me, and so he became friends with several of my students and postdoctoral fellows. One of them tried to crystallize a class I HLA protein, with advice from Don of course, but without much success. The class of graduate students that had entered in 1978 included Pamela Bjorkman. Like all of our graduate students, she did at least three rotations in her ﬁrst year 18

Strominger

to try to decide which lab to join. She was interested in protein chemistry or crystallography. One of her rotations was in my lab. One of my graduate students, Jim Kaufman, who was a year or two ahead of Pam and was working on MHC class II proteins, persuaded her that the crystallography of HLA proteins was a dynamite project. Don, who had only recently been promoted to professor, and I had talked about this project casually over the years. He was interested in surface glycoproteins and had been working on inﬂuenza hemagglutinin so it was a natural extension of his interest. Despite the small amounts of HLA protein that could be obtained from cultured LCL and the huge effort required, it seemed like an attractive project to Pam, so at the end of the year she elected to begin her thesis work in my lab with the view of doing the crystal structure in Don’s lab if she succeeded in crystallizing one of these proteins. One other interesting event is worth noting in terms of the forces that guide science in our country now. My NIH research grant to work on MHC proteins was up for its ﬁve-year renewal. The main focus of the renewal application was the three-dimensional structure of HLA proteins. It was turned down by Study Section (now called Scientiﬁc Review Group) with a score of around 275, the lowest score I had received in 25 years of applications for research work on the penicillin problem, the HLA problem, and other projects I describe below. This grant at that time was the main support of my lab, and if it was not funded I would have to close up shop. At this point I learned that Study Sections do not make funding decisions. They only make recommendations to Institute Council, who together with the institute director make the funding decisions. I appealed to Council and asked them to overrule the Study Section. They did not, but instead they funded the grant for one additional year and asked me to revise the application and resubmit. What was the basis for the Study Section score? My memory is a little dim 25 years later, but I do remember two things. One was a

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

reverberation of the review of my ﬁrst application to work on HLA proteins (the amount that could be obtained was too small to make crystallization and subsequent analysis by Xray crystallography feasible). Neither I nor the reviewers could know that the techniques for X-ray crystallography would improve dramatically in the next few years. The other criticism that I remember was that the application was not sufﬁciently well written or focused. I went back and looked at the application, of course, and it did not look that bad to me. I did not see much to change, but I did go through it and improved its language and focus. I also asked on resubmission that it be reviewed by a different Study Section. The new Study Section gave it a dramatically different score, in my memory around 110, well within the funding range at that time. The message is that even an established investigator has difﬁculty being funded when he wants to branch out and do something new, different, and difﬁcult. If established investigators are not encouraged to undertake important, difﬁcult problems, then who is going to tackle them? Something needs to be done to improve the system in that respect for both junior and senior investigators. Pamela Bjorkman worked in my lab for about a year or a year and a half. By that time, she had crystals of HLA-A2 (MAC, the ﬁrst polymorphic leukocyte antigen that had been detected by Jean Dausset). The crystals were small, thin, and fragile. She then moved to the Wiley lab to learn X-ray crystallography, and she completed her thesis as a joint student with Don and me. The thin crystals did not do well in a conventional X-ray source, i.e., they disintegrated so that a complete data set could not be obtained. However, sufﬁcient information was obtained for a ﬁrst publication (59). Importantly, the data showed that the molecule had approximate twofold rotational symmetry. In the illustration that Jim Kaufman created comparing MHC class I and MHC class II proteins, only one of several possible alignments for the three domains of the heavy chain of the MHC class I protein was shown, the one favored by Jim that made

MHC class I look much like MHC class II. For example, in one alignment not shown in Jim’s illustration, the three domains of the heavy chain would be linear. Pam’s initial Xray data showing twofold rotational symmetry supported Jim’s guess and certainly informed the illustration. The next important event was the advent of synchrotron radiation sources for protein crystallography. One of the ﬁrst available was DESY at Hamburg, Germany, so Pamela and Bill Bennett, one of Don’s postdocs, made frequent transatlantic trips to collect data there. Later, data were collected at the Cornell High Energy Synchrotron Source (CHESS) when it became available for protein crystallography, which of course was much more convenient. High-intensity X-rays produced by a synchrotron source were much less destructive to the small, thin crystals of HLA-A2, so complete native and heavy atom derivative data sets could be obtained, and, with important help from other members of Don’s laboratory, particularly Mark Saper, the structure was solved (60). One seemingly trivial but important problem was also solved. The technique for identifying amino acids at each step of degradation of a protein or peptide was not as precise in the late 1970s as it is today, and a few errors in the sequence were corrected by the sequence obtained from a genomic clone (61). The X-ray data ﬁt the sequence that was determined. It was a bombshell, now so well known to most of you that it does not need any description other than the structure that I show in Figure 2. Two striking features of the structure are the “extra density” that lies in a cleft between the two α-helices, a mixture of foreign and self peptides that are presented to the immune system and the location of polymorphic residues. The presence of the extra density and the need to establish which part of the X-ray diffraction pattern was due to the extra density and which part to the main chain of the HLA-A2 molecule were the last and most difﬁcult parts of the structure to establish. Nearly all the polymorphic residues that www.annualreviews.org • My Scientific Journey

19

ANRV270-IY24-01

ARI

13 February 2006

11:14

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

characterized MHC class I sequences were pointing into the peptide-binding cleft (62)! Thus, it was clear that the polymorphism was designed to establish in the population a variety of MHC proteins that could bind a very

large number of peptides. The evolutionary force that had driven it was the need to protect the population as a whole from catastrophic variation in an infectious agent that might otherwise decimate the population, or from differences in the geographic distribution of infectious agents. Next, Tom Garrett, along with Pam and Mark, solved the structure of HLA-Aw68, which differed from HLA-A2 by only 13 residues and demonstrated the existence of speciﬁcity pockets that accommodate the side chains of peptide antigens (anchor residues) (63). Joan Gorga solubilized and crystallized HLA-B27, and its structure was solved by Dean Madden (64, 65). It was notable because ˚ permitthe high-resolution data set (2.1 A) ted details of the mode of peptide binding to ←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 2 Structures of class I and class II MHC proteins as determined by 1994. MHC structures are in the left column and MHC class II are in the right column. (a) Ribbon diagrams are shown. The clefts between the two helices at the top are the sites of peptide binding (60, 62, 69, 71). (b–e) Peptides in the clefts of class I and class II MHC proteins viewed from the top and looking down into the clefts. (b) van der Waals surface representation of the mixture of peptides in the clefts of HLA-A2 and HLA-DR1 (60, 69, 106). (c) Single peptide complexes of the inﬂuenza nucleoprotein peptide 91–99 in the cleft of HLA-A68 and of the inﬂuenza hemagglutinin peptide 306–318 in the cleft of HLA-DR1 (66, 71). (d) Side views show the location of anchor residues of the peptides that extend into pockets in the structure of both classes of molecules. Those at P1 and P9 in both class I and class II proteins are particularly evident, but other anchor residues in each class are also important. (e) Mechanism of peptide binding to class I and class II MHC proteins. Hydrogen bonds with the main chains of the bound peptides are shown: class I (HLA-A2) and class II (HLA-DR1). Note that in class I hydrogen bonds are located at the two ends, whereas in class II they are distributed all along the length of the peptide (107–109). These ﬁgures are the result of 15 years of collaboration between members of Don Wiley’s laboratory and my laboratory between 1979 and 1994. Don died accidentally in 2001, and this ﬁgure honors his memory.

20

Strominger

ANRV270-IY24-01

ARI

13 February 2006

11:14

Table 2 Yields of purified MHC antigensa Antigen

Yield mg/10 g cells

DR

12

DQ

2

DP

0.2

Class I

0.5

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

a

First published in Reference 67.

be described. The single-peptide complex of HLA-Aw68 binding an inﬂuenza virus peptide was also solved (66). While all of this work was going on, people in my lab were carefully examining the possibility of crystallizing a class II MHC protein. The ﬁrst problem was to separate the three isotypes, HLA-DR, -DQ, and -DP. Immunoafﬁnity chromatography and monoclonal antibodies speciﬁc for each of the isotypes made that possible (67). Fortunately, the yield of the major MHC class II isotype HLA-DR was much greater than that of MHC class I isotypes (Table 2), but HLA-DR was resistant to papain proteolysis. After that problem was solved, seven allotypes of HLADR were crystallized by Joan Gorga (68). All the crystals were the same, relatively small and fragile, and HLA-DR1 was chosen for further study, probably because it had the number 1. Progress was again slow, but eventually a structure was obtained from data assembled from three different types of crystal, the native molecule (Jerry Brown), a superantigen complex of it (Ted Jardetzky), and a single peptide complex of hemagglutinin 306–318 bound to HLA-DR1 that had been obtained in the meantime using recombinant DNA technology to produce it (Larry Stern) (all three in Don’s lab) (69–71) (Figure 2). There are many differences between MHC class II and MHC class I proteins, but in outline the molecules look remarkably similar. The major difference is that the peptide-binding cleft is open at both ends in MHC class II, so that the peptide extends through the cleft, consistent with the fact that long peptides had been obtained after denaturation of HLA-DR pro-

teins and sequenced (71a,b). Thirteen amino acids lay within the cleft, the central nine residues being bound by anchor residues in pockets (as is true for the MHC class I proteins) and by substantial hydrogen bonds between the peptide backbone and the MHC class II protein. Much more work followed on the structures of both class I and class II MHC proteins, both in our laboratories and those of others, but none of it changed the outlines in the structures that had been determined. I am often asked how Don and I interacted during these exciting years. Don and I had very different styles and interests, and maybe that is why the collaboration was so successful. He was very focused and precise, and I was much more interested in where the problem was going to go. In the beginning, Don wanted me to work directly with him and learn to be a crystallographer, but he may have soon realized that I was a slow learner. I knew some of the language from Max Perutz’ course, and I learned much, much more from Don, his students, and his post-docs. However, I focused on the proteins. After Pam Bjorkman, Mark Saper, Tom Garrett, and others solved the structures of HLA-A2 and HLA-Aw68, products of alleles at the HLA-A locus, Joan Gorga solubilized and crystallized HLA-B27, and Dean Madden (in Don’s lab) solved its structure. Joan, together with Vaclav Horejsi and David Johnson, separated the three MHC class II isotypes and then solubilized and crystallized the HLA-DR proteins. Jerry Brown, Ted Jardetsky, and Larry Stern (all in Don’s lab) did the difﬁcult structure determination that was ﬁnally solved by combining data from the three different crystals. Don became one of the outstanding structural biologists in the world, and I was very fortunate to have him as a natural collaborator within the department. Don’s ofﬁce and lab were on the ﬁrst ﬂoor of the building, and mine were on the fourth. When I came into the building in the morning and passed the door that led to his lab and ofﬁce, I would often stop for a few minutes, most often with nothing particular in mind, but www.annualreviews.org • My Scientific Journey

21

ARI

13 February 2006

11:14

sometimes with some speciﬁc question. I always came away stimulated and with new insights. Don died accidentally in 2001 (71c). He left an enormous intellectual hole for me and for our department. I loved the way he did science. My own interests turned after that to attempting to use the structural information to further our understanding of autoimmune diseases. Nearly all these diseases are linked to speciﬁc allotypes of class II proteins (72). An important drug for the treatment of multiple sclerosis (MS) (Copolymer 1, Copaxone® , Glatiramer acetate) was developed by Michael Sela, Ruth Arnon, and their colleagues in Israel (73) and is in wide use for treatment today. Although it reduces the frequency of relapses by only 30%, this reduction is highly signiﬁcant for patients with this disease. This disease is linked to HLA-DR2, and by using the structural information obtained for the manner in which MBP(85−99) (a myelin basic protein region that is thought to be encephalitogenic in MS and is thus a major candidate autoantigen for the disease) is bound to the MHC class II protein, new copolymers were designed that were more effective in binding to HLA-DR2 and in inhibiting MBP(85−99) -speciﬁc, HLA-DR2-restricted T cell lines. Importantly, they are also more effective in ameliorating experimental autoimmune encephalomyelitis, the rodent model of MS, using several different mouse models and modes of administration of the copolymers (74–77). However, mice are not humans, and only clinical trials will answer the question whether these new molecules are improvements over Copolymer 1. Work on these materials is ongoing.

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

NATURAL KILLER CELLS Soon after the identiﬁcation of many new genes in the class III region of the human MHC, Moretta and colleagues, who had been studying human NK cells, reported that a resistance gene that regulated NK cell activity was linked to the MHC (78). They sug22

Strominger

gested that we collaborate in identifying this gene, believing that it might be one of the newly discovered genes. However, immunogenetic studies established that the resistance gene was located in the large 2-mb region between HLA-B and HLA-A (79). Although several genes had been identiﬁed in this region, only one of them, HLA-C, was known to be functional and expressed on the cell surface. Moretta and colleagues had also cloned NK cells from human peripheral blood and found that the two dominant sets of clones were mutually exclusive in their ability to lyse NK targets, i.e., cells that were lysed by group 1 clones were not lysed by group 2 clones and vice versa (80). A few target cells were resistant to both group I and group II clones. Marco Colonna, then a postdoctoral fellow, noticed that HLA-C was also dimorphic (81). One group of HLA-C allotypes had serine at position 77 and asparagine at position 80, whereas the second group of allotypes had asparagine 77 and lysine 80. He quickly established the relationship between the two groups of HLAC allotypes and the two groups of human NK clones, i.e., NK 1 clones were inhibited by members of HLA-C allotype group II, and NK 2 clones were inhibited by members of HLA-C allotype group molecules (82). Later, residue 80 was shown to be critical for the inhibitory function of HLA-C (83). These studies greatly clariﬁed how NK clones recognize (or fail to recognize) target cells. Later, Marco cloned the NK inhibitory receptors, now called KIR2DL1, 2, or 3 (as did another group) (84, 85). He also showed that the serological speciﬁcities Bw4 and Bw6 that had been deﬁned by van Rood and his colleagues corresponded to a similar dimorphism of HLA-B allotypes at residue 80 that were related to inhibition of lysis by other types of NK clones (86). The receptor for this inhibition, KIR3DL1, was subsequently cloned (87). My laboratory has continued to be involved in studies of human NK cells, principally in trying to understand the formation and structure of inhibitory and activating NK synapses and their relationships to each other

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

(88, 89; K. Krzewski, X. Chen, J.S. Orange, and J. Strominger, manuscript submitted). In addition, we are engaged in studies of human decidual NK cells, which represent 70%–90% of the lymphocytes found in the uterine decidua during pregnancy (91–93). These cells are likely involved in maternal fetal tolerance. The enigma of acceptance of the hemiallogeneic fetal graft by the pregnant female was pointed out by Peter Medawar more than 50 years ago, and its explanation is still incomplete. Both of these topics also represent current research projects in my laboratory.

THE DANA-FARBER CANCER INSTITUTE, 1974-PRESENT Soon after I moved to Harvard in 1968, Emil (Tom) Frei, my classmate at Yale Medical School and a close friend, became the chief of medicine of what was then called the Children’s Cancer Research Foundation, founded by Sidney Farber. Dr. Farber was the director, but he died a few years later, and Tom then became the director. A large new research building that included a small research hospital (the Dana Building) was under construction. It was initially called the Sidney Farber Cancer Institute, but the name was later changed to the Dana-Farber Cancer Institute as a result of a large contribution by the Dana Foundation. The new building had several hundred thousand square feet of research space, and stafﬁng it was a huge opportunity and responsibility. Around that time, I had begun to work on EBV, which had been discovered only a few years earlier as an agent closely associated with Burkitt’s lymphoma, a disease endemic in the malaria belt of Africa. EBV was soon found in sporadic cases in the United States, in endemic malarial regions of the world, and in other regions. EBV was particularly useful in research because it transformed human lymphocytes in vitro and thus provided a means of establishing cell lines that express a variety of HLA haplotypes, including homozygous cell lines. EBV is a very large herpes virus, and the mechanism by which it transformed lym-

phocyte was unknown. One of my graduate students, Roger Yocum, thought it would be interesting to study, and I thought it would be an interesting diversiﬁcation of our work. However, soon after we began Jim Watson, then a member of the department, became concerned that growing the virus in our relatively poor facilities could be dangerous, even though the virus is found in the throats of nearly all individuals. Neither the potential for increased tumorigenicity as a result of virus variation nor the potential for large doses of virus in aerosols to produce disease were well understood. He was uncomfortable with my growing it in our building. I then asked Tom whether he had a small lab I could use for growing EBV at the Farber, as the institute is called. He responded by asking me to become the institute’s director of Basic Sciences and to help him get it off to a good start. After some discussion, I agreed to do it for three years. It was not an easy job. Sidney Farber had not been especially careful in appointing staff, very few of whom had Harvard appointments. One ground rule we established was that all future appointments would need to be Harvard faculty appointments. There was also tension between excellent clinical scientists and the basic scientists that Tom and I both wanted to bring into the institute. I had, and still have, the strong view that clinical science prospers in the presence of outstanding basic science, and vice versa. I am very proud of the excellent trajectory for science that was established by the appointments made in those three years. That trajectory has been maintained until the present time, so that the Farber has become an outstanding scientiﬁc institution as well as an outstanding center for clinical research and treatment. At the end of three years, I returned full-time to my position in Cambridge, Massachusetts, but Tom asked me to keep a laboratory at the Farber, and excellent work was carried out there by postdoctoral fellows, some of whom became faculty members. Notable among those contributions were the www.annualreviews.org • My Scientific Journey

23

ARI

13 February 2006

11:14

cloning of the EBV genome (Jim Skare) (94); the identiﬁcation of a very large polycistronic message derived from EBV and its differential splicing, together with the subsequent identiﬁcation of the complex promoter usage by EBV (Sam Speck) (95); the identiﬁcation of C3d as the EBV receptor (Joyce Fingeroth) (96); the identiﬁcation of VLA (very late antigens, now called the β1-integrin family of molecules) by Martin Hemler (97); the identiﬁcation of a second T cell receptor, the γδ T cell receptor, by Mike Brenner and Mike Krangel (98); the identiﬁcation of the two subunits of the TAP transporter as well as the ﬁrst genetic map of the entire MHC, including the identiﬁcation of many new genes by Thomas Spies (99); and most recently the discovery of the role of caspases and nitric oxide in the maturation of dendritic cells and the light this discovery shed on the intracellular transport of proteins, including MHC class II, by Laura Santambrogio and Siew-Heng Wong (100, 101). An extremely fruitful and pleasant collaboration with Steve Burakoff resulted in the ﬁrst suggestion that MHC class II–restricted T cell clones are CD4+ , whereas MHC class I–restricted clones are CD8+ (102). This collaboration was my introduction to cellular immunology, and I learned an enormous amount.

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

NIH R35 CA 47554 (OUTSTANDING INVESTIGATOR GRANT) Having taken several digs at NIH several times above, I have to conclude by expressing my deep appreciation of the support I have received over my career, most speciﬁcally the support of the Outstanding Investigator Grant. In 1976, when I ﬁnally had organized a laboratory at the Farber, I had seven or eight R01 grants to support the two laboratories. An enormous amount of time was being spent on writing up competing renewals and noncompeting renewals. I learned about the OIG awards and applied. When I re-

24

Strominger

ceived the award, program ofﬁcers at the National Cancer Institute (NCI) were extremely generous. It was an umbrella grant that replaced all my existing grants, including those from other institutes and my EBV grant from NCI. The yearly noncompeting renewal process was greatly simpliﬁed, and after seven years the competing renewal was a review of progress made under that grant by a panel constituted for that purpose. In many ways, the OIGs were equivalent to the investigatorships offered by the Howard Hughes Medical Institute, but they were not encumbered by the necessity of attending annual meetings or including NCI in the title lines of papers as one of the institutions from which the work had come. My grant could not have been administered in a better way. After 13 years, however, NCI terminated the program. I am not sure why. I believe that programs of this type would greatly simplify the administrative burdens now placed on investigators. The OIG and NCI played a very large part in my success. Immunoland has truly been a wonderland for me since I entered the ﬁeld as a biochemist in about 1970. During this period, two extremely important immunological tools were introduced, the generation of monoclonal antibodies by George Kohler and Cesar Milstein (103) and the development of the ﬂuorescence activated cell sorter (FACS) by Leonard Herzenberg (104). Without these inventions, progress would have been much, much slower. Developmental processes, cellcell interactions, signaling processes, and biochemical mechanisms in the immune system were all made more accessible to investigation during the period that I worked in this ﬁeld. I was very fortunate to be there at the beginning. I conclude by thanking and expressing my deep gratitude to the many postdoctoral fellows, graduate and undergraduate students, research assistants, administrative assistants, laboratory aides, and collaborators who worked with me during this adventure.

ANRV270-IY24-01

ARI

13 February 2006

11:14

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LITERATURE CITED 1. Strominger J. 1948. A national science foundation. Yale. J. Biol. Med. 20:577 2. Park JT. 1952. Uridine-5 -pyrophosphate derivatives. II. Isolation from Staphylococcus aureus. J. Biol. Chem. 194:877–84 3. Reissig JL, Strominger JL, Leloir LF. 1955. A modiﬁed colorimetric method for the estimation of N-acetylamino sugars. J. Biol. Chem. 217:959–66 4. Strominger JL. 1957. Microbial uridine-5 -pyrophosphate N-acetylamino sugar compounds. I. Biology of the penicillin-induced accumulation. J. Biol. Chem. 224:509– 23 5. Strominger JL, Heppel LA, Maxwell ES. 1959. Nucleoside monophosphate kinases. I. Transphosphorylation between adenosine triphosphate and nucleoside monophosphates. Biochim. Biophys. Acta 32:412–21 6. Strominger JL, Kalckar HM, Axelrod J, Maxwell ES. 1954. Enzymatic oxidation of uridine diphosphate glucose to uridine diphosphate glucuronic acid. J. Am. Chem. Soc. 76: 6411 7. Anderson EP, Isselbacher KJ, Kalckar HM. 1957. Defect in uptake of galactose-1phosphate into liver nucleotides in congenital galactosemia. Science 125:113–14 8. Strominger JL. 1959. The amino acid sequence in the uridine nucleotide-peptide from Staphylococcus aureus. Comp. Rend. Trav. Lab. Carlsberg 31:181 9. Park JT, Strominger JL. 1957. Mode of action of penicillin. Science 125:99–101 10. Strominger JL, Park JT, Thompson RE. 1959. Composition of the cell wall of Staphylococcus aureus: its relation to the mechanism of action of penicillin. J. Biol. Chem. 234:3263–68 11. Nathenson SG, Strominger JL. 1961. Effects of penicillin on the biosynthesis of the cell walls of Escherichia coli and Staphylococcus aureus. J. Pharmacol. Exp. Ther. 131:1–6 12. Strominger JL, Ghuysen JM. 1967. Mechanisms of enzymatic bacteriaolysis. Cell walls of bacteria are solubilized by action of either speciﬁc carbohydrases or speciﬁc peptidases. Science 156:213–21 13. Strominger JL. 1970. Penicillin-sensitive enzymatic reactions in bacterial cell wall synthesis. In Harvey Lectures, pp. 179–213. New York: Academic 14. Meadow PM, Anderson JS, Strominger JL. 1964. Enzymatic polymerization of UDPacetylmuramyl.l-ala.d-glu.l-lys.d-ala.d-ala and UDP-acetylglucosamine by a particulate enzyme from Staphylococcus aureus and its inhibition by antibiotics. Biochem. Biophys. Res. Commun. 14:382–87 15. Strominger JL, Okazaki T, Okazaki R. 1963. Oxidation and reduction of nucleotidelinked sugars. In The Enzymes, ed. P Boyer, H Lardy, K Myrback, p. 161. New York: Academic 16. Kobayashi KS, Chamaillard M, Ogura Y, Henegariu O, Inohara N, et al. 2005. Nod2dependent regulation of innate and adaptive immunity in the intestinal tract. Science 307:731–34 17. Juergens WG, Sanderson AR, Strominger JL. 1963. Chemical basis for an immunological speciﬁcity of a strain of Staphylococcus aureus. J. Exp. Med. 117:925–35 18. Krueger L, Luederitz O, Strominger JL, Westphal O. 1962. [On the immunochemistry of O antigens of Enterobacteriaceae. VII. The relation of hexoses and 6-desoxyhexoses in Salmonella lipopolysaccharides to the D and L group.]. Biochem Z 335:548–58 19. Izaki K, Matsuhashi M, Strominger JL. 1966. Glycopeptide transpeptidase and D-alanine carboxypeptidase: penicillin-sensitive enzymatic reactions. Proc. Natl. Acad. Sci. USA 55:656–63

www.annualreviews.org • My Scientific Journey

25

ARI

13 February 2006

11:14

20. Tipper DJ, Strominger JL. 1965. Mechanism of action of penicillins: a proposal based on their structural similarity to acyl-D-alanyl-D-alanine. Proc. Natl. Acad. Sci. USA 54:1133– 41 21. Lee W, McDonough MA, Kotra L, Li ZH, Silvaggi NR, et al. 2001. A 1.2-A snapshot of the ﬁnal step of bacterial cell wall biosynthesis. Proc. Natl. Acad. Sci. USA 98:1427– 31 22. Anderson JS, Matsuhashi M, Haskin MA, Strominger JL. 1965. Lipid-Phosphoacetylmuramyl-pentapeptide and lipid-phosphodisaccharide-pentapeptide: presumed membrane transport intermediates in cell wall synthesis. Proc. Natl. Acad. Sci. USA 53:881– 89 23. Higashi Y, Strominger JL, Sweeley CC. 1967. Structure of a lipid intermediate in cell wall peptidoglycan synthesis: a derivative of a C55 isoprenoid alcohol. Proc. Natl. Acad. Sci. USA 57:1878–84 24. Matsuhashi M, Strominger JL. 1964. Thymidine diphosphate 4-acetamido-4, 6dideoxyhexoses. I. Enzymatic synthesis by strains of Escherichia coli. J. Biol. Chem. 239:2454–63 25. Roberts WS, Petit JF, Strominger JL. 1968. Biosynthesis of the peptidoglycan of bacterial cell walls. 8. Speciﬁcity in the utilization of L-alanyl transfer ribonucleic acid for interpeptide bridge synthesis in Arthrobacter crystallopoietes. J. Biol. Chem. 243:768–72 26. Strominger JL, Threnn RH. 1959. Accumulation of a uridine nucleotide in Staphylococcus aureus as the consequence of lysine deprivation. Biochim. Biophys. Acta 36:83–92 27. Adam A, Petit JF, Lefrancier P, Lederer E. 1981. Muramyl peptides. Chemical structure, biological activity and mechanism of action. Mol. Cell. Biochem. 41:27–47 28. Janeway CA Jr, Medzhitov R. 2002. Innate immune recognition. Annu. Rev. Immunol. 20:197–216 29. Gorer PA. 1937. The genetic and antigenic basis for tumor transplantation. J. Pathol. Bacteriol. 44:691–97 30. Snell GD, Lyman S, Gorer PA. 1948. Studies on the genetic and antigenetic basis of tumor transplantation. Linkage between a histocompatibility gene and “fused” in mice. Proc. R. Soc. London Ser. B Biol. Sci. 135:499–505 31. Nathenson SG, Davies DA. 1966. Solubilization and partial puriﬁcation of mouse histocompatibility antigens from a membranous lipoprotein fraction. Proc. Natl. Acad. Sci. USA 56:476–83 32. Sanderson AR, Strominger JL, Nathenson SG. 1962. Chemical structure of teichoic acid from Staphylococcus aureus, strain Copenhagen. J. Biol. Chem. 237:3603–13 33. Sanderson AR, Batchelor JR. 1968. Transplantation antigens from human spleens. Nature 219:184–86 34. Turner MJ, Cresswell P, Parham P, Strominger JL, Mann DL, Sanderson AR. 1975. Puriﬁcation of papain-solubilized histocompatibility antigens from a cultured human lymphoblastoid line, RPMI 4265. J. Biol. Chem. 250:4512–19 35. Cresswell P, Turner MJ, Strominger JL. 1973. Papain-solubilized HL-A antigens from cultured human lymphocytes contain two peptide fragments. Proc. Natl. Acad. Sci. USA 70:1603–7 36. Grey HM, Kubo RT, Colon SM, Poulik MD, Cresswell P, et al. 1973. The small subunit of HL-A antigens is beta 2-microglobulin. J. Exp. Med. 138:1608–12 37. Cresswell P, Springer T, Strominger JL, Turner MJ, Grey HM, Kubo RT. 1974. Immunological identity of the small subunit of HL-A antigens and beta2-microglobulin and its turnover on the cell membrane. Proc. Natl. Acad. Sci. USA 71:2123–27

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

26

Strominger

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

38. Springer TA, Strominger JL. 1976. Detergent-soluble HLA antigens contain a hydrophilic region at the COOH-terminus and a penultimate hydrophobic region. Proc. Natl. Acad. Sci. USA 73:2481–85 39. Orr HT, Lancet D, Robb RJ, Lopez de Castro JA, Strominger JL. 1979. The heavy chain of human histocompatibility antigen HLA-B7 contains an immunoglobulin-like region. Nature 282:266–70 40. Orr HT, Lopez de Castro JA, Lancet D, Strominger JL. 1979. Complete amino acid sequence of a papain-solubilized human histocompatibility antigen, HLA-B7. 2. Sequence determination and search for homologies. Biochemistry 18:5711–20 41. Humphreys RE, McCune JM, Chess L, Herrman HC, Malenka DJ, et al. 1976. Isolation and immunologic characterization of a human B-lymphocyte-speciﬁc, cell surface antigen. J. Exp. Med. 144:99–112 42. Kaufman JF, Auffray C, Korman AJ, Shackelford DA, Strominger J. 1984. The class II molecules of the human and murine major histocompatibility complex. Cell 36:1–13 43. Shackelford DA, Mann DL, van Rood JJ, Ferrara GB, Strominger JL. 1981. Human B-cell alloantigens DC1, MT1, and LB12 are identical to each other but distinct from the HLA-DR antigen. Proc. Natl. Acad. Sci. USA 78:4566–70 44. Ploegh HL, Orr HT, Strominger JL. 1980. Molecular cloning of a human histocompatibility antigen cDNA fragment. Proc. Natl. Acad. Sci. USA 77:6081–85 45. Orr HT, Bach FH, Ploegh HL, Strominger JL, Kavathas P, DeMars R. 1982. Use of HLA loss mutants to analyse the structure of the human major histocompatibility complex. Nature 296:454–56 46. Koller BH, Geraghty DE, DeMars R, Duvick L, Rich SS, Orr HT. 1989. Chromosomal organization of the human major histocompatibility complex class I gene family. J. Exp. Med. 169:469–80 47. Korman AJ, Knudsen PJ, Kaufman JF, Strominger JL. 1982. cDNA clones for the heavy chain of HLA-DR antigens obtained after immunopuriﬁcation of polysomes by monoclonal antibody. Proc. Natl. Acad. Sci. USA 79:1844–48 48. Korman AJ, Auffray C, Schamboeck A, Strominger JL. 1982. The amino acid sequence and gene organization of the heavy chain of the HLA-DR antigen: homology to immunoglobulins. Proc. Natl. Acad. Sci. USA 79:6013–17 49. Auffray C, Kuo J, DeMars R, Strominger JL. 1983. A minimum of four human class II alpha-chain genes are encoded in the HLA region of chromosome 6. Nature 304:174–77 50. Roux-Dosseto M, Auffray C, Lillie JW, Boss JM, Cohen D, et al. 1983. Genetic mapping of a human class II antigen beta-chain cDNA clone to the SB region of the HLA complex. Proc. Natl. Acad. Sci. USA 80:6036–40 51. Steinmetz M, Minard K, Horvath S, McNicholas J, Srelinger J, et al. 1982. A molecular map of the immune response region from the major histocompatibility complex of the mouse. Nature 300:35–42 52. Okada K, Boss JM, Prentice H, Spies T, Mengler R, et al. 1985. Gene organization of DC and DX subregions of the human major histocompatibility complex. Proc. Natl. Acad. Sci. USA 82:3410–14 53. Spies T, Sorrentino R, Boss JM, Okada K, Strominger JL. 1985. Structural organization of the DR subregion of the human major histocompatibility complex. Proc. Natl. Acad. Sci. USA 82:5165–69 54. Spies T, Morton CC, Nedospasov SA, Fiers W, Pious D, Strominger JL. 1986. Genes for the tumor necrosis factors α and β are linked to the human major histocompatibility complex. Proc. Natl. Acad. Sci. USA 83:8699–702 www.annualreviews.org • My Scientific Journey

27

ARI

13 February 2006

11:14

55. Carroll MC, Katzman P, Alicot EM, Koller BH, Geraghty DE, et al. 1987. Linkage map of the human major histocompatibility complex including the tumor necrosis factor genes. Proc. Natl. Acad. Sci. USA 84:8535–39 56. Spies T, Blanck G, Bresnahan M, Sands J, Strominger JL. 1989. A new cluster of genes within the human major histocompatibility complex. Science 243:214 –17 57. Bahram S, Spies T. 1996. The MIC gene family. Res. Immunol. 147:328–33 58. Orr HT, Lopez de Castro JA, Parham P, Ploegh HL, Strominger JL. 1979. Comparison of amino acid sequences of two human histocompatibility antigens, HLA-A2 and HLA-B7: location of putative alloantigenic sites. Proc. Natl. Acad. Sci. USA 76:4395– 99 59. Bjorkman PJ, Strominger JL, Wiley DC. 1985. Crystallization and X-ray diffraction studies on the histocompatibility antigens HLA-A2 and HLA-A28 from human cell membranes. J. Mol. Biol. 186:205–10 60. Bjorkman PJ, Saper MA, Samraoui B, Bennett WS, Strominger JL, Wiley DC. 1987. Structure of the human class I histocompatibility antigen, HLA-A2. Nature 329:506– 12 61. Koller BH, Orr HT. 1985. Cloning and complete sequence of an HLA-A2 gene: analysis of two HLA-A alleles at the nucleotide level. J. Immunol. 134:2727–33 62. Bjorkman PJ, Saper MA, Samraoui B, Bennett WS, Strominger JL, Wiley DC. 1987. The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature 329:512–18 63. Garrett TP, Saper MA, Bjorkman PJ, Strominger JL, Wiley DC. 1989. Speciﬁcity pockets for the side chains of peptide antigens in HLA-Aw68. Nature 342:692–96 64. Gorga JC, Madden DR, Prendergast JK, Wiley DC, Strominger JL. 1992. Crystallization and preliminary X-ray diffraction studies of the human major histocompatibility antigen HLA-B27. Proteins 12:87–90 65. Madden DR, Gorga JC, Strominger JL, Wiley DC. 1992. The three-dimensional structure of HLA-B27 at 2.1 A˚ resolution suggests a general mechanism for tight peptide binding to MHC. Cell 70:1035–48 66. Silver ML, Guo HC, Strominger JL, Wiley DC. 1992. Atomic structure of a human MHC molecule presenting an inﬂuenza virus peptide. Nature 360:367–69 67. Gorga JC, Horejsi V, Johnson DR, Raghupathy R, Strominger JL. 1987. Puriﬁcation and characterization of class II histocompatibility antigens from a homozygous human B cell line. J. Biol. Chem. 262:16087–94 68. Gorga JC, Brown JH, Jardetzky T, Wiley DC, Strominger JL. 1991. Crystallization of HLA-DR antigens. Res. Immunol. 142:401–7 69. Brown JH, Jardetzky TS, Gorga JC, Stern LJ, Urban RG, et al. 1993. Three-dimensional structure of the human class II histocompatibility antigen HLA-DR1. Nature 364:33– 39 70. Jardetzky TS, Brown JH, Gorga JC, Stern LJ, Urban RG, et al. 1994. Three-dimensional structure of a human class II histocompatibility molecule complexed with superantigen. Nature 368:711–18 71. Stern LJ, Brown JH, Jardetzky TS, Gorga JC, Urban RG, et al. 1994. Crystal structure of the human class II MHC protein HLA-DR1 complexed with an inﬂuenza virus peptide. Nature 368:215–21 71a. Chicz RM, Urban RG, Lane WS, Gorga JC, Stern LJ, et al. 1992. Predominant naturally processed peptides bound to HLA-DR1 are derived from MHC-related molecules and are heterogeneous in size. Nature 358:764–68

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

28

Strominger

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

71b. Chicz RM, Urban RG, Gorga JC, Vignali DAA, Lane WS, Strominger JL. 1993. Speciﬁcity and promiscuity among naturally processed peptides bound to HLA-DR alleles. J. Exp. Med. 178:27–47 71c. Strominger JL. 2002. Don Craig Wiley (1944–2001): a reminiscence. Nat. Immunol. 2:103–4 72. Svejgaard A, Platz P, Ryder LP. 1983. HLA and disease 1982—a survey. Immunol. Rev. 70:193–218 73. Teitelbaum D, Webb C, Meshorer A, Arnon R, Sela M. 1973. Suppression by several synthetic polypeptides of experimental allergic encephalomyelitis induced in guinea pigs and rabbits with bovine and human basic encephalitogen. Eur. J. Immunol. 3:273–79 74. Fridkis-Hareli M, Santambrogio L, Stern JN, Fugger L, Brosnan C, Strominger JL. 2002. Novel synthetic amino acid copolymers that inhibit autoantigen-speciﬁc T cell responses and suppress experimental autoimmune encephalomyelitis. J. Clin. Invest. 109:1635–43 75. Stern JN, Illes Z, Reddy J, Keskin DB, Sheu E, et al. 2004. Amelioration of proteolipid protein 139–151-induced encephalomyelitis in SJL mice by modiﬁed amino acid copolymers and their mechanisms. Proc. Natl. Acad. Sci. USA 101:11743–48 76. Illes Z, Stern JN, Reddy J, Waldner H, Mycko MP, et al. 2004. Modiﬁed amino acid copolymers suppress myelin basic protein 85–99-induced encephalomyelitis in humanized mice through different effects on T cells. Proc. Natl. Acad. Sci. USA 101:11749–54 77. Stern JN, Illes Z, Reddy J, Keskin DB, Fridkis-Hareli M, et al. 2005. Peptide 15-mers of deﬁned sequence that substitute for random amino acid copolymers in amelioration of experimental autoimmune encephalomyelitis. Proc. Natl. Acad. Sci. USA 102:1620–25 78. Ciccone E, Pende D, Viale O, Tambussi G, Ferrini S, et al. 1990. Speciﬁc recognition of human CD3-CD16+ natural killer cells requires the expression of an autosomic recessive gene on target cells. J. Exp. Med. 172:47–52 79. Ciccone E, Colonna M, Viale O, Pende D, Di Donato C, et al. 1990. Susceptibility or resistance to lysis by alloreactive natural killer cells is governed by a gene in the human major histocompatibility complex between BF and HLA-B. Proc. Natl. Acad. Sci. USA 87:9794–97 (and correction in Proc. Natl. Acad. USA 88:5477) 80. Ciccone E, Pende D, Viale O, Di Donato C, Tripodi G, et al. 1992. Evidence of a natural killer (NK) cell repertoire for (allo) antigen recognition: deﬁnition of ﬁve distinct NKdetermined allospeciﬁcities in humans. J. Exp. Med. 175:709–18 81. Colonna M, Spies T, Strominger JL, Ciccone E, Moretta A, et al. 1992. Alloantigen recognition by two human natural killer cell clones is associated with HLA-C or a closely linked gene. Proc. Natl. Acad. Sci. USA 89:7983–85 82. Colonna M, Borsellino G, Falco M, Ferrara GB, Strominger JL. 1993. HLA-C is the inhibitory ligand that determines dominant resistance to lysis by NK1- and NK2-speciﬁc natural killer cells. Proc. Natl. Acad. Sci. USA 90:12000–4 83. Mandelboim O, Reyburn HT, Vales-Gomez M, Pazmany L, Colonna M, et al. 1996. Protection from lysis by natural killer cells of group 1 and 2 speciﬁcity is mediated by residue 80 in human histocompatibility leukocyte antigen C alleles and also occurs with empty major histocompatibility complex molecules. J. Exp. Med. 184:913–22 84. Colonna M, Samaridis J. 1995. Cloning of immunoglobulin-superfamily members associated with HLA-C and HLA-B recognition by human natural killer cells. Science 268:405– 8 85. Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, et al. 1995. Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains. Immunity 2:439–49 www.annualreviews.org • My Scientific Journey

29

ARI

13 February 2006

11:14

86. Cella M, Longo A, Ferrara GB, Strominger JL, Colonna M. 1994. NK3-speciﬁc natural killer cells are selectively inhibited by Bw4-positive HLA alleles with isoleucine 80. J. Exp. Med. 180:1235–42 87. D’Andrea A, Chang C, Franz-Bacon K, McClanahan T, Phillips JH, Lanier LL. 1995. Molecular cloning of NKB1. A natural killer cell receptor for HLA-B allotypes. J. Immunol. 155:2306–10 88. Davis DM, Chiu I, Fassett M, Cohen GB, Mandelboim O, Strominger JL. 1999. The human natural killer cell immune synapse. Proc. Natl. Acad. Sci. USA 96:15062–67 89. Orange JS, Harris KE, Andzelm MM, Valter MM, Geha RS, Strominger JL. 2003. The mature activating natural killer cell immunologic synapse is formed in distinct stages. Proc. Natl. Acad. Sci. USA 100:14151–56 90. Deleted in proof 91. Orange JS, Fassett MS, Koopman LA, Boyson JE, Strominger JL. 2002. Viral evasion of natural killer cells. Nat. Immunol. 3:1006–12 92. Koopman LA, Kopcow HD, Rybalov B, Boyson JE, Orange JS, et al. 2003. Human decidual natural killer cells are a unique NK cell subset with immunomodulatory potential. J. Exp. Med. 198:1201–12 93. Kopcow HD, Allan DS, Ge B, Rybalov B, Andzelm MM, et al. 2005. Human decidual NK cells form immature activating synapses and are not cytotoxic. Proc. Natl. Acad. Sci. USA. 43:15563–68 94. Skare J, Strominger JL. 1980. Cloning and mapping of BamHi endonuclease fragments of DNA from the transforming B95-8 strain of Epstein-Barr virus. Proc. Natl. Acad. Sci. USA 77:3860–64 95. Schaefer BC, Woisetschlaeger M, Strominger JL, Speck SH. 1991. Exclusive expression of Epstein-Barr virus nuclear antigen 1 in Burkitt lymphoma arises from a third promoter, distinct from the promoters used in latently infected lymphocytes. Proc. Natl. Acad. Sci. USA 88:6550–54 96. Fingeroth JD, Weis JJ, Tedder TF, Strominger JL, Biro PA, Fearon DT. 1984. EpsteinBarr virus receptor of human B lymphocytes is the C3d receptor CR2. Proc. Natl. Acad. Sci. USA 81:4510–14 97. Hemler ME, Ware CF, Strominger JL. 1983. Characterization of a novel differentiation antigen complex recognize by a monoclonal antibody (A-1A5): unique activation-speciﬁc molecular forms on stimulated T cells. J. Immunol. 131:334–40 98. Brenner MB, McLean J, Dialynas DP, Strominger JL, Smith JA, et al. 1986. Identiﬁcation of a putative second T-cell receptor. Nature 322:145–49 99. Spies T, Bresnahan M, Strominger JL. 1989. Human major histocompatibility complex contains a minimum of 19 genes between the complement cluster and HLA-B. Proc. Natl. Acad. Sci. USA 86:8955–58 100. Wong SH, Santambrogio L, Strominger JL. 2004. Caspases and nitric oxide broadly regulate dendritic cell maturation and surface expression of class II MHC proteins. Proc. Natl. Acad. Sci. USA 101:17783–88 101. Santambrogio L, Potolicchio I, Fessler SP, Wong SH, Raposo G, Strominger JL. 2005. Involvement of caspase-cleaved and intact adaptor protein 1 complex in endosomal remodeling in maturing dendritic cells. Nat. Immunol. 6:1020–28 102. Krensky AM, Clayberger C, Reiss CS, Strominger JL, Burakoff SJ. 1982. Speciﬁcity of OKT4+ cytotoxic T lymphocyte clones. J. Immunol. 129:2001–3 103. Kohler G, Milstein C. 1975. Continuous cultures of fused cells secreting antibody of predeﬁned speciﬁcity. Nature 256:495–97

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

30

Strominger

Annu. Rev. Immunol. 2006.24:1-31. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-01

ARI

13 February 2006

11:14

104. Herzenberg LA, Herzenberg LA. 2004. Genetics, FACS, immunology, and redox: a tale of two lives intertwined. Annu. Rev. Immunol. 22:1–31 105. Ploegh HL, Orr HT, Strominger JL. 1981. Major histocompatibility antigens: the human (HLA-A, -B, -C) and murine (H-2K, H-2D) class I molecules. Cell 24:287–99 106. Chicz RM, Urban RG, Gorga JC, Vignali DA, Lane WS, Strominger JL. 1993. Speciﬁcity and promiscuity among naturally processed peptides bound to HLA-DR alleles. J. Exp. Med. 178:27–47 107. Stern LJ, Wiley DC. 1994. Antigenic peptide binding by class I and class II histocompatibility proteins. Structure 2:245–51 108. Madden DR. 1995. The three-dimensional structure of peptide-MHC complexes. Annu. Rev. Immunol. 13:587–622 109. Latron F, Pazmany L, Morrison J, Moots R, Saper MA, et al. 1992. A critical role for conserved residues in the cleft of HLA-A2 in presentation of a nonapeptide to T cells. Science 257:964–67

www.annualreviews.org • My Scientific Journey

31

ANRV270-IY24-02

ARI

13 February 2006

12:28

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh,1 Nacksung Kim,2 Yuho Kadono,1 Jaerang Rho,3 Soo Young Lee,4 Joseph Lorenzo,5 and Yongwon Choi1 1

Department of Pathology and Laboratory Medicine, AFCRI, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104; email: [email protected], [email protected], [email protected]

2

MRC for Gene Regulation, Chonnam National University Medical School, Gwangju 501–746, Korea; email: [email protected]

3

Department of Microbiology and RCTCG, Chungnam National University, Daejon 305–764, Korea; email: [email protected]

4

Division of Molecular Life Sciences, CCSR, Ewha Womans University, Seoul 120–750, Korea; email: [email protected]

5

Department of Medicine, University of Connecticut Health Center, Farmington, Connecticut 06030-1317; email: [email protected]

Annu. Rev. Immunol. 2006. 24:33–63 First published online as a Review in Advance on October 12, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090646 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0033$20.00

Key Words T lymphocyte, dendritic cell, osteoblast, osteoclast, TRANCE-RANK, costimulation

Abstract Studies of bone and the immune system have converged in recent years under the banner of osteoimmunology. The immune system is spawned in the bone marrow reservoir, and investigators now recognize that important niches also exist there for memory lymphocytes. At the same time, various factors produced during immune responses are capable of profoundly affecting regulation of bone. Mechanisms have evolved to prevent excessive interference by the immune system with bone homeostasis, yet pathologic bone loss is a common sequela associated with autoimmunity and cancer. There are also developmental links, or parallels, between bone and the immune system. Cells that regulate bone turnover share a common precursor with inﬂammatory immune cells and may restrict themselves anatomically, in part by utilizing a signaling network analogous to lymphocyte costimulation. Efforts are currently under way to further characterize how these two organ systems overlap and to develop therapeutic strategies that beneﬁt from this understanding. 33

ANRV270-IY24-02

ARI

13 February 2006

12:28

INTRODUCTION HSC: hematopoietic stem cell

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Osteoblast (OB): bone-producing cell that also provides signals for osteoclast formation Osteoclast (OC): specialized multinucleated giant cell that resorbs bone TNF: tumor necrosis factor TNF-related activation-induced cytokine (TRANCE): a TNF family cytokine that is essential for osteoclast development Osteoporosis: a condition characterized by decreased bone mass and density that causes bones to become fragile

34

The emergence of the niche ﬁeld of osteoimmunology represents a conceptual rethinking of multiple phenomena, relating biological events in bone and the immune system (1). The root of exploration of this interplay begins with the basic understanding that bone provides a microenvironment that is critical for the development of the hematopoietic stem cells (HSCs), from which all cells of the mammalian immune system derive, and that various immunoregulatory cytokines inﬂuence the fate of bone cells. The reasons bone is an ideal anatomic microenvironment for HSC maintenance and differentiation have become clearer with recent data indicating that bone matrix– generating osteoblasts (OBs) provide key factors to the development of the HSC niche. There is growing evidence that bone continues to play a role in adaptive immunity at steps beyond lymphocyte development. We now know that long-lived memory T and B cells return to specialized niches in the bone marrow. Why these cells are retained in the bone and what molecular and environmental factors they rely on while there represent a series of challenging questions that will clearly beneﬁt from experiments designed in the context of osteoimmunology. How the immune system exerts its inﬂuence on bone is equally interesting. Ontogenically, skeletal development proceeds independently of early development of the immune system. Hence, it is unlikely that there are developmental inﬂuences of the immune system on skeletal and marrow cavity formation. However, bone homeostasis and remodeling occur throughout life in all bony animals. Anatomically, bone marrow spaces are loosely compartmentalized, allowing immune cells and bone cells to interact and inﬂuence each other (Figure 1). Hence, bone homeostasis is often inﬂuenced by immune responses, particularly when the immune system has been activated or becomes diseased. During pathological conditions like arthritis, inﬁl-

Walsh et al.

trating lymphocytes and other mononuclear cells provide several key factors that inﬂuence bone metabolism by altering the balance between bone-forming OBs and bone-resorbing osteoclasts (OCs). In the past, whether these interactions also inﬂuence normal bone homeostasis had been unclear. However, the discovery that activated T cells express the TNF superfamily member TRANCE, coupled with the subsequent ﬁnding that TRANCE is a key differentiation factor for OCs, represents critical evidence that a link exists between normal immune cell function and bone metabolism. During their lifetimes, mammals are challenged with various infectious agents, which results in a gradual change in the composition of the T cell compartment toward an accumulation of TRANCEexpressing memory cells that preferentially reside in bone. Hence, with age the immune system might exert a greater inﬂuence on bone homeostasis. Finally, how relevant are osteoimmunologic approaches in biomedical research to the treatment of human disease and the maintenance of good health? Although most research in osteoimmunology currently focuses on inﬂammatory bone diseases, such as those associated with osteoarthritis or rheumatoid arthritis (RA), it is important to consider the vast public health implications arising from common metabolic bone diseases, such as osteoporosis, that may be caused by or associated with inﬂammatory molecules. To prevent and treat these conditions, it will be necessary to develop anti-inﬂammatory agents that have a high degree of speciﬁcity so that both the immune and bone systems may function with minimal complications. With these issues in mind, we ﬁrst brieﬂy review bone development and remodeling and then focus on several key areas of crosstalk between the bone and immune system that we believe will most beneﬁt from interdisciplinary approaches aimed at elucidating underlying cellular and molecular mechanisms of action. The work thus far of many groups on the physiologic and clinical relevance of

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

Figure 1 Schematic diagram of the bone microenvironment as a loosely compartmentalized lymphoid organ: T∗ (memory T cells and circulating T cells), B∗ (B cells, differentiation of which occurs via interaction with stromal cells; memory B cells also interact with stromal cells; in addition, there are circulating mature B cells), stromal cells (bone marrow stromal cells are of mesenchymal origin, but not fully characterized), M∗ (monocyte and its derivatives), and osteoid [newly formed, but not yet calciﬁed matrix, composed mostly of type I collagen (∼90%) and noncollagenous proteins (∼10%)].

TRANCE-RANK signaling has served as a cornerstone of osteoimmunology research. As such, we provide a current review of the role of TRANCE in the osteoimmune system. Additionally, we review the overlapping signaling networks within cells of the immune system and bone and attempt to build a conceptual framework of parallel signaling systems in immune and bone cells.

A BRIEF INTRODUCTION TO BONE Bone Development The skeleton is among the largest organs in the body. It is composed of a mineralized framework that is maintained by a complex cellular network (2). In addition to providing structural integrity, the skeleton is a

www.annualreviews.org • Insights into Osteoimmunology

Receptor activator of NF-κB (RANK): the signaling receptor for TRANCE

35

ANRV270-IY24-02

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Osteoid: uncalciﬁed bone matrix produced by osteoblasts and consisting mainly of collagen, but also of some noncollagenous proteins BMP: bone morphogenetic protein

36

12:28

storehouse for calcium, a critical ion for a variety of metabolic processes, and it is the site of hematopoiesis. Bone development occurs along two pathways. Endochondral bone, which includes the long bones and vertebrae, begins in the embryo as a cartilaginous template (3) (Figure 2). A second type of bone, called intramembranous bone (e.g., the ﬂat bones of the skull, scapula, and ileum), forms directly from the condensation of mesenchymal cells, which directly differentiate into OBs (3). A second level of organization in endochondral bone is its division into cortical and trabecular or cancellous bone. Cortical bone forms the outer surface of endochondral bones and provides the structural integrity for many of the long bones. It forms up to 80% of the skeleton and is typically dense bone with a well-organized pattern of collagen ﬁbrils that are aligned along stress lines to provide bone with maximum strength (2, 3). Trabecular bone is thinner and less well organized, and it is primarily found traversing the bone marrow space. However, in some bones with a high degree of trabecular bone, like vertebrae, trabecular bone provides much of the structural integrity. A major function of trabecular bone is to provide a large surface area for metabolic processes. Bone turnover, which consists of bone resorption (removal) and its replacement with new bone, occurs much more frequently in trabecular bone than in cortical bone. OBs initially produce an osteoid matrix that is calciﬁed extracellularly. The major structural protein of bone is type I collagen, which provides bone with a resistance to fracture in a way that is similar to the effect of reinforcing bars in modern reinforced concrete buildings. In addition, osteoid contains a large number of other noncollagen proteins that have a variety of critical functions in bone. The mineral crystal of bone is hydroxyapetite, which is a calcium-phosphate salt, containing hydroxyl ions.

Walsh et al.

Bone Cells OBs are derived from a mesenchymal progenitor cell that is multipotential and can also differentiate into marrow stromal cells and adipocytes (4) (Figure 3). The signals that regulate the decision of mesenchymal progenitor cells to form OBs are not fully understood. However, a number of critical paracrine signals and transcription factors have been identiﬁed. These include the transcription factors Runx2 and osterix, which when absent prevent OB formation, and members of the bone morphogenetic protein (BMP) family (5–7), which initiate the signals for OB differentiation. Most recently, it was demonstrated that Wnt signaling pathways are involved in the decision of the mesenchymal progenitor cell to become either an adipocyte or an OB (8–12). As matrix calciﬁes under the inﬂuence of the OB-produced enzyme bone-speciﬁc alkaline phosphatase, a portion of the OBs are entrapped in the calciﬁed matrix and persist in bone as unique cells called osteocytes. These cells are believed to sense mechanical force on bone and to send signals via cellular processes, termed canaliculi (13), that connect osteocytes to each other and to OBs on the surface of bone. OCs are specialized multinucleated giant cells that resorb bone (14). They are hematopoietic in origin and derive from a myeloid precursor that also gives rise to macrophages and dendritic cells (DCs) (Figure 3). The signals that stimulate OCs to form and resorb bone involve a series of transcription factors and paracrine cytokines, which are discussed in detail later in this review. OCs attach themselves to bone through a specialized structure called the sealing zone. This structure allows them to create a resorption space that is isolated from the extracellular space. OCs can acidify the resorption space to solubilize the mineral component of bone (14). To remove the organic components of bone, OCs produce lysosomal enzymes, including cathepsin K, that are released into

ARI

13 February 2006

12:28

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

Figure 2 Schematic diagram of endochondral bone development. During embryogenesis, the initial step in long bone development is the patterning of the future bone by cartilage (1). Shortly thereafter, a collar of intramembranous bone forms around the center of the cartilaginous shaft (2), and simultaneously the remaining cartilage near the center of the developing bone calciﬁes (3). The calciﬁed cartilage is then removed (4), and vascular cells invade the developing marrow space to establish a blood supply. The next step in this process is the production of mineralized bone by OBs (5). As the development of the embryo progresses, mineralized tissue, produced by OBs, replaces cartilage in the majority of the embryonic bone (6–7). However, an area of cartilage is maintained at either end of growing bones after birth. This forms a structure called the epiphyseal plate or growth plate, which in humans facilitates bone growth through childhood and into adolescence. Linear growth in humans ends after puberty when the growth plate is lost as a result of the actions of sex steroid hormones (8). www.annualreviews.org • Insights into Osteoimmunology

37

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

12:28

Figure 3 Schematic summary of bone cell differentiation. Mesenchymal stem cells, which also give rise to myoblasts, adipocytes, chondrocytes, and some as yet uncharacterized stromal cells (also called marrow stromal cells, marrow ﬁbroblasts, or the reticular network), differentiate into preosteoblasts and then become OBs on the bone surface. OBs either incorporate into bone as osteocytes or remain on the surface as lining cells. A common myeloid lineage precursor, which can also give rise to macrophages and dendritic cells, commits to becoming a preosteoclast and then fuses to become a mature, multinucleated osteoclast.

the resorption space (14). To facilitate the resorption process, OCs polarize their structure and form a unique element called the rufﬂed border, which allows the surface area available for active transport of H+ ions through a unique vacuolar proton pump. The products of resorption traverse the OC by a process that is termed transcytosis and leave the OC through the basolateral membrane opposite the resorption space. OCs are highly motile, move across the bone surface, and resorb relatively large areas of bone. OCs die by apoptotic processes that appear to be regulated by

38

Walsh et al.

paracrine-acting cytokines and possibly factors in bone matrix (14).

The Coupling of Bone Resorption and Formation The skeleton is a dynamic organ that is constantly remodeling. To accomplish this function there must be a tight coupling between bone resorption and formation. Although investigators have proposed a local coupling factor linking bone resorption to subsequent formation, its nature remains elusive

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

(15). During growth and continuing to midadulthood in humans, bone mass increases (16). Genetics plays a major role in the peak bone mass that is achieved. Sex steroid hormones are also critical regulators of the skeleton (2). Loss of either estrogens in women or androgens in men is associated with an enhancement of resorption rates in bone without an equivalent increase in bone formation. In women, loss of estrogens with menopause causes roughly a doubling of the rate of bone loss and increases the risk of developing osteoporosis. Because women generally live longer and have a lower peak bone mass than men, they are more prone to develop osteoporosis (2).

INTERPLAY BETWEEN OBs AND THE IMMUNE SYSTEM The Role of OBs in the HSC Niche OBs on the endosteal surface of bone, which is adjacent to the marrow cavity, function as critical support cells for HSCs in bone marrow (17–19). Investigators have shown that HSCs are adjacent to OBs and that their number is ∼2.3-fold higher in mice upon deletion of the BMP receptor 1A (BMPR1A). Significantly, BMPR1A-deﬁcient mice also have a similar increase in OB number (17). The OBs involved are on the marrow surface of the endosteum and are likely early in their commitment to the OB lineage. Adhesion of HSC and OBs appears to be mediated by interaction of N-cadherin and β-catenin (17). Similarly, it was demonstrated that expansion of the OB population in bone by stimulation of the PTH/PTHrP (parathyroid hormone/parathyroid hormone–related protein) receptor on osteoblastic cells increased the number of HSCs in bone marrow (18). This effect appeared to be mediated by Jagged-1-Notch-1 signaling because Jagged-1 levels were increased in mice with OB-targeted activation of the PTH/PTHrP receptor. In addition, the increase in the number of HSCs in cultures of cells from trans-

Hormone and Bone Health Estrogen therapy for the treatment of osteoporosis in women had been the standard of care until the release in 2002 of the ﬁndings of the Women’s Health Initiative. This study, which was among the largest ever funded by the United States National Institutes of Health, demonstrated for the ﬁrst time that oral estrogen replacement therapy prevented the development of osteoporotic fractures in postmenopausal women. However, the study also showed that this therapy, both with and without progesterone, was associated with an increased risk of thromboembolic disease and strokes. In addition, it was reported that women who took the combination of oral estrogen plus progesterone had an increased risk of breast cancer and myocardial infarctions. Although estrogen therapy is effective for the prevention of postmenopausal osteoporosis, the United States Food and Drug Administration has recommended that it should only be considered for the treatment of women at signiﬁcant risk of osteoporosis who cannot take nonestrogen medications, and it should be used at the lowest doses and for the shortest duration possible to reach treatment goals.

genic mice with OB-targeted activation of the PTH/PTHrP receptor was abrogated by inhibitors of Notch signaling. In a converse experiment, researchers found that targeted destruction of osteoblastic cells in mice led to a decrease in HSCs in bone marrow (19). Interactions of HSCs and OBs are also mediated by interactions of Tie2 on HSCs and Angiopoietin-1 on OBs. This signaling system inhibits cell division in HSCs while maintaining their capacity for self-renewal (20).

Bone as a Reservoir for Memory B and T Cells The bone marrow has long been recognized as the site of early lymphocyte development, but more recent ﬁndings indicate that endstage and memory lymphocytes preferentially inhabit the bone marrow. Upon completing differentiation in germinal centers, antibodyproducing B cells, or plasma cells, downregulate CXCR5 and upregulate CXCR4, www.annualreviews.org • Insights into Osteoimmunology

39

ARI

13 February 2006

12:28

facilitating migration from the spleen to the bone marrow, where stromal cells highly express the CXCR4 ligand, CXCL12 (21–23). Bone marrow stromal cells, which appear to be different from mature OBs, retain plasma cells via expression of adhesion molecules VCAM-1 (vascular cell adhesion molecule 1) and E-selectin (22, 23). BrdU incorporation studies have demonstrated that plasma cells found in the bone marrow have life spans typically in excess of 90 days (21, 24), suggesting the existence in bone marrow of specialized survival niches. Although it has been reported that individual bone marrow plasma cells exhibit no intrinsic survival advantage over splenic plasma cells (24), bone marrow may be capable, unlike the red pulp of the spleen, of providing the necessary space for plasma cells to thrive as a population (25). Given the short half-life of bone marrow plasma cells when cultured in vitro, the bone marrow microenvironment may contain a combination of factors, including IL-5, IL-6, SDF-1α, TNF-α, and CD44 ligands, that are required to sustain resident bone marrow plasma cells (26). Of considerable signiﬁcance is the role of bone as a source of the B cell survival factor BAFF/BLYS (B cell– activating factor/B lymphocyte stimulator). A BAFF/BLYS receptor, BCMA (B cell maturation factor), has recently been shown to be critical for long-term plasma cell survival (27, 28). Recent ﬁndings in multiple myeloma patients show that in addition to neutrophils and monocytes, OCs serve as a greater source of BAFF/BLYS than do bone marrow stromal cells (29). Although there is little evidence that the bone marrow microenvironment is required for optimal function of plasma cells, additional research is required to clearly determine the anatomical advantage for plasma cells of relocating from secondary lymphoid organs to this site. T lymphocyte precursor cells leave the bone marrow and relocate to the thymus for further differentiation. Mature T cells function in the secondary lymphoid organs and at sites of infection. Investigators have re-

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

40

Walsh et al.

cently shown, however, that in the absence of secondary lymphoid tissues the bone marrow can support productive cytotoxic T cell and memory cell generation (30). Although the importance of bone marrow as a secondary lymphoid organ during normal immune responses requires further study, it has potential relevance to host defense against bacterial infections in bone (e.g., osteomyelitis), where OBs may facilitate immune responses by producing a variety of immunomodulatory molecules (31). Memory T cells form after the expansion and contraction of both CD4+ and CD8+ primary response T cells. In normal mice, memory CD8+ T cells preferentially reside and homeostatically proliferate in the bone marrow (32). One homing study indicates that long-lived central memory CD8+ T cells have a special afﬁnity for the bone marrow, employing VCAM-1 for sticking and L-selectin for rolling within the bone marrow microvessels (33). The advantage of central memory CD8+ T cells relocating to the bone marrow may involve the presence of the crucial survival factor IL-15 (30, 33).

Regulation of OBs by Immune Cells and Cytokines A variety of cytokines are known to regulate osteoblastic cells. TNF-α inhibits the differentiation of OBs (34). IL-1, TNF-α, and IFN-γ inhibit collagen synthesis in OBs (35– 38). IL-4 and IL-13 suppress prostaglandin synthesis in bone and are reported to be chemoattractants for OBs (39, 40). IL-4 acts as a direct stimulator of proliferation and inhibitor of differentiation in an osteoblastic cell line (41). Similarly, IL-4-overexpressing mice exhibit a decrease in bone formation and differentiated OBs on the bone surface (42). The role of cytokines in OB apoptosis has also been studied. TNF-α is potently proapoptotic for OBs (43). Activated T lymphocytes also produce products that drive differentiation of human bone marrow stromal cells toward an osteoblastic phenotype (44). B7-H3 is an Ig

ANRV270-IY24-02

ARI

13 February 2006

12:28

superfamily member that is expressed on the surface of antigen-presenting cells (APCs). Recently, B7-H3 was found to be expressed on developing OBs, with its expression increasing during cell maturation (45). Furthermore, B7-H3-knockout mice have decreased cortical bone mineral density compared with littermate controls (45).

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

INTERPLAY BETWEEN OCs AND THE IMMUNE SYSTEM The ﬁrst observation that immune cells could inﬂuence the activity of OCs came from the ﬁnding that supernatant from normal human phytohemagglutinin-stimulated peripheral blood monocytes contained factors that stimulated bone resorption (46). This activity was named OC-activating factor (OAF). When it was eventually puriﬁed and sequenced, the principal stimulator of bone resorption in these crude OAF preparations was found to be the cytokine IL-1 (47). Subsequently, a long list of cytokines have been identiﬁed to have effects on bone. Stimulators of bone resorption include IL-1, TNF-α, IL-6, IL-11, IL-15, and IL-17. Inhibitors of resorption include IL-4, IL-10, IL-13, IL-18, GM-CSF, and IFN-γ. TGF (transforming growth factor)-β and prostaglandins can have either stimulatory or inhibitory effects on resorption, depending on the conditions under which these factors are examined (48). Production of cytokines by immune cells has been linked to human diseases that involve bone. Perhaps the most extensive studies have been on the role of cytokines in the development of osteolytic lesions observed in RA and other inﬂammatory bone diseases, including periodontal disease. Although the immunologic perspective toward RA is typically limited to how RA is initiated (i.e., by failure of immunological tolerance) and its resultant synovitis (the conventional deﬁnition of disease culmination), a hallmark of RA is the rapid erosion of periarticular bone, which is often followed by general secondary osteoporosis (or osteopenia). OCs are found to be

prevalent at the site of focal erosion and are critical for bone erosion (49, 50). In addition to inﬂammatory bone diseases, altered immune responses or cytokine production may cause other osteolytic diseases. Estrogen withdrawal after menopause is associated with a rapid and sustained increase in the rate at which bone is lost. This phenomenon seems to result from an increase in bone resorption that is not met by an equivalent increase in bone formation. Interestingly, activated T cells may cause rapid bone loss under conditions of estrogen deﬁciency by enhancing TNF-α production (51). In a series of experiments involving ovariectomy (OVX)-induced bone loss in mice, an animal model for postmenopausal bone disease, it was reported that nude mice did not lose bone mass after OVX, suggesting that T cells are critical for this response (51). However, similar experiments using nude rats (52), RAG2- or TCR-α-deﬁcient mice, and SCID rats demonstrated OVX-induced trabecular bone loss that was equivalent with that seen in wild-type mice (Y. Choi, Y. Kadono, and J. Lorenzo, unpublished data). Curiously, loss of cortical bone upon OVX was different between T cell–deﬁcient and wild-type models and depended on the bone that was examined. These results suggest there may be compartmental and bone-speciﬁc effects of T cell depletion on OVX-induced bone loss. Additional experiments are required to determine how T cells are involved in this response of bone. These studies will likely require the use of mutant mouse models deﬁcient in speciﬁc immunoregulatory molecules to mechanistically examine the causes of OVX-induced bone loss. Special consideration must be given to the genetic background of any mutant mice examined as part of future studies. The reduction in bone density observed as a result of OVX is roughly only 30%–50%, whereas the difference in average bone density observed between 129/J and C57BL/6 mice is signiﬁcant, with 129/J exhibiting ∼40% higher bone density than C57BL/6. Together, 129/J and www.annualreviews.org • Insights into Osteoimmunology

41

ANRV270-IY24-02

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Osteoprotegerin (OPG): the decoy receptor for TRANCE found in soluble form Osteopetrosis: increased bone density resulting from an imbalance between the formation and breakdown of bone

12:28

C57BL/6 mice constitute most of the knockout mice analyzed. The role of cytokines in malignancyrelated bone disease has also been studied extensively (53). Hematological malignancies such as lymphoma, multiple myeloma, and adult T cell leukemia are associated with increased OC formation and activity, possibly through dysregulation of various cytokines, including IL-1, TNF-α, and IL-6. Unlike bone disease associated with solid tumors, which is typically mediated by PTHrP, hematological malignancies are often characterized by an uncoupling of resorption from formation and the frequent development of purely lytic bone lesions. It should be noted, however, that most cytokines believed to play a role in regulating bone cells are produced by nonimmune cells, like ﬁbroblasts, as well as by immune cells, and exert pleiotropic effects on various cell types. Despite extensive research over the past two decades on this topic, the molecular and cellular signiﬁcance of these cytokines in vivo, speciﬁcally in the context of an immune response, has only recently begun to be elucidated with the identiﬁcation of the TRANCE-RANK-OPG axis (54–60).

THE TRANCE-RANK-OPG AXIS AND OSTEOIMMUNOLOGY Characterization of the functions of TRANCE and its receptors [RANK and osteoprotegerin (OPG)] have contributed signiﬁcantly to the emergence of osteoimmunology, speciﬁcally with respect to examination of the interplay between active immunity and maintenance of bone homeostasis (58–60). Because there are a number of recent reviews on the diverse physiologic function of the TRANCERANK-OPG axis (58, 59), we focus here on its role in the context of osteoimmunology (Figure 4).

42

Walsh et al.

Regulation of Bone Homeostasis by the TRANCE-RANK-OPG Axis Bone homeostasis is maintained via functional balance of two cell types: OBs, which build bone, and OCs, which resorb bone (14, 60). In an ongoing cycle, OCs remove bone, and subsequently OBs ﬁll the cavity with new bone. This balance enables the continuous remodeling of the bone matrix necessary to maintain skeletal strength and a reservoir for hematopoiesis. Much recent effort has gone into determining the developmental processes underlying OC differentiation and activation. OPG was initially identiﬁed as a soluble decoy-like factor, capable of inhibiting osteoclastogenesis in vitro (61, 62) and inducing osteopetrosis when transgenically overexpressed in mice (62). Furthermore, OPGdeﬁcient mice were described as osteoporotic and found to have an excess of OCs (63). Interestingly, these mice are also susceptible to arterial calciﬁcation, highlighting a potential genetic link between osteoporosis and vascular calciﬁcation (63). The gene identiﬁed as encoding the ligand for OPG was determined to be identical to the gene originally characterized as encoding the activated T cell factor TRANCE (54–56). TRANCE is capable of inducing OC differentiation, maturation, and activation in vitro and, importantly, can do so in the absence of bone marrow stromal cells (55– 57). Many well-known osteotropic factors, including IL-1, IL-6, and IL-11, are now believed to exert most of their osteoclastogenic activity by inducing TRANCE expression on OBs (60, 64). Not surprisingly, TRANCEdeﬁcient mice are severely osteopetrotic owing to a cell nonautonomous defect in OC development (65, 66). These mice also exhibit failed tooth eruption, a common defect associated with developmental osteopetrosis, and diversion of hematopoiesis to the spleen and liver because a functional bone marrow cavity fails to form in the absence of OCs (65, 66). RANK, the signaling receptor for TRANCE, was initially identiﬁed through

ARI

13 February 2006

12:28

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

Figure 4 A model for the cellular interactions linking T cell immunity and bone homeostasis. Pathogenic stimuli or self antigens are phagocytosed and presented to naive T cells by DCs. T cells provide activating signals to DCs through CD40L and in return receive optimal activating and costimulatory signals through MHC:TCR and B7:CD28 interactions, respectively. The activated T cells are induced to express TRANCE, which provides further activating and survival signals to the DCs. The DCs may negatively regulate TRANCE:RANK signaling through upregulation of the TRANCE decoy receptor, OPG. Inﬂammatory cytokines (IL-1, TNF-α) produced during successful T cell immune responses, as well as calciotropic factors (PGE2 or VitD3), induce TRANCE expression by OBs, which cooperate with effector T cells to induce OC differentiation via provision of TRANCE to OC precursors. TRANCE signaling in mature OCs induces bone-resorbing function. OBs block TRANCE binding through secretion of OPG, whereas IFN-γ and IL-4 produced by effector T cells inhibit RANK signaling. Without proper regulation, excessive bone resorption leads to osteoporosis, arthritic joint erosion, and periodontal tooth loss. www.annualreviews.org • Insights into Osteoimmunology

43

ANRV270-IY24-02

ARI

13 February 2006

TRAF: TNF receptor–associated factor

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TRAF6: a TRAF family adapter essential for RANK signaling as well as IL-1R/TLR family signaling

44

12:28

a large-scale analysis of genes expressed in DCs (67). RANK expression at the RNA level is detected in most cell types or tissues examined (67). RANK-deﬁcient mice were demonstrated to phenocopy the defect in OC development observed in the TRANCEknockout mouse, conﬁrming the exclusive speciﬁcity of TRANCE for OC-expressed RANK (68). In humans, gain-of-function mutations in RANK are associated with familial expansile osteolysis and with expansible skeletal hyperphosphatasia, which is characterized by increased OC and OB activity, resulting in fragile bones, pain, and deformity (69–73). Efforts aimed at elucidating the signaling mechanisms involved in TRANCEmediated osteoclastogenesis have been informative (58, 60, 64). RANK signal transduction is mediated by adapter proteins called TNF receptor–associated factors (TRAFs) (74–76). Of the six known TRAFs, RANK interacts with TRAFs 1, 2, 3, and 5 in a membranedistal region of the cytoplasmic tail and with TRAF6 at a distinct membrane-proximal ProX-Glu-X-X-(aromatic/acid residue) binding motif (74–77). Genetic experiments show that TRAF6-deﬁcient mice have severe osteopetrosis, implying that the key signals sent through RANK in OC precursors are mediated by the adapter molecule TRAF6 (78, 79; Y. Choi, N. Kim, and Y. Kadono, unpublished data). Downstream of TRAF6, TRANCE signaling in OCs activates PI3K, TAK1, c-Src, JNK1, p44/42 ERK, p38 MAPK, Akt/PKB, and mTOR, and subsequently a series of transcription factors including NF-κB, c-Fos, Fra-1, and NFATc1. This aspect of TRANCE signaling has been recently reviewed elsewhere (58–60, 64, 80). In addition to the signaling pathways mentioned above, TRANCE stimulation also triggers reactive oxygen species (ROS) production (81). ROS, such as superoxide anions, hydroxyl radicals, and H2 O2 , have been associated with many cellular responses, including metabolic bone diseases found in aged osteoporotic women (82). Walsh et al.

Recent reports suggest that ROS act as a key second messenger during osteoclastogenesis (81), such that TRANCE stimulation induces the production of ROS in OC precursors via the small GTPase Rac1 and the ROSinducing factor NADPH oxidase (Nox) 1. How ROS cross-regulates the signaling pathways necessary for OC differentiation is unclear, although one interesting hypothesis is that ROS may potentiate mitogen-activated protein kinase (MAPK) activation by inactivating protein tyrosine phosphatase activity in a manner similar to mechanisms recently described in B cells (83).

Modulation of Immunity by the TRANCE-RANK-OPG Axis The signiﬁcance of TRANCE-RANK-OPG signaling in regulating the immune system continues to emerge. Initial studies of TRANCE- and RANK-deﬁcient mice demonstrated the importance of these signals for secondary lymphoid organ development, as these animals display a lack of peripheral lymph nodes and abnormalities in B cell follicle formation and marginal zone integrity in the spleen (58, 59). In this section, however, we focus on the role TRANCE-RANK plays in shaping the immune response in the adult immune system. To date, most reported data indicate that TRANCE modulates immunity through DCs (Figure 4). DCs are the most potent professional APCs and are required to initiate T cell–mediated immunity in vivo (84). DCs differentiate from the hematopoietic monocyte/macrophage progenitor cell lineage and, as close relatives of OCs, can be generated in vitro by treating a common precursor cell with GM-CSF. GM-CSF suppresses c-Fos and Fra-1 (85, 86), which are key transcription factors for OC differentiation. These results highlight a mechanism of developmental divergence between these two cell types. Upon receipt of inﬂammatory or activating stimuli, DCs home to the T cell areas of the lymph nodes to activate antigen-speciﬁc

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

T cells. Productive activation relies on numerous DC-speciﬁc factors, including alteration of the chemokine receptor repertoire, upregulation of costimulatory molecules, and cytokine production. These modiﬁcations are induced by exogenous inﬂammatory stimuli as well as by signals transmitted by TNF family members, TNF-α and CD40L. TRANCE signaling has also been implicated in DC function, particularly with regard to regulation of DC survival. Activated DCs are relatively short-lived cells, with a halflife as low as 1–2 days upon arrival in the lymph node (87), yet in situ imaging studies suggest that individual T-DC couplings may last 37 h or longer (88–90). TRANCEprolonged DC survival is attributed to upregulation of the antiapoptotic protein Bcl-xL (91) through a pathway requiring the NF-κB components p50 and c-Rel (92). Treatment of DCs with TRANCE also activates the antiapoptotic serine/threonine kinase, Akt/PKB, through recruitment of PI3-K by TRAF6 and Cbl-b to RANK, in a mechanism dependent on the kinase activity of c-Src (75, 93). TRANCE-prolonged DC survival also has in vivo relevance, as pretreatment of peptidepulsed DCs with TRANCE prior to subcutaneous injection into recipient mice results in signiﬁcantly elevated DC persistence in draining lymph nodes and in enhanced Th1 cytokine production and T cell memory formation (94). DC vectors intended for use in immunotherapy persist longer when pretreated with TRANCE (95), and enforced autocrine TRANCE-RANK signaling but not CD40L-CD40 signaling on DCs enhances antitumor immunity (96). Opg−/− DCs potentiate in vitro mixed lymphocyte reaction (MLR), despite CD86, MHC class II, and antigen-presentation levels identical to syngeneic opg+/− DCs (97). Blockade of TRANCE signaling in vivo results in a slightly reduced CD4+ T cell response to LCMV infection, although the response is severely inhibited in the absence of CD40 signaling (98). These experiments highlight the requirement for TNF family

member signaling in the generation of antiviral immunity, as well as the degree to which TRANCE-RANK and CD40L-CD40 function overlap. However, physiologic signaling through RANK is more limited in scope than that through CD40 in that treatment of immature DCs with TRANCE cannot initiate activation, and TRANCE signaling does not complement the cd40−/− defect in germinal center formation and B cell afﬁnity maturation (91, 98). This disconnect is likely not explained by intrinsic signaling differences, as RANK and CD40 activate the same set of signaling cascades, but instead is explained by differential expression patterns and kinetics. For example, on T cells CD40L is rapidly and transiently expressed and is limited only to the CD4+ subset (99). In contrast, TRANCE is expressed on both CD4+ and CD8+ T cells (100) and is capable of binding both its functional (RANK) and decoy (OPG) receptors. These interactions are also likely to succeed CD40L-CD40 signaling, as CD40L is a key inducer of RANK and OPG expression by DCs (101). The physiologic role of CD40L-CD40 versus TRANCE-RANK signaling in DC function may, therefore, depend on the phase of the immune response. CD40L-CD40 signaling may be more prominent during the initiation and effector phases, when many cellular components of the immune system are strongly activated. By contrast, TRANCE-RANK signaling may be more important during the waning phases to ensure that T memory formation is established and then to wind down remaining T-DC interactions, possibly through OPG interference with TRANCE signaling. The severe phenotype of TRANCE- and RANKknockout mice has thus far not allowed a thorough examination of the role of TRANCE in memory cell formation. Evidence also suggests that TRANCE may be important for survival of interstitial DCs engaged in antigen surveillance during the interim period separating immune responses. Human CD34+ immature DCs express both TRANCE and RANK and can therefore www.annualreviews.org • Insights into Osteoimmunology

45

ARI

13 February 2006

12:28

provide an autocrine survival signal. Peripheral maturation of these DCs leads to a downregulation of TRANCE, suggesting a requirement for an independent source of TRANCE to validate DC activation (102). TRANCE may also be involved in actively inducing tolerance. TRANCE signaling has been directly implicated in the induction of oral tolerance in mice. Feeding low-dose ovalbumin to mice concomitant with intravenous TRANCE treatment results in T cells refractory to rechallenge and correlates with in vitro production of the suppressive cytokine IL-10 by mucosal DCs (103). Another study has demonstrated that TRANCEmediated signaling is required to prevent the onset of autoimmune disease in a TNF-αinducible mouse model of diabetes and that blockade of TRANCE-RANK interactions parallel a diminution of CD4+ CD25+ regulatory lymphocytes, which are necessary to prevent CTL-mediated islet cell destruction (104). In a recent study of murine cardiac allograft tolerance, TRANCE-RANK signaling was shown to be important for the generation of regulatory T cells via intratracheal delivery of alloantigen (105). It remains unclear, though, whether TRANCE directly triggers T lymphocyte suppression or, alternatively, acts through DC intermediaries. TRANCE is also induced preferentially, among key costimulatory molecules, on T cells activated by tolerogenic DCs (106). Further study of this issue should yield insights into the generation and maintenance of T lymphocyte tolerance. In addition to regulation of DCs, TRANCE might inﬂuence B cell development. In OPG-deﬁcient mice there is an expansion of pro-B cells in the bone marrow, whereas the opposite has been observed in TRANCE- or RANK-deﬁcient mice (65, 68, 97, 107), suggesting that TRANCE-RANK interaction might be involved in the proliferation of pro-B cells. Of interest, pro-B cells also expand in the bone marrow of ovariectomized mice (108), in which TRANCE expression on OBs is thought to be increased. Future studies are required to elucidate the

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

46

Walsh et al.

molecular mechanisms of how TRANCE might regulate the fate of pro-B cells and what the immunological consequences of this regulation might be.

Autoimmunity, Bone, and TRANCE: The Birth of Osteoimmunology TRANCE expression on T lymphocytes is induced upon T cell receptor engagement and depends on Ca2+ mobilization (54, 109). Initial experiments demonstrated that activated T lymphocytes, or even supernatants from activated T lymphocyte cultures, could support osteoclastogenesis in vitro (110). It was subsequently observed that mice lacking CTLA4, in which T lymphocytes are systemically activated, exhibit an osteoporotic phenotype associated with increased OC numbers. Transfer of ctla-4−/− T lymphocytes into rag2−/− mice leads to decreased bone density over time, which can be prevented by OPG treatment. This ﬁnding indicated that activated T cells can disrupt bone homeostasis by modulating TRANCE expression (110), although whether T cell–derived TRANCE per se is responsible for aberrant bone metabolism is unclear. In a complementary study, transgenic overexpression of TRANCE restricted to T or T/B lymphocytes was sufﬁcient to partially correct the osteopetrotic phenotype observed in TRANCE-deﬁcient mice (66; Y. Choi & N. Kim, unpublished data). Together, these data deﬁnitively showed the ability of lymphocytes, by expression of TRANCE, to regulate bone homeostasis in vivo and conﬁrmed a bona ﬁde interplay between the adaptive immune system and bone metabolism, giving birth to the ﬁeld of osteoimmunology (Figure 4). In human arthritis, inﬂammation of the synovial joints is accompanied by bone and cartilage destruction. Various animal models have been established for the study of arthritis, and the role of TRANCE in their pathogenesis has been investigated. Treating adjuvantinduced arthritis in Lewis rats with OPG had no discernible effect on inﬂammation but

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

prevented bone loss and cartilage destruction (110). These experiments could not resolve whether preservation of cartilage was an indirect beneﬁt of inhibiting bone erosion or was due to independent mechanisms. A subsequent study demonstrated that bone loss and cartilage destruction were independent in an arthritis model induced by transfer of serum from K/BxN transgenic mice, where T cell activity is not required for onset of disease (111). When K/BxN serum was transferred into TRANCE-deﬁcient mice, inﬂammation and cartilage destruction were comparable to control recipients, but bone erosion was greatly reduced (111). These ﬁndings reinforced the notion that TRANCE per se mediates induction of bone destruction by OCs in animal models of autoimmune arthritis. Examination of the cellular constituents of synovial ﬂuid collected from human arthritis patients revealed that all local T lymphocytes expressed TRANCE, establishing the clinical relevance of the connection between arthritis and immunologically derived TRANCE (110). Recently, investigators have demonstrated that TRANCE, in combination with M-CSF (macrophage colony stimulating factor), can induce transdifferentiation of immature DCs to the OC lineage and that this process is signiﬁcantly enhanced by RA synovial ﬂuid, potentially identifying another mechanism for disease-related bone destruction (112). Periodontitis, induced by infection with various subgingival bacteria, is a major cause of tooth loss and is associated with increased risk for heart failure and stroke (113, 114). To examine the etiology of the disease, peripheral blood lymphocytes (PBL) from patients with localized juvenile periodontitis (LJP) were transferred into rag2−/− mice, which were then orally inoculated with the Gramnegative bacterium Actinobacillus actinomycetemcomitans (114). LJP was recapitulated in the recipient animals and was accompanied by accumulation of OCs at the alveolar sockets (114). Treatment with OPG inhibited the OC inﬁltration and bone damage (114).

In vitro stimulation of PBL showed that TRANCE was induced on CD4+ T lymphocytes activated with A. actinomycetemcomitans antigens and that disease was attenuated when the same cells were speciﬁcally depleted from recipient mice (114). This study demonstrated the importance of CD4+ T lymphocytes in the pathogenesis of periodontitis, specifically with regard to disease-related bone destruction. Bone loss has long been recognized as an extraintestinal complication of inﬂammatory disorders of the gut, such as Crohn’s disease and ulcerative colitis (115). One recent study found that patients with these diseases have elevated levels of serum OPG, which derive from the site of inﬂammation and inversely correlate with severity of bone loss (116), whereas another study found that Crohn’s disease patients have elevated levels of both OPG and soluble TRANCE (117). Mechanistic insight into this link is provided by a study demonstrating that OPG treatment of mice suffering from IL-2-deﬁciency-induced ulcerative colitis results not only in reduced osteopenia, but also in mitigation of colitis owing to reduced colonic DC survival (118). In addition to arthritis, periodontal disease, and inﬂammatory bowel disease, pathologic bone loss is observed in patients suffering from other autoimmune diseases (diabetes mellitus and lupus erythematosus), chronic viral infections (HIV), allergic diseases (asthma), and metastatic breast and lung cancers (49, 50, 119). The contribution to pathogeneses by osteoimmunologic factors merits further investigation and may provide viable therapeutic options for alleviating painful sequelae associated with a variety of conditions. Although autoimmunity is, in some cases, associated with bone loss, each productive T cell response does not have such a deleterious outcome. T cells also secrete cytokines, such as IFN-γ, IL-4, and TGF-β, that inhibit the pro-osteoclastogenic effects of TRANCE (48–50) (Figure 4). The role of the Th1 cytokine IFN-γ, in particular, appears to be www.annualreviews.org • Insights into Osteoimmunology

47

ANRV270-IY24-02

ARI

13 February 2006

12:28

crucial in preventing T lymphocyte–mediated osteoclastogenesis (120). TGF-β is characterized as both an osteotropic and immunosuppressive cytokine. Although the largest repository of latent TGF-β is in bone, its role in OC formation is complex and insufﬁciently understood (14). TGF-β downmodulates TRANCE expression in OBs, thereby negatively impacting their ability to mediate osteoclastogenesis in culture (121). However, TGF-β has also been shown to potentiate TRANCE expression in activated T lymphocytes (109) and enhance osteoclastogenesis in cultures supplemented with soluble TRANCE (121). Additional studies are necessary to determine whether TGF-β uses multiple regulatory mechanisms, and if so, what disparate purposes they might serve. Given the variety of T lymphocyte–associated cytokines with osteotropic function, it will also be useful to clarify the correlation between Th1/Th2 cytokine polarization and any attendant osteoimmunologic bone destruction.

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TLR: Toll-like receptor

TOLL-LIKE RECEPTORS, INFLAMMATION, AND BONE METABOLISM Toll-like receptors (TLRs) are members of an ancient receptor family that shares homology with IL-1R and are critical activators of the innate immune response (122). They are most highly expressed on APCs such as DCs, macrophages, and B cells, but some members are expressed on a diverse array of tissues. Ligation of these receptors by conserved microbial molecules or endogenous danger factors results in the upregulation of costimulatory molecules and the elaboration of inﬂammatory cytokines in preparation for an adaptive immune response. TLR signaling is mediated by the adaptors MyD88, TRAF6, and TRIF, which activate various downstream signaling pathways, including IKK-NF-κB, MAPK, and IRF (122). Because macrophages and DCs share a common progenitor with OCs, it is not surprising that TLR expression is also detected 48

Walsh et al.

on bone cells (123–125). Direct signaling of various TLRs (including TLR4) on OC precursors inhibits TRANCE-mediated osteoclastogenesis (123). The data that microbial products inhibit OC differentiation via TLRs are counterintuitive because bacterial infection can cause inﬂammatory bone diseases such as periodontitis, osteomyelitis, and bacterial arthritis (126). Bone mineral density is reduced in such diseases because of excessive bone resorption by OCs. In addition, lipopolysaccharide may be a potent stimulator of bone loss by causing an increase in the number of OCs in mice. Moreover, TLR activation can enhance OB-mediated OC differentiation by inducing TRANCE and TNF-α on OBs (124, 125, 127). The basis for the apparent discrepancy between TLR stimulation as a potent negative regulator of osteoclastogenesis and the association of bacterial infection with excessive bone resorption by OCs remains unclear. As described earlier, alveolar bone destruction in periodontitis caused by infection of Gramnegative bacteria is mediated by enhanced osteoclastogenesis owing to T cell responses and subsequent upregulation of TRANCE (114). In the same study, bacterial infection of immunodeﬁcient (SCID) mice did not lead to signiﬁcant levels of alveolar bone loss, suggesting that bacterial products do not have a direct role in osteoclastogenesis because SCID mice have no known defect in OC precursors or OBs (114). Therefore, bone loss associated with bacterial infection is likely an indirect outcome of exacerbated T cell responses. Similar to macrophages or DCs, OC precursors also produce proinﬂammatory cytokines, such as TNF-α, in response to various TLR ligands (123). Moreover, although TLR stimulation inhibits OC differentiation, OC precursors treated with TLR ligands still retain high levels of phagocytic activity, which is a major host-defense mechanism for the clearance of bacterial infection. Therefore, the net outcome of TLR stimulation in OC precursors is likely the enhancement of

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

immune responses toward achieving bacterial clearance. This enhancement of immune responses can be achieved by promoting cytokine production from precursor cells and by inhibiting their differentiation into nonphagocytic, nonimmune cells, such as mature OCs. Thus, interaction of these microbial products with TLRs on OC precursors appears to favor the role of OC precursors as part of the proinﬂammatory system by inhibiting their differentiation into mature OCs and by promoting the production of inﬂammatory cytokines. However, because these cells can differentiate into mature OCs if TLR ligands are removed (123), after a microbial infection is cleared the presence of residual, activated T cells can lead to the differentiation of phagocytic precursors into mature, boneresorbing OCs. In addition, TNF-α produced by OC precursors upon TLR stimulation can enhance osteoclastic bone resorption. TLRs are thus likely to regulate the balance of immune responses and bone metabolism during acute attacks of vertebrate hosts by various microbes. However, physiologic in vivo stimulation of TLRs, which are expressed on various cells, may result in different effects on bone metabolism depending on the nature of the given immune responses. In addition, ongoing stimulation of TLRs by commensal bacteria might affect bone metabolism. In support of this idea, recent data show that mice deﬁcient in mediators of the TLR/IL-1R signaling pathway (MyD88 or IRAK-M) exhibit an altered bone metabolism, although it is not clear whether the defects are due to the signals from TLRs or IL-1R (128, 129).

AN OSTEOIMMUNOLOGICAL SIGNALING NETWORK AND GENE REGULATORY MECHANISM In addition to interplay between cellular constituents of the immune system and bone, there are also numerous parallels between the signaling networks used by the cells of each

system. There are cases both of common, shared pathways and of analogous signaling mechanisms, which activate specialized gene targets via system-speciﬁc mediators. In this section, we touch brieﬂy on the better characterized shared pathways (58–60, 64, 80) and focus on some of the novel osteoimmunologic signaling mechanisms that have recently been identiﬁed (Figure 5).

Costimulation The formation and activation of OCs are processes tightly regulated by OBs, which provide at least two known essential factors for osteoclastogenesis, TRANCE and M-CSF. In addition, stromal cells produce various osteotropic factors that inﬂuence OB-induced osteoclastogenesis of bone marrow precursors. These factors can be divided into two groups: those that inﬂuence activity of OBs (e.g., TNF-α that induces TRANCE expression in OBs), and those that affect the OC precursors or OCs per se. A series of experiments showed that M-CSF and TRANCE together appear to be sufﬁcient to induce the differentiation of bone marrow precursors, spleen cells, or blood monocytes to become mature OCs in vitro. However, the expression of M-CSF, TRANCE, and their receptors is not limited to bone cells. For example, M-CSF and TRANCE are important cytokines for the activity/viability of macrophages and DCs. Despite this pleiotropy, OCs are not found in soft tissues, raising the question of why the same set of signaling receptors leads to different functional outcomes in different anatomical environments. One possibility is the existence of costimulatory molecule(s) present only in bone. Alternatively, there could be a powerful inhibitor of osteoclastogenesis in soft tissues that is not found in bone. To address this question, we proposed the hypothesis a few years ago that there exists a mechanism in preosteoclasts analogous to the costimulation requirement for T cell activation (130). Hence, our hypothesis proposed that OC differentiation is controlled www.annualreviews.org • Insights into Osteoimmunology

49

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

12:28

Figure 5 A model for a signaling network necessary for TRANCE-induced OC differentiation and its analog in T cell activation. Factors shaded yellowish have been shown experimentally to be important for OC differentiation. Those shaded green are molecules potentially involved in OC differentiation based on their analogous roles in the TCR signaling network. Src family kinases (SFK) necessary for phosphorylation of ITAMs are not identiﬁed. Positive feedback circuits are marked as (+) and shown in red. TREM is constitutively expressed on pOCs (and monocytes), while OSCAR (OC-associated receptor) expression is further upregulated by TRANCE stimulation. The putative ligand for TREM is expressed on pOCs, whereas OSCAR-L is expressed on OBs. The connection to integrin signaling and the cytoskeleton is proposed based on the analogy to T cell signaling requirements. The term BAAF [bone and age associated factor(s)] is proposed to describe putative costimulatory factors for OCs speciﬁc to bone. For simplicity, negative regulators of the signaling network are not illustrated but are described in the text.

not only by two essential factors, M-CSF and TRANCE (analogous to MHC/antigen complexes interacting with TCR/CD4 or TCR/CD8), but also by other nonessential but critical costimulatory molecules (analogous to B7 family proteins interacting with CD28) (131). Because the in vivo concen50

Walsh et al.

trations of M-CSF and TRANCE produced by OBs in response to bone-resorbing hormones is likely to be much lower than is provided in in vitro experiments, costimulatory molecules are likely to inﬂuence physiologic differentiation of OCs in a manner analogous to T cell activation, whereby signals from the

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

costimulatory receptor CD28 complement requisite signals from the TCR complex (130, 131). In addition, as with T cells, the requirement for a particular set of costimulatory factors/receptors for OCs should vary depending on the microenvironment. Cells expressing ligands for costimulatory receptors expressed on OCs also vary, but those interacting with OCs themselves, such as OBs, most often provide costimulation (analogous to DC provision of B7 family proteins or TNF family proteins like 4–1BBL to T cells) (130). The signals resulting from the interaction of costimulatory factors and their receptors on OC precursors determine the efﬁcacy of the signals from the essential osteoclastogenic receptor, RANK (similar to TCR/CD4 or TCR/CD8 for T cells), and the sum of the two will determine the quality of OC differentiation and activation. In support of our costimulation hypothesis, we have identiﬁed a novel cell surface receptor, OSCAR, which is preferentially expressed on OCs, and we have shown that in addition to normal TRANCE-RANK signaling, interaction of OSCAR with its putative ligand (OSCAR-L) is important for OB-induced OC differentiation (131). Moreover, OSCAR-L expression is most prevalent on osteoblastic cells (131). Therefore, the OSCAR receptor/ligand pairing could be characterized as a putative costimulation receptor/factor for efﬁcient OC differentiation, and may provide bone-speciﬁc costimulation required for the differentiation of OCs in conjunction with the essential factors M-CSF and TRANCE. This signaling combination may provide a mechanistic explanation of why OCs are found only on bone surfaces in vivo (Figure 5). Although the nature of bone-speciﬁc costimulatory molecules, such as OSCAR-L, requires further study, a series of recent experiments have supported our costimulation hypothesis (132, 133). For OC development in vivo, some surface receptors on OC precursors, such as PIR-A, OSCAR, TREM2, and SIRPβ1, associate with ITAM-containing molecules, DAP12 and FcRγ, and provide

necessary costimulation and activation of Ca2+ signaling (132, 133). Hence, although a single deﬁciency for either DAP12 or FcRγ results in only minor OC defects, double deﬁciency results in severe osteopetrosis (132, 133). Additional analysis of mutant mice suggests that these receptors activate calcineurin via Syk and PLCγ (132–134). More signaling proteins have been identiﬁed in lymphocytes that bridge Syk (or ZAP-70) and PLCγ, and lead to Ca2+ activation (135, 136). Indeed, Gab2 has recently been shown to be critical for generation of functional OCs (137). It will not be surprising if additional molecules (or family members) crucial to T cell signaling are identiﬁed as playing an equivalent role in OC differentiation (Figure 5). However, it is important to point out that osteopetrosis in DAP12/FcRγ doubledeﬁcient mice is much less severe than that in TRANCE- or RANK-knockout mice, and that, in contrast to TRANCE- or RANKknockout mice, these animals exhibit signiﬁcant numbers of OCs. (132, 133). This is consistent with our hypothesis that costimulatory receptors for OC differentiation are not essential and that multiple redundancies probably exist (131). Sustained Ca2+ mobilization is necessary for OC differentiation because NFATc1 activation is absolutely required for the process (138). The NFAT family of transcription factors was originally identiﬁed as a set of regulators of gene transcription in activated T cells (139). Recently, researchers found that RANK signaling induces expression of the NFAT family member NFATc1 (NFAT2) and that this factor is critical for OC development, as NFATc1-deﬁcient precursor cells exhibit an absolute failure to differentiate into OCs (138). Like other NFAT family members, the induction and activation of NFATc1 rely on the calcium-regulated phosphatase, calcineurin, thereby explaining negative effects of calcineurin inhibitors like FK506 and cyclosporine on osteoclastogenesis. The ability of NFATc1 to regulate its own expression indicates the existence of an www.annualreviews.org • Insights into Osteoimmunology

OSCAR: osteoclast-associated receptor NFAT: nuclear factor of activated T cells

51

ARI

13 February 2006

12:28

autonomic feedback loop, although initial triggering of NFATc1 induction is mediated by TRAF6 and c-fos via TRANCE-RANK stimulation (138). Thus, Ca2+ signaling via costimulatory receptors on pre-OCs is critical for ampliﬁcation of NFATc1 activity to a level sufﬁcient for OC differentiation. Interestingly, NFATc1, in conjunction with MITF and PU.1, transactivates OSCAR expression during TRANCE-induced OC differentiation (Y. Choi & N. Kim, unpublished data). This suggests that there is a positive feedback circuit from TRANCE to NFATc1 via costimulatory receptors, such as OSCAR, during OC differentiation, which ensures a high level of NFATc1 activity (Figure 5). Key to the analogy with lymphocyte costimulation, RANK, like TCR, is still the primary, requisite receptor, the absence of which renders the secondary receptors inconsequential to osteoclastogenesis. One issue that remains to be worked out is a greater understanding of why this system has evolved, and whether there exists a parallel state in OC development that mimics anergy, or induced tolerance, as observed in lymphocytes.

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

Countervailing Osteoclastogenic Signaling In addition to OPG (and factors like TGFβ, which induce TRANCE downregulation), there are other negative regulators of RANK signaling in OCs. Although T cells express TRANCE, there is a negative correlation between T lymphocyte activation and signaling through RANK, apparently due to T cell– derived IFN-γ (120). Signaling through the IFN-γR on OCs or OC precursors leads to rapid proteasomal degradation of TRAF6 and to abortive differentiation and function (120). In this way a productive immune response is prevented from having an overlapping, deleterious effect on bone in the surrounding environment. Another regulatory mechanism involves negative feedback induced by RANK itself. Activation of the fos gene by TRANCE 52

Walsh et al.

leads to upregulation of IFN-β, which mediates a feedback mechanism blocking further c-Fos-dependent activity (140). As such, mice deﬁcient for the IFN-α/β receptor (IFNAR1) suffer from an osteoporotic phenotype characterized by an increase in OCs (140). Promoter characterization showed that TRANCE-mediated upregulation of IFN-β utilizes AP-1-binding sites and that c-Fosdeﬁcient OC precursors are incapable of inducing IFN-β production (140). To facilitate OC development, therefore, OC precursors need to upregulate the cytokine signaling regulator SOCS3 to inhibit IFN-mediated suppression (141–143). Interferons are cytokines critical to inducing productive immune responses against viruses (IFN-α/β) as well as parasites and bacteria (IFN-γ) (144–146). Type I (IFN-α/β) and type II (IFN-γ) interferons signal through different surface receptors and activate distinct DNA recognition sites [interferon-stimulated response elements (ISRE) and gamma interferon activation sites (GAS), respectively], but they use a common mediator of gene transcription, Stat1. In immune cells, IFN-activated Stat1 induces antiviral and inﬂammatory gene transcription, although in OC precursors Stat1-dependent mechanisms mitigate osteoclastogenesis. As IFNs are associated with inﬂammation and active T cell immunity, which is itself associated with enhanced osteoclastogenic factors, it is notable that they are involved in mitigating the effects of T cells on bone erosion. Interestingly, Stat1-deﬁcient mice exhibit increased OB differentiation and bone mass, as Stat1 plays a role in vivo in attenuating the critical OB transcription factor Runx2 (147). It will be interesting to determine whether the bone phenotype in Stat1-deﬁcient mice is consistent under different immunologic conditions, or whether under inﬂammatory conditions a greater role for Stat1-mediated inhibition of osteoclastogenesis emerges. Finally, since ITAM-containing molecules provide critical costimulatory signals for OC differentiation, future studies are required

ANRV270-IY24-02

ARI

13 February 2006

12:28

to determine whether ITIM-containing molecules counteract costimulatory signals for OC differentiation via DAP12/FcRγ.

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Quantitatively Distinctive Utilization of TRAF6 by TRANCE It remains puzzling why TRANCE-RANKTRAF6 signaling is so critical for osteoclastogenesis in vivo. Other immune receptors, including CD40 and IL-1R/TLR, are expressed on OC precursors and use TRAF6 to activate overlapping signaling cascades but do not induce osteoclastogenesis. These observations caused us to question whether qualitative or quantitative differences exist between the TRAF6-mediated signals induced by TRANCE/RANK versus other ligandreceptor pairs. Our recent data show that stimulation via overexpressed wild-type CD40 can induce osteoclastogenesis (148). RANK contains three TRAF6-binding sites on its cytoplasmic tail, whereas CD40 has one. Stimulation through modiﬁed CD40 molecules containing additional TRAF6-binding sites in the cytoplasmic tail showed a dose-dependent increase in osteoclastogenesis. Moreover, precursors overexpressing TRAF6 alone differentiate into OCs in the absence of additional signals from TRANCE. Thus, our results suggest that differences in the osteoclastogenic capacity of RANK versus other TRAF6-associated receptors may stem, in part, from a quantitative difference in TRAF6-mediated signals. Similar mechanisms of controlling distinct biological outcomes by growth/differentiation factors, despite their use in overlapping signaling cascades, have been reported (149, 150). For example, although both epidermal growth factor (EGF) and nerve growth factor (NGF) use the same kinase cascade, only NGF can induce neuronal differentiation. When EGF receptor was overexpressed, it induced neuronal differentiation similar to NGF (149). In the case of OC differentiation, any TRAF6-binding receptor might have the potential, but only RANK

can meet the threshold to induce osteoclastogenesis. Whether OC differentiation occurs through various TRAF6-utilizing receptors also correlates with the induction and persistence of NFATc1 activation. In normal OC differentiation, TRANCE-RANK interaction, in conjunction with costimulation via ITAM-containing molecules, sets the level of NFATc1, determining whether the differentiation process becomes irreversible. However, this concept also suggests that as long as NFATc1 is induced beyond the threshold level, then OC differentiation can be mediated by factors other than TRANCE-RANK. In support of this idea, we have recently found that RANK-deﬁcient splenic cells can become bone-resorbing OCs in vitro when they are stimulated by a cocktail of cytokines (Y. Choi, N. Kim, Y. Kadono, J. Lorenzo, unpublished data). Whether such an outcome is possible in vivo, speciﬁcally under abnormal or pathological conditions, requires further investigation.

CONCLUSIONS The ﬁelds of immunology and bone biology have matured such that key cellular and molecular mechanisms governing the homeostasis of the individual systems are largely understood. However, despite extensive crossregulation between bone metabolism and the immune system, the mechanisms by which one regulates the other, and the biological implications of such interactions, are poorly understood. We believe that this lack of understanding is due in part to the challenges typically associated with crossing disciplinary boundaries that form naturally during the separate evolutions of ﬁelds like modern immunology and bone biology. It is difﬁcult enough for scientists/physicians to keep abreast of advances in multiple ﬁelds, but even more so to develop the knowledge base, skills, and materials necessary to address important issues. Therefore, it will be critical to create an environment conducive to the study of www.annualreviews.org • Insights into Osteoimmunology

53

ARI

13 February 2006

12:28

intersystem crosstalk. Awareness of intersystem crosstalk will no doubt contribute to our understanding of how both bone and the immune system are regulated in a physiologic context, both at the molecular level and at the level of organ systems. Moreover, this endeavor will lead to better treatments for human diseases involving both systems, including various inﬂammatory and metabolic bone diseases, as well as tumor-induced bone lysis. Many of these pathologic processes are major targets for therapeutic intervention and are being pursued in the absence of solid scientiﬁc understanding of the molecular and cellular processes underpinning these interactions. According to the ﬁrst-ever report by

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

the U.S. Surgeon General on bone health, by 2020 one in two Americans over age 50 will be at risk for fractures from osteoporosis or low bone mass. These secondary health concerns become more prominent as people not only live longer but also expect to remain active in old age. Future preventative treatments for chronic bone-related diseases that are often associated with inﬂammation and that impact quality of life will require a high degree of speciﬁcity, especially if tailored for a segment of the population already suffering from, or vulnerable to, other age-related ailments. We believe these issues place osteoimmunology in a position of unique clinical signiﬁcance.

SUMMARY POINTS 1. Bone cells are inﬂuenced by cytokines and cell surface proteins that are expressed on lymphocytes. 2. Bone cells provide key factors necessary for HSC, B cell differentiation and memory, and T cell memory. 3. Identiﬁcation of the TRANCE-RANK-OPG axis clearly established the physiological connection between immune responses and bone metabolism, thus conﬁrming the importance of various osteotropic cytokines for bone metabolism. 4. Identiﬁcation of effector cytokines produced by activated T cells as inhibitors of TRANCE-induced osteoclast differentiation explains why normal T cell responses do not overtly affect normal bone metabolism. 5. Identiﬁcation of costimulatory receptors and their signaling pathways conﬁrms the molecular parallel between T cell activation and osteoclast differentiation.

FUTURE ISSUES TO BE RESOLVED 1. What is the nature of bone marrow stromal cells and what factors are necessary for their interaction with hematopoietic lineage cells? 2. Why are osteoclasts found only on the surface of bone? What are the ligands for costimulatory receptors found on preosteoclasts? 3. To what extent do normal immune responses affect bone homeostasis? And what are the consequences of bone metabolism to adaptive immune responses? 4. How does the interplay between bone and the immune system change with age?

54

Walsh et al.

ANRV270-IY24-02

ARI

13 February 2006

12:28

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LITERATURE CITED 1. Arron JR, Choi Y. 2000. Bone versus immune system. Nature 408:535–36 2. Raisz LG, Kream BE, Lorenzo JA. 2002. Metabolic bone disease. In Williams Textbook of Endocrinology, ed. RH Williams, PR Larsen, HM Kronenberg, KS Polosky, pp. 1373– 410. Philadelphia: WB Saunders 3. Ross MH, Kaye GI, Pawlina W. 2003. Bone. In Histology: A Text and Atlas, ed. MH Ross, GI Kaye, W Pawlina, pp. 180–213. Philadelphia: Lippincott Williams & Wilkins 4. Aubin JE. 2001. Regulation of osteoblast formation and function. Rev. Endocr. Metab. Disord. 2:81–94 5. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. 1997. Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 89:747–54 6. Nakashima K, Zhou X, Kunkel G, Zhang Z, Deng JM, et al. 2002. The novel zinc ﬁngercontaining transcription factor osterix is required for osteoblast differentiation and bone formation. Cell 108:17–29 7. Celeste AJ, Iannazzi JA, Taylor RC, Hewick RM, Rosen V, et al. 1990. Identiﬁcation of transforming growth factor β family members present in bone-inductive protein puriﬁed from bovine bone. Proc. Natl. Acad. Sci. USA 87:9843–47 8. Gong Y, Slee RB, Fukai N, Rawadi G, Roman-Roman S, et al. 2001. LDL receptor-related protein 5 (LRP5) affects bone accrual and eye development. Cell 107:513–23 9. Little RD, Carulli JP, Del Mastro RG, Dupuis J, Osborne M, et al. 2002. A mutation in the LDL receptor-related protein 5 gene results in the autosomal dominant high-bone-mass trait. Am. J. Hum. Genet. 70:11–19 10. Kato M, Patel MS, Levasseur R, Lobov I, Chang BH, et al. 2002. Cbfa1-independent decrease in osteoblast proliferation, osteopenia, and persistent embryonic eye vascularization in mice deﬁcient in Lrp5, a Wnt coreceptor. J. Cell Biol. 157:303–14 11. Boyden LM, Mao J, Belsky J, Mitzner L, Farhi A, et al. 2002. High bone density due to a mutation in LDL-receptor-related protein 5. N. Engl. J. Med. 346:1513–21 12. Bennett CN, Longo KA, Wright WS, Suva LJ, Lane TF, et al. 2005. Regulation of osteoblastogenesis and bone mass by Wnt10b. Proc. Natl. Acad. Sci. USA 102:3324–29 13. Knothe Tate ML, Adamson JR, Tami AE, Bauer TW. 2004. The osteocyte. Int. J. Biochem. Cell Biol. 36:1–8 14. Teitelbaum SL. 2000. Bone resorption by osteoclasts. Science 289:1504–8 15. Howard GA, Bottemiller BL, Turner RT, Rader JI, Baylink DJ. 1981. Parathyroid hormone stimulates bone formation and resorption in organ culture: evidence for a coupling mechanism. Proc. Natl. Acad. Sci. USA 78:3204–8 16. Riggs BL, Melton LJ 3rd. 1986. Involutional osteoporosis. N. Engl. J. Med. 314:1676–86 17. Zhang J, Niu C, Ye L, Huang H, He X, et al. 2003. Identification of the haematopoietic stem cell niche and control of the niche size. Nature 425:836–41 18. Calvi LM, Adams GB, Weibrecht KW, Weber JM, Olson DP, et al. 2003. Osteoblastic cells regulate the haematopoietic stem cell niche. Nature 425:841–46 19. Visnjic D, Kalajzic Z, Rowe DW, Katavic V, Lorenzo J, Aguila HL. 2004. Hematopoiesis is severely altered in mice with an induced osteoblast deﬁciency. Blood 103:3258–64

17, 18, and 20. All three were key papers showing the molecular link between OBs and the HSC niche.

20. Arai F, Hirao A, Ohmura M, Sato H, Matsuoka S, et al. 2004. Tie2/angiopoietin-1 signaling regulates hematopoietic stem cell quiescence in the bone marrow niche. Cell 118:149–61 21. Manz RA, Arce S, Cassese G, Hauser AE, Hiepe F, Radbruch A. 2002. Humoral immunity and long-lived plasma cells. Curr. Opin. Immunol. 14:517–21 www.annualreviews.org • Insights into Osteoimmunology

55

ARI

13 February 2006

12:28

22. Shapiro-Shelef M, Calame K. 2005. Regulation of plasma-cell development. Nat. Rev. Immunol. 5:230–42 23. Tokoyoda K, Egawa T, Sugiyama T, Choi BI, Nagasawa T. 2004. Cellular niches controlling B lymphocyte behavior within bone marrow during development. Immunity 20:707– 18 24. Slifka MK, Antia R, Whitmire JK, Ahmed R. 1998. Humoral immunity due to long-lived plasma cells. Immunity 8:363–72 25. Sze DM, Toellner KM, Garcia de Vinuesa C, Taylor DR, MacLennan IC. 2000. Intrinsic constraint on plasmablast growth and extrinsic limits of plasma cell survival. J. Exp. Med. 192:813–21 26. Cassese G, Arce S, Hauser AE, Lehnert K, Moewes B, et al. 2003. Plasma cell survival is mediated by synergistic effects of cytokines and adhesion-dependent signals. J. Immunol. 171:1684–90 27. O’Connor BP, Gleeson MW, Noelle RJ, Erickson LD. 2003. The rise and fall of longlived humoral immunity: terminal differentiation of plasma cells in health and disease. Immunol. Rev. 194:61–76 28. O’Connor BP, Raman VS, Erickson LD, Cook WJ, Weaver LK, et al. 2004. BCMA is essential for the survival of long-lived bone marrow plasma cells. J. Exp. Med. 199:91–98 29. Moreaux J, Cremer FW, Reme T, Raab M, Mahtouk K, et al. 2005. The level of TACI gene expression in myeloma cells is associated with a signature of microenvironment dependence versus a plasmablastic signature. Blood 106:1021–30 30. Feuerer M, Beckhove P, Garbi N, Mahnke Y, Limmer A, et al. 2003. Bone marrow as a priming site for T-cell responses to blood-borne antigen. Nat. Med. 9:1151–57 31. Marriott I. 2004. Osteoblast responses to bacterial pathogens: a previously unappreciated role for bone-forming cells in host defense and disease progression. Immunol. Res. 30:291– 308 32. Becker TC, Coley SM, Wherry EJ, Ahmed R. 2005. Bone marrow is a preferred site for homeostatic proliferation of memory CD8 T cells. J. Immunol. 174:1269–73 33. Mazo IB, Honczarenko M, Leung H, Cavanagh LL, Bonasio R, et al. 2005. Bone marrow is a major reservoir and site of recruitment for central memory CD8+ T cells. Immunity 22:259–70 34. Gilbert L, He X, Farmer P, Boden S, Kozlowski M, et al. 2000. Inhibition of osteoblast differentiation by tumor necrosis factor-α. Endocrinology 141:3956–64 35. Canalis E. 1986. Interleukin-1 has independent effects on deoxyribonucleic acid and collagen synthesis in cultures of rat calvariae. Endocrinology 118:74–81 36. Stashenko P, Dewhirst FE, Rooney ML, Desjardins LA, Heeley JD. 1987. Interleukin-1β is a potent inhibitor of bone formation in vitro. J. Bone Miner. Res. 2:559–65 37. Centrella M, McCarthy TL, Canalis E. 1988. Tumor necrosis factor-α inhibits collagen synthesis and alkaline phosphatase activity independently of its effect on deoxyribonucleic acid synthesis in osteoblast-enriched bone cell cultures. Endocrinology 123:1442–48 38. Smith DD, Gowen M, Mundy GR. 1987. Effects of interferon-γ and other cytokines on collagen synthesis in fetal rat bone cultures. Endocrinology 120:2494–99 39. Onoe Y, Miyaura C, Kaminakayashiki T, Nagai Y, Noguchi K, et al. 1996. IL-13 and IL-4 inhibit bone resorption by suppressing cyclooxygenase-2-dependent prostaglandin synthesis in osteoblasts. J. Immunol. 156:758–64 40. Lind M, Deleuran B, Yssel H, Fink-Eriksen E, Thestrup-Pedersen K. 1995. IL-4 and IL-13, but not IL-10, are chemotactic factors for human osteoblasts. Cytokine 7:78–82 41. Ura K, Morimoto I, Watanabe K, Saito K, Yanagihara N, Eto S. 2000. Interleukin (IL)-4

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

56

Walsh et al.

ANRV270-IY24-02

42. 43.

44.

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

45.

46.

47.

48. 49. 50. 51.

52. 53. 54.

55.

56.

57.

58. 59.

ARI

13 February 2006

12:28

and IL-13 inhibit the differentiation of murine osteoblastic MC3T3-E1 cells. Endocr. J. 47:293–302 Lewis DB, Liggitt HD, Effmann EL, Motley ST, Teitelbaum SL, et al. 1993. Osteoporosis induced in mice by overproduction of interleukin 4. Proc. Natl. Acad. Sci. USA 90:11618–22 Jilka RL, Weinstein RS, Bellido T, Parﬁtt AM, Manolagas SC. 1998. Osteoblast programmed cell death (apoptosis): modulation by growth factors and cytokines. J. Bone Miner. Res. 13:793–802 Rifas L, Arackal S, Weitzmann MN. 2003. Inﬂammatory T cells rapidly induce differentiation of human bone marrow stromal cells into mature osteoblasts. J. Cell Biochem. 88:650–59 Suh WK, Wang SX, Jheon AH, Moreno L, Yoshinaga SK, et al. 2004. The immune regulatory protein B7-H3 promotes osteoblast differentiation and bone mineralization. Proc. Natl. Acad. Sci. USA 101:12969–73 Horton JE, Raisz LG, Simmons HA, Oppenheim JJ, Mergenhagen SE. 1972. Bone resorbing activity in supernatant ﬂuid from cultured human peripheral blood leukocytes. Science 177:793–95 Dewhirst FE, Stashenko PP, Mole JE, Tsurumachi T. 1985. Puriﬁcation and partial sequence of human osteoclast-activating factor: identity with interleukin 1 β. J. Immunol. 135:2562–68 Martin TJ, Romas E, Gillespie MT. 1998. Interleukins in the control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8:107–23 Goldring SR, Gravallese EM. 2000. Pathogenesis of bone erosions in rheumatoid arthritis. Curr. Opin. Rheumatol. 12:195–99 O’Gradaigh D, Compston JE. 2004. T-cell involvement in osteoclast biology: implications for rheumatoid bone erosion. Rheumatology (Oxford) 43:122–30 Cenci S, Weitzmann MN, Roggia C, Namba N, Novack D, et al. 2000. Estrogen deﬁciency induces bone loss by enhancing T-cell production of TNF-α. J. Clin. Invest. 106:1229– 37 Sass DA, Liss T, Bowman AR, Rucinski B, Popoff SN, et al. 1997. The role of the Tlymphocyte in estrogen deﬁciency osteopenia. J. Bone Miner. Res. 12:479–86 Guise TA, Mundy GR. 1998. Cancer and bone. Endocr. Rev. 19:18–54 Wong BR, Rho J, Arron J, Robinson E, Orlinick J, et al. 1997. TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-Jun N-terminal kinase in T cells. J. Biol. Chem. 272:25190–94 Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, et al. 1998. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95:3597– 602 Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, et al. 1998. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93:165–76 Fuller K, Wong B, Fox S, Choi Y, Chambers TJ. 1998. TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J. Exp. Med. 188:997–1001 Walsh MC, Choi Y. 2003. Biology of the TRANCE axis. Cytokine Growth Factor Rev. 14:251–63 Theill LE, Boyle WJ, Penninger JM. 2002. RANK-L and RANK: T cells, bone loss, and mammalian evolution. Annu. Rev. Immunol. 20:795–823 www.annualreviews.org • Insights into Osteoimmunology

54. This key report identifies TRANCE as a novel TNF family member on activated T cells. 55 and 56. Both are key reports showing TRANCE as the long-sought OC differentiation factor.

57. This paper shows that TRANCE is a survival factor for OCs.

57

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

66 and 110. Key papers establishing that activated T cells and lymphocytederived TRANCE contribute to OC differentiation in vivo.

58

12:28

60. Suda T, Takahashi N, Udagawa N, Jimi E, Gillespie MT, Martin TJ. 1999. Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20:345–57 61. Tsuda E, Goto M, Mochizuki S, Yano K, Kobayashi F, et al. 1997. Isolation of a novel cytokine from human ﬁbroblasts that speciﬁcally inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234:137–42 62. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, et al. 1997. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89:309–19 63. Bucay N, Sarosi I, Dunstan CR, Morony S, Tarpley J, et al. 1998. osteoprotegerin-deﬁcient mice develop early onset osteoporosis and arterial calciﬁcation. Genes Dev. 12:1260– 68 64. Boyle WJ, Simonet WS, Lacey DL. 2003. Osteoclast differentiation and activation. Nature 423:337–42 65. Kong YY, Yoshida H, Sarosi I, Tan HL, Timms E, et al. 1999. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397:315–23 66. Kim N, Odgren PR, Kim DK, Marks SC Jr, Choi Y. 2000. Diverse roles of the tumor necrosis factor family member TRANCE in skeletal physiology revealed by TRANCE deficiency and partial rescue by a lymphocyte-expressed TRANCE transgene. Proc. Natl. Acad. Sci. USA 97:10905–10 67. Anderson DM, Maraskovsky E, Billingsley WL, Dougall WC, Tometsko ME, et al. 1997. A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390:175–79 68. Dougall WC, Glaccum M, Charrier K, Rohrbach K, Brasel K, et al. 1999. RANK is essential for osteoclast and lymph node development. Genes Dev. 13:2412–24 69. Hughes AE, Ralston SH, Marken J, Bell C, MacPherson H, et al. 2000. Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nat. Genet. 24:45–48 70. Whyte MP, Mumm S. 2004. Heritable disorders of the RANKL/OPG/RANK signaling pathway. J. Musculoskelet. Neuronal Interact.. 4:254–67 71. Whyte MP, Obrecht SE, Finnegan PM, Jones JL, Podgornik MN, et al. 2002. Osteoprotegerin deﬁciency and juvenile Paget’s disease. N. Engl. J. Med. 347:175–84 72. Whyte MP, Reinus WR, Podgornik MN, Mills BG. 2002. Familial expansile osteolysis (excessive RANK effect) in a 5-generation American kindred. Medicine (Baltimore) 81:101– 21 73. Whyte MP, Hughes AE. 2002. Expansile skeletal hyperphosphatasia is caused by a 15base pair tandem duplication in TNFRSF11A encoding RANK and is allelic to familial expansile osteolysis. J. Bone Miner. Res. 17:26–29 74. Darnay BG, Haridas V, Ni J, Moore PA, Aggarwal BB. 1998. Characterization of the intracellular domain of receptor activator of NF-κB (RANK). Interaction with tumor necrosis factor receptor-associated factors and activation of NF-κB and c-Jun N-terminal kinase. J. Biol. Chem. 273:20551–55 75. Wong BR, Besser D, Kim N, Arron JR, Vologodskaia M, et al. 1999. TRANCE, a TNF family member, activates Akt/PKB through a signaling complex involving TRAF6 and c-Src. Mol. Cell 4:1041–49 76. Wong BR, Josien R, Lee SY, Vologodskaia M, Steinman RM, Choi Y. 1998. The TRAF family of signal transducers mediates NF-κB activation by the TRANCE receptor. J. Biol. Chem. 273:28355–59 Walsh et al.

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

77. Ye H, Arron JR, Lamothe B, Cirilli M, Kobayashi T, et al. 2002. Distinct molecular mechanism for initiating TRAF6 signalling. Nature 418:443–47 78. Naito A, Azuma S, Tanaka S, Miyazaki T, Takaki S, et al. 1999. Severe osteopetrosis, defective interleukin-1 signalling and lymph node organogenesis in TRAF6-deﬁcient mice. Genes Cells 4:353–62 79. Lomaga MA, Yeh WC, Sarosi I, Duncan GS, Furlonger C, et al. 1999. TRAF6 deﬁciency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13:1015–24 80. Takayanagi H. 2005. Mechanistic insight into osteoclast differentiation in osteoimmunology. J. Mol. Med. 83:170–79 81. Lee NK, Choi YG, Baik JY, Han SY, Jeong DW, et al. 2005. A crucial role for reactive oxygen species in RANKL-induced osteoclast differentiation. Blood 106:852–59 82. Maggio D, Barabani M, Pierandrei M, Polidori MC, Catani M, et al. 2003. Marked decrease in plasma antioxidants in aged osteoporotic women: results of a cross-sectional study. J. Clin. Endocrinol. Metab. 88:1523–27 83. Singh DK, Kumar D, Siddiqui Z, Basu SK, Kumar V, Rao KV. 2005. The strength of receptor signaling is centrally controlled through a cooperative loop between Ca2+ and an oxidant signal. Cell 121:281–93 84. Steinman RM, Bonifaz L, Fujii S, Liu K, Bonnyay D, et al. 2005. The innate functions of dendritic cells in peripheral lymphoid tissues. Adv. Exp. Med. Biol. 560:83–97 85. Matsuo K, Owens JM, Tonko M, Elliott C, Chambers TJ, Wagner EF. 2000. Fosl1 is a transcriptional target of c-Fos during osteoclast differentiation. Nat. Genet. 24:184–87 86. Miyamoto T, Ohneda O, Arai F, Iwamoto K, Okada S, et al. 2001. Bifurcation of osteoclasts and dendritic cells from common progenitors. Blood 98:2544–54 87. Kamath AT, Henri S, Battye F, Tough DF, Shortman K. 2002. Developmental kinetics and lifespan of dendritic cells in mouse lymphoid organs. Blood 100:1734–41 88. Stoll S, Delon J, Brotz TM, Germain RN. 2002. Dynamic imaging of T cell-dendritic cell interactions in lymph nodes. Science 296:1873–76 89. Shakhar G, Lindquist RL, Skokos D, Dudziak D, Huang JH, et al. 2005. Stable T celldendritic cell interactions precede the development of both tolerance and immunity in vivo. Nat. Immunol. 6:707–14 90. Lindquist RL, Shakhar G, Dudziak D, Wardemann H, Eisenreich T, et al. 2004. Visualizing dendritic cell networks in vivo. Nat. Immunol.. 5:1243–50 91. Wong BR, Josien R, Lee SY, Sauter B, Li HL, et al. 1997. TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor. J. Exp. Med. 186:2075–80 92. Ouaaz F, Arron J, Zheng Y, Choi Y, Beg AA. 2002. Dendritic cell development and survival require distinct NF-κB subunits. Immunity 16:257–70 93. Arron JR, Vologodskaia M, Wong BR, Naramura M, Kim N, et al. 2001. A positive regulatory role for Cbl family proteins in tumor necrosis factor-related activation-induced cytokine (TRANCE) and CD40L-mediated Akt activation. J. Biol. Chem. 276:30011–17 94. Josien R, Li HL, Ingulli E, Sarma S, Wong BR, et al. 2000. TRANCE, a tumor necrosis factor family member, enhances the longevity and adjuvant properties of dendritic cells in vivo. J. Exp. Med. 191:495–502 95. Zhong L, Granelli-Piperno A, Choi Y, Steinman RM. 1999. Recombinant adenovirus is an efﬁcient and non-perturbing genetic vector for human dendritic cells. Eur. J. Immunol. 29:964–72 www.annualreviews.org • Insights into Osteoimmunology

67 and 91. Key reports identifying RANK as the TRANCE receptor and showing TRANCE is a survival factor for DCs.

59

ARI

13 February 2006

12:28

96. Wiethe C, Dittmar K, Doan T, Lindenmaier W, Tindle R. 2003. Enhanced effector and memory CTL responses generated by incorporation of receptor activator of NF-κB (RANK)/RANK ligand costimulatory molecules into dendritic cell immunogens expressing a human tumor-speciﬁc antigen. J. Immunol. 171:4121–30 97. Yun TJ, Tallquist MD, Aicher A, Rafferty KL, Marshall AJ, et al. 2001. Osteoprotegerin, a crucial regulator of bone metabolism, also regulates B cell development and function. J. Immunol. 166:1482–91 98. Bachmann MF, Wong BR, Josien R, Steinman RM, Oxenius A, Choi Y. 1999. TRANCE, a tumor necrosis factor family member critical for CD40 ligand-independent T helper cell activation. J. Exp. Med. 189:1025–31 99. Roy M, Waldschmidt T, Aruffo A, Ledbetter JA, Noelle RJ. 1993. The regulation of the expression of gp39, the CD40 ligand, on normal and cloned CD4+ T cells. J. Immunol. 151:2497–510 100. Josien R, Wong BR, Li HL, Steinman RM, Choi Y. 1999. TRANCE, a TNF family member, is differentially expressed on T cell subsets and induces cytokine production in dendritic cells. J. Immunol. 162:2562–68 101. Yun TJ, Chaudhary PM, Shu GL, Frazer JK, Ewings MK, et al. 1998. OPG/FDCR-1, a TNF receptor family member, is expressed in lymphoid cells and is up-regulated by ligating CD40. J. Immunol. 161:6113–21 102. Cremer I, Dieu-Nosjean MC, Marechal S, Dezutter-Dambuyant C, Goddard S, et al. 2002. Long-lived immature dendritic cells mediated by TRANCE-RANK interaction. Blood 100:3646–55 103. Williamson E, Bilsborough JM, Viney JL. 2002. Regulation of mucosal dendritic cell function by receptor activator of NF-κB (RANK)/RANK ligand interactions: impact on tolerance induction. J. Immunol. 169:3606–12 104. Green EA, Choi Y, Flavell RA. 2002. Pancreatic lymph node-derived CD4+ CD25+ Treg cells: highly potent regulators of diabetes that require TRANCE-RANK signals. Immunity 16:183–91 105. Aramaki O, Inoue F, Takayama T, Shimazu M, Kitajima M, et al. 2005. Various costimulatory pathways are essential for induction of regulatory cells by intratracheal delivery of alloantigen. Transplant Proc. 37:1934–36 106. Hochweller K, Anderton SM. 2005. Kinetics of costimulatory molecule expression by T cells and dendritic cells during the induction of tolerance versus immunity in vivo. Eur. J. Immunol. 35:1086–96 107. Kim D, Mebius RE, MacMicking JD, Jung S, Cupedo T, et al. 2000. Regulation of peripheral lymph node genesis by the tumor necrosis factor family member TRANCE. J. Exp. Med. 192:1467–78 108. Masuzawa T, Miyaura C, Onoe Y, Kusano K, Ohta H, et al. 1994. Estrogen deﬁciency stimulates B lymphopoiesis in mouse bone marrow. J. Clin. Invest. 94:1090–97 109. Wang R, Zhang L, Zhang X, Moreno J, Celluzzi C, et al. 2002. Regulation of activationinduced receptor activator of NF-κB ligand (RANKL) expression in T cells. Eur. J. Immunol. 32:1090–98 110. Kong YY, Feige U, Sarosi I, Bolon B, Tafuri A, et al. 1999. Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402:304–9 111. Pettit AR, Ji H, von Stechow D, Muller R, Goldring SR, et al. 2001. TRANCE/RANKL knockout mice are protected from bone erosion in a serum transfer model of arthritis. Am. J. Pathol. 159:1689–99

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

60

Walsh et al.

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-02

ARI

13 February 2006

12:28

112. Rivollier A, Mazzorana M, Tebib J, Piperno M, Aitsiselmi T, et al. 2004. Immature dendritic cell transdifferentiation into osteoclasts: a novel pathway sustained by the rheumatoid arthritis microenvironment. Blood 104:4029–37 113. Clark WB, Loe H. 1993. Mechanisms of initiation and progression of periodontal disease. Periodontology 2000 2:72–82 114. Teng YT, Nguyen H, Gao X, Kong YY, Gorczynski RM, et al. 2000. Functional human Tcell immunity and osteoprotegerin ligand control alveolar bone destruction in periodontal infection. J. Clin. Invest. 106:R59–67 115. Lopez I, Buchman AL. 2000. Metabolic bone disease in IBD. Curr. Gastroenterol. Rep. 2:317–22 116. Moschen AR, Kaser A, Enrich B, Ludwiczek O, Gabriel M, et al. 2005. The RANKL/OPG system is activated in inﬂammatory bowel disease and relates to the state of bone loss. Gut 54:479–87 117. Franchimont N, Reenaers C, Lambert C, Belaiche J, Bours V, et al. 2004. Increased expression of receptor activator of NF-κB ligand (RANKL), its receptor RANK and its decoy receptor osteoprotegerin in the colon of Crohn’s disease patients. Clin. Exp. Immunol. 138:491–98 118. Ashcroft AJ, Cruickshank SM, Croucher PI, Perry MJ, Rollinson S, et al. 2003. Colonic dendritic cells, intestinal inﬂammation, and T cell-mediated bone destruction are modulated by recombinant osteoprotegerin. Immunity 19:849–61 119. Mundy GR. 2002. Metastasis to bone: causes, consequences and therapeutic opportunities. Nat. Rev. Cancer 2:584–93 120. Takayanagi H, Ogasawara K, Hida S, Chiba T, Murata S, et al. 2000. T-cellmediated regulation of osteoclastogenesis by signalling cross-talk between RANKL and IFN-γ. Nature 408:600–5 121. Quinn JM, Itoh K, Udagawa N, Hausler K, Yasuda H, et al. 2001. Transforming growth factor β affects osteoclast differentiation via direct and indirect actions. J. Bone Miner. Res. 16:1787–94 122. Takeda K, Akira S. 2004. TLR signaling pathways. Semin. Immunol. 16:3–9 123. Takami M, Kim N, Rho J, Choi Y. 2002. Stimulation by Toll-like receptors inhibits osteoclast differentiation. J. Immunol. 169:1516–23 124. Kikuchi T, Matsuguchi T, Tsuboi N, Mitani A, Tanaka S, et al. 2001. Gene expression of osteoclast differentiation factor is induced by lipopolysaccharide in mouse osteoblasts via Toll-like receptors. J. Immunol. 166:3574–79 125. Hayashi S, Yamada T, Tsuneto M, Yamane T, Takahashi M, et al. 2003. Distinct osteoclast precursors in the bone marrow and extramedullary organs characterized by responsiveness to Toll-like receptor ligands and TNF-α. J. Immunol. 171:5130–39 126. Nair SP, Meghji S, Wilson M, Reddi K, White P, Henderson B. 1996. Bacterially induced bone destruction: mechanisms and misconceptions. Infect. Immun. 64:2371– 80 127. Zou W, Schwartz H, Endres S, Hartmann G, Bar-Shavit Z. 2002. CpG oligonucleotides: novel regulators of osteoclast differentiation. FASEB J. 16:274–82 128. Sato N, Takahashi N, Suda K, Nakamura M, Yamaki M, et al. 2004. MyD88 but not TRIF is essential for osteoclastogenesis induced by lipopolysaccharide, diacyl lipopeptide, and IL-1α. J. Exp. Med. 200:601–11 129. Li H, Cuartas E, Cui W, Choi Y, Crawford TD, et al. 2005. IL-1 receptor-associated kinase M is a central regulator of osteoclast differentiation and activation. J. Exp. Med. 201:1169–77 www.annualreviews.org • Insights into Osteoimmunology

120. This paper shows a mechanism for how most normal T cell immune responses fail to elicit aberrant OC differentiation.

61

ANRV270-IY24-02

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

131. A key paper for the costimulation hypothesis of OC differentiation.

132 and 133. Key papers describing ITAM-containing adapters and how their associated receptors serve as costimulatory receptors for OC differentiation.

138. This paper demonstrated the importance of NFATc1 for OC differentiation.

62

12:28

130. Allison JP, Krummel MF. 1995. The yin and yang of T cell costimulation. Science 270:932– 33 131. Kim N, Takami M, Rho J, Josien R, Choi Y. 2002. A novel member of the leukocyte receptor complex regulates osteoclast differentiation. J. Exp. Med. 195:201–9 132. Koga T, Inui M, Inoue K, Kim S, Suematsu A, et al. 2004. Costimulatory signals mediated by the ITAM motif cooperate with RANKL for bone homeostasis. Nature 428:758–63 133. Mocsai A, Humphrey MB, Van Ziffle JA, Hu Y, Burghardt A, et al. 2004. The immunomodulatory adapter proteins DAP12 and Fc receptor γ-chain (FcRγ) regulate development of functional osteoclasts through the Syk tyrosine kinase. Proc. Natl. Acad. Sci. USA 101:6158–63 134. Faccio R, Teitelbaum SL, Fujikawa K, Chappel J, Zallone A, et al. 2005. Vav3 regulates osteoclast function and bone mass. Nat. Med. 11:284–90 135. Samelson LE. 2002. Signal transduction mediated by the T cell antigen receptor: the role of adapter proteins. Annu. Rev. Immunol. 20:371–94 136. Rudd CE, Raab M. 2003. Independent CD28 signaling via VAV and SLP-76: a model for in trans costimulation. Immunol. Rev. 192:32–41 137. Wada T, Nakashima T, Oliveira-Dos-Santos AJ, Gasser J, Hara H, et al. 2005. The molecular scaffold Gab2 is a crucial component of RANK signaling and osteoclastogenesis. Nat. Med. 11:394–99 138. Takayanagi H, Kim S, Koga T, Nishina H, Isshiki M, et al. 2002. Induction and activation of the transcription factor NFATc1 (NFAT2) integrate RANKL signaling in terminal differentiation of osteoclasts. Dev. Cell 3:889–901 139. Crabtree GR. 1999. Generic signals and speciﬁc outcomes: signaling through Ca2+ , calcineurin, and NF-AT. Cell 96:611–14 140. Takayanagi H, Kim S, Matsuo K, Suzuki H, Suzuki T, et al. 2002. RANKL maintains bone homeostasis through c-Fos-dependent induction of interferon-β. Nature 416:744– 49 141. Hayashi T, Kaneda T, Toyama Y, Kumegawa M, Hakeda Y. 2002. Regulation of receptor activator of NF-κB ligand-induced osteoclastogenesis by endogenous interferon-β (INF-β) and suppressors of cytokine signaling (SOCS). The possible counteracting role of SOCSs in IFN-β-inhibited osteoclast formation. J. Biol. Chem. 277:27880–86 142. Fox SW, Haque SJ, Lovibond AC, Chambers TJ. 2003. The possible role of TGF-βinduced suppressors of cytokine signaling expression in osteoclast/macrophage lineage commitment in vitro. J. Immunol. 170:3679–87 143. Ohishi M, Matsumura Y, Aki D, Mashima R, Taniguchi K, et al. 2005. Suppressors of cytokine signaling-1 and -3 regulate osteoclastogenesis in the presence of inﬂammatory cytokines. J. Immunol. 174:3024–31 144. Levy DE, Darnell JE Jr. 2002. Stats: transcriptional control and biological impact. Nat. Rev. Mol. Cell Biol. 3:651–62 145. Ramana CV, Gil MP, Schreiber RD, Stark GR. 2002. Stat1-dependent and -independent pathways in IFN-γ-dependent signaling. Trends Immunol. 23:96–101 146. Taniguchi T, Takaoka A. 2002. The interferon-α/β system in antiviral responses: a multimodal machinery of gene regulation by the IRF family of transcription factors. Curr. Opin. Immunol. 14:111–16 147. Kim S, Koga T, Isobe M, Kern BE, Yokochi T, et al. 2003. Stat1 functions as a cytoplasmic attenuator of Runx2 in the transcriptional program of osteoblast differentiation. Genes Dev. 17:1979–91 Walsh et al.

ANRV270-IY24-02

ARI

13 February 2006

12:28

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

148. Kadono Y, Okada F, Perchonock C, Jang HD, Lee SY, et al. 2005. Strength of TRAF6 signalling determines osteoclastogenesis. EMBO Rep. 6:171–76 149. Traverse S, Seedorf K, Paterson H, Marshall CJ, Cohen P, Ullrich A. 1994. EGF triggers neuronal differentiation of PC12 cells that overexpress the EGF receptor. Curr. Biol. 4:694–701 150. Vaudry D, Stork PJ, Lazarovici P, Eiden LE. 2002. Signaling pathways for PC12 cell differentiation: making the right connections. Science 296:1648–49

www.annualreviews.org • Insights into Osteoimmunology

63

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:33-63. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-03

ARI

13 February 2006

11:54

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaqu´ın Madrenas The FOCIS Center for Clinical Immunology and Immunotherapeutics, Robarts Research Institute, and Department of Microbiology and Immunology and Department of Medicine, University of Western Ontario, London, Ontario, Canada, N6A 5K8; email: [email protected]

Annu. Rev. Immunol. 2006. 24:65–97 First published online as a Review in Advance on November 8, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090535 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0065$20.00

Key Words costimulation, T cell activation, immunological synapse, PP2A, B7, signal plasticity

Abstract Within the paradigm of the two-signal model of lymphocyte activation, the interest in costimulation has witnessed a remarkable emergence in the past few years with the discovery of a large array of molecules that can serve this role, including some with an inhibitory function. Interest has been further enhanced by the realization of these molecules’ potential as targets to modulate clinical immune responses. Although the therapeutic translation of mechanistic knowledge in costimulatory molecules has been relatively straightforward, the capacity to target their inhibitory counterparts has remained limited. This limited capacity is particularly apparent in the case of the cytotoxic T lymphocyte-associated antigen-4 (CTLA-4), a major negative regulator of T cell responses. Because there have been several previous comprehensive reviews on the function of this molecule, we focus here on the physiological implications of its structural features. Such an exercise may ultimately help us to design immunotherapeutic agents that target CTLA-4.

65

ANRV270-IY24-03

ARI

13 February 2006

11:54

INTRODUCTION

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CTLA-4: cytotoxic T lymphocyteassociated antigen-4

Cytotoxic T lymphocyte-associated antigen4 (CTLA-4), also known as CD152, was initially described during a search for cytotoxic genes using subtractive hybridization (1). It is a type 1 transmembrane glycoprotein of the immunoglobin superfamily (1), 223 amino acids (aa) in length, with a 35 aa signal peptide (1, 2), and it is found as a covalent homodimer of 41–43 kDa (3, 4). The extracellular architecture of CTLA-4 is characterized by a single IgV-like domain containing the B71 (CD80)/B7-2 (CD86) ligand-binding site (1, 2, 4–6). Homodimerization of CTLA-4 is mediated by cysteine-dependent bonding at position 122 in the stalk region and by Nglycosylation at positions 78 and 110 (4, 7). The cytoplasmic portion of CTLA-4 is 36 aa in length and lacks any intrinsic enzymatic activity (8). It contains a lysine-rich motif located proximal to the membrane (9), two tyrosine residues at positions 165 and 182, and a proline-rich region starting at position 169 (8). Altogether, these areas have been implicated in the association of CTLA-4 with a variety of signaling molecules modulating its function (9–12). The simplicity of this structure disguises its critical function as a negative regulator of T cell–mediated immune responses (13–16).

MOLECULAR GENETICS OF CTLA-4 The CTLA-4 gene is located on chromosome 2 in humans and on chromosome 1 in mice (2, 17). This gene consists of 4 exons: Exons 1 and 2 encode the extracellular portion of the molecule, with exon 1 containing the leader peptide sequence and exon 2 the ligand-binding site; exon 3 encodes the transmembrane region, and exon 4 codes for the cytoplasmic tail (2, 17) (Figure 1). The resulting transcript can undergo alternative splicing, which is different between human and mouse CTLA-4. For human CTLA-4, one can detect (a) full-length mRNA containing 66

Teft

·

Kirchhof

·

Madrenas

exons 1 to 4, (b) a transcript coding for soluble CTLA-4 (sCTLA-4) that does not contain exon 3, and (c) a transcript coding only for exons 1 and 4 (1, 2, 17–21). Mouse T cells can express an additional CTLA-4 transcript that is known as ligand-independent CTLA-4 (liCTLA-4), containing exons 1, 3, and 4 (18). The CTLA-4 gene is mainly expressed in T cells upon activation (22), although it is constitutively expressed in the CD4+ CD25+ regulatory T cell (Treg) subset (3, 22–24). Memory T cells have a relatively large pool of intracellular CTLA-4, which is rapidly expressed on the cell surface upon activation compared to resting naive T cells, and CTLA4 is retained on the surface of memory T cells for a considerably longer time than is the case on activated naive T cells (25). Interestingly, mouse resting memory T cells express higher levels of liCTLA-4, which is downregulated upon activation (19). CTLA-4 expression is also detectable at the RNA and protein levels in resting thymocytes (1, 26) and is upregulated in all thymocyte populations upon in vivo activation with anti-CD3 antibodies (Abs) (27). Further reports have also shown CTLA-4 gene expression in a variety of cells, including B cells, monocytes, granulocytes, CD34+ stem cells, placental ﬁbroblasts, and mouse embryonic cells and embryoid bodies (28–32). The precise function of CTLA-4 in non-T cells is unknown. However, signaling to these cells by CTLA-4 ligation or to other cells via B7 could have a so far unknown regulatory role. The transcriptional regulation of CTLA4 gene expression is only partially known. It is likely initiated about 335 bp upstream from the start codon (17, 23) and may be dependent on NFAT (nuclear factor of activated T cells) because modulation of NFAT levels correlates directly with CTLA-4 expression (33, 34). Furthermore, inhibition of NFAT activation by cyclosporine A causes a marked decrease in CTLA-4 gene transcription (34). Sequence analysis of the 5 upstream region of the CTLA-4 gene revealed several transcriptional regulatory elements, including sites for

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

Figure 1 The CTLA-4 gene and its various splice variants. The CTLA-4 gene consists of four exons that can give rise to four different splice variants. Each exon codes for a domain in the protein as indicated in the top of the ﬁgure. Humans express the full-length CTLA-4, exon 1–4 CTLA-4, and soluble CTLA-4 splice variants, whereas mice additionally express the ligand-independent CTLA-4.

AP-1, STAT, GATA-1, NF-κB, Oct-1, and an IL-4 negative regulatory element (17, 22, 34). Also, CTLA-4 expression is upregulated by cAMP, suggesting a cAMP-dependent regulation of CTLA-4 gene transcription (35). In T cells, CTLA-4 mRNA is detected after 1 h of T cell receptor (TCR) engagement and peaks around 24–36 h after stimulation (3, 22, 23). The translational controls and machinery responsible for the stabilization of CTLA-4 mRNA have not been completely elucidated. Investigators believe that the 3 UTR of CTLA-4 plays a major role in its stability because the 3 UTR contains three AUUUA motifs implicated as binding sites for proteins that mediate RNA degradation and as such determine the half-life of the mRNA (34, 36). Of interest, soluble CTLA-4 mRNA has

a shorter half-life than full-length CTLA-4 mRNA (37). The stability of CTLA-4 mRNA has also been tied to TCR signaling and costimulation. Consistent with this, the half-life of CTLA-4 mRNA is 4.6 h under conditions of TCR signaling alone and increases to 8.9 h in the presence of CD28-mediated costimulation (22, 34, 38). Expression of CTLA-4 protein on the surface of T cells is not detected until 24–48 h after activation (22). This export event involves a signal peptide that ensures proper targeting to the rough endoplasmic reticulum and subsequent localization within the membrane, and the leader peptide is cleaved off between serine-35 and glutamine-36 (1, 2). Once at the surface, CTLA-4 can be either retained or rapidly internalized by endocytosis and targeted to lysosomes, depending on its www.annualreviews.org • CTLA- 4 Function

Costimulation: also known as “second signal,” costimulation refers to a group of biochemical events that, in addition to antigen receptor–dependent signals, are required for full activation of T and B lymphocytes

67

ANRV270-IY24-03

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

T cell activation: changes in gene expression that lead to the differentiation of a T lymphocyte into an effector cell or a memory cell, resulting from signals emanating from cell surface receptors (e.g., TCR) Polymorphism: high degree of allelic variation for a given gene

68

11:54

molecular interactions at the surface through mechanisms discussed below. Upon internalization, it is quickly degraded with a half-life of 2 h in activated T cells (38–41). Recycling in the absence of TCR ligation has not been formally demonstrated, but TCR signaling can induce CTLA-4 recycling from the lysosomes to the surface of the T cell, providing an alternative method to increase surface expression of CTLA-4 in a timely manner (41).

THE EVOLUTIONARY BIOLOGY OF CTLA-4 Soon after CTLA-4 was identiﬁed (1), it was cloned and mapped to the same chromosomal region as CD28 in both the mouse and human genome (1, 2, 23, 42). At the protein level, CTLA-4 and CD28 share approximately 30% identity (43). Owing to the genetic and molecular similarities between CD28 and CTLA-4, the evolution of CTLA-4 has been attributed to a duplication event of an ancient costimulatory gene (23, 42, 44). There is an increasing degree of conservation of CTLA-4 from the extracellular domain to the intracellular tail. The intracellular tail of CTLA-4 is 100% conserved among mammalian species, as is the cysteine (C) residue mediating covalent bonding between individual CTLA-4 chains, although there is greater variability in the extracellular V domain (17) (Figure 2). This suggests that there is a selective pressure to conserve the mechanisms that mediate CTLA-4 dimerization and signaling through its cytoplasmic tail. CTLA-4 is expressed in humans, primates, dogs, cats, cows, sheep, rats, and mice (1, 2, 45, 46). Recent reports of a putative CTLA-4-like molecule in chickens and in rainbow trout require functional characterization (NCBI accession numbers XM 421960.1 and AY789436.1, respectively) but may provide additional insights into the evolution of the CTLA-4 molecule and the selective pressure for its mechanisms of action (Figure 3). These two homologs have cytoplasmic tails lacking many of the features seen in the wellTeft

·

Kirchhof

·

Madrenas

characterized mammalian CTLA-4. Speciﬁcally, trout CTLA-4 lacks tyrosine residues in its tail, and chicken CTLA-4 contains only one of the conserved tyrosine residues in the YVKM motif. This is consistent with the idea that the tail of CTLA-4 may have evolved from these species to take advantage of the signaling pathways present in the T cell, in addition to a more ancestral mechanism of sequestration of costimulatory ligands (see below for details). Experimental data are required to determine if the magnitude of upregulation of CTLA-4 is comparable among species. In humans, CTLA-4 is not expressed on the surface of naive, resting T cells. Only after activation is it expressed on the surface at low levels (less than 10% of total CTLA-4) (38, 47). One can argue that, given the lack of conventional signaling domains in the cytoplasmic tail of trout and chicken CTLA-4, this molecule should be massively upregulated upon T cell activation in these two nonmammalian species to compete with CD28-dependent costimulation (48).

CTLA-4 POLYMORPHISMS Given the evolutionary conservation of CTLA-4 and its differential expression in T cell subsets, transcriptional or translational changes in its expression may have serious immunological consequences. This has justiﬁed the search for polymorphisms and their link with immune-mediated diseases. To date, four main polymorphisms of the CTLA-4 gene have been identiﬁed and studied in the context of autoimmune disorders. Single nucleotide polymorphisms have been noted in human CTLA-4 at positions –1722, –1661, and –318, all in the regulatory/promoter region, and at position +49 in exon 1, which results in the substitution of an alanine for a threonine in the signal sequence of CTLA-4 (49–53). The A49G polymorphism is the only polymorphism that changes the primary amino acid sequence of CTLA-4. In vitro studies of A49G CTLA-4 have shown that this mutant form

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

Figure 2 Alignment of reported mammalian CTLA-4 protein sequences. Amino acids in red indicate conserved residues among mammalian species; amino acids in blue indicate those with conserved biochemical properties. Speciﬁc areas of potential functional relevance (all of them conserved across mammalian species) are within yellow boxes. The greatest divergence of sequence occurs at the N terminus, which codes for the extracellular portion of the molecule excluding the B7-binding region. The C-terminal cytoplasmic tail of the molecule is 100% conserved among mammals.

of CTLA-4 is aberrantly processed in the endoplasmic reticulum, leading to reduced surface expression (54). In contrast, the C-318T polymorphism has been associated with higher promoter activity and subsequently increased CTLA-4 expression (52). The other two polymorphisms, T-1772C and A-1661G, are less characterized, and further investigation is required to deﬁne their effects on CTLA-4 expression and association with immunological pathologies. In addition, a dinucleotide repeat polymorphism in the 3 UTR of CTLA-4 has been identiﬁed (55). Patients with longer AT repeats in the 3 UTR

have T cells that have higher proliferative responses when stimulated with anti-CD3/antiCD28 (56). The correlation between all these polymorphisms of CTLA-4 and susceptibility to autoimmune disorders has been contentious, with some reports indicating strong correlation and others disputing such a correlation. More importantly, how these polymorphisms affect CTLA-4 function is unclear. They may affect CTLA-4 processing and intracellular trafﬁcking, or they may affect oligomerization and surface retention through indirect effects on CTLA-4 posttranslation modiﬁcations. www.annualreviews.org • CTLA- 4 Function

Oligomerization: structurally organized arrangement of individual molecules of a protein beyond dimers

69

ANRV270-IY24-03

ARI

13 February 2006

11:54

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CTLA-4 AS AN INHIBITORY RECEPTOR IN T CELLS: IN VIVO EVIDENCE

Figure 3 Dendrogram indicating the degree of homology between CTLA-4 of various species. Note that human CTLA-4 most closely resembles chimpanzee CTLA-4 and has the greatest divergence from rainbow trout CTLA-4.

An additional polymorphism detected in mice is linked to the expression of liCTLA4. The expression of this alternatively spliced form is dependent on a single nucleotide polymorphism at base 77 of exon 2 (18). In diabetes-resistant mice, base 77 of exon 2 is an adenine, whereas in diabetes-susceptible mice base 77 of exon 2 is a guanine (18). Ueda et al. (18) predict that this polymorphism works by modulating alternative splicing of the CTLA-4 transcript so that the guanine favors the inclusion of exon 2 in the ﬁnal transcript, whereas the adenine does not. For comparison, the human CTLA-4 base 77 of exon 2 is a thymine that also appears to favor the transcription/translation of full-length CTLA-4 versus the liCTLA4 splice variant, which may explain why liCTLA-4 has not been detected in humans (18). 70

Teft

·

Kirchhof

·

Madrenas

The structural similarities between CTLA-4 and CD28 suggested that the former was another costimulatory receptor. Early in vitro evidence, using anti-CTLA-4 monoclonal Abs, showed that ligation of CTLA-4 enhanced T cell proliferation and IL-2 production (57–59). Subsequently, it was demonstrated that crosslinking of CTLA-4 downregulated T cell proliferation and IL-2 production, whereas blockade of CTLA-4 with Fabs had an enhancing effect on T cell activation. Together, these data pointed toward a negative role for CTLA-4 in T cell activation (60, 61). Further evidence for CTLA-4 as an inhibitory receptor came from the phenotype of CTLA-4-deﬁcient mice (62, 63). These mice are born healthy. The thymus of CTLA-4deﬁcient mice remains normal in size, alterations in thymocyte populations are not detected until late stages of their life (62–64), and negative selection seems mostly normal (65), although CTLA-4 may ﬁne tune negative selection (66). However, 5–6 days after birth a large percentage of T cells become activated and inﬁltrate nonlymphoid tissues, resulting in the death of the mice by 3–4 weeks. Pathology in these mice reveals substantial tissue destruction with severe myocarditis and pancreatitis, lymphadenopathy and splenomegaly, with their peripheral lymphoid organs containing 5–10 times the normal amount of lymphocytes. These mice also exhibit elevated levels of serum immunoglobulin owing to an increase in B cell activation (63). CTLA-4-deﬁcient T cells display a phenotype characteristic of an activated T cell (CD69+ , CD25+ , CD44hi ). The increase in T cell activation is not restricted to a particular subset of cells, as the ratio of CD4+ to CD8+ T cells remains constant and the TCR repertoire is unaltered. CTLA-4−/− T cells are capable of spontaneous proliferation in vitro and secrete increased amounts

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

of interferon-γ (IFN-γ), interleukin-4 (IL-4), and granulocyte-macrophage colony-stimulating factor (GM-CSF) upon TCR engagement compared with normal mice (62). Additionally, the number of T cells progressing through the cell cycle was 3–4 times higher in CTLA-4-deﬁcient mice compared with normal littermate controls, as measured by BrdU labeling (64). Such a robust proliferation of CTLA-4-deﬁcient T cells is not due to a defect in apoptosis, as these cells are susceptible to Fas-mediated apoptosis (65). Together, this phenotype suggests a generalized expansion of self-reactive T cells that are no longer held in check, implying that CTLA-4 is involved in negative regulation of T cell activation. Subsequent studies have corroborated this interpretation and shown that CTLA-4 may act by terminating cytokine production and proliferation (67). These results indicate the essential role of CTLA-4 in the negative regulation of T cell activation as well as its requirement for maintenance of T cell homeostasis. Experimental data generated from TCRtransgenic mice suggest that antigen-speciﬁc TCR signaling is required for the initiation of T cell activation in CTLA-4-deﬁcient mice (65). In addition to TCR stimulation, CTLA4-deﬁcient T cells require costimulation. Recombinant CTLA-4-Ig can effectively block CD28:B7 interactions by binding to both B7-1 and B7-2, thereby preventing CD28dependent costimulation (68). Treatment of CTLA-4−/− mice with daily injections of CTLA-4-Ig protected these animals from fatal lymphoproliferation and multi-organ inﬁltration in vivo (64, 69). The absolute numbers of CD4+ and CD8+ T cells remained normal, and these cells retained a naive phenotype (64). Isolated splenocytes and lymph node cells failed to proliferate spontaneously in vitro compared with cells isolated from isotype-treated CTLA-4-deﬁcient mice. Removal of CTLA-4-Ig treatment resulted in the rapid induction of T cell activation and the onset of the lymphoproliferative disease (69). These results rule out a defect in the TCR-

dependent signaling machinery as a cause of the phenotype of CTLA-4−/− mice. An observation that is still puzzling is that the CTLA-4 deﬁciency in T cells is not cell autonomous. Although lack of CTLA-4 correlates with disregulated T cell activation and homeostasis, reconstituted RAG-deﬁcient mice receiving equal numbers of wild-type T cells and CTLA-4-deﬁcient T cells do not develop lymphoproliferation and remain mostly healthy (70, 71). In the presence of CTLA-4-expressing T cells, CTLA-4−/− T cells undergo normal activation, expansion, and contraction when challenged with acute, chronic, and persistent infections (72). Together, these data suggest that although CTLA-4 may provide inhibitory signals in a cell autonomous fashion, additional factors (potentially immunosuppressive cytokines) provided by all or at least a subset of CTLA-4expressing T cells (potentially Tregs) can protect mice from the lymphoproliferative disorder seen in CTLA-4-deﬁcient mice (72, 73). In summary, most of the observations of CTLA-4 function point clearly toward a negative regulatory role on TCR-mediated T cell activation. Such an effect can be revealed by coligation of CTLA-4 with the TCR or, alternatively, by blockade of B7-dependent costimulation with CTLA-4. Either one of these strategies leads to a decrease in T cell responses as measured by in vitro readouts or in vivo functional outcomes.

T cell homeostasis: regulation of steady state numbers of T lymphocytes

CTLA-4 LIGANDS: B7-1 AND B7-2 The structural similarities and sequence homology between CTLA-4 and CD28 extend to shared natural ligands: the B7 molecules B7-1 (CD80) and B7-2 (CD86) (74–76). However, a key feature that may explain CTLA-4’s mechanism of action is that CTLA-4 binds B7-1 and B7-2 with greater afﬁnity and avidity compared with CD28 (68, 77). Although initial measurements suggested a 100-fold increased afﬁnity and avidity for CTLA-4:B7 interactions, more recent www.annualreviews.org • CTLA- 4 Function

71

ANRV270-IY24-03

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

APC: antigen-presenting cell Immunological synapse (IS): an organized interface between a T lymphocyte and an APC, involving a central core in which engaged antigen receptors and costimulatory molecules are clustered, and a peripheral ring composed of engaged adhesion molecules

72

11:54

data using surface plasmon resonance experiments revealed that the interactions between these molecules are 10-fold weaker owing to rapid dissociation rates. CTLA-4 binds B7-1 and B7-2 with dissociation constants (Kd) of 0.2 and 2.6 μM, respectively, whereas CD28 binds B7-1 and B7-2 with Kd values of 4 and 20 μM, respectively (78–80). The higher afﬁnity and avidity of CTLA-4 over CD28 for the B7 molecules suggest that inhibitory signals are dominant over costimulatory signals. As CTLA-4 is primarily an intracellular molecule and is maximally expressed 30- to 50-fold less than CD28 on the cell surface following activation, the stronger CTLA-4:B7 interaction compensates for the bias towards CD28 costimulation (59). The higher avidity of CTLA-4 for B7 is due to the structural topology of the CTLA-4 homodimer that allows for bivalent binding to B7 (81). In contrast, CD28 is only able to bind B7 monovalently (78, 82). An additional factor to consider is that the binding of CTLA-4 and of CD28 to both B7-1 and B7-2 occurs via the same MYPPPY motif located on exon 2 (83). Thus, the increased CTLA-4 afﬁnity in the context of the conservation of the MYPPPY sequence suggests that residues outside of the MYPPPY motif play a role in stabilizing interactions with B7 molecules (78). Most studies in the literature have considered the binding of CTLA-4 to B7-1 functionally equivalent to the binding to B7-2. However, recent data have revealed differences in the response to CTLA-4 ligation by each of these molecules individually. It is important to point out that B7-1 and B7-2, although similar, are distinct molecules that have unique proﬁles of expression. The expression patterns of the B7 ligands mimic that of CD28 and CTLA-4. B7-2 is constitutively expressed on resting antigen-presenting cells (APCs) and is upregulated rapidly after activation, reaching peak levels at 24 h (84–86). In contrast, B7-1 expression is induced after activation, peaking around 48–72 h (84). Researchers interpreted the differences in temporal regulation of these molecules by conTeft

·

Kirchhof

·

Madrenas

cluding that B7-2 is the predominant CD28 (costimulatory) ligand and B7-1 is the predominant CTLA-4 (inhibitory) ligand. In addition, the binding properties of B7-1 and B7-2 to CTLA-4 and CD28 are such that the ratio of CTLA-4 engagement to CD28 engagement in binding to B7-1 is approximately 20 to 1, compared with a ratio of 8 to 1 in binding to B7-2 (78, 80). In humans, because B7-1 is maximally expressed upon activation of APCs, whereas B7-2 is constitutively expressed, it follows that CTLA-4 interactions with B7-1 will be more prominent later in the activation process, mediating stronger inhibitory capacities upon CTLA-4 engagement (84–86). In addition, CD28 homodimers, but not CTLA-4 homodimers, are predicted to be monovalent, suggesting differences in how these receptors interact with their ligands (78, 87). Crystallographic data are consistent in showing that CTLA4 homodimers interact with alternating B7-1 homodimers, forming a lattice-like structure (81). In contrast, homodimerization of B7-2 is considered unlikely, thereby limiting the extent of a similar lattice with CTLA-4 (78, 82). At the cellular level, recent data show that B7-2 stabilizes CD28 at the immunological synapse (IS), whereas B7-1 preferentially recruits CTLA-4 to the synapse (88). On the basis of this type of data, the CTLA-4/CD28binding paradigm has been reﬁned to propose that B7-1 and B7-2 are individual ligands with potentially independent functions, as shown by their differential interactions with CTLA4 and CD28 during the timeline of an immune response. Such a paradigm helps explain the different effects of B7-1 versus B7-2 blockade on the progression of autoimmunity (89). Although in the past ﬁve years new members of the B7 family have been identiﬁed and shown to have costimulatory and coinhibitory function (reviewed in 43, 90), only B7-1 and B7-2 behave as true endogenous ligands of CTLA-4. However, the potential for additional ligands remains, based on data pointing to alternative forms of CTLA-4 with B7independent effects and to an inherent signal

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

plasticity of CTLA-4 revealed by recombinant ligands. Speciﬁcally, Kuchroo’s group (19) has revealed the existence of a splice variant of CTLA-4 that lacks exon 2 and thus does not have a B7-binding domain. This liCTLA4 is expressed at higher levels in resting memory T cells and appears to be functional as an attenuator of T cell activation. Similar B7independent CTLA-4 functions have been revealed by Bluestone’s group (91). Additionally, our group has shown that one can convert CTLA-4 from inhibitor to activator of T cells with a recombinant, bispeciﬁc tandem ScFv molecule named 24:26, raising the possibility of an endogenous ligand having a similar role (see below) (92). Similar activating effects of CTLA-4 ligation have been reported by Liu’s group (57) and more recently by Nandi’s group (93). The idea of a B7-independent function of CTLA-4 is also suggested by data linking liCTLA-4 both to resistance to autoimmune diabetes (19) and to decreased T cell proliferation and IFN-γ production in response to APC/anti-CD3 stimulation (19, 91). Similar conclusions have been reached by looking at the effect of reconstitution of CTLA-4−/− mice with a mutant CTLA-4 lacking a functional B7-binding site. The mechanism by which these B7-independent forms may work is unclear, but it has been linked to decreased phosphorylation of TCR-ζ and decreased TCR signaling (19). In fact, in these studies liCTLA-4 was more potent at inhibiting TCR-ζ phosphorylation than was full-length CTLA-4. The increased signaling capacity is illustrated by the fact that diabetes-resistant mice have higher expression of liCTLA-4 in memory/regulatory T cells than do diabetessusceptible mice (19).

CTLA-4 DIMERIZATION AND HIGHER ORDER OLIGOMERIZATION The crystal structure of CTLA-4 bound to B7-1 suggests that the bivalent binding of CTLA-4 to different B7-1 dimers

would lead to the formation of a repeating lattice-like structure of alternating CTLA-4/ B7 molecules (81, 94–96). Such an arrangement is contingent on CTLA-4 dimerization. Because CTLA-4 function may be determined by dimerization and subsequent high-order CTLA-4 arrangement, one may assume that these structural characteristics would be tightly regulated. This seems to be the case. CTLA-4 dimerization in solution is mediated by intermolecular disulﬁde bond formation through the C residue at position 122 [located in the extracellular portion of CTLA-4 (4)]. The CTLA-4 dimer is further stabilized by the interface between each monomer through 4 hydrogen bonds, 48 van der Waals contacts, and 15 water-mediated hydrogen bonds (5). However, in vivo, regulation of CTLA-4 dimerization is more complex and involves a hierarchical process that includes, in addition to C122-dependent disulﬁde bonding, N-glycosylation at N78 and N110 and engagement to B7 (7, 97) (Figure 4). Interestingly, both the C122 responsible for intermolecular disulﬁde bonding and the N-glycosylation sites are evolutionarily conserved among species, suggesting that the formation of the B7-dependent, CTLA-4-dimer-based oligomers is important for the maintenance of CTLA-4 function. Such an arrangement may be important for the mechanism of T cell inactivation either by facilitating the initiation of CTLA4-mediated signaling or the sequestration of B7 by CTLA-4, or both (see below). It is important to note that the current paradigm for receptor-mediated signaling involves either ligand-induced conformational changes (98) or ligand-induced dimerization (99). However, CTLA-4 does not undergo any detectable conformation change upon B7 binding (81), and prior to ligation, it already exists as a nonfunctional covalent homodimer (4, 7), suggesting that other mechanisms of signal initiation may be involved. Upon interaction with the APC, CTLA-4 on the T cell surface accumulates at the IS (for reviews see 100, 101). Such redistribution www.annualreviews.org • CTLA- 4 Function

Plasticity: the ability to generate different or polar outcomes from the same functional unit (either molecule, cell, or organism)

73

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

Figure 4 Hierarchical regulation of CTLA-4 dimerization and oligomer formation. Dimerization of CTLA-4 is primarily dependent on an intermolecular disulﬁde bond at cysteine 122 that is formed in the endoplasmic reticulum. Molecules lacking this disulﬁde bond undergo aberrant glycosylation at N78 and N110 in the Golgi that helps stabilize CTLA-4 dimers. CTLA-4 molecules deﬁcient in disulﬁde bonding and N-glycosylation can still dimerize upon engagement with B7 at the T cell surface. Together, this elaborate regulation suggests the importance of CTLA-4 oligomerization for its function.

involves increased partitioning of CTLA-4 to lipid rafts and is dependent on the strength of TCR signaling (100, 102, 103). Allison’s group (88) has shown that the recruitment of CTLA-4 to the IS is highly dependent on B7-1, with little contribution of B7-2. This contrasts with the recruitment of CD28 to the IS that is mostly dependent on B7-2 (88). However, in the absence of both B7-1 and B7-2, there is still low but signiﬁcant relocation of CTLA-4 to the IS (88), suggesting that CTLA-4 may still play a role in the negative regulation of the T cell activation in a B7independent fashion. However, whether this form oligomerizes when located at the IS is unclear.

MTOC: microtubule organizing center

74

Teft

·

Kirchhof

·

Madrenas

INTRACELLULAR TRAFFICKING OF CTLA-4 AND LOCALIZATION AT THE IMMUNOLOGICAL SYNAPSE One remarkable feature of CTLA-4 expression is that most CTLA-4 molecules remain intracellular in vesicles that localize close to the microtubule organizing center (MTOC) (37, 38). Only a small pool is expressed on the T cell surface, in a very dynamic way. This proﬁle of distribution is regulated through an interaction of the cytoplasmic tail of CTLA-4 with AP-1 during its intracellular trafﬁcking as the protein leaves the Golgi (40) and is regulated by phospholipase D and ADP ribosylation factor-1 (104) and, more importantly,

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

Figure 5 Schematic representation of CTLA-4 export pathways to the T cell surface under resting conditions. Upon entering the Golgi, CTLA-4 interacts with AP-1, which mediates its trafﬁc to the cell surface, through a mechanism that may be regulated by phospholipase D (PLD) and ADP ribosylation factor-1 (ARF). CTLA-4 can also be directed to endosomes and lysosomes for storage or degradation. Surface CTLA-4 can interact with AP-2, when its cytoplasmic domain tyrosine residues are not phosphorylated, and this mediates its internalization. Whether it can undergo recycling is still unknown.

through its association with the μ2 subunit of the clathrin-associated adapter protein AP2 (12, 105, 106) (Figure 5). This latter interaction is dependent on the Y165 residue in the cytoplasmic domain, and is regulated by its phosphorylation, which prevents binding to the AP-2 adapter and blocks endocytosis (12, 106). Phosphorylation of this tyrosine residue requires Lck-dependent and ZAP-70-dependent TCR signals and leads to retention of CTLA-4 on the cell surface (107) (Figure 6). The interaction between a T cell and an APC displaying the appropriate antigenic

peptide leads to the formation of a distinct structure at the interface between both cells known as the IS (108–110). The MTOC relocalizes proximal to the IS, and the pool of CTLA-4 also relocates to this T cell:APC interface (102, 111, 112). At this point, surface expression of CTLA-4 is modulated by phosphorylation of the Y165 VKM motif (3, 12, 107), and the strength of TCR signaling dictates CTLA-4 surface expression (102), consistent with the observation that TCR signaling leads to tyrosine phosphorylation of CTLA-4 and its retention on the cell surface (107) (Figure 6). www.annualreviews.org • CTLA- 4 Function

75

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

11:54

Figure 6 Schematic representation of CTLA-4 export pathways and retention at the plasma membrane under conditions of T cell activation. Upon activation of the T cell, CTLA-4 is expressed and its RNA stabilized. The resulting protein is exported to the surface through the Golgi in association with AP-1. In addition, upon TCR signaling surface CTLA-4 is tyrosine phosphorylated (circled P), preventing the binding of AP-2 to its cytoplasmic tail and its internalization. The net result is increased surface CTLA-4 expression and retention. In addition, activation may mediate the recycling of internalized CTLA-4 from endosomes or lysosomes back to the cell surface.

The IS includes an accumulation of membrane microdomains known as lipid rafts, both on T cells and APCs (113–115). Lipid rafts are cholesterol- and glycoprotein-enriched microdomains that are involved in signalosome assembly. Although the majority of CTLA-4 in T cells localizes in the detergent-soluble fraction or nonlipid raft fraction, a portion of CTLA-4 partitions within lipid raft microdomains (103, 116). Upon CTLA-4 activation, the amount of CTLA-4 within the lipid raft fraction increases as CTLA-4 localizes to the IS (103, 116). Such a proﬁle of distribution is necessary but does not seem 76

Teft

·

Kirchhof

·

Madrenas

sufﬁcient for T cell inactivation by CTLA-4. It is important to note that the partitioning of CTLA-4 into lipid rafts depends on its cytoplasmic tail and not on its transmembrane domain because tailless CTLA-4 partitions much less in lipid rafts despite being expressed on the surface at much higher levels (103).

MECHANISMS OF T CELL INACTIVATION BY CTLA-4 T cell inhibition by CTLA-4 is achieved through two mechanisms: competitive antagonism of CD28 signals and direct negative

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

signaling (48). Engagement of CD28 molecules by B7 during early T cell responses initiates a costimulatory signaling cascade that is required for optimal T cell activation through the TCR. Such coordinated signaling leads to reorientation of intracellular CTLA-4 toward the site of TCR engagement, an increase in its surface expression, and clustering within lipid rafts at the IS (102, 112). At the IS, CTLA-4 is available to compete with CD28 for B7 engagement. Because CTLA-4 has greater avidity and afﬁnity for B7-1/B7-2 than does CD28 (48, 94, 117), CTLA-4 will likely sequester B7 molecules, thereby reducing the amount of CD28-dependent costimulation available to the cell. This mechanism does not require the cytoplasmic tail of CTLA-4, as has been shown by different groups (see 48, 118). B7 sequestration at the IS provides a simple explanation for the inhibitory function of CTLA-4 and is fully operational in vitro (48). However, its relevance and sufﬁciency in vivo is still debatable. First, it is contingent on CTLA-4 being expressed at high levels on the T cell surface compared to CD28 (requiring a 3:1 or higher ratio of CTLA-4:CD28) (48), levels not often detected on the surface. But one can argue that such levels could be achieved in relative terms owing to the focal accumulation of CTLA-4 at the IS. A second caveat is that mice expressing CTLA-4 molecules that lack the cytoplasmic tail (and are thus unable to signal) still show a lymphoproliferative phenotype, albeit a much less aggressive one, suggesting that antagonism of CD28 costimulation is not a sufﬁcient mechanism for CTLA-4 function in vivo (48, 119, 120). However, it is important to note that proper partitioning within lipid rafts at the IS requires the cytoplasmic domain of CTLA-4 (103). Thus, tailless CTLA-4 may not be able to properly provide competitive antagonism for CD28 signals. A third point is that the inhibitory function of CTLA-4 still occurs in the presence of nonlimiting CD28 costimulation. Notably, CTLA-4-mediated inactivation of T cells only occurs upon

TCR-CTLA-4 coligation (121). Thus, ligation of the TCR and CTLA-4 must occur in cis (both signals being provided by the same APC). In contrast, CD28-mediated costimulation can be provided in trans, suggesting that sequestration of B7 molecules at the IS may not be sufﬁcient to block costimulatory signals received by the T cell. A fourth point is that CTLA-4 has also been shown to inhibit some T cell responses in the absence of CD28 expression (120, 122). Based on these arguments, we lean toward the idea that antagonism of CD28-mediated costimulation is not the only mechanism of CTLA-4 function, but may be operational in vivo under some circumstances, leading not only to the inhibition of T cell activation but also to the induction of anergy owing to limited costimulation or to regulation of the activity of cyclin-dependent kinase 2 and p27kip1 (123–128). The second mechanism of CTLA-4mediated T cell inactivation is the delivery of an inhibitory signal through its cytoplasmic tail. The evidence for this mechanism derives from observations that antibody cocrosslinking of CTLA-4 and TCR in the presence of nonlimiting CD28 costimulation is sufﬁcient to induce cell-cycle arrest and inhibition of IL-2 (107, 129). In contrast to antagonism of CD28 costimulation, this mechanism is operational when CTLA-4 is expressed at low levels on the T cell surface. CTLA-4-mediated signaling is dependent on the presence of its cytoplasmic tail (48). The precise signals delivered by CTLA4 and the level of action remain uncertain (see 130). Three main models have been proposed: (a) that CTLA-4 inhibits the early events of TCR signaling, (b) that CTLA-4 inhibits events downstream of TCR and CD28 signaling, and (c) that CTLA-4 directly inhibits CD28-dependent signaling. The precise pathways that have been linked to these models are discussed below. Independently of which signaling pathway is used, CTLA4 signaling downregulates cytokine production by inhibiting the accumulation of AP-1, NF-κB, and NFAT in the nucleus of activated www.annualreviews.org • CTLA- 4 Function

77

ANRV270-IY24-03

ARI

13 February 2006

ERK: extracellular signal-regulated kinase SHP: SH2 domain-containing tyrosine phosphatase

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LAT: linker for activation of T cells

78

11:54

T cells (121, 131). CTLA-4 also regulates cell proliferation by inhibiting cyclin D3, cyclindependent kinases 4 and 6, and the degradation of cell-cycle inhibitor p27kip1 (132). In addition to direct effects on TCR/CD28 signaling, two additional processes during T cell activation may directly or indirectly be affected by CTLA-4 signaling. One is the regulation of cell surface antigen receptor and signaling molecule availability through Cbl-b. A recent study has suggested that CD28 and CTLA-4 control T cell activation and proliferation through the regulation of the E3 ubiquitin ligase Cbl-b (133). Cbl-b-knockout mice are prone to autoimmune diseases, demonstrating an important role for this molecule in the regulation of immune responses (134, 135). Interestingly, the effects of CD28 costimulation and CTLA-4 inhibition are uncoupled in Cbl-b-deﬁcient T cells, suggesting that the function of CD28 and CTLA-4 depends on the expression of Cblb. Cbl-b is known to regulate the activity of Vav (134–137). CD28 costimulation leads to the downregulation of Cbl-b at the transcriptional and post-transcriptional levels. B7 ligation of CTLA-4 results in the re-expression of Cbl-b and regulates its expression primarily at the transcriptional level. Further, CTLA4-deﬁcient T cells have signiﬁcantly reduced levels of Cbl-b (133). Therefore, CD28 and CTLA-4 may control the threshold for T cell activation by regulating the expression of Cbl-b. The other function that may be affected by CTLA-4 is cytoskeletal reorganization, impacting many diverse responses such as cell mobility and adherence, by regulating the activity of the small G protein Rap1. CD28 costimulation has been linked to inhibition of Rap1 activity, leading to enhanced levels of extracellular signal-regulated kinase (ERK) activation (138). A recent study (139) has shown that CTLA-4 has a reduced capacity to inhibit T cell responses in mice transgenic for Rap1Gap1, a negative regulator of Rap1. Furthermore, CTLA-4 was shown to activate Rap1, suggesting that CTLA-4 may act to Teft

·

Kirchhof

·

Madrenas

inhibit T cell activation by modulating the CD28-dependent activity of Rap1 (139).

MOLECULAR INTERACTIONS THROUGH THE CYTOPLASMIC TAIL OF CTLA-4: CLUES TO CTLA-4 SIGNALING The cytoplasmic domain of CTLA-4 is unique from other known receptors with inhibitory activity. The tail of CTLA-4 lacks intrinsic enzymatic activity and does not contain a bona ﬁde immune receptor tyrosinebased inhibitory motif (ITIM). However, it contains many other potential protein:protein interacting motifs. The initial studies to identify the molecular interactions of the cytoplasmic domain of CTLA-4 concentrated on two tyrosine residues located at positions 165 and 182 that could provide a putative SH2 domain-binding site. Both Y165 and Y182 can be phosphorylated in vitro by src kinases Lck, fyn, and lyn, as well as by resting lymphocyte kinase (Rlk) and by Janus kinase 2 ( Jak2) (106, 140–143). Such a potential SH2 domain-binding site in addition to the increase in tyrosine phosphorylation observed in CTLA-4-deﬁcient T cells, suggested that an SH2 domain-containing tyrosine phosphatase might mediate CTLA-4 signaling (144). This guided the identiﬁcation of SH2 domain-containing tyrosine phosphatase-2 (SHP-2) as a potential mediator through association to the Y165 VKM motif in the tail of CTLA-4 (105, 145). Phosphorylation of Y165 would recruit SHP-2 and downregulate early TCR signaling through decreasing tyrosine phosphorylation of TCRζ and LAT (linker for activation of T cells) and the Ras regulator p52SHC (140, 142, 145, 146). CTLA-4 crosslinking was also reported to decrease tyrosine phosphorylation of TCRζ and LAT, events that correlated with the formation of a multimolecular complex containing CTLA-4, SHP-2, and Lck (10). Formation of this complex did not require tyrosine phosphorylation of CTLA-4, however. Despite these results, the ability of SHP-2

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

to mediate CTLA-4-dependent T cell inhibition remains controversial. CTLA-4 can inhibit signals from CD3-ε as well as from TCR-ζ (147). Mutation of Y165 or Y182 residues has revealed that neither of these residues is required for CTLA-4-mediated T cell inhibition (107, 148). The inhibitory effects of CTLA-4 are enhanced when the Y165 residue is mutated as a result of higher surface expression owing to its inability to be internalized (107). These results suggest that negative signaling through CTLA-4 is likely mediated through phosphotyrosine-independent interactions. Even when it is phosphorylated, several studies have reported that CTLA4 does not directly associate with SHP-2 or the SH2 domains of SHP-2 (149, 150), and coimmunoprecipitation of SHP-2 with CTLA-4 has not been reproduced in different systems, including models using human T cells (107, 118, 151). In addition, SHP-2 enhances rather than inhibits Ras- and mitogen-activated protein kinase (MAPK)dependent signaling (152). It is important to note, though, that there is consistency in the detection of phosphatase activity associated to the cytoplasmic regions of CTLA-4, although not associated to SHP-2 but rather to SHP-1 (153) or protein phosphatase 2A (PP2A) (9, 154). An alternative body of work points to CTLA-4 signals working downstream of the TCR. This proposal was initially based on the observation of the lack of effect of antibodymediated ligation of CTLA-4 on phosphorylation of TCR-ζ and ZAP-70 (107, 151, 155). These studies concluded that CTLA4 decreases the level of ERK and Jun-NH2terminal kinase ( JNK) activity, leading to inhibition of IL-2 production (155). More recently, using primary murine T cells, Rudd’s group (156) has shown that ligation of CTLA4 with B7-1/B7-2 could differentially regulate members of the MAPK family by inhibiting ERK activity while maintaining JNK activity. Several molecules have been linked to the inhibitory effect of CTLA-4 on ERK-1/-2 ac-

tivation. One is Rap1, whose activation would lead to inhibition of ERK-1/-2 in a Rasindependent manner (157, 158). No information is currently available on the pathway linking CTLA-4 to Rap1. Another molecule is the serine/threonine phosphatase PP2A, reported independently by Thompson’s and Kuchroo’s and our group using yeast twohybrid screening (9, 154). PP2A accounts for close to 1% of all cellular proteins and provides the majority of serine/threonine phosphatase activity within eukaryotic cells (159). PP2A is a trimeric holoenzyme composed of a core enzyme and a regulatory B subunit (PP2AB). The core enzyme is made of a catalytic subunit (PP2AC) and a scaffolding subunit (PP2AA) (159–161). PP2AA is composed of 15 tandem repeats of 39–41 aa that conform to the HEAT (Huntington/elongation/A subunit/TOR) motif. The repeats are composed of two antiparallel helices connected by a short intrarepeat loop; the 15 repeats stack together to form an elongated, left-handed, rod-like architecture that resembles a hook (162). The intrarepeat loops are highly hydrophobic, providing a suitable surface for protein:protein interactions. Mutational analysis of PP2AA identiﬁed that PP2AB binds the intrarepeat loops of HEAT motifs 1–10 in the N terminus, whereas PP2AC binds repeats 11–15 in the C terminus (163, 164). In its interaction with CTLA-4, the A subunit of PP2A binds the lysine-rich motif located at lysine residues 152, 155, and 156 of the juxtamembrane region of the CTLA4 tail, and the C subunit binds the Y165 residue (9, 154). The binding sites for the A and C subunits within the tail of CTLA-4 are 14 aa apart, which correlates exactly with the length of each alpha-helix within PP2AA. Putting this information together, we propose a model of CTLA-4:PP2A interaction in which the CTLA-4 homodimer has a PP2A trimer bound to each tail. PP2AA would bind the lysine-rich motif on the juxtamembrane portion of CTLA-4 and adopt an extended conformation from which PP2AC and PP2AB would hang, the former binding the tyrosine www.annualreviews.org • CTLA- 4 Function

79

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

11:54

Figure 7 A model for the interaction between CTLA-4 and the serine/threonine phosphatase PP2A. The A and C subunits of the trimeric PP2A interact with the cytoplasmic domain of CTLA-4 on regions centered at lysine residues 152, 155, and 156 and tyrosine residue 165. The length of the two amphipatic helices of PP2A is 14 aa, the same distance between the ﬁrst lysine residue and tyrosine residue 165. Numbers of the HEAT domains of PP2AA are indicated. The regulatory PP2AB subunit would be accessible to interact with other molecules. The aa highlighted in yellow represent putative PP2A contact residues with CTLA-4.

165 on that tail and the latter being able to interact with other molecules (Figure 7). What is the functional relevance of the association of PP2A to CTLA-4? Newly synthesized CTLA-4 associates with PP2A, and under these conditions the inhibitory function of CTLA-4 is inactive (9). Upon TCR-CTLA-4 coligation, the A subunit is tyrosine phosphorylated, and PP2A dissociates from CTLA-4 (Figure 8). This correlates with T cell inactivation. Mutations of the lysine-rich motif that render CTLA-4 unable to bind PP2A enhance its inhibitory function, implying that 80

Teft

·

Kirchhof

·

Madrenas

PP2A when associated with CTLA-4 blocks negative signaling through this receptor (9). Interestingly, PP2A is also able to associate with CD28 (154). The association between PP2A and CD28 is also present in resting conditions, and under these conditions PP2A appears to be enzymatically active. Upon T cell activation, CD28 is tyrosine phosphorylated by Lck, and this correlates with its dissociation from PP2A (154). Thus, PP2A seems to act as a negative regulator of the function of membrane receptors to which it associates.

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

Figure 8 Mechanisms of CTLA-4-mediated T cell inactivation: crosstalk between CTLA-4 and CD28. Following T cell activation, CD28 and CTLA-4 are sequentially (t1 and t2 ) phosphorylated in a Lck/ZAP-70-dependent manner by different kinases. Phosphorylated CD28 activates PKB/Akt in a PI3K-dependent manner, leading to downregulation of the E3 ubiquitin ligase Cbl-b and the small G protein Rap1. Engagement of CTLA-4 with B7 correlates with dissociation of PP2A from its tail in a phosphorylation-dependent manner (circled P). This dissociation may cause inhibition of the PKB/Akt signaling pathway and/or enhance the activity of Cbl-b and Rap1. Together, these events may mediate the inhibition of downstream signaling events. The symmetry between CD28 and CTLA-4 signaling interactions suggests that these two receptors may regulate each other through sequestration/blockade of their respective functions.

The mechanism of receptor inhibition by PP2A (e.g., CTLA-4, CD28) and its precise role in CTLA-4-mediated signaling is unknown. PP2A can inactivate protein kinase B (PKB)/Akt (165, 166), a major component of the CD28 signaling pathway dependent on phosphatidylinositol 3-kinase (PI3K) (167–169). Thus, we favor a model in which, when CTLA-4 is expressed but not ligated,

PP2A is bound to CTLA-4 and is not able to inactivate Akt. Under these conditions, CTLA-4 is not inhibitory. This may be relevant in memory T cells, in which CTLA4 is expressed under resting conditions and may control the threshold of activation (25, 170). However, upon ligation of CTLA-4 with the TCR, PP2A would dissociate from CTLA-4 and be free to inactivate Akt, thus www.annualreviews.org • CTLA- 4 Function

81

ARI

13 February 2006

11:54

mediating the inhibitory function of CTLA4 (Figure 8). Consistent with this model, genomic analysis of TCR-CTLA-4 coligation has shown that the effect of CTLA-4mediated T cell inactivation is to downregulate CD28-dependent gene transcription (171). Because PP2A, PI3K, and Grb2 all bind the YMNM motif of CD28, the dissociation and potential inactivation of PP2A may be necessary to induce CD28-mediated costimulation by allowing the binding of PI3K or Grb2. Consistent with this model are the observations that okadaic acid, a potent inhibitor of PP2A, can partially inhibit IL-2 production upon TCR/CD28 ligation, that overexpression of inactive PP2A can increase IL-2 production (154), and that T cell activation through CD28 is enhanced in CTLA-4deﬁcient T cells (172). Therefore, CTLA-4 may inhibit T cell activation by downregulating CD28 signaling. This downregulation may occur through transfer of PP2A from CTLA-4 to CD28 or, alternatively, by direct phosphatase activity of PP2A on downstream steps of CD28 signaling such as Akt. According to the model outlined above, there is a remarkable symmetry in the signaling pathways used by CD28 and by CTLA-4. PI3K associates with the YVKM motif in the cytoplasmic tail of CTLA-4 (11, 141). Also, the SH2 binding domain of the p85 subunit of PI3K associates with a similar YMNM motif in CD28 when phosphorylated (173). The interaction between PI3K and CD28 leads to the recruitment and activation of Akt (PKB), a regulator of many downstream signaling events (reviewed in 174). The ability of both of these receptors to interact with PI3K suggests that CTLA-4 could potentially exchange PP2A and PI3K, sequestering PI3K and thereby reducing its availability to mediate CD28-dependent costimulation. This process may not be the only way this interaction works because CTLA-4 in intracellular vesicles may be phosphorylated by src kinases, allowing association with PI3K (175). The cytoplasmic tail of CTLA-4 also contains a potential docking site for SH3

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

82

Teft

·

Kirchhof

·

Madrenas

domain-containing proteins. This is centered at a proline-rich (PPTEP) motif starting at residue 169. No evidence for a direct SH3 domain-mediated protein interaction has been reported to date. Furthermore, mutation of these proline residues did not affect the inhibitory function of CTLA-4 (48), calling into question a functional role for a conventional interaction with this motif. In summary, the precise nature of the signals transmitted through CTLA-4 remains controversial. Based on the experimental data available, we propose that CTLA-4 and CD28 share a set of mechanisms that regulate themselves and counterregulate each other. Within this framework, the data point to the serine/threonine phosphatase PP2A as the most likely player of this role, inhibiting the function of the receptor to which it is associated through an unknown mechanism, and at the same time enhancing signaling by the other receptor through a Lck/PI3K/Aktdependent manner. Interestingly, current evidence also implies that the mechanism of action of CTLA-4 is distinct from that of the two other negative regulatory receptors reported in T cells: PD-1 (programmed death 1) and BTLA (B and T lymphocyte attenuator) (176, 177; M.L. Baroja, B.M. Carreno, and J. Madrenas, unpublished observations). In contrast to CTLA-4, both PD-1 and BTLA are expressed as a monomer on the cell surface, suggesting that these two receptors are not governed by the same structural constraints as CTLA-4 to exert their inhibitory activities (178), and both of them use SHP-1 and SHP-2 (177, 179, 180).

SIGNALING PLASTICITY OF CTLA-4 Although most of the evidence establishes that CTLA-4 is a key negative regulator of T cell responses, emerging data suggest that it has an inherent functional plasticity. This plasticity is illustrated by the capacity of CTLA-4 to act as a positive regulator or as a negative regulator of T cell activation, depending on the

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

conditions of engagement and on the signals it transduces to the T cell. Such plasticity was implied by the very early experiments showing that CTLA-4 and CD28 could act synergistically to enhance T cell proliferation (59). Later, interactions between B7-1 and CTLA4 were shown to enhance T cell clonal expansion in the absence of CD28 expression (57). The potential of B7-1 acting on a second costimulatory molecule under these conditions was unlikely because intact or Fab CTLA-4 mAbs completely blocked T cell expansion. Additionally, ligation of CTLA-4 with mutant B7-1 that had lost its ability to bind CD28 also resulted in T cell clonal expansion. As suggested by Yang Liu, results based on the use of CTLA-4 blocking Abs may be misinterpreted because these Abs could be agonists rather than antagonists of CTLA-4 function (181). Further evidence for an activating signal through CTLA-4 comes from studies on thymocyte development, in which CTLA4 seemed to contribute to the activation of double-positive thymocytes because CTLA-4 blockade resulted in the inhibition of doublepositive thymocyte deletion (182). In addition, increased surface expression of CTLA-4 in Th2 effector cells renders these cells resistant to activation-induced cell death (183). This seems to occur by recruitment of PI3K and the activation of Akt, leading to the suppression of FasL expression and the induction of the survival factor Bcl-2. These studies further suggest that CTLA-4 may indeed have the ability to induce a positive signal to enhance T cell responses but do not demonstrate CTLA-4 signaling in the absence of other confounding factors. Direct evidence of the signal plasticity of CTLA-4 was obtained with a recombinant ligand of this receptor. This ligand is a bispeciﬁc tandem single chain Fv (ScFv) molecule directed against two different human-speciﬁc epitopes of CTLA-4 recognized by antiCTLA-4 24 and 26 Abs (thus named 24:26) (92). Remarkably, 24:26 by itself activates CTLA-4-expressing primary human T cells.

Although 24:26 effectively blocks the interaction with B7, its effect is independent of B7 and occurs without concomitant TCR signaling, ruling out blockade of CTLA-4 as a mechanism of this paradoxical response. The effects of 24:26-mediated T cell activation are dependent on the expression of the TCR/CD3 complex and Lck and can be further enhanced by CD28 costimulation. The ability of CTLA-4 to deliver an activating signal upon engagement with 24:26 is the result of CTLA-4 triggering a Lck-dependent, TCR-ζ-dependent signaling cascade with phosphorylation of TCRζ and LAT, and activation of ZAP-70, PLC-γ1, ERK-1/-2, and Ca2+ ﬂuxing (92). This signal transduction event is likely triggered by ligand-induced stabilization of the PP2A:CTLA-4 association, which then activates bystander Lck, in contrast to CTLA-4 ligation with B7, which triggers dissociation of PP2A from CTLA-4 (92) (Figure 9). Consistent with this, CTLA-4 mutants that are unable to bind PP2A do not respond to 24:26, whereas okadaic acid blocks the IL-2 response to 24:26 in a dose-dependent manner. The realization of the signaling plasticity of CTLA-4 puts forth a new paradigm on how CTLA-4 can regulate immune responses. According to this paradigm, PP2A is the critical regulator of CTLA-4 function both as an inhibitor and as an activator (9, 92). In conditions of TCR-CTLA-4 coligation, PP2A is phosphorylated and subsequently dissociates from the tail of CTLA-4, leading to T cell inhibition, whereas engagement of CTLA-4 with 24:26 stabilizes the association of PP2A, leading to T cell activation. We are currently investigating the functional importance of ligand-dependent differential association of PP2A to CTLA-4. So far, three factors appear to be important for the signaling plasticity of CTLA-4. First, the bivalency of CTLA-4 homodimers on the cell surface is important for the oligomerization of CTLA-4. Second, the ligand-dependent oligomerization of CTLA-4 seems to play a critical role in the ability of CTLA-4 to www.annualreviews.org • CTLA- 4 Function

83

ARI

13 February 2006

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

11:54

Figure 9 Signaling plasticity of CTLA-4 correlates with its differential association with PP2A. TCR-CD28 coligation results in T cell activation and induction of CTLA-4 expression. Twenty-four to 48 h after activation, CTLA-4 is expressed on the T cell surface associated with PP2A, which blocks its inhibitory function. Upon TCR-CTLA-4 coligation, PP2A is phosphorylated (circled P) and dissociates from CTLA-4. This correlates with CTLA-4 negative signaling and inhibition of T cell activation. Alternatively, CTLA-4 ligation with 24:26, by itself, stabilizes the association between PP2A and CTLA-4, and this correlates with CTLA-4-mediated T cell activation in a Lck-dependent, TCR-ζ-dependent manner.

associate with signaling molecules. Lastly, the differential association of CTLA-4 with PP2A is key in regulating the function of CTLA-4 as either an inhibitor or activator of T cell responses. Preliminary evidence indicates that the ability of CTLA-4 to act as an inhibitor or an activator of T cells seems to correlate with the formation of structurally distinct oligomers on the cell surface in response to engagement with B7 or with this recombinant ligand (W.A. Teft and J. Madrenas, unpublished observations). The existence of an endogenous ligand with analogous function to 84

Teft

·

Kirchhof

·

Madrenas

the response seen with 24:26 is currently under investigation. Moreover, such plasticity is important when considering the development of novel therapeutic agents to enhance immunity (e.g., adjuvants). One should note that the plasticity of CTLA-4 signaling refers to the ability of CTLA-4 to deliver a positive or a negative signal, and it should not be confused with the ability to modulate a polyclonal immune response by CTLA-4 blockade. This blockade can cause immune deviation, can have differential effects on CD4+ T cells versus CD8+

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

T cells, and can affect Tregs. Immune deviation is based on the representation of highafﬁnity T cell clones that would otherwise be inactivated owing to their more rapid upregulation of CTLA-4. Such an effect may translate into immune deviation from Th2 to Th1 responses (73) that can be achieved by direct modulation of cytokine signaling (184). Th1 and Th2 subsets have similar kinetics of CTLA-4 upregulation upon activation and respond similarly to CTLA-4 blockade. However, Th2 cells express signiﬁcantly higher levels of CTLA-4 than do Th1 cells, although the ratio of intracellular to surface expression is constant between both subsets (47, 183). CTLA-4 seems to downregulate predominantly Th2 cell differentiation (185, 186). Therefore, lack of CTLA-4 function correlates with Th2 cell differentiation (187, 188), and blockade of CTLA-4 function resulted in the polarization of the naive CD4+ T cells to Th1 cells (73, 189). One needs to keep in mind that the timing of modulation of CTLA-4 function and the strength of the TCR signal may determine the functional outcome (184, 190–192). For example, CTLA-4 blockade in the presence of strong TCR signals preferentially results in enhanced Th2 immune responses and loss of Th1 responses, whereas T cells receiving weak TCR signals and CTLA-4 blockade can expand both Th1 and Th2 populations (73). Another differential effect of blockade of CTLA-4 can result from differential sensitivity of CD4+ T cells versus CD8+ T cells (193). For example, the effect of CTLA-4 on

CD8+ T cells seems to be of lesser magnitude because CTLA-4−/− , MHC class I–restricted TCR transgenic mice remain healthy for a signiﬁcantly longer time than CTLA-4−/− mice, although they eventually develop a lymphoproliferative disease that appears to be mediated by a small remaining CD4+ T cell subset (65, 194). Furthermore, CD8+ T cells from CTLA-4-deﬁcient mice expressing a transgenic lymphocytic choriomeningitis virus (LCMV)-speciﬁc TCR are able to mount an immune response similar to normal mice upon LCMV infection in vivo (195). Under these types of circumstances involving polyclonal immune responses, CTLA-4 works as a threshold modulator favoring the expansion of some clones over others. Finally, CTLA-4 may also modulate the function of Tregs or APCs through B7. In some studies, in vitro blockade of CTLA-4 on Tregs abrogated the inhibition of proliferation of CTLA-4-deﬁcient responding T cells in a TGF-β-independent manner (196). In contrast, Tregs from CTLA-4-deﬁcient mice were still able to suppress responder T cells in vitro in the presence or absence of CTLA-4 blockade. In human Tregs, CTLA-4 expression has been linked to enhanced suppressor activity and higher expression of FoxP3 compared with CD4+ CD25+ CTLA-4− Tregs, and blockade of CTLA-4 in this subset resulted in a signiﬁcant but incomplete loss of Treg suppressor activity (197). CTLA-4 may also mediate the suppression of T cells through a direct interaction between CTLA4 on Tregs and B7 molecules on DCs to induce tryptophan catabolism (198–200).

SUMMARY POINTS 1. CTLA-4 is an activation-induced homodimeric glycoprotein receptor on T cells and interacts with the B7-1 (CD80)/B7-2 (CD86) ligands on the surface of an APC. 2. When engaged by B7, CTLA-4 plays a key role as a negative regulator of T cell activation, leading to downregulation of T cell responses and to the preservation of T cell homeostasis and peripheral tolerance.

www.annualreviews.org • CTLA- 4 Function

85

ANRV270-IY24-03

ARI

13 February 2006

11:54

3. The mechanism of T cell inactivation by CTLA-4 involves antagonism of CD28dependent costimulation and direct negative signaling through its cytoplasmic tail. Both mechanisms may be operational through a hierarchical regulation of CTLA-4 oligomerization within lipid rafts at the immunological synapse. 4. The pathway involved in CTLA-4 signaling remains uncertain. Emerging data indicate that it is primarily regulated by the serine/threonine phosphatase PP2A acting downstream of early TCR and CD28 signaling, either by direct competition with PI3K or by inhibition of the PKB/Akt pathway.

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

5. CTLA-4 has an intrinsic plasticity for signaling. This concept refers to the ability of CTLA-4 to actively deliver, by itself, a negative signal or a positive signal to a T cell, depending on the ligand it engages and its association/dissociation to PP2A. When engaged by B7, it dissociates from PP2A and acts as an inactivator of T cells. When engaged by recombinant ligands (e.g., 24:26), its association to PP2A is stabilized and it acts as an activator of T cells. Such plasticity may have important implications in the development of immunotherapeutic agents targeting CTLA-4.

ACKNOWLEDGMENTS We thank our colleagues working in the ﬁeld of CTLA-4, in particular Drs. M.L. Baroja, B.M. Carreno, M. Collins, V. Kuchroo, and V. Ling, and the members of the Madrenas laboratory for many very useful discussions. We apologize in advance to those investigators whose work on CTLA-4 could not be explicitly referenced here owing to space limitations. This work was supported by operating grants from the Canadian Institutes of Health Research (CIHR), the Kidney Foundation of Canada, and the London Health Sciences Center Multi-Organ Transplant Program. Wendy A. Teft holds a CIHR doctoral award, Mark G. Kirchhof holds a MD/PhD studentship from CIHR, and Joaqu´ın Madrenas holds a Canada Research Chair in Transplantation and Immunobiology. We dedicate this review article to the memory of Dr. Cynthia A. Chambers.

LITERATURE CITED 1. Brunet JF, Denizot F, Luciani MF, Roux-Dosseto M, Suzan M, et al. 1987. A new member of the immunoglobulin superfamily—CTLA-4. Nature 328:267–70 2. Dariavach P, Mattei MG, Golstein P, Lefranc MP. 1988. Human Ig superfamily CTLA4 gene: chromosomal localization and identity of protein sequence between murine and human CTLA-4 cytoplasmic domains. Eur. J. Immunol. 18:1901–5 3. Lindsten T, Lee KP, Harris ES, Petryniak B, Craighead N, et al. 1993. Characterization of CTLA-4 structure and expression on human T cells. J. Immunol. 151:3489–99 4. Linsley PS, Nadler SG, Bajorath J, Peach R, Leung HT, et al. 1995. Binding stoichiometry of the cytotoxic T lymphocyte-associated molecule-4 (CTLA-4). A disulﬁde-linked homodimer binds two CD86 molecules. J. Biol. Chem. 270:15417–24 5. Ostrov DA, Shi W, Schwartz JC, Almo SC, Nathenson SG. 2000. Structure of murine CTLA-4 and its role in modulating T cell responsiveness. Science 290:816–19 86

Teft

·

Kirchhof

·

Madrenas

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

6. Metzler WJ, Bajorath J, Fenderson W, Shaw SY, Constantine KL, et al. 1997. Solution structure of human CTLA-4 and delineation of a CD80/CD86 binding site conserved in CD28. Nat. Struct. Biol. 4:527–31 7. Darlington PJ, Kirchhof MG, Criado G, Sondhi J, Madrenas J. 2005. Hierarchical regulation of CTLA-4 dimer-based lattice formation and its biological relevance for T cell inactivation. J. Immunol. 175:996–1004 8. Baroja ML, Darlington PJ, Carreno BM, Madrenas J. 2000. Inhibition of T cell activation by CTLA-4: truths and red herrings. Mod. Asp. Immunobiol. 1:169–73 9. Baroja ML, Vijayakrishnan L, Bettelli E, Darlington PJ, Chau TA, et al. 2002. Inhibition of CTLA-4 function by the regulatory subunit of serine/threonine phosphatase 2A. J. Immunol. 168:5070–78 10. Lee KM, Chuang E, Grifﬁn M, Khattri R, Hong DK, et al. 1998. Molecular basis of T cell inactivation by CTLA-4. Science 282:2263–66 11. Schneider H, Prasad KV, Shoelson SE, Rudd CE. 1995. CTLA-4 binding to the lipid kinase phosphatidylinositol 3-kinase in T cells. J. Exp. Med. 181:351–55 12. Shiratori T, Miyatake S, Ohno H, Nakaseko C, Isono K, et al. 1997. Tyrosine phosphorylation controls internalization of CTLA-4 by regulating its interaction with clathrinassociated adaptor complex AP-2. Immunity 6:583–89 13. Chambers CA, Kuhns MS, Egen JG, Allison JP. 2001. CTLA-4-mediated inhibition in regulation of T cell responses: mechanisms and manipulation in tumor immunotherapy. Annu. Rev. Immunol. 19:565–94 14. Salomon B, Bluestone JA. 2001. Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation. Annu. Rev. Immunol. 19:225–52 15. Rudd CE, Martin M, Schneider H. 2002. CTLA-4 negative signaling via lipid rafts: a new perspective. Sci. STKE 2002:PE18 16. Alegre ML, Frauwirth KA, Thompson CB. 2001. T-cell regulation by CD28 and CTLA-4. Nat. Rev. Immunol. 1:220–28 17. Ling V, Wu PW, Finnerty HF, Sharpe AH, Gray GS, Collins M. 1999. Complete sequence determination of the mouse and human CTLA4 gene loci: cross-species DNA sequence similarity beyond exon borders. Genomics 60:341–55 18. Ueda H, Howson JM, Esposito L, Heward J, Snook H, et al. 2003. Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423:506–11 19. Vijayakrishnan L, Slavik JM, Illes Z, Greenwald RJ, Rainbow D, et al. 2004. An autoimmune disease-associated CTLA-4 splice variant lacking the B7 binding domain signals negatively in T cells. Immunity 20:563–75 20. Magistrelli G, Jeannin P, Herbault N, Benoit DC, Gauchat JF, et al. 1999. A soluble form of CTLA-4 generated by alternative splicing is expressed by nonstimulated human T cells. Eur. J. Immunol. 29:3596–602 21. Oaks MK, Hallett KM, Penwell RT, Stauber EC, Warren SJ, Tector AJ. 2000. A native soluble form of CTLA-4. Cell. Immunol. 201:144–53 22. Perkins D, Wang Z, Donovan C, He H, Mark D, et al. 1996. Regulation of CTLA-4 expression during T cell activation. J. Immunol. 156:4154–59 23. Harper K, Balzano C, Rouvier E, Mattei MG, Luciani MF, Golstein P. 1991. CTLA-4 and CD28 activated lymphocyte molecules are closely related in both mouse and human as to sequence, message expression, gene structure, and chromosomal location. J. Immunol. 147:1037–44 24. Takahashi T, Tagami T, Yamazaki S, Uede T, Shimizu J, et al. 2000. Immunologic selftolerance maintained by CD25+ CD4+ regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4. J. Exp. Med. 192:303–10 www.annualreviews.org • CTLA- 4 Function

87

ARI

13 February 2006

11:54

25. Jago CB, Yates J, Camara NO, Lechler RI, Lombardi G. 2004. Differential expression of CTLA-4 among T cell subsets. Clin. Exp. Immunol. 136:463–71 26. Wagner DH Jr, Hagman J, Linsley PS, Hodsdon W, Freed JH, Newell MK. 1996. Rescue of thymocytes from glucocorticoid-induced cell death mediated by CD28/CTLA-4 costimulatory interactions with B7-1/B7-2. J. Exp. Med. 184:1631–38 27. Cilio CM, Daws MR, Malashicheva A, Sentman CL, Holmberg D. 1998. Cytotoxic T lymphocyte antigen 4 is induced in the thymus upon in vivo activation and its blockade prevents anti-CD3-mediated depletion of thymocytes. J. Exp. Med. 188:1239–46 28. Pioli C, Gatta L, Ubaldi V, Doria G. 2000. Inhibition of IgG1 and IgE production by stimulation of the B cell CTLA-4 receptor. J. Immunol. 165:5530–36 29. Pistillo MP, Tazzari PL, Palmisano GL, Pierri I, Bolognesi A, et al. 2003. CTLA-4 is not restricted to the lymphoid cell lineage and can function as a target molecule for apoptosis induction of leukemic cells. Blood 101:202–9 30. Kaufman KA, Bowen JA, Tsai AF, Bluestone JA, Hunt JS, Ober C. 1999. The CTLA-4 gene is expressed in placental ﬁbroblasts. Mol. Hum. Reprod. 5:84–87 31. Wang XB, Giscombe R, Yan Z, Heiden T, Xu D, Lefvert AK. 2002. Expression of CTLA-4 by human monocytes. Scand. J. Immunol. 55:53–60 32. Ling V, Munroe RC, Murphy EA, Gray GS. 1998. Embryonic stem cells and embryoid bodies express lymphocyte costimulatory molecules. Exp. Cell Res. 241:55–65 33. Miller RE, Fayen JD, Mohammad SF, Stein K, Kadereit S, et al. 2002. Reduced CTLA4 protein and messenger RNA expression in umbilical cord blood T lymphocytes. Exp. Hematol. 30:738–44 34. Finn PW, He H, Wang Y, Wang Z, Guan G, et al. 1997. Synergistic induction of CTLA-4 expression by costimulation with TCR plus CD28 signals mediated by increased transcription and messenger ribonucleic acid stability. J. Immunol. 158:4074–81 35. Vendetti S, Riccomi A, Sacchi A, Gatta L, Pioli C, De Magistris MT. 2002. Cyclic adenosine 5’-monophosphate and calcium induce CD152 (CTLA-4) up-regulation in resting CD4+ T lymphocytes. J. Immunol. 169:6231–35 36. Shaw G, Kamen R. 1986. A conserved AU sequence from the 3 untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46:659–67 37. Wang XB, Zheng CY, Giscombe R, Lefvert AK. 2001. Regulation of surface and intracellular expression of CTLA-4 on human peripheral T cells. Scand. J. Immunol. 54:453–58 38. Alegre ML, Noel PJ, Eisfelder BJ, Chuang E, Clark MR, et al. 1996. Regulation of surface and intracellular expression of CTLA4 on mouse T cells. J. Immunol. 157:4762–70 39. Chuang E, Alegre ML, Duckett CS, Noel PJ, Vander Heiden MG, Thompson CB. 1997. Interaction of CTLA-4 with the clathrin-associated protein AP50 results in ligandindependent endocytosis that limits cell surface expression. J. Immunol. 159:144–51 40. Schneider H, Martin M, Agarraberes FA, Yin L, Rapoport I, et al. 1999. Cytolytic T lymphocyte-associated antigen-4 and the TCRζ/CD3 complex, but not CD28, interact with clathrin adaptor complexes AP-1 and AP-2. J. Immunol. 163:1868–79 41. Iida T, Ohno H, Nakaseko C, Sakuma M, Takeda-Ezaki M, et al. 2000. Regulation of cell surface expression of CTLA-4 by secretion of CTLA-4-containing lysosomes upon activation of CD4+ T cells. J. Immunol. 165:5062–68 42. Lafage-Pochitaloff M, Costello R, Couez D, Simonetti J, Mannoni P, et al. 1990. Human CD28 and CTLA-4 Ig superfamily genes are located on chromosome 2 at bands q33-q34. Immunogenetics 31:198–201 43. Carreno BM, Collins M. 2002. The B7 family of ligands and its receptors: new pathways for costimulation and inhibition of immune responses. Annu. Rev. Immunol. 20:29–53

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

88

Teft

·

Kirchhof

·

Madrenas

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

44. Howard TA, Rochelle JM, Seldin MF. 1991. CD28 and CTLA-4, two related members of the Ig supergene family, are tightly linked on proximal mouse chromosome 1. Immunogenetics 33:74–76 45. Choi IS, Hash SM, Collisson EW. 2000. Molecular cloning and expression of feline CD28 and CTLA-4 cDNA. Vet. Immunol. Immunopathol. 76:45–59 46. Palmisano GL, Tazzari PL, Cozzi E, Bolognesi A, Polito L, et al. 2004. Expression of CTLA-4 in nonhuman primate lymphocytes and its use as a potential target for speciﬁc immunotoxin-mediated apoptosis: results of in vitro studies. Clin. Exp. Immunol. 135:259– 66 47. Alegre ML, Shiels H, Thompson CB, Gajewski TF. 1998. Expression and function of CTLA-4 in Th1 and Th2 cells. J. Immunol. 161:3347–56 48. Carreno BM, Bennett F, Chau TA, Ling V, Luxenberg D, et al. 2000. CTLA-4 (CD152) can inhibit T cell activation by two different mechanisms depending on its level of cell surface expression. J. Immunol. 165:1352–56 49. Nistico L, Buzzetti R, Pritchard LE, Van der Auwera B, Giovannini C, et al. 1996. The CTLA-4 gene region of chromosome 2q33 is linked to, and associated with, type 1 diabetes. Belgian Diabetes Registry. Hum. Mol. Genet. 5:1075–80 50. Deichmann K, Heinzmann A, Bruggenolte E, Forster J, Kuehr J. 1996. An Mse I RFLP in the human CTLA4 promotor. Biochem. Biophys. Res. Commun. 225:817–18 51. Donner H, Rau H, Walﬁsh PG, Braun J, Siegmund T, et al. 1997. CTLA4 alanine-17 confers genetic susceptibility to Graves’ disease and to type 1 diabetes mellitus. J. Clin. Endocrinol. Metab. 82:143–46 52. Wang XB, Zhao X, Giscombe R, Lefvert AK. 2002. A CTLA-4 gene polymorphism at position -318 in the promoter region affects the expression of protein. Genes Immun. 3:233–34 53. Kristiansen OP, Larsen ZM, Pociot F. 2000. CTLA-4 in autoimmune diseases—a general susceptibility gene to autoimmunity? Genes Immun. 1:170–84 54. Anjos S, Nguyen A, Ounissi-Benkalha H, Tessier MC, Polychronakos C. 2002. A common autoimmunity predisposing signal peptide variant of the cytotoxic T-lymphocyte antigen 4 results in inefﬁcient glycosylation of the susceptibility allele. J. Biol. Chem. 277:46478–86 55. Huang D, Giscombe R, Zhou Y, Pirskanen R, Lefvert AK. 2000. Dinucleotide repeat expansion in the CTLA-4 gene leads to T cell hyper-reactivity via the CD28 pathway in myasthenia gravis. J. Neuroimmunol. 105:69–77 56. Wang XB, Kakoulidou M, Giscombe R, Qiu Q, Huang D, et al. 2002. Abnormal expression of CTLA-4 by T cells from patients with myasthenia gravis: effect of an AT-rich gene sequence. J. Neuroimmunol. 130:224–32 57. Wu Y, Guo Y, Huang A, Zheng P, Liu Y. 1997. CTLA-4-B7 interaction is sufﬁcient to costimulate T cell clonal expansion. J. Exp. Med. 185:1327–35 58. Damle NK, Klussman K, Leytze G, Myrdal S, Aruffo A, et al. 1994. Costimulation of T lymphocytes with integrin ligands intercellular adhesion molecule-1 or vascular cell adhesion molecule-1 induces functional expression of CTLA-4, a second receptor for B7. J. Immunol. 152:2686–97 59. Linsley PS, Greene JL, Tan P, Bradshaw J, Ledbetter JA, et al. 1992. Coexpression and functional cooperation of CTLA-4 and CD28 on activated T lymphocytes. J. Exp. Med. 176:1595–604 60. Walunas TL, Lenschow DJ, Bakker CY, Linsley PS, Freeman GJ, et al. 1994. CTLA-4 can function as a negative regulator of T cell activation. Immunity 1:405–13 61. Krummel MF, Allison JP. 1995. CD28 and CTLA-4 have opposing effects on the response of T cells to stimulation. J. Exp. Med. 182:459–65 www.annualreviews.org • CTLA- 4 Function

89

ARI

13 February 2006

11:54

62. Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA, Sharpe AH. 1995. Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3:541–47 63. Waterhouse P, Penninger JM, Timms E, Wakeham A, Shahinian A, et al. 1995. Lymphoproliferative disorders with early lethality in mice deﬁcient in CTLA-4. Science 270:985– 88 64. Chambers CA, Cado D, Truong T, Allison JP. 1997. Thymocyte development is normal in CTLA-4-deﬁcient mice. Proc. Natl. Acad. Sci. USA 94:9296–301 65. Waterhouse P, Bachmann MF, Penninger JM, Ohashi PS, Mak TW. 1997. Normal thymic selection, normal viability and decreased lymphoproliferation in T cell receptor-transgenic CTLA-4-deﬁcient mice. Eur. J. Immunol. 27:1887–92 66. Takahashi S, Kataoka H, Hara S, Yokosuka T, Takase K, et al. 2005. In vivo overexpression of CTLA-4 suppresses lymphoproliferative diseases and thymic negative selection. Eur. J. Immunol. 35:399–407 67. Allison JP, Krummel MF. 1995. The yin and yang of T cell costimulation. Science 270:932– 33 68. Linsley PS, Brady W, Urnes M, Grosmaire LS, Damle NK, Ledbetter JA. 1991. CTLA-4 is a second receptor for the B cell activation antigen B7. J. Exp. Med. 174:561–69 69. Tivol EA, Boyd SD, McKeon S, Borriello F, Nickerson P, et al. 1997. CTLA4Ig prevents lymphoproliferation and fatal multiorgan tissue destruction in CTLA-4-deﬁcient mice. J. Immunol. 158:5091–94 70. Bachmann MF, Kohler G, Ecabert B, Mak TW, Kopf M. 1999. Cutting edge: Lymphoproliferative disease in the absence of CTLA-4 is not T cell autonomous. J. Immunol. 163:1128–31 71. Tivol EA, Gorski J. 2002. Re-establishing peripheral tolerance in the absence of CTLA-4: complementation by wild-type T cells points to an indirect role for CTLA-4. J. Immunol. 169:1852–58 72. Bachmann MF, Gallimore A, Jones E, Ecabert B, Acha-Orbea H, Kopf M. 2001. Normal pathogen-speciﬁc immune responses mounted by CTLA-4-deﬁcient T cells: a paradigm reconsidered. Eur. J. Immunol. 31:450–58 73. Anderson DE, Bieganowska KD, Bar-Or A, Oliveira EM, Carreno B, et al. 2000. Paradoxical inhibition of T-cell function in response to CTLA-4 blockade; heterogeneity within the human T-cell population. Nat. Med. 6:211–14 74. Linsley PS, Clark EA, Ledbetter JA. 1990. T-cell antigen CD28 mediates adhesion with B cells by interacting with activation antigen B7/BB-1. Proc. Natl. Acad. Sci. USA 87:5031– 35 75. Freeman GJ, Borriello F, Hodes RJ, Reiser H, Gribben JG, et al. 1993. Murine B7-2, an alternative CTLA4 counter-receptor that costimulates T cell proliferation and interleukin 2 production. J. Exp. Med. 178:2185–92 76. Freeman GJ, Gribben JG, Boussiotis VA, Ng JW, Restivo VA Jr, et al. 1993. Cloning of B7-2: a CTLA-4 counter-receptor that costimulates human T cell proliferation. Science 262:909–11 77. Linsley PS, Greene JL, Brady W, Bajorath J, Ledbetter JA, Peach R. 1994. Human B7-1 (CD80) and B7-2 (CD86) bind with similar avidities but distinct kinetics to CD28 and CTLA-4 receptors. Immunity 1:793–801 78. Collins AV, Brodie DW, Gilbert RJ, Iaboni A, Manso-Sancho R, et al. 2002. The interaction properties of costimulatory molecules revisited. Immunity 17:201–10

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

90

Teft

·

Kirchhof

·

Madrenas

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

79. van der Merwe PA, Bodian DL, Daenke S, Linsley P, Davis SJ. 1997. CD80 (B7-1) binds both CD28 and CTLA-4 with a low afﬁnity and very fast kinetics. J. Exp. Med. 185:393– 403 80. van der Merwe PA, Davis SJ. 2003. Molecular interactions mediating T cell antigen recognition. Annu. Rev. Immunol. 21:659–84 81. Stamper CC, Zhang Y, Tobin JF, Erbe DV, Ikemizu S, et al. 2001. Crystal structure of the B7-1/CTLA-4 complex that inhibits human immune responses. Nature 410:608–11 82. Zhang X, Schwartz JC, Almo SC, Nathenson SG. 2003. Crystal structure of the receptorbinding domain of human B7-2: insights into organization and signaling. Proc. Natl. Acad. Sci. USA 100:2586–91 83. Balzano C, Buonavista N, Rouvier E, Golstein P. 1992. CTLA-4 and CD28: similar proteins, neighbouring genes. Int. J. Cancer Suppl. 7:28–32 84. Lenschow DJ, Su GH, Zuckerman LA, Nabavi N, Jellis CL, et al. 1993. Expression and functional signiﬁcance of an additional ligand for CTLA-4. Proc. Natl. Acad. Sci. USA 90:11054–58 85. Hathcock KS, Laszlo G, Pucillo C, Linsley P, Hodes RJ. 1994. Comparative analysis of B7-1 and B7-2 costimulatory ligands: expression and function. J. Exp. Med. 180:631–40 86. Lenschow DJ, Walunas TL, Bluestone JA. 1996. CD28/B7 system of T cell costimulation. Annu. Rev. Immunol. 14:233–58 87. Evans EJ, Esnouf RM, Manso-Sancho R, Gilbert RJ, James JR, et al. 2005. Crystal structure of a soluble CD28-Fab complex. Nat. Immunol. 6:271–79 88. Pentcheva-Hoang T, Egen JG, Wojnoonski K, Allison JP. 2004. B7-1 and B7-2 selectively recruit CTLA-4 and CD28 to the immunological synapse. Immunity 21:401–13 89. Kuchroo VK, Das MP, Brown JA, Ranger AM, Zamvil SS, et al. 1995. B7-1 and B7-2 costimulatory molecules activate differentially the Th1/Th2 developmental pathways: application to autoimmune disease therapy. Cell 80:707–18 90. Chen L. 2004. Co-inhibitory molecules of the B7-CD28 family in the control of T-cell immunity. Nat. Rev. Immunol. 4:336–47 91. Chikuma S, Abbas AK, Bluestone JA. 2005. B7-independent inhibition of T cells by CTLA-4. J. Immunol. 175:177–81 92. Madrenas J, Chau LA, Teft WA, Wu PW, Jussif J, et al. 2004. Conversion of CTLA-4 from inhibitor to activator of T cells with a bispeciﬁc tandem single-chain Fv ligand. J. Immunol. 172:5948–56 93. Mukherjee S, Ahmed A, Nandi D. 2005. CTLA4-CD80/CD86 interactions on primary mouse CD4+ T cells integrate signal-strength information to modulate activation with Concanavalin A. J. Leukoc. Biol. 78:144–57 94. Greene JL, Leytze GM, Emswiler J, Peach R, Bajorath J, et al. 1996. Covalent dimerization of CD28/CTLA-4 and oligomerization of CD80/CD86 regulate T cell costimulatory interactions. J. Biol. Chem. 271:26762–71 95. Schwartz JC, Zhang X, Fedorov AA, Nathenson SG, Almo SC. 2001. Structural basis for co-stimulation by the human CTLA-4/B7-2 complex. Nature 410:604–8 96. Schwartz JC, Zhang X, Nathenson SG, Almo SC. 2002. Structural mechanisms of costimulation. Nat. Immunol. 3:427–34 97. Chun T, Choi HJ, Chung YH. 2004. Two different forms of human CTLA-4 proteins following peripheral T cell activation. Immunol. Lett. 91:213–20 98. Luo RZ, Beniac DR, Fernandes A, Yip CC, Ottensmeyer FP. 1999. Quaternary structure of the insulin-insulin receptor complex. Science 285:1077–80 www.annualreviews.org • CTLA- 4 Function

91

ARI

13 February 2006

11:54

99. Cunningham BC, Ultsch M, De Vos AM, Mulkerrin MG, Clauser KR, Wells JA. 1991. Dimerization of the extracellular domain of the human growth hormone receptor by a single hormone molecule. Science 254:821–25 100. Baroja ML, Madrenas J. 2003. Viewpoint: therapeutic implications of CTLA-4 compartmentalization. Am. J. Transplant. 3:919–26 101. Chikuma S, Bluestone JA. 2003. CTLA-4 and tolerance: the biochemical point of view. Immunol. Res. 28:241–53 102. Egen JG, Allison JP. 2002. Cytotoxic T lymphocyte antigen-4 accumulation in the immunological synapse is regulated by TCR signal strength. Immunity 16:23–35 103. Darlington PJ, Baroja ML, Chau TA, Siu E, Ling V, et al. 2002. Surface cytotoxic T lymphocyte-associated antigen 4 partitions within lipid rafts and relocates to the immunological synapse under conditions of inhibition of T cell activation. J. Exp. Med. 195:1337–47 104. Mead KI, Zheng Y, Manzotti CN, Perry LC, Liu MK, et al. 2005. Exocytosis of CTLA-4 is dependent on phospholipase D and ADP ribosylation factor-1 and stimulated during activation of regulatory T cells. J. Immunol. 174:4803–11 105. Zhang Y, Allison JP. 1997. Interaction of CTLA-4 with AP50, a clathrin-coated pit adaptor protein. Proc. Natl. Acad. Sci. USA 94:9273–78 106. Bradshaw JD, Lu P, Leytze G, Rodgers J, Schieven GL, et al. 1997. Interaction of the cytoplasmic tail of CTLA-4 (CD152) with a clathrin-associated protein is negatively regulated by tyrosine phosphorylation. Biochemistry 36:15975–82 107. Baroja ML, Luxenberg D, Chau T, Ling V, Strathdee CA, et al. 2000. The inhibitory function of CTLA-4 does not require its tyrosine phosphorylation. J. Immunol. 164:49–55 108. Davis DM, Dustin ML. 2004. What is the importance of the immunological synapse? Trends Immunol. 25:323–27 109. Jacobelli J, Andres PG, Boisvert J, Krummel MF. 2004. New views of the immunological synapse: variations in assembly and function. Curr. Opin. Immunol. 16:345–52 110. Dustin ML. 2003. Coordination of T cell activation and migration through formation of the immunological synapse. Ann. NY Acad. Sci. 987:51–59 111. Kupfer A, Swain SL, Singer SJ. 1987. The speciﬁc direct interaction of helper T cells and antigen-presenting B cells. II. Reorientation of the microtubule organizing center and reorganization of the membrane-associated cytoskeleton inside the bound helper T cells. J. Exp. Med. 165:1565–80 112. Linsley PS, Bradshaw J, Greene J, Peach R, Bennett KL, Mittler RS. 1996. Intracellular trafﬁcking of CTLA-4 and focal localization towards sites of TCR engagement. Immunity 4:535–43 113. Taner SB, Onfelt B, Pirinen NJ, McCann FE, Magee AI, Davis DM. 2004. Control of immune responses by trafﬁcking cell surface proteins, vesicles and lipid rafts to and from the immunological synapse. Trafﬁc 5:651–61 114. Vogt AB, Spindeldreher S, Kropshofer H. 2002. Clustering of MHC-peptide complexes prior to their engagement in the immunological synapse: lipid raft and tetraspan microdomains. Immunol. Rev. 189:136–51 115. Miceli MC, Moran M, Chung CD, Patel VP, Low T, Zinnanti W. 2001. Co-stimulation and counter-stimulation: lipid raft clustering controls TCR signaling and functional outcomes. Semin. Immunol. 13:115–28 116. Chikuma S, Imboden JB, Bluestone JA. 2003. Negative regulation of T cell receptor-lipid raft interaction by cytotoxic T lymphocyte-associated antigen 4. J. Exp. Med. 197:129–35

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

92

Teft

·

Kirchhof

·

Madrenas

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

117. Leung HT, Bradshaw J, Cleaveland JS, Linsley PS. 1995. Cytotoxic T lymphocyteassociated molecule-4, a high-avidity receptor for CD80 and CD86, contains an intracellular localization motif in its cytoplasmic tail. J. Biol. Chem. 270:25107–14 118. Cinek T, Sadra A, Imboden JB. 2000. Cutting edge: tyrosine-independent transmission of inhibitory signals by CTLA-4. J. Immunol. 164:5–8 119. Masteller EL, Chuang E, Mullen AC, Reiner SL, Thompson CB. 2000. Structural analysis of CTLA-4 function in vivo. J. Immunol. 164:5319–27 120. Fallarino F, Fields PE, Gajewski TF. 1998. B7-1 engagement of cytotoxic T lymphocyte antigen 4 inhibits T cell activation in the absence of CD28. J. Exp. Med. 188:205–10 121. Fraser JH, Rincon M, McCoy KD, Le Gros G. 1999. CTLA4 ligation attenuates AP-1, NFAT and NF-κB activity in activated T cells. Eur. J. Immunol. 29:838–44 122. Lin H, Rathmell JC, Gray GS, Thompson CB, Leiden JM, Alegre ML. 1998. Cytotoxic T lymphocyte antigen 4 (CTLA4) blockade accelerates the acute rejection of cardiac allografts in CD28-deﬁcient mice: CTLA4 can function independently of CD28. J. Exp. Med. 188:199–204 123. Greenwald RJ, Boussiotis VA, Lorsbach RB, Abbas AK, Sharpe AH. 2001. CTLA-4 regulates induction of anergy in vivo. Immunity 14:145–55 124. Wells AD, Walsh MC, Bluestone JA, Turka LA. 2001. Signaling through CD28 and CTLA-4 controls two distinct forms of T cell anergy. J. Clin. Invest. 108:895–903 125. Vanasek TL, Khoruts A, Zell T, Mueller DL. 2001. Antagonistic roles for CTLA-4 and the mammalian target of rapamycin in the regulation of clonal anergy: enhanced cell cycle progression promotes recall antigen responsiveness. J. Immunol. 167:5636–44 126. Eagar TN, Karandikar NJ, Bluestone JA, Miller SD. 2002. The role of CTLA-4 in induction and maintenance of peripheral T cell tolerance. Eur. J. Immunol. 32:972–81 127. Eagar TN, Turley DM, Padilla J, Karandikar NJ, Tan L, et al. 2004. CTLA-4 regulates expansion and differentiation of Th1 cells following induction of peripheral T cell tolerance. J. Immunol. 172:7442–50 128. Schwartz RH. 2003. T cell anergy. Annu. Rev. Immunol. 21:305–34 129. Krummel MF, Allison JP. 1996. CTLA-4 engagement inhibits IL-2 accumulation and cell cycle progression upon activation of resting T cells. J. Exp. Med. 183:2533–40 130. Rudd CE, Schneider H. 2003. Unifying concepts in CD28, ICOS and CTLA4 co-receptor signalling. Nat. Rev. Immunol.. 3:544–56 131. Olsson C, Riesbeck K, Dohlsten M, Michaelsson E. 1999. CTLA-4 ligation suppresses CD28-induced NF-κB and AP-1 activity in mouse T cell blasts. J. Biol. Chem. 274:14400– 5 132. Brunner MC, Chambers CA, Chan FK, Hanke J, Winoto A, Allison JP. 1999. CTLA-4mediated inhibition of early events of T cell proliferation. J. Immunol. 162:5813–20 133. Li D, Gal I, Vermes C, Alegre ML, Chong AS, et al. 2004. Cutting edge: Cbl-b: one of the key molecules tuning CD28- and CTLA-4-mediated T cell costimulation. J. Immunol. 173:7135–39 134. Bachmaier K, Krawczyk C, Kozieradzki I, Kong YY, Sasaki T, et al. 2000. Negative regulation of lymphocyte activation and autoimmunity by the molecular adaptor Cbl-b. Nature 403:211–16 135. Chiang YJ, Kole HK, Brown K, Naramura M, Fukuhara S, et al. 2000. Cbl-b regulates the CD28 dependence of T-cell activation. Nature 403:216–20 136. Michel F, Mangino G, Attal-Bonnefoy G, Tuosto L, Alcover A, et al. 2000. CD28 utilizes Vav-1 to enhance TCR-proximal signaling and NF-AT activation. J. Immunol. 165:3820– 29 www.annualreviews.org • CTLA- 4 Function

93

ARI

13 February 2006

11:54

137. Salojin KV, Zhang J, Delovitch TL. 1999. TCR and CD28 are coupled via ZAP-70 to the activation of the Vav/Rac-1-/PAK-1/p38 MAPK signaling pathway. J. Immunol. 163:844– 53 138. Carey KD, Dillon TJ, Schmitt JM, Baird AM, Holdorf AD, et al. 2000. CD28 and the tyrosine kinase lck stimulate mitogen-activated protein kinase activity in T cells via inhibition of the small G protein Rap1. Mol. Cell. Biol. 20:8409–19 139. Dillon TJ, Carey KD, Wetzel SA, Parker DC, Stork PJ. 2005. Regulation of the small GTPase Rap1 and extracellular signal-regulated kinases by the costimulatory molecule CTLA-4. Mol. Cell. Biol. 25:4117–28 140. Miyatake S, Nakaseko C, Umemori H, Yamamoto T, Saito T. 1998. Src family tyrosine kinases associate with and phosphorylate CTLA-4 (CD152). Biochem. Biophys. Res. Commun. 249:444–48 141. Schneider H, Schwartzberg PL, Rudd CE. 1998. Resting lymphocyte kinase (Rlk/Txk) phosphorylates the YVKM motif and regulates PI 3-kinase binding to T-cell antigen CTLA-4. Biochem. Biophys. Res. Commun. 252:14–19 142. Chuang E, Lee KM, Robbins MD, Duerr JM, Alegre ML, et al. 1999. Regulation of cytotoxic T lymphocyte-associated molecule-4 by Src kinases. J. Immunol. 162:1270–77 143. Chikuma S, Murakami M, Tanaka K, Uede T. 2000. Janus kinase 2 is associated with a box 1-like motif and phosphorylates a critical tyrosine residue in the cytoplasmic region of cytotoxic T lymphocyte associated molecule-4. J. Cell. Biochem. 78:241–50 144. Chambers CA, Allison JP. 1996. The role of tyrosine phosphorylation and PTP-1C in CTLA-4 signal transduction. Eur. J. Immunol. 26:3224–29 145. Marengere LE, Waterhouse P, Duncan GS, Mittrucker HW, Feng GS, Mak TW. 1996. Regulation of T cell receptor signaling by tyrosine phosphatase SYP association with CTLA-4. Science 272:1170–73 146. Yi LA, Hajialiasgar S, Chuang E. 2004. Tyrosine-mediated inhibitory signals contribute to CTLA-4 function in vivo. Int. Immunol. 16:539–47 147. Siu E, Carreno BM, Madrenas J. 2003. TCR subunit speciﬁcity of CTLA-4-mediated signaling. J. Leukoc. Biol. 74:1102–7 148. Nakaseko C, Miyatake S, Iida T, Hara S, Abe R, et al. 1999. Cytotoxic T lymphocyte antigen 4 (CTLA-4) engagement delivers an inhibitory signal through the membraneproximal region in the absence of the tyrosine motif in the cytoplasmic tail. J. Exp. Med. 190:765–74 149. Tonks NK, Neel BG. 1996. From form to function: signaling by protein tyrosine phosphatases. Cell 87:365–68 150. Schneider H, Rudd CE. 2000. Tyrosine phosphatase SHP-2 binding to CTLA-4: absence of direct YVKM/YFIP motif recognition. Biochem. Biophys. Res. Commun. 269:279–83 151. Schneider H, da Rocha Dias S, Hu H, Rudd CE. 2001. A regulatory role for cytoplasmic YVKM motif in CTLA-4 inhibition of TCR signaling. Eur. J. Immunol. 31:2042–50 152. Frearson JA, Alexander DR. 1998. The phosphotyrosine phosphatase SHP-2 participates in a multimeric signaling complex and regulates T cell receptor (TCR) coupling to the Ras/mitogen-activated protein kinase (MAPK) pathway in Jurkat T cells. J. Exp. Med. 187:1417–26 153. Guntermann C, Alexander DR. 2002. CTLA-4 suppresses proximal TCR signaling in resting human CD4+ T cells by inhibiting ZAP-70 Tyr319 phosphorylation: a potential role for tyrosine phosphatases. J. Immunol. 168:4420–29 154. Chuang E, Fisher TS, Morgan RW, Robbins MD, Duerr JM, et al. 2000. The CD28 and CTLA-4 receptors associate with the serine/threonine phosphatase PP2A. Immunity 13:313–22

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

94

Teft

·

Kirchhof

·

Madrenas

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

155. Calvo CR, Amsen D, Kruisbeek AM. 1997. Cytotoxic T lymphocyte antigen 4 (CTLA-4) interferes with extracellular signal-regulated kinase (ERK) and Jun NH2-terminal kinase (JNK) activation, but does not affect phosphorylation of T cell receptor ζ and ZAP70. J. Exp. Med. 186:1645–53 156. Schneider H, Mandelbrot DA, Greenwald RJ, Ng F, Lechler R, et al. 2002. Cutting edge: CTLA-4 (CD152) differentially regulates mitogen-activated protein kinases (extracellular signal-regulated kinase and c-Jun N-terminal kinase) in CD4+ T cells from receptor/ligand-deﬁcient mice. J. Immunol. 169:3475–79 157. Bos JL, de Bruyn K, Enserink J, Kuiperij B, Rangarajan S, et al. 2003. The role of Rap1 in integrin-mediated cell adhesion. Biochem. Soc. Trans. 31:83–86 158. Stork PJ, Schmitt JM. 2002. Crosstalk between cAMP and MAP kinase signaling in the regulation of cell proliferation. Trends Cell Biol. 12:258–66 159. Sontag E. 2001. Protein phosphatase 2A: the Trojan Horse of cellular signaling. Cell. Signal. 13:7–16 160. Lechward K, Awotunde OS, Swiatek W, Muszynska G. 2001. Protein phosphatase 2A: variety of forms and diversity of functions. Acta Biochim. Pol. 48:921–33 161. Janssens V, Goris J. 2001. Protein phosphatase 2A: a highly regulated family of serine/threonine phosphatases implicated in cell growth and signalling. Biochem. J. 353:417– 39 162. Groves MR, Hanlon N, Turowski P, Hemmings BA, Barford D. 1999. The structure of the protein phosphatase 2A PR65/A subunit reveals the conformation of its 15 tandemly repeated HEAT motifs. Cell 96:99–110 163. Ruediger R, Hentz M, Fait J, Mumby M, Walter G. 1994. Molecular model of the A subunit of protein phosphatase 2A: interaction with other subunits and tumor antigens. J. Virol. 68:123–29 164. Kremmer E, Ohst K, Kiefer J, Brewis N, Walter G. 1997. Separation of PP2A core enzyme and holoenzyme with monoclonal antibodies against the regulatory A subunit: abundant expression of both forms in cells. Mol. Cell. Biol. 17:1692–701 165. Cazzolli R, Carpenter L, Biden TJ, Schmitz-Peiffer C. 2001. A role for protein phosphatase 2A-like activity, but not atypical protein kinase Cζ, in the inhibition of protein kinase B/Akt and glycogen synthesis by palmitate. Diabetes 50:2210–18 166. Resjo S, Goransson O, Harndahl L, Zolnierowicz S, Manganiello V, Degerman E. 2002. Protein phosphatase 2A is the main phosphatase involved in the regulation of protein kinase B in rat adipocytes. Cell. Signal. 14:231–38 167. Kane LP, Andres PG, Howland KC, Abbas AK, Weiss A. 2001. Akt provides the CD28 costimulatory signal for up-regulation of IL-2 and IFN-γ but not TH2 cytokines. Nat. Immunol. 2:37–44 168. Cantrell D. 2002. Protein kinase B (Akt) regulation and function in T lymphocytes. Semin. Immunol. 14:19–26 169. Appleman LJ, van Puijenbroek AA, Shu KM, Nadler LM, Boussiotis VA. 2002. CD28 costimulation mediates down-regulation of p27kip1 and cell cycle progression by activation of the PI3K/PKB signaling pathway in primary human T cells. J. Immunol. 168:2729–36 170. Metz DP, Farber DL, Taylor T, Bottomly K. 1998. Differential role of CTLA-4 in regulation of resting memory versus naive CD4 T cell activation. J. Immunol. 161:5855– 61 171. Riley JL, Mao M, Kobayashi S, Biery M, Burchard J, et al. 2002. Modulation of TCRinduced transcriptional proﬁles by ligation of CD28, ICOS, and CTLA-4 receptors. Proc. Natl. Acad. Sci. USA 99:11790–95 www.annualreviews.org • CTLA- 4 Function

95

ARI

13 February 2006

11:54

172. Mandelbrot DA, McAdam AJ, Sharpe AH. 1999. B7-1 or B7-2 is required to produce the lymphoproliferative phenotype in mice lacking cytotoxic T lymphocyte-associated antigen 4 (CTLA-4). J. Exp. Med. 189:435–40 173. Stein PH, Fraser JD, Weiss A. 1994. The cytoplasmic domain of CD28 is both necessary and sufﬁcient for costimulation of interleukin-2 secretion and association with phosphatidylinositol 3 -kinase. Mol. Cell. Biol. 14:3392–402 174. Deane JA, Fruman DA. 2004. Phosphoinositide 3-kinase: diverse roles in immune cell activation. Annu. Rev. Immunol. 22:563–98 175. Hu H, Rudd CE, Schneider H. 2001. Src kinases Fyn and Lck facilitate the accumulation of phosphorylated CTLA-4 and its association with PI-3 kinase in intracellular compartments of T-cells. Biochem. Biophys. Res. Commun. 288:573–78 176. Freeman GJ, Long AJ, Iwai Y, Bourque K, Chernova T, et al. 2000. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation. J. Exp. Med. 192:1027–34 177. Watanabe N, Gavrieli M, Sedy JR, Yang J, Fallarino F, et al. 2003. BTLA is a lymphocyte inhibitory receptor with similarities to CTLA-4 and PD-1. Nat. Immunol. 4:670–79 178. Zhang X, Schwartz JC, Guo X, Bhatia S, Cao E, et al. 2004. Structural and functional analysis of the costimulatory receptor programmed death-1. Immunity 20:337–47 179. Shlapatska LM, Mikhalap SV, Berdova AG, Zelensky OM, Yun TJ, et al. 2001. CD150 association with either the SH2-containing inositol phosphatase or the SH2-containing protein tyrosine phosphatase is regulated by the adaptor protein SH2D1A. J. Immunol. 166:5480–87 180. Okazaki T, Maeda A, Nishimura H, Kurosaki T, Honjo T. 2001. PD-1 immunoreceptor inhibits B cell receptor-mediated signaling by recruiting src homology 2-domaincontaining tyrosine phosphatase 2 to phosphotyrosine. Proc. Natl. Acad. Sci. USA 98: 13866–71 181. Liu Y. 1997. Is CTLA-4 a negative regulator for T-cell activation? Immunol. Today 18:569– 72 182. Kwon H, Jun HS, Khil LY, Yoon JW. 2004. Role of CTLA-4 in the activation of singleand double-positive thymocytes. J. Immunol. 173:6645–53 183. Pandiyan P, Gartner D, Soezeri O, Radbruch A, Schulze-Osthoff K, Brunner-Weinzierl MC. 2004. CD152 (CTLA-4) determines the unequal resistance of Th1 and Th2 cells against activation-induced cell death by a mechanism requiring PI3 kinase function. J. Exp. Med. 199:831–42 184. Bour-Jordan H, Grogan JL, Tang Q, Auger JA, Locksley RM, Bluestone JA. 2003. CTLA4 regulates the requirement for cytokine-induced signals in T(H)2 lineage commitment. Nat. Immunol. 4:182–88 185. Khattri R, Auger JA, Grifﬁn MD, Sharpe AH, Bluestone JA. 1999. Lymphoproliferative disorder in CTLA-4 knockout mice is characterized by CD28-regulated activation of Th2 responses. J. Immunol. 162:5784–91 186. Oosterwegel MA, Mandelbrot DA, Boyd SD, Lorsbach RB, Jarrett DY, et al. 1999. The role of CTLA-4 in regulating Th2 differentiation. J. Immunol. 163:2634–39 187. Constant S, Pfeiffer C, Woodard A, Pasqualini T, Bottomly K. 1995. Extent of T cell receptor ligation can determine the functional differentiation of naive CD4+ T cells. J. Exp. Med. 182:1591–96 188. Hosken NA, Shibuya K, Heath AW, Murphy KM, O’Garra A. 1995. The effect of antigen dose on CD4+ T helper cell phenotype development in a T cell receptor-αβ-transgenic model. J. Exp. Med. 182:1579–84

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

96

Teft

·

Kirchhof

·

Madrenas

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-03

ARI

13 February 2006

11:54

189. Kato T, Nariuchi H. 2000. Polarization of naive CD4+ T cells toward the Th1 subset by CTLA-4 costimulation. J. Immunol. 164:3554–62 190. Perrin PJ, Maldonado JH, Davis TA, June CH, Racke MK. 1996. CTLA-4 blockade enhances clinical disease and cytokine production during experimental allergic encephalomyelitis. J. Immunol. 157:1333–36 191. Hurwitz AA, Sullivan TJ, Krummel MF, Sobel RA, Allison JP. 1997. Speciﬁc blockade of CTLA-4/B7 interactions results in exacerbated clinical and histologic disease in an actively-induced model of experimental allergic encephalomyelitis. J. Neuroimmunol. 73:57–62 192. Karandikar NJ, Vanderlugt CL, Walunas TL, Miller SD, Bluestone JA. 1996. CTLA-4: a negative regulator of autoimmune disease. J. Exp. Med. 184:783–88 193. Chambers CA, Kuhns MS, Allison JP. 1999. Cytotoxic T lymphocyte antigen-4 (CTLA4) regulates primary and secondary peptide-speciﬁc CD4+ T cell responses. Proc. Natl. Acad. Sci. USA 96:8603–8 194. Chambers CA, Sullivan TJ, Truong T, Allison JP. 1998. Secondary but not primary T cell responses are enhanced in CTLA-4-deﬁcient CD8+ T cells. Eur. J. Immunol. 28:3137–43 195. Bachmann MF, Waterhouse P, Speiser DE, McKall-Faienza K, Mak TW, Ohashi PS. 1998. Normal responsiveness of CTLA-4-deﬁcient anti-viral cytotoxic T cells. J. Immunol. 160:95–100 196. Tang Q, Boden EK, Henriksen KJ, Bour-Jordan H, Bi M, Bluestone JA. 2004. Distinct roles of CTLA-4 and TGF-β in CD4+ CD25+ regulatory T cell function. Eur. J. Immunol. 34:2996–3005 197. Birebent B, Lorho R, Lechartier H, de Guibert S, Alizadeh M, et al. 2004. Suppressive properties of human CD4+ CD25+ regulatory T cells are dependent on CTLA-4 expression. Eur. J. Immunol. 34:3485–96 198. Grohmann U, Orabona C, Fallarino F, Vacca C, Calcinaro F, et al. 2002. CTLA-4-Ig regulates tryptophan catabolism in vivo. Nat. Immunol. 3:1097–101 199. Grohmann U, Fallarino F, Puccetti P. 2003. Tolerance, DCs and tryptophan: much ado about IDO. Trends Immunol. 24:242–48 200. Fallarino F, Grohmann U, Hwang KW, Orabona C, Vacca C, et al. 2003. Modulation of tryptophan catabolism by regulatory T cells. Nat. Immunol. 4:1206–12

www.annualreviews.org • CTLA- 4 Function

97

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:65-97. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li,1 Yisong Y. Wan,1 Shomyseh Sanjabi,1 Anna-Karin L. Robertson,1 and Richard A. Flavell1,2 1

Section of Immunobiology, 2 Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, Connecticut 06520; email: [email protected], [email protected], [email protected], [email protected], richard.ﬂ[email protected]

Annu. Rev. Immunol. 2006. 24:99–146 First published online as a Review in Advance on November 8, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090737 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0099$20.00

Key Words TGF-β, Smad, tolerance, inﬂammation, immunopathology

Abstract Transforming growth factor-β (TGF-β) is a potent regulatory cytokine with diverse effects on hemopoietic cells. The pivotal function of TGF-β in the immune system is to maintain tolerance via the regulation of lymphocyte proliferation, differentiation, and survival. In addition, TGF-β controls the initiation and resolution of inﬂammatory responses through the regulation of chemotaxis, activation, and survival of lymphocytes, natural killer cells, dendritic cells, macrophages, mast cells, and granulocytes. The regulatory activity of TGF-β is modulated by the cell differentiation state and by the presence of inﬂammatory cytokines and costimulatory molecules. Collectively, TGF-β inhibits the development of immunopathology to self or nonharmful antigens without compromising immune responses to pathogens. This review highlights the ﬁndings that have advanced our understanding of TGF-β in the immune system and in disease.

99

ANRV270-IY24-04

ARI

23 February 2006

15:9

INTRODUCTION

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Tolerance: a physiological state in which the immune system does not mount destructive responses to antigens Regulatory T cells (Tregs): T cell subsets that possess immune suppression activities that are essential for maintaining self-tolerance and for controlling pathological immune responses.

The immune system has evolved to mount robust responses to pathogens while maintaining tolerance to self or innocuous antigens such as those in commensal bacteria. Multiple mechanisms operate to ensure this normal immunological function. Deletion of highafﬁnity self-reactive T and B cell clones during their development creates a peripheral lymphocyte repertoire that preferentially recognizes nonself-antigens (1, 2). Innate immune recognition of and activation by pathogenassociated molecules further directs immune responses to foreign antigens (3). Nonetheless, no absolute demarcation exists for these selection- or recognition-based mechanisms. There is clear evidence that self-reactive lymphocytes are present in the periphery of normal individuals (4). Self components can also

activate innate immune receptors and sometimes precipitate the development of autoimmune diseases (5, 6). Recently, regulatory T cells (Tregs) and cytokines have been recognized as essential components of peripheral tolerance mechanisms (7, 8). Transforming growth factor-β (TGF-β) is one such regulatory cytokine with a pivotal role in regulating immune responses (9, 10) (Figure 1). TGF-βs are regulatory molecules with pleiotropic effects on cell proliferation, differentiation, migration, and survival that affect multiple biological processes, including development, carcinogenesis, ﬁbrosis, wound healing, and immune responses (11). The TGF-β system originated about one billion years ago, before the divergence of arthropods from vertebrates, and has been progressively adapted to regulate newly emerged

Figure 1 TGF-β regulation of tolerance and immunity. TGF-β signaling is coupled to other environmental changes, allowing it to act as a maintainer of peripheral tolerance and immune homeostasis during steady state in a healthy host. There are times when it is essential to initiate immune responses to provide protection against foreign pathogens. TGF-β is a key regulator of the immune responses, as it controls the threshold of activation and aids in chemotaxis. By regulating the function and survival of activated leukocytes and the healing process, TGF-β also helps to resolve the immune response. An immune response can result in the development of immunological memory, which provides long-term protection. The role of TGF-β in the generation and maintenance of memory cells is an area that requires more investigation. Some of the processes affected by TGF-β during each phase of the immune response are listed. 100

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

systems in higher organisms (12). The ﬁrst observation of TGF-β regulation of immune cell functions was made two decades ago (13). In the 1990s, generation and analysis of TGFβ1−/− mice established a central role for this cytokine in inhibiting inﬂammation and autoimmune diseases and has fostered growing interest in this cytokine in the immune system (14, 15). Identiﬁcation of TGF-β receptors and Smad factors as mediators of receptor signaling pathways has also shed light on the molecular mechanisms of TGF-β regulation of immune responses (16, 17). In the 2000s, development of mouse models with cell type– speciﬁc inactivation of TGF-β signaling has started to reveal the regulatory network of the TGF-β pathway in vivo (18–22). In this review, we discuss the multifaceted roles of TGF-β in the immune system with an emphasis on its regulation of peripheral leukocyte functions.

TGF-β EXPRESSION AND ACTIVATION TGF-βs belong to the TGF-β superfamily, with additional members including bone morphogenetic proteins, activins, and growth differentiation factors (23). There are three homologous TGF-β isoforms in mammals, TGF-β1, TGF-β2, and TGF-β3, encoded by different genes (24). TGF-β1 is the predominant isoform expressed in the immune system, but all three isoforms have similar properties in vitro. TGF-β is synthesized as a prepro-TGF-β precursor. The pre region contains a signal peptide, and pro-TGF-β is processed in the Golgi by a furin-like peptidase (25), where the N-terminus of the immature protein, the propeptide, is removed. A homodimer of this new protein, called the latencyassociated protein (LAP), is noncovalently associated with a homodimer of mature TGFβ.This complex is called latent TGF-β or the small latent complex (SLC). The SLC can be secreted as such or in association with latentTGF-β-binding protein (LTBP) as a large

TGF-β AND WOUND HEALING Tissue damage often occurs during pathological immune responses due to inﬂammation, where neutrophils enter the wound site to remove pathogens and damaged tissue through phagocytosis, and macrophages (Ms) appear to continue this process and secrete more growth factors, including TGF-β. Following pathogen clearance, ﬁbroblasts migrate in to deposit collagen that is cross-linked and organized to form an extracellular matrix during the ﬁnal healing phase. For these highly organized processes to occur, numerous cell-signaling events are required, and TGF-β is an important one. As a ﬁbrotic factor, TGF-β facilitates wound healing during resolution of immune responses and many other wound healing processes, e.g., incisional and excisional procedure, punch, ulcers, and bone damage. Discussion of the role of TGF-β in wound healing is beyond the scope of this review. It is nevertheless a critical regulator for immune resolution.

latent complex (LLC). LTBP plays an important role in targeting TGF-β to the extracellular matrix (26). TGF-β cannot bind to its receptors in its latent form but needs to be liberated from the constraints of LAP and LTBP. In vitro, this can be achieved using extremes of pH, heat, or several proteases (26). In vivo, the mechanisms for activation are less clear, but several models have been proposed, including proteolytic activation by transglutaminase, conformational change of LAP through physical interaction with thrombospondin, and mechanical traction via αvβ6 integrins on epithelial cells (27–31). Little is known about the transcriptional regulation of TGF-β1. The TGF-β1 promoter contains activator protein 1 (AP-1) binding sites where c-jun and c-fos induced by TGF-β1 itself bind to and stimulate TGF-β1 production (32–34). A recent study has shown that the AP-1 activators c-Jun-NH2-terminal kinase 1 and 2 (JNK1 and JNK2) inhibit TGF-β1 expression in ﬁbroblasts (35), suggesting that AP-1 may play a dual role in regulating TGF-β1 transcription. Cytokines can also modulate TGF-β1 expression in

www.annualreviews.org • Regulation of Immune Responses by TGF-β

Smad: a family of intracellular transcription factors activated by the TGF-β family of receptors through phosphorylation. The nomenclature is a combination of the orthologs identiﬁed in C. elegans (Sma) and Drosophila (Mad) LAP: latency-associated protein SLC: small latent complex LTBP: latent-TGFβ-binding protein LLC: large latent complex

101

ANRV270-IY24-04

ARI

23 February 2006

ALK: activin receptor like kinase

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

SBE: Smad-binding element

15:9

immune cells (see below), although the mechanisms, in most cases, are uncharacterized. In many cases, TGF-β1 message levels do not correlate with the amount of secreted protein, suggesting post-transcriptional and/or post-translational regulation of TGF-β1 expression (36). The most abundant TGF-β1 transcript has a long 5 untranslated region that negatively regulates TGF-β1 translation, the molecular mechanism of which is still not completely understood (37). The regulation of TGF-β2 and TGF-β3 expression is distinct from that of TGF-β1 and is mostly under the control of developmental or hormonal signals (38).

TGF-β RECEPTOR AND SIGNALING The TGF-β superfamily mediates its biological functions via binding to type I and II transmembrane serine/threonine kinase receptors. Although more than 35 TGF family members have been identiﬁed, only ﬁve type I (activin receptor like kinase, ALK family) and seven type II receptors have been reported (23). TGF-β mediates its functions mostly via ALK5 and TGF-β receptor II (TGFβRII) (Figure 2). The active forms of TGF-β initially engage TGFβRII. TGF-β1 and TGFβ3 bind to TGFβRII with high afﬁnity, but strong binding of TGF-β2 to this receptor only occurs in the presence of membranebound betaglycan, also known as TGFβRIII (39). Although ALK5 is not required for the initial binding of TGF-β, it is necessary for the signaling. Active TGF-β dimer binds to the tetrameric ALK5 and TGFβRII receptor complex to initiate cell signaling (40). In endothelial cells, TGF-β can use an alternative type I receptor, ALK1, for signaling, but its involvement in immune cell functions remains unknown (41). TGF-β family signaling is best characterized as a linear pathway where ligand binding brings the type II receptor near and then activates the type I receptor, which subsequently phosphorylates intracellular Smad 102

Li et al.

proteins whose nuclear localization is required for the transcriptional regulation of target genes (40, 42) (Figure 2). The eight vertebrate Smads identiﬁed so far are grouped into three categories: ﬁve receptor-associated Smads (R-Smad1, 2, 3, 5, and 8), one common Smad (Co-Smad4), and two inhibitory Smads (I-Smad6 and 7). R-Smads are sequestered in the cytoplasm in the absence of signaling. Upon phosphorylation by ALK5, R-Smad2 and 3 associate with Co-Smad4 and translocate into the nucleus to bind to SBE (Smad-binding element) (43–46). Although Co-Smad4 facilitates nuclear translocation of R-Smads, R-Smad can translocate into the nucleus without Co-Smad4 (42, 47). Unlike R-Smads, I-Smad7 is not phosphorylated upon TGF-β activation (48). Instead of being a transmitter for TGF-β signaling, I-Smad7 suppresses TGF-β signaling through at least two mechanisms: (a) competing with R-Smads for the binding to ALK5 and (b) recruiting Smurf-containing E3 ubiquitinase complexes to degrade ALK5 (49, 50) (Figure 2). Activated Smad protein complexes bind to DNA weakly, and their highafﬁnity DNA binding is achieved by associating with a large number of other transcription factors (51–53). To control gene expression, Smad complexes recruit coactivators that contain histone-acetyl transferase (HAT) activity, e.g., CBP/p300; or histone-deacetylase (HDAC) activity–containing corepressors, e.g., Sno/Ski, to activate or repress target genes, respectively (52, 54, 55) (Figure 2). Studies using cells that are deﬁcient in Co-Smad4 or that express dominant-negative R-Smads or mutated ALK5 that is defective in signaling support the existence of Smad-independent TGF-β signaling pathways (56, 57) (Figure 2). Indeed, rapid activation of Ras-Erk, TAK-MKK4-JNK, TAKMKK3/6-p38, Rho-Rac-cdc42 MAPK, and PI3K-Akt pathways were observed when cells were treated with TGF-β, although detailed mechanisms are not yet fully understood (54). Furthermore, cross-talk between MAPKs and Smad proteins appears to modulate TGF-β

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

Figure 2 TGF-β transduces signals through Smad-dependent and Smad-independent pathways. Dimeric TGF-β becomes activated following dissociation from LAP and LTBP and binds to and activates tetrameric TGF-β receptor complex consisting of ALK5 and TGFβRII. Activated ALK5 phosphorylates Smad2 and Smad3, which translocate into the nucleus in a complex with Smad4. Smad complex binds to a target promoter in association with other transcription factors (TFs) and regulates gene expression via recruiting HAT or HDAC. One of TGF-β’s targets is Smad7, which downregulates TGF-β signaling by competing with Smad2 and Smad3 for ALK5 binding and by degrading ALK5 through recruiting Smurf-containing E3 ubiquitin ligase complexes. In addition to Smads, PI3K and various MAPKs are also activated by TGF-β. Ras-ERK also cross-talks with Smads. Furthermore, activated TGF-β receptors modulate the function of FKBP12, TRIP-1, PP2A, and eIFs by direct association.

responses (58–60). In addition, TGF-β receptors activate FKBP12, TRIP-1, and PP2A and regulate certain translation initiation complexes, such as eIF2a and eIF3 through direct protein binding (61–63). The signiﬁcance of these pathways in regulating immune cell functions remains to be established.

TGF-β REGULATION OF IMMUNE CELL FUNCTION TGF-β and T Lymphocytes TGF-β exerts the greatest impact on T lymphocytes, as is evident in mice with a T cell– speciﬁc blockade of TGF-β signaling (18).

www.annualreviews.org • Regulation of Immune Responses by TGF-β

103

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

Figure 3 TGF-β regulation of T cell responses. TGF-β blocks T cell proliferation by inhibiting IL-2 production via Smad3 and by downregulating the expression of cyclinD2 and E, CDK4, and c-myc. TGF-β inhibits differentiation of Th1 and Th2 cells through suppressing the expression or function of T-bet/Stat4 and GATA-3/NFAT. Mechanisms involved in TGF-β inhibition of CTL differentiation are not well understood, although inhibition of c-myc and T-bet expression is suggested. TGF-β induces FoxP3 expression and the generation of Tregs. TGF-β also prevents T cell activation-induced cell death (AICD) through inhibiting c-myc-mediated FasL expression and other unknown mechanisms.

cki: cycle-dependent kinase inhibitor

TGF-β affects T cell proliferation, differentiation, and survival (64, 65) (Figure 3). The differentiation states of target cells and the presence of additional regulatory signals, including costimulatory molecules and inﬂammatory cytokines, inﬂuence TGF-β regulation of T cell activity. The ﬁnal outcome of TGF-β on T cells is therefore context dependent. TGF-β also induces the FoxP3expressing CD4+ CD25+ Tregs and can therefore indirectly inﬂuence T cell activation (Figure 3). TGF-β regulation of T cell proliferation. A crucial function for TGF-β in T cell biology was ﬁrst demonstrated by its antiproliferation activity on T cells in vitro (13). TGF-β uses multiple pathways to inhibit T cell proliferation, including inhibiting the expression of the T cell mitogenic cytokine interleukin (IL)-2 which is most likely via suppression of IL-2 transcription. Functional analysis of cis-regulatory elements of the murine IL-2 promoter region showed that TGF-β blocks

104

Li et al.

IL-2 transcription via inhibition of the activity of an octamer-binding enhancer (66). Smad3−/− murine T cells are resistant to TGF-β-induced inhibition of IL-2 expression (67). Interestingly, one SBE was mapped to the –105 region upstream of the human IL-2 promoter, through which Smads cooperate with the transcriptional corepressor Tob to inhibit IL-2 transcription (68). Inclusion of IL-2 in T cell culture relieves TGF-β inhibition of T cell proliferation in a dose-dependent manner (13). Nevertheless, even with the highest dose of IL-2, TGF-βinduced reduction of T cell proliferation was observed (13). TGF-β inhibits proliferation of various cell types through mechanisms directly targeting cell cycle regulators, including upregulation of cyclin-dependent kinase inhibitors (cki) p15, p21, or p27, and downregulation of c-myc (69–72). A recent study showed that TGF-β inhibits T cell proliferation in all three cki knockouts, demonstrating a dispensable or redundant role for the ckis in the TGF-β pathway in T cells (73).

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

TGF-β blocks the expression of c-myc, cyclin D2, and cyclin E in T cell lines (74–76), as well as cdk4 expression in primary CD4+ T cells (73). The signiﬁcance of these molecules in TGF-β inhibition of T cell proliferation warrants further investigation. Smad3 is essential for TGF-β blockade of IL-2 production in both CD4+ and CD8+ T cells (77). Nevertheless, although TGF-β inhibition of T cell receptor (TCR)-mediated CD4+ T cell proliferation is abrogated in Smad3−/− T cells, its inhibition of CD8+ T cell proliferation remains intact (77). These observations revealed a Smad3- and IL-2-independent antiproliferative pathway in CD8+ T cells, the molecular mechanism of which is currently unknown. TGF-β inhibition of T cell proliferation is inﬂuenced by the differentiation status of T cells and the integrated signals from cytokines and costimulatory molecules on T cells. Although TGF-β inhibits naive T cell proliferation, it has minimal effect on activated T cells, which correlates with reduced TGFβRII expression (78). Interestingly, IL-10 treatment enhances TGFβRII expression and restores TGF-β responsiveness on activated T cells (78). The presence of CD28 costimulation attenuates TGF-β inhibition of TCR-mediated naive CD4+ T cell proliferation (79). Although the expression of IL-2 is strongly inhibited by TGF-β even with CD28 stimulation, IL-2 is required for the proliferation of TCR/CD28/TGF-βstimulated naive T cells (79). It appears that CD28 costimulation synergizes with low IL-2 to drive T cell proliferation, which overrides the inhibitory effects of TGF-β. It is conceivable that TGF-β inhibition of naive T cell activation in the absence of costimulation blunts T cell responses to self-antigens under steady state, whereas reversal of TGF-β inhibition by strong costimulatory signals, which are often associated with infection, limits the suppressive activity of TGF-β during a normal immune response. Thymocytes isolated from TGF-β1deﬁcient mice hyperproliferated upon stim-

ulation (80), which might contribute to the development of their autoimmune phenotype. In transgenic mice, the expression of a dominant-negative mutant of TGFβRII under the CD2 promoter speciﬁcally inhibits TGF-β signaling in T cells (19). These mice develop a progressive CD8+ T cell lymphoproliferative disorder that later transforms into lymphoma (81). Recently, an important function for Smad3 in inhibiting leukemogenesis was revealed: A compound mutation of p27 cki and Smad3 in mice results in the development of T cell lymphoma (82). Interestingly, Smad3 protein is also absent or reduced in several cases of human T cell acute lymphoblastic leukemia (82). Thus, TGF-β inhibition of T cell proliferation appears to be an important mechanism to maintain T cell homeostasis and to prevent lymphoproliferative disorder.

TCR: T cell receptor

TGF-β regulation of helper T (Th) cell differentiation. Activation of naive CD4+ T cells under polarization conditions leads to their differentiation into helper T cells (83). T helper 1 (Th1) cells express interferon-γ (IFN-γ) and lymphotoxin-α (LT-α), which mobilize the cellular arm of the immune system to combat intracellular pathogens. Th2 cells secrete cytokines, including IL-4, IL-13, and IL-5, which are essential for optimal antibody production and elimination of extracellular pathogens. Dysregulated CD4+ Th cells can induce immunopathology. Excessive Th1 responses are associated with various autoimmune and inﬂammatory disorders, whereas enhanced Th2 cytokine production is involved in atopic diseases, including allergic asthma. TGF-β is a potent regulator of effector T cell differentiation, and it generally inhibits the acquisition of Th cell functions (64). TGF-β-mediated inhibition of Th cell differentiation occurs even in the presence of IL-2, in which case T cell proliferation is largely unaffected (84). These observations demonstrate that TGF-β can target both proliferation and differentiation programs in T cells independently.

www.annualreviews.org • Regulation of Immune Responses by TGF-β

105

ARI

23 February 2006

15:9

Effector T cell differentiation is orchestrated by signals from TCR, costimulatory molecules, and cytokine receptors (83). These integrated signals induce the expression of lineage-commitment transcription factors that drive Th cell differentiation. Transcription factors T-bet and Stat4 specify Th1 cell fate, whereas GATA-3 and Stat6 induce Th2 cell differentiation (83). Several reports show that TGF-β prevents the development of Th2 cells by inhibiting the expression of GATA-3 (85, 86). However, these studies do not reveal whether TGF-β directly inhibits GATA-3 expression. One recent study showed that TGF-β interferes with TCR and costimulatory receptor signaling pathways by inhibiting the activation of the Tec kinase Itk and the calcium inﬂux elicited by TCR/CD28 stimulation in CD4+ T cells (87). Impaired calcium signaling compromises nuclear translocation of nuclear factor of activated T cells (NFAT) transcription factors and the expression of GATA-3 (87). TGF-β blockade of Th1 cell differentiation is associated with reduced IL-12 receptor β2 (IL-12Rβ2) and T-bet expression (88, 89). T-bet is required for the induction of IL12Rβ2 (90). Therefore, reduced IL-12Rβ2 levels upon TGF-β treatment is likely due to its inhibition of T-bet expression, which is dependent on the IFN-γ/Stat1 pathway (90). T-bet also induces IFN-γ expression (91) and provides a positive feedback loop to stabilize Th1 cell fate. Whether TGF-β primarily targets IFN-γ or T-bet to inhibit Th1 cell differentiation is unclear. TGF-β also inhibits the expression of Stat4 (92), a transcription factor activated by IL-12. It appears that Stat4 downregulation blocks IFN-γ expression at priming, whereas reduced T-bet expression inhibits IFN-γ production at recall responses (92). Thus, TGF-β can target multiple signaling pathways to regulate Th1 cell differentiation. Similar to T cell proliferation, TGF-β regulation of Th cell differentiation is inﬂuenced by cell differentiation state and other signals. Fully differentiated Th2 cells are re-

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

106

Li et al.

fractory to TGF-β inhibition of cytokine production, whereas IFN-γ expression in differentiated Th1 cells is inhibited by TGF-β (93). Nevertheless, early IFN-γ production by Th1 cells attenuated the inhibitory effects of TGF-β, as IFN-γ neutralization leads to a complete blockade of Th1 cell differentiation (84). IFN-γ is known to induce Smad7 expression and consequent inhibition of TGF-β signaling (94). Whether the same mechanism operates in Th1 cells remains to be tested. Whereas TGF-β inhibits the production of most effector cytokines in CD4+ T cells, ectopic expression of TGF-β1 in T cells results in enhanced production of the antiinﬂammatory cytokine IL-10 (95). This effect is more profound under Th1 conditions and is associated with TGF-β activation of the IL-10 promoter via Smad4 (95). These ﬁndings are consistent with the observation that these two suppressive cytokines are coexpressed at sites of inﬂammation (96). TGF-β also induces the expression of IL-9 in CD4+ T cells (97), an important cytokine involved in the Th2 responses that induce mast cells. The in vivo signiﬁcance of this pathway remains unknown. The prominent inhibitory activity of TGF-β on Th cell differentiation in vitro is supported by the in vivo phenotypes of TGF-β1−/− mice and mice with T cell– speciﬁc blockade of TGF-β signaling. TGFβ1-deﬁcient mice in a mixed genetic background develop a multifocal inﬂammatory disease associated with increased inﬂammatory cytokine production (14, 15). Similar CD4+ T cell activation and differentiation was observed when TGF-β1−/− mice were backcrossed onto the BALB/c background (98). Signiﬁcantly, depletion of CD4+ T cells or crossing TGF-β1−/− mice onto an MHC class II–null background inhibits inﬂammation (99), demonstrating an essential role for CD4+ T cells in promoting disease development. Whether T cells themselves are direct targets of TGF-β1 is not revealed from these studies because TGF-β1 can modulate the activity of multiple cell types. Expression

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

of a dominant-negative form of TGFβRII from the CD4 promoter (CD4-DNR) blocks TGF-β signaling in T cells (18). These mice develop an autoimmune inﬂammatory phenotype associated with CD4+ T cell differentiation into Th1 or Th2 cells (18). Another study demonstrated that transgenic expression of Smad7 in T cells attenuates TGF-β signaling in T cells and renders mice hyperresponsive to antigen-induced airway inﬂammation associated with enhanced Th2 cell differentiation (20). These observations reveal that TGF-β is an essential regulator of Th1 and Th2 cell responses in vivo. TGF-β regulation of cytotoxic T lymphocyte differentiation. CD8+ T cells represent an important arm of adaptive immunity to intracellular pathogens and tumors. Dysregulated CD8+ T cells mediate pathogenesis in various autoimmune diseases. CD8+ T cells exert their effector functions through production of inﬂammatory cytokines such as IFN-γ and LT-α, and acquisition of cytolytic activities mediated by perforin/granzyme and death receptor pathways. Compared with CD4+ T cells, naive CD8+ T cells acquire effector T cell functions more readily upon antigenic stimulation (100) and therefore require stringent regulation. In addition to proliferation, TGF-β potently inhibits CD8+ T cell differentiation. Early studies showed that CD8+ T cells activated in the presence of TGF-β do not acquire cytotoxic T lymphocyte (CTL) functions (101), which is likely due to TGF-β inhibition of perforin expression in activated CD8+ T cells (102). TGF-β also inhibits Fas ligand (FasL) expression in T cell lines (75), thereby affecting the death receptor cytotoxic pathway of CD8+ T cells. As in Th1 cells, the presence of TGF-β greatly attenuates IFN-γ expression in CD8+ T cells (103, 104). Thus, TGF-β inhibits the expression of multiple effector molecules of CTLs. Although not well understood, several molecular mechanisms involved in TGF-β-mediated inhibition of CTL effector functions have been

suggested. IFN-γ expression and CTL differentiation is regulated by two T-box transcription factors, eomesodermin and T-bet (105). Similar to Th1 cells, TGF-β inhibition of IFN-γ expression in CD8+ T cells is associated with reduced T-bet expression (104). TGF-β inhibition of FasL expression appears to be a consequence of blockade of cmyc expression (75). In addition, TGF-β inhibition of perforin expression was observed upon activation of resting T cells (102), implying perforin as a direct TGF-β target. TGFβ regulation of CD8+ T cell differentiation is context dependent. 4-1BB is a costimulatory molecule induced in activated CD4+ and CD8+ T cells (106). 4-1BB costimulation reverses TGF-β inhibition of CTL differentiation (107), which is further enhanced by IL-12 but reduced by IL-4 (107). Therefore, TGF-β inhibition of CD8+ T cell differentiation is modulated by costimulatory receptor and cytokine signaling pathways. TGF-β regulation of CD8+ T cell differentiation has been further illuminated by in vivo studies with TGF-β1−/− mice and CD4-DNR mice. Elimination of CD8+ T cells alleviated the inﬂammatory disease developed in TGF-β1−/− mice (108). The CD4 promoter used in CD4-DNR mice lacks the CD8 silencer (18); therefore, TGF-β signaling in both CD4+ and CD8+ T cells is blocked in these mice. CD8+ T cells in this model spontaneously acquire an activated phenotype and produce effector cytokines (18). Remarkably, CD4-DNR mice also mount a strong antitumor response (109), associated with the expansion and enhanced activities of tumorspeciﬁc CTLs (109). Thus, TGF-β essentially regulates CD8+ T cell differentiation in vivo.

DNR: dominant-negative TGF-β receptor II CTL: cytotoxic T lymphocytes

TGF-β regulation of Tregs. The existence of T cell subsets that actively suppress immune responses and maintain immunological tolerance was documented in the 1970s (110). These suppressor T cells are now termed regulatory T cells (Tregs). The bestcharacterized Tregs are the naturally occurring Tregs that are CD4+ CD25+ and

www.annualreviews.org • Regulation of Immune Responses by TGF-β

107

ANRV270-IY24-04

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Activation-induced cell death (AICD): T cells proliferate in response to antigen stimulation and undergo apoptosis to eliminate excess numbers of activated T cells

108

15:9

comprise 5%–10% of peripheral CD4+ T cells (7). FoxP3, a forkhead family transcription factor, is a lineage speciﬁcation factor for Tregs (111). Tregs are generated during thymic development as well as in the periphery during homeostatic proliferation (7, 112, 113). In addition to Tregs, mature T cells are able to acquire immunoregulatory functions upon activation, such as CD4+ Th3 cells generated in oral tolerance models (114). The role of TGF-β in the development, maintenance, and induction of Tregs has attracted much attention recently. Investigators have shown that TGF-β1 can convert CD4+ CD25− T cells to Tregs in vitro (115– 119). Overexpression of TGF-β1 in the islets of the pancreas also expands the Treg population that protects nonobese diabetic (NOD) mice from type 1 diabetes (120). In both cases, the induction of regulatory activities correlates with increased expression of FoxP3. In fact, only FoxP3-expressing T cells possess suppressive activities (119). The function of TGF-β in Treg generation/maintenance under physiological conditions remains controversial and poorly understood. Blockade of TGF-β signaling in Tregs inhibits their proliferation in mice treated with dextran sulfate, a drug that induces colitis. These Tregs also fail to protect mice from dextran sulfate–induced colitis (121). In another report, peripheral but not thymic Treg numbers were found reduced in 8- to 10-day old TGF-β1−/− mice (122), which suggests a function for endogenous TGF-β1 in the maintenance of the peripheral Treg population. However, an earlier study showed no defect in Treg development in TGF-β1−/− mice (123). In another recent report, thymic Tregs from CD4-DNR mice retain their ability to inhibit colitis, and TGF-β1−/− TCR transgenic T cells possess a normal number of FoxP3-expressing Tregs (124). The reasons for these discrepancies are unknown but may relate to the different experimental systems and mouse genetic backgrounds used in these studies. In addition, CD25 was used as a marker for Tregs. Because CD25 is also exLi et al.

pressed by activated T cells, Tregs identiﬁed by CD25 staining might be contaminated by effector T cells. With the development of GFP- and RFP-FoxP3-knockin mice (119, 125), these potential complications can be circumvented. TGF-β regulation of T cell survival. Acquired immune response to foreign antigens is a multistep process that involves clonal expansion of antigen-speciﬁc T cells followed by contraction; most of these are eliminated by apoptosis, but some activated T cells develop into memory T cells. There are two predominant apoptotic pathways in T cells: the intrinsic pathway that is regulated by the proand antiapoptotic Bcl-2 family members; and the death receptor pathway, including CD95 (APO-1/Fas), that induces apoptosis via direct caspase activation (126, 127). TGF-β promotes the survival of T cells during T cell expansion and differentiation. TGF-β inhibition of T cell activation (see above) is inﬂuenced by the T cell differentiation status and integrated signals from cytokines and costimulatory molecules. Therefore, in the absence of CD28 costimulation, TGF-β inhibits TCR-stimulated proliferation of naive T cells. However, in the presence of CD28, TGF-β inhibits T cell apoptosis and promotes T cell expansion (79, 128). TGF-β1 also promotes the survival of T cells with memory/effector phenotypes (78, 79, 93, 129, 130). Additionally, IL-2 and TGF-β synergize to block activation-induced cell death (AICD) in both Th1 and Th2 CD4+ T cells (129, 131–133). The mechanism of TGF-β-mediated T cell survival is not well understood; however, several studies have shown that TGF-β inhibits Fas-induced T cell apoptosis (134, 135). The antiapoptotic effect of TGF-β was associated with reduced c-myc expression that resulted in reduced levels of FasL (75). Together, these studies demonstrate that TGF-β can promote T cell survival and effector T cell expansion. The ﬁrst in vivo demonstration of TGFβ’s role in T cell survival came from the

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

analysis of TGF-β1−/− mice (136). Spontaneous apoptosis was found in both thymic and peripheral T cells in the absence of TGF-β1 (136). When TGF-β1−/− mice were treated with anti-CD3 antibody, massive thymic and peripheral T cell apoptosis were observed (136). Addition of TGF-β to T cell culture does not correct the increased cell death in TGF-β1−/− cells, and the apoptotic phenotype was not found in Smad3−/− mice. On the basis of these observations, investigators suggested an antiapoptotic role for intracellular TGF-β1 (136). In another study, splenic T cells from TGF-β1−/− mice show elevated IFN-γ in response to TCR stimulation and undergo enhanced apoptosis (137). IFNγ precipitates T cell AICD by inducing caspase 8 expression (138). Therefore, increased T cell apoptosis in TGF-β1−/− mice may also be a consequence of enhanced IFN-γ production. It remains to be determined whether TGF-β regulation of T cell survival represents an independent pathway or is secondary to its regulation of T cell activation.

T cell production of TGF-β. Among the three family members of TGF-β, TGF-β1 is the predominant isoform expressed in leukocytes (9). Early studies showed that activated T cells produce TGF-β1 (13). Later studies of oral tolerance models revealed a subset of CD4+ T cells, named Th3 cells, that express TGF-β1 as an effector cytokine (114). TGF-β1 produced by Th3 cells has been proposed to inhibit Th1 and Th2 cell differentiation and provide help for B cells in IgA production (139). Despite considerable study, the mechanisms by which the recently identiﬁed CD4+ CD25+ Tregs suppress immune responses remain poorly understood. Some Tregs express surface TGF-β1, which investigators have suggested mediates their suppressor function (140–142). However, TGF-β1−/− Tregs still inhibit proliferation of responder T cells in vitro, indicating that TGF-β1 synthesis by Tregs is not required for suppression under these culture conditions

(143). The protective activity of Tregs against inﬂammatory bowel disease (IBD) induced by transfer of CD4+ CD45RBhigh T cells into SCID mice is reversed by the treatment of the recipients with anti-TGF-β antibody (144). However, in an autoimmune gastritis model, Treg-mediated protection is not affected by the same treatment (143). Even in the same colitis model, TGF-β1−/− Tregs were shown to be protective or inactive depending on experimental systems (123, 124, 145). Therefore, the function of Treg-produced TGF-β1 remains controversial. To determine a deﬁnitive function of T cell–produced TGF-β1 in vivo, we have generated a strain of TGF-β1 conditional knockout mice. Mice that lack TGF-β1 selectively in T cells develop an autoimmune inﬂammatory disease characterized by multiorgan leukocytic inﬁltration and circulating autoreactive antibodies, which is associated with spontaneous T cell activation and effector T cell differentiation. TGF-β1 produced by Tregs, as well as by naive/activated T cells, is required to protect mice from IBD (M.O. Li and R.A. Flavell, unpublished). These observations reveal a speciﬁc function for T cell–produced TGF-β1 in suppressing T cell activation and inﬂammatory disease. Cell signaling pathways involved in TGFβ1 production in T cells remain largely uncharacterized. Engagement of the inhibitory receptor CTLA-4 was described as inducing TGF-β1 production (146–148), although this was disputed in another report (149). CD69 is an early leukocyte activation molecule. Cross-linking of CD69 induces TGF-β1 production in both CD4+ and CD8+ T cells (150). Interestingly, CD69−/− mice develop exaggerated collagen-induced arthritis (151), which is normally inhibited by TGF-β1. Th2 polarization conditions enhance and Th1 conditions inhibit TGF-β1 production in vitro (152). Consistent with this, CD4+ T cells from T-bet-deﬁcient mice produce more TGF-β1 than wild-type T cells (153). In addition, T cells produce TGF-β1 when

www.annualreviews.org • Regulation of Immune Responses by TGF-β

109

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

Figure 4 TGF-β regulation of B cell responses. TGF-β inhibits B cell proliferation through induction of Id3, p27, and p21 and inhibition of c-myc, calcyclin, and ATM. TGF-β blocks IL-4-induced Stat6 activation via SOCS1 and SOCS3 and B cell receptor–induced Src activation through CD72 and SHIP-1. TGF-β promotes IgA class switching through Smad3/4 and CBFα3-mediated transactivation of germline α-promoter. TGF-β also induces apoptosis in immature and resting B cells through induction of Id3 and Bim, and inhibition of NF-κB and c-myc.

Class switching: a process in which the heavy chain of an antibody is switched from the IgM isotype to IgG, IgA, or IgE isotypes

they undergo apoptosis (154), which might downregulate immune responses to apoptotic cell-associated antigens.

TGF-β and B Lymphocytes TGF-β is an important regulator of B lymphocyte activity, exempliﬁed by the phenotype of mice with a B cell–speciﬁc blockade of TGF-β signaling (21). TGF-β inhibits B cell proliferation, induces apoptosis of immature or resting B cells, and blocks B cell activation and class switching to most isotypes except for IgA (Figure 4). Thus, similar to T cells, TGF-β has both inhibitory and stimulatory effects on B cells (155, 156). TGF-β regulation of B cell proliferation. TGF-β inhibits proliferation of B lymphocyte progenitors, which is mediated partly by the induction of TGF-β target gene Id3, a helixloop-helix transcription factor (157). Overexpression of Smad7 reverses TGF-β-induced Id3 expression, suggesting a requirement of the Smad pathway in this regulation (157).

110

Li et al.

TGF-β also potently inhibits proliferation of mature B cells in vitro, which is associated with cell cycle blockade at G1/S (158–160). TGF-β represses cyclin A expression and inactivates cdk2, possibly via the upregulation of the cki p27 (161, 162). Recently, global gene changes regulated by TGF-β in primary B cells have been characterized. TGF-β induces the expression of the cki p21 and suppresses the expression of c-myc, calcyclin, and ATM (163), which might also collectively induce growth arrest. Studies with Smad3−/− B cells reveal that TGF-β inhibition of B cell proliferation is Smad3 independent (16, 17). The signaling pathway that leads to the inhibition of mature B cell proliferation remains to be identiﬁed. Mature splenic B cells from conditional B cell TGFβRIIknockout mice show increased BrdU labeling (21), revealing an important function for TGF-β in inhibiting B cell proliferation in vivo. As in T cells, TGF-β regulation of B cell proliferation is modulated by other signaling pathways. CD40, a member of the tumor

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

necrosis factor (TNF) receptor superfamily, plays an important role in promoting B cell survival and immunoglobulin (Ig) class switching. Engagement of CD40 induces Smad7 expression and protects B cells from TGF-β-induced growth inhibition (164).

attenuating B cell responses to low-afﬁnity/ -avidity antigens. Interestingly, anti-DNA antibody titers also increase in these mice (21), suggesting that TGF-β is an important regulator of B cell tolerance to self-antigens in vivo.

TGF-β regulation of B cell differentiation. TGF-β is an important regulator of B cell activation and differentiation. Early studies showed that exogenously administrated TGF-β inhibits the expression of both light chains and the secreted form of the heavy chain (165, 166). TGF-β also blocks class switching to most IgG isotypes (158). In contrast, TGF-β promotes switching to IgA and IgG2b in mouse B cells and IgA in human B cells in vitro (167–171). However, although TGF-β signaling in B cells is essential for IgA switching in mice (see below), it is dispensable for IgG2b production (21). These in vivo studies reveal a general inhibitory function for TGF-β on antibody production with the exception of IgA. IL-4 produced by Th2 cells is an important mitogenic and survival factor for B cells and promotes isotype switching to IgG1 and IgE via the Jak/Stat pathway. Signiﬁcantly, TGFβ attenuates IL-4-induced Stat6 phosphorylation in B cells (163), which correlates with enhanced expression of SOCS1 and SOCS3 (163), negative regulators of the Jak/Stat pathway. TGF-β also induces the expression of CD72 and inositol-pyrophosphate phosphatase SHIP-1 (163). The CD72/SHIP-1 complex plays an important role in antagonizing the activity of Src kinases activated upon B cell receptor stimulation. Therefore, TGFβ targets both the cytokine and B cell receptor pathways to regulate B cell activation and differentiation. Consistent with TGF-β inhibition of the production of most antibody isotypes, conditional inactivation of TGF-β signaling in B cells results in elevated serum Ig in mice (21). These mice also develop enhanced IgG3 responses to a normally weak antigen (21), revealing an important function for TGF-β in

TGF-β regulation of IgA production. Secretory IgA is important for the prevention of microbial infection in mucosa (172). TGF-β has a unique role in promoting the differentiation of IgA-secreting plasma cells. Upon optimal antigen and cytokine stimulation, the antiproliferative activity of TGF-β subsides, and TGF-β strongly promotes IgA production (173). TGF-β-induced IgA class switching is associated with increased transcription of germline α-transcripts (174). The identiﬁcation of a TGF-β-responsive tandem repeat sequence in the mouse germline α-promoter led to the characterization of transcription factors involved in this regulation (175). Two Smad-binding sites and two core binding factor elements are necessary and, when dimerized, sufﬁcient for the TGF-β response (175). In vitro studies showed that Smad3/4 and CBFα3 [also known as acute myeloid leukemia (AML)-2 or Runx3] bind to these sites, respectively (176– 178). The Smad3/4 and CBFα3 complexes may further recruit the transcriptional coactivator p300 to augment transcription of the germline α-promoter (179, 180). However, Smad3−/− mice have normal levels of serum IgA (16, 17). The mechanism of Smad3independent regulation of IgA production in vivo is unknown. Mice with a blockade of TGF-β signaling in B cells are almost devoid of serum IgA (21). The TGF-β pathway is also required for the IgA response to mucosally administered antigens (orally or intranasally) (181). TGF-β1−/− mice have IgA only at reduced levels (182). Maternally derived TGF-β1 could be responsible for this partial phenotype (183). Alternatively, TGF-β2 or TGF-β3 may also be involved in promoting IgA switching.

www.annualreviews.org • Regulation of Immune Responses by TGF-β

111

ARI

23 February 2006

15:9

TGF-β regulation of B cell survival. TGF-β induces apoptosis in immature B cells and resting B cells (184, 185). TGF-βmediated apoptosis in the mouse immature B cell line, WEHI 231, is associated with decreased expression of c-myc (186). TGF-β also induces NF-κB inhibitor IκBα and inhibits NF-κB activation (186). Ectopic expression of the c-Rel subunit of NF-κB or cmyc overrides apoptosis induced by TGF-β1 (186), revealing the functional importance of these two targets. In WEHI 231 cells, TGF-β also induces expression of the BH3-only protein Bim (187), which is a proapoptotic member of the Bcl-2 family. It remains to be established whether downregulation of NF-κB and c-myc causes induction of Bim or whether they represent separate apoptotic pathways. TGF-β induces apoptosis in B lymphocyte progenitors, which is mediated by TGF-β target gene Id3 (157). Interestingly, both c-myc and Id3 regulate B cell proliferation. Therefore, the antiproliferation pathway and the apoptotic pathway may overlap signiﬁcantly. Mice with a B cell blockade of TGF-β signaling have increased peritoneal B-1 cells (21), a unique B cell subset responsible for producing natural Igs. However, B-1 cells show reduced BrdU labeling (21), suggesting that increased B-1 cell pools are mediated by enhanced survival of these cells. The B cell receptor repertoire of B-1 cells recognizes shared microbial structures and/or selfantigens (188, 189). TGF-β signaling may therefore be involved in B cell negative selection to eliminate high-afﬁnity self-reactive B-1 cell clones.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

B cell production of TGF-β. Resting human B cells express TGF-β1 mRNA but secrete low amounts of TGF-β1 protein (158). Upon stimulation with Staphylococcus aureus Cowan, B cells produce high amounts of TGF-β1 protein with little alteration of TGF-β1 mRNA (158). Polyclonal stimulation of B cells with LPS or anti-IgM antibody also induces TGF-β1 production. Neutralization of TGF-β1 from LPS-stimulated 112

Li et al.

B cells results in a signiﬁcant decrease of IgG2a, IgG3, IgG1, and IgE secretion, but not IgM secretion (190). Restoration of TGFβ1 revealed that low doses of TGF-β1 promote, whereas high doses inhibit, antibody secretion (190). These observations suggest that autocrine TGF-β1 may stimulate B cells to secrete antibody. The signiﬁcance of this pathway remains unknown in vivo. Stimulation of human B cells with anti-CD40 antibody induced TGF-β production and IgA class switching, which was ablated by anti-TGF-β antibody (191), suggesting that B cell–produced autocrine TGF-β may promote IgA class switching. Studies of autoimmune MRL/lpr mice have revealed that a fraction of B cell– produced TGF-β1 associates with IgG autoantibodies (192). The antibody-bound TGF-β1 is consistently more active than the uncomplexed TGF-β1 in vivo and is responsible for the increased susceptibility of these mice to bacterial infection (192). In light of the potent suppressive function of TGF-β1 in autoimmunity (see below), increased TGFβ1 levels in MRL/lpr mice might represent a negative feedback mechanism to attenuate disease development.

TGF-β and Natural Killer Cells Natural killer (NK) cells are lymphoid cells that participate in innate immunity and in early defense against a variety of infections and cancers. NK cells express cell surface activating and inhibitory receptors that allow them to respond to microbial products, cytokines, stress signals, and inducible molecules expressed in cancer cells (193, 194). Thus, without requiring prior activation or immunization, NK cells recognize and kill infected or tumorigenic cells and rapidly produce chemokines and cytokines. IFN-γ produced by NK cells is considered essential for stimulation of Th1 responses (22, 195). TGFβ is a potent inhibitor of NK cell functions via attenuating its cytolytic activity and IFNγ production (196–199). The importance of

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

TGF-β signaling in NK cells is supported by a recent study of transgenic mice in which TGF-β signaling is blocked in both NK cells and dendritic cells (DCs) (22). Interestingly, a selective inhibitory function of TGF-β in NK cell homeostasis was also observed (22). TGF-β therefore controls the homeostasis of NK cells and suppresses their cytokine production and cytolytic activity.

and the downregulation of NKp30 results in inhibition of DC killing by TGF-β1treated NK cells (206). Downmodulation of NKG2D, which is associated with elevated levels of TGF-β1, was also observed in cancer patients (207). Thus, by regulating the expression of activating receptors on NK cells, TGF-β can inﬂuence their cytolytic functions.

TGF-β regulation of cytokine production by NK cells. TGF-β is a potent antagonist of IL-12-induced IFN-γ expression in NK cells (199, 200), and its expression coincides with the downregulation of NK cell responses during viral infection (201). TGF-β also inhibits IFN-γ production by human NK cells stimulated with IL-2 (202, 203). A recent study further demonstrated an essential role for TGF-β in inhibiting NK IFN-γ expression in vivo (22), also showing that NK cells rather than DCs are the major innate sources of IFN-γ. Stimulation of TGF-β-resistant NK cells results in higher IFN-γ production, which promotes Th1 differentiation (22). The mechanisms whereby TGF-β antagonizes NK cell production of IFN-γ are unknown. TGF-β downregulates the expression of IFN-α receptor and the α chain of the IL-2 receptor on human NK cells (198, 202), which might be involved in TGF-β inhibition of cytokine signaling pathways in NK cells.

NK cell production of TGF-β. NK cells have been shown to constitutively produce active and latent TGF-β1 (208). In addition, anti-CD2 antibody stimulation strongly enhances CD3− CD56+ NK cell production of active TGF-β1 (208). Stimulation of NK cells with anti-CD69 antibody induces TGF-β1 production (150). Signiﬁcantly, CD69−/− mice mount enhanced NK responses to tumors (150). Thus, NK cells may represent an important in vivo source of TGF-β1 to inhibit immune responses.

DCs are professional antigen-presenting cells (APCs) with an essential role in immunity and tolerance (209). TGF-β1 is required to support the development of Langerhans cells (LCs), resident DCs present within epithelial cells in the epidermis. TGF-β also regulates the maturation of differentiated DCs and DCmediated T cell responses (210).

TGF-β regulation of cytolytic function of NK cells. The ﬁrst demonstration of TGFβ inhibition of NK cell cytolytic function, in which TGF-β suppressed cytolytic activity of IFN-α-stimulated NK cells, was made two decades ago (198). Targeted killing by NK cells depends on the engagement of activating receptors and coreceptors on NK cells, among which the NK cell–speciﬁc NKp46, NKp30, and NKp44, collectively termed natural cytotoxicity receptors (204), and NKG2D (205) appear to play a major role in NK cell– mediated cytotoxicity. TGF-β1 inhibits expression of NKp30 and NKG2D receptors,

TGF-β regulation of Langerhans cell development. In vitro, TGF-β1 is required for the differentiation of LCs from hematopoietic precursor cells in serum-free medium supplemented with GM-CSF, TNF-α, and stem cell factor (211). In the absence of TGF-β1, monocytes are generated (211). The addition of TGF-β1 to GM-CSF- and IL-4supplemented cultures of human monocytes induces the expression of LC markers such as E-cadherin and CD1a (212). These studies and others (213, 214) establish that TGF-β1 promotes LC generation from monocytes. TGF-β1 stimulation of LC development

TGF-β and Dendritic Cells

www.annualreviews.org • Regulation of Immune Responses by TGF-β

113

ARI

23 February 2006

15:9

from hematopoietic progenitor cells is also associated with protection of DC progenitor cells from apoptosis (215). TGF-β1 therefore regulates both cell differentiation and cell survival during LC development. The importance of TGF-β1 in LC development is underscored by the observation that TGF-β1−/− mice lack LCs (216). Transfer of TGF-β1-deﬁcient bone marrow cells into lethally irradiated recipients results in the differentiation of donor LCs (217), demonstrating that paracrine sources of TGF-β1 are sufﬁcient to support LC development. Conversely, reconstitution of SCID TGF-β1deﬁcient mice with TGF-β1 heterozygous bone marrow also leads to the generation of LCs (218). These observations suggest that TGF-β1 functions via both paracrine and autocrine mechanisms to induce LC differentiation. Id2, a helix-loop-helix transcription factor, is strongly upregulated during differentiation of DCs, including LCs (219). TGF-β1 induces Id2 expression in DCs (219). Signiﬁcantly, Id2−/− mice lack LCs (219), suggesting that Id2 is an essential TGF-β1 target in promoting LC differentiation.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

TGF-β regulation of DC maturation and function. DCs differentiated from human hematopoietic progenitor cells or monocytes in the presence of TGF-β1 express predominantly intracellular MHC class II, low levels of CD1d, and costimulatory molecules including CD80, CD83, and CD86 (220– 222), features characteristic of in vivo LCs. These observations are consistent with murine studies in which TGF-β1 inhibits maturation of DCs differentiated from bone marrow cells with GM-CSF (223), demonstrating that TGF-β1 promotes the generation of DCs with an immature phenotype. In astrocytes, TGF-β1 inhibits the transcription of class II transactivator (CIITA) and consequently MHC class II expression, which depends on the presence of Smad3 (224). Whether TGF-β uses the same pathway to regulate MHC class II expression in DCs remains to be tested. 114

Li et al.

TGF-β also regulates the antigenpresentation function of differentiated DCs in vitro. Stimulation of DCs with bacterial components, inﬂammatory cytokines, or costimulatory receptors upregulates MHC class II and costimulatory molecules and cytokines such as IL-12, which induce T cell activation and differentiation. Inclusion of TGF-β in LPS-stimulated DC culture inhibits the expression of MHC class II and costimulatory molecules, which attenuates the antigenpresentation function of DCs (221). In addition, DC maturation and IL-12 production induced by inﬂammatory cytokines IL-1 and TNF-α are inhibited by TGF-β (221). In contrast, CD40 ligand or anti-CD40 antibody-induced DC maturation or cytokine production is not affected by TGF-β (221, 225). CD40 ligand represents a key T helper signal for DC maturation. Lack of TGF-β inhibition of CD40 ligand–induced DC maturation implies that TGF-β1 primarily suppresses noncognate DC activation. DC production of TGF-β. Immunostaining of epidermal sections revealed that TGFβ1 is strongly associated with LCs (226). DCs differentiated from human CD34+ cord blood cells also express high levels of TGFβ1 mRNA (227). In mouse, TGF-β1 is highly expressed in bone marrow–derived CD11c+ CD86− immature DCs (228). The function of DC-produced TGF-β1 remains unknown. Under steady state, presentation of antigens by immature DCs results in T cell unresponsiveness (229). It is conceivable that DC TGF-β1 may act on T cells in a paracrine manner to induce T cell tolerance. DC TGFβ1 may also be involved in the maintenance of the immature state of DCs via an autocrine pathway. To this end, it is interesting to note that TGF-β1 is expressed at low levels in bone marrow–derived CD11c+ CD86+ mature DCs (228).

TGF-β and Macrophages Macrophages (Ms) are professional phagocytes with a critical role in clearing

ANRV270-IY24-04

ARI

23 February 2006

15:9

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

apoptotic cells (230). They also phagocytose microbes and produce inﬂammatory mediators upon microbial challenge. In addition, Ms present antigens to T cells and regulate adaptive immunity. TGF-β regulation of the monocyte/M lineage appears to depend on the differentiation stage of the cells. Generally, TGF-β stimulates cells at the resting state (monocytes), whereas activated cells (Ms) are inhibited (231). TGF-β regulation of monocyte recruitment and production of inflammatory mediators. TGF-β recruits monocytes to the site of injury or inﬂammation via multiple mechanisms: It acts as a chemoattractant for monocytes (232–234); it induces adhesion molecules, including LFA-1 and the ﬁbronectin receptor on monocytes, enabling their attachment to extracellular matrix (235, 236); and it induces matrix metalloproteinases (MMPs), which can dissolve vascular membranes and facilitate monocyte transmigration (235). In addition, TGF-β potentiates inﬂammation through induction of IL-1, IL-6, and leukotriene C4 synthase in monocytes (232, 237, 238). These observations reveal a proinﬂammatory function for TGF-β on monocytes. TGF-β regulation of macrophage phagocytosis. Once monocytes differentiate into Ms, TGF-β functions mostly as an inhibitory molecule. Ms express surface receptors involved in the phagocytosis of bacteria as well as of aged or apoptotic cells. CD36 and SR-A are two such scavenger receptors, the expression of which is downregulated by TGF-β (239, 240). The ability of Ms to phagocytose microbes is greatly enhanced by IgG opsonization. Expression of two of the IgG receptors, FcγRI and FcγRIII, and of the common γ-subunit is downregulated by TGF-β in murine myeloid cells and in human THP-1 cells, which results in reduced phagocytosis of IgG-coated particles (241).

TGF-β regulation of macrophage activation. Stimulation of Ms with IFN-γ and/or microbial products such as LPS induces their activation. In vitro, TGF-β inhibits the expression of several LPS-induced inﬂammatory mediators such as TNF-α and MMP12 as well as chemokines including MIP-1α and MIP-2 (242, 243). Reactive oxygen and nitrogen species produced by activated Ms are important mediators against invading microbes. TGF-β downregulates the production of nitric oxide (NO) and superoxide ion and inhibits the expression of inducible NO synthase (iNOS) in activated Ms (244, 245). These inhibitory activities of TGF-β on Ms may resolve inﬂammation and prevent the development of immunopathology. Toll-like receptors (TLRs) are innate immune receptors involved in the recognition of microbes (246). LPS is recognized by TLR4 with the assistance of CD14. TGF-β inhibits CD14 expression in LPSstimulated Ms and may therefore attenuate TLR4 signaling (247). MyD88 is a key adapter molecule downstream of TLRs. TGF-β inhibits MyD88-dependent TLR signaling pathways in Ms by promoting its degradation (248). Overexpression of Smad3 blocks LPS-induced iNOS and MMP-12 promoter activity in Ms, whereas a dominantnegative mutant of Smad3 alleviates the TGF-β inhibition, suggesting that Smad3 is a critical mediator of TGF-β inhibition of M activation (243). TGF-β regulation of macrophage antigen presentation. IFN-γ-induced expression of MHC class II molecules in Ms is inhibited by TGF-β via the attenuation of CIITA (249). In addition, TGF-β inhibits expression of the costimulatory molecule CD40 and the inﬂammatory cytokine IL-12p40, which collectively results in the inhibition of the antigenpresentation function of Ms (250, 251). This inhibition may play an important role in resolving an ongoing immune response by diminishing secondary stimulation of T cells at the site of infection.

www.annualreviews.org • Regulation of Immune Responses by TGF-β

115

ANRV270-IY24-04

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

PMN: polymorphonuclear neutrophils

15:9

Macrophage production of TGF-β. It has been reported that phagocytosis of apoptotic cells leads to TGF-β1 secretion, which inhibits the production of inﬂammatory cytokines and chemokines in Ms including IL1β, TNF-α, GM-CSF, and IL-8 (252, 253). Exposure of phosphatidylserine on the outer leaﬂet of the plasma membrane represents a key signal for the recognition of apoptotic cells by Ms (254). Liposomes containing phosphatidylserine induce TGF-β1 in Ms in vitro (255), and instillation of apoptotic cells or phosphatidylserine-containing liposomes enhances TGF-β1 production associated with reduced inﬂammation in mice, illustrating the in vivo relevance for this pathway (256, 257). Infection of Ms with certain microbes triggers the production of TGF-β1, which modulates the host responses (see below). TGF-β1 production in Ms is additionally regulated by other cytokines: IL-9, a Th2 cytokine, promotes TGF-β1 production in LPS-stimulated human peripheral blood monocytes (258); and the p40 subunit of IL-12 inhibits TGF-β1 production in Ms, as p40−/− Ms secrete a large amount of TGF-β1 (259).

TGF-β and Mast Cells Mast cells are recognized mostly for their role in allergy and asthma, but they are also important in host defense against pathogens and in autoimmune diseases. They act mainly through the release of effector molecules such as histamine, proteases, and TNF-α. TGF-β has a stimulatory effect on the initial phase of mast cell–dependent immune responses and may also regulate their functions at later phases. TGF-β regulation of mast cell recruitment and activation. TGF-β is a potent chemoattractant for mast cells (260, 261). In murine bone marrow–derived mast cells, TGF-β induces integrin α7 expression, which facilitates their intraepithelial

116

Li et al.

migration through binding to laminin-1 (262). Mast cells are activated mainly through the cross-binding of IgE to FcεR. TGF-β attenuates the expression of FcεR receptor in cultured human and mouse mast cells by blocking protein translation (263). Data regarding the effects of TGF-β on mast cell proliferation, histamine release, and the secretion of TNFα and other inﬂammatory mediators are contradictory, possibly due to heterogeneity of mast cells derived from different species and anatomical sites (264–268). Mast cell production of TGF-β. Early studies showed that mast cells or cell lines produce TGF-β1, which induces collagen synthesis in ﬁbroblasts (269, 270). The Th2 cytokine IL-9 also induces TGF-β1 production from peritoneal mast cells or D36 mast cells (264, 268). Mast cells produce chymase-1 that activates TGF-β1. Both molecules are stored in secretory granules, which implies that mast cell stimulation and degranulation may result in both the secretion and the activation of TGF-β1 (271).

TGF-β and Granulocytes Polymorphonuclear neutrophils (PMN) are professional phagocytes essential for ﬁrstline defense against microorganisms (272). Together with proteases liberated from the granules, reactive oxygen species derived from superoxide (which is generated after PMN activation) is used to kill ingested microbes. Eosinophils are granulocytes known for their role in immune responses to infection with helminthic parasitic worms (273). They secrete inﬂammatory mediators, including lipid mediators, cationic proteins, cytokines, chemokines, and components of the oxidative burst, all of which can also cause tissue damage in inﬂammatory diseases (274). Many of the eosinophil-speciﬁc mediators act directly on mast cells and aggravate allergic inﬂammation.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

TGF-β regulation of granulocyte recruitment. TGF-β is a potent chemoattractant for PMN, acting at femtomolar levels (275, 276), and may play an important role in recruiting cells to sites of infection or injury. However, TNF-α-induced endothelial cell expression of IL-8, another PMN chemoattractant, is inhibited by TGF-β (277). TGF-β also downregulates the expression of Eselection and VCAM-1 in endothelial cells (278, 279) and inhibits PMN transmigration in TNF-α-activated endothelial monolayers (279). It appears that TGF-β acts initially to stimulate an immune response by recruiting PMN, but when inﬂammation is ongoing with secretion of proinﬂammatory mediators, TGF-β limits the inﬂux of leukocytes through other mechanisms. An alternative interpretation is that the chemotactic property of TGF-β is important for PMN movement in tissues, whereas it inhibits their migration across the endothelium (279). Smad3−/− PMN are impaired in their chemotactic responses toward TGF-β (16), which is associated with chronic mucosal infections in these mice. Therefore, TGF-β-mediated PMN chemotaxis may represent an important host defense mechanism. TGF-β has also been demonstrated to act as a chemoattractant for human eosinophils (280). The in vivo signiﬁcance of this pathway remains to be determined. TGF-β regulation of granulocyte function. TGF-β regulation of PMN activation and effector functions such as phagocytosis, degranulation, and respiratory burst is controversial and requires further investigation (276, 281, 282). IL-13 plays a crucial role in activation and survival of eosinophils, and surface expression of the IL-13Rα1 is enhanced by TGF-β in vitro (283). These observations suggest a stimulatory function for TGF-β on eosinophil-mediated inﬂammation. In contrast, TGF-β reduces the viability and the release of peroxidase in eosinophils isolated from peripheral human blood (284). These ﬁndings imply a complex

role for TGF-β in the regulation of eosinophil functions. Granulocyte production of TGF-β. PMN-derived TGF-β1 may play a role in synovial effusions (285). TGF-β1 produced by eosinophils may be involved in ﬁbrosis and the remodeling of asthmatic airways by inducing collagen and other extracellular matrix proteins by smooth muscle cells (SMCs) and ﬁbroblasts (274). IL-5 is an important cytokine for eosinophil differentiation, activation, and recruitment. IL-5 induces TGF-β1 production in eosinophils in vitro (286). Eosinophils from IL-5 transgenic mice concordantly secrete large amounts of TGFβ1 (286). Consistent with these ﬁndings, deﬁciency of IL-5 or IL-5R or treatment with anti-IL-5 antibody leads to a parallel reduction of eosinophils, TGF-β1, and ﬁbrotic lesions in models of airway remodeling (287, 288). These observations suggest that IL-5-induced TGF-β1 in eosinophils may play a critical role in promoting ﬁbrosis in asthma.

Chemotaxis: a phenomenon in which mobile cells direct their movement to the concentration gradient of chemicals. Chemotaxis is crucial for the recruitment of inﬂammatory cells to sites of infection or injury

TGF-β AND DISEASE TGF-β Regulation of Autoimmune Diseases Autoimmune diseases, including more than 70 chronic disorders affecting 3%–5% of the population, are characterized by loss of immunological tolerance to self-antigens. Studies of animal models and human patients have revealed a critical function for TGF-β in regulating leukocyte functions in autoimmune diseases (289). Systemic lupus erythematosus. Systemic lupus erythematosus (SLE) is a B and T cell autoimmune disorder characterized by B cell hyperactivity and activation of autoreactive T cells. Autoantibodies and immune complexes, together with pathogenic T cells, induce pathology in multiple organs, including skin, blood vessels, lung, joints, and kidney.

www.annualreviews.org • Regulation of Immune Responses by TGF-β

117

ARI

23 February 2006

15:9

Interestingly, TGF-β1−/− mice develop immunopathology resembling that of SLE, including formation of autoantibodies and IgG deposits in their renal glomeruli (14, 15, 290, 291). Consistent with this ﬁnding, decreased production of TGF-β1 was found in some SLE patients (292). Nonetheless, decreased production of active TGF-β in SLE patients does not correlate with disease severity (293). SLE is a multifactorial disease with genetic and environmental factors. MRL/lpr mice develop systemic autoimmune disease mimicking human lupus. It is paradoxical that these mice have elevated levels of circulating TGF-β1 (294). Increased TGFβ1 suppresses T cell and PMN functions, which may account for the susceptibility to infection of these mice (192, 294). Interestingly, overexpression of TGF-β1 in MRL/lpr mice inhibits the development of SLE (295). These observations suggest that TGF-β1 elaboration during evolution of autoimmunity in these mice may represent a negative feedback mechanism of the host to restore self-tolerance.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

Rheumatoid arthritis. Rheumatoid arthritis (RA) is a T cell–dependent, antibodymediated inﬂammatory disease that targets the joints. Collagen-induced arthritis (CIA) is an experimental model sharing several clinical features with RA (296). Regulation of RA by TGF-β1 was shown to be site and context dependent, owing to its pleiotropic activities. Systemic administration of TGFβ1 to mice inhibits CIA (297), whereas its local administration to joints induces synovitis and aggravates disease (234). Similarly, blocking endogenous TGF-β1 by the systemic injection of anti-TGF-β antibody exacerbates CIA in mice (297), whereas the local blockade of TGF-β1 ameliorates ongoing inﬂammation (298). TGF-β signaling in T cells regulates CIA because inhibition of TGF-β signaling in T cells renders mice more susceptible to the disease; this is associated with increased production of IFN-γ 118

Li et al.

and TNF-α (299). Inhibition of T cell activation and differentiation by endogenous TGFβ is therefore required for the maintenance of joint integrity. CD69 is an early activation molecule expressed in different leukocyte subsets. CD69 cross-linking induces TGFβ1 production in T cells as well as in NK cells and Ms (151). Signiﬁcantly, CD69−/− mice develop severe CIA associated with decreased TGF-β1 production (151). These studies suggest that the CD69 pathway is required to inhibit CIA by regulating TGF-β1 production. Insulin-dependent (type 1) diabetes mellitus. Type 1 diabetes (T1DM) is an autoimmune disease with a progressive T cell– mediated destruction of insulin-producing β cells in the pancreatic islets. NOD mice develop spontaneous T1DM resembling human disease. Several laboratories have developed transgenic models to study TGF-β1 regulation of T1DM in NOD mice. Constitutive expression of TGF-β1 in β cells under the rat insulin promoter inhibits T1DM, which investigators propose is caused by immune deviation from Th1 to Th2 responses (300) and increased T cell apoptosis (301). Expression of TGF-β1 in α cells with the transgene under the control of rat glucagon promoter also blocks disease development (302). In addition, these mice are resistant to passive transfer of disease with diabetogenic T cells (302), which demonstrates that paracrine TGF-β1 is sufﬁcient to protect β cells from killing by CD4+ and/or CD8+ effector T cells. Overproduction of TGF-β1 in the islets induces ﬁbrosis and may complicate the analysis of disease development (301). Expression of TGF-β1 in a temporally controlled inducible system during either preinsulitis, or priming, or the destruction phase of the disease protects mice from T1DM (120). Signiﬁcantly, overproduction of TGF-β1 expands FoxP3expressing Tregs in the islets, which inhibits T1DM (120). These observations suggest that TGF-β may use multiple mechanisms to inhibit disease development.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

Endogenously produced TGF-β1 also regulates autoimmune T1DM. Treatment with nonmitogenic anti-CD3 antibody induces long-term remission of T1DM in NOD mice (303). Interestingly, the protective effect is abrogated by neutralizing anti-TGF-β antibody (303). T cells are likely a key source of TGF-β, as anti-CD3 antibody treatment stimulates CD4+ T cells to produce TGFβ1 (303). A humanized anti-CD3 antibody is currently in clinical trial for autoimmune T1DM (304). It will be interesting to investigate whether the same TGF-β1-dependent mechanism is also involved in inhibiting human disease. Multiple sclerosis. Multiple sclerosis (MS) is an autoimmune disorder affecting the central nervous system. Experimental autoimmune encephalomyelitis (EAE) serves as an experimental model for MS. Early studies showed that administration of exogenous TGF-β1 to mice or treatment of myelin basic protein (MBP)-speciﬁc T cells with TGF-β is able to prevent or inhibit EAE (305–307). Increased expression of TGF-β1 mRNA or protein is associated with remission of the disease (308, 309). Administration of neutralizing antibody to TGF-β also enhances the clinical severity of the disease (310). Together, these studies suggest that endogenous TGF-β1 is involved in protection from EAE. The sources and targets of TGF-β1 in the regulation of EAE remain to be deﬁned. TGF-β1 is associated with the induction of oral tolerance to EAE antigens. Oral administration of high doses of MBP results in clonal deletion or anergy, whereas treatment with low doses of MBP induces suppressive T cells (114). These T cells, named Th3 cells, secrete high amounts of TGF-β1 and protect mice from EAE (114). Anti-TGFβ1 antibody treatment abrogates the protection (114), suggesting that TGF-β1 functions as an effector cytokine of Th3 cells. Nasal administration of MBP also induces TGFβ1-producing Tregs in rats (311), which prevents the relapse of EAE. These observations

demonstrate a critical function for TGF-β1 in the induction of tolerance to antigens administrated via the mucosal route.

TGF-β Regulation of Tumor Immunity Cancers result from accumulated genetic and epigenetic changes in tumor cells, which confer growth advantages. At least three stages are involved in cancer development: (a) induction, the accumulation of mutations induced by carcinogens; (b) promotion, a process triggered by mechanical or chemical injury that is not mutagenic by itself; and (c) progression, the invasive growth that results in aggressive metastasis (312, 313). Most tumor cells are clonal in origin, and during cancer progression new subpopulations arise with alterations in their morphology, hormone dependence, migration property, and proteome composition (314, 315). Consequently, most tumors acquire new antigens. Their invasive activity also fosters an inﬂammatory environment, which potentiates antitumor immune responses involving both the innate arm (NK cells, PMN, and Ms) and the adaptive arm (CD8+ and CD4+ T cells) of the immune system (316, 317). During their evolution, tumors that produce increased TGF-β or promote TGFβ production by surrounding cells not only advance tumor progression but also allow tumors to evade immune surveillance (318– 320). A recent study showed that RNAi inhibition of TGF-β expression in glioma cells restores NKG2D expression in NK and CD8+ T cells and that the expression of the NKG2D ligand MICA in tumor cells is associated with decreased tumor growth in mice (321). On the other hand, overexpression of TGFβ1 in a highly immunogenic murine tumor suppresses the antitumor immune response (322). These observations demonstrate that TGF-β produced by tumor cells is involved in their immune evasion. Studies with CD4-DNR mice reveal that TGF-β is required for suppressing T cell–mediated tumor

www.annualreviews.org • Regulation of Immune Responses by TGF-β

119

ARI

23 February 2006

15:9

rejection (109). These mice were resistant to challenge with two highly aggressive tumors, EL-4 and B16-F10 (109). TGF-β signaling in CD8+ but not in CD4+ T cells was essential for tumor immune evasion (109). Tregs have been shown to promote tumorigenesis (323), and abrogation of Tregs resulted in enhanced antitumor immunity (324–326). Interestingly, the inhibitory function of Tregs requires TGF-β signaling in CD8+ T cells (327). In addition to the inhibition of effector T cell functions, TGF-β may promote tumor growth through modulating Tregs. Most tumor cells produce large amounts of TGF-β that may induce Treg differentiation. In support of this notion, a recent study demonstrated that FoxP3-expressing Tregs increase in the peripheral blood of tumorbearing patients (328) and decrease following therapy (329). It will therefore be interesting to study whether TGF-β signaling is involved in the generation of tumor-promoted Tregs.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

TGF-β Regulation of Atherosclerosis Atherosclerosis is an inﬂammatory disease that is a response to inﬁltration, retention, and oxidation of low-density lipoprotein (LDL) in the arterial intima beneath the endothelial cell layer (330). Lipid-laden M-derived foam cells as well as collagen-producing SMCs comprise the bulk of the lesion. Inﬂammatory cells, including activated T cells and APCs such as Ms and mature DCs, affect both the size and vulnerability of the lesion. Vulnerable plaques with little collagen, high lipid content, and a high density of inﬂammatory cells are prone to rupture and may have detrimental consequences such as formation of thrombi, which may lead to myocardial infarction or stroke (330). IFN-γ, TNF-α, IL-18, and IL-12 secreted by inﬂammatory cells within the lesion promote the formation of vulnerable plaques (331, 332). Neutralization of TGF-β increased lesion vulnerability in mouse mod120

Li et al.

els, supporting an antiatherosclerotic function of TGF-β (333, 334). Apolipoprotein E−/− mice bred with CD4-DNR mice exhibited a dramatic increase of large vulnerable lesions, despite decreased cholesterol levels in serum (335). IFN-γ mRNA increased more than 100-fold in the aortas, concomitant with less collagen, more T cells, and increased expression of MHC class II in the lesions (335). Reconstitution of LDL receptor−/− mice with bone marrow from CD2-DNR mice similarly led to the formation of vulnerable lesions but with minimal effects on the lesion size (336). Thus, T cells are central targets of the antiatherosclerotic property of TGFβ. Proatherogenic T cells likely drive atherogenesis by affecting endothelial cells, Ms, and SMCs. Inhibition of T cells by TGFβ dampens these atherogenic mechanisms (335). In addition, TGF-β may directly target endothelial cells, DCs, Ms, and SMCs to regulate atherogenesis. In endothelial cells, TGF-β inhibits expression of adhesion molecules and chemokines that partake in recruitment of leukocytes (278, 279). The expression of the scavenger receptors CD36 and SR-A in Ms are downregulated by TGF-β, and cholesterol efﬂux is increased through TGF-β regulation of ATP-binding cassette transporter-1 (ABC1) (239, 240, 337). Collectively, this may lead to reduced foam cell formation. Because TGF-β reduces antigen presentation by Ms and DCs, disease-promoting adaptive immune responses against proatherogenic antigens are likely inhibited by TGF-β. In addition, the formation of stable lesions may be promoted by TGF-β induction of collagen and TIMP (tissue inhibitor of matrix metalloproteinase) synthesis in SMCs (338, 339). Therefore, TGF-β may target multiple cell types to prevent the development of vulnerable atherosclerotic plaques. Interestingly, higher levels of TGF-β1 are present in human asymptomatic lesions than in lesions from symptomatic patients (340). Lower serum levels of active TGF-β1 have also been found in patients with advanced atherosclerosis

ANRV270-IY24-04

ARI

23 February 2006

15:9

(341), suggesting that TGF-β1 may regulate atherogenesis in human patients.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TGF-β Regulation of Asthma Asthma is a complex inﬂammatory disease dependent on Th2-type immune responses and is characterized by airway hyperresponsiveness (AHS). In a murine model of asthma, overexpression of TGF-β1 in ovalbumin (OVA)-speciﬁc Th cells abolishes AHS and airway inﬂammation induced by OVAspeciﬁc Th2 cells (342). A recent study of TGF-β1 heterozygous mice, which produce lower levels of TGF-β1 than wild-type mice, showed exacerbated asthmatic immunopathology (343). These studies demonstrate that endogenous TGF-β1 is an important suppressor of asthma in mice. T cells appear to be a key target of TGF-β1 in this model. Transgenic expression of Smad7 under the control of a distal Lck promoter blocks TGF-β signaling in mature T cells (20). Signiﬁcantly, both OVA-induced AHS and airway inﬂammation is enhanced in the transgenic mice associated with high production of Th2 cytokines (20). Whether TGF-β regulates human asthma remains to be established. Levels of TGF-β1 in the airways of asthmatics were found higher than in normal controls (344). It is conceivable that elevated TGF-β1 in these patients may represent a negative feedback mechanism to control airway inﬂammation. Because TGFβ1 is important in healing, increased TGF-β1 production may be involved in the repair of asthmatic airways. On the other hand, TGFβ1 may also induce ﬁbrosis to exaggerate disease development (345).

TGF-β Regulation of Inflammatory Bowel Disease The inﬂammatory bowel diseases (IBD), comprising Crohn’s disease and ulcerative colitis, are chronic inﬂammatory diseases of the gastrointestinal tract (346). Development of IBD is often a consequence of a loss of tolerance to-

ward commensal bacterial ﬂora (346). Blockade of TGF-β signaling in T cells (CD4DNR) results in T cell activation and induction of IBD in mice (18), demonstrating an essential role for TGF-β signaling in T cells in inhibiting colitis. In a murine model of colitis, protection from IBD induced by transfer of CD4+ CD45RBhigh T cells into SCID mice by CD4+ CD45RBlow or CD4+ CD25+ Tregs is abrogated by administration of neutralizing anti-TGF-β antibody (144). Consistent with this, colitis induced by transfer of CD4+ CD45RBhigh T cells isolated from CD4-DNR mice escapes control by Tregs (124). The cellular sources of TGF-β1 in regulating colitis remain controversial. In a recent study, CD4+ CD25+ cells isolated from TGFβ1−/− mice failed to prevent colitis (145), suggesting an essential function for Treg TGF-β1. In another study, Tregs isolated from TCR transgenic TGF-β1−/− mice were capable of suppressing colitis (124). However, suppression was still abrogated by antiTGF-β antibody (124), demonstrating that TGF-β produced by cells other than Tregs is also involved in inhibiting colitis. In light of the mouse studies, it seems paradoxical that TGF-β1 levels are elevated in the gut of patients with IBD (347). Interestingly, T cells from these patients are refractory to the inhibitory activity of TGF-β1, which is associated with increased expression of Smad7 (348), a negative regulator of TGF-β signaling. These observations suggest that defective TGF-β signaling may be involved in human IBD.

TGF-β Regulation of Infectious Diseases Copious evidence suggests that TGF-β controls many interactions between pathogens and their hosts. TGF-β has the potential to inﬂuence all phases of pathogenesis and plays an important role in the regulation of immune responses to pathogens, inﬂammation, and wound healing. TGF-β is also a common

www.annualreviews.org • Regulation of Immune Responses by TGF-β

121

ARI

23 February 2006

15:9

molecule exploited by pathogens to facilitate entry, replication, and persistence in the host. In general, most M pathogens have evolved mechanisms to induce TGF-β production, and TGF-β production in turn suppresses the killing activity of Ms, enhances intracellular proliferation of the pathogen, and thus favors virulence. However, in some organisms, TGF-β plays a protective role in host resistance against infection. Therefore, the role of TGF-β in the resolution or exacerbation of infection depends on the nature of the immune response necessary to control parasite replication and disease.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

Parasites Trypanosome. Trypanosoma cruzi is one parasite that makes direct use of the host’s TGF-β signaling pathway. It uses TGF-β receptors I and II for successful entry into mammalian cells (349). Epithelial cells lacking TGF-β receptors are resistant to T. cruzi infection, and infectivity is restored following transfection of functional TGF-β receptors (349). T. cruzi can also induce expression of a TGF-β-responsive reporter gene, suggesting that the parasite might directly activate the TGF-β signaling pathway to facilitate entry into mammalian cells (349). Spleen cells from T. cruzi–infected mice have higher levels of bioactive TGF-β (350). TGF-β treatment of mouse and human Ms blocks IFNγ-mediated inhibition of parasite growth, and TGF-β-treated mice develop higher parasite loads and die faster than control mice (350). An alpha-fetoprotein homolog of T. cruzi can induce TGF-β from mouse splenocytes (351). The African trypanosome T. brucei, which does not invade host cells, might nevertheless possess a functional homolog of this TGF-β-activating moiety because this parasite has also been shown to release a factor that induces TGF-β mRNA expression (352). Leishmania. Leishmania can survive and replicate in the phagolysosome of Ms. Mice 122

Li et al.

resistant to Leishmania major develop a Th1type response with high levels of IL-2 and IFN-γ, whereas susceptible strains of mice develop a Th2-like response with accompanying increased production of IL-4 and IL-10. NK-produced IFN-γ is responsible for activating Ms to express high levels of iNOS and toxic NO, a molecule with the potent ability to kill intracellular parasites (353). Treatment with anti-IFN-γ abrogates the resistance to L. major in the C3H/HeN mouse strain, whereas administration of IFN-γ offers protection in the susceptible BALB/c strain (354, 355), suggesting that IFN-γ is the most important cytokine for resistance to L. major. Virulent strains of Leishmania have developed mechanisms to induce Ms to produce high levels of active TGF-β, whereas nonpathogenic strains that produce low-grade infection induce relatively low levels of active TGF-β (356). TGF-β made by infected Ms suppresses NO production (244) and can inﬂuence T cell differentiation by inhibiting the production of TNF-α and IFN-γ. Leishmania species can also activate TGF-β using parasite-derived cathepsins to increase local concentrations of active TGF-β to modulate local iNOS and arginase levels (357). Systemic administration of TGF-β increases the infectivity of relatively avirulent strains of Leishmania, and in vivo neutralization of TGF-β is sufﬁcient to inhibit the development of fatal Leishmaniasis in susceptible BALB/c mice (356). When BALB/c mice expressing CD4DNR were challenged with L. major, they arrested infection as effectively as genetically resistant C57BL/6 mice, suggesting susceptibility to L. major infection requires TGF-β signaling in T cells (89). In another study, blockade of TGF-β signaling in NK cells caused the generation of large numbers of NK cells capable of producing considerable IFN-γ, which led to Th1 polarization and protection from L. major infection in BALB/c mice (22). Collectively, these in vivo studies show that TGF-β-mediated inhibition of Th1 differentiation is involved in Leishmaniasis in susceptible mice.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

Toxoplasma gondii. Toxoplasma is another obligate intracellular parasite that infects Ms. As in the case of Leishmania, infection of mouse Ms with T. gondi results in the release of TGF-β, which was associated with the downregulation of TNF-α and its receptors (358, 359). Elevated TGF-β levels were found in spleen cells from SCID mice following exposure to killed T. gondi or after oral exposure to live T. gondi (200). In the same model, TGFβ could inhibit the protective effects of IL-12, and anti-TGF-β antibody partially protected mice from lethal infection (200). From these studies, there appear to be close parallels in how TGF-β regulates infection by a number of protozoan pathogens.

Malaria. In murine malaria infections, TNF-α, IFN-γ, and NO are associated with parasite clearance. TNF-α and IFN-γ act synergistically to activate Ms to phagocytose malaria-infected red blood cells and to release NO. Overproduction of IFN-γ, TNF-α, or LT-α predisposes the host to severe pathology. IL-10 and TGF-β play a key role in limiting malaria pathology (360). When infected mice were treated with neutralizing antibody to TGF-β, the virulence of lethal Plasmodium berghei strains was exacerbated, and a normally resolving chabaudi infection became lethal (361). However, when mice infected with P. berghei were treated with recombinant TGF-β, IL-10 was upregulated, TNF-α was downregulated, parasite replication was slowed, and survival time was extended (361). Mortality is increased in IL-10−/− mice when they are infected with the normally nonlethal P. chabaudi (362), and mortality is further exacerbated by neutralization of TGF-β (363). Moreover, metalloproteinases and a thrombospondin-like molecule from malaria activate TGF-β (364). These studies suggest that TGF-β is associated with protection rather than disease in murine models of malaria.

Bacteria Mycobacteria. Mycobacteria are obligate intracellular pathogens of Ms that cause tuberculosis, leprosy, and opportunistic infections due to immunosuppression. Similar to most protozoan infections, mycobacteria induce M production of active TGF-β and suppress their antibacterial activity to aid their pathogenesis. Both puriﬁed protein derivative and lipoarabinomannan, a cell wall component of tuberculosis, induce TGF-β from human peripheral blood mononuclear cell (PBMC)-derived Ms (365, 366). Listeria. Listeria monocytogenes is a facultative intracellular bacterium and a strong inducer of Th1 response (367). Cytokines such as IFN-γ, TNF-α, and IL-6 play an important role in host resistance to Listeria, and TGF-β plays a protective role in Listeria infection (368). Mice injected with antiTGF-β antibody became susceptible to listeriosis, whereas the administration of human platelet-derived TGF-β1 enhanced their resistance, even though endogenous levels of IFN-γ, TNF-α, and IL-6 were reduced in these mice (368). The mechanism by which TGF-β confers resistance to lethal doses of L. monocytogenes in mice is not yet clear and requires more investigation.

Yeast Candida albicans is an opportunistic fungal pathogen that can cause morbidity and mortality in immunocompromised hosts. An effective Th1 response and production of IFN-γ is required for clearance of C. albicans in mouse models of infection. As in Listeria infection, studies demonstrate that TGF-β plays a beneﬁcial role in acquired resistance against C. albicans infection in mice. Administration of recombinant TGF-β to mice susceptible to C. albicans had a protective effect and delayed the progression of disease (369). In contrast, local TGF-β suppresses

www.annualreviews.org • Regulation of Immune Responses by TGF-β

123

ANRV270-IY24-04

ARI

23 February 2006

15:9

host defense in immunocompromised hosts (370). Taken together, these data suggest that TGF-β may play a protective or suppressive role depending on the host’s immune status.

Virus

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

There are reports of increased TGF-β during many viral infections, such as inﬂuenza, HIV, hepatitis B virus (HBV), hepatitis C virus (HCV), and others. The neuraminidase glycoprotein of inﬂuenza A and B viruses have been shown to directly activate latent TGF-β (371). In the case of HCV, investigators observed enhanced TGF-β production and lack of TGF-β responses (372) and showed that HCV core protein directly upregulates TGFβ1 transcript (373). HCV proteins can also perturb TGF-β signaling via their interac-

tions with Smad molecules (374). Similarly, HIV Tat can induce TGF-β1 in antigenstimulated PBMCs, monocytes, T cells, and astrocytes, and this induction may be in part responsible for the immunosuppressive effects of HIV-1 Tat (375). TGF-β1 also enhances M responsiveness to SDF-1α stimulation and susceptibility to HIV-1 by selectively increasing expression of CXCR4 and enhancing the entry of HIV-1 into human monocyte-derived Ms (376).

SUMMARY AND FUTURE PERSPECTIVES The ﬁrst study of TGF-β on immune cell function was carried out two decades ago (13). TGF-β is now known for its activities on virtually all leukocyte lineages (Figure 5).

Figure 5 Pleiotropic effects of TGF-β on leukocytes. All leukocytes produce and respond to TGF-β. The yin-yang symbol illustrates the fact that TGF-β exerts both stimulatory and inhibitory effects on immune cells. Selected immunological processes regulated by TGF-β are depicted (MC, mast cell; EO, eosinophil; MO/M, monocyte/macrophage). 124

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

The dominant role of TGF-β in the immune system is to induce tolerance, as well as to contain and resolve inﬂammation (Figure 1). TGF-β also positively regulates immune responses by recruiting leukocytes at the initial phase, promoting T cell survival, and inducing IgA class switching in B cells. The activity of TGF-β is inﬂuenced by the environmental milieu. Notably, inﬂammatory cytokines such as IFN-γ and costimulatory receptors including CD28, 4-1BB, and CD40 counteract the suppressive activity of TGF-β. This ensures robust immune responses to pathogens while maintaining tolerance to self or innocuous antigens. Disturbance of this balance results in various immune disorders, and TGF-β appears to be a key regulator of these diseases. What questions remain to be addressed? First, our understanding of the cellular mechanisms of immune regulation by TGF-β remains incomplete. Cell type-speciﬁc inactivation of TGF-β signaling has revealed the activities of TGF-β in T, B, and NK cells.

Similar studies can be extended to other leukocyte lineages. All leukocytes produce at least one form of TGF-β, the function and regulation of which is mostly unknown. Generation of TGF-β conditional knockout mouse models will help address this issue. Second, the molecular mechanisms of TGFβ regulation of immune cell functions need further investigation. A recent study in epithelial cells has revealed versatile interactions between signaling molecules of the TGF-β pathway and those of the other signaling pathways (377). To understand how immune cells read and respond to the combinatorial inputs from TGF-β and environmental cues remains a big challenge. Third, translation of our knowledge on TGF-β to cure immune disorders is largely unexplored. Owing to its pleiotropic activities, future attempts to harness the TGF-β system for therapeutic purposes need to focus on speciﬁc targeting strategies. We are conﬁdent that these exciting areas will enthrall investigators for at least another two decades.

ACKNOWLEDGMENTS Because of space restrictions, we were able to cite only a fraction of the relevant literature. We apologize to any colleagues whose contributions may not be appropriately acknowledged in this review or those we could cite only in secondary references. We are grateful to Elizabeth E. Eynon, Sean T. Kim, and Martin Kriegel for reading the manuscript and for incisive and ¨ helpful comments and to Goran K. Hansson for helpful suggestions. M.O. Li is an American Cancer Society fellow, Y.Y. Wan and S. Sanjabi are Cancer Research Institute fellows, and A.K.L. Robertson is a Wenner-Gren Fellow. R.A. Flavell is an investigator of the Howard Hughes Medical Institute. Work from this laboratory has been supported by grants from ADA (7-03-RA-23) and NIDDK (DK53015, DK51665).

LITERATURE CITED 1. Kappler JW, Roehm N, Marrack P. 1987. T cell tolerance by clonal elimination in the thymus. Cell 49:273–80 2. Goodnow CC, Cyster JG, Hartley SB, Bell SE, Cooke MP, et al. 1995. Self-tolerance checkpoints in B lymphocyte development. Adv. Immunol. 59:279–368 3. Medzhitov R, Janeway CA Jr. 1997. Innate immunity: the virtues of a nonclonal system of recognition. Cell 91:295–98 4. Guerder S, Picarella DE, Linsley PS, Flavell RA. 1994. Costimulator B7-1 confers antigen-presenting-cell function to parenchymal tissue and in conjunction with tumor www.annualreviews.org • Regulation of Immune Responses by TGF-β

125

ANRV270-IY24-04

ARI

23 February 2006

5.

6.

7.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

8. 9. 10. 11. 12. First study of TGF-β on immune cells.

First demonstration of the requirement of TGF-β1 for immune tolerance.

13.

14.

15.

16.

17.

Demonstrated TGF-β regulation of T cell activity in vivo.

Showed TGF-β regulation of B cell activity in vivo.

18. 19.

20.

21. Study of TGF-β regulation of NK cell activity in vivo.

126

22.

15:9

necrosis factor α leads to autoimmunity in transgenic mice. Proc. Natl. Acad. Sci. USA 91:5138–42 Li M, Carpio DF, Zheng Y, Bruzzo P, Singh V, et al. 2001. An essential role of the NF-κB/Toll-like receptor pathway in induction of inﬂammatory and tissue-repair gene expression by necrotic cells. J. Immunol. 166:7128–35 Leadbetter EA, Rifkin IR, Hohlbaum AM, Beaudette BC, Shlomchik MJ, MarshakRothstein A. 2002. Chromatin-IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors. Nature 416:603–7 Sakaguchi S. 2004. Naturally arising CD4+ regulatory T cells for immunologic selftolerance and negative control of immune responses. Annu. Rev. Immunol. 22:531–62 Moore KW, de Waal Malefyt R, Coffman RL, O’Garra A. 2001. Interleukin-10 and the interleukin-10 receptor. Annu. Rev. Immunol. 19:683–765 Letterio JJ, Roberts AB. 1998. Regulation of immune responses by TGF-β. Annu. Rev. Immunol. 16:137–61 Wahl SM. 1994. Transforming growth factor β: the good, the bad, and the ugly. J. Exp. Med. 180:1587–90 Blobe GC, Schiemann WP, Lodish HF. 2000. Role of transforming growth factor β in human disease. N. Engl. J. Med. 342:1350–58 Newfeld SJ, Wisotzkey RG, Kumar S. 1999. Molecular evolution of a developmental pathway: phylogenetic analyses of transforming growth factor-β family ligands, receptors and Smad signal transducers. Genetics 152:783–95 Kehrl JH, Wakefield LM, Roberts AB, Jakowlew S, Alvarez-Mon M, et al. 1986. Production of transforming growth factor β by human T lymphocytes and its potential role in the regulation of T cell growth. J. Exp. Med. 163:1037–50 Shull MM, Ormsby I, Kier AB, Pawlowski S, Diebold RJ, et al. 1992. Targeted disruption of the mouse transforming growth factor-β1 gene results in multifocal inflammatory disease. Nature 359:693–99 Kulkarni AB, Huh CG, Becker D, Geiser A, Lyght M, et al. 1993. Transforming growth factor β1 null mutation in mice causes excessive inﬂammatory response and early death. Proc. Natl. Acad. Sci. USA 90:770–74 Yang X, Letterio JJ, Lechleider RJ, Chen L, Hayman R, et al. 1999. Targeted disruption of SMAD3 results in impaired mucosal immunity and diminished T cell responsiveness to TGF-β. EMBO J. 18:1280–91 Datto MB, Frederick JP, Pan L, Borton AJ, Zhuang Y, et al. 1999. Targeted disruption of Smad3 reveals an essential role in transforming growth factor β-mediated signal transduction. Mol. Cell. Biol. 19:2495–504 Gorelik L, Flavell RA. 2000. Abrogation of TGFβ signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 12:171–81 Lucas PJ, Kim SJ, Melby SJ, Gress RE. 2000. Disruption of T cell homeostasis in mice expressing a T cell-speciﬁc dominant negative transforming growth factor β II receptor. J. Exp. Med. 191:1187–96 Nakao A, Miike S, Hatano M, Okumura K, Tokuhisa T, et al. 2000. Blockade of transforming growth factor β/Smad signaling in T cells by overexpression of Smad7 enhances antigen-induced airway inﬂammation and airway reactivity. J. Exp. Med. 192:151–58 Cazac BB, Roes J. 2000. TGF-β receptor controls B cell responsiveness and induction of IgA in vivo. Immunity 13:443–51 Laouar Y, Sutterwala FS, Gorelik L, Flavell RA. 2005. Transforming growth factorβ controls T helper type 1 cell development through regulation of natural killer cell interferon-γ. Nat. Immunol. 6:600–7

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

23. Chang H, Brown CW, Matzuk MM. 2002. Genetic analysis of the mammalian transforming growth factor-β superfamily. Endocr. Rev. 23:787–823 24. Govinden R, Bhoola KD. 2003. Genealogy, expression, and cellular function of transforming growth factor-β. Pharmacol. Ther. 98:257–65 25. Dubois CM, Laprise MH, Blanchette F, Gentry LE, Leduc R. 1995. Processing of transforming growth factor β1 precursor by human furin convertase. J. Biol. Chem. 270:10618– 24 26. Annes JP, Munger JS, Rifkin DB. 2003. Making sense of latent TGF β activation. J. Cell Sci. 116:217–24 27. Nunes I, Shapiro RL, Rifkin DB. 1995. Characterization of latent TGF-β activation by murine peritoneal macrophages. J. Immunol. 155:1450–59 28. Crawford SE, Stellmach V, Murphy-Ullrich JE, Ribeiro SM, Lawler J, et al. 1998. Thrombospondin-1 is a major activator of TGF-β1 in vivo. Cell 93:1159–70 29. Yehualaeshet T, O’Connor R, Green-Johnson J, Mai S, Silverstein R, et al. 1999. Activation of rat alveolar macrophage-derived latent transforming growth factor β-1 by plasmin requires interaction with thrombospondin-1 and its cell surface receptor, CD36. Am. J. Pathol. 155:841–51 30. Munger JS, Huang X, Kawakatsu H, Grifﬁths MJ, Dalton SL, et al. 1999. The integrin αvβ6 binds and activates latent TGFβ1: a mechanism for regulating pulmonary inﬂammation and ﬁbrosis. Cell 96:319–28 31. Annes JP, Chen Y, Munger JS, Rifkin DB. 2004. Integrin αVβ6-mediated activation of latent TGF-β requires the latent TGF-β binding protein-1. J. Cell Biol. 165:723–34 32. Van Obberghen-Schilling E, Roche NS, Flanders KC, Sporn MB, Roberts AB. 1988. Transforming growth factor β1 positively regulates its own expression in normal and transformed cells. J. Biol. Chem. 263:7741–46 33. Kim SJ, Jeang KT, Glick AB, Sporn MB, Roberts AB. 1989. Promoter sequences of the human transforming growth factor-β1 gene responsive to transforming growth factor-β1 autoinduction. J. Biol. Chem. 264:7041–45 34. Kim SJ, Angel P, Lafyatis R, Hattori K, Kim KY, et al. 1990. Autoinduction of transforming growth factor β1 is mediated by the AP-1 complex. Mol. Cell. Biol. 10:1492–97 35. Ventura JJ, Kennedy NJ, Flavell RA, Davis RJ. 2004. JNK regulates autocrine expression of TGF-β1. Mol. Cell 15:269–78 36. Kim SJ, Park K, Koeller D, Kim KY, Wakeﬁeld LM, et al. 1992. Post-transcriptional regulation of the human transforming growth factor-β1 gene. J. Biol. Chem. 267:13702– 7 37. Allison RS, Mumy ML, Wakeﬁeld LM. 1998. Translational control elements in the major human transforming growth factor-β1 mRNA. Growth Factors 16:89–100 38. Roberts AB, Sporn MB. 1992. Differential expression of the TGF-β isoforms in embryogenesis suggests speciﬁc roles in developing and adult tissues. Mol. Reprod. Dev. 32:91–98 39. Brown CB, Boyer AS, Runyan RB, Barnett JV. 1999. Requirement of type III TGF-β receptor for endocardial cell transformation in the heart. Science 283:2080–82 40. Massague J. 1998. TGF-β signal transduction. Annu. Rev. Biochem. 67:753–91 41. ten Dijke P, Hill CS. 2004. New insights into TGF-β-Smad signalling. Trends Biochem. Sci. 29:265–73 42. Huse M, Muir TW, Xu L, Chen YG, Kuriyan J, Massague J. 2001. The TGF β receptor activation process: an inhibitor- to substrate-binding switch. Mol. Cell 8:671–82 43. Inman GJ, Nicolas FJ, Hill CS. 2002. Nucleocytoplasmic shuttling of Smads 2, 3, and 4 permits sensing of TGF-β receptor activity. Mol. Cell 10:283–94 www.annualreviews.org • Regulation of Immune Responses by TGF-β

127

ARI

23 February 2006

15:9

44. Johnson K, Kirkpatrick H, Comer A, Hoffmann FM, Laughon A. 1999. Interaction of Smad complexes with tripartite DNA-binding sites. J. Biol. Chem. 274:20709–16 45. Shi Y, Wang YF, Jayaraman L, Yang H, Massague J, Pavletich NP. 1998. Crystal structure of a Smad MH1 domain bound to DNA: insights on DNA binding in TGF-β signaling. Cell 94:585–94 46. Zawel L, Dai JL, Buckhaults P, Zhou S, Kinzler KW, et al. 1998. Human Smad3 and Smad4 are sequence-speciﬁc transcription activators. Mol. Cell 1:611–67 47. Liu F, Pouponnot C, Massague J. 1997. Dual role of the Smad4/DPC4 tumor suppressor in TGFβ-inducible transcriptional complexes. Genes Dev. 11:3157–67 48. Nakao A, Afrakhte M, Moren A, Nakayama T, Christian JL, et al. 1997. Identiﬁcation of Smad7, a TGFβ-inducible antagonist of TGF-β signalling. Nature 389:631–35 49. Kavsak P, Rasmussen RK, Causing CG, Bonni S, Zhu H, et al. 2000. Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF β receptor for degradation. Mol. Cell 6:1365–75 50. Ebisawa T, Fukuchi M, Murakami G, Chiba T, Tanaka K, et al. 2001. Smurf1 interacts with transforming growth factor-β type I receptor through Smad7 and induces receptor degradation. J. Biol. Chem. 276:12477–80 51. Attisano L, Wrana JL. 2000. Smads as transcriptional co-modulators. Curr. Opin. Cell Biol. 12:235–43 52. Massague J, Wotton D. 2000. Transcriptional control by the TGF-β/Smad signaling system. EMBO J. 19:1745–54 53. ten Dijke P, Miyazono K, Heldin CH. 2000. Signaling inputs converge on nuclear effectors in TGF-β signaling. Trends Biochem. Sci. 25:64–70 54. Derynck R, Zhang YE. 2003. Smad-dependent and Smad-independent pathways in TGF-β family signalling. Nature 425:577–84 55. Shi Y, Massague J. 2003. Mechanisms of TGF-β signaling from cell membrane to the nucleus. Cell 113:685–700 56. Engel ME, McDonnell MA, Law BK, Moses HL. 1999. Interdependent SMAD and JNK signaling in transforming growth factor-β-mediated transcription. J. Biol. Chem. 274:37413–20 57. Yu L, Hebert MC, Zhang YE. 2002. TGF-β receptor-activated p38 MAP kinase mediates Smad-independent TGF-β responses. EMBO J. 21:3749–59 58. Kretzschmar M, Doody J, Timokhina I, Massague J. 1999. A mechanism of repression of TGFβ/Smad signaling by oncogenic Ras. Genes Dev. 13:804–16 59. Funaba M, Zimmerman CM, Mathews LS. 2002. Modulation of Smad2-mediated signaling by extracellular signal-regulated kinase. J. Biol. Chem. 277:41361–68 60. Blanchette F, Rivard N, Rudd P, Grondin F, Attisano L, Dubois CM. 2001. Cross-talk between the p42/p44 MAP kinase and Smad pathways in transforming growth factor β1-induced furin gene transactivation. J. Biol. Chem. 276:33986–94 61. Choy L, Derynck R. 1998. The type II transforming growth factor (TGF)-β receptorinteracting protein TRIP-1 acts as a modulator of the TGF-β response. J. Biol. Chem. 273:31455–62 62. McGonigle S, Beall MJ, Pearce EJ. 2002. Eukaryotic initiation factor 2α subunit associates with TGFβ receptors and 14–3–3ε and acts as a modulator of the TGFβ response. Biochemistry 41:579–87 63. Griswold-Prenner I, Kamibayashi C, Maruoka EM, Mumby MC, Derynck R. 1998. Physical and functional interactions between type I transforming growth factor β receptors and Bα, a WD-40 repeat subunit of phosphatase 2A. Mol. Cell. Biol. 18:6595–604

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

128

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

64. Gorelik L, Flavell RA. 2002. Transforming growth factor-β in T-cell biology. Nat. Rev. Immunol 2:46–53 65. Cerwenka A, Swain SL. 1999. TGF-β1: immunosuppressant and viability factor for T lymphocytes. Microbes Infect. 1:1291–96 66. Brabletz T, Pfeuffer I, Schorr E, Siebelt F, Wirth T, Serﬂing E. 1993. Transforming growth factor β and cyclosporin A inhibit the inducible activity of the interleukin-2 gene in T cells through a noncanonical octamer-binding site. Mol. Cell. Biol. 13:1155–62 67. McKarns SC, Schwartz RH, Kaminski NE. 2004. Smad3 is essential for TGF-β1 to suppress IL-2 production and TCR-induced proliferation, but not IL-2-induced proliferation. J. Immunol. 172:4275–84 68. Tzachanis D, Freeman GJ, Hirano N, van Puijenbroek AA, Delfs MW, et al. 2001. Tob is a negative regulator of activation that is expressed in anergic and quiescent T cells. Nat. Immunol. 2:1174–82 69. Hannon GJ, Beach D. 1994. p15INK4B is a potential effector of TGF-β-induced cell cycle arrest. Nature 371:257–61 70. Datto MB, Li Y, Panus JF, Howe DJ, Xiong Y, Wang XF. 1995. Transforming growth factor β induces the cyclin-dependent kinase inhibitor p21 through a p53-independent mechanism. Proc. Natl. Acad. Sci. USA 92:5545–49 71. Polyak K, Kato JY, Solomon MJ, Sherr CJ, Massague J, et al. 1994. p27Kip1, a cyclin-Cdk inhibitor, links transforming growth factor-β and contact inhibition to cell cycle arrest. Genes Dev. 8:9–22 72. Coffey RJ Jr, Bascom CC, Sipes NJ, Graves-Deal R, Weissman BE, et al. 1988. Selective inhibition of growth-related gene expression in murine keratinocytes by transforming growth factor β. Mol. Cell. Biol. 8:3088–93 73. Wolfraim LA, Walz TM, James Z, Fernandez T, Letterio JJ. 2004. p21Cip1 and p27Kip1 act in synergy to alter the sensitivity of naive T cells to TGF-β-mediated G1 arrest through modulation of IL-2 responsiveness. J. Immunol. 173:3093–102 74. Ruegemer JJ, Ho SN, Augustine JA, Schlager JW, Bell MP, et al. 1990. Regulatory effects of transforming growth factor-β on IL-2- and IL-4-dependent T cell-cycle progression. J. Immunol. 144:1767–76 75. Genestier L, Kasibhatla S, Brunner T, Green DR. 1999. Transforming growth factor β1 inhibits Fas ligand expression and subsequent activation-induced cell death in T cells via downregulation of c-Myc. J. Exp. Med. 189:231–39 76. Nelson BH, Martyak TP, Thompson LJ, Moon JJ, Wang T. 2003. Uncoupling of promitogenic and antiapoptotic functions of IL-2 by Smad-dependent TGF-β signaling. J. Immunol. 170:5563–70 77. McKarns SC, Schwartz RH. 2005. Distinct effects of TGF-β1 on CD4+ and CD8+ T cell survival, division, and IL-2 production: a role for T cell intrinsic Smad3. J. Immunol. 174:2071–83 78. Cottrez F, Groux H. 2001. Regulation of TGF-β response during T cell activation is modulated by IL-10. J. Immunol. 167:773–78 79. Sung JL, Lin JT, Gorham JD. 2003. CD28 co-stimulation regulates the effect of transforming growth factor-β1 on the proliferation of naive CD4+ T cells. Int. Immunol. 3:233–45 80. Bommireddy R, Ormsby I, Yin M, Boivin GP, Babcock GF, Doetschman T. 2003. TGFβ1 inhibits Ca2+ -calcineurin-mediated activation in thymocytes. J. Immunol. 170:3645–52 81. Lucas PJ, McNeil N, Hilgenfeld E, Choudhury B, Kim SJ, et al. 2004. Transforming growth factor-β pathway serves as a primary tumor suppressor in CD8+ T cell tumorigenesis. Cancer Res. 64:6524–29 www.annualreviews.org • Regulation of Immune Responses by TGF-β

129

ARI

23 February 2006

15:9

82. Wolfraim LA, Fernandez TM, Mamura M, Fuller WL, Kumar R, et al. 2004. Loss of Smad3 in acute T-cell lymphoblastic leukemia. N. Engl. J. Med. 351:552–59 83. Murphy KM, Reiner SL. 2002. The lineage decisions of helper T cells. Nat. Rev. Immunol. 2:933–44 84. Sad S, Mosmann TR. 1994. Single IL-2-secreting precursor CD4 T cell can develop into either Th1 or Th2 cytokine secretion phenotype. J. Immunol. 153:3514–22 85. Gorelik L, Fields PE, Flavell RA. 2000. Cutting edge: TGF-β inhibits Th type 2 development through inhibition of GATA-3 expression. J. Immunol. 165:4773–77 86. Heath VL, Murphy EE, Crain C, Tomlinson MG, O’Garra A. 2000. TGF-β1 downregulates Th2 development and results in decreased IL-4-induced STAT6 activation and GATA-3 expression. Eur. J. Immunol. 30:2639–49 87. Chen CH, Seguin-Devaux C, Burke NA, Oriss TB, Watkins SC, et al. 2003. Transforming growth factor β blocks Tec kinase phosphorylation, Ca2+ inﬂux, and NFATc translocation causing inhibition of T cell differentiation. J. Exp. Med. 197:1689–99 88. Gorham JD, Guler ML, Fenoglio D, Gubler U, Murphy KM. 1998. Low dose TGF-β attenuates IL-12 responsiveness in murine Th cells. J. Immunol. 161:1664–70 89. Gorelik L, Constant S, Flavell RA. 2002. Mechanism of transforming growth factor βinduced inhibition of T helper type 1 differentiation. J. Exp. Med. 195:1499–505 90. Afkarian M, Sedy JR, Yang J, Jacobson NG, Cereb N, et al. 2002. T-bet is a STAT1induced regulator of IL-12R expression in naive CD4+ T cells. Nat. Immunol. 3:549–57 91. Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH. 2000. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100:655–69 92. Lin JT, Martin SL, Xia L, Gorham JD. 2005. TGF-β1 uses distinct mechanisms to inhibit IFN-γ expression in CD4+ T cells at priming and at recall: differential involvement of Stat4 and T-bet. J. Immunol. 174:5950–58 93. Ludviksson BR, Seegers D, Resnick AS, Strober W. 2000. The effect of TGF-β1 on immune responses of naive versus memory CD4+ Th1/Th2 T cells. Eur. J. Immunol. 30:2101–11 94. Ulloa L, Doody J, Massague J. 1999. Inhibition of transforming growth factor-β/SMAD signalling by the interferon-γ/STAT pathway. Nature 397:710–13 95. Kitani A, Fuss I, Nakamura K, Kumaki F, Usui T, Strober W. 2003. Transforming growth factor (TGF)-β1-producing regulatory T cells induce Smad-mediated interleukin 10 secretion that facilitates coordinated immunoregulatory activity and amelioration of TGFβ1-mediated ﬁbrosis. J. Exp. Med. 198:1179–88 96. Fuss IJ, Boirivant M, Lacy B, Strober W. 2002. The interrelated roles of TGF-β and IL-10 in the regulation of experimental colitis. J. Immunol. 168:900–8 97. Schmitt E, Germann T, Goedert S, Hoehn P, Huels C, et al. 1994. IL-9 production of naive CD4+ T cells depends on IL-2, is synergistically enhanced by a combination of TGF-β and IL-4, and is inhibited by IFN-γ. J. Immunol. 153:3989–96 98. Rudner LA, Lin JT, Park IK, Cates JM, Dyer DA, et al. 2003. Necroinﬂammatory liver disease in BALB/c background, TGF-β1-deﬁcient mice requires CD4+ T cells. J. Immunol. 170:4785–92 99. Letterio JJ, Geiser AG, Kulkarni AB, Dang H, Kong L, et al. 1996. Autoimmunity associated with TGF-β1-deﬁciency in mice is dependent on MHC class II antigen expression. J. Clin. Invest. 98:2109–19 100. Seder RA, Ahmed R. 2003. Similarities and differences in CD4+ and CD8+ effector and memory T cell generation. Nat. Immunol. 4:835–42

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

130

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

101. Ranges GE, Figari IS, Espevik T, Palladino MA Jr. 1987. Inhibition of cytotoxic T cell development by transforming growth factor β and reversal by recombinant tumor necrosis factor α. J. Exp. Med. 166:991–98 102. Smyth MJ, Strobl SL, Young HA, Ortaldo JR, Ochoa AC. 1991. Regulation of lymphokine-activated killer activity and pore-forming protein gene expression in human peripheral blood CD8+ T lymphocytes. Inhibition by transforming growth factor-β. J. Immunol. 146:3289–97 103. Bonig H, Banning U, Hannen M, Kim YM, Verheyen J, et al. 1999. Transforming growth factor-β1 suppresses interleukin-15-mediated interferon-γ production in human T lymphocytes. Scand. J. Immunol. 50:612–18 104. Ahmadzadeh M, Rosenberg SA. 2005. TGF-β1 attenuates the acquisition and expression of effector function by tumor antigen-speciﬁc human memory CD8 T cells. J. Immunol. 174:5215–23 105. Glimcher LH, Townsend MJ, Sullivan BM, Lord GM. 2004. Recent developments in the transcriptional regulation of cytolytic effector cells. Nat. Rev. Immunol. 4:900–11 106. Croft M. 2003. Co-stimulatory members of the TNFR family: keys to effective T-cell immunity? Nat. Rev. Immunol. 3:609–20 107. Kim YJ, Stringﬁeld TM, Chen Y, Broxmeyer HE. 2005. Modulation of cord blood CD8+ T-cell effector differentiation by TGF-β1 and 4-1BB costimulation. Blood 105:274–81 108. Kobayashi S, Yoshida K, Ward JM, Letterio JJ, Longenecker G, et al. 1999. β2 microglobulin-deﬁcient background ameliorates lethal phenotype of the TGF-β1 null mouse. J. Immunol. 163:4013–19 109. Gorelik L, Flavell RA. 2001. Immune-mediated eradication of tumors through the blockade of transforming growth factor-β signaling in T cells. Nat. Med. 7:1118–22 110. Gershon RK. 1975. A disquisition on suppressor T cells. Transplant Rev. 26:170–85 111. Fontenot JD, Rudensky AY. 2005. A well adapted regulatory contrivance: regulatory T cell development and the forkhead family transcription factor Foxp3. Nat. Immunol. 6:331–37 112. Zelenay S, Lopes-Carvalho T, Caramalho I, Moraes-Fontes MF, Rebelo M, Demengeot J. 2005. Foxp3+ CD25− CD4 T cells constitute a reservoir of committed regulatory cells that regain CD25 expression upon homeostatic expansion. Proc. Natl. Acad. Sci. USA 102:4091–96 113. Liang S, Alard P, Zhao Y, Parnell S, Clark SL, Kosiewicz MM. 2005. Conversion of CD4+ CD25− cells into CD4+ CD25+ regulatory T cells in vivo requires B7 costimulation, but not the thymus. J. Exp. Med. 201:127–37 114. Chen Y, Kuchroo VK, Inobe J, Hafler DA, Weiner HL. 1994. Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis. Science 265:1237–40 115. Horwitz DA, Zheng SG, Gray JD. 2003. The role of the combination of IL-2 and TGF-β or IL-10 in the generation and function of CD4+ CD25+ and CD8+ regulatory T cell subsets. J. Leukoc. Biol. 74:471–78 116. Chen W, Jin W, Hardegen N, Lei KJ, Li L, et al. 2003. Conversion of peripheral CD4+ CD25− naive T cells to CD4+ CD25+ regulatory T cells by TGF-β induction of transcription factor Foxp3. J. Exp. Med. 198:1875–86 117. Fantini MC, Becker C, Monteleone G, Pallone F, Galle PR, Neurath MF. 2004. Cutting edge: TGF-β induces a regulatory phenotype in CD4+ CD25− T cells through Foxp3 induction and down-regulation of Smad7. J. Immunol. 172:5149–53 118. Zheng SG, Wang JH, Gray JD, Soucier H, Horwitz DA. 2004. Natural and induced CD4+ CD25+ cells educate CD4+ CD25− cells to develop suppressive activity: the role of IL-2, TGF-β, and IL-10. J. Immunol. 172:5213–21 www.annualreviews.org • Regulation of Immune Responses by TGF-β

Study of TGF-β regulation of oral tolerance.

First demonstration of TGF-β induction of FoxP3 and regulatory T cell generation.

131

ARI

23 February 2006

15:9

119. Wan YY, Flavell RA. 2005. Identifying Foxp3-expressing suppressor T cells with a bicistronic reporter. Proc. Natl. Acad. Sci. USA 102:5126–31 120. Peng Y, Laouar Y, Li MO, Green EA, Flavell RA. 2004. TGF-β regulates in vivo expansion of Foxp3-expressing CD4+ CD25+ regulatory T cells responsible for protection against diabetes. Proc. Natl. Acad. Sci. USA 101:4572–77 121. Huber S, Schramm C, Lehr HA, Mann A, Schmitt S, et al. 2004. Cutting edge: TGF-β signaling is required for the in vivo expansion and immunosuppressive capacity of regulatory CD4+ CD25+ T cells. J. Immunol. 173:6526–31 122. Marie JC, Letterio JJ, Gavin M, Rudensky AY. 2005. TGF-β1 maintains suppressor function and Foxp3 expression in CD4+ CD25+ regulatory T cells. J. Exp. Med. 201:1061– 67 123. Mamura M, Lee W, Sullivan TJ, Felici A, Sowers AL, et al. 2004. CD28 disruption exacerbates inﬂammation in Tgf-β1−/− mice: in vivo suppression by CD4+ CD25+ regulatory T cells independent of autocrine TGF-β1. Blood 103:4594–601 124. Fahlen L, Read S, Gorelik L, Hurst SD, Coffman RL, et al. 2005. T cells that cannot respond to TGF-β escape control by CD4+ CD25+ regulatory T cells. J. Exp. Med. 201:737–46 125. Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, Rudensky AY. 2005. Regulatory T cell lineage speciﬁcation by the forkhead transcription factor Foxp3. Immunity 22:329–41 126. Green DR, Droin N, Pinkoski M. 2003. Activation-induced cell death in T cells. Immunol. Rev. 193:70–81 127. Marrack P, Kappler J. 2004. Control of T cell viability. Annu. Rev. Immunol. 22:765–87 128. Gunnlaugsdottir B, Maggadottir SM, Ludviksson BR. 2005. Anti-CD28-induced costimulation and TCR avidity regulates the differential effect of TGF-β1 on CD4+ and CD8+ naive human T-cells. Int. Immunol. 17:35–44 129. Swain SL, Huston G, Tonkonogy S, Weinberg A. 1991. Transforming growth factor-β and IL-4 cause helper T cell precursors to develop into distinct effector helper cells that differ in lymphokine secretion pattern and cell surface phenotype. J. Immunol. 147:2991–3000 130. Weinberg AD, Whitham R, Swain SL, Morrison WJ, Wyrick G, et al. 1992. Transforming growth factor-β enhances the in vivo effector function and memory phenotype of antigen-speciﬁc T helper cells in experimental autoimmune encephalomyelitis. J. Immunol. 148:2109–17 131. de Jong R, van Lier RA, Ruscetti FW, Schmitt C, Debre P, Mossalayi MD. 1994. Differential effect of transforming growth factor-β1 on the activation of human naive and memory CD4+ T lymphocytes. Int. Immunol. 6:631–38 132. Zhang X, Giangreco L, Broome HE, Dargan CM, Swain SL. 1995. Control of CD4 effector fate: transforming growth factor β1 and interleukin 2 synergize to prevent apoptosis and promote effector expansion. J. Exp. Med. 182:699–709 133. Swain SL, Croft M, Dubey C, Haynes L, Rogers P, et al. 1996. From naive to memory T cells. Immunol. Rev. 150:143–67 134. Rich S, Van Nood N, Lee HM. 1996. Role of alpha 5 beta 1 integrin in TGF-beta 1costimulated CD8+ T cell growth and apoptosis. J. Immunol. 157:2916–23 135. Cerwenka A, Kovar H, Majdic O, Holter W. 1996. Fas- and activation-induced apoptosis are reduced in human T cells preactivated in the presence of TGF-β1. J. Immunol. 156:459–64 136. Chen W, Jin W, Tian H, Sicurello P, Frank M, et al. 2001. Requirement for transforming growth factor β1 in controlling T cell apoptosis. J. Exp. Med. 194:439–53

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

132

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

137. Bommireddy R, Saxena V, Ormsby I, Yin M, Boivin GP, et al. 2003. TGF-β1 regulates lymphocyte homeostasis by preventing activation and subsequent apoptosis of peripheral lymphocytes. J. Immunol. 170:4612–22 138. Refaeli Y, Van Parijs L, Alexander SI, Abbas AK. 2002. Interferon γ is required for activation-induced death of T lymphocytes. J. Exp. Med. 196:999–1005 139. Weiner HL. 2001. Induction and mechanism of action of transforming growth factor-βsecreting Th3 regulatory cells. Immunol. Rev. 182:207–14 140. Nakamura K, Kitani A, Strober W. 2001. Cell contact-dependent immunosuppression by CD4+ CD25+ regulatory T cells is mediated by cell surface-bound transforming growth factor β. J. Exp. Med. 194:629–44 141. Oida T, Zhang X, Goto M, Hachimura S, Totsuka M, et al. 2003. CD4+ CD25− T cells that express latency-associated peptide on the surface suppress CD4+ CD45RBhigh -induced colitis by a TGF-β-dependent mechanism. J. Immunol. 170:2516–22 142. Green EA, Gorelik L, McGregor CM, Tran EH, Flavell RA. 2003. CD4+ CD25+ T regulatory cells control anti-islet CD8+ T cells through TGF-β-TGF-β receptor interactions in type 1 diabetes. Proc. Natl. Acad. Sci. USA 100:10878–83 143. Piccirillo CA, Letterio JJ, Thornton AM, McHugh RS, Mamura M, et al. 2002. CD4+ CD25+ regulatory T cells can mediate suppressor function in the absence of transforming growth factor β1 production and responsiveness. J. Exp. Med. 196:237–46 144. Powrie F, Carlino J, Leach MW, Mauze S, Coffman RL. 1996. A critical role for transforming growth factor-β but not interleukin 4 in the suppression of T helper type 1mediated colitis by CD45RBlow CD4+ T cells. J. Exp. Med. 183:2669–74 145. Nakamura K, Kitani A, Fuss I, Pedersen A, Harada N, et al. 2004. TGF-β1 plays an important role in the mechanism of CD4+ CD25+ regulatory T cell activity in both humans and mice. J. Immunol. 172:834–42 146. Chen W, Jin W, Wahl SM. 1998. Engagement of cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) induces transforming growth factor β (TGF-β) production by murine CD4+ T cells. J. Exp. Med. 188:1849–57 147. Gomes NA, Gattass CR, Barreto-De-Souza V, Wilson ME, DosReis GA. 2000. TGF-β mediates CTLA-4 suppression of cellular immunity in murine kalaazar. J. Immunol. 164:2001–8 148. Kitani A, Chua K, Nakamura K, Strober W. 2000. Activated self-MHC-reactive T cells have the cytokine phenotype of Th3/T regulatory cell 1 T cells. J. Immunol. 165:691–702 149. Sullivan TJ, Letterio JJ, van Elsas A, Mamura M, van Amelsfort J, et al. 2001. Lack of a role for transforming growth factor-β in cytotoxic T lymphocyte antigen-4-mediated inhibition of T cell activation. Proc. Natl. Acad. Sci. USA 98:2587–92 150. Esplugues E, Sancho D, Vega-Ramos J, Martinez C, Syrbe U, et al. 2003. Enhanced antitumor immunity in mice deﬁcient in CD69. J. Exp. Med. 197:1093–106 151. Sancho D, Gomez M, Viedma F, Esplugues E, Gordon-Alonso M, et al. 2003. CD69 downregulates autoimmune reactivity through active transforming growth factor-β production in collagen-induced arthritis. J. Clin. Invest. 112:872–82 152. Seder RA, Marth T, Sieve MC, Strober W, Letterio JJ, et al. 1998. Factors involved in the differentiation of TGF-β-producing cells from naive CD4+ T cells: IL-4 and IFN-γ have opposing effects, while TGF-β positively regulates its own production. J. Immunol. 160:5719–28 153. Neurath MF, Weigmann B, Finotto S, Glickman J, Nieuwenhuis E, et al. 2002. The transcription factor T-bet regulates mucosal T cell activation in experimental colitis and Crohn’s disease. J. Exp. Med. 195:1129–43 www.annualreviews.org • Regulation of Immune Responses by TGF-β

133

ARI

23 February 2006

15:9

154. Chen W, Frank ME, Jin W, Wahl SM. 2001. TGF-β released by apoptotic T cells contributes to an immunosuppressive milieu. Immunity 14:715–25 155. Stavnezer J. 1995. Regulation of antibody production and class switching by TGF-β. J. Immunol. 155:1647–51 156. Lebman DA, Edmiston JS. 1999. The role of TGF-β in growth, differentiation, and maturation of B lymphocytes. Microbes Infect. 1:1297–304 157. Kee BL, Rivera RR, Murre C. 2001. Id3 inhibits B lymphocyte progenitor growth and survival in response to TGF-β. Nat. Immunol. 2:242–47 158. Kehrl JH, Roberts AB, Wakeﬁeld LM, Jakowlew S, Sporn MB, Fauci AS. 1986. Transforming growth factor β is an important immunomodulatory protein for human B lymphocytes. J. Immunol. 137:3855–60 159. Petit-Koskas E, Genot E, Lawrence D, Kolb JP. 1988. Inhibition of the proliferative response of human B lymphocytes to B cell growth factor by transforming growth factorβ. Eur. J. Immunol. 18:111–16 160. Smeland EB, Blomhoff HK, Holte H, Ruud E, Beiske K, et al. 1987. Transforming growth factor type β (TGF β) inhibits G1 to S transition, but not activation of human B lymphocytes. Exp. Cell Res. 171:213–22 161. Bouchard C, Fridman WH, Sautes C. 1997. Effect of TGF-β1 on cell cycle regulatory proteins in LPS-stimulated normal mouse B lymphocytes. J. Immunol. 159:4155–64 162. Kamesaki H, Nishizawa K, Michaud GY, Cossman J, Kiyono T. 1998. TGF-β1 induces the cyclin-dependent kinase inhibitor p27Kip1 mRNA and protein in murine B cells. J. Immunol. 160:770–77 163. Roes J, Choi BK, Cazac BB. 2003. Redirection of B cell responsiveness by transforming growth factor β receptor. Proc. Natl. Acad. Sci. USA 100:7241–46 164. Patil S, Wildey GM, Brown TL, Choy L, Derynck R, Howe PH. 2000. Smad7 is induced by CD40 and protects WEHI 231 B-lymphocytes from transforming growth factor-βinduced growth inhibition and apoptosis. J. Biol. Chem. 275:38363–70 165. Briskin M, Kuwabara MD, Sigman DS, Wall R. 1988. Induction of κ transcription by interferon-γ without activation of NF-κB. Science 242:1036–37 166. Kehrl JH, Thevenin C, Rieckmann P, Fauci AS. 1991. Transforming growth factor-β suppresses human B lymphocyte Ig production by inhibiting synthesis and the switch from the membrane form to the secreted form of Ig mRNA. J. Immunol. 146:4016–23 167. Lebman DA, Lee FD, Coffman RL. 1990. Mechanism for transforming growth factor β and IL-2 enhancement of IgA expression in lipopolysaccharide-stimulated B cell cultures. J. Immunol. 144:952–59 168. Kim PH, Kagnoff MF. 1990. Transforming growth factor beta 1 increases IgA isotype switching at the clonal level. J. Immunol. 145:3773–78 169. Defrance T, Vanbervliet B, Briere F, Durand I, Rousset F, Banchereau J. 1992. Interleukin 10 and transforming growth factor β cooperate to induce anti-CD40-activated naive human B cells to secrete immunoglobulin A. J. Exp. Med. 175:671–82 170. van Vlasselaer P, Punnonen J, de Vries JE. 1992. Transforming growth factor-β directs IgA switching in human B cells. J. Immunol. 148:2062–67 171. McIntyre TM, Klinman DR, Rothman P, Lugo M, Dasch JR, et al. 1993. Transforming growth factor β1 selectivity stimulates immunoglobulin G2b secretion by lipopolysaccharide-activated murine B cells. J. Exp. Med. 177:1031–37 172. Fagarasan S, Honjo T. 2003. Intestinal IgA synthesis: regulation of front-line body defences. Nat. Rev. Immunol. 3:63–72 173. McIntyre TM, Kehry MR, Snapper CM. 1995. Novel in vitro model for high-rate IgA class switching. J. Immunol. 154:3156–61

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

134

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

174. Shockett P, Stavnezer J. 1991. Effect of cytokines on switching to IgA and α germline transcripts in the B lymphoma I.29 μ: transforming growth factor-β activates transcription of the unrearranged Cα gene. J. Immunol. 147:4374–83 175. Lin YC, Stavnezer J. 1992. Regulation of transcription of the germ-line Ig α constant region gene by an ATF element and by novel transforming growth factor-β1-responsive elements. J. Immunol. 149:2914–25 176. Shi MJ, Stavnezer J. 1998. CBF α3 (AML2) is induced by TGF-β1 to bind and activate the mouse germline Ig α promoter. J. Immunol. 161:6751–60 177. Hanai J, Chen LF, Kanno T, Ohtani-Fujita N, Kim WY, et al. 1999. Interaction and functional cooperation of PEBP2/CBF with Smads. Synergistic induction of the immunoglobulin germline Cα promoter. J. Biol. Chem. 274:31577–82 178. Pardali E, Xie XQ, Tsapogas P, Itoh S, Arvanitidis K, et al. 2000. Smad and AML proteins synergistically confer transforming growth factor β1 responsiveness to human germ-line IgA genes. J. Biol. Chem. 275:3552–60 179. Park SR, Lee JH, Kim PH. 2001. Smad3 and Smad4 mediate transforming growth factorβ1-induced IgA expression in murine B lymphocytes. Eur. J. Immunol. 31:1706–15 180. Park SR, Lee EK, Kim BC, Kim PH. 2003. p300 cooperates with Smad3/4 and Runx3 in TGFβ1-induced IgA isotype expression. Eur. J. Immunol. 33:3386–92 181. Borsutzky S, Cazac BB, Roes J, Guzman CA. 2004. TGF-β receptor signaling is critical for mucosal IgA responses. J. Immunol. 173:3305–9 182. van Ginkel FW, Wahl SM, Kearney JF, Kweon MN, Fujihashi K, et al. 1999. Partial IgAdeﬁciency with increased Th2-type cytokines in TGF-β1 knockout mice. J. Immunol. 163:1951–57 183. Letterio JJ, Geiser AG, Kulkarni AB, Roche NS, Sporn MB, Roberts AB. 1994. Maternal rescue of transforming growth factor-β1 null mice. Science 264:1936–38 184. Holder MJ, Knox K, Gordon J. 1992. Factors modifying survival pathways of germinal center B cells. Glucocorticoids and transforming growth factor-β, but not cyclosporin A or anti-CD19, block surface immunoglobulin-mediated rescue from apoptosis. Eur. J. Immunol. 22:2725–28 185. Lomo J, Blomhoff HK, Beiske K, Stokke T, Smeland EB. 1995. TGF-β1 and cyclic AMP promote apoptosis in resting human B lymphocytes. J. Immunol. 154:1634–43 186. Arsura M, Wu M, Sonenshein GE. 1996. TGF β1 inhibits NF-κB/Rel activity inducing apoptosis of B cells: transcriptional activation of IκBα. Immunity 5:31–40 187. Wildey GM, Patil S, Howe PH. 2003. Smad3 potentiates transforming growth factor β (TGFβ)-induced apoptosis and expression of the BH3-only protein Bim in WEHI 231 B lymphocytes. J. Biol. Chem. 278:18069–77 188. Hayakawa K, Asano M, Shinton SA, Gui M, Allman D, et al. 1999. Positive selection of natural autoreactive B cells. Science 285:113–16 189. Lam KP, Rajewsky K. 1999. B cell antigen receptor speciﬁcity and surface density together determine B-1 versus B-2 cell development. J. Exp. Med. 190:471–77 190. Snapper CM, Waegell W, Beernink H, Dasch JR. 1993. Transforming growth factor-β1 is required for secretion of IgG of all subclasses by LPS-activated murine B cells in vitro. J. Immunol. 151:4625–36 191. Zan H, Cerutti A, Dramitinos P, Schaffer A, Casali P. 1998. CD40 engagement triggers switching to IgA1 and IgA2 in human B cells through induction of endogenous TGF-β: evidence for TGF-β but not IL-10-dependent direct Sμ → Sα and sequential Sμ → Sγ, Sγ → Sα DNA recombination. J. Immunol. 161:5217–25 www.annualreviews.org • Regulation of Immune Responses by TGF-β

135

ARI

23 February 2006

15:9

192. Caver TE, O’Sullivan FX, Gold LI, Gresham HD. 1996. Intracellular demonstration of active TGFβ1 in B cells and plasma cells of autoimmune mice. IgG-bound TGFβ1 suppresses neutrophil function and host defense against Staphylococcus aureus infection. J. Clin. Invest. 98:2496–506 193. Biron CA, Nguyen KB, Pien GC, Cousens LP, Salazar-Mather TP. 1999. Natural killer cells in antiviral defense: function and regulation by innate cytokines. Annu. Rev. Immunol. 17:189–220 194. Hamerman JA, Ogasawara K, Lanier LL. 2005. NK cells in innate immunity. Curr. Opin. Immunol. 17:29–35 195. Martin-Fontecha A, Thomsen LL, Brett S, Gerard C, Lipp M, et al. 2004. Induced recruitment of NK cells to lymph nodes provides IFN-γ for TH 1 priming. Nat. Immunol. 5:1260–65 196. Perussia B, Chan SH, D’Andrea A, Tsuji K, Santoli D, et al. 1992. Natural killer (NK) cell stimulatory factor or IL-12 has differential effects on the proliferation of TCR-alpha beta+, TCR-gamma delta+ T lymphocytes, and NK cells. J. Immunol. 149:3495–502 197. Tripp CS, Wolf SF, Unanue ER. 1993. Interleukin 12 and tumor necrosis factor α are costimulators of interferon γ production by natural killer cells in severe combined immunodeﬁciency mice with listeriosis, and interleukin 10 is a physiologic antagonist. Proc. Natl. Acad. Sci. USA 90:3725–29 198. Rook AH, Kehrl JH, Wakeﬁeld LM, Roberts AB, Sporn MB, et al. 1986. Effects of transforming growth factor β on the functions of natural killer cells: depressed cytolytic activity and blunting of interferon responsiveness. J. Immunol. 136:3916–20 199. Bellone G, Aste-Amezaga M, Trinchieri G, Rodeck U. 1995. Regulation of NK cell functions by TGF-β 1. J. Immunol. 155:1066–73 200. Hunter CA, Bermudez L, Beernink H, Waegell W, Remington JS. 1995. Transforming growth factor-β inhibits interleukin-12-induced production of interferon-γ by natural killer cells: a role for transforming growth factor-β in the regulation of T cell-independent resistance to Toxoplasma gondii. Eur. J. Immunol. 25:994–1000 201. Su HC, Ishikawa R, Biron CA. 1993. Transforming growth factor-β expression and natural killer cell responses during virus infection of normal, nude, and SCID mice. J. Immunol. 151:4874–90 202. Ortaldo JR, Mason AT, O’Shea JJ, Smyth MJ, Falk LA, et al. 1991. Mechanistic studies of transforming growth factor-β inhibition of IL-2-dependent activation of CD3-large granular lymphocyte functions. Regulation of IL-2R β (p75) signal transduction. J. Immunol. 146:3791–98 203. Malygin AM, Meri S, Timonen T. 1993. Regulation of natural killer cell activity by transforming growth factor-β and prostaglandin E2. Scand. J. Immunol. 37:71–76 204. Moretta A, Biassoni R, Bottino C, Mingari MC, Moretta L. 2000. Natural cytotoxicity receptors that trigger human NK-cell-mediated cytolysis. Immunol. Today 21:228–34 205. Bauer S, Groh V, Wu J, Steinle A, Phillips JH, et al. 1999. Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science 285:727–29 206. Castriconi R, Cantoni C, Della Chiesa M, Vitale M, Marcenaro E, et al. 2003. Transforming growth factor β1 inhibits expression of NKp30 and NKG2D receptors: consequences for the NK-mediated killing of dendritic cells. Proc. Natl. Acad. Sci. USA 100:4120– 25 207. Lee JC, Lee KM, Kim DW, Heo DS. 2004. Elevated TGF-β1 secretion and downmodulation of NKG2D underlies impaired NK cytotoxicity in cancer patients. J. Immunol. 172:7335–40

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

136

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

208. Gray JD, Hirokawa M, Ohtsuka K, Horwitz DA. 1998. Generation of an inhibitory circuit involving CD8+ T cells, IL-2, and NK cell-derived TGF-β: contrasting effects of antiCD2 and anti-CD3. J. Immunol. 160:2248–54 209. Steinman RM, Hawiger D, Nussenzweig MC. 2003. Tolerogenic dendritic cells. Annu. Rev. Immunol. 21:685–711 210. Strobl H, Knapp W. 1999. TGF-β1 regulation of dendritic cells. Microbes Infect. 1:1283– 90 211. Strobl H, Riedl E, Scheinecker C, Bello-Fernandez C, Pickl WF, et al. 1996. TGF-β1 promotes in vitro development of dendritic cells from CD34+ hemopoietic progenitors. J. Immunol. 157:1499–507 212. Geissmann F, Prost C, Monnet JP, Dy M, Brousse N, Hermine O. 1998. Transforming growth factor β1, in the presence of granulocyte/macrophage colony-stimulating factor and interleukin 4, induces differentiation of human peripheral blood monocytes into dendritic Langerhans cells. J. Exp. Med. 187:961–66 213. Jaksits S, Kriehuber E, Charbonnier AS, Rappersberger K, Stingl G, Maurer D. 1999. CD34+ cell-derived CD14+ precursor cells develop into Langerhans cells in a TGF-β1dependent manner. J. Immunol. 163:4869–77 214. Zhang Y, Zhang YY, Ogata M, Chen P, Harada A, et al. 1999. Transforming growth factorβ1 polarizes murine hematopoietic progenitor cells to generate Langerhans cell-like dendritic cells through a monocyte/macrophage differentiation pathway. Blood 93:1208–20 215. Riedl E, Strobl H, Majdic O, Knapp W. 1997. TGF-β1 promotes in vitro generation of dendritic cells by protecting progenitor cells from apoptosis. J. Immunol. 158:1591–97 216. Borkowski TA, Letterio JJ, Farr AG, Udey MC. 1996. A role for endogenous transforming growth factor β1 in Langerhans cell biology: the skin of transforming growth factor β1 null mice is devoid of epidermal Langerhans cells. J. Exp. Med. 184:2417–22 217. Borkowski TA, Letterio JJ, Mackall CL, Saitoh A, Wang XJ, et al. 1997. A role for TGFβ1 in Langerhans cell biology. Further characterization of the epidermal Langerhans cell defect in TGFβ1 null mice. J. Clin. Invest. 100:575–81 218. Thomas RM, Belsito DV, Huang C, Chen LZ, Ormsby I, et al. 2001. Appearance of Langerhans cells in the epidermis of Tgfb1−/− SCID mice: paracrine and autocrine effects of transforming growth factor-β1 and -β2(1). J. Invest. Dermatol. 117:1574–80 219. Hacker C, Kirsch RD, Ju XS, Hieronymus T, Gust TC, et al. 2003. Transcriptional proﬁling identiﬁes Id2 function in dendritic cell development. Nat. Immunol. 4:380–86 220. Strobl H, Bello-Fernandez C, Riedl E, Pickl WF, Majdic O, et al. 1997. ﬂt3 ligand in cooperation with transforming growth factor-β1 potentiates in vitro development of Langerhans-type dendritic cells and allows single-cell dendritic cell cluster formation under serum-free conditions. Blood 90:1425–34 221. Geissmann F, Revy P, Regnault A, Lepelletier Y, Dy M, et al. 1999. TGF-β1 prevents the noncognate maturation of human dendritic Langerhans cells. J. Immunol. 162:4567–75 222. Ronger-Savle S, Valladeau J, Claudy A, Schmitt D, Peguet-Navarro J, et al. 2005. TGFβ inhibits CD1d expression on dendritic cells. J. Invest. Dermatol. 124:116–18 223. Yamaguchi Y, Tsumura H, Miwa M, Inaba K. 1997. Contrasting effects of TGF-β1 and TNF-α on the development of dendritic cells from progenitors in mouse bone marrow. Stem Cells 15:144–53 224. Dong Y, Tang L, Letterio JJ, Benveniste EN. 2001. The Smad3 protein is involved in TGF-β inhibition of class II transactivator and class II MHC expression. J. Immunol. 167:311–19 www.annualreviews.org • Regulation of Immune Responses by TGF-β

An in vivo demonstration of TGF-β1 in Langerhans cell development.

137

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

First demonstration of TGF-β induction of chemotaxis.

138

15:9

225. Tada Y, Asahina A, Fujita H, Mitsui H, Torii H, et al. 2004. Differential effects of LPS and TGF-β on the production of IL-6 and IL-12 by Langerhans cells, splenic dendritic cells, and macrophages. Cytokine 25:155–61 226. Gruschwitz MS, Hornstein OP. 1992. Expression of transforming growth factor type β on human epidermal dendritic cells. J. Invest. Dermatol. 99:114–16 227. de Saint-Vis B, Fugier-Vivier I, Massacrier C, Gaillard C, Vanbervliet B, et al. 1998. The cytokine proﬁle expressed by human dendritic cells is dependent on cell subtype and mode of activation. J. Immunol. 160:1666–76 228. Morelli AE, Zahorchak AF, Larregina AT, Colvin BL, Logar AJ, et al. 2001. Cytokine production by mouse myeloid dendritic cells in relation to differentiation and terminal maturation induced by lipopolysaccharide or CD40 ligation. Blood 98:1512–23 229. Steinman RM, Hawiger D, Liu K, Bonifaz L, Bonnyay D, et al. 2003. Dendritic cell function in vivo during the steady state: a role in peripheral tolerance. Ann. NY Acad. Sci. 987:15–25 230. Gregory CD, Devitt A. 2004. The macrophage and the apoptotic cell: an innate immune interaction viewed simplistically?Immunology 113:1–14 231. Ashcroft GS. 1999. Bidirectional regulation of macrophage function by TGF-β. Microbes Infect. 1:1275–82 232. Wahl SM, Hunt DA, Wakefield LM, McCartney-Francis N, Wahl LM, et al. 1987. Transforming growth factor type β induces monocyte chemotaxis and growth factor production. Proc. Natl. Acad. Sci. USA 84:5788–92 233. Wiseman DM, Polverini PJ, Kamp DW, Leibovich SJ. 1988. Transforming growth factorβ (TGF β) is chemotactic for human monocytes and induces their expression of angiogenic activity. Biochem. Biophys. Res. Commun. 157:793–800 234. Allen JB, Manthey CL, Hand AR, Ohura K, Ellingsworth L, Wahl SM. 1990. Rapid onset synovial inﬂammation and hyperplasia induced by transforming growth factor β. J. Exp. Med. 171:231–47 235. Wahl SM, Allen JB, Weeks BS, Wong HL, Klotman PE. 1993. Transforming growth factor β enhances integrin expression and type IV collagenase secretion in human monocytes. Proc. Natl. Acad. Sci. USA 90:4577–81 236. Bauvois B, Rouillard D, Sanceau J, Wietzerbin J. 1992. IFN-γ and transforming growth factor-β1 differently regulate ﬁbronectin and laminin receptors of human differentiating monocytic cells. J. Immunol. 148:3912–19 237. Turner M, Chantry D, Feldmann M. 1990. Transforming growth factor β induces the production of interleukin 6 by human peripheral blood mononuclear cells. Cytokine 2:211– 16 238. Riddick CA, Serio KJ, Hodulik CR, Ring WL, Regan MS, Bigby TD. 1999. TGF-β increases leukotriene C4 synthase expression in the monocyte-like cell line, THP-1. J. Immunol. 162:1101–7 239. Bottalico LA, Wager RE, Agellon LB, Assoian RK, Tabas I. 1991. Transforming growth factor-β1 inhibits scavenger receptor activity in THP-1 human macrophages. J. Biol. Chem. 266:22866–71 240. Han J, Hajjar DP, Tauras JM, Feng J, Gotto AM Jr, Nicholson AC. 2000. Transforming growth factor-β1 (TGF-β1) and TGF-β2 decrease expression of CD36, the type B scavenger receptor, through mitogen-activated protein kinase phosphorylation of peroxisome proliferator-activated receptor-γ. J. Biol. Chem. 275:1241–46 241. Tridandapani S, Wardrop R, Baran CP, Wang Y, Opalek JM, et al. 2003. TGF-β1 suppresses [correction of supresses] myeloid Fcγ receptor function by regulating the expression and function of the common γ-subunit. J. Immunol. 170:4572–77 Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

242. Bogdan C, Paik J, Vodovotz Y, Nathan C. 1992. Contrasting mechanisms for suppression of macrophage cytokine release by transforming growth factor-β and interleukin-10. J. Biol. Chem. 267:23301–8 243. Werner F, Jain MK, Feinberg MW, Sibinga NE, Pellacani A, et al. 2000. Transforming growth factor-β1 inhibition of macrophage activation is mediated via Smad3. J. Biol. Chem. 275:36653–58 244. Vodovotz Y, Bogdan C, Paik J, Xie QW, Nathan C. 1993. Mechanisms of suppression of macrophage nitric oxide release by transforming growth factor β. J. Exp. Med. 178:605–13 245. Tsunawaki S, Sporn M, Ding A, Nathan C. 1988. Deactivation of macrophages by transforming growth factor-β. Nature 334:260–62 246. Janeway CA Jr, Medzhitov R. 2002. Innate immune recognition. Annu. Rev. Immunol. 20:197–216 247. Imai K, Takeshita A, Hanazawa S. 2000. Transforming growth factor-β inhibits lipopolysaccharide-stimulated expression of inﬂammatory cytokines in mouse macrophages through downregulation of activation protein 1 and CD14 receptor expression. Infect. Immun. 68:2418–23 248. Naiki Y, Michelsen KS, Zhang W, Chen S, Doherty TM, Arditi M. 2005. Transforming growth factor-β differentially inhibits MyD88-dependent, but not TRAM- and TRIFdependent, lipopolysaccharide-induced TLR4 signaling. J. Biol. Chem. 280:5491–95 249. Nandan D, Reiner NE. 1997. TGF-β attenuates the class II transactivator and reveals an accessory pathway of IFN-γ action. J. Immunol. 158:1095–101 250. Takeuchi M, Alard P, Streilein JW. 1998. TGF-β promotes immune deviation by altering accessory signals of antigen-presenting cells. J. Immunol. 160:1589–97 251. Du C, Sriram S. 1998. Mechanism of inhibition of LPS-induced IL-12p40 production by IL-10 and TGF-β in ANA-1 cells. J. Leukoc. Biol. 64:92–97 252. Fadok VA, Bratton DL, Konowal A, Freed PW, Westcott JY, Henson PM. 1998. Macrophages that have ingested apoptotic cells in vitro inhibit proinﬂammatory cytokine production through autocrine/paracrine mechanisms involving TGF-β, PGE2, and PAF. J. Clin. Invest. 101:890–98 253. McDonald PP, Fadok VA, Bratton D, Henson PM. 1999. Transcriptional and translational regulation of inﬂammatory mediator production by endogenous TGF-β in macrophages that have ingested apoptotic cells. J. Immunol. 163:6164–72 254. Fadok VA, Voelker DR, Campbell PA, Cohen JJ, Bratton DL, Henson PM. 1992. Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers speciﬁc recognition and removal by macrophages. J. Immunol. 148:2207–16 255. Fadok VA, Bratton DL, Rose DM, Pearson A, Ezekewitz RA, Henson PM. 2000. A receptor for phosphatidylserine-speciﬁc clearance of apoptotic cells. Nature 405:85–90 256. Huynh ML, Fadok VA, Henson PM. 2002. Phosphatidylserine-dependent ingestion of apoptotic cells promotes TGF-β1 secretion and the resolution of inﬂammation. J. Clin. Invest. 109:41–50 257. Otsuka M, Tsuchiya S, Aramaki Y. 2004. Involvement of ERK, a MAP kinase, in the production of TGF-β by macrophages treated with liposomes composed of phosphatidylserine. Biochem. Biophys. Res. Commun. 324:1400–5 258. Pilette C, Ouadrhiri Y, Van Snick J, Renauld JC, Staquet P, et al. 2002. IL-9 inhibits oxidative burst and TNF-α release in lipopolysaccharide-stimulated human monocytes through TGF-β. J. Immunol. 168:4103–11 259. Bastos KR, Alvarez JM, Marinho CR, Rizzo LV, Lima MR. 2002. Macrophages from IL-12p40-deﬁcient mice have a bias toward the M2 activation proﬁle. J. Leukoc. Biol. 71:271–78 www.annualreviews.org • Regulation of Immune Responses by TGF-β

Identified TGF-β as an inhibitor of macrophage activation.

139

ARI

23 February 2006

15:9

260. Gruber BL, Marchese MJ, Kew RR. 1994. Transforming growth factor-β1 mediates mast cell chemotaxis. J. Immunol. 152:5860–67 261. Olsson N, Piek E, ten Dijke P, Nilsson G. 2000. Human mast cell migration in response to members of the transforming growth factor-β family. J. Leukoc. Biol. 67:350–56 262. Rosbottom A, Scudamore CL, von der Mark H, Thornton EM, Wright SH, Miller HR. 2002. TGF-β1 regulates adhesion of mucosal mast cell homologues to laminin-1 through expression of integrin α7 J. Immunol. 169:5689–95 263. Gomez G, Ramirez CD, Rivera J, Patel M, Norozian F, et al. 2005. TGF-β1 inhibits mast cell FcεRI expression. J. Immunol. 174:5987–93 264. Mesples B, Fontaine RH, Lelievre V, Launay JM, Gressens P. 2005. Neuronal TGFβ1 mediates IL-9/mast cell interaction and exacerbates excitotoxicity in newborn mice. Neurobiol. Dis. 18:193–205 265. Bissonnette EY, Enciso JA, Befus AD. 1997. TGF-β1 inhibits the release of histamine and tumor necrosis factor-α from mast cells through an autocrine pathway. Am. J. Respir. Cell Mol. Biol. 16:275–82 266. Gebhardt T, Lorentz A, Detmer F, Trautwein C, Bektas H, et al. 2005. Growth, phenotype, and function of human intestinal mast cells are tightly regulated by transforming growth factor β1. Gut 54:928–34 267. Broide DH, Wasserman SI, Alvaro-Gracia J, Zvaiﬂer NJ, Firestein GS. 1989. Transforming growth factor-β1 selectively inhibits IL-3-dependent mast cell proliferation without affecting mast cell function or differentiation. J. Immunol. 143:1591–97 268. Miller HR, Wright SH, Knight PA, Thornton EM. 1999. A novel function for transforming growth factor-β1: upregulation of the expression and the IgE-independent extracellular release of a mucosal mast cell granule-speciﬁc β-chymase, mouse mast cell protease-1. Blood 93:3473–86 269. Pennington DW, Lopez AR, Thomas PS, Peck C, Gold WM. 1992. Dog mastocytoma cells produce transforming growth factor β1. J. Clin. Invest. 90:35–41 270. Gordon JR, Galli SJ. 1994. Promotion of mouse ﬁbroblast collagen gene expression by mast cells stimulated via the FcεRI. Role for mast cell-derived transforming growth factor β and tumor necrosis factor α. J. Exp. Med. 180:2027–37 271. Lindstedt KA, Wang Y, Shiota N, Saarinen J, Hyytiainen M, et al. 2001. Activation of paracrine TGF-β1 signaling upon stimulation and degranulation of rat serosal mast cells: a novel function for chymase. FASEB J. 15:1377–88 272. Kobayashi SD, Voyich JM, DeLeo FR. 2003. Regulation of the neutrophil-mediated inﬂammatory response to infection. Microbes Infect. 5:1337–44 273. Klion AD, Nutman TB. 2004. The role of eosinophils in host defense against helminth parasites. J. Allergy Clin. Immunol. 113:30–37 274. Munitz A, Levi-Schaffer F. 2004. Eosinophils: ‘new’ roles for ‘old’ cells. Allergy 59:268–75 275. Reibman J, Meixler S, Lee TC, Gold LI, Cronstein BN, et al. 1991. Transforming growth factor β1, a potent chemoattractant for human neutrophils, bypasses classic signaltransduction pathways. Proc. Natl. Acad. Sci. USA 88:6805–9 276. Brandes ME, Mai UE, Ohura K, Wahl SM. 1991. Type I transforming growth factor-β receptors on neutrophils mediate chemotaxis to transforming growth factor-β. J. Immunol. 147:1600–6 277. Chen CC, Manning AM. 1996. TGF-β1, IL-10 and IL-4 differentially modulate the cytokine-induced expression of IL-6 and IL-8 in human endothelial cells. Cytokine 8:58– 65

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

140

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

278. Park SK, Yang WS, Lee SK, Ahn H, Park JS, et al. 2000. TGF-β1 down-regulates inﬂammatory cytokine-induced VCAM-1 expression in cultured human glomerular endothelial cells. Nephrol. Dial. Transplant. 15:596–604 279. Smith WB, Noack L, Khew-Goodall Y, Isenmann S, Vadas MA, Gamble JR. 1996. Transforming growth factor-β1 inhibits the production of IL-8 and the transmigration of neutrophils through activated endothelium. J. Immunol. 157:360–68 280. Luttmann W, Franz P, Matthys H, Virchow JC Jr. 1998. Effects of TGF-β on eosinophil chemotaxis. Scand. J. Immunol. 47:127–30 281. Lagraoui M, Gagnon L. 1997. Enhancement of human neutrophil survival and activation by TGF-β1. Cell. Mol. Biol. 43:313–18 282. Balazovich KJ, Fernandez R, Hinkovska-Galcheva V, Suchard SJ, Boxer LA. 1996. Transforming growth factor-β1 stimulates degranulation and oxidant release by adherent human neutrophils. J. Leukoc. Biol. 60:772–77 283. Myrtek D, Knoll M, Matthiesen T, Krause S, Lohrmann J, et al. 2004. Expression of interleukin-13 receptor alpha 1-subunit on peripheral blood eosinophils is regulated by cytokines. Immunology 112:597–604 284. Alam R, Forsythe P, Stafford S, Fukuda Y. 1994. Transforming growth factor β abrogates the effects of hematopoietins on eosinophils and induces their apoptosis. J. Exp. Med. 179:1041–45 285. Fava RA, Olsen NJ, Postlethwaite AE, Broadley KN, Davidson JM, et al. 1991. Transforming growth factor β1 (TGF-β1) induced neutrophil recruitment to synovial tissues: implications for TGF-β-driven synovial inﬂammation and hyperplasia. J. Exp. Med. 173:1121–32 286. Kobayashi T, Iijima K, Kita H. 2003. Marked airway eosinophilia prevents development of airway hyper-responsiveness during an allergic response in IL-5 transgenic mice. J. Immunol. 170:5756–63 287. Tanaka H, Komai M, Nagao K, Ishizaki M, Kajiwara D, et al. 2004. Role of interleukin-5 and eosinophils in allergen-induced airway remodeling in mice. Am. J. Respir. Cell Mol. Biol. 31:62–68 288. Cho JY, Miller M, Baek KJ, Han JW, Nayar J, et al. 2004. Inhibition of airway remodeling in IL-5-deﬁcient mice. J. Clin. Invest. 113:551–60 289. Prud’homme GJ, Piccirillo CA. 2000. The inhibitory effects of transforming growth factor-β-1 (TGF-β1) in autoimmune diseases. J. Autoimmun. 14:23–42 290. Dang H, Geiser AG, Letterio JJ, Nakabayashi T, Kong L, et al. 1995. SLE-like autoantibodies and Sjogren’s syndrome-like lymphoproliferation in TGF-β knockout mice. J. Immunol. 155:3205–12 291. Yaswen L, Kulkarni AB, Fredrickson T, Mittleman B, Schiffman R, et al. 1996. Autoimmune manifestations in the transforming growth factor-β1 knockout mouse. Blood 87:1439–45 292. Ohtsuka K, Gray JD, Stimmler MM, Toro B, Horwitz DA. 1998. Decreased production of TGF-β by lymphocytes from patients with systemic lupus erythematosus. J. Immunol. 160:2539–45 293. Ohtsuka K, Gray JD, Stimmler MM, Horwitz DA. 1999. The relationship between defects in lymphocyte production of transforming growth factor-β1 in systemic lupus erythematosus and disease activity or severity. Lupus 8:90–94 294. Lowrance JH, O’Sullivan FX, Caver TE, Waegell W, Gresham HD. 1994. Spontaneous elaboration of transforming growth factor β suppresses host defense against bacterial infection in autoimmune MRL/lpr mice. J. Exp. Med. 180:1693–703 www.annualreviews.org • Regulation of Immune Responses by TGF-β

141

ARI

23 February 2006

15:9

295. Raz E, Dudler J, Lotz M, Baird SM, Berry CC, et al. 1995. Modulation of disease activity in murine systemic lupus erythematosus by cytokine gene delivery. Lupus 4:286–92 296. Holmdahl R, Bockermann R, Backlund J, Yamada H. 2002. The molecular pathogenesis of collagen-induced arthritis in mice—a model for rheumatoid arthritis. Ageing Res. Rev. 1:135–47 297. Thorbecke GJ, Shah R, Leu CH, Kuruvilla AP, Hardison AM, Palladino MA. 1992. Involvement of endogenous tumor necrosis factor α and transforming growth factor β during induction of collagen type II arthritis in mice. Proc. Natl. Acad. Sci. USA 89:7375– 79 298. Wahl SM, Allen JB, Costa GL, Wong HL, Dasch JR. 1993. Reversal of acute and chronic synovial inﬂammation by anti-transforming growth factor β. J. Exp. Med. 177:225–30 299. Schramm C, Kriegsmann J, Protschka M, Huber S, Hansen T, et al. 2004. Susceptibility to collagen-induced arthritis is modulated by TGFβ responsiveness of T cells. Arthritis Res. Ther. 6:R114–19 300. King C, Davies J, Mueller R, Lee MS, Krahl T, et al. 1998. TGF-β1 alters APC preference, polarizing islet antigen responses toward a Th2 phenotype. Immunity 8:601–13 301. Grewal IS, Grewal KD, Wong FS, Wang H, Picarella DE, et al. 2002. Expression of transgene encoded TGF-β in islets prevents autoimmune diabetes in NOD mice by a local mechanism. J. Autoimmun. 19:9–22 302. Moritani M, Yoshimoto K, Wong SF, Tanaka C, Yamaoka T, et al. 1998. Abrogation of autoimmune diabetes in nonobese diabetic mice and protection against effector lymphocytes by transgenic paracrine TGF-β1. J. Clin. Invest. 102:499–506 303. Belghith M, Bluestone JA, Barriot S, Megret J, Bach JF, Chatenoud L. 2003. TGF-βdependent mechanisms mediate restoration of self-tolerance induced by antibodies to CD3 in overt autoimmune diabetes. Nat. Med. 9:1202–8 304. Herold KC, Hagopian W, Auger JA, Poumian-Ruiz E, Taylor L, et al. 2002. Anti-CD3 monoclonal antibody in new-onset type 1 diabetes mellitus. N. Engl. J. Med. 346:1692–98 305. Johns LD, Flanders KC, Ranges GE, Sriram S. 1991. Successful treatment of experimental allergic encephalomyelitis with transforming growth factor-β1. J. Immunol. 147:1792–96 306. Racke MK, Dhib-Jalbut S, Cannella B, Albert PS, Raine CS, McFarlin DE. 1991. Prevention and treatment of chronic relapsing experimental allergic encephalomyelitis by transforming growth factor-β1. J. Immunol. 146:3012–17 307. Kuruvilla AP, Shah R, Hochwald GM, Liggitt HD, Palladino MA, Thorbecke GJ. 1991. Protective effect of transforming growth factor β1 on experimental autoimmune diseases in mice. Proc. Natl. Acad. Sci. USA 88:2918–21 308. Racke MK, Cannella B, Albert P, Sporn M, Raine CS, McFarlin DE. 1992. Evidence of endogenous regulatory function of transforming growth factor-β1 in experimental allergic encephalomyelitis. Int. Immunol. 4:615–20 309. Issazadeh S, Mustafa M, Ljungdahl A, Hojeberg B, Dagerlind A, et al. 1995. Interferon γ, interleukin 4 and transforming growth factor β in experimental autoimmune encephalomyelitis in Lewis rats: dynamics of cellular mRNA expression in the central nervous system and lymphoid cells. J. Neurosci. Res. 40:579–90 310. Johns LD, Sriram S. 1993. Experimental allergic encephalomyelitis: neutralizing antibody to TGF β1 enhances the clinical severity of the disease. J. Neuroimmunol. 47:1–7 311. Bai XF, Shi FD, Xiao BG, Li HL, van der Meide PH, Link H. 1997. Nasal administration of myelin basic protein prevents relapsing experimental autoimmune encephalomyelitis in DA rats by activating regulatory cells expressing IL-4 and TGF-β mRNA. J. Neuroimmunol. 80:65–75

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

142

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

312. Nowell PC. 1976. The clonal evolution of tumor cell populations. Science 194:23–28 313. Foulds L. 1954. The experimental study of tumor progression: a review. Cancer Res. 14:327–39 314. Kinzler KW, Vogelstein B. 1996. Lessons from hereditary colorectal cancer. Cell 87:159– 70 315. Weinberg RA. 1996. How cancer arises. Sci. Am. 275:62–70 316. Pardoll D. 2003. Does the immune system see tumors as foreign or self?Annu. Rev. Immunol. 21:807–39 317. Dunn GP, Old LJ, Schreiber RD. 2004. The three Es of cancer immunoediting. Annu. Rev. Immunol. 22:329–60 318. Chang HL, Gillett N, Figari I, Lopez AR, Palladino MA, Derynck R. 1993. Increased transforming growth factor β expression inhibits cell proliferation in vitro, yet increases tumorigenicity and tumor growth of Meth A sarcoma cells. Cancer Res. 53:4391–98 319. Stander M, Naumann U, Dumitrescu L, Heneka M, Loschmann P, et al. 1998. Decorin gene transfer-mediated suppression of TGF-β synthesis abrogates experimental malignant glioma growth in vivo. Gene Ther. 5:1187–94 320. Fakhrai H, Dorigo O, Shawler DL, Lin H, Mercola D, et al. 1996. Eradication of established intracranial rat gliomas by transforming growth factor β antisense gene therapy. Proc. Natl. Acad. Sci. USA 93:2909–14 321. Friese MA, Wischhusen J, Wick W, Weiler M, Eisele G, et al. 2004. RNA interference targeting transforming growth factor-β enhances NKG2D-mediated antiglioma immune response, inhibits glioma cell migration and invasiveness, and abrogates tumorigenicity in vivo. Cancer Res. 64:7596–603 322. Torre-Amione G, Beauchamp RD, Koeppen H, Park BH, et al. 1990. A highly immunogenic tumor transfected with a murine transforming growth factor type β1 cDNA escapes immune surveillance. Proc. Natl. Acad. Sci. USA 87:1486–90 323. Nishikawa H, Kato T, Tawara I, Takemitsu T, Saito K, et al. 2005. Accelerated chemically induced tumor development mediated by CD4+ CD25+ regulatory T cells in wild-type hosts. Proc. Natl. Acad. Sci. USA 102:9253–57 324. Shimizu J, Yamazaki S, Sakaguchi S. 1999. Induction of tumor immunity by removing CD25+ CD4+ T cells: a common basis between tumor immunity and autoimmunity. J. Immunol. 163:5211–18 325. Onizuka S, Tawara I, Shimizu J, Sakaguchi S, Fujita T, Nakayama E. 1999. Tumor rejection by in vivo administration of anti-CD25 (interleukin-2 receptor α) monoclonal antibody. Cancer Res. 59:3128–33 326. Sutmuller RP, van Duivenvoorde LM, van Elsas A, Schumacher TN, Wildenberg ME, et al. 2001. Synergism of cytotoxic T lymphocyte-associated antigen 4 blockade and depletion of CD25+ regulatory T cells in antitumor therapy reveals alternative pathways for suppression of autoreactive cytotoxic T lymphocyte responses. J. Exp. Med. 194:823–32 327. Chen ML, Pittet MJ, Gorelik L, Flavell RA, Weissleder R, et al. 2005. Regulatory T cells suppress tumor-speciﬁc CD8 T cell cytotoxicity through TGF-β signals in vivo. Proc. Natl. Acad. Sci. USA 102:419–24 328. Ormandy LA, Hillemann T, Wedemeyer H, Manns MP, Greten TF, Korangy F. 2005. Increased populations of regulatory T cells in peripheral blood of patients with hepatocellular carcinoma. Cancer Res. 65:2457–64 329. Beyer M, Kochanek M, Darabi K, Popov A, Jensen M, et al. 2005. Reduced frequencies and suppressive function of CD4+ CD25high regulatory T cells in patients with chronic lymphocytic leukemia after therapy with ﬂudarabine. Blood 106:2018–25 www.annualreviews.org • Regulation of Immune Responses by TGF-β

143

ARI

23 February 2006

15:9

330. Hansson GK. 2005. Inﬂammation, atherosclerosis, and coronary artery disease. N. Engl. J. Med. 352:1685–95 331. Binder CJ, Chang MK, Shaw PX, Miller YI, Hartvigsen K, et al. 2002. Innate and acquired immunity in atherogenesis. Nat. Med. 8:1218–26 332. Libby P. 2002. Inﬂammation in atherosclerosis. Nature 420:868–74 333. Mallat Z, Gojova A, Marchiol-Fournigault C, Esposito B, Kamate C, et al. 2001. Inhibition of transforming growth factor-β signaling accelerates atherosclerosis and induces an unstable plaque phenotype in mice. Circ. Res. 89:930–34 334. Lutgens E, Gijbels M, Smook M, Heeringa P, Gotwals P, et al. 2002. Transforming growth factor-β mediates balance between inﬂammation and ﬁbrosis during plaque progression. Arterioscler. Thromb. Vasc. Biol. 22:975–82 335. Robertson AK, Rudling M, Zhou X, Gorelik L, Flavell RA, Hansson GK. 2003. Disruption of TGF-β signaling in T cells accelerates atherosclerosis. J. Clin. Invest. 112:1342–50 336. Gojova A, Brun V, Esposito B, Cottrez F, Gourdy P, et al. 2003. Speciﬁc abrogation of transforming growth factor-β signaling in T cells alters atherosclerotic lesion size and composition in mice. Blood 102:4052–58 337. Argmann CA, Van Den Diepstraten CH, Sawyez CG, Edwards JY, Hegele RA, et al. 2001. Transforming growth factor-β1 inhibits macrophage cholesteryl ester accumulation induced by native and oxidized VLDL remnants. Arterioscler. Thromb. Vasc. Biol. 21:2011– 18 338. Amento EP, Ehsani N, Palmer H, Libby P. 1991. Cytokines and growth factors positively and negatively regulate interstitial collagen gene expression in human vascular smooth muscle cells. Arterioscler. Thromb. 11:1223–30 339. Fabunmi RP, Baker AH, Murray EJ, Booth RF, Newby AC. 1996. Divergent regulation by growth factors and cytokines of 95 kDa and 72 kDa gelatinases and tissue inhibitors or metalloproteinases-1, -2, and -3 in rabbit aortic smooth muscle cells. Biochem. J. 315(Pt. 1):335–42 340. Cipollone F, Fazia M, Mincione G, Iezzi A, Pini B, et al. 2004. Increased expression of transforming growth factor-β1 as a stabilizing factor in human atherosclerotic plaques. Stroke 35:2253–57 341. Grainger DJ, Kemp PR, Metcalfe JC, Liu AC, Lawn RM, et al. 1995. The serum concentration of active transforming growth factor-β is severely depressed in advanced atherosclerosis. Nat. Med. 1:74–79 342. Hansen G, McIntire JJ, Yeung VP, Berry G, Thorbecke GJ, et al. 2000. CD4+ T helper cells engineered to produce latent TGF-β1 reverse allergen-induced airway hyperreactivity and inﬂammation. J. Clin. Invest. 105:61–70 343. Scherf W, Burdach S, Hansen G. 2005. Reduced expression of transforming growth factor β1 exacerbates pathology in an experimental asthma model. Eur. J. Immunol. 35:198–206 344. Redington AE, Madden J, Frew AJ, Djukanovic R, Roche WR, et al. 1997. Transforming growth factor-β1 in asthma. Measurement in bronchoalveolar lavage ﬂuid. Am. J. Respir. Crit. Care Med. 156:642–47 345. Branton MH, Kopp JB. 1999. TGF-β and ﬁbrosis. Microbes Infect. 1:1349–65 346. Bouma G, Strober W. 2003. The immunological and genetic basis of inﬂammatory bowel disease. Nat. Rev. Immunol. 3:521–33 347. Monteleone G, Pallone F, MacDonald TT. 2004. Smad7 in TGF-β-mediated negative regulation of gut inﬂammation. Trends Immunol. 25:513–17 348. Monteleone G, Kumberova A, Croft NM, McKenzie C, Steer HW, MacDonald TT. 2001. Blocking Smad7 restores TGF-β1 signaling in chronic inﬂammatory bowel disease. J. Clin. Invest. 108:601–9

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

144

Li et al.

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

ARI

23 February 2006

15:9

349. Ming M, Ewen ME, Pereira ME. 1995. Trypanosome invasion of mammalian cells requires activation of the TGF β signaling pathway. Cell 82:287–96 350. Silva JS, Twardzik DR, Reed SG. 1991. Regulation of Trypanosoma cruzi infections in vitro and in vivo by transforming growth factor β (TGF-β). J. Exp. Med. 174:539–45 351. Hansen DS, Villacres-Eriksson M, Akerblom L, Hellman U, Segura E, et al. 1998. An immunoafﬁnity-puriﬁed Trypanosoma cruzi antigen suppresses cellular proliferation through a TGF-β-mediated mechanism. Scand. J. Immunol. 47:509–16 352. Olsson T, Bakhiet M, Hojeberg B, Ljungdahl A, Edlund C, et al. 1993. CD8 is critically involved in lymphocyte activation by a T. brucei brucei-released molecule. Cell 72:715–27 353. Sacks D, Noben-Trauth N. 2002. The immunology of susceptibility and resistance to Leishmania major in mice. Nat. Rev. Immunol. 2:845–58 354. Scott P. 1991. IFN-γ modulates the early development of Th1 and Th2 responses in a murine model of cutaneous leishmaniasis. J. Immunol. 147:3149–55 355. Belosevic M, Finbloom DS, Van Der Meide PH, Slayter MV, Nacy CA. 1989. Administration of monoclonal anti-IFN-γ antibodies in vivo abrogates natural resistance of C3H/HeN mice to infection with Leishmania major. J. Immunol. 143:266–74 356. Barral A, Barral-Netto M, Yong EC, Brownell CE, Twardzik DR, Reed SG. 1993. Transforming growth factor β as a virulence mechanism for Leishmania braziliensis. Proc. Natl. Acad. Sci. USA 90:3442–46 357. Gantt KR, Schultz-Cherry S, Rodriguez N, Jeronimo SM, Nascimento ET, et al. 2003. Activation of TGF-β by Leishmania chagasi: importance for parasite survival in macrophages. J. Immunol. 170:2613–20 358. Langermans JA, Nibbering PH, Van Vuren-Van Der Hulst ME, Van Furth R. 2001. Transforming growth factor-β suppresses interferon-γ-induced toxoplasmastatic activity in murine macrophages by inhibition of tumour necrosis factor-α production. Parasite Immunol. 23:169–75 359. Bermudez LE, Covaro G, Remington J. 1993. Infection of murine macrophages with Toxoplasma gondii is associated with release of transforming growth factor β and downregulation of expression of tumor necrosis factor receptors. Infect. Immun. 61:4126–30 360. Omer FM, Kurtzhals JA, Riley EM. 2000. Maintaining the immunological balance in parasitic infections: a role for TGF-β? Parasitol. Today 16:18–23 361. Omer FM, Riley EM. 1998. Transforming growth factor β production is inversely correlated with severity of murine malaria infection. J. Exp. Med. 188:39–48 362. Li C, Corraliza I, Langhorne J. 1999. A defect in interleukin-10 leads to enhanced malarial disease in Plasmodium chabaudi chabaudi infection in mice. Infect. Immun. 67:4435–42 363. Li C, Sanni LA, Omer F, Riley E, Langhorne J. 2003. Pathology of Plasmodium chabaudi chabaudi infection and mortality in interleukin-10-deﬁcient mice are ameliorated by antitumor necrosis factor α and exacerbated by anti-transforming growth factor β antibodies. Infect. Immun. 71:4850–56 364. Omer FM, de Souza JB, Corran PH, Sultan AA, Riley EM. 2003. Activation of transforming growth factor β by malaria parasite-derived metalloproteinases and a thrombospondin-like molecule. J. Exp. Med. 198:1817–27 365. Toossi Z, Young TG, Averill LE, Hamilton BD, Shiratsuchi H, Ellner JJ. 1995. Induction of transforming growth factor β1 by puriﬁed protein derivative of Mycobacterium tuberculosis. Infect. Immun. 63:224–28 366. Dahl KE, Shiratsuchi H, Hamilton BD, Ellner JJ, Toossi Z. 1996. Selective induction of transforming growth factor β in human monocytes by lipoarabinomannan of Mycobacterium tuberculosis. Infect. Immun. 64:399–405 www.annualreviews.org • Regulation of Immune Responses by TGF-β

145

ARI

23 February 2006

15:9

367. Pamer EG. 2004. Immune responses to Listeria monocytogenes. Nat. Rev. Immunol. 4:812– 23 368. Nakane A, Asano M, Sasaki S, Nishikawa S, Miura T, et al. 1996. Transforming growth factor β is protective in host resistance against Listeria monocytogenes infection in mice. Infect. Immun. 64:3901–4 369. Spaccapelo R, Romani L, Tonnetti L, Cenci E, Mencacci A, et al. 1995. TGF-β is important in determining the in vivo patterns of susceptibility or resistance in mice infected with Candida albicans. J. Immunol. 155:1349–60 370. Letterio JJ, Lehrnbecher T, Pollack G, Walsh TJ, Chanock SJ. 2001. Invasive candidiasis stimulates hepatocyte and monocyte production of active transforming growth factor β. Infect. Immun. 69:5115–20 371. Schultz-Cherry S, Hinshaw VS. 1996. Inﬂuenza virus neuraminidase activates latent transforming growth factor β. J. Virol. 70:8624–29 372. Cecere A, Marotta F, Vangieri B, Tancredi L, Gattoni A. 2004. Progressive liver injury in chronic hepatitis C infection is related to altered cellular immune response and to different citokine proﬁle. Panminerva Med. 46:171–87 373. Taniguchi H, Kato N, Otsuka M, Goto T, Yoshida H, et al. 2004. Hepatitis C virus core protein upregulates transforming growth factor-β1 transcription. J. Med. Virol. 72:52–59 374. Cheng PL, Chang MH, Chao CH, Lee YH. 2004. Hepatitis C viral proteins interact with Smad3 and differentially regulate TGF-β/Smad3-mediated transcriptional activation. Oncogene 23:7821–38 375. Reinhold D, Wrenger S, Kahne T, Ansorge S. 1999. HIV-1 Tat: immunosuppression via TGF-β1 induction. Immunol. Today 20:384–85 376. Chen S, Tuttle DL, Oshier JT, Knot HJ, Streit WJ, et al. 2005. Transforming growth factor-β1 increases CXCR4 expression, stromal-derived factor-1α-stimulated signalling and human immunodeﬁciency virus-1 entry in human monocyte-derived macrophages. Immunology 114:565–74 377. Barrios-Rodiles M, Brown KR, Ozdamar B, Bose R, Liu Z, et al. 2005. High-throughput mapping of a dynamic signaling network in mammalian cells. Science 307:S1621–25

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-04

146

Li et al.

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:99-146. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-05

ARI

23 February 2006

15:12

The Eosinophil Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Marc E. Rothenberg and Simon P. Hogan Division of Allergy and Immunology, Department of Pediatrics, Cincinnati Children’s Hospital Medical Center, University of Cincinnati College of Medicine, Cincinnati, Ohio 45229; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:147–74 First published online as a Review in Advance on November 4, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090720 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0147$20.00

Key Words allergy, asthma, cellular trafﬁcking, chemokines, mucosal immunity

Abstract Eosinophils have been considered end-stage cells involved in host protection against parasites. However, numerous lines of evidence have now changed this perspective by showing that eosinophils are pleiotropic multifunctional leukocytes involved in initiation and propagation of diverse inﬂammatory responses, as well as modulators of innate and adaptive immunity. In this review, we summarize the biology of eosinophils, focusing on the growing properties of eosinophil-derived products, including the constituents of their granules as well as the mechanisms by which they release their pleiotropic mediators. We examine new views on the role of eosinophils in homeostatic function, including developmental biology and innate and adaptive immunity (as well as interaction with mast cells and T cells). The molecular steps involved in eosinophil development and trafﬁcking are described, with special attention to the important role of the transcription factor GATA-1, the eosinophil-selective cytokine IL-5, and the eotaxin subfamily of chemokines. We also review the role of eosinophils in disease processes, including infections, asthma, and gastrointestinal disorders, and new data concerning genetically engineered eosinophil-deﬁcient mice. Finally, strategies for targeted therapeutic intervention in eosinophil-mediated mucosal diseases are conceptualized.

147

ANRV270-IY24-05

ARI

23 February 2006

15:12

INTRODUCTION

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Eotaxin family of chemokines: the group of related eosinophil-selective chemoattractant proteins eotaxin-1, eotaxin-2, and eotaxin-3 Major basic protein (MBP): a major protein in eosinophil granules EPO: eosinophil peroxidase

148

Eosinophils are multifunctional leukocytes implicated in the pathogenesis of numerous inﬂammatory processes, including parasitic helminth infections and allergic diseases (1– 3). In response to diverse stimuli, eosinophils are recruited from the circulation into inﬂammatory foci, where they modulate immune responses through an array of mechanisms. Triggering of eosinophils by engagement of receptors for cytokines, immunoglobulins, and complement can lead to the secretion of an array of proinﬂammatory cytokines [IL-2, IL-4, IL-5, IL-10, IL-12, IL-13, IL-16, IL18, and TGF (transforming growth factor)α/β], chemokines (RANTES and eotaxin1), and lipid mediators [platelet-activating factor and leukotriene C4 (LTC4)] (4) (Figure 1). These molecules have proinﬂammatory effects, including upregulation of adhesion systems, modulation of cellular trafﬁcking, and activation and regulation of vascular permeability, mucus secretion, and smooth muscle constriction. Eosinophils can initiate antigen-speciﬁc immune responses by acting as antigen-presenting cells (APCs). Furthermore, eosinophils can serve as major effector cells inducing tissue damage and dysfunction by releasing toxic granule proteins and lipid mediators (5). In this review, we summarize the biology of eosinophils, focusing on the growing properties of eosinophil-derived products, including the constituents of their granules as well as the mechanisms by which they release their pleiotropic mediators. We examine new views on the role of eosinophils in homeostatic function, including developmental biology and innate and adaptive immunity (including interaction with mast cells and T cells). The molecular steps involved in eosinophil development and trafﬁcking are described, with special attention to the important role of the transcription factor GATA1 and the eosinophil-selective cytokine IL-5 and the eotaxin subfamily of chemokines. Furthermore, we review the role of eosinophils

Rothenberg

·

Hogan

in disease processes, including infections, asthma, and gastrointestinal disorders. We also review new data concerning genetically engineered eosinophil-deﬁcient mice. Finally, strategies for targeted therapeutic intervention in eosinophil-mediated diseases are conceptualized.

EOSINOPHIL GRANULE PROTEINS Eosinophils secrete an array of cytotoxic granule cationic proteins [major basic protein (MBP), eosinophil cationic protein (ECP), eosinophil peroxidase (EPO), and eosinophilderived neurotoxin (EDN)] that are capable of inducing tissue damage and dysfunction (5). Eosinophil granules contain a crystalloid core composed of MBP-1 (and MBP2) and a matrix composed of ECP, EDN, and EPO (5). MBP, EPO, and ECP are toxic to a variety of tissues, including heart, brain, and bronchial epithelium (6–9). ECP and EDN are ribonucleases and have been shown to possess antiviral activity, and ECP causes voltage-insensitive, ion-selective toxic pores in the membranes of target cells, possibly facilitating the entry of other cytotoxic molecules (10–13). ECP also has a number of additional noncytotoxic activities, including suppression of T cell proliferative responses and immunoglobulin synthesis by B cells, induction of mast cell degranulation, and stimulation of airway mucus secretion and glycosaminoglycan production by human ﬁbroblasts (14). MBP directly alters smooth muscle contraction responses by dysregulating vagal muscarinic M2 and M3 receptor function and by inducing mast cell and basophil degranulation (15–17). MBP has recently been implicated in regulating peripheral nerve plasticity (18). EPO, which constitutes ∼25% of the total protein mass of speciﬁc granules, catalyzes the oxidation of pseudohalides [thyiocyanate (SCN− )], halides [chloride (Cl− ), bromide (Br− ), iodide (I− )], and nitric oxide (nitrite) to form highly reactive oxygen species

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

Figure 1 Schematic diagram of an eosinophil and its multifunctional effects. Eosinophils are bilobed granulocytes with eosinophilic staining secondary granules. The secondary granules contain four primary cationic proteins, designated eosinophil peroxidase (EPO), major basic protein (MBP), eosinophil cationic protein (ECP), and eosinophil-derived neurotoxin (EDN). All four proteins are cytotoxic molecules; in addition, ECP and EDN are ribonucleases. Eosinophils respond to diverse stimuli, including nonspeciﬁc tissue injury, infections, allografts, allergens, and tumors. In addition to releasing their preformed cationic proteins, eosinophils can also release a variety of cytokines, chemokines, lipid mediators, and neuromodulators. Eosinophils directly communicate with T cells and mast cells in a bidirectional manner. Eosinophils activate T cells by serving as APCs, and eosinophil-derived MBP is a mast cell secretagogue. Eosinophils can also regulate T cell polarization through synthesis of indoleamine 2,3-dioxygenase (IDO), an enzyme involved in oxidative metabolism of tryptophan, catalyzing the conversion of tryptophan to kynurenines (KYN), a regulator of Th1/Th2 balance.

(hypohalous acids) and reactive nitrogen metabolites (perioxynitrate). These molecules oxidize nucleophilic targets on proteins, promoting oxidative stress and subsequent cell death by apoptosis and necrosis (19–21). Eosinophils predominantly secrete their granule protein by regulated exocytosis and degranulation (22). In a process of piecemeal degranulation, eosinophils selectively release components of their speciﬁc granules (23). For example, activation of human eosinophils by IFN-γ promotes the mobilization of granule-derived RANTES to the cell periphery without inducing cationic protein release (24, 25). Regulated exocytosis occurs by the formation of a docking complex com-

posed of soluble N-ethylmaleimide-sensitive factor attachment protein (SNAP) receptors (SNAREs) located on the vesicle (v-SNAREs) and the target membrane (t-SNAREs). SNAREs are classiﬁed into two categories based on the presence of a conserved amino acid (arginine [R] or glutamine [Q]). Human eosinophils express the Q-SNAREs SNAP-23 and syntaxin-4, which are predominantly localized to the plasma membrane (26), and the R-SNARE VAMP (vesicle-associated membrane protein)-2, which is localized to cytoplasmic secretory vesicles. It is postulated that receptor-coupled activation of eosinophils leads to rapid mobilization of cytoplasmic vesicles to the plasma membrane, leading to www.annualreviews.org • The Eosinophil

SNARE: soluble N-ethylmaleimidesensitive factor attachment protein (SNAP) receptor

149

ANRV270-IY24-05

ARI

23 February 2006

15:12

the formation of a SNARE complex (VAMP2/SNAP-23/syntaxin-4) and subsequent mediator release (22).

EOSINOPHILS AND HOMEOSTATIC FUNCTION

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Early clinical investigations have demonstrated an association between eosinophils and parasitic infections, leading investigators to hypothesize that eosinophils were classical end-stage effector cells involved in host defense (27). However, in recent years, eosinophils have been shown to be involved in numerous biological processes, including postpubertal mammary gland development (28), estrus cycling (29, 30), organ transplantation (31), viral infection (13), allergic inﬂammatory responses, and neoplasia (32).

Eosinophils and Reproduction Eosinophils are a prevalent cell population in the female reproductive tract, with numbers reaching maximum levels at estrus. Eosinophils are predominantly localized to the endometrial stroma subadjacent to the luminal and glandular epithelium and at the endometrial-myometrial junction (28). Eosinophil recruitment into the uterus is regulated by IL-5; however, while uterine eosinophil numbers are depleted in IL5-deﬁcient mice, a residual population of eosinophils is still present, and their localization in the subepithelial stroma is comparable to wild-type mice, suggesting that IL-5independent mechanisms regulate the tissuespeciﬁc recruitment of eosinophils into the uterus (30). Consistent with this notion, in response to ovarian steroid hormones, the expression of the eosinophil-active chemokines eotaxin-1, RANTES, and MIP-1α is upregulated, paralleling eosinophil inﬁltration into the uterus (29, 33, 34). Indeed, eotaxin-1deﬁcient mice not only have a deﬁciency of uterine eosinophils, but also have a two-week delay in the onset of estrus, along with a delay in the ﬁrst age of parturition, suggesting a role 150

Rothenberg

·

Hogan

for eosinophils in preparing the mature uterus for pregnancy (35). Furthermore, eosinophils inﬁltrate the endometrium following copulation (36), and investigators have postulated that this cell may have a role in blastocyst implantation and protection against infection; however, this has yet to be proven (37, 38). Interestingly, eosinophil MBP is ectopically expressed by the uterus during pregnancy, but this is not directly related to eosinophils (39). Eosinophils have also been implicated in postnatal mammary gland development (40). Eosinophils reside in the postnatal developing mammary gland and are predominantly localized around the head of the terminal end buds. The expression level of eotaxin-1 mRNA is low between zero and four weeks of age; however, it is signiﬁcantly increased in the mammary gland at ﬁve weeks of age. Notably, increased expression of eotaxin-1 at this time coincides with eosinophil inﬁltration into the head of the terminal end bud (40). Depletion of eosinophils from the postnatal mammary gland by deletion of the eotaxin-1 gene results in reduction in terminal end bud formation and reduced branching complexity of the ductal tree (40). It is likely that eosinophils regulate mammary gland ductal outgrowth through local secretion of eosinophil-derived TGF-β (40).

Thymic Eosinophils Eosinophils migrate into the thymus during the neonatal period, localizing to the corticomedullary region and reaching maximum levels by two weeks of age. Interestingly, their absolute levels are approximately equivalent to that of thymic dendritic cells (41). In mice, a second inﬂux of eosinophils is observed at 16 weeks of age, corresponding to the commencement of thymic involution. Eosinophils localize to the medullary region. Thymic eosinophils express high levels of MHC class II molecules and moderate levels of MHC class I and the costimulatory molecules CD86 (B7.2) and CD30L (CD153) (Figure 2). Furthermore, thymic eosinophils

ARI

23 February 2006

15:12

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

Figure 2 Eosinophil surface markers. This schematic diagram lists surface molecules expressed by human eosinophils. Molecules have been listed generally based on convincing evidence for their expression as assessed by ﬂow cytometry or inferred by cellular responsiveness to speciﬁc stimuli. Cluster designation (CD) for particular molecules is indicated based on the most recent classiﬁcation (www.ncbi.nlm.nih. gov/prow/).

are CD11b/CD11c double positive and appear to be activated as they lose expression of GL-1 and CD62L and upregulate CD25 and CD69 surface expression. Analysis of thymic eosinophil cytokine production reveals that eosinophils express mRNA for the proinﬂammatory cytokines TNF-α, TGF-β, IL1α, and IL-6 and the Th2-cytokines IL-4 and IL-13 (41). Notably, the recruitment of eosinophils into the thymus is regulated by

eotaxin-1, which is constitutively expressed in the thymus (42). It has been postulated that eosinophils are associated with MHC class I–restricted thymocyte deletion. Consistent with this notion, the biphasic recruitment of eosinophils and their anatomical localization within discrete compartments of the thymus coincide with negative selection of double-positive thymocytes (41). Employing an experimental model www.annualreviews.org • The Eosinophil

151

ARI

23 February 2006

15:12

of acute negative selection, researchers have demonstrated increased thymic eosinophil levels in MHC class I–restricted male (HY) antigen T cell receptor (TCR) transgenic mice following cognate peptide injection. In addition, eosinophils are associated with clusters of apoptotic bodies, suggesting eosinophil-mediated MHC class I–restricted thymocyte deletion. Thymic eosinophils have the capacity to promote thymocyte apoptosis as they express costimulatory molecules that are involved in clonal deletions, such as CD30 ligand (CD153) and CD66 (41). Additionally, eosinophils may induce thymocyte apoptosis through free radicals, as thymic eosinophils express high levels of NADPH oxidase activity; notably, developing thymocytes have increased sensitivity to free radicals owing to the downregulation of Cu2+ /Zn2+ superoxide dismutase.

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

EOSINOPHILS AND IMMUNE REGULATION In recent years, investigators have shown that eosinophils can perform numerous immune functions, including antigen presentation (43, 44) and exacerbation of inﬂammatory responses through their capacity to release a range of largely preformed cytokines and lipid mediators (2, 5).

Antigen Presentation Recent clinical and experimental investigations have shown that eosinophils can function as APCs (Figure 1). Eosinophils can process and present a variety of microbial, viral, and parasitic antigens. (45). In addition, granulocyte-macrophage colony stimulating factor (GM-CSF)-treated eosinophils promote T cell proliferation in response to staphylococcal superantigen (Staphylococcus enterotoxins A, B, and E) stimulation (46). Furthermore, eosinophils incubated with human rhinovirus-16 promote rhinovirus-16speciﬁc T cell proliferation and IFN-γ secretion (47). Eosinophils can also effectively 152

Rothenberg

·

Hogan

present soluble antigens to CD4+ T cells, thereby promoting T cell proliferation and polarization. Adoptive transfer of antigenpulsed eosinophils results in eosinophildependent T cell proliferation (44). Furthermore, addition of antigen to eosinophil and T cell cocultures promotes heightened T cell proliferative responses (43). The capacity of eosinophils to present antigen has been debated in some publications. It is interesting to note that the failure of eosinophils to present antigen may be related to the methods used for isolating eosinophils. For example, lysis of erythrocytes with ammonium chloride, an inhibitor of lysosome acidiﬁcation (needed for antigen presentation), negatively correlates with eosinophil antigen presentation activity (43, 48). Eosinophils secrete an array of cytokines (IL-2, IL-4, IL-6, IL-10, IL-12) capable of promoting T cell proliferation, activation, and Th1/Th2 polarization (4, 43, 44, 49) (Figure 1). Recent attention has been drawn to the ability of murine eosinophils to produce IL-4. Employing mice with enhanced green ﬂuorescent protein (GFP) in the IL-4 gene locus (4get mice), investigators have demonstrated that eosinophils are a primary source of GFP following parasitic infection or anti-IgD treatment (a strong Th2 stimulator). Notably, although the IL4 gene locus is transcriptionally active in eosinophils, the amount of IL-4 protein production appears to be lower than in T cells and basophils (50–52). Furthermore, murine eosinophils promote IL-4, IL-5, and IL-13 secretion by CD4+ T cells (44). Eosinophils can also regulate T cell polarization through their synthesis of indoleamine 2,3-dioxygenase (IDO), an enzyme involved in oxidative metabolism of tryptophan, converting tryptophan to kynurenines (KYN). KYN regulates Th1 and Th2 imbalance by promoting Th1 cell apoptosis (53). The eosinophilmediated T cell proliferative and cytokine secretion responses are dependent on costimulation. Indeed, blockade of CD80, CD86, and CTLA-4 by neutralizing antibodies inhibits

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

eosinophil-elicited T cell proliferation and cytokine secretion (45). Fluorescent labeling studies revealed that eosinophils instilled into the trachea of mice trafﬁc into the draining peritracheal lymph nodes and localize to the T cell–rich paracortical regions (B cell zones) within 24 h (43). Employing models of allergic airway disease and gastrointestinal allergy, investigators have demonstrated that inhalation of antigen promotes eosinophil homing to the draining endotracheal lymph nodes and Peyer’s patches (44, 54–56). Interestingly, a recent investigation suggests that eosinophils can only promote proliferation of effector T cells but not naive T cells (48). Moreover, eosinophils pulsed with OVA peptide and cocultured with OVAspeciﬁc TCR transgenic T cells (D011.10 T cells) induced effector T cell proliferation; however, when cocultured with naive CD4+ T cells, no T cell proliferation was observed. It is tempting to speculate that eosinophils trafﬁc to draining lymph nodes to recruit activated effector T cells and promote proliferation of effector T cells.

Mast Cell Regulation A substantial body of literature has emerged demonstrating that eosinophils have the capacity to regulate mast cell function (Figure 1). Notably, human umbilical cord blood–derived mast cells can be activated by MBP to release histamine, PGD-2, GM-CSF, TNF-α, and IL-8 (57). The activation of mast cells by MBP elicits not only exocytosis, but also eicosanoid generation and cytokine production, both of which are prominent responses following FcRI-dependent activation of mast cells (57). Incubation of rat peritoneal mast cells with native MBP, EPO, and ECP (but not EDN) results in concentration-dependent histamine release (15). Several studies have shown that MBP induces mast cell activation via a pathway similar to that observed with other polybasic compounds such as substance P, com-

pound 48/80, and bradykinin (16). Freshly isolated human lung mast cells are resistant to IgE-independent activation; however, highly puriﬁed lung mast cells cocultured with human lung ﬁbroblasts are sensitive to IgEindependent activation by MBP (57). Interestingly, activation of eosinophils with the mast cell protease chymase promotes production of eosinophil-derived stem cell factor, a critical mast cell growth factor. Eosinophils also produce nerve growth factor (NGF) (58), a cytokine not only involved in survival and functional maintenance of sympathetic neurons but also in immune regulation. For example, NGF promotes mast cell survival and activation (59, 60). NGF is preformed in eosinophils and acts in an autocrine fashion by activating release of EPO (58). EPO activates rat peritoneal muscles to release histamine, suggesting a role for eosinophil-derived NGF in mast cell–eosinophil interactions. Thus, eosinophils and mast cells communicate in a bidirectional fashion.

EOSINOPHIL DEVELOPMENT Eosinophils are produced in the bone marrow from pluripotential stem cells, which ﬁrst differentiate into a hybrid precursor with shared properties of basophils and eosinophils and then into a separate eosinophil lineage (61). Eosinophil lineage speciﬁcation is dictated by the interplay of at least three classes of transcription factors, including GATA-1 (a zinc family ﬁnger member), PU.1 (an ETS family member), and C/EBP members (CCAAT/enhancer-binding protein family) (62–64) (Figure 3). Although these transcription factors are expressed in a variety of hematopoietic lineages, their mechanism of action in eosinophils is unique. In particular, graded expression of PU.1 speciﬁes distinct cell lineage fates, with low levels specifying lymphocytic and high levels myeloid differentiation (65–67). Although GATA-1 and PU.1 antagonize each other’s function in most cell types, they have synergistic activity in regulating eosinophil lineage speciﬁcation (and www.annualreviews.org • The Eosinophil

153

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

15:12

Figure 3 Schematic representation of eosinophil trafﬁcking. Eosinophils develop in the bone marrow, where they differentiate from hematopoietic progenitor cells into mature eosinophils under the control of critical transcription factors, especially GATA-1. The eosinophilopoietins IL-3, IL-5, and GM-CSF regulate eosinophil expansion, especially in conditions of hypereosinophilia. Eosinophil migration out of the bone marrow into the circulation is primarily regulated by IL-5. Circulating eosinophils subsequently interact with the endothelium by processes involving rolling, adhesion, and diapedesis. Depending on the target organ, eosinophils cross the endothelium into tissues by a regulated process involving the coordinated interaction between networks involving the chemokine eotaxin-1, eosinophil adhesion molecules (α4 β1 , α4 β7 , αm β2 , αL β2 ), and adhesion receptors on the endothelium (MAdCAM-1, VCAM-1, and ICAM-1). Under homeostatic conditions, eosinophils trafﬁc into the thymus, mammary gland, uterus, and most prominently into the gastrointestinal tract.

eosinophil granule protein transcription) (67). The speciﬁcity of these factors for eosinophils is conserved across species, as C/EBP factors and GATA-1 drive differentiation of chicken progenitor cells into eosinophils (62). Of these 154

Rothenberg

·

Hogan

transcription factors, GATA-1 is clearly the most important for eosinophil lineage speciﬁcation, as revealed by the loss of the eosinophil lineage in mice harboring a targeted deletion of the high-afﬁnity GATA-binding site in

ANRV270-IY24-05

ARI

23 February 2006

15:12

Figure 4

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Double GATA site. Sequence alignment of the palindromic double GATA high-afﬁnity binding site identiﬁed in a hypersensitivity region of the murine GATA-1 regulatory locus with the dual GATA sites found in the promoters of three human eosinophil lineage-selective genes, including the human IL-5Rα gene promoter 1, human MBP promoter 2, and the human CCR3 regulatory exon-1. This ﬁgure was kindly provided by Drs. J. Du and S. Ackerman, University of Illinois, Chicago.

the GATA-1 promoter (68) and by eosinophil differentiation experiments in vitro (69). In particular, the speciﬁc activity of GATA-1 in eosinophils but not other GATA-1+ lineages (mast cells, megakaryocytes, and erythroid cells) appears to be mediated by a high-afﬁnity palindromic (or double) GATA site (67). This double GATA site is present in the downstream GATA-1 promoter and also in the regulatory regions of eosinophil-speciﬁc genes, including the eotaxin receptor CC chemokine receptor-3 (CCR3), MBP, and the IL-5 receptor alpha (IL-5Rα) gene (Figure 4), and it accounts for eosinophil-speciﬁc gene expression (67, 68, 70). For example, the tandem double GATA site in the human MBP-P2 promoter is required for both promoter activity in human eosinophil cell lines and for the synergistic transactivation by GATA-1 and PU.1 (67). Three cytokines, IL-3, IL-5, and GMCSF, are particularly important in regulating eosinophil development (71–74) (Figure 3). These eosinophilopoietins likely provide permissive proliferative and differentiation signals following the instructive signals speciﬁed by the transcription factors GATA-1, PU.1, and C/EBPs. These cytokines are encoded by closely linked genes on chromosome 5q31. They bind to receptors that share a common beta chain and have unique alpha chains (75). Of these three cytokines, IL-5 is the most speciﬁc to the eosinophil lineage and is responsible for selective differentiation of eosinophils (76). IL-5 also stimulates the release of eosinophils from the bone

marrow into the peripheral circulation (77). The critical role of IL-5 in the production of eosinophils is best demonstrated by genetic manipulation of mice. Overproduction of IL-5 in transgenic mice results in profound eosinophilia (78–81), and deletion of the IL-5 gene causes a marked reduction of eosinophils in the blood and lungs after allergen challenge (82, 83). The overproduction of one or a combination of these three cytokines occurs in humans with eosinophilia, and diseases with selective eosinophilia are often accompanied by overproduction of IL-5 (84). The critical role of IL-5 in regulating eosinophils in humans has been demonstrated by several clinical trials with humanized anti-IL-5 antibody; this currently unapproved drug dramatically lowers eosinophil levels in the blood and to a lesser extent in the inﬂamed lung (85–87).

Double GATA site: a high-afﬁnity palindromic (or double) GATA site located in the regulatory region of eosinophil-speciﬁc genes CCR3: CC chemokine receptor-3, the major eosinophil chemokine receptor that binds the eotaxin family of chemokines Eosinophilopoietins: IL-3, IL-5, and GM-CSF

EOSINOPHIL TRAFFICKING Under baseline conditions, most eosinophils trafﬁc into the gastrointestinal tract where they normally reside within the lamina propria of all segments except the esophagus (88) (Figure 3). The gastrointestinal eosinophil is the predominant population of eosinophils. Under baseline conditions, eosinophil levels in the gastrointestinal tract occur independently of lymphocytes and enteric ﬂora, indicating unique regulation compared with other leukocytes (88). Indeed, the recruitment of gastrointestinal eosinophils is regulated by the constitutive expression of eotaxin-1, www.annualreviews.org • The Eosinophil

155

ARI

23 February 2006

15:12

as demonstrated by the marked decrease of this population of eosinophils in eotaxin-1deﬁcient mice. The importance of eotaxin-1 in regulating the baseline level of eosinophils is reinforced by the observation that mice with the targeted deletion of CCR3 (but not eotaxin-2-deﬁcient mice) also have a deﬁciency in gastrointestinal eosinophils (89, 90). In addition to trafﬁcking into the gastrointestinal tract, under homeostatic conditions, eosinophils home into the thymus, mammary gland, and uterus, also under the regulation of eotaxin-1 (40, 91) (Figure 3). Of note, trafﬁcking into the uterus is regulated by estrogen, as eosinophil and eotaxin-1 levels cycle along with estrus (29). The trafﬁcking of eosinophils into inﬂammatory sites involves a number of cytokines (in particular, Th2 and endothelial cell products IL-4, IL-5, and IL-13) (92–94), adhesion molecules (e.g., β1-, β2-, and β7integrins) (95), chemokines (e.g., RANTES and the eotaxins) (96), and other recently identiﬁed molecules (e.g., acidic mammalian chitinase) (97). Tissue eosinophils likely can survive for at least two weeks based on in vitro observations (92). Of the cytokines implicated in modulating leukocyte recruitment, only IL-5 and the eotaxins selectively regulate eosinophil trafﬁcking (98). IL-5 regulates growth, differentiation, activation, and survival of eosinophils and provides an essential signal for the expansion and mobilization of eosinophils from the bone marrow into the lung following allergen exposure (77). However, antigen-induced tissue eosinophilia can occur independently of IL-5, as demonstrated by residual tissue eosinophils in trials using anti-IL-5 in patients with asthma (86) and using IL-5-deﬁcient mice (82, 99). Recent studies have demonstrated an important role for the eotaxin subfamily of chemokines in eosinophil recruitment to the lung (96). Eotaxin was initially discovered using a biological assay in guinea pigs designed to identify the molecules responsible for allergeninduced eosinophil accumulation in the lungs (98, 100, 101). Subsequently, using genomic

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

156

Rothenberg

·

Hogan

analyses, two additional chemokines were identiﬁed in the human genome that encode for CC chemokines with eosinophilselective chemoattractant activity and have thus been designated eotaxin-2 and eotaxin3 (96). Eotaxin-2 and eotaxin-3 are only distantly related to eotaxin-1 because they are only ∼30% identical in sequence and are located in a different chromosomal position (102, 103). The speciﬁc activity of all eotaxins is mediated by the selective expression of the seven-transmembrane spanning, G protein–coupled receptor CCR3, primarily expressed on eosinophils (104–106). Notably, the eotaxin chemokines cooperate with IL-5 in the induction of tissue eosinophilia. IL-5 increases the pool of eotaxin-responsive cells and primes eosinophils to respond to CCR3 ligands (96). Furthermore, when given exogenously, eotaxins cooperate with IL-5 to induce substantial production of IL-13 in the lung (96). The ﬁnding that IL-4 and IL-13 are potent inducers of the eotaxin chemokines by a STAT6-dependent pathway provides an integrated mechanism to explain the eosinophilia associated with Th2 responses (96). Recent studies have identiﬁed that eosinophil recruitment to the lung is dependent on STAT6 and a bone marrow–derived lung tissue resident non-T or non-B cell (51); in particular, eotaxin-2 production by airway macrophages likely accounts for this (90, 107). Of further interest, recently CCR3 has been shown also to deliver a powerful negative signal in eosinophils, depending on the ligand engaged. For example, pretreatment with the chemokine Mig inhibits eosinophil responses by a CCR3- and Rac2-dependent mechanism (108). Using eotaxin-1 and eotaxin-2 single- and double-gene-deﬁcient mice or neutralizing antibodies, investigators have shown that both chemokines have nonoverlapping roles in regulating the temporal and regional distribution of eosinophils in an allergic inﬂammatory site (90, 109, 110). In a standard experimental asthma model induced by systemic sensitization with OVA/alum followed

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

by respiratory OVA challenge, only a modest reduction in lung eosinophils was found in CCR3-deﬁcient mice (89). However, when the same CCR3-deﬁcient mouse line was subjected to experimental asthma induction by epicutaneous OVA sensitization, there was a marked deﬁciency of lung and bronchoalveolar lavage eosinophils (111). It was proposed that these apparently conﬂicting results may be related to the sensitization protocol (111), but the reason for this apparent discrepancy remains unclear. Notably, another CCR3-deﬁcient mouse strain has recently been shown to have a profound reduction in eosinophil recruitment to the lung in the standard OVA/alum systemic sensitization model (107). Substantial preclinical evidence now supports a role for the eotaxin chemokines in human allergic disease (96). Experimental induction of cutaneous and pulmonary latephase responses in humans has revealed that the eotaxin chemokines are produced by tissue resident cells (e.g., respiratory epithelial cells and skin ﬁbroblasts) and allergeninduced inﬁltrative cells (e.g., macrophages and eosinophils). Following allergen challenge in the human lung, eotaxin-1 is induced early (6 h) and correlates with early eosinophil recruitment; in contrast, eotaxin-2 correlates with eosinophil accumulation at 24 h (96). In another study, eotaxin-1 and eotaxin-2 mRNA was increased in patients with asthma compared with normal controls; however, there was no further increase following allergen challenge (96). In contrast, eotaxin-3 mRNA was dramatically enhanced 24 h after allergen challenge (96). The chemoattractant activity of the bronchoalveolar lavage ﬂuid from patients with asthma is inhibited by antibodies against RANTES, MCP (monocyte chemoattractant protein)-3, MCP-4, and eotaxin-1 (96). Further support for an important role of eotaxin-1 in human asthma is derived from analysis of a single nucleotide polymorphism (SNP) in the eotaxin-1 gene. A naturally occurring mutation encoding for a change in the last amino acid in the signal peptide

(alanine→threonine) results in less effective cellular secretion of eotaxin-1 in vitro and in vivo (112). Notably, this SNP is associated with reduced levels of circulating eotaxin-1 and eosinophils and improved lung function (e.g., FEV1) (112). Furthermore, a SNP in the eotaxin-3 gene is associated with atopy in a Korean population (113). Recently, the activity of eotaxin-1 and eotaxin-2 in humans has been investigated by injection of these chemokines into the skin of humans; both eotaxin-1 and eotaxin-2 induce an immediate wheal and ﬂare response associated with mast cell degranulation and subsequent inﬁltrations by eosinophils, basophils, and neutrophils (114). The inﬁltration by neutrophils is likely to be mediated indirectly by the mast cell degranulation. These results provide substantial evidence that the biological activities attributed to eotaxins in animals are conserved in humans. Eosinophils express numerous adhesion molecules, and most attention has focused on their highly expressed integrins, including α4 β7 , the CD18 family of molecules (β2integrins), and the very late antigen (VLA)4 molecules (β1-integrins) (95) (Figure 1). The CD18 family of molecules includes lymphocyte function antigen (LFA)-1 and Mac-1 that interact with endothelial cells via intercellular adhesion molecule (ICAM)-1. VLA4 interacts with endothelium via vascular cell adhesion molecule (VCAM)-1, as well as ﬁbronectin. The α4 β7 integrin interacts with the mucosal addressin cell adhesion molecule (MAdCAM)-1 expressed by vascular endothelium in the intestinal tract. These integrins have variable roles in eosinophil trafﬁcking during inﬂammation, but the role of speciﬁc adhesion molecules in the baseline homing of eosinophils into the gastrointestinal tract has yet to be elucidated. For example, in β7 genetargeted mice, there is a delay and reduced magnitude in the development of intestinal eosinophilia following Trichinella spiralis infection (115) and when the eotaxin-1 intestine transgene is expressed, but no changes in the baseline level of small intestine eosinophils www.annualreviews.org • The Eosinophil

157

ARI

23 February 2006

15:12

(81). Analysis of anti-β1-treated mice or VLA-4-deﬁcient mice has shown the critical participation of this family of molecules in regulating eosinophil homing to the allergic lung (116–118). Indeed, eotaxin-1-stimulated eosinophils have increased expression and avidity of VLA-4 (119). It has also become clear that engagement of eosinophil adhesion molecules with their ligands not only induces a proadhesive pathway, but also activates expression of a series of proinﬂammatory genes within eosinophils, including GMCSF, that then propagate eosinophil survival by a paracrine pathway. Numerous other pathways for regulating eosinophil accumulation and trafﬁcking are operational in various inﬂammatory models. However, recently several lines of evidence have focused attention on the importance of arachidonic acid metabolites, especially leukotriene B4 (LTB4), the cysteinyl leukotrienes (LTC4, LTCD4, and LTE4), and prostaglandin (PG) D2. Notably, cysteinyl leukotriene type 1 receptor antagonists (now approved for asthma therapy) reduce blood and lung eosinophilia. Mice with the targeted deletion of the LTB4 receptor also have markedly reduced allergen-induced lung eosinophilia (120). Furthermore, eosinophils express high levels of a high-afﬁnity PGD2 type 2 receptor. Interestingly, this receptor is also expressed by basophils and Th2 cells [and is now designated chemoattractant receptor Th2 cells (CRTH2)] and appears to co-mediate Th2 cell and eosinophil/basophil recruitment (121). Eosinophils have recently also been shown to express high levels of the histamine receptor 4 (H4) that mediates eosinophil chemoattraction and activation in vitro (122).

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ROLE OF EOSINOPHILS IN DISEASE Infections The beneﬁcial function of eosinophils has been primarily attributed to their ability to 158

Rothenberg

·

Hogan

defend the host against parasitic helminths. This is based on several lines of evidence, including (a) the ability of eosinophils to mediate antibody- (or complement-) dependent cellular toxicity against helminths in vitro (27), (b) the observation that eosinophil levels increase during helminthic infections and that eosinophils aggregate and degranulate in the local vicinity of damaged parasites in vivo, and (c) the results in experimental parasite infected mice that have been depleted of eosinophils by IL-5 neutralization and/or gene targeting (123). Murine studies are particularly problematic because mice are not the natural hosts of many of the experimental parasites; nevertheless, in some primary infection models, a role for IL-5 in protective immunity has been suggested following infection with Strongyloides venezuelensis, Strongyloides ratti, Nippostrongyloides brasiliensis, and Heligmosomoides polygyrus (123, 124). These in vivo studies need to be interpreted with caution because IL-5 neutralization may have effects on other IL-5 receptor bearing cells (including murine B cells, human basophils, and possibly human respiratory smooth muscle cells) (76, 125–127). Other approaches, including analysis of CCR3- and eotaxin-1deﬁcient mice, have recently demonstrated a role for eosinophils in the encystment of larvae in Trichinella spiralis and in controlling the Brugia malayi microﬁlariae, respectively (128, 129). Perhaps analysis of the recently generated eosinophil-deﬁcient mice following experimental parasitic infection will provide further compelling evidence that eosinophils participate in host defense against parasites. Thus, although the debate continues, it seems likely that eosinophils participate in the protective immunity against selected helminths. Evidence is emerging that eosinophils may also have a protective role in other infections, especially against RNA viruses such as respiratory syncytial virus (RSV) and the related natural rodent pathogen, pneumonia virus of mice (PVM), in vivo (13, 130). Notably, eosinophil granule proteins include abundant

ANRV270-IY24-05

ARI

23 February 2006

15:12

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ribonucleases [such as human ECP and EDN, and at least 11 eosinophil-associated ribonuclease (EAR) orthologs in mice] that degrade single-stranded RNA containing viruses (13). In fact, ECP and EDN are the most divergent coding sequences in the entire human genome (compared with other primates) (13). Despite their divergence, they have conserved ribonuclease activity across species, strongly implicating evolutionary pressure to preserve this critical enzymatic activity.

Asthma Elevated levels of eosinophil granule proteins (e.g., MBP) have been found in bronchoalveolar lavage ﬂuid from patients with asthma, and importantly these concentrations are sufﬁcient to induce cytotoxicity of a variety of host tissue, including respiratory epithelial cells in vitro (3). Direct degranulation of mast cells and basophils, triggered by MBP, is thought also to be involved in disease pathogenesis (3). In addition to being cytotoxic, MBP directly increases smooth muscle reactivity by causing dysfunction of vagal muscarinic M2 receptors, which is thought to contribute to the development of airway hyperreactivity (AHR), a cardinal feature of asthma (131). Additionally, eosinophils generate large amounts of the cysteinyl leukotrienes (132). Of note, eosinophil granule proteins contain all the biochemical machinery necessary to synthesize cysteinyl leukotrienes (132). These mediators lead to increased vascular permeability and mucus secretion and are potent smooth muscle constrictors. Indeed, inhibitors of cysteinyl leukotrienes are effective therapeutic agents for the treatment of allergic airway disease. Multiple studies employing experimental models of asthma (primarily in mice, guinea pigs, and monkeys) have demonstrated that neutralization of IL-5 can block various aspects of asthma (82, 133). Although extensive investigations have implicated the eosinophil as a central effector cell in asthma and an important clinical target for the resolution

of this disease, the role of this granulocyte in the development and exacerbation of asthma pathogenesis has been controversial. This controversy stems in part from distinctions between human asthma and experimental murine models of asthma. For example, in contrast to human asthma, mice with eosinophil lung disease triggered by allergens or helminthic infection have variable levels of eosinophil degranulation (50, 134). In experimental models, inhibition of the actions of IL-5 consistently suppresses pulmonary eosinophilia in response to antigen inhalation; however, this effect does not always correlate with a reduction of AHR (135). This dichotomy is highlighted by ﬁndings in allergic IL-5-deﬁcient mice of the C57BL/6 strain (82) that do not develop antigen-induced AHR, whereas IL-5-deﬁcient BALB/c mice develop enhanced reactivity independent of this factor (136). Although eosinophil trafﬁcking to the allergic lung is profoundly attenuated in IL-5-deﬁcient mice or in those treated with anti-IL-5 antibodies in comparison to wild-type responses (137–139), a marked residual tissue eosinophilia can persist in these mice after allergen inhalation (82, 140, 141). Furthermore, the degree of residual tissue eosinophilia is under genetic regulation, as lung eosinophilia is 10- to 100-fold greater in the BALB/c strain, where AHR persists, compared with the C57BL/6 strain, where AHR is abolished in the absence of IL-5 (82, 99, 138). Studies with transgenic mice overexpressing IL-5 (in T cells, lung epithelial cells, or enterocytes) have demonstrated that overexpression of IL-5 is sufﬁcient for the development of eosinophilia (78–81, 142); however, elevated levels of eosinophils are not universally associated with the development of asthma-like changes in the lung. Indeed, clinical studies in patients have shown that AHR correlates with mast cell localization near pulmonary nerves, whereas pulmonary eosinophilia relates more strongly with chronic cough (143). However, depletion of murine eosinophils (by administration www.annualreviews.org • The Eosinophil

Airway hyperresponsiveness or hyperreactivity (AHR): increased constriction of the airways to various stimuli such as methacholine

159

ANRV270-IY24-05

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Airway remodeling: microscopic changes (e.g., goblet cell metaplasia, collagen deposition, smooth muscle hyperplasia) in the lungs associated with functional alterations in lung function PHIL mice: genetically engineered eosinophil-deﬁcient mice produced by insertion of the diphtheria toxin A chain into the EPO gene locus dbl-GATA-1 mice: genetically engineered eosinophil-deﬁcient mice produced by deleting the high-afﬁnity double GATA site in the GATA-1 promoter

15:12

of complement-ﬁxing antibodies against CCR3) has demonstrated an important role for eosinophils in the development of asthmaassociated AHR (144); a role for other CCR3+ cells was not ruled out, but there was no evidence for CCR3 expression by non-eosinophils (144). Accordingly, a humanized antibody against IL-5 has recently been tested for asthma (85). In the early studies with this reagent, patients with mild to moderate asthma were shown to have a drop in their circulating and sputum eosinophil levels (85); however, no clinical beneﬁt (e.g., improvement in FEV1) was demonstrated. This result prompted some investigators to conclude that eosinophils were not effector cells in human asthma (85); however, the anti-IL5 study was not properly designed to address the efﬁcacy of this drug (145). In support of these preclinical studies, a very recent study has demonstrated that anti-IL-5 in humans blocks lung eosinophil recruitment by only 55% (146), providing evidence that accessory molecules (in addition to IL-5) regulate lung eosinophilia. Thus, anti-IL-5 treatment does not completely resolve tissue eosinophilia in the allergic lung, and therefore this cell may still contribute to disease pathogenesis even in

Table 1 Effect of eosinophil depletion on experimental asthma parametersa Mouse line

PHIL

Δdbl-GATA

Asthma parameter BALF eosinophils Lung tissue eosinophils BALF mononuclear cells AHR Mucus production Collagen deposition Th2 antibody production Th2 cytokines

+ + NE + + ND ND +

+ + NE NE NE + NE NE

a If the genetic manipulation of the mouse resulted in protection from or reduction in severity of the asthma parameter, the parameter is labeled with a “+”. If there was no change in the asthma parameter between the genetically modiﬁed mouse and wild-type control mice, the parameter is labeled with “NE” for no effect. If the asthma parameter was not measured, the parameter is labeled with “ND” for not determined.

160

Rothenberg

·

Hogan

the presence of IL-5 neutralization. With the discovery of the eotaxins, and the ﬁnding that IL-5 cooperates with eotaxins in regulating eosinophil tissue recruitment, it became critical to determine if the lack of efﬁcacy of antiIL-5 in humans was related to the inability of this drug to block eosinophil tissue recruitment or to the noneffector role of eosinophils. One possibility is that local chemokine systems (eotaxins) can operate independently of IL-5 to recruit eosinophils into the allergic lung. Studies with eotaxin-1 gene-targeted mice, IL-5 gene-targeted mice, and eotaxin1/IL-5 double-gene-targeted mice have revealed an independent and synergistic role for both of these molecules in regulating the tissue level of eosinophils in the asthmatic lung and in the induction of AHR (139). Although early studies with anti-IL-5 in human asthma have continued to ﬁnd no improvement in airﬂow measurements (FEV1), pathological markers of chronic airway remodeling (e.g., deposition of tenascin, procollagen III, and lumican) are improved by anti-IL-5 (146). Decreased levels of TGF-β in the bronchoalveolar lavage ﬂuid following anti-IL-5 treatment have been found, suggesting that eosinophilderived TGF-β regulates lung remodeling. In support for a role of eosinophils in the pathogenesis of human asthma, a very recent study has demonstrated improved clinical outcome when asthma treatment decisions are based on monitoring sputum eosinophil counts rather than conventional guidelines from the British Thoracic Society (147). Recently, two different lines of eosinophildeﬁcient mice were developed (see Table 1 and Eosinophil-Deﬁcient Mice). Lee et al. (148) targeted the depletion of eosinophils by using an eosinophil-speciﬁc promoter to drive expression of a cytocidal protein, diphtheria toxin A chain. The eosinophil-deﬁcient character of these mice (called PHIL mice) was assessed by examination of peripheral blood and by immunohistochemistry of tissues with abundant resident populations (e.g., bone marrow, uterus, small intestine, and thymus) using antibodies speciﬁc for eosinophil

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

granule proteins. In comparison, Yu et al. (68) developed mice harboring a deletion of a high-afﬁnity GATA-binding site in the GATA-1 promoter (dbl-GATA) which led to the speciﬁc ablation of the eosinophil lineage. RT-PCR analysis of gene expression in the bone marrow of the dbl-GATA mice revealed no expression of EPO, but expression of MBP was only partially reduced and CCR3 expression remained unchanged. Nevertheless, eosinophil deﬁciency in these mice was veriﬁed by morphological observation of cells from the blood, bone marrow, and spleen. Using both lines of eosinophildeﬁcient mice, eosinophils were shown to have an integral role in experimental allergic asthma. However, their speciﬁc contribution toward allergen-induced AHR and mucus cell metaplasia was different (Table 1). Perhaps dbl-GATA mice have residual eosinophils or unappreciated hematological abnormalities, or alternatively, diptheria toxin treatment of PHIL mice may induce toxic effects on noneosinophils; these and other explanations for the distinct results will hopefully be uncovered soon. It should be noted that dbl-GATA mice had impaired development of lung remodeling in a chronic model of asthma, consistent with the results of anti-IL-5 in patients with asthma. Taken together, compelling evidence now exists that eosinophils are prominent effector cells in eliciting multiple parameters of experimental asthma.

Gastrointestinal Disorders The accumulation of eosinophils in the gastrointestinal tract is a common feature of numerous disorders, such as drug reactions, helminth infections, hypereosinophilic syndromes, eosinophilic gastroenteritis, allergic colitis, inﬂammatory bowel disease, and gastroesophageal reﬂux disease (150). A subset of these diseases, referred to as primary eosinophil-associated gastrointestinal disorders (EGID), includes eosinophilic esophagitis (EE), eosinophilic gastritis, and

Eosinophil-Deficient Mice Two different lines of eosinophil-deﬁcient mice have recently been developed. One group targeted the depletion of eosinophils using an eosinophil-speciﬁc promoter (the EPO gene) to drive expression of a cytocidal protein diphtheria toxin A (13). These mice (called PHIL) are protected from the development of AHR in a model of experimental asthma. Another group developed mice harboring a deletion of the high-afﬁnity GATA-binding site in the GATA-1 promoter (dbl-GATA); this led to the speciﬁc ablation of the eosinophil lineage even when these mice were crossed with IL-5 transgenic mice (48). The dbl-GATA mice are protected from features of airway remodeling but not AHR in an experimental model of asthma. It is anticipated that these newly generated eosinophil-deﬁcient mouse lines will transform eosinophil research over the next decade, especially because the dbl-GATA mice are now commercially available from Jackson Laboratories, Inc.

eosinophilic gastroenteritis. These are hypersensitivity disorders that lie in the middle of a spectrum ranging from anaphylaxis to Celiac disease (150). EGID usually occurs independently of peripheral blood eosinophilia, indicating the signiﬁcance of gastrointestinalspeciﬁc mechanisms for regulating eosinophil levels. Indeed, in murine models of EGID, a deﬁnitive role for eosinophils and eotaxin-1 has been demonstrated. Notably, eosinophils are frequently associated near damaged enteric nerves, and indeed eotaxin-1-deﬁcient mice are protected from this feature of disease. EE is distinguished from gastroesophageal reﬂux disease by several important differences, including the relatively higher prevalence of atopy, dysphagia, male gender, familial inheritance, degree of proximal esophagitis, and intensity of esophageal pathology [e.g., epithelial hyperplasia and eosinophil density (generally >24 eosinophils/high power ﬁeld)] (150). Consistent with the high rate of atopic respiratory disease in patients with EE, experimental EE develops in mice following respiratory allergen exposure or following intratracheal IL-13 delivery (151). These www.annualreviews.org • The Eosinophil

Hypereosinophilic syndrome: a group of disorders characterized by severely elevated blood eosinophil levels and end-organ damage EGID: eosinophil-associated gastrointestinal disorders EE: eosinophilic esophagitis

161

ANRV270-IY24-05

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

FIP1L1-PDGFRA: activated tyrosine kinase fusion gene product that occurs in hypereosinophilic syndrome owing to an 800-kb interstitial deletion in chromosome 4

15:12

results establish an intimate immunological connection between the lung and esophagus. The epithelial hyperplasia associated with EE and the level of esophageal eosinophils is attenuated in IL-5-deﬁcient mice (152), providing strong evidence that eosinophils are effector cells in this gastrointestinal disease. Indeed, a recent preliminary evaluation of humanized anti-IL-5 in patients with EE demonstrates lowering of esophageal eosinophil levels. Supporting a connection between allergic responses in the lung and gastrointestinal tract, eotaxin-1 intestine transgenic mice not only develop intestinal eosinophilia but also AHR by an IL-13dependent mechanism (153). Thus, increased expression of eotaxin-1 in the gastrointestinal compartment can lead to increased CD4+ T cell–derived Th2 lymphocyte-cytokine production that drives aberrant immunophysiological responses in distant noninﬂamed mucosal tissue (the lung). These results provide a possible explanation for the altered lung function seen in some patients with inﬂammatory gastrointestinal disorders.

ANTI-EOSINOPHIL THERAPEUTICS Numerous drugs inhibit eosinophil production or eosinophil-derived products. They include glucocorticoids, myelosuppressive drugs, leukotriene synthesis or receptor antagonists, tyrosine kinase inhibitors, IFN-α, and humanized anti-IL-5 antibodies. The etiology of the primary disease often speciﬁes the best therapeutic strategy. For example, a subset of patients with hypereosinophilic syndrome have an 800-kb interstitial deletion on chromosome 4 (4q12) that results in the fusion of an unknown gene FIP1L1 with the platelet-derived growth factor receptor-α (PDGFRA) gene (154, 155). This fusion gene produces a constitutively active tyrosine kinase (PDGFRA) that is exquisitely sensitive to the inhibitor imatinib mesylate, which is now approved for the treatment of several malig162

Rothenberg

·

Hogan

nancies (GleevecTM ). Although PDGFRA is not normally active in hematopoietic cells, the activated kinase renders cells growth factor independent, perhaps by activating STAT5 signal transduction. Thus, eosinophilic patients with FIP1L1-PDGFRA+ disease are now treated with GleevecTM as ﬁrst-line therapy (156). In addition, a variety of other activated tyrosine kinases have just been associated with hypereosinophilic syndromes, including PDGFRB, Janus kinase-2, and ﬁbroblast growth factor receptor-1. In most other individuals, glucocorticoids are the most effective agents for reducing eosinophilia (3). They suppress the transcription of a number of genes for inﬂammatory mediators, including the genes for IL-3, IL-4, IL-5, GM-CSF, and various chemokines including the eotaxins. Recently, the main action of glucocorticoids on eosinophil-active cytokines has been shown to involve mRNA destabilization, thus reducing the half-life of cytokines such as eotaxins (157). In addition, glucocorticoids inhibit the cytokinedependent survival of eosinophils (158). Systemic or topical (inhaled or intranasal) glucocorticoid treatment typically causes a rapid reduction in eosinophils, but some patients are glucocorticoid resistant and maintain eosinophilia despite high doses (159). The mechanism of glucocorticoid resistance is unclear, but a reduced level of glucocorticoid receptors and alterations in transcription factor activator protein (AP)-1 appear to be at least partially responsible (159). Glucocorticoid-resistant patients sometimes require other therapy such as myelosuppressive drugs (hydroxyurea, vincristine) or IFN-α (3). IFN-α can be especially helpful because it inhibits eosinophil degranulation and effector function (160). Notably, patients with myeloproliferative variants of hypereosinophilic syndrome can often go into remission with IFN-α therapy. Cyclophilins (e.g., cyclosporine A) have also been used because they block the transcription of numerous eosinophil-active cytokines (e.g.,

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

IL-5, GM-CSF) (3). Recently, lidocaine has been shown to shorten eosinophil survival, and its effects mimic those of glucocorticoids and are noncytotoxic (161). Indeed, an early clinical trial has shown that nebulized lidocaine is safe and effective in subjects with asthma (162). Drugs that interfere with eosinophil chemotactic signals include recently approved leukotriene antagonists and inhibitors. 5lipoxygenase inhibition (e.g., zileuton) blocks the rate-limiting step in leukotriene synthesis and inhibits the generation of the eosinophil chemoattractant, LTB4, and the cysteinyl leukotrienes (163). Cysteinyl leukotriene receptor antagonists block the muscle contraction and increased vascular permeability mediated by leukocyte-derived leukotrienes (164). Some of the third generation antihistamines inhibit the vacuolization (165) and accumulation (166) of eosinophils after allergen challenge and directly inhibit eosinophils in vitro (165, 167). Cromoglycate and nedocromil inhibit the effector function of eosinophils, such as antibody-dependent cellular cytotoxicity (167). The identiﬁcation of molecules that specifically regulate eosinophil function and/or production offers new therapeutic strategies in the pipeline. Agents that interrupt eosinophil adhesion to the endothelium through the interaction of CD18/ICAM-1 (168) or VLA-4 /VCAM-1 may be useful (169, 170). Indeed, antibodies that block these pathways have recently been approved for other diseases, but their anti-eosinophil activity has yet to be determined (171). Antibodies against IL-5, now humanized by two different pharmaceutical companies, are under active clinical investigation (172, 173). Although their utility for asthma may be limited owing to redundant pathways, anti-IL-5 is particularly promising for hypereosinophilic syndromes. Numerous inhibitors of the eotaxin/CCR3 pathway, including small molecule inhibitors of CCR3 and a human anti-eotaxin-1 antibody, are being developed (96). Early results with a phase I trial of human anti-eotaxin-1 antibody in

patients with allergic rhinitis have demonstrated the ability of this apparently safe drug to lower levels of nasal eosinophils and to improve nasal patency (96). Anti-human IL-13 antibody is now in preclinical trials (174) and looks promising for lowering tissue eosinophil levels. Finally, a recently identiﬁed eosinophil surface molecule Siglec-8 may offer a therapeutic opportunity (175). Siglec-8 is a member of the sialic acid–binding lectin family and contains ITIMs (immunoreceptor tyrosinebased inhibitory motifs) that can induce efﬁcient eosinophil apoptosis when engaged by anti-Siglec-8 crosslinking antibodies. Siglec8 as well as CCR3 and CRTH2 are coexpressed by other cells involved in Th2 responses, including Th2 cells, mast cells, and basophils. Thus, agents that block these receptors may be particularly useful for allergic disorders.

PERSPECTIVE Historically, eosinophils have been considered end-stage cells involved in host protection against parasites. However, numerous lines of evidence have now changed this perspective by showing that eosinophils are pleiotropic multifunctional leukocytes involved in initiation and propagation of diverse inﬂammatory responses, as well as modulators of adaptive immunity by directly activating T cells. As normal constituents of the mucosal immune system, particularly in the gastrointestinal tract, eosinophils are likely to have a physiological function. Indeed, eosinophils have been implicated in innate immunity by being an early and possibly instrumental source of cytokines (e.g., IL-4) and have a role in developmental processes such as mammary gland development. Analysis of recently generated genetically engineered eosinophil-deﬁcient mice will soon answer critical questions concerning the true involvement of this cell type in a variety of processes. Breakthroughs in identifying key eosinophil regulatory cytokines such as IL-5 and the eotaxin subfamily of chemokines www.annualreviews.org • The Eosinophil

163

ANRV270-IY24-05

ARI

23 February 2006

15:12

have uncovered mechanisms that selectively regulate eosinophil production and localization at baseline and during inﬂammatory responses. In particular, an integrated mechanism involving Th2 cell–derived IL-5 regulating eosinophil expansion in the bone marrow and blood and Th2 cell–derived IL-13 reg-

ulating eotaxin production now explains the means by which T cells regulate eosinophils. Based on these ﬁndings, targeted therapy against key eosinophil regulators (e.g., humanized anti-IL-5 and CCR3 antagonists) will likely transform medical management of eosinophilic patients.

DISCLOSURE STATEMENT

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

M.R. has consulted on an advisory board for GlaxoSmithKline, received stock and an honorarium for consulting for Ception Therapeutics, received a speaker’s honorarium from Merck and Tanox, and has consulted and received a research grant from Cambridge Antibody Technology.

ACKNOWLEDGMENTS Dr. Rothenberg’s laboratory is indebted to the following grants and/or organizations that have partly ﬁnanced the work presented in this review: NIAID R01 AI045898, NIAID R01 AI057803, NHLBI P01 HL076383, FDA FD-R-002313, the Campaign Urging Research for Eosinophilic Disorders (CURED), and the Burroughs Wellcome Fund. The authors thank numerous instrumental colleagues who have contributed to the ideas formulated in this review, including Drs. K. Frank Austen, Fred Finkelman, Paul Foster, Ian Young, Nives Zimmermann, Anil Mishra, Steven Ackerman, James Lee, and Gerald Gleich, the dedicated laboratory workers, and Andrea Lippelman for editorial assistance. The authors are grateful to Drs. Jamie and Nancy Lee, Craig Gerard, and Alison Humbles for sharing their innovative reagents, and to the International Eosinophil Society that recently hosted a meeting where some of the information in this review was formulated.

LITERATURE CITED 1. Gleich GJ, Loegering DA. 1984. Immunobiology of eosinophils. Annu. Rev. Immunol. 2:429–59 2. Weller PF. 1994. Eosinophils: structure and functions. Curr. Opin. Immunol. 6:85–90 3. Rothenberg ME. 1998. Eosinophilia. N. Engl. J. Med. 338:1592–600 4. Kita H. 1996. The eosinophil: a cytokine-producing cell? J. Allergy Clin. Immunol. 97:889– 92 5. Gleich GJ, Adolphson CR. 1986. The eosinophilic leukocyte: structure and function. Adv. Immunol. 39:177–253 6. Tai P-C, Hayes DJ, Clark JB, Spry CJF. 1982. Toxic effects of eosinophil secretion products on isolated rat heart cells in vitro. Biochem. J. 204:75–80 7. Venge P, Dahl R, Hallgren R, Olsson I. 1980. Cationic proteins of human eosinophils and their role in the inﬂammatory reaction. In The Eosinophil in Health and Disease, ed. AAF Mahmoud, KF Austen, pp. 1131–42. New York: Grune & Stratton 8. Frigas E, Loegering DA, Gleich GJ. 1980. Cytotoxic effects of the guinea pig eosinophil major basic protein on tracheal epithelium. Lab. Invest. 42:35–43 9. Gleich GJ, Frigas E, Loegering DA, Wassom DL, Steinmuller D. 1979. The cytotoxic properties of the eosinophil major basic protein. J. Immunol. 123:2925 10. Young JD, Peterson CG, Venge P, Cohn ZA. 1986. Mechanism of membrane damage mediated by human eosinophil cationic protein. Nature 321:613–16 164

Rothenberg

·

Hogan

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

11. Slifman NR, Loegering DA, McKean DJ, Gleich GJ. 1986. Ribonuclease activity associated with human eosinophil-derived neurotoxin and eosinophil cationic protein. J. Immunol. 137:2913–17 12. Gleich GJ, Loegering DA, Bell MP, Checkel JL, Ackerman SJ, McKean DJ. 1986. Biochemical and functional similarities between human eosinophil-derived neurotoxin and eosinophil cationic protein: homology with ribonuclease. Proc. Natl. Acad. Sci. USA 83:3146–50 13. Rosenberg HF, Domachowske JB. 2001. Eosinophils, eosinophil ribonucleases, and their role in host defense against respiratory virus pathogens. J. Leukoc. Biol. 70:691– 98 14. Venge P, Bystrom J, Carlson M, Hakansson L, Karawacjzyk M, et al. 1999. Eosinophil cationic protein (ECP): molecular and biological properties and the use of ECP as a marker of eosinophil activation in disease. Clin. Exp. Allergy 29:1172–86 15. Zheutlin LM, Ackerman SJ, Gleich GJ, Thomas LL. 1984. Stimulation of basophil and rat mast cell histamine release by eosinophil granule-derived cationic proteins. J. Immunol. 133:2180–85 16. Piliponsky AM, Pickholtz D, Gleich GJ, Levi-Schaffer F. 2001. Human eosinophils induce histamine release from antigen-activated rat peritoneal mast cells: a possible role for mast cells in late-phase allergic reactions. J. Allergy Clin. Immunol. 107:993–1000 17. Jacoby DB, Costello RM, Fryer AD. 2001. Eosinophil recruitment to the airway nerves. J. Allergy Clin. Immunol. 107:211–18 18. Morgan RK, Costello RW, Durcan N, Kingham PJ, Gleich GJ, et al. 2005. Diverse effects of eosinophil cationic granule proteins on IMR-32 nerve cell signalling and survival. Am. J. Respir. Cell Mol. Biol. 33:169–77 19. Agosti JM, Altman LC, Ayars GH, Loegering DA, Gleich GJ, Klebanoff SJ. 1987. The injurious effect of eosinophil peroxidase, hydrogen peroxide, and halides on pneumocytes in vitro. J. Allergy Clin. Immunol. 79:496–504 20. Wu W, Chen Y, Hazen SL. 1999. Eosinophil peroxidase, nitrates, protein tyrosyl residues. Implications for oxidative damage by nitrating intermediates in eosinophilic inﬂammatory disorders. J. Biol. Chem. 274:25933–44 21. MacPherson JC, Comhair SA, Erzurum SC, Klein DF, Lipscomb MF, et al. 2001. Eosinophils are a major source of nitric oxide-derived oxidants in severe asthma: Characterization of pathways available to eosinophils for generating reactive nitrogen species. J. Immunol. 166:5763–72 22. Logan MR, Odemuyiwa SO, Moqbel R. 2003. Understanding exocytosis in immune and inﬂammatory cells: the molecular basis of mediator secretion. J. Allergy Clin. Immunol. 111:923–32 23. Dvorak AM, Furitsu T, Letourneau L, Ishizaka T, Ackerman SJ. 1991. Mature eosinophils stimulated to develop in human cord blood mononuclear cell cultures supplemented with recombinant human interleukin-5. Part I. Piecemeal degranulation of speciﬁc granules and distribution of Charcot-Leyden crystal protein. Am. J. Pathol. 138:69– 82 24. Lacy P, Mahmudi-Azer S, Bablitz B, Hagen SC, Velazquez JR, et al. 1999. Rapid mobilization of intracellularly stored RANTES in response to interferon-γ in human eosinophils. Blood 94:23–32 25. Bandeira-Melo C, Gillard G, Ghiran I, Weller PF. 2000. EliCell: a gel-phase dual antibody capture and detection assay to measure cytokine release from eosinophils. J. Immunol. Methods 244:105–15 www.annualreviews.org • The Eosinophil

165

ARI

23 February 2006

15:12

26. Logan MR, Lacy P, Bablitz B, Moqbel R. 2002. Expression of eosinophil target SNAREs as potential cognate receptors for vesicle-associated membrane protein-2 in exocytosis. J. Allergy Clin. Immunol. 109:299–306 27. Butterworth AE. 1977. The eosinophil and its role in immunity to helminth infection. Curr. Top. Microbiol. Immunol. 77:127–68 28. Sferruzzi-Perri AN, Robertson SA, Dent LA. 2003. Interleukin-5 transgene expression and eosinophilia are associated with retarded mammary gland development in mice. Biol. Reprod. 69:224–33 29. Gouon-Evans V, Pollard JW. 2001. Eotaxin is required for eosinophil homing into the stroma of the pubertal and cycling uterus. Endocrinology 142:4515–21 30. Robertson SA, Mau VJ, Young IG, Matthaei KI. 2000. Uterine eosinophils and reproductive performance in interleukin 5-deﬁcient mice. J. Reprod. Fertil. 120:423–32 31. Nagral A, Ben-Ari Z, Dhillon AP, Burroughs AK. 1998. Eosinophils in acute cellular rejection in liver allografts. Liver Transpl. Surg. 4:355–62 32. Tepper RI, Coffman RL, Leder P. 1992. An eosinophil-dependent mechanism for the antitumor effect of interleukin-4. Science 257:548–51 33. Robertson SA, Allanson M, Mau VJ. 1998. Molecular regulation of uterine recruitment during early pregnancy in the mouse. Trophobl. Res. 11:101–20 34. Zhang J, Lathbury LJ, Salamonsen LA. 2000. Expression of the chemokine eotaxin and its receptor, CCR3, in human endometrium. Biol. Reprod. 62:404–11 35. Gouon-Evans V, Lin EY, Pollard JW. 2002. Requirement of macrophages and eosinophils and their cytokines/chemokines for mammary gland development. Breast Cancer Res. 4:155–64 36. McMaster MT, Newton RC, Dey SK, Andrews GK. 1992. Activation and distribution of inﬂammatory cells in the mouse uterus during the preimplantation period. J. Immunol. 148:1699–705 37. Robertson SA, Mau VJ, Hudson SA, Tremellen KP. 1997. Cytokine-leukocyte networks and the establishment of preganancy in the mouse. J. Reprod. Fertil. 107:265–77 38. De M, Choudhuri R, Wood GW. 1991. Determination of the number and distribution of macrophages, lymphocytes and granulocytes in the mouse uterus from mating through implantation. J. Leukoc. Biol. 50:252–62 39. Wagner JM, Hustin J, Bonno M, Kephart GM, Gurian KV, Gleich GJ. 1994. Pregnancyassociated major basic protein: deposition of protein and expression of mRNA at the maternal-fetal junction in early and late gestation. Placenta 15:625–40 40. Gouon-Evans V, Rothenberg ME, Pollard JW. 2000. Postnatal mammary gland development requires macrophages and eosinophils. Development 127:2269–82 41. Throsby M, Herbelin A, Pleau JM, Dardenne M. 2000. CD11c+ eosinophils in the murine thymus: developmental regulation and recruitment upon MHC class I-restricted thymocyte deletion. J. Immunol. 165:1965–75 42. Matthews AN, Friend DS, Zimmermann N, Saraﬁ MN, Luster AD, et al. 1998. Eotaxin is required for the baseline level of tissue eosinophils. Proc. Natl. Acad. Sci. USA 95:6273– 78 43. Shi HZ, Humbles A, Gerard C, Jin Z, Weller PF. 2000. Lymph node trafﬁcking and antigen presentation by endobronchial eosinophils. J. Clin. Invest. 105:945–53 44. MacKenzie JR, Mattes J, Dent LA, Foster PS. 2001. Eosinophils promote allergic disease of the lung by regulating CD4+ Th2 lymphocyte function. J. Immunol. 167:3146–55 45. Shi H. 2004. Eosinophils function as antigen-presenting cells. J. Leukoc. Biol. 76:520– 27

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

166

Rothenberg

·

Hogan

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

46. Mawhorter SD, Kazura JW, Boom WH. 1994. Human eosinophils as antigen-presenting cells: relative efﬁciency for superantigen- and antigen-induced CD4+ T-cell proliferation. Immunology 81:584–91 47. Handzel ZT, Busse WW, Sedgwick JB, Vrtis R, Lee WM, et al. 1998. Eosinophils bind rhinovirus and activate virus-speciﬁc T cells. J. Immunol. 160:1279–84 48. van Rijt LS, Vos N, Hijdra D, De Vries VC, Hoogsteden HC, Lambrecht BN. 2003. Airways eosinophils accumulate in the mediastinal lymph nodes but lack antigen-presenting potential for naive T-cells. J. Immunol. 171:3372–78 49. Lacy P, Moqbel R. 2000. Eosinophil cytokines. Chem. Immunol. 76:134–55 50. Shinkai K, Mohrs M, Locksley RM. 2002. Helper T cells regulate type-2 innate immunity in vivo. Nature 420:825–29 51. Voehringer D, Shinkai K, Locksley RM. 2004. Type 2 immunity reﬂects orchestrated recruitment of cells committed to IL-4 production. Immunity 20:267–77 52. Khodoun MV, Orekhova T, Potter C, Morris S, Finkelman FD. 2004. Basophils initiate IL-4 production during a memory T-dependent response. J. Exp. Med. 200:857–70 53. Odemuyiwa SO, Ghahary A, Li Y, Puttagunta L, Lee JE, et al. 2004. Human eosinophils regulate T cell subset selection through indoleamine 2,3-dioxyenase. J. Immunol. 173:5909–13 54. Korsgren M, Erjefalt JS, Korsgren O, Sundler F, Persson CGA. 1997. Allergic eosinophilrich inﬂammation develops in lungs and airways of B cell-deﬁcient mice. J. Exp. Med. 185:885–92 55. Hogan SP, Mishra A, Brandt EB, Royalty MP, Pope SM, et al. 2001. A pathological function for eotaxin and eosinophils in eosinophilic gastrointestinal inﬂammation. Nat. Immunol. 2:353–60 56. Mishra A, Hogan SP, Brandt EB, Rothenberg ME. 2000. Peyer’s patch eosinophils: identiﬁcation, characterization, and regulation by mucosal allergen exposure, interleukin-5, and eotaxin. Blood 96:1538–44 57. Piliponsky AM, Gleich GJ, Bar I, Levi-Schaffer F. 2002. Effects of eosinophils on mast cells: a new pathway for the perpetuation of allergic inﬂammation. Mol. Immunol. 38:1369 58. Solomon A, Aloe L, Pe’er J, Frucht-Pery J, Bonini S, Levi-Schaffer F. 1998. Nerve growth factor is preformed in and activates human peripheral blood eosinophils. J. Allergy Clin. Immunol. 102:454–60 59. Bullock ED, Johnson EM Jr 1996. Nerve growth factor induces the expression of certain cytokine genes and bcl-2 in mast cells. Potential role in survival promotion. J. Biol. Chem. 271:27500–8 60. Horigome K, Bullock ED, Johnson EM Jr. 1994. Effects of nerve growth factor on rat peritoneal mast cells. Survival promotion and immediate-early gene induction. J. Biol. Chem. 269:2695–702 61. Boyce JA, Friend D, Matsumoto R, Austen KF, Owen WF. 1995. Differentiation in vitro of hybrid eosinophil/basophil granulocytes: autocrine function of an eosinophil developmental intermediate. J. Exp. Med. 182:49–57 62. McNagny K, Graf T. 2002. Making eosinophils through subtle shifts in transcription factor expression. J. Exp. Med. 195:F43–47 63. Nerlov C, Graf T. 1998. PU.1 induces myeloid lineage commitment in multipotent hematopoietic progenitors. Genes Dev. 12:2403–12 64. Nerlov C, McNagny KM, Doderlein G, Kowenz-Leutz E, Graf T. 1998. Distinct C/EBP functions are required for eosinophil lineage commitment and maturation. Genes Dev. 12:2413–23 www.annualreviews.org • The Eosinophil

167

ARI

23 February 2006

15:12

65. DeKoter RP, Singh H. 2000. Regulation of B lymphocyte and macrophage development by graded expression of PU.1. Science 288:1439–41 66. Walsh JC, DeKoter RP, Lee HJ, Smith ED, Lancki DW, et al. 2002. Cooperative and antagonistic interplay between PU.1 and GATA-2 in the speciﬁcation of myeloid cell fates. Immunity 17:665–76 67. Du J, Stankiewicz MJ, Liu Y, Xi Q, Schmitz JE, et al. 2002. Novel combinatorial interactions of GATA-1, PU.1, and C/EBP isoforms regulate transcription of the gene encoding eosinophil granule major basic protein. J. Biol. Chem. 277:43481– 94 68. Yu C, Cantor AB, Yang H, Browne C, Wells RA, et al. 2002. Targeted deletion of a high-afﬁnity GATA-binding site in the GATA-1 promoter leads to selective loss of the eosinophil lineage in vivo. J. Exp. Med. 195:1387–95 69. Hirasawa R, Shimizu R, Takahashi S, Osawa M, Takayanagi S, et al. 2002. Essential and instructive roles of GATA factors in eosinophil development. J. Exp. Med. 195:1379– 86 70. Zimmermann N, Daugherty BL, Kavanaugh JL, El-Awar FY, Moulton EA, Rothenberg ME. 2000. Analysis of the CC chemokine receptor 3 gene reveals a complex 5 exon organization, a functional role for untranslated exon 1, and a broadly active promoter with eosinophil-selective elements. Blood 96:2346–54 71. Lopez AF, Begley CG, Williamson DJ, Warren DJ, Vadas MA, Sanderson CJ. 1986. Murine eosinophil differentiation factor. An eosinophil-speciﬁc colony-stimulating factor with activity for human cells. J. Exp. Med. 163:1085–99 72. Rothenberg ME, Pomerantz JL, Owen WF, Avraham S, Soberman RJ, et al. 1988. Characterization of a human eosinophil proteoglycan, and augmentation of its biosynthesis and size by interleukin 3, interleukin 5, and granulocyte/macrophage colony stimulating factor. J. Biol. Chem. 263:13901–8 73. Lopez AF, Sanderson CJ, Gamble JR, Campbell HD, Young IG, Vadas MA. 1988. Recombinant human interleukin 5 is a selective activator of human eosinophil function. J. Exp. Med. 167:219–24 74. Takatsu K, Takaki S, Hitoshi Y. 1994. Interleukin-5 and its receptor system: implications in the immune system and inﬂammation. Adv. Immunol. 57:45–90 75. Vadas M, Lopez A, Gamble J, Khew-Goodall Y, Smith W, et al. 1994. Cytokines and allergy. J. Allergy Clin. Immunol. 94:1289–93 76. Sanderson CJ. 1992. Interleukin-5, eosinophils, and disease. Blood 79:3101–9 77. Collins PD, Marleau S, Grifﬁths-Johnson DA, Jose PJ, Williams TJ. 1995. Cooperation between interleukin-5 and the chemokine eotaxin to induce eosinophil accumulation in vivo. J. Exp. Med. 182:1169–74 78. Dent LA, Strath M, Mellor AL, Sanderson CJ. 1990. Eosinophilia in transgenic mice expressing interleukin 5. J. Exp. Med. 172:1425–31 79. Tominaga A, Takaki S, Koyama N, Katoh S, Matsumoto R, et al. 1991. Transgenic mice expressing a B cell growth and differentiation factor (interleukin 5) develop eosinophilia and autoantibody production. J. Exp. Med. 173:429–39 80. Lee J, McGarry M, Farmer S, Denzler K, Larson K, et al. 1997. Interleukin-5 expression in the lung epithelium of transgenic mice leads to pulmonary changes pathogmomonic of asthma. J. Exp. Med. 185:2143–56 81. Mishra A, Hogan SP, Brandt EB, Wagner N, Crossman MW, et al. 2002. Enterocyte expression of the eotaxin and interleukin-5 transgenes induces compartmentalized dysregulation of eosinophil trafﬁcking. J. Biol. Chem. 277:4406–12

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

168

Rothenberg

·

Hogan

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

82. Foster P, Hogan S, Ramsay A, Matthaei K, Young I. 1996. Interleukin-5 deﬁciency abolishes eosinophilia, airway hyperreactivity and lung damage in a mouse asthma model. J. Exp. Med. 183:195–201 83. Kopf M, Brombacher F, Hodgkin PD, Ramsay AJ, Milbourne EA, et al. 1996. IL-5deﬁcient mice have a developmental defect in CD5+ B-1 cells and lack eosinophilia but have normal antibody and cytotoxic T cell responses. Immunity 4:15–24 84. Owen WF, Rothenberg ME, Petersen J, Weller PF, Silberstein D, et al. 1989. Interleukin 5 and phenotypically altered eosinophils in the blood of patients with the idiopathic hypereosinophilic syndrome. J. Exp. Med. 170:343–48 85. Leckie MJ, ten Brinke A, Khan J, Diamant Z, O’Connor BJ, et al. 2000. Effects of an interleukin-5 blocking monoclonal antibody on eosinophils, airway hyper-responsiveness, and the late asthmatic response. Lancet 356:2144–48 86. Flood-Page P, Phipps S, Menzies-Gow A, Ong YE, Kay AB. 2003. Effect of intravenous administration of an anti-IL-5 (mepolizumab) on allergen-induced tissue eosinophilia, the late-phase allergic reaction and the expression of a marker of repair/remodeling in human atopic subjects. J. Allergy Clin. Immunol. 111:S261 87. Kips JC, O’Connor BJ, Langley SJ, Woodcock AA, Kerstjens HA, et al. 2003. Effect of SCH55700, a humanized anti-human interleukin-5 antibody, in severe persistent asthma: a pilot study. Am. J. Respir. Crit. Care Med. 167:1655–59 88. Mishra A, Hogan SP, Lee JJ, Foster PS, Rothenberg ME. 1999. Fundamental signals that regulate eosinophil homing to the gastrointestinal tract. J. Clin. Invest. 103:1719– 27 89. Humbles AA, Lu B, Friend DS, Okinaga S, Lora J, et al. 2002. The murine CCR3 receptor regulates both the role of eosinophils and mast cells in allergen-induced airway inﬂammation and hyperresponsiveness. Proc. Natl. Acad. Sci. USA 99:1479–84 90. Pope SM, Fulkerson PC, Blanchard C, Akei HS, Nikolaidis NM, et al. 2005. Identiﬁcation of a cooperative mechanism involving interleukin-13 and eotaxin-2 in experimental allergic lung inﬂammation. J. Biol. Chem. 280:13952–61 91. Rothenberg ME, Mishra A, Brandt EB, Hogan SP. 2001. Gastrointestinal eosinophils in health and disease. Adv. Immunol. 78:291–328 92. Rothenberg ME, Owen WF Jr, Silberstein DS, Soberman RJ, Austen KF, Stevens RL. 1987. Eosinophils co-cultured with endothelial cells have increased survival and functional properties. Science 237:645–47 93. Sher A, Coffman RL, Hieny S, Cheever AW. 1990. Ablation of eosinophil and IgE responses with anti-IL-5 or anti-IL-4 antibodies fails to affect immunity against Schistosoma mansoni in the mouse. J. Immunol. 145:3911–16 94. Horie S, Okubo Y, Hossain M, Sato E, Nomura H, et al. 1997. Interleukin-13 but not interleukin-4 prolongs eosinophil survival and induces eosinophil chemotaxis. Intern. Med. 36:179–85 95. Bochner BS, Schleimer RP. 1994. The role of adhesion molecules in human eosinophil and basophil recruitment. J. Allergy Clin. Immunol. 94:427–38 96. Zimmermann N, Hershey GK, Foster PS, Rothenberg ME. 2003. Chemokines in asthma: cooperative interaction between chemokines and IL-13. J. Allergy Clin. Immunol. 111:227– 42 97. Zhu Z, Zheng T, Homer RJ, Kim YK, Chen NY, et al. 2004. Acidic mammalian chitinase in asthmatic Th2 inﬂammation and IL-13 pathway activation. Science 304:1678–82 98. Rankin SM, Conroy DM, Williams TJ. 2000. Eotaxin and eosinophil recruitment: implications for human disease. Mol. Med. Today 6:20–27 www.annualreviews.org • The Eosinophil

169

ARI

23 February 2006

15:12

99. Hogan SP, Koskinen A, Foster PS. 1997. Interleukin-5 and eosinophils induce airway damage and bronchial hyperreactivity during allergic airway inﬂammation in BALB/c mice. Immunol. Cell Biol. 75:284–88 100. Jose PJ, Grifﬁths-Johnson DA, Collins PD, Walsh DT, Moqbel R, et al. 1994. Eotaxin: a potent eosinophil chemoattractant cytokine detected in a guinea pig model of allergic airways inﬂammation. J. Exp. Med. 179:881–87 101. Rothenberg ME, Luster AD, Lilly CM, Drazen JM, Leder P. 1995. Constitutive and allergen-induced expression of eotaxin mRNA in the guinea pig lung. J. Exp. Med. 181:1211–16 102. Zimmermann N, Hogan SP, Mishra A, Brandt EB, Bodette TR, et al. 2000. Murine eotaxin-2: a constitutive eosinophil chemokine induced by allergen challenge and IL-4 overexpression. J. Immunol. 165:5839–46 103. Shinkai A, Yoshisue H, Koike M, Shoji E, Nakagawa S, et al. 1999. A novel human CC chemokine, eotaxin-3, which is expressed in IL-4-stimulated vascular endothelial cells, exhibits potent activity toward eosinophils. J. Immunol. 163:1602–10 104. Murphy PM. 1994. The molecular biology of leukocyte chemoattractant receptors. Annu. Rev. Immunol. 12:593–633 105. Ponath PD, Qin S, Post TW, Wang J, Wu L, et al. 1996. Molecular cloning and characterization of a human eotaxin receptor expressed selectively on eosinophils. J. Exp. Med. 183:2437–48 106. Daugherty BL, Siciliano SJ, Demartino JA, Malkowitz L, Sirotina A, Springer MS. 1996. Cloning, expression, and characterization of the human eosinophil eotaxin receptor. J. Exp. Med. 183:2349–54 107. Pope SM, Zimmermann N, Stringer KF, Karow ML, Rothenberg ME. 2005. The eotaxin chemokines and CCR3 are fundemental regulators of allergen-induced pulmonary eosinophilia. J. Immunol. 175:5341–50 108. Fulkerson PC, Zhu H, Williams DA, Zimmermann N, Rothenberg ME. 2005. CXCL9 inhibits eosinophil responses by a CCR3- and Rac2-dependent mechanism. Blood 106:436– 43 109. Rothenberg ME, MacLean JA, Pearlman E, Luster AD, Leder P. 1997. Targeted disruption of the chemokine eotaxin partially reduces antigen-induced tissue eosinophilia. J. Exp. Med. 185:785–90 110. Gonzalo JA, Lloyd CM, Wen D, Albar JP, Wells TN, et al. 1998. The coordinated action of CC chemokines in the lung orchestrates allergic inﬂammation and airway hyperresponsiveness. J. Exp. Med. 188:157–67 111. Ma W, Bryce PJ, Humbles AA, Laouini D, Yalcindag A, et al. 2002. CCR3 is essential for skin eosinophilia and airway hyperresponsiveness in a murine model of allergic skin inﬂammation. J. Clin. Invest. 109:621–28 112. Nakamura H, Luster AD, Nakamura T, In KH, Sonna LA, et al. 2001. Variant eotaxin: its effects on the asthma phenotype. J. Allergy Clin. Immunol. 108:946–53 113. Chae SC, Park YR, Oh GJ, Lee JH, Chung HT. 2005. The suggestive association of eotaxin-2 and eotaxin-3 gene polymorphisms in Korean population with allergic rhinitis. Immunogenetics 56:760–64 114. Menzies-Gow A, Ying S, Sabroe I, Stubbs VL, Soler D, et al. 2002. Eotaxin (CCL11) and eotaxin-2 (CCL24) induce recruitment of eosinophils, basophils, neutrophils, and macrophages as well as features of early- and late-phase allergic reactions following cutaneous injection in human atopic and nonatopic volunteers. J. Immunol. 169:2712– 18

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

170

Rothenberg

·

Hogan

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

115. Artis D, Humphreys NE, Potten CS, Wagner N, Muller W, et al. 2000. β7 integrindeﬁcient mice: delayed leukocyte recruitment and attenuated protective immunity in the small intestine during enteric helminthic infection. Eur. J. Immunol. 30:1656–64 116. Pretolani M, Rufﬁe C, Lapa e Silva JR, Joseph D, Lobb RR, Vargaftig BB. 1994. Antibody to very late activation antigen 4 prevents antigen-induced bronchial hyperreactivity and cellular inﬁltration in the guinea pig airways. J. Exp. Med. 180:795–805 117. Nakajima H, Sano H, Nishimura T, Yoshida S, Iwamoto I. 1994. Role of vascular cell adhesion molecule 1/very late activation antigen 4 and intercellular adhesion molecule 1/lymphocyte function-associated antigen 1 interactions in antigen-induced eosinophil and T cell recruitment into the tissue. J. Exp. Med. 179:1145–54 118. Gonzalo JA, Lloyd CM, Kremer L, Finger E, Martinez AC, et al. 1996. Eosinophil recruitment to the lung in a murine model of allergic inﬂammation. The role of T cells, chemokines, and adhesion receptors. J. Clin. Invest. 98:2332–45 119. Jia GQ, Gonzalo JA, Hidalgo A, Wagner D, Cybulsky M, Gutierrez-Ramos JC. 1999. Selective eosinophil transendothelial migration triggered by eotaxin via modulation of Mac-1/ICAM-1 and VLA-4/VCAM-1 interactions. Int. Immunol. 11:1–10 120. Tager AM, Dufour JH, Goodarzi K, Bercury SD, von Andrian UH, Luster AD. 2000. BLTR mediates leukotriene B4 -induced chemotaxis and adhesion and plays a dominant role in eosinophil accumulation in a murine model of peritonitis. J. Exp. Med. 192:439–46 121. Hirai H, Tanaka K, Yoshie O, Ogawa K, Kenmotsu K, et al. 2001. Prostaglandin D2 selectively induces chemotaxis in T helper type 2 cells, eosinophils, and basophils via seven-transmembrane receptor CRTH2. J. Exp. Med. 193:255–61 122. O’Reilly M, Alpert R, Jenkinson S, Gladue RP, Foo S, et al. 2002. Identiﬁcation of a histamine H4 receptor on human eosinophils—role in eosinophil chemotaxis. J. Recept. Signal Transduct. Res. 22:431–48 123. Behm CA, Ovington KS. 2000. The role of eosinophils in parasitic helminth infections: insights from genetically modiﬁed mice. Parasitol. Today 16:202–9 124. Korenaga M, Hitoshi Y, Yamaguchi N, Sato Y, Takatsu K, Tada I. 1991. The role of interleukin-5 in protective immunity to Strongyloides venezuelensis infection in mice. Immunology 72:502–7 125. Erickson LD, Foy TM, Waldschmidt TJ. 2001. Murine B1 B cells require IL-5 for optimal T cell-dependent activation. J. Immunol. 166:1531–39 126. Bischoff SC, Brunner T, De Weck AL, Dahinden CA. 1990. Interleukin 5 modiﬁes histamine release and leukotriene generation by human basophils in response to diverse agonists. J. Exp. Med. 172:1577–82 127. Hakonarson H, Maskeri N, Carter C, Chuang S, Grunstein MM. 1999. Autocrine interaction between IL-5 and IL-1β mediates altered responsiveness of atopic asthmatic sensitized airway smooth muscle. J. Clin. Invest. 104:657–67 128. Gurish MF, Humbles A, Tao H, Finkelstein S, Boyce JA, et al. 2002. CCR3 is required for tissue eosinophilia and larval cytotoxicity after infection with Trichinella spiralis. J. Immunol. 168:5730–36 129. Simons JE, Rothenberg ME, Lawrence RA. 2005. Eotaxin-1-regulated eosinophils have a critical role in innate immunity against experimental Brugia malayi infection. Eur. J. Immunol. 35:189–97 130. Adamko DJ, Yost BL, Gleich GJ, Fryer AD, Jacoby DB. 1999. Ovalbumin sensitization changes the inﬂammatory response to subsequent parainﬂuenza infection: Eosinophils mediate airway hyperresponsiveness, M2 muscarinic receptor dysfunction and antiviral effects. J. Exp. Med. 1999:1465–77 www.annualreviews.org • The Eosinophil

171

ARI

23 February 2006

15:12

131. Jacoby DB, Gleich GJ, Fryer AD. 1993. Human eosinophil major basic protein is an endogenous allosteric antagonist at the inhibitory muscarinic M2 receptor. J. Clin. Invest. 91:1314–18 132. Bandeira-Melo C, Bozza PT, Weller PF. 2002. The cellular biology of eosinophil eicosanoid formation and function. J. Allergy Clin. Immunol. 109:393–400 133. Hamelmann E, Gelfand EW. 2001. IL-5-induced airway eosinophilia—the key to asthma? Immunol. Rev. 179:182–91 134. Denzler KL, Farmer SC, Crosby JR, Borchers MT, Cieslewicz G, et al. 2000. Eosinophil major basic protein-1 does not contribute to allergen-induced airways pathologies in mouse models of asthma. J. Immunol. 165:5509–17 135. Boyce JA, Austen KF. 2005. No audible wheezing: nuggets and conundrums from mouse asthma models. J. Exp. Med. 201:1869–73 136. Hogan S, Koskinen A, Foster P. 1997. Interleukin-5 and eosinophils induce airway damage and bronchial hyperreactivity during allergic airway inﬂammation in BALB/c mice. Immunol. Cell Biol. 75:284–88 137. Foster P, Mould A, Yang M, Mackenzie J, Mattes J, et al. 2001. Elemental signals regulating eosinophil accumulation in the lung. Immunol. Rev. 179:173–81 138. Hogan SP, Koskinen A, Matthaei KI, Young IG, Foster PS. 1998. Interleukin-5-producing CD4+ T cells play a pivotal role in aeroallergen-induced eosinophilia, bronchial hyperreactivity, and lung damage in mice. Am. J. Respir. Crit. Care Med. 157:210–18 139. Mattes J, Yang M, Mahalingam S, Kuehr J, Webb DC, et al. 2002. Intrinsic defect in T cell production of interleukin (IL)-13 in the absence of both IL-5 and eotaxin precludes the development of eosinophilia and airways hyperreactivity in experimental asthma. J. Exp. Med. 195:1433–44 140. Corry D, Folkesson H, Warnock M, Erle D, Matthay M, et al. 1996. Interleukin 4, but not interleukin 5 or eosinophils, is required in a murine model of acute airway hyperreactivity. J. Exp. Med. 183:109–17 141. Hamelmann E, Oshiba A, Loader J, Larsen GL, Gleich G, et al. 1997. Antiinterleukin-5 antibody prevents airway hyperresponsiveness in a murine model of airway sensitization. Am. J. Respir. Crit. Care Med. 155:819–25 142. Lee NA, McGarry MP, Larson KA, Horton MA, Kristensen AB, Lee JJ. 1997. Expression of IL-5 in thymocytes/T cells leads to the development of a massive eosinophilia, extramedullary eosinophilopoiesis, and unique histopathologies. J. Immunol. 158:1332– 44 143. Brightling CE, Bradding P, Symon FA, Holgate ST, Wardlaw AJ, Pavord ID. 2002. Mastcell inﬁltration of airway smooth muscle in asthma. N. Engl. J. Med. 346:1699–705 144. Justice JP, Borchers MT, Crosby JR, Hines EM, Shen HH, et al. 2003. Ablation of eosinophils leads to a reduction of allergen-induced pulmonary pathology. Am. J. Physiol. Lung Cell Mol. Physiol. 284:L169–78 145. O’Byrne PM, Inman MD, Parameswaran K. 2001. The trials and tribulations of IL-5, eosinophils, and allergic asthma. J. Allergy Clin. Immunol. 108:503–8 146. Flood-Page P, Menzies-Gow A, Phipps S, Ying S, Wangoo A, et al. 2003. Anti-IL-5 treatment reduces deposition of ECM proteins in the bronchial subepithelial basement membrane of mild atopic asthmatics. J. Clin. Invest. 112:1029–36 147. Green RH, Brightling CE, McKenna S, Hargadon B, Parker D, et al. 2002. Asthma exacerbations and sputum eosinophil counts: a randomised controlled trial. Lancet 360:1715–21 148. Lee JJ, Dimina D, Macias MP, Ochkur SI, McGarry MP, et al. 2004. Deﬁning a link with asthma in mice congenitally deﬁcient in eosinophils. Science 305:1773–76

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

172

Rothenberg

·

Hogan

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

ARI

23 February 2006

15:12

149. Humbles AA, Lloyd CM, McMillan SJ, Friend DS, Xanthou G, et al. 2004. A critical role for eosinophils in allergic airways remodeling. Science 305:1776–79 150. Rothenberg ME. 2004. Eosinophilic gastrointestinal disorders (EGID). J. Allergy Clin. Immunol. 113:11–28 151. Mishra A, Rothenberg ME. 2003. Intratracheal IL-13 induces eosinophilic esophagitis by an IL-5, eotaxin-1, and STAT6-dependent mechanism. Gastroenterology 125:1419– 27 152. Mishra A, Hogan SP, Brandt EB, Rothenberg ME. 2001. An etiological role for aeroallergens and eosinophils in experimental esophagitis. J. Clin. Invest. 107:83–90 153. Forbes E, Smart VE, D’Aprile A, Henry P, Yang M, et al. 2004. T helper-2 immunity regulates bronchial hyperresponsiveness in eosinophil-associated gastrointestinal disease in mice. Gastroenterology 127:105–18 154. Cools J, DeAngelo DJ, Gotlib J, Stover EH, Legare RD, et al. 2003. A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome. N. Engl. J. Med. 348:1201–14 155. Cools J, Stover EH, Wlodarska I, Marynen P, Gilliland DG. 2004. The FIP1L1-PDGFRα kinase in hypereosinophilic syndrome and chronic eosinophilic leukemia. Curr. Opin. Hematol. 11:51–57 156. Gleich GJ, Leiferman KM, Pardanani A, Tefferi A, Butterﬁeld JH. 2002. Treatment of hypereosinophilic syndrome with imatinib mesilate. Lancet 359:1577–78 157. Stellato C, Matsukura S, Fal A, White J, Beck LA, et al. 1999. Differential regulation of epithelial-derived C-C chemokine expression by IL-4 and the glucocorticoid budesonide. J. Immunol. 163:5624–32 158. Schleimer RP, Bochner BS. 1994. The effect of glucocorticoids on human eosinophils. J. Allergy Clin. Immunol. 94:1202–13 159. Barnes PJ, Adcock IM. 1995. Steroid resistance in asthma. QJM 88:455–68 160. Aldebert D, Lamkhioued B, Desaint C, Gounni AS, Goldman M, et al. 1996. Eosinophils express a functional receptor for interferon α: inhibitory role of interferon α on the release of mediators. Blood 87:2354–60 161. Bankers-Fulbright JL, Kephart GM, Loegering DA, Bradford AL, Okada S, et al. 1998. Sulfonylureas inhibit cytokine-induced eosinophil survival and activation. J. Immunol. 160:5546–53 162. Hunt LW, Frigas E, Butterﬁeld JH, Kita H, Blomgren J, et al. 2004. Treatment of asthma with nebulized lidocaine: a randomized, placebo-controlled study. J. Allergy Clin. Immunol. 113:853–59 163. Kane GC, Pollice M, Kim CJ, Cohn J, Dworski RT, et al. 1996. A controlled trial of the effect of the 5-lipoxygenase inhibitor, zileuton, on lung inﬂammation produced by segmental antigen challenge in human beings. J. Allergy Clin. Immunol. 97:646– 54 164. Gaddy JN, Margolskee DJ, Bush RK, Williams VC, Busse WW. 1992. Bronchodilation with a potent and selective leukotriene D4 (LTD4) receptor antagonist (MK-571) in patients with asthma. Am. Rev. Respir. Dis. 146:358–63 165. Snyman JR, Sommers DK, Gregorowski MD, Boraine H. 1992. Effect of cetirizine, ketotifen and chlorpheniramine on the dynamics of the cutaneous hypersensitivity reaction: a comparative study. Eur. J. Clin. Pharmacol. 42:359–62 166. Redier H, Chanez P, De Vos C, Rifai N, Clauzel AM, et al. 1992. Inhibitory effect of cetirizine on the bronchial eosinophil recruitment induced by allergen inhalation challenge in allergic patients with asthma. J. Allergy Clin. Immunol. 90:215–24 www.annualreviews.org • The Eosinophil

173

ARI

23 February 2006

15:12

167. Rand TH, Lopez AF, Gamble JR, Vadas MA. 1988. Nedocromil sodium and cromolyn (sodium cromoglycate) selectively inhibit antibody-dependent granulocyte-mediated cytotoxicity. Int. Arch. Allergy Appl. Immunol. 87:151–58 168. Wegner CD, Gundel RH, Reilly P, Haynes N, Letts LG, Rothlein R. 1990. Intercellular adhesion molecule-1 (ICAM-1) in the pathogenesis of asthma. Science 247:456–59 169. Weg VB, Williams TJ, Lobb RR, Nourshargh S. 1993. A monoclonal antibody recognizing very late activation antigen-4 inhibits eosinophil accumulation in vivo. J. Exp. Med. 177:561–66 170. Kuijpers TW, Mul EP, Blom M, Kovach NL, Gaeta FC, et al. 1993. Freezing adhesion molecules in a state of high-avidity binding blocks eosinophil migration. J. Exp. Med. 178:279–84 171. von Andrian UH, Engelhardt B. 2003. α4 integrins as therapeutic targets in autoimmune disease. N. Engl. J. Med. 348:68–72 172. Mauser PJ, Pitman AM, Fernandez X, Foran SK, Adams GK 3rd, et al. 1995. Effects of an antibody to interleukin-5 in a monkey model of asthma. Am. J. Respir. Crit. Care Med. 152:467–72 173. Egan RW, Athwahl D, Chou CC, Emtage S, Jehn CH, et al. 1995. Inhibition of pulmonary eosinophilia and hyperreactivity by antibodies to interleukin-5. Int. Arch. Allergy Immunol. 107:321–22 174. Blanchard C, Mishra A, Saito-Akei H, Monk P, Anderson I, Rothenberg ME. 2005. Inhibition of human interleukin-13-induced respiratory and oesophageal inﬂammation by anti-human-interleukin-13 antibody (CAT-354). Clin. Exp. Allergy 35:1096–103 175. Nutku E, Aizawa H, Hudson SA, Bochner BS. 2003. Ligation of Siglec-8: a selective mechanism for induction of human eosinophil apoptosis. Blood 101:5014–20

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-05

174

Rothenberg

·

Hogan

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:147-174. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-06

ARI

23 February 2006

15:18

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Human T Cell Responses Against Melanoma Thierry Boon,1,2 Pierre G. Coulie,2 Benoˆıt J. Van den Eynde,1,2 and Pierre van der Bruggen1,2 1

Ludwig Institute for Cancer Research, Brussels Branch, B-1200 Brussels, Belgium; email: [email protected], [email protected], [email protected]

2

Cellular Genetics Unit, Universit´e de Louvain, B-1200 Brussels, Belgium; email: [email protected]

Annu. Rev. Immunol. 2006. 24:175–208 First published online as a Review in Advance on December 1, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090733 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0175$20.00

Key Words vaccination, cancer, T lymphocyte

Abstract Many antigens recognized by autologous T lymphocytes have been identiﬁed on human melanoma. Melanoma patients usually mount a spontaneous T cell response against their tumor. But at some point, the responder T cells become ineffective, probably because of a local immunosuppressive process occurring at the tumor sites. Therapeutic vaccination of metastatic melanoma patients with these antigens is followed by tumor regressions only in a small minority of the patients. The T cell responses to the vaccines show correlation with the tumor regressions. The local immunosuppression may be the cause of the lack of vaccination effectiveness that is observed in most patients. In patients who do respond to the vaccine, the antivaccine T cells probably succeed in reversing focally this immunosuppression and trigger a broad activation of other antitumor T cells, which proceed to destroy the tumor.

175

ANRV270-IY24-06

ARI

23 February 2006

15:18

INTRODUCTION

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MAGE: melanoma antigen-encoding gene

The large diversity of mouse tumor systems has predictably led to the elucidation of a wide variety of interesting mechanisms of tumor attack by the immune system, as well as of mechanisms of tumor resistance and escape (1). We suspect that for each type of human tumor, there is a subset of this wide array of processes that critically inﬂuences the tumor’s evolution and its response to immunotherapy. This crucial set of processes can only be identiﬁed by studying in depth each type of human tumor. With a few exceptions, we have restricted this review to observations made on human melanoma patients regarding antitumoral T cell responses that occur either spontaneously or following vaccination with tumor antigens. Even with this limitation, this brief review is far from exhaustive.

ANTIGENS RECOGNIZED ON MELANOMA BY AUTOLOGOUS T CELLS Genes Encoding Melanoma Antigens Human tumors bear antigens that can be recognized by autologous T lymphocytes. Some of these antigens appear to be tumor speciﬁc, whereas others are also present on normal tissues. Two main genetic mechanisms produce tumor-speciﬁc antigens: (a) Point mutations occurring in tumor cells can generate antigenic peptides either by enabling them to bind to the groove of MHC molecules or by generating new epitopes on these peptides. (b) In tumor cells, a gene is expressed that is silent in normal cells. On human melanoma, autologous T cells recognize tumor-speciﬁc antigens produced by both mechanisms: the antigens caused by mutational events and those encoded by cancer-germline genes. Somewhat surprisingly, in view of what is expected of natural tolerance, it is also possible to observe or generate autologous T cell responses against differentiation antigens common to melanoma and normal melanocytes. 176

Boon et al.

Antigens encoded by cancer-germline genes. A small number of gene families are expressed in male germline cells and in many tumors, but not in other normal adult tissues. Because male germline cells do not express human leukocyte antigen (HLA) molecules on their surface, they do not express antigens that can be recognized by T cells (2). As a result, the antigens encoded by these cancergermline genes are strictly tumor-speciﬁc T cell targets. These antigens are shared by many tumors. The prototype of the cancer-germline genes is the melanoma antigen-encoding (MAGE) gene family (3–5). This family is composed of 24 genes that range across three subfamilies: MAGE-A, -B, and -C, located in three different regions of the X chromosome. The expression of MAGE genes in tumors appears to be triggered by the demethylation of their promoter, apparently as a consequence of the widespread demethylation process that occurs in many tumors (6). Six additional MAGE gene families have been identiﬁed in the human genome (5, 7). These genes are expressed in normal tissues. However, because of their low similarity with the sequences of the MAGE-A, -B, and C family, not a single MAGE-A, -B, or -C antigenic peptide is encoded by one of the ubiquitously expressed MAGE genes, leaving intact the strict tumor speciﬁcity of the antigens encoded by the MAGE-A, -B, and -C genes. Other important cancer-germline gene families are the BAGE (8), GAGE (9), LAGE/NY-ESO-1 (10, 11), and SSX families (12, 13). Table 1 describes the main cancergermline genes and antigens that are discussed in this review. Investigators often state that antigens encoded by cancer-germline genes represent “self ” antigens and that tolerance must therefore be broken in the course of immunization against these antigens. In one sense, cancer-germline genes are indeed self in that they are normal genes of the human genome. However, the strict tumoral speciﬁcity of the

ANRV270-IY24-06

ARI

Table 1

23 February 2006

Main antigenic peptides used to vaccinate melanoma patientsa

Gene

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

15:18

Frequency of expression in metastatic melanoma

Presenting HLA

Peptide sequence

Position in the protein

MAGE-A1

46%

A2 A3

MAGE-A3

74%

A1 DP4

MAGE-A4

28%

MAGE-A10

47%

MAGE-A12

62%

Cw7

MAGE-C2

59%

A2

ALKDVEERV

NY-ESO-1/LAGE-2

28%

A2

SLLMWITQ(A,I,L,V)

Tyrosinase

>90%

A2

YMDGTMSQV

Melan-A/MART-1

>90%

A2

ELAGIGILTV

26–35 (A27L)

gp100

>90%

A2

IMDQVPFSV

209–217 (T210M)

50%

A2

VLPDVFIRC

(intron)

GNT-V a

KVLEYVIKV SLFRAVITK

278–286 96–104

EVDPIGHLY KKLLTQHFVQENYLEY

168–176 243–258

A2

GVYDGREHTV

230–239

A2

GLYDGMEHL

254–262

VRIGHLYIL

170–178 336–344 157–165 (C165A,I,L,V) 369–377

More information can be found at http://www.cancerimmunity.org/peptidedatabase/Tcellepitopes.htm.

antigens is hardly conveyed by the word self. The critical issue is whether there exists some measurable degree of natural immunological tolerance against these antigens. In our opinion this question has not yet been answered. One ought to be able to examine this issue with the mouse P1A gene, which codes for a mastocytoma P815 antigen and is also a cancer-germline gene (14). Normal DBA/2 mice mount a CD8 response to this antigen. Transgenic mice that express the P1A gene in all their tissues are clearly incapable of a cytotoxic T lymphocyte (CTL) response to the relevant antigens (15). But the crucial question has not yet been answered, namely, whether knockout mice will mount anti-P1A T cell responses more readily than normal mice. We expect these mice to be available in the near future. For gene P1A, as well as for MAGE genes, Kyewski and coworkers (16, 17) have shown that some sporadic gene expression occurs in medullary thymic epithelial cells, thereby opening the possibility that some form of thymic tolerance might exist. Antigens resulting from point mutations. Many antigens recognized by T cells on human tumors are encoded by a sequence con-

taining a point mutation that originated in the tumor cells (18). Many of the point mutations that have been identiﬁed because they generated a tumor antigen appear to have an oncogenic function. One mutation found in melanoma in a cyclin-dependent kinase molecule (CDK4) was also found in a familial form of melanoma (19, 20). The mutated antigens that have been identiﬁed after autologous mixed lymphocyte/tumor cell cultures (MLTC) were invariably found to result from mutations that are present on a single tumor or on very few. Thus, these mutated antigens are strictly tumor speciﬁc but not shared. This limits their use as potential vaccines. It is remarkable that MLTC have not produced T cells recognizing antigens encoded by mutated regions that are very frequent in cancer cells, such as mutated ras or p53. However, a mutated N-ras antigen was found to be recognized by tumor-inﬁltrating lymphocytes (TILs) (21). Promising candidates for shared mutated antigens are encoded by B-RAF, a member of the mitogen-activated protein (MAP) kinase cascade. The same activating mutation of B-RAF is found in more than 60% of

www.annualreviews.org • Human T Cell Responses Against Melanoma

TIL: tumor-inﬁltrating lymphocyte

177

ANRV270-IY24-06

ARI

23 February 2006

15:18

melanomas (22). Recently, CD4 T cells were shown to recognize a B-Raf peptide encoded by the mutated region. These CD4 T cells also recognized cells expressing the mutated B-RAF gene (23).

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Antigens encoded by genes overexpressed in tumors. The gene PRAME was identiﬁed as coding for a melanoma antigen recognized by autologous CTL. It is expressed at a high level in almost all melanomas and in many other tumors (24, 25). Survivin, an antiapoptotic protein of the inhibitor of apoptosis protein (IAP) family, was reported to be overexpressed in tumors, and melanoma cells are lysed by antisurvivin CTL (26, 27). However, no comparison of the level of survivin in tumor cells and in proliferating normal cells has been published. Telomerase has been suggested as a good target for immunotherapy, as its presence is essential for the proliferation of tumor cells. An antigenic peptide encoded by telomerase was recognized by CTL (28). However, other reports claimed that CTL of patients vaccinated with the antigenic peptide recognized cells pulsed with the peptide but did not recognize tumor cells expressing telomerase (29, 30). The antigens encoded by these genes are shared between many tumors, but they are not tumor speciﬁc, and the risk of vaccinating with these antigens must be weighed carefully.

Antigens encoded by differentiation genes. Melanoma patients have T lymphocytes capable of recognizing antigens encoded by normal melanocytic differentiation genes, such as those encoded by tyrosinase, Melan-A/Mart-1, gp100/pMel17, tyrosinase-related protein (TRP)-1, and TRP-2 (31–34). It was surprising to observe that natural tolerance against such antigens was not tighter. Vaccination with such antigens can produce vitiligo, i.e., elimination of normal melanocytes in some areas of the skin (35, 36). 178

Boon et al.

Processing of Tumor Antigens Antigens presented by HLA class I molecules. Most tumor antigens presented by class I molecules appear to be processed in tumor cells by the normal pathway of protein destruction by the proteasome, followed by the transport of the produced peptides into the endoplasmic reticulum by TAP (transporters associated with antigen presentation) proteins. The antigenic peptides then bind into the groove of MHC class I molecules and their N-terminal end can undergo a ﬁnal trimming by aminopeptidases. It is becoming increasingly clear that some antigenic peptides are processed much better in cells harboring immunoproteasomes, whereas others are processed better in cells with standard proteasomes (37–39). The endopeptidase activities of these proteasomes differ by their speciﬁcity, the immunoproteasome cleaving better after hydrophobic residues and the standard proteasome after acidic residues. This can affect not only the cleavage producing the proper C-terminus, but also the degree of destruction of the antigenic peptide by internal cleavage. Such proteasome preferences could have consequences for the CTL attack of tumor cells. In a noninﬂammatory environment, tumor cells are expected to express only standard proteasomes and, accordingly, only the tumor antigens produced by these proteasomes. However, upon immune attack of the tumor, IFN-γ ought to be produced, resulting in the replacement of standard proteasomes by immunoproteasomes, thereby eliminating those antigens that are produced by the former and replacing them by those produced by the latter. This could render the tumor cells resistant to some CTL. However, many tumor antigens are expressed on most tumors, and a number of them are processed by both types of proteasomes. Proteasome preferences also have implications for antitumoral vaccination. The mature dendritic cells (DCs), which present antigens to T lymphocytes in lymph nodes, express immunoproteasomes (37, 40,

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

41). This means that for some antigens that are processed only by standard proteasomes, it is preferable to vaccinate with the peptide or with the recombinant virus carrying a minigene coding for the peptide rather than with the protein or a recombinant virus coding for the whole protein. A substantial minority (∼20%) of the tumor antigens are the result of a deviation from the genetic expression process that produces the standard protein. The deviation can occur at the transcription stage, as is the case for the tumor-speciﬁc antigen that results from the action of a cryptic promoter located in an intron of gene GNT-V (42). This produces a protein encoded by an intronic sequence and out-of-frame exonic sequences. Such a cryptic promoter can be located on the reverse strand of an intron and produce an antisense transcript that also produces antigenic peptides (43). A second mechanism is the translation of incompletely spliced messages, resulting in antigenic peptides spanning an exon and an intronic region (44–47). The deviation can also occur at the translational stage, with the translation of small alternative open reading frames (48–54). Posttranslational modiﬁcation such as deamidation of asparagine can also occur (55). A recent development in the area of these atypical antigens was the startling observation that peptide sequences that are noncontiguous on a protein sequence can be joined together to form an antigenic peptide (56, 57). This peptide splicing occurs in the proteasome, where one free cleavage product forms a peptide bond with another one that is still bound to the catalytic threonine (57). This enables the splicing of peptide fragments located distantly on the protein either in the direct order or in the reverse order.

Antigens presented by HLA class II molecules. Constitutive expression of MHC class II molecules is normally restricted to DCs, macrophages, and B lymphocytes. But it also occurs frequently in several types of tumor, particularly melanoma (58). Moreover,

tumors usually express class II molecules when IFN-γ is present (59). Two different pathways lead to presentation of antigenic peptides on HLA class II molecules. Through the endocytic pathway, the presenting cells can internalize external cell debris or proteins, as well as their own secreted and membrane proteins (60–62). Through the endogenous pathway, intracellular proteins reach the endosomal compartment, either after passage through the endoplasmic reticulum for proteins having appropriate signal and targeting sequences, or by direct inclusion of cytosolic elements in endosomes (autophagy) (63–65). The melanocyte protein tyrosinase follows the second pathway, being directed through its targeting sequence to melanosomes, organelles that are derived from the endosomes (66, 67). Protein MAGE-A3 does not contain a signal sequence. Accordingly, CD4 T cell clones recognizing two MAGE-A3 peptides presented by HLA-DR13 molecules could not recognize DR13 melanoma cells expressing MAGE-A3 (68). Nevertheless, three other MAGE-A3 antigenic peptides were found to be recognized by CD4 T cell clones on HLADP4, DR1, or DR11 tumor cells expressing MAGE-A3 (69–71). This presumably occurs through autophagy. An optimal anticancer vaccine may require the cooperation of CD4+ Th cells and CD8+ CTLs. Various strategies have been explored to obtain presentation of both HLA class I– and class II–restricted peptides derived from the same vaccine candidate. One interesting possibility for cytosolic proteins is the manipulation of the HLA class II presentation pathway. This manipulation can be achieved by using vectors where the protein-coding sequence is linked to a sequence encoding the targeting motif of an endosomal or lysosomal protein, such as the invariant chain, the lysosome-associated membrane protein-1 (LAMP-1), or DC-LAMP (72–78). Interestingly, targeting MAGE-A3 to the HLA class II–processing pathway induced concurrent presentation of class I– and class II–restricted

www.annualreviews.org • Human T Cell Responses Against Melanoma

LAMP: lysosome-associated membrane protein

179

ANRV270-IY24-06

ARI

23 February 2006

DriPs: defective ribosomal products

15:18

peptides to CD8+ and CD4+ T cells, respectively. We observed that the HLA class I presentation by the DCs electroporated with the chimeric constructs was more efﬁcient than that by peptide-pulsed DCs. The reason for this is not clear. It could be due to the translation of an abnormal protein, inducing an increased formation of defective ribosomal products (DriPs) that are addressed to the proteasome (79).

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Optimization of Peptide Immunogenicity For immunization with antigenic peptides, an important line of research has been the substitution of some amino acids to obtain peptides with stronger binding to HLA molecules. This has been applied ﬁrst to the gp100209−217 peptide, ITDQVPFSV, which is presented by HLA-A2. The threonine in position 2 was replaced by methionine. The modiﬁed peptide was much more immunogenic in HLA-A2 transgenic mice (80, 81). Among T lymphocytes collected from patients vaccinated with the modiﬁed peptide, some recognized both the modiﬁed and the normal peptide, whereas others recognized only the modiﬁed peptide (82, 83). This approach was also applied to MelanA.A2 peptide EAAGIGILTV, where the alanine in position 2 was replaced by a leucine residue (84, 85). This increased the ability of the peptide to trigger CTL lysis and to stimulate PBL in vitro (86). T lymphocytes from melanoma patients were isolated with HLA tetramers containing the modiﬁed peptide. Some of these T lymphocytes recognized targets presenting the normal antigen, but others recognized only targets pulsed with the modiﬁed peptide (87). However, almost all T cells collected from patients vaccinated with the modiﬁed peptide recognized melanoma cell lines (88). For the NY-ESO-1 peptide SLLMWITQC, the C-terminal residue was replaced by valine, leucine, or isoleucine. Here, again, these peptides showed better 180

Boon et al.

binding to HLA and were more readily recognized by CTL (89, 90). In conclusion, the usefulness of vaccination with modiﬁed peptides showing increased immunogenicity has to be weighed carefully against the risk of having a signiﬁcant subset of responder T cells that fail to recognize the tumor cells.

SPONTANEOUS T CELL RESPONSES Do spontaneous antitumoral T cell responses exist? Are they beneﬁcial to some patients? To the extent that they exist, why do they fail to reject the tumors in many instances? Do they inﬂuence the state of the tumor and its sensitivity to immunotherapy by selecting for resistance?

Existence of Spontaneous T Cell Responses Evidence that strongly suggested the existence of a spontaneous antitumoral T cell response in melanoma patients predated the identiﬁcation of human tumor antigens. Rosenberg and associates (91) carried out a considerable amount of work with TILs. Single cell suspensions derived from metastases, including invaded lymph nodes that contained lymphocytes and an undeﬁned number of tumor cells, were cultured from 30–60 days in the presence of IL-2. This procedure often produced as many as 1010 –1011 T lymphocytes. Many of these lymphocyte preparations were found to exert cytolytic activity on fresh tumor cell preparations, albeit under conditions in which stringent speciﬁcity controls could not be realized (92). Adoptive transfer of these lymphocytes together with high-dose IL-2 treatment produced objective tumor responses in a higher proportion of the patients than that observed with IL-2 treatment alone (91, 93). Our early work on the identiﬁcation of antigens recognized on melanoma by autologous T cells provided unquestionable proof of the antitumoral speciﬁcity of the T cells

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

found in the blood of melanoma patients (4). However, these patients had received various immunotherapy regimens, and no frequency comparison was made between normal and cancerous individuals. So, even though in retrospect we believe that, with the techniques available at that time, we would not have obtained antitumor T cell clones unless a response had occurred in these patients, our work did not clearly demonstrate the existence of such a spontaneous response. At that time, we also carried out a systematic evaluation of the antitumoral T cell response of melanoma patients, using autologous MLTC under limiting dilution conditions (94). Microcultures were considered to contain antitumor effectors when they exerted a lytic activity that was much higher on autologous tumor cells than on K562 or autologous T cell blasts. Frequencies of antitumor T cells ranging from 3 × 10−5 to 10−3 of CD8 T cells were observed in a group of six patients. The ability to conﬁrm the existence of a spontaneous antitumoral T cell response was hampered by the impossibility of providing a baseline with parallel tests run with autologous cell lines on noncancerous patients, and by the lack of knowledge of the target antigens. Parmiani and collaborators (95) compared frequencies of antitumor T cells in the blood and in metastatic sites (invaded lymph nodes and subcutaneous tumors). Frequencies ranging from 2 × 10−3 to 10−4 of CD8+ T cells were observed in the blood. In subcutaneous tumors the frequencies observed were about tenfold higher (95). They also examined melanoma metastases and found signiﬁcant differences in the frequencies of some T cell receptor (TCR)-V genes in metastases relative to the blood, suggesting that some T cells were enriched in the tumors (96). Analysis of TCR sequences indicated that some TCRs were present at high frequency. More recently, a limiting dilution analysis of tumor-speciﬁc CTLp in the blood was carried out on six melanoma patients before

vaccination with various antigens. All patients had high frequencies of antitumor CTLp ranging from 6 × 10−5 to 2 × 10−3 of the CD8 T cells (97). The antitumoral speciﬁcity of these responses now appears ﬁrmly established because in one patient the response was analyzed at a clonal level, and almost all the clones were found to be directed against antigens encoded either by the cancer-germline gene MAGE-C2 or by the melanocyte differentiation gene gp100. Letsch et al. (98) found in an ex vivo ELISPOT test that 4/5 melanoma patients had a large number of T cells that produced IFN-γ upon stimulation with an autologous cell line, with frequencies ranging from 5 × 10−3 to 10−4 of CD8 T cells.

TCR: T cell receptor CTLp: cytotoxic T lymphocyte precursor

Spontaneous responses against tumor antigens encoded by cancer-germline genes. The background level of T cell precursors against MAGE-encoded antigens in normal donors appears to be situated ∼4 × 10−7 of CD8+ T cells (see below) (99). Spontaneous (i.e., prevaccination) ampliﬁcation of anti-MAGE-3.A1 CD8 T cells was observed in a very small proportion of metastatic melanoma patients: 2/45 (100, 101). Similar observations were made for other tumorspeciﬁc antigens such as GNT-V.A2. But some other antigens of this type appear to elicit spontaneous responses more frequently. J¨ager et al. (102, 103) demonstrated that in 10/27 patients with tumors expressing NY-ESO-1, and only in those, a spontaneous CTL response against NY-ESO-1 occurred, which was strongly correlated with the occurrence of an antibody response against the NYESO-1 protein. Valmori et al. (104) showed that TILs contained anti-NY-ESO-1 CTL and found that a patient had in the blood a frequency of 4 × 10−4 of CD8+ T cells that were directed against a NY-ESO-1 peptide. Some melanoma patients have spontaneous responses against an out-of-frame antigenic peptide encoded by LAGE-1 (105). Hercend and colleagues (106–108) observed an oligoclonal ampliﬁcation of T cells

www.annualreviews.org • Human T Cell Responses Against Melanoma

181

ANRV270-IY24-06

ARI

23 February 2006

15:18

in a spontaneously regressing melanoma lesion. Interestingly, the most frequent T cell clonotype present in the tumor was not the dominant clone in the TIL culture. Nevertheless, it proved possible to obtain a permanent T cell clone with the dominant TIL clonotype, and the clone recognized an antigen encoded by gene MAGE-A6 (106–108). A detailed analysis of the blood of one melanoma patient before vaccination indicated the presence of ﬁve ampliﬁed CTL clonotypes recognizing peptides encoded by gene MAGE-C2 (97). The frequencies of these CTL ranged from 2 × 10−6 to 4 × 10−5 of the blood CD8+ cells. These CTL were found to be highly enriched in a skin metastasis, where they reached frequencies of 0.1% (109). No response was observed in this patient against antigens encoded by other cancer-germline genes, even though these were expressed by the tumor at a similar level as MAGE-C2. A spontaneous anti-MAGE-C2 response has been found in another patient, suggesting that, like NY-ESO-1, MAGE-C2 is a preferred target for spontaneous T cell responses. These responses against MAGE-C2 antigens target several different peptides presented by several different HLA molecules. Wang et al. (110) reported that a LAGE1 antigen was the target of tumor-speciﬁc CD4+ regulatory T cell (Treg) clones derived from the TILs of a melanoma patient. Phenotypic and functional analyses demonstrated that the CD4+ Tregs expressed CD25 and GITR (glucocorticoid-induced TNF receptor family–related) molecules and exerted a suppressive activity on the proliferative response of naive CD4+ T cells to anti-CD3 antibody treatment (110).

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Treg: regulatory T cell

Spontaneous responses against mutated antigens. In the study mentioned above by Hercend et al. (108), one of the frequent T cell clonotypes present in the tumor was found to recognize a mutated antigen encoded by a myosin gene (111). A mutated tyrosine phosphatase was the target of a CD4 T cell clone isolated from the TILs of a melanoma patient 182

Boon et al.

(112). Sensi et al. (113) found that about 0.4% of the CD8+ T cells present in a lymph node invaded by a melanoma were speciﬁc for a unique antigen generated by a point mutation in the peroxiredoxin 5 gene. Spontaneous responses against differentiation antigens. TILs of melanoma patients were shown to contain not only CD8 T cells but also CD4 T cells recognizing melanocyte differentiation antigens (114, 115). A very interesting and exceptional situation is that of the Melan-A/Mart-126−35 antigenic peptide, which is recognized by T cells on HLA-A2. This antigenic peptide is located in the putative transmembrane region of the 118 amino acid Melan-A protein. Only in this region were Melan-A antigenic peptides identiﬁed. High frequencies (∼0.1%) of anti-Melan-A.A2 CD8 T cells are found in the blood of most melanoma patients. Still higher frequencies, of the order of several percent, are found among CD8 T cells in invaded lymph nodes or cutaneous metastases (116). But healthy individuals also harbor in their blood a frequency of anti-Melan-A.A2 of about 0.005%, a frequency that is two orders of magnitude higher than that observed for naive precursors against MAGE antigens. Romero and colleagues (116) showed that these lymphocytes, which have a naive phenotype of CCR7+ CD45RA+ , appear to have multiplied little since their exit from the thymus as judged from their content of TCR excision circles and their telomere length (117). The same group identiﬁed a large number of cross-reactive peptides in a protein data search, suggesting that presentation of these antigenic peptides in the thymus could account for the extraordinary high frequency of anti-Melan-A T cells (87). Considering the frequency in healthy individuals, can one conclude that a spontaneous Melan-A response occurs in melanoma patients? The answer appears to be yes in many patients, who show a frequency of CTL precursors that is 20-fold higher than in noncancerous individuals. Moreover, one observes, particularly in

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

advanced patients, that a signiﬁcant fraction of the anti-Melan-A T cells have a memory effector phenotype of CCR7+ CD45RO+ , whereas in healthy individuals all these T cells have a naive phenotype (118). The situation is very different for the other melanocyte differentiation antigens, which appear to stimulate spontaneous CTL responses much less often. Peter Lee (119) screened 11 melanoma patients for the presence of T cells recognizing the tyrosinase369−377 peptide. Two were positive, with frequencies of 0.2 and 2% of the CD8+ cells. Frequencies around 0.2% were also observed in 2/6 patients analyzed by another group (120). We screened a dozen melanoma patients and found no spontaneous response against this tyrosinase peptide or against the gp100209−217 peptide (P.G. Coulie, unpublished data).

Benefit to the Patient Several authors have examined the correlation between intratumoral T cell inﬁltrates and clinical evolution. T cell inﬁltration in primary melanoma was reported to correlate with a better eight-year survival rate (121, 122). The inﬁltration by T lymphocytes of tumor cell areas in an invaded lymph node was also found to correspond with a better prognosis (123). However, in advanced melanoma patients (stage IV), the presence of antiMelan-A- or antityrosinase-speciﬁc T cells in the blood was not correlated with better survival (124). Similar observations were made in prostatic adenocarcinomas (125). In colorectal cancer, a study showed that invasion of CD8+ T cells inside cancer cell nests signiﬁcantly correlated with a favorable prognosis, whereas inﬁltrates at the tumor margin or in the stroma did not (126). The same conclusion was reached for esophageal carcinomas (127). In a study on epithelial ovarian cancer collected at the time of surgical debulking prior to the chemotherapy, Zhang et al. (128) reported a strong correlation between the presence of an intratu-

moral T cell inﬁltrate and both disease-free and overall survival. Interestingly, the level of RNA for cytokines, such as monokine induced by IFN-γ, macrophage-derived chemokine, and secondary lymphoid-tissue chemokine, was markedly higher in those inﬁltrated tumors that were from patients with a better progression-free survival. An interesting observation, which did not involve a truly spontaneous T cell response, was made in chronic myelogenous leukemia. IFN-α2b treatment produced about 60% overall responses. In 11 out of 12 patients who responded, CTL directed against a peptide encoded by proteinase 3 appeared following the treatment. None of the seven patients who did not respond to the treatment showed these CTL. Proteinase 3 is expressed in the azurophil granules of the normal myeloid cells and is overexpressed twoto ﬁvefold in some leukemia cells. These observations suggest that the effect of IFN-α2b may be mediated by a T cell response (129). We tend to conclude that spontaneous T cell responses correlate with a better prognosis, even though the correlation may be obscured in some circumstances by the fact that this T cell response must be initiated at one point in the evolution of the tumor and therefore must be more frequently observed in more advanced tumors. From this correlation we cannot, however, rigorously conclude that the T cell inﬁltration is the cause of a better evolution of the tumor.

Tumor Resistance and T Cell Anergy On the basis of the observations described above, one may conclude that most metastatic melanoma patients are progressing despite having mounted a quantitatively signiﬁcant T cell response against their tumor, resulting in a CTL precursor frequency of 10−4 to 10−3 of CD8 T cells in the blood. The lack of effectiveness of the antitumoral T cells does not seem to be due to their failure to reach the tumor sites. They are present there in numbers that are not very different from those observed in metastases that are regressing following

www.annualreviews.org • Human T Cell Responses Against Melanoma

183

ARI

23 February 2006

15:18

immunotherapy, and they are considerably enriched relative to the blood (109). Considering that antitumor T cells are clearly capable of rallying to the tumor site, the spontaneous antitumoral T cell response must become ineffective at one point, either because the tumor cells have become insensitive to the effector cells or because the effector cells themselves have become unable to be stimulated by the antigen or to exert their function, a state we refer to as anergy. This anergy could result from inhibitory processes elicited by tumor cells that were selected by the immune response. A mechanism of tumor escape is the selection of tumor cell variants that no longer present the antigens that are the targets of the effector T cells. The existence of such tumor cell variants has been very well documented (130). The loss of antigen can be due to a defect in the presentation system. General loss of expression of HLA class I molecules, often due to mutation in the β-2 microglobulin gene or to loss of expression of one HLA haplotype or of a single HLA allele, has also been observed on many samples of tumors of different histological types (131). HLA loss or downregulation were observed in a signiﬁcant fraction of melanoma primary and metastatic samples (132). A clear example of partial HLA loss followed by evolution of the T cell response to target antigens presented by other HLA molecules has been reported recently (133). Antigen loss can also result from a mutation or deletion in the gene coding for the antigenic peptide. This has been well documented in mouse systems but to our knowledge not often in human tumors. One apparent exception is the switching off of the genes coding for melanoma differentiation antigens. However, we have observed that such genes can be readily switched on and off in cultured cells, and it is therefore difﬁcult to assign their arrest to immunoselection. Is antigen loss a major mechanism of tumor escape from spontaneous T cell response, and is it a major source of ineffec-

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

184

Boon et al.

tiveness of immunotherapy? In our opinion, it is not presently the major cause of failure of antitumoral vaccination strategies in melanoma. When we isolate tumor cell lines from metastatic samples either before or after a vaccine therapy, we ﬁnd in more than 90% of instances that these cells are capable of restimulating in vitro autologous T lymphocytes and are sensitive to speciﬁc lysis by the resulting CTL. This does not exclude, however, that if at one point antitumoral vaccination becomes effective in causing complete tumor regressions in a large fraction of the patients, relapses will be observed that result from antigen loss due to immunoselection. If one admits that the spontaneous antitumoral T cell response becomes blocked, a ﬁrst issue is whether there is systemic anergy or whether the anergy is a local phenomenon occurring at the tumor site. Are antitumoral T cells present in the blood less readily reactivatable in vivo than memory T cells directed against, e.g., viral antigens? Do they show signs of impaired lytic functionality? This issue can be analyzed by capturing with tetramers blood T cells that have the appropriate speciﬁcity and then studying their properties. Contradictory observations have been reported. Lee et al. (119) reported that antityrosinase CD8+ T cells in the blood of a melanoma patient were not readily reactivatable in vitro. In contrast, Zippelius et al. (134) found that CCR7− CD8+ T cells directed against the MelanA.A2 antigen and present in the blood expressed IFN-γ upon antigenic stimulations and contained perforin and granzyme. This ﬁnding was in sharp contrast to ﬁndings with TILs collected from tumor sites: Anti-MelanA.A2 CCR7− CD8+ T cells did not synthesize IFN-γ upon short-term restimulation with antigen, whereas anti-cytomegalovirus T cells present in metastases could be stimulated (134). There is additional evidence that antitumoral T cells present in metastases are anergic. Recently, Bronte et al. (135) showed that T cells present in prostatic carcinoma

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

locations did not upregulate CD25, CD69, or CD137 upon in situ phytohemagglutinin- or phorbol myristate acetate–ionomycin stimulation. In contrast, T cells present in normal prostate could be stimulated. Interestingly, T cells inside prostatic carcinomas appeared to lack polarization of their cytotoxic granules, which ought to have prevented their lytic ability (135). This phenomenon had been previously described in mouse tumors (136). To conclude, we feel that there is credible evidence for local T cell anergy in tumor locations, whereas more evidence should be obtained to establish if there exists some degree of systemic anergy. What is the cause of local anergy? There seem to be two main possibilities. Either the tumor cells, or the stroma inﬂuenced by the tumor cells, anergize the potential responder T cells directly, or they generate suppressive T cells, now called regulatory T cells (Tregs), that in turn anergize the responder T cells. During the past few years, the concept of Tregs has been revisited, largely owing to the work of Sakaguchi (137, 138). In humans, CD4+ CD25+ T cells present in the blood can exert an inhibitory effect in vitro on the CD4+ CD25− T cell population and T cell clones (139–141). This ﬁnding is leading to a reexploration of the concept that antitumoral immune responses might be blocked by such Tregs. T cell clones that were derived from TILs and directed against a LAGE and a mutated antigen were found to have the functional proﬁle of Treg cells (110, 142). Another systemic suppressive mechanism has been proposed: Observations in renal carcinoma patients showed that a subset of human polymorphonuclear granulocytes produced arginase, resulting in a low arginine level in the blood and depressed T cell functions (143). There is no lack of candidate agents that tumor cells could use to directly or indirectly anergize responder T cells. In mouse systems, TGF-β could play a major role in inhibiting antitumoral immune responses, but, to our knowledge, ﬁrm evidence that it plays a

determining role in the evolution of human tumors is not yet available (144). Many human tumor cells express IDO (indoleamine 2,3-dioxygenase), an enzyme also expressed in placenta, which degrades tryptophan (145). Cells expressing IDO serve as a tryptophan sink that deprives surrounding T lymphocytes of tryptophan and impairs their function. Bronte et al. (135) observed that inhibitors of arginase and nitric oxide synthase restored the functionality of TILs. Human melanoma cells secrete galectin-1, which inhibits T cells (146), and also produce IL-10, which could inhibit production of proinﬂammatory cytokines such as TNF-α, IFN-γ, and IL-2 (147). Human melanoma cells, particularly in the presence of IFN-γ, express the B7-H1 surface molecule (148). This molecule, which binds to the PD-1 receptor on T cells, causes increased apoptosis.

IDO: indoleamine 2,3-dioxygenase

T CELL RESPONSES AGAINST VACCINE ANTIGENS Quantitative Evaluation of T Cell Responses The efﬁcacy of the T cell response to a tumor antigen vaccine ought to depend on the number of activated T lymphocytes and on their functional properties. Even though it appears very plausible that T cells with highafﬁnity receptors might be more effective and might even be required for an antitumoral effect (149), there is no evidence regarding the set of functional properties that are required for antivaccine T lymphocytes to be effective. This leaves us with only a quantitative evaluation of the T cell responses to vaccines. The evaluation of the frequency of vaccine-induced CD8+ T cells can be performed ex vivo, that is, with T cells that are tested immediately upon their collection from the blood. Alternatively, it can be performed after an in vitro restimulation period of several days, which results in the proliferation of the antivaccine T cells. The lymphocytes are usually drawn from the blood, but they can also be collected from tumor sites.

www.annualreviews.org • Human T Cell Responses Against Melanoma

185

ARI

23 February 2006

15:18

CD8 responses. Ex vivo frequency assays allow for a direct evaluation of the frequency of antivaccine T cells. Three assays are commonly used. The tetramer assay measures the fraction of cells expressing TCRs speciﬁc for a given antigen, regardless of the functional characteristics of these T cells. The lymphocytes can then be counted and sorted while staying alive (150). Tetramers are soluble complexes of recombinant MHC molecules folded in the presence of an antigenic peptide and biotinylated. When these complexes are bound to ﬂuorescinated avidin, which can bind four biotin moieties, the name tetramer is appropriate. But when the ﬂuorescent reagent is phycoerythrin, which spontaneously polymerizes, it brings several avidin molecules together so that the reagent is really a multimer. Nevertheless, we use the word tetramer throughout. With most HLA class I tetramers, the lower detection limit of the test is ∼3 × 10−4 of the CD8+ cells. A second assay that is frequently used is the ELISPOT assay for IFN-γ secretion. Antibodies are used to detect the cytokine produced by individual cells after a short (6–48 h) period of restimulation with the antigen. The test has been adapted to the study of the response of melanoma patients by Wolfel and colleagues (151), who greatly simpliﬁed and standardized the reading procedure by using computer-assisted image analysis (152). When the assay is performed with unfractionated peripheral blood mononuclear cells (PBMC), the sensitivity of the assay is about 5 × 10−4 of the CD8+ T cells. A third assay also evaluates the number of cells producing IFN-γ upon contact with the antigen. During the last 4–18 h of stimulation, brefeldin is added to block the export of the cytokine, which then accumulates in the cell. The lymphocytes are ﬁxed, permeabilized, and stained with antibody. The fraction of positive cells is evaluated by ﬂow cytometry. The sensitivity of this assay is similar to that of ELISPOT. As with ELISPOT, the lymphocytes are lost during the assay, preventing any further assessment of their speciﬁcity.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

186

Boon et al.

Because the frequency of antivaccine responder T cells is often lower than the detection limit of the assays just mentioned, it is usually necessary to submit the collected lymphocytes to an in vitro restimulation with the vaccine antigen during 5–14 days. During this restimulation, the precursor lymphocytes (CTLp) that recognize the antigen can multiply by several hundredfold. This ampliﬁcation can be followed, as for the ex vivo procedure, by a tetramer, ELISPOT, or intracellular cytokine assay. After in vitro restimulation, it is also possible to use a lysis assay on a target cell presenting the antigen. For a proper evaluation of the frequency of antivaccine T cell precursors in the original sample, it is essential to conduct the restimulation process under limiting dilution conditions, for all that can be deduced from the positive scoring of a restimulated culture is that it contained at least one relevant T cell precursor to start with. We have used a limiting dilution restimulation followed by a tetramer assay and the cloning of the positive microcultures. A lytic assay is then performed with each T cell clone, not only on peptide-pulsed cells but also on target cells expressing the appropriate antigen-encoding gene and HLA molecule. This approach is long and labor intensive, but it provides a stringent control of the speciﬁcity of the putative antivaccine T cells as well as the possibility of assessing their TCR sequence. This in turn provides the possibility of setting up clonotypic polymerase chain reaction (PCR) assays to reevaluate directly precursor frequencies in the blood and to compare these with the estimates obtained by the restimulation-tetramer procedure. The comparison has shown that, with a few exceptions, the results were remarkably close, leading to the conclusion that at least half of the antivaccine T cell precursors multiply in vitro under the restimulation conditions. However, if a precursor were totally refractory to our in vitro restimulation conditions, we would fail to notice its existence. We also observed that T cell precursors belonging to the same clonotype can proliferate at vastly different

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

rates in independent microcultures, conﬁrming the necessity of the limiting dilution approach for a frequency evaluation. We believe that one can detect very lowfrequency antivaccine T cell responses that are barely above the naive precursor frequency by evaluating the presence of repeated clonotypes. As stated below, we ﬁnd that in a noncancerous individual, the frequency of precursors of T cells recognizing a MAGEA3 antigen presented by HLA-A1 is 4 × 10−7 CD8+ T cells, with a TCR repertoire of >100 clonotypes (100). On this basis, when sequence analysis indicates that at least 3/10 or 4/20 T cell clones derived from independent microcultures have the same TCR, we conclude that an ampliﬁed clonotype is present, which represents a response to the vaccine. CD4 responses. Tetramers of class II HLA molecules have proven much more arduous to produce than their class I equivalents, presumably because of the intrinsic structural instability of class II molecules in solution. A ﬁrst use of these tetramers was the evaluation of an anti-inﬂuenza response with a DR4 tetramer (153). We produced a DP4 tetramer that can be used for ex vivo staining of PBMC. The positive cells are sorted by ﬂow cytometry. The sorted cells are then cloned by limiting dilution. After multiplication, the speciﬁcity of the T cell clones is assessed by their ability to secrete IFN-γ upon contact with EBVtransformed B cells pulsed with the peptide or EBV-B cells transduced with a retrovirus carrying the protein coding sequence fused to a truncated invariant chain. This approach has a high sensitivity: Frequencies as low as 10−6 of CD4+ T cells have been ascertained (154). Ex vivo ELISPOT approaches involving release of IFN-γ, IL-2, and IL-4 following stimulation with the relevant peptide have also been used (155). Sensitivity of 3 × 10−4 of CD4+ T cells can be reached. The IFN-γELISPOT test can also be used after in vitro restimulation with a peptide, but here again,

a limiting dilution approach is required for a frequency evaluation. It is also important to analyze CD4+ T cell responses in patients vaccinated with protein under conditions in which the target antigenic peptides are unknown. One approach is a brief stimulation with DCs pulsed with protein, followed by an IFN-γ secretion assay involving the immediate recapture of the cytokine secreted by one lymphocyte with bispeciﬁc antibodies coating the cell. The cells are then stained with a ﬂuorescent anti-IFN-γ antibody and sorted by ﬂow cytometry. Cloning and functional analysis are performed as described above (156).

The Antivaccine Responses Responses against antigens encoded by cancer-germline genes. Early investigations regarding the CTL responses of patients vaccinated with MAGE antigens indicated that these responses were not of a high level (157). To evaluate weak responses, it appeared necessary ﬁrst to obtain a precise description of the anti-MAGE T cell precursors in noncancerous individuals. The precursor T cell frequency against the MAGE-A3168−176 peptide presented by HLA-A1 (MAGE-3.A1) was analyzed in detail (99, 100). The naive precursor frequency in the blood was estimated at 4 × 10−7 of CD8+ T cells and was found to be remarkably similar in different individuals. Independent T cell clones of one individual were tested for their TCRβ chain sequences, and only one repeat was found among 23 clones. The statistical analysis of this result led to the conclusion that the diversity of the naive repertoire against the MAGE-3.A1 antigen is very likely to exceed 100. Considering that the total number of CD8+ T cells in humans is about 4 × 1010 , one can therefore reasonably represent the naive CTLp as comprising about 100 clones of 160 cells each (Figure 1). Naive CTLp frequencies for another MAGE-A3 antigen were ﬁrst evaluated to be ∼10−6 of CD8 T cells (99), but later studies suggested somewhat lower

www.annualreviews.org • Human T Cell Responses Against Melanoma

187

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

Figure 1 Frequencies and diversity of anti-MAGE-3.A1 CTLs. Different clonotypes are represented by circled numbers.

frequencies of ∼5 × 10−7 . We have also observed frequencies of ∼5 × 10−7 for several other antigens, such as MAGE-A1278−286 , tyrosinase.A2, GNT-V.A2, MAGE-A1254−262 , and MAGE-C2336−344 , in prevaccination samples of melanoma patients (158; P.G. Coulie, unpublished data). Vaccines including antigen MAGE-3.A1 have been applied in various modalities to metastatic melanoma patients with detectable disease. In such patients, spontaneous T cell responses against the MAGE-3.A1 antigen are rare, as we observed only two such cases out of 45 (100, 101). Following vaccination with the peptide without adjuvant, seven patients showing evidence of tumor regression were analyzed, and a CTL response was observed in only two patients, with frequencies of 5 × 10−6 and 4 × 10−5 , respectively (159, 160). Other patients were vaccinated with a recombinant ALVAC (canarypox) virus containing the MAGE-3.A1 coding sequence. Among four patients showing evidence of tumor regression, anti-MAGE-3.A1 CTL responses were observed in three, with frequen188

Boon et al.

cies ranging from 3 × 10−3 to 7 × 10−7 , the latter being considered as a response because of the presence of a repeated clonotype (100, 161). For all the patients vaccinated with peptide or ALVAC, TCR analysis showed that the CTL response was monoclonal. Another constant feature of the anti-MAGE-3.A1 T cell responses is their stability: Once detected, the T cells appear to remain in the same frequency range, even several months following the last revaccination. Metastatic melanoma patients were also immunized with monocyte-derived mature DCs pulsed with the MAGE-3.A1 peptide (162). Analyses of three regressor patients showed a CTL response in all three, with frequencies ranging from 3 × 10−6 to 10−3 (163). In contrast with the results obtained with peptide or ALVAC, the responses of these three patients were polyclonal. No response could be detected in ﬁve vaccinated patients who did not show tumor regression. In another study involving vaccination with CD34+ progenitor-derived DCs pulsed with peptide MAGE-A3271−279 , ﬁve patients out of

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

eight appeared to have a T cell response (164). On the basis of these results, one may conclude that when peptide-pulsed DCs stimulate a response against antigens encoded by cancer-germline genes, the responses are somewhat stronger than those elicited by peptide alone or by ALVAC, but there are still many patients who do not show a T cell response. Another possible method for increasing the immunogenicity of cancer-germlineencoded antigens is to combine them with immunological adjuvants. In a study in which peptide MAGE-A3271−279 was injected with IL-12 into four patients, ELISPOT assays after restimulation (which was not performed under limiting dilution conditions) suggested a response, as prevaccination tests were negative (165). No response was observed in two patients who received the peptide without IL-12. Using the same detection approach, responses were reported for peptides MAGE-A1243−258 and MAGE-A10254−262 injected with incomplete Freund’s adjuvant (IFA) and GM-CSF (166). In another study, no CTL response was observed in seven melanoma patients vaccinated with peptide MAGE-A12170−178 with IFA (167). Considering the strong T cell responses obtained with the Melan-A/MART-1 peptide combined with IFA and CpG (see below), it will be interesting to test this combination of adjuvants with MAGE peptides. Finally, the possibility of providing help to the CD8 response by vaccinating with antigens recognized by CD4+ T cells has been explored. We injected four patients with peptide MAGE-3.A1 together with peptide MAGEA3243−258 , presented by HLA-DP4, and observed no response (P. van der Bruggen, unpublished data). A recombinant MAGE-A3 protein mixed with the adjuvant AS02B was administered to 24 patients. All patients except one mounted an IgG response to the protein, indicating a stimulation of CD4+ T cells (168). When the protein was used without adjuvant, only 1/24 patients produced an anti-MAGE-A3 IgG response, conﬁrm-

ing the importance of the adjuvant for the in vivo stimulation of CD4+ T cells (169). It would be interesting to analyze the adjuvant effect of this protein-AS02B combination on CD8 responses against coadministered class I MAGE peptides. A few experiments were performed to document the CD4 responses to the MAGE-A3 protein. In a patient vaccinated with the protein alone, anti-MAGE-A3 CD4+ T cells were detected at a frequency of 1.5 × 10−5 of blood CD4+ cells, representing at least an 80-fold increase over the prevaccination frequency (156). Most of these CD4+ cells recognized the same peptide, presented by HLA-DR1. Class II MAGE peptides have also been combined with DCs. HLA-DP4restricted CD4+ T cells recognizing peptide MAGE-A3243−258 were found in 7 × 10−4 of CD4+ blood T cells after vaccination with peptide-pulsed DCs (154, 155). The evaluation of the responses of patients vaccinated with the NY-ESO-1.A2 peptide is complicated by the fact that, at the time of vaccination, about half the patients with NY-ESO-1+ tumors had already made a spontaneous T cell response against this antigen. Vaccination of patients with preexisting responses did not seem to lead to a clear increase in the CTL frequency. Vaccination of the other patients clearly produced T cell responses (102). A recent report suggests that melanoma patients with no detectable disease who were vaccinated with NY-ESO-1 protein and ISCOMATRIX, a saponin-based adjuvant, had CD8 and CD4 responses against a wide range of NY-ESO-1-encoded peptides (170).

IFA: incomplete Freund’s adjuvant

Vaccination with melanocyte differentiation antigens. Several groups reported the induction of CD8 T cell responses after vaccination of melanoma patients with the Melan-A/MART-126−35(27L) peptide administered with IFA (171–173). Speiser et al. (174) compared responses after vaccination with the peptide in IFA with or without CpG oligodeoxynucleotides. With CpG, more patients (8/8 versus 4/8) had a detectable

www.annualreviews.org • Human T Cell Responses Against Melanoma

189

ARI

23 February 2006

15:18

increase in the frequency of anti-Melan-A CD8 T cells, and the percentages of antiMelan-A T cells among the CD8 were tenfold higher, reaching 1% in several patients. This is higher than the frequency observed after any of the following vaccinations: DC pulsed with a set of peptides including Melan-A26−35 (175); a recombinant vaccinia virus encoding antigenic peptides together with B7.1 and B7.2 molecules (176); a recombinant modiﬁed Ankara vaccinia virus encoding seven antigenic peptides (177); peptide and IL-12 (178); or peptides with IFA and an anti-CTLA-4 monoclonal antibody (179). Peptide gp100209−217(210M) has also been used in several vaccination trials. In one trial, the peptide was injected in IFA, and patients also received IL-2. Signiﬁcant T cell responses were observed (180). In another study in which a higher dose of peptide was combined with IFA, ex vivo tetramer analysis also indicated that almost all patients made a strong T cell response, with an average of 0.3% of CD8, whereas prevaccination frequencies were 0.02% (181). The peptide was also injected into patients with IL-12. Strong T cell responses with frequencies ranging from 0.05% to 2.5% of CD8 T cells were observed. In all patients, anti-gp100 T cells were undetectable in prevaccination samples (182). Slingluff and coworkers (183, 184) performed interesting studies in which the T cell responses observed in lymph nodes were compared with those observed in the blood following vaccination with a number of peptides, including gp100209−217 . T cell responses were observed more frequently in lymph nodes draining the vaccination site than in blood. In a study by Schaed et al. (185), ex vivo ELISPOT assays were used to compare postvaccination frequencies of T cells against peptide tyrosinase369−377 administered with either IFA, the puriﬁed saponin QS-21, or GM-CSF. None of the nine patients immunized with IFA developed a detectable response, whereas 3/9 and 2/8 did so after vaccination with QS-

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

190

Boon et al.

21 and GM-CSF, respectively. Immunization with DCs pulsed with various differentiation peptides produced T cell responses that appeared to be similar to those obtained with the modalities described above (155, 164, 186). Thus, it seems that the highest responses obtained against differentiation antigens are higher than the highest responses obtained against antigens encoded by cancer-germline genes.

RELATIONSHIP BETWEEN T CELL RESPONSES AND TUMOR REGRESSIONS Vaccination of metastatic melanoma patients with deﬁned antigens has up to now produced only modest results in terms of beneﬁt to the patients. Only a small minority of patients (5%–10%) has shown an objective clinical response. We have vaccinated with MAGE-A3 antigens metastatic melanoma patients with detectable disease using various modalities ranging from the MAGE-3.A1 antigenic peptide alone, to peptide with adjuvant, protein with adjuvant, and recombinant ALVAC. Of a total of 111 patients included in these various studies, 8 have shown objective clinical responses (159, 168, 169, 187). But some evidence of tumor regression, usually mixed responses, has been observed in 12 additional patients. We accept that these regressions are not clinically signiﬁcant, but we feel that they must be taken into account for the elucidation of the process by which the vaccines are occasionally effective. The results obtained by many groups using other antigens encoded by cancer-germline genes (MAGE, NY-ESO-1), or differentiation antigens (tyrosinase, Melan-A/Mart-1, gp100), or even allogenic melanoma lysates appear to produce tumor responses in the same frequency range (165, 178, 184, 188, 189). Rosenberg and coworkers (190) reported an even lower average objective clinical response frequency of 2.6% out of 444 treated patients. Clinical trials with DCs pulsed with peptide do not appear to produce markedly

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

better tumor response rates (162, 186, 191– 194). This appraisal should, however, be received with some caution, as the small number of patients included in most studies precludes concluding either that one vaccination modality is superior to others or that it is not. Melanoma tumors occasionally regress spontaneously. But in metastatic melanoma these regressions appear to occur rarely, at a frequency situated between 0.1% and 1% (195). Infrequent as they are, the objective clinical responses and the other tumor regressions that occur following vaccination are above this frequency of spontaneous regression. Moreover, they occur in the weeks or months following onset of vaccination. Effects can also be observed in an adjuvant setting: Keilholz and coworkers (196) observed that among nine patients who had presented more than three new metastases in the year before vaccination, two patients showed no relapse for more than two years (196). We tentatively conclude that the regressions observed following vaccination are produced by the vaccine. But it is difﬁcult to exclude the possibility that patients included in therapeutic vaccination trials represent a biased favorable subset. For instance, there may be a selection of patients with slowly progressing tumors. The low proportion of objective clinical responses does not appear to be due to the fact that 90%–95% of melanoma tumors are absolutely resistant to T cell attack. Rosenberg et al. (197) have shown that following adoptive transfer of a massive number of antitumor CTL (1011 ), objective clinical responses are observed in about 50% of the patients. Before the adoptive transfer, these patients received a treatment with cyclophosphamide aimed at eliminating the bulk of Tregs. It is also noteworthy that B. Dr´eno and coworkers (198) have shown rigorously in a randomized trial that adoptive transfer of TILs in an adjuvant setting was signiﬁcantly extending the duration of relapse-free survival of patients with only one invaded lymph node.

Is there a correlation between the occurrence of a detectable antivaccine T cell response and tumor regression? For patients vaccinated with ALVAC-MAGE-3.A1, we observed a deﬁnite correlation, as 3/4 of patients with evidence of tumor regression showed an antivaccine T cell response, in contrast to 1/17 patients without tumor regression (100, 187). In a small group of patients vaccinated with DCs, all patients with regressions showed a T cell response (162, 163). In contrast, for patients vaccinated with peptide alone, we see antivaccine T cell responses only in a minority of the patients showing a regression. If we consider all the patients having received various modalities of MAGE-3.A1 antigen, we obtain a total of 10/20 for responder patients and 1/30 for nonresponders. There is therefore a clear correlation between T cell responses and tumor regressions. Attempts have been made to improve correlations by including various parameters in addition to antivaccine T cell frequencies in the blood. Figdor et al. (193) observed in patients vaccinated with mature DC pulsed with gp100 and tyrosinase peptides that all the patients produced delayed-type hypersensitivity (DTH) following challenge with those pulsed DC. Very interestingly, only in the patients showing evidence of tumor regression were anti-gp100 or antityrosinase CTL precursors present at the DTH site (193). Banchereau et al. (186) used a compound score involving a statistical analysis of the results of direct and recall assays of all the vaccine antigens. Speiser et al. (199) found that tumor regression following vaccination with the Melan-A.A2 peptide in IFA correlated well with a compound score involving antivaccine T cell frequency increase in the blood and increase in CD45RA− or CD28− anti-Melan-A T cells. In many MAGE-vaccinated patients who show tumor regression, the frequency of antivaccine T cells is very low. One wonders how such a low number of T cells could provide the main component of the speciﬁc effectors destroying the tumor cells. This led us to

www.annualreviews.org • Human T Cell Responses Against Melanoma

DTH: delayed-type hypersensitivity

191

ARI

23 February 2006

15:18

evaluate in some vaccinated patients the total frequency of T cells that recognized the autologous tumor, hereafter labeled antitumor T cells. In addition, we analyzed not only the T cells present in the blood but also those present at tumor sites. A detailed analysis was carried out on one patient, EB81, who showed complete regression of all of a large number of skin metastases, but had a progressing lymph node removed in the course of vaccination. A tumor cell line was obtained from the lymph node and it was used to stimulate CD8 T cells. We observed that the vaccine had elicited a frequency of about 3 × 10−6 anti-MAGE-3.A1 T cells, and that after vaccination antitumor T cells (deﬁned by their ability to lyse autologous tumor cells and not EBV-B cells or NK target K562) were present at a frequency of 10−3 . Moreover these antitumor T cells were already present at a similar high frequency before vaccination. Five other patients were analyzed, and all presented the same high frequency of antitumor T cells before as well as after vaccination (97). The antitumor T cells of patient EB81 were cloned, and the antigens of 13/15 clones could be identiﬁed. Three were directed against gp100 antigens and ten were directed against various antigenic peptides, all encoded by gene MAGE-C2. Gene MAGE-C2 is a cancer-germline gene belonging to a different MAGE family than MAGE-A3, the source of the vaccine antigen. Thus, the spontaneous antitumor T cell response of this patient was directed against the type of melanoma antigens that are used in the vaccines. PCR assays speciﬁc for TCR sequences made it possible to establish the frequency of antivaccine and antitumor T cell clones in a number of metastases. Whereas antivaccine T cells showed a modest enrichment in tumor sites relative to the blood, some antitumor T cell clones showed an enrichment by several hundredfold, reaching frequencies in the tumor of several percent of CD8 T cells. It appears that at least 50% of the T cells present at tumor sites of patient EB81 belong to a small number of antitumor T cell clones. The same antitu-

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

192

Boon et al.

mor T cells were found at high frequency in different metastases (109). These results have recently been conﬁrmed with another vaccinated patient. On the basis of these observations, we propose the following scenario for the occasional tumor regressions that are observed following vaccination (Figure 2). The melanoma metastases contain highly enriched antitumor T cells before vaccination. These T cells have become ineffective, presumably because of immunosuppressive factors produced by the tumor. In some vaccinated patients, antivaccine T cells migrate to the tumor and are capable of resisting the immunosuppressive environment long enough to attack some tumor cells and reverse focally the immunosuppression. This provides conditions that enable a much larger number of antitumor T cells to be stimulated and to proliferate. These antitumor T cells then serve as speciﬁc effectors to eliminate the bulk of the tumor cells. Thus, the antivaccine T cells only serve as a spark that triggers the process leading to the activation of the antitumor T cells. These activated antitumor T cells could either be derived from the blocked T cells that were already present in the tumor or result from a new wave of activation of naive precursors. Our analysis of patient EB81 indicated that regressing tumors contained T cell clones that were already present before vaccination, but that the most enriched T cell clones were not. It is therefore possible that a new wave of antitumor T cells constitutes an important component of the effectors of the tumor rejection. It is noteworthy that most of the antigens recognized by the new wave of antitumor T cells were already recognized by the prevaccination set of antitumor T cells. If correct, this hypothetical scenario leads to a reconsideration of the correlation between tumor regression and antivaccine T cell response. Insofar as the regression process involves a new stimulation and expansion process of antitumor lymphocytes, this process can also apply to the antivaccine T cells that have migrated to the tumor. Accordingly, the

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

Figure 2 Hypothetical scenario for tumor regression following vaccination. T lymphocytes are represented by circles, with different colors indicating different antigenic speciﬁcities. Inactivation of these T cells is represented by faded out colors.

frequency of antivaccine T cells that is measured following vaccination would be the result of a two-stage ampliﬁcation: a ﬁrst one stimulated by the vaccine and a second one by the tumor. Therefore, the observation of a higher level of antivaccine T cells in a regressing patient would not necessarily be due to fact that in this patient the vaccine produced a higher response and therefore his tumor regressed. The antivaccine response may well not have been higher in this patient, but somehow it was capable of initiating a regression process that in turn resulted in the restimulation that produced a higher level of antivaccine T cells. In conclusion, therapeutic success following vaccination may not depend on the number of T cells produced directly by the vaccine, but rather on the production of a T cell clone with functional properties that enable it to migrate to the tumor and resist the

local immunosuppressive environment long enough to initiate a regression process. The view that the main effectors of tumor regression are antitumor cells distinct from the antivaccine T cells has other implications. One relates to the issue of tumor escape by antigen-loss variants. If antivaccine T cells were the only effectors of rejection, the loss of expression of the vaccine antigen by some tumor cells would ensure escape of the tumor. However, if many other antitumor effectors come into play, a general loss of antigen expression is required for escape. Another implication relates to the collaboration between CD4+ T helper cells and CD8+ effector T cells. A vast body of evidence supports the importance of this collaboration, which seems particularly important to the production of memory T cells (200). Accordingly, it has been claimed that effective

www.annualreviews.org • Human T Cell Responses Against Melanoma

193

ARI

23 February 2006

15:18

vaccination of cancer patients requires the combined use of antigens presented by MHC class II and class I molecules. However, in our experience vaccines containing peptide MAGE-A3243−258 , which is presented by HLA-DP4, in addition to peptide MAGE3.A1, did not prove to be clearly superior to the vaccine containing only MAGE-3.A1. Other groups have included helper peptides or proteins, such as keyhole limpet hemocyanin, tetanus toxoid, or the “PADRE” peptide, in their vaccines (164, 166, 193, 201). In our view, it is impossible to conclude unequivocally that this improved the clinical responses. One explanation may be that if a regression process can be triggered by the antivaccine CD8, then the antitumor T cells that come into play comprise several CD4+ as well as CD8+ T cell clones, thus providing the required helper effect. To achieve therapeutic success, investigators will probably need to understand the cause of the local immunosuppression in the tumors and ﬁnd counteracting agents. As stated above, the list of possible immunosuppressive agents present in tumors is considerable. But it will be important to ﬁnd whether, for each type of tumor, there is a prevalent immunosuppressive agent. Just as many types of tumors have preferred oncogenic pathways that differ from one type of tumor to another, each type of tumor may also have preferred immunosuppressive processes that we must identify to achieve therapeutic success. A combination of vaccination with agents that reduce this immunosuppressive process may considerably increase the proportion of patients showing clinical responses following immunization. Recent immunotherapy trials have included cyclophosphamide pretreatment aimed at eliminating the bulk of the Tregs (197). Vaccination com-

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

bined with anti-CTLA-4 antibodies has been applied to a small number of patients. In one study, 3/14 patients showed a clinical response, but many patients have suffered from severe autoimmune side effects (202). In another study, the tumor response rate did not seem to be very different from that obtained with other modalities (179). Therapeutic vaccination of cancer has not yet proved to be effective enough to become a generally applied cancer treatment. But considering its remarkably low toxicity, even in patients who have shown clinical response, it would be regrettable to abandon the effort at this stage without trying to understand the mechanisms of tumor resistance and ﬁnd the ways to overcome them. We can envision two extreme situations. In the ﬁrst, all patients would have tumors with a similar degree of tumor resistance. The difference between the regressors and the others would be due to a more effective response to the vaccine. We do not believe that melanoma patients suffer from a degree of general immunosuppression, which we believe is restricted to very late-stage patients who are not included in most studies (203). Therefore, the difference in the quality of the response would be due to a chance event determining, for instance, the functional properties of the unique or the few responder T cell clones elicited by the vaccine. In that case, it will be essential to understand what this crucial functional property is. At the other extreme, the antivaccine T cell responses would be similar in all patients, but the level of resistance of the tumors would vary considerably. In that case, investigators would need to identify the main component of this resistance and ﬁnd ways to counteract it. Analysis of gene expression in tumors using microarrays may be an important way to achieve this goal (204).

DISCLOSURE STATEMENT The Ludwig Institute owns several patents relative to tumor antigens discovered by the authors. The authors are entitled to a share of the royalties received by the Ludwig Institute for the licensing of the patents. 194

Boon et al.

ANRV270-IY24-06

ARI

23 February 2006

15:18

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LITERATURE CITED 1. Pardoll D. 2003. Does the immune system see tumors as foreign or self? Annu. Rev. Immunol. 21:807–39 2. Haas GG Jr, D’Cruz OJ, De Bault LE. 1988. Distribution of human leukocyte antigenABC and -D/DR antigens in the unﬁxed human testis. Am. J. Reprod. Immunol. Microbiol. 18:47–51 3. De Plaen E, Arden K, Traversari C, Gaforio JJ, Szikora JP, et al. 1994. Structure, chromosomal localization, and expression of 12 genes of the MAGE family. Immunogenetics 40:360–69 4. van der Bruggen P, Traversari C, Chomez P, Lurquin C, De Plaen E, et al. 1991. A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma. Science 254:1643–47 5. Chomez P, De Backer O, Bertrand M, De Plaen E, Boon T, Lucas S. 2001. An overview of the MAGE gene family with the identiﬁcation of all human members of the family. Cancer Res. 61:5544–51 6. De Smet C, De Backer O, Faraoni I, Lurquin C, Brasseur F, Boon T. 1996. The activation of human gene MAGE-1 in tumor cells is correlated with genome-wide demethylation. Proc. Natl. Acad. Sci. USA 93:7149–53 7. Lucas S, Brasseur F, Boon T. 1999. A new MAGE gene with ubiquitous expression does not code for known MAGE antigens recognized by T cells. Cancer Res. 59:4100–3 8. Ruault M, van der Bruggen P, Brun ME, Boyle S, Roizes G, De Sario A. 2002. New BAGE (B melanoma antigen) genes mapping to the juxtacentromeric regions of human chromosomes 13 and 21 have a cancer/testis expression proﬁle. Eur. J. Hum. Genet. 10:833– 40 9. De Backer O, Arden KC, Boretti M, Vantomme V, De Smet C, et al. 1999. Characterization of the GAGE genes that are expressed in various human cancers and in normal testis. Cancer Res. 59:3157–65 10. Lethe B, Lucas S, Michaux L, De Smet C, Godelaine D, et al. 1998. LAGE-1, a new gene with tumor speciﬁcity. Int. J. Cancer 76:903–8 11. Chen YT, Scanlan MJ, Sahin U, Tureci O, Gure AO, et al. 1997. A testicular antigen aberrantly expressed in human cancers detected by autologous antibody screening. Proc. Natl. Acad. Sci. USA 94:1914–18 12. Gure AO, Tureci O, Sahin U, Tsang S, Scanlan MJ, et al. 1997. SSX: a multigene family with several members transcribed in normal testis and human cancer. Int. J. Cancer 72:965– 71 13. Gure AO, Wei IJ, Old LJ, Chen YT. 2002. The SSX gene family: characterization of 9 complete genes. Int. J. Cancer 101:448–53 14. Van den Eynde B, Lethe B, Van Pel A, De Plaen E, Boon T. 1991. The gene coding for a major tumor rejection antigen of tumor P815 is identical to the normal gene of syngeneic DBA/2 mice. J. Exp. Med. 173:1373–84 15. Brandle D, Bilsborough J, Rulicke T, Uyttenhove C, Boon T, Van den Eynde BJ. 1998. The shared tumor-speciﬁc antigen encoded by mouse gene P1A is a target not only for cytolytic T lymphocytes but also for tumor rejection. Eur. J. Immunol. 28:4010– 19 16. Gotter J, Brors B, Hergenhahn M, Kyewski B. 2004. Medullary epithelial cells of the human thymus express a highly diverse selection of tissue-speciﬁc genes colocalized in chromosomal clusters. J. Exp. Med. 199:155–66 www.annualreviews.org • Human T Cell Responses Against Melanoma

195

ARI

23 February 2006

15:18

17. Derbinski J, Schulte A, Kyewski B, Klein L. 2001. Promiscuous gene expression in medullary thymic epithelial cells mirrors the peripheral self. Nat. Immunol. 2:1032–39 18. Van den Eynde B, van der Bruggen P. 2005. Peptide database of T-cell deﬁned tumor antigens. Cancer Immun. http://www.cancerimmunity.org/peptidedatabase/Tcellepitopes.htm 19. Wolfel T, Hauer M, Schneider J, Serrano M, Wolfel C, et al. 1995. A p16INK4ainsensitive CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma. Science 269:1281–84 20. Zuo L, Weger J, Yang Q, Goldstein AM, Tucker MA, et al. 1996. Germline mutations in the p16INK4a binding domain of CDK4 in familial melanoma. Nat. Genet. 12:97–99 21. Linard B, Bezieau S, Benlalam H, Labarriere N, Guilloux Y, et al. 2002. A ras-mutated peptide targeted by CTL inﬁltrating a human melanoma lesion. J. Immunol. 168:4802–8 22. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, et al. 2002. Mutations of the BRAF gene in human cancer. Nature 417:949–54 23. Sharkey MS, Lizee G, Gonzales MI, Patel S, Topalian SL. 2004. CD4+ T-cell recognition of mutated B-RAF in melanoma patients harboring the V599E mutation. Cancer Res. 64:1595–99 24. Ikeda H, Lethe B, Lehmann F, van Baren N, Baurain JF, et al. 1997. Characterization of an antigen that is recognized on a melanoma showing partial HLA loss by CTL expressing an NK inhibitory receptor. Immunity 6:199–208 25. Kessler JH, Beekman NJ, Bres-Vloemans SA, Verdijk P, van Veelen PA, et al. 2001. Efﬁcient identiﬁcation of novel HLA-A(∗ )0201-presented cytotoxic T lymphocyte epitopes in the widely expressed tumor antigen PRAME by proteasome-mediated digestion analysis. J. Exp. Med. 193:73–88 26. Schmitz M, Diestelkoetter P, Weigle B, Schmachtenberg F, Stevanovic S, et al. 2000. Generation of survivin-speciﬁc CD8+ T effector cells by dendritic cells pulsed with protein or selected peptides. Cancer Res. 60:4845–49 27. Schmidt SM, Schag K, Muller MR, Weck MM, Appel S, et al. 2003. Survivin is a shared tumor-associated antigen expressed in a broad variety of malignancies and recognized by speciﬁc cytotoxic T cells. Blood 102:571–76 28. Vonderheide RH, Hahn WC, Schultze JL, Nadler LM. 1999. The telomerase catalytic subunit is a widely expressed tumor-associated antigen recognized by cytotoxic T lymphocytes. Immunity 10:673–79 29. Parkhurst MR, Riley JP, Igarashi T, Li Y, Robbins PF, Rosenberg SA. 2004. Immunization of patients with the hTERT:540–548 peptide induces peptide-reactive T lymphocytes that do not recognize tumors endogenously expressing telomerase. Clin. Cancer Res. 10:4688– 98 30. Ayyoub M, Migliaccio M, Guillaume P, Lienard D, Cerottini JC, et al. 2001. Lack of tumor recognition by hTERT peptide 540–548-speciﬁc CD8+ T cells from melanoma patients reveals inefﬁcient antigen processing. Eur. J. Immunol. 31:2642–51 31. Brichard V, Van Pel A, Wolfel T, Wolfel C, De Plaen E, et al. 1993. The tyrosinase gene codes for an antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanomas. J. Exp. Med. 178:489–95 32. Coulie PG, Brichard V, Van Pel A, Wolfel T, Schneider J, et al. 1994. A new gene coding for a differentiation antigen recognized by autologous cytolytic T lymphocytes on HLAA2 melanomas. J. Exp. Med. 180:35–42 33. Bakker AB, Schreurs MW, de Boer AJ, Kawakami Y, Rosenberg SA, et al. 1994. Melanocyte lineage-speciﬁc antigen gp100 is recognized by melanoma-derived tumorinﬁltrating lymphocytes. J. Exp. Med. 179:1005–9

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

196

Boon et al.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

34. Kawakami Y, Eliyahu S, Sakaguchi K, Robbins PF, Rivoltini L, et al. 1994. Identiﬁcation of the immunodominant peptides of the MART-1 human melanoma antigen recognized by the majority of HLA-A2-restricted tumor inﬁltrating lymphocytes. J. Exp. Med. 180:347– 52 35. Yee C, Thompson JA, Roche P, Byrd DR, Lee PP, et al. 2000. Melanocyte destruction after antigen-speciﬁc immunotherapy of melanoma: direct evidence of T cell-mediated vitiligo. J. Exp. Med. 192:1637–44 36. Kawakami Y, Robbins PF, Wang RF, Parkhurst M, Kang X, Rosenberg SA. 1998. The use of melanosomal proteins in the immunotherapy of melanoma. J. Immunother. 21:237–46 37. Morel S, Levy F, Burlet-Schiltz O, Brasseur F, Probst-Kepper M, et al. 2000. Processing of some antigens by the standard proteasome but not by the immunoproteasome results in poor presentation by dendritic cells. Immunity 12:107–17 38. Schultz ES, Chapiro J, Lurquin C, Claverol S, Burlet-Schiltz O, et al. 2002. The production of a new MAGE-3 peptide presented to cytolytic T lymphocytes by HLA-B40 requires the immunoproteasome. J. Exp. Med. 195:391–99 39. Van den Eynde BJ, Morel S. 2001. Differential processing of class-I-restricted epitopes by the standard proteasome and the immunoproteasome. Curr. Opin. Immunol. 13:147–53 40. Macagno A, Gilliet M, Sallusto F, Lanzavecchia A, Nestle FO, Groettrup M. 1999. Dendritic cells up-regulate immunoproteasomes and the proteasome regulator PA28 during maturation. Eur. J. Immunol. 29:4037–42 41. Macagno A, Kuehn L, de Giuli R, Groettrup M. 2001. Pronounced up-regulation of the PA28α/β proteasome regulator but little increase in the steady-state content of immunoproteasome during dendritic cell maturation. Eur. J. Immunol. 31:3271–80 42. Guilloux Y, Lucas S, Brichard VG, Van Pel A, Viret C, et al. 1996. A peptide recognized by human cytolytic T lymphocytes on HLA-A2 melanomas is encoded by an intron sequence of the N-acetylglucosaminyltransferase V gene. J. Exp. Med. 183:1173–83 43. Van den Eynde BJ, Gaugler B, Probst-Kepper M, Michaux L, Devuyst O, et al. 1999. A new antigen recognized by cytolytic T lymphocytes on a human kidney tumor results from reverse strand transcription. J. Exp. Med. 190:1793–800 44. Coulie PG, Lehmann F, Lethe B, Herman J, Lurquin C, et al. 1995. A mutated intron sequence codes for an antigenic peptide recognized by cytolytic T lymphocytes on a human melanoma. Proc. Natl. Acad. Sci. USA 92:7976–80 45. Lupetti R, Pisarra P, Verrecchia A, Farina C, Nicolini G, et al. 1998. Translation of a retained intron in tyrosinase-related protein (TRP) 2 mRNA generates a new cytotoxic T lymphocyte (CTL)-deﬁned and shared human melanoma antigen not expressed in normal cells of the melanocytic lineage. J. Exp. Med. 188:1005–16 46. Robbins PF, El-Gamil M, Li YF, Fitzgerald EB, Kawakami Y, Rosenberg SA. 1997. The intronic region of an incompletely spliced gp100 gene transcript encodes an epitope recognized by melanoma-reactive tumor-inﬁltrating lymphocytes. J. Immunol. 159: 303–8 47. Harada M, Li YF, El-Gamil M, Ohnmacht GA, Rosenberg SA, Robbins PF. 2001. Melanoma-Reactive CD8+ T cells recognize a novel tumor antigen expressed in a wide variety of tumor types. J. Immunother. 24:323–33 48. Wang RF, Appella E, Kawakami Y, Kang X, Rosenberg SA. 1996. Identiﬁcation of TRP-2 as a human tumor antigen recognized by cytotoxic T lymphocytes. J. Exp. Med. 184:2207– 16 49. Ronsin C, Chung-Scott V, Poullion I, Aknouche N, Gaudin C, Triebel F. 1999. A nonAUG-deﬁned alternative open reading frame of the intestinal carboxyl esterase mRNA www.annualreviews.org • Human T Cell Responses Against Melanoma

197

ANRV270-IY24-06

ARI

23 February 2006

50.

51.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

52.

53.

54.

55.

56. 57. 58. 59.

60.

61.

62.

63.

64.

198

15:18

generates an epitope recognized by renal cell carcinoma-reactive tumor-inﬁltrating lymphocytes in situ. J. Immunol. 163:483–90 Probst-Kepper M, Stroobant V, Kridel R, Gaugler B, Landry C, et al. 2001. An alternative open reading frame of the human macrophage colony-stimulating factor gene is independently translated and codes for an antigenic peptide of 14 amino acids recognized by tumor-inﬁltrating CD8 T lymphocytes. J. Exp. Med. 193:1189–98 Rosenberg SA, Tong-On P, Li Y, Riley JP, El-Gamil M, et al. 2002. Identiﬁcation of BING-4 cancer antigen translated from an alternative open reading frame of a gene in the extended MHC class II region using lymphocytes from a patient with a durable complete regression following immunotherapy. J. Immunol. 168:2402–7 Aarnoudse CA, van den Doel PB, Heemskerk B, Schrier PI. 1999. Interleukin-2-induced, melanoma-speciﬁc T cells recognize CAMEL, an unexpected translation product of LAGE-1. Int. J. Cancer 82:442–48 Wang RF, Johnston SL, Zeng G, Topalian SL, Schwartzentruber DJ, Rosenberg SA. 1998. A breast and melanoma-shared tumor antigen: T cell responses to antigenic peptides translated from different open reading frames. J. Immunol. 161:3598–606 Slager EH, van der Minne CE, Goudsmit J, van Oers JM, Kostense S, et al. 2004. Induction of CAMEL/NY-ESO-ORF2-speciﬁc CD8+ T cells upon stimulation with dendritic cells infected with a modiﬁed Ad5 vector expressing a chimeric Ad5/35 ﬁber. Cancer Gene Ther. 11:227–36 Skipper JC, Hendrickson RC, Gulden PH, Brichard V, Van Pel A, et al. 1996. An HLA-A2restricted tyrosinase antigen on melanoma cells results from posttranslational modiﬁcation and suggests a novel pathway for processing of membrane proteins. J. Exp. Med. 183:527– 34 Hanada K, Yewdell JW, Yang JC. 2004. Immune recognition of a human renal cancer antigen through post-translational protein splicing. Nature 427:252–56 Vigneron N, Stroobant V, Chapiro J, Ooms A, Degiovanni G, et al. 2004. An antigenic peptide produced by peptide splicing in the proteasome. Science 304:587–90 Ruiter DJ, Mattijssen V, Broecker EB, Ferrone S. 1991. MHC antigens in human melanomas. Semin. Cancer Biol. 2:35–45 Shaw AR, Chan JK, Reid S, Seehafer J. 1985. HLA-DR synthesis induction and expression in HLA-DR-negative carcinoma cell lines of diverse origins by interferon-γ but not by interferon-β. J. Natl. Cancer Inst. 74:1261–68 Chen BP, Madrigal A, Parham P. 1990. Cytotoxic T cell recognition of an endogenous class I HLA peptide presented by a class II HLA molecule. J. Exp. Med. 172:779– 88 Chicz RM, Urban RG, Gorga JC, Vignali DA, Lane WS, Strominger JL. 1993. Speciﬁcity and promiscuity among naturally processed peptides bound to HLA-DR alleles. J. Exp. Med. 178:27–47 Weiss S, Bogen B. 1989. B-lymphoma cells process and present their endogenous immunoglobulin to major histocompatibility complex-restricted T cells. Proc. Natl. Acad. Sci. USA 86:282–86 Dengjel J, Schoor O, Fischer R, Reich M, Kraus M, et al. 2005. Autophagy promotes MHC class II presentation of peptides from intracellular source proteins. Proc. Natl. Acad. Sci. USA 102:7922–27 Nimmerjahn F, Milosevic S, Behrends U, Jaffee EM, Pardoll DM, et al. 2003. Major histocompatibility complex class II-restricted presentation of a cytosolic antigen by autophagy. Eur. J. Immunol. 33:1250–59

Boon et al.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

65. Paludan C, Schmid D, Landthaler M, Vockerodt M, Kube D, et al. 2005. Endogenous MHC class II processing of a viral nuclear antigen after autophagy. Science 307:593–96 66. Vijayasaradhi S, Xu Y, Bouchard B, Houghton AN. 1995. Intracellular sorting and targeting of melanosomal membrane proteins: identiﬁcation of signals for sorting of the human brown locus protein, gp75. J. Cell Biol. 130:807–20 67. Wang S, Bartido S, Yang G, Qin J, Moroi Y, et al. 1999. A role for a melanosome transport signal in accessing the MHC class II presentation pathway and in eliciting CD4+ T cell responses. J. Immunol. 163:5820–26 68. Chaux P, Vantomme V, Stroobant V, Thielemans K, Corthals J, et al. 1999. Identiﬁcation of MAGE-3 epitopes presented by HLA-DR molecules to CD4+ T lymphocytes. J. Exp. Med. 189:767–78 69. Manici S, Sturniolo T, Imro MA, Hammer J, Sinigaglia F, et al. 1999. Melanoma cells present a MAGE-3 epitope to CD4+ cytotoxic T cells in association with histocompatibility leukocyte antigen DR11. J. Exp. Med. 189:871–76 70. Schultz ES, Lethe B, Cambiaso CL, Van Snick J, Chaux P, et al. 2000. A MAGE-A3 peptide presented by HLA-DP4 is recognized on tumor cells by CD4+ cytolytic T lymphocytes. Cancer Res. 60:6272–75 71. Zhang Y, Chaux P, Stroobant V, Eggermont AM, Corthals J, et al. 2003. A MAGE-3 peptide presented by HLA-DR1 to CD4+ T cells that were isolated from a melanoma patient vaccinated with a MAGE-3 protein. J. Immunol. 171:219–25 72. Lin KY, Guarnieri FG, Staveley-O’Carroll KF, Levitsky HI, August JT, et al. 1996. Treatment of established tumors with a novel vaccine that enhances major histocompatibility class II presentation of tumor antigen. Cancer Res. 56:21–26 73. Lin X, Dashti A, Schinazi RF, Tang J. 1993. Intracellular diversion of glycoprotein GP160 of human immunodeﬁciency virus to lysosomes as a strategy of AIDS gene therapy. FASEB J. 7:1070–80 74. Rowell JF, Ruff AL, Guarnieri FG, Staveley-O’Carroll K, Lin X, et al. 1995. Lysosomeassociated membrane protein-1-mediated targeting of the HIV-1 envelope protein to an endosomal/lysosomal compartment enhances its presentation to MHC class II-restricted T cells. J. Immunol. 155:1818–28 75. Ruff AL, Guarnieri FG, Staveley-O’Carroll K, Siliciano RF, August JT. 1997. The enhanced immune response to the HIV gp160/LAMP chimeric gene product targeted to the lysosome membrane protein trafﬁcking pathway. J. Biol. Chem. 272:8671–78 76. Thomson SA, Burrows SR, Misko IS, Moss DJ, Coupar BE, Khanna R. 1998. Targeting a polyepitope protein incorporating multiple class II-restricted viral epitopes to the secretory/endocytic pathway facilitates immune recognition by CD4+ cytotoxic T lymphocytes: a novel approach to vaccine design. J. Virol. 72:2246–52 77. Wu TC, Guarnieri FG, Staveley-O’Carroll KF, Viscidi RP, Levitsky HI, et al. 1995. Engineering an intracellular pathway for major histocompatibility complex class II presentation of antigens. Proc. Natl. Acad. Sci. USA 92:11671–75 78. Bonehill A, Heirman C, Tuyaerts S, Michiels A, Breckpot K, et al. 2004. Messenger RNAelectroporated dendritic cells presenting MAGE-A3 simultaneously in HLA class I and class II molecules. J. Immunol. 172:6649–57 79. Yewdell JW, Anton LC, Bennink JR. 1996. Defective ribosomal products (DRiPs): a major source of antigenic peptides for MHC class I molecules? J. Immunol. 157:1823–26 80. Yu Z, Theoret MR, Touloukian CE, Surman DR, Garman SC, et al. 2004. Poor immunogenicity of a self/tumor antigen derives from peptide-MHC-I instability and is independent of tolerance. J. Clin. Invest. 114:551–59 www.annualreviews.org • Human T Cell Responses Against Melanoma

199

ARI

23 February 2006

15:18

81. Parkhurst MR, Salgaller ML, Southwood S, Robbins PF, Sette A, et al. 1996. Improved induction of melanoma-reactive CTL with peptides from the melanoma antigen gp100 modiﬁed at HLA-A∗ 0201-binding residues. J. Immunol. 157:2539–48 82. Pass HA, Schwarz SL, Wunderlich JR, Rosenberg SA. 1998. Immunization of patients with melanoma peptide vaccines: immunologic assessment using the ELISPOT assay. Cancer J. Sci. Am. 4:316–23 83. Clay TM, Custer MC, McKee MD, Parkhurst M, Robbins PF, et al. 1999. Changes in the ﬁne speciﬁcity of gp100209 –217 -reactive T cells in patients following vaccination with a peptide modiﬁed at an HLA-A2.1 anchor residue. J. Immunol. 162:1749–55 84. Valmori D, Fonteneau JF, Lizana CM, Gervois N, Lienard D, et al. 1998. Enhanced generation of speciﬁc tumor-reactive CTL in vitro by selected Melan-A/MART-1 immunodominant peptide analogues. J. Immunol. 160:1750–58 85. Rivoltini L, Squarcina P, Loftus DJ, Castelli C, Tarsini P, et al. 1999. A superagonist variant of peptide MART1/Melan A27–35 elicits anti-melanoma CD8+ T cells with enhanced functional characteristics: implication for more effective immunotherapy. Cancer Res. 59:301–6 86. Valmori D, Levy F, Miconnet I, Zajac P, Spagnoli GC, et al. 2000. Induction of potent antitumor CTL responses by recombinant vaccinia encoding a melan-A peptide analogue. J. Immunol. 164:1125–31 87. Dutoit V, Rubio-Godoy V, Pittet MJ, Zippelius A, Dietrich PY, et al. 2002. Degeneracy of antigen recognition as the molecular basis for the high frequency of naive A2/Melan-A peptide multimer+ CD8+ T cells in humans. J. Exp. Med. 196:207–16 88. Ayyoub M, Zippelius A, Pittet MJ, Rimoldi D, Valmori D, et al. 2003. Activation of human melanoma reactive CD8+ T cells by vaccination with an immunogenic peptide analog derived from Melan-A/melanoma antigen recognized by T cells. Clin. Cancer Res. 9:669–77 89. Chen JL, Dunbar PR, Gileadi U, Jager E, Gnjatic S, et al. 2000. Identiﬁcation of NYESO-1 peptide analogues capable of improved stimulation of tumor-reactive CTL. J. Immunol. 165:948–55 90. Romero P, Dutoit V, Rubio-Godoy V, Lienard D, Speiser D, et al. 2001. CD8+ T-cell response to NY-ESO-1: relative antigenicity and in vitro immunogenicity of natural and analogue sequences. Clin. Cancer Res. 7:S766–72 91. Rosenberg SA, Packard BS, Aebersold PM, Solomon D, Topalian SL, et al. 1988. Use of tumor-inﬁltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma. A preliminary report. N. Engl. J. Med. 319:1676–80 92. Yannelli JR, Hyatt C, McConnell S, Hines K, Jacknin L, et al. 1996. Growth of tumorinﬁltrating lymphocytes from human solid cancers: summary of a 5-year experience. Int. J. Cancer 65:413–21 93. Rosenberg SA. 2001. Progress in the development of immunotherapy for the treatment of patients with cancer. J. Intern. Med. 250:462–75 94. Coulie PG, Somville M, Lehmann F, Hainaut P, Brasseur F, et al. 1992. Precursor frequency analysis of human cytolytic T lymphocytes directed against autologous melanoma cells. Int. J. Cancer 50:289–97 95. Mazzocchi A, Belli F, Mascheroni L, Vegetti C, Parmiani G, Anichini A. 1994. Frequency of cytotoxic T lymphocyte precursors (CTLp) interacting with autologous tumor via the T-cell receptor: limiting dilution analysis of speciﬁc CTLp in peripheral blood and tumor-invaded lymph nodes of melanoma patients. Int. J. Cancer 58:330– 39

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

200

Boon et al.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

96. Sensi M, Parmiani G. 1995. Analysis of TCR usage in human tumors: a new tool for assessing tumor-speciﬁc immune responses. Immunol. Today 16:588–95 97. Germeau C, Ma W, Schiavetti F, Lurquin C, Henry E, et al. 2005. High frequency of antitumor T cells in the blood of melanoma patients before and after vaccination with tumor antigens. J. Exp. Med. 201:241–48 98. Letsch A, Keilholz U, Schadendorf D, Nagorsen D, Schmittel A, et al. 2000. High frequencies of circulating melanoma-reactive CD8+ T cells in patients with advanced melanoma. Int. J. Cancer 87:659–64 99. Chaux P, Vantomme V, Coulie P, Boon T, van der Bruggen P. 1998. Estimation of the frequencies of anti-MAGE-3 cytolytic T-lymphocyte precursors in blood from individuals without cancer. Int. J. Cancer 77:538–42 100. Lonchay C, van der Bruggen P, Connerotte T, Hanagiri T, Coulie P, et al. 2004. Correlation between tumor regression and T cell responses in melanoma patients vaccinated with a MAGE antigen. Proc. Natl. Acad. Sci. USA 101(Suppl. 2):14631–38 101. Hanagiri T, van Baren N, Neyns B, Boon T, Coulie PG. 2006. Analysis of one of the rare melanoma patients with a spontaneous CTL response to a MAGE-A3 peptide presented by HLA-A1. Cancer Immunol. Immunother. 55:178–84 102. Jager E, Nagata Y, Gnjatic S, Wada H, Stockert E, et al. 2000. Monitoring CD8 T cell responses to NY-ESO-1: correlation of humoral and cellular immune responses. Proc. Natl. Acad. Sci. USA 97:4760–65 103. Jager E, Gnjatic S, Nagata Y, Stockert E, Jager D, et al. 2000. Induction of primary NY-ESO-1 immunity: CD8+ T lymphocyte and antibody responses in peptidevaccinated patients with NY-ESO-1+ cancers. Proc. Natl. Acad. Sci. USA 97:12198– 203 104. Valmori D, Dutoit V, Lienard D, Rimoldi D, Pittet MJ, et al. 2000. Naturally occurring human lymphocyte antigen-A2 restricted CD8+ T-cell response to the cancer testis antigen NY-ESO-1 in melanoma patients. Cancer Res. 60:4499–506 105. Grifﬁoen M, Borghi M, Schrier PI, Osanto S. 2001. Detection and quantiﬁcation of CD8+ T cells speciﬁc for HLA-A∗ 0201-binding melanoma and viral peptides by the IFN-γ-ELISPOT assay. Int. J. Cancer 93:549–55 106. Zorn E, Hercend T. 1999. A MAGE-6-encoded peptide is recognized by expanded lymphocytes inﬁltrating a spontaneously regressing human primary melanoma lesion. Eur. J. Immunol. 29:602–7 107. Ferradini L, Mackensen A, Genevee C, Bosq J, Duvillard P, et al. 1993. Analysis of T cell receptor variability in tumor-inﬁltrating lymphocytes from a human regressive melanoma. Evidence for in situ T cell clonal expansion. J. Clin. Invest. 91:1183–90 108. Mackensen A, Ferradini L, Carcelain G, Triebel F, Faure F, et al. 1993. Evidence for in situ ampliﬁcation of cytotoxic T-lymphocytes with antitumor activity in a human regressive melanoma. Cancer Res. 53:3569–73 109. Lurquin C, Lethe B, De Plaen E, Corbiere V, Theate I, et al. 2005. Contrasting frequencies of antitumor and anti-vaccine T cells in metastases of a melanoma patient vaccinated with a MAGE tumor antigen. J. Exp. Med. 201:249–57 110. Wang HY, Lee DA, Peng G, Guo Z, Li Y, et al. 2004. Tumor-speciﬁc human CD4+ regulatory T cells and their ligands: implications for immunotherapy. Immunity 20:107– 18 111. Zorn E, Hercend T. 1999. A natural cytotoxic T cell response in a spontaneously regressing human melanoma targets a neoantigen resulting from a somatic point mutation. Eur. J. Immunol. 29:592–601 www.annualreviews.org • Human T Cell Responses Against Melanoma

201

ARI

23 February 2006

15:18

112. Novellino L, Renkvist N, Rini F, Mazzocchi A, Rivoltini L, et al. 2003. Identiﬁcation of a mutated receptor-like protein tyrosine phosphatase κ as a novel, class II HLA-restricted melanoma antigen. J. Immunol. 170:6363–70 113. Sensi M, Nicolini G, Zanon M, Colombo C, Molla A, et al. 2005. Immunogenicity without immunoselection: a mutant but functional antioxidant enzyme retained in a human metastatic melanoma and targeted by CD8+ T cells with a memory phenotype. Cancer Res. 65:632–40 114. Robbins PF, El-Gamil M, Li YF, Zeng G, Dudley M, Rosenberg SA. 2002. Multiple HLA class II-restricted melanocyte differentiation antigens are recognized by tumor-inﬁltrating lymphocytes from a patient with melanoma. J. Immunol. 169:6036–47 115. Topalian SL, Gonzales MI, Parkhurst M, Li YF, Southwood S, et al. 1996. Melanomaspeciﬁc CD4+ T cells recognize nonmutated HLA-DR-restricted tyrosinase epitopes. J. Exp. Med. 183:1965–71 116. Romero P, Valmori D, Pittet MJ, Zippelius A, Rimoldi D, et al. 2002. Antigenicity and immunogenicity of Melan-A/MART-1 derived peptides as targets for tumor reactive CTL in human melanoma. Immunol. Rev. 188:81–96 117. Zippelius A, Pittet MJ, Batard P, Rufer N, de Smedt M, et al. 2002. Thymic selection generates a large T cell pool recognizing a self-peptide in humans. J. Exp. Med. 195:485–94 118. Dunbar PR, Smith CL, Chao D, Salio M, Shepherd D, et al. 2000. A shift in the phenotype of melan-A-speciﬁc CTL identiﬁes melanoma patients with an active tumor-speciﬁc immune response. J. Immunol. 165:6644–52 119. Lee PP, Yee C, Savage PA, Fong L, Brockstedt D, et al. 1999. Characterization of circulating T cells speciﬁc for tumor-associated antigens in melanoma patients. Nat. Med. 5:677–85 120. Palermo B, Campanelli R, Mantovani S, Lantelme E, Manganoni AM, et al. 2001. Diverse expansion potential and heterogeneous avidity in tumor-associated antigen-speciﬁc T lymphocytes from primary melanoma patients. Eur. J. Immunol. 31:412–20 121. Halpern AC, Schuchter LM. 1997. Prognostic models in melanoma. Semin. Oncol. 24:S2–7 122. Clemente CG, Mihm MC Jr, Bufalino R, Zurrida S, Collini P, Cascinelli N. 1996. Prognostic value of tumor inﬁltrating lymphocytes in the vertical growth phase of primary cutaneous melanoma. Cancer 77:1303–10 123. Mihm MC Jr, Clemente CG, Cascinelli N. 1996. Tumor inﬁltrating lymphocytes in lymph node melanoma metastases: a histopathologic prognostic indicator and an expression of local immune response. Lab. Invest. 74:43–47 124. van Oijen M, Bins A, Elias S, Sein J, Weder P, et al. 2004. On the role of melanoma-speciﬁc CD8+ T-cell immunity in disease progression of advanced-stage melanoma patients. Clin. Cancer Res. 10:4754–60 125. Vesalainen S, Lipponen P, Talja M, Syrjanen K. 1994. Histological grade, perineural inﬁltration, tumour-inﬁltrating lymphocytes and apoptosis as determinants of long-term prognosis in prostatic adenocarcinoma. Eur. J. Cancer 30A:1797–803 126. Naito Y, Saito K, Shiiba K, Ohuchi A, Saigenji K, et al. 1998. CD8+ T cells inﬁltrated within cancer cell nests as a prognostic factor in human colorectal cancer. Cancer Res. 58:3491–94 127. Schumacher K, Haensch W, Roefzaad C, Schlag PM. 2001. Prognostic signiﬁcance of activated CD8+ T cell inﬁltrations within esophageal carcinomas. Cancer Res. 61:3932–36 128. Zhang L, Conejo-Garcia JR, Katsaros D, Gimotty PA, Massobrio M, et al. 2003. Intratumoral T cells, recurrence, and survival in epithelial ovarian cancer. N. Engl. J. Med. 348:203–13

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

202

Boon et al.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

129. Molldrem JJ, Lee PP, Wang C, Felio K, Kantarjian HM, et al. 2000. Evidence that speciﬁc T lymphocytes may participate in the elimination of chronic myelogenous leukemia. Nat. Med. 6:1018–23 130. Marincola FM, Jaffee EM, Hicklin DJ, Ferrone S. 2000. Escape of human solid tumors from T-cell recognition: molecular mechanisms and functional signiﬁcance. Adv. Immunol. 74:181–273 131. Garcia-Lora A, Algarra I, Garrido F. 2003. MHC class I antigens, immune surveillance, and tumor immune escape. J. Cell. Physiol. 195:346–55 132. Kageshita T, Ishihara T, Campoli M, Ferrone S. 2005. Selective monomorphic and polymorphic HLA class I antigenic determinant loss in surgically removed melanoma lesions. Tissue Antigens 65:419–28 133. Yamshchikov GV, Mullins DW, Chang CC, Ogino T, Thompson L, et al. 2005. Sequential immune escape and shifting of T cell responses in a long-term survivor of melanoma. J. Immunol. 174:6863–71 134. Zippelius A, Batard P, Rubio-Godoy V, Bioley G, Lienard D, et al. 2004. Effector function of human tumor-speciﬁc CD8 T cells in melanoma lesions: a state of local functional tolerance. Cancer Res. 64:2865–73 135. Bronte V, Kasic T, Gri G, Gallana K, Borsellino G, et al. 2005. Boosting antitumor responses of T lymphocytes inﬁltrating human prostate cancers. J. Exp. Med. 201:1257– 68 136. Radoja S, Saio M, Schaer D, Koneru M, Vukmanovic S, Frey AB. 2001. CD8+ tumorinﬁltrating T cells are deﬁcient in perforin-mediated cytolytic activity due to defective microtubule-organizing center mobilization and lytic granule exocytosis. J. Immunol. 167:5042–51 137. Sakaguchi S, Fukuma K, Kuribayashi K, Masuda T. 1985. Organ-speciﬁc autoimmune diseases induced in mice by elimination of T cell subset. I. Evidence for the active participation of T cells in natural self-tolerance; deﬁcit of a T cell subset as a possible cause of autoimmune disease. J. Exp. Med. 161:72–87 138. Sakaguchi S. 2004. Naturally arising CD4+ regulatory T cells for immunologic selftolerance and negative control of immune responses. Annu. Rev. Immunol. 22:531– 62 139. Dieckmann D, Plottner H, Berchtold S, Berger T, Schuler G. 2001. Ex vivo isolation and characterization of CD4+ CD25+ T cells with regulatory properties from human blood. J. Exp. Med. 193:1303–10 140. Levings MK, Sangregorio R, Roncarolo MG. 2001. Human CD25+ CD4+ T regulatory cells suppress naive and memory T cell proliferation and can be expanded in vitro without loss of function. J. Exp. Med. 193:1295–302 141. Jonuleit H, Schmitt E, Stassen M, Tuettenberg A, Knop J, Enk AH. 2001. Identiﬁcation and functional characterization of human CD4+ CD25+ T cells with regulatory properties isolated from peripheral blood. J. Exp. Med. 193:1285–94 142. Wang HY, Peng G, Guo Z, Shevach EM, Wang RF. 2005. Recognition of a new ARTC1 peptide ligand uniquely expressed in tumor cells by antigen-speciﬁc CD4+ regulatory T cells. J. Immunol. 174:2661–70 143. Zea AH, Rodriguez PC, Atkins MB, Hernandez C, Signoretti S, et al. 2005. Arginaseproducing myeloid suppressor cells in renal cell carcinoma patients: a mechanism of tumor evasion. Cancer Res. 65:3044–48 144. Gorelik L, Flavell RA. 2001. Immune-mediated eradication of tumors through the blockade of transforming growth factor-β signaling in T cells. Nat. Med. 7:1118–22 www.annualreviews.org • Human T Cell Responses Against Melanoma

203

ARI

23 February 2006

15:18

145. Uyttenhove C, Pilotte L, Theate I, Stroobant V, Colau D, et al. 2003. Evidence for a tumoral immune resistance mechanism based on tryptophan degradation by indoleamine 2,3-dioxygenase. Nat. Med. 9:1269–74 146. Rubinstein N, Alvarez M, Zwirner NW, Toscano MA, Ilarregui JM, et al. 2004. Targeted inhibition of galectin-1 gene expression in tumor cells results in heightened T cellmediated rejection: a potential mechanism of tumor-immune privilege. Cancer Cell 5:241– 51 147. Chen Q, Daniel V, Maher DW, Hersey P. 1994. Production of IL-10 by melanoma cells: examination of its role in immunosuppression mediated by melanoma. Int. J. Cancer 56:755–60 148. Dong H, Strome SE, Salomao DR, Tamura H, Hirano F, et al. 2002. Tumor-associated B7-H1 promotes T-cell apoptosis: a potential mechanism of immune evasion. Nat. Med. 8:793–800 149. Speiser DE, Kyburz D, Stubi U, Hengartner H, Zinkernagel RM. 1992. Discrepancy between in vitro measurable and in vivo virus neutralizing cytotoxic T cell reactivities. Low T cell receptor speciﬁcity and avidity sufﬁcient for in vitro proliferation or cytotoxicity to peptide-coated target cells but not for in vivo protection. J. Immunol. 149:972– 80 150. Altman JD, Moss PA, Goulder PJ, Barouch DH, McHeyzer-Williams MG, et al. 1996. Phenotypic analysis of antigen-speciﬁc T lymphocytes. Science 274:94–96 151. Herr W, Schneider J, Lohse AW, Meyer zum Buschenfelde KH, Wolfel T. 1996. Detection and quantiﬁcation of blood-derived CD8+ T lymphocytes secreting tumor necrosis factor α in response to HLA-A2.1-binding melanoma and viral peptide antigens. J. Immunol. Methods 191:131–42 152. Herr W, Linn B, Leister N, Wandel E, Meyer zum Buschenfelde KH, Wolfel T. 1997. The use of computer-assisted video image analysis for the quantiﬁcation of CD8+ T lymphocytes producing tumor necrosis factor α spots in response to peptide antigens. J. Immunol. Methods 203:141–52 153. Novak EJ, Liu AW, Nepom GT, Kwok WW. 1999. MHC class II tetramers identify peptide-speciﬁc human CD4+ T cells proliferating in response to inﬂuenza A antigen. J. Clin. Invest. 104:R63–67 154. Zhang Y, Renkvist N, Sun Z, Schuler-Thurner B, Glaichenhaus N, et al. 2005. A polyclonal anti-vaccine CD4 T cell response detected with HLA-DP4 multimers in a melanoma patient vaccinated with MAGE-3.DP4-peptide-pulsed dendritic cells. Eur. J. Immunol. 35:1066–75 155. Schuler-Thurner B, Schultz ES, Berger TG, Weinlich G, Ebner S, et al. 2002. Rapid induction of tumor-speciﬁc type 1 T helper cells in metastatic melanoma patients by vaccination with mature, cryopreserved, peptide-loaded monocyte-derived dendritic cells. J. Exp. Med. 195:1279–88 156. Zhang Y, Sun Z, Nicolay H, Meyer RG, Renkvist N, et al. 2005. Monitoring of antivaccine CD4 T cell frequencies in melanoma patients vaccinated with a MAGE-3 protein. J. Immunol. 174:2404–11 157. Marchand M, Weynants P, Rankin E, Arienti F, Belli F, et al. 1995. Tumor regression responses in melanoma patients treated with a peptide encoded by gene MAGE-3. Int. J. Cancer 63:883–85 158. Ottaviani S, Zhang Y, Boon T, van der Bruggen P. 2005. A MAGE-1 antigenic peptide recognized by human cytolytic T lymphocytes on HLA-A2 tumor cells. Cancer Immunol. Immunother. 54:1214–20

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

204

Boon et al.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

159. Marchand M, van Baren N, Weynants P, Brichard V, Dreno B, et al. 1999. Tumor regressions observed in patients with metastatic melanoma treated with an antigenic peptide encoded by gene MAGE-3 and presented by HLA-A1. Int. J. Cancer 80:219–30 160. Coulie PG, Karanikas V, Colau D, Lurquin C, Landry C, et al. 2001. A monoclonal cytolytic T-lymphocyte response observed in a melanoma patient vaccinated with a tumorspeciﬁc antigenic peptide encoded by gene MAGE-3. Proc. Natl. Acad. Sci. USA 98:10290– 95 161. Karanikas V, Lurquin C, Colau D, van Baren N, De Smet C, et al. 2003. Monoclonal anti-MAGE-3 CTL responses in melanoma patients displaying tumor regression after vaccination with a recombinant canarypox virus. J. Immunol. 171:4898–904 162. Thurner B, Haendle I, Roder C, Dieckmann D, Keikavoussi P, et al. 1999. Vaccination with mage-3A1 peptide-pulsed mature, monocyte-derived dendritic cells expands speciﬁc cytotoxic T cells and induces regression of some metastases in advanced stage IV melanoma. J. Exp. Med. 190:1669–78 163. Godelaine D, Carrasco J, Lucas S, Karanikas V, Schuler-Thurner B, et al. 2003. Polyclonal CTL responses observed in melanoma patients vaccinated with dendritic cells pulsed with a MAGE-3.A1 peptide. J. Immunol. 171:4893–97 164. Paczesny S, Banchereau J, Wittkowski KM, Saracino G, Fay J, Palucka AK. 2004. Expansion of melanoma-speciﬁc cytolytic CD8+ T cell precursors in patients with metastatic melanoma vaccinated with CD34+ progenitor-derived dendritic cells. J. Exp. Med. 199:1503–11 165. Gajewski TF, Fallarino F, Ashikari A, Sherman M. 2001. Immunization of HLA-A2+ melanoma patients with MAGE-3 or MelanA peptide-pulsed autologous peripheral blood mononuclear cells plus recombinant human interleukin 12. Clin. Cancer Res. 7:S895–901 166. Chianese-Bullock KA, Pressley J, Garbee C, Hibbitts S, Murphy C, et al. 2005. MAGEA1-, MAGE-A10-, and gp100-derived peptides are immunogenic when combined with granulocyte-macrophage colony-stimulating factor and montanide ISA-51 adjuvant and administered as part of a multipeptide vaccine for melanoma. J. Immunol. 174:3080–86 167. Bettinotti MP, Panelli MC, Ruppe E, Mocellin S, Phan GQ, et al. 2003. Clinical and immunological evaluation of patients with metastatic melanoma undergoing immunization with the HLA-Cw∗ 0702-associated epitope MAGE-A12:170–178. Int. J. Cancer 105:210– 16 168. Marchand M, Punt CJ, Aamdal S, Escudier B, Kruit WH, et al. 2003. Immunisation of metastatic cancer patients with MAGE-3 protein combined with adjuvant SBAS-2: a clinical report. Eur. J. Cancer 39:70–77 169. Kruit WH, van Ojik HH, Brichard VG, Escudier B, Dorval T, et al. 2005. Phase 1/2 study of subcutaneous and intradermal immunization with a recombinant MAGE-3 protein in patients with detectable metastatic melanoma. Int. J. Cancer 117:596–604 170. Davis ID, Chen W, Jackson H, Parente P, Shackleton M, et al. 2004. Recombinant NY-ESO-1 protein with ISCOMATRIX adjuvant induces broad integrated antibody and CD4+ and CD8+ T cell responses in humans. Proc. Natl. Acad. Sci. USA 101:10697– 702 171. Wang F, Bade E, Kuniyoshi C, Spears L, Jeffery G, et al. 1999. Phase I trial of a MART-1 peptide vaccine with incomplete Freund’s adjuvant for resected high-risk melanoma. Clin. Cancer Res. 5:2756–65 172. Pittet MJ, Speiser DE, Lienard D, Valmori D, Guillaume P, et al. 2001. Expansion and functional maturation of human tumor antigen-speciﬁc CD8+ T cells after vaccination with antigenic peptide. Clin. Cancer Res. 7:S796–803 www.annualreviews.org • Human T Cell Responses Against Melanoma

205

ARI

23 February 2006

15:18

173. Cormier JN, Salgaller ML, Prevette T, Barracchini KC, Rivoltini L, et al. 1997. Enhancement of cellular immunity in melanoma patients immunized with a peptide from MART-1/Melan A. Cancer J. Sci. Am. 3:37–44 174. Speiser DE, Lienard D, Rufer N, Rubio-Godoy V, Rimoldi D, et al. 2005. Rapid and strong human CD8+ T cell responses to vaccination with peptide, IFA, and CpG oligodeoxynucleotide 7909. J. Clin. Invest. 115:739–46 175. Mackensen A, Herbst B, Chen JL, Kohler G, Noppen C, et al. 2000. Phase I study in melanoma patients of a vaccine with peptide-pulsed dendritic cells generated in vitro from CD34+ hematopoietic progenitor cells. Int. J. Cancer 86:385–92 176. Oertli D, Marti WR, Zajac P, Noppen C, Kocher T, et al. 2002. Rapid induction of speciﬁc cytotoxic T lymphocytes against melanoma-associated antigens by a recombinant vaccinia virus vector expressing multiple immunodominant epitopes and costimulatory molecules in vivo. Hum. Gene Ther. 13:569–75 177. Smith CL, Dunbar PR, Mirza F, Palmowski MJ, Shepherd D, et al. 2005. Recombinant modiﬁed vaccinia Ankara primes functionally activated CTL speciﬁc for a melanoma tumor antigen epitope in melanoma patients with a high risk of disease recurrence. Int. J. Cancer 113:259–66 178. Cebon J, Jager E, Shackleton MJ, Gibbs P, Davis ID, et al. 2003. Two phase I studies of low dose recombinant human IL-12 with Melan-A and inﬂuenza peptides in subjects with advanced malignant melanoma. Cancer Immun. 3:7–12 179. Sanderson K, Scotland R, Lee P, Liu D, Groshen S, et al. 2005. Autoimmunity in a phase I trial of a fully human anti-cytotoxic T-lymphocyte antigen-4 monoclonal antibody with multiple melanoma peptides and Montanide ISA 51 for patients with resected stages III and IV melanoma. J. Clin. Oncol. 23:741–50 180. Rosenberg SA, Yang JC, Schwartzentruber DJ, Hwu P, Marincola FM, et al. 1998. Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma. Nat. Med. 4:321–27 181. Smith JW 2nd, Walker EB, Fox BA, Haley D, Wisner KP, et al. 2003. Adjuvant immunization of HLA-A2-positive melanoma patients with a modiﬁed gp100 peptide induces peptide-speciﬁc CD8+ T-cell responses. J. Clin. Oncol. 21:1562–73 182. Lee P, Wang F, Kuniyoshi J, Rubio V, Stuges T, et al. 2001. Effects of interleukin-12 on the immune response to a multipeptide vaccine for resected metastatic melanoma. J. Clin. Oncol. 19:3836–47 183. Yamshchikov GV, Barnd DL, Eastham S, Galavotti H, Patterson JW, et al. 2001. Evaluation of peptide vaccine immunogenicity in draining lymph nodes and peripheral blood of melanoma patients. Int. J. Cancer 92:703–11 184. Slingluff CL Jr, Petroni GR, Yamshchikov GV, Hibbitts S, Grosh WW, et al. 2004. Immunologic and clinical outcomes of vaccination with a multiepitope melanoma peptide vaccine plus low-dose interleukin-2 administered either concurrently or on a delayed schedule. J. Clin. Oncol. 22:4474–85 185. Schaed SG, Klimek VM, Panageas KS, Musselli CM, Butterworth L, et al. 2002. T-cell responses against tyrosinase 368–376 (370D) peptide in HLA∗ A0201+ melanoma patients: randomized trial comparing incomplete Freund’s adjuvant, granulocyte macrophage colony-stimulating factor, and QS-21 as immunological adjuvants. Clin. Cancer Res. 8:967– 72 186. Banchereau J, Palucka AK, Dhodapkar M, Burkeholder S, Taquet N, et al. 2001. Immune and clinical responses in patients with metastatic melanoma to CD34+ progenitor-derived dendritic cell vaccine. Cancer Res. 61:6451–58

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

206

Boon et al.

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-06

ARI

23 February 2006

15:18

187. van Baren N, Bonnet M-C, Dr´eno B, Khammari A, Dorval T, et al. 2006. Tumoral and immunological responses following vaccination of metastatic melanoma patients with an ALVAC virus encoding two Mage antigens recognized by T cells. J. Clin. Oncol. In press 188. Slingluff CL Jr, Petroni GR, Yamshchikov GV, Barnd DL, Eastham S, et al. 2003. Clinical and immunologic results of a randomized phase II trial of vaccination using four melanoma peptides either administered in granulocyte-macrophage colony-stimulating factor in adjuvant or pulsed on dendritic cells. J. Clin. Oncol. 21:4016–26 189. Mitchell MS, Harel W, Kempf RA, Hu E, Kan-Mitchell J, et al. 1990. Active-speciﬁc immunotherapy for melanoma. J. Clin. Oncol. 8:856–69 190. Rosenberg SA, Yang JC, Restifo NP. 2004. Cancer immunotherapy: moving beyond current vaccines. Nat. Med. 10:909–15 191. Hersey P, Menzies SW, Halliday GM, Nguyen T, Farrelly ML, et al. 2004. Phase I/II study of treatment with dendritic cell vaccines in patients with disseminated melanoma. Cancer Immunol. Immunother. 53:125–34 192. Butterﬁeld LH, Ribas A, Dissette VB, Amarnani SN, Vu HT, et al. 2003. Determinant spreading associated with clinical response in dendritic cell-based immunotherapy for malignant melanoma. Clin. Cancer Res. 9:998–1008 193. de Vries IJ, Lesterhuis WJ, Scharenborg NM, Engelen LP, Ruiter DJ, et al. 2003. Maturation of dendritic cells is a prerequisite for inducing immune responses in advanced melanoma patients. Clin. Cancer Res. 9:5091–100 194. Lau R, Wang F, Jeffery G, Marty V, Kuniyoshi J, et al. 2001. Phase I trial of intravenous peptide-pulsed dendritic cells in patients with metastatic melanoma. J. Immunother. 24:66– 78 195. Baldo M, Schiavon M, Cicogna PA, Boccato P, Mazzoleni F. 1992. Spontaneous regression of subcutaneous metastasis of cutaneous melanoma. Plast. Reconstr. Surg. 90:1073– 76 196. Letsch A, Keilholz U, Fluck M, Nagorsen D, Asemissen AM, et al. 2005. Peptide vaccination after repeated resection of metastases can induce a prolonged relapse-free interval in melanoma patients. Int. J. Cancer 114:936–41 197. Rosenberg SA, Dudley ME. 2004. Cancer regression in patients with metastatic melanoma after the transfer of autologous antitumor lymphocytes. Proc. Natl. Acad. Sci. USA 101(Suppl. 2):14639–45 198. Dr´eno B, Nguyen JM, Khammari A, Pandolﬁno MC, Tessier MH, et al. 2002. Randomized trial of adoptive transfer of melanoma tumor-inﬁltrating lymphocytes as adjuvant therapy for stage III melanoma. Cancer Immunol. Immunother. 51:539–46 199. Lienard D, Rimoldi D, Marchand M, Dietrich PY, van Baren N, et al. 2004. Ex vivo detectable activation of Melan-A-speciﬁc T cells correlating with inﬂammatory skin reactions in melanoma patients vaccinated with peptides in IFA. Cancer Immun. 4:4 200. Bevan MJ. 2004. Helping the CD8+ T-cell response. Nat. Rev. Immunol. 4:595–602 201. Weber JS, Hua FL, Spears L, Marty V, Kuniyoshi C, Celis E. 1999. A phase I trial of an HLA-A1 restricted MAGE-3 epitope peptide with incomplete Freund’s adjuvant in patients with resected high-risk melanoma. J. Immunother. 22:431–40 202. Phan GQ, Yang JC, Sherry RM, Hwu P, Topalian SL, et al. 2003. Cancer regression and autoimmunity induced by cytotoxic T lymphocyte-associated antigen 4 blockade in patients with metastatic melanoma. Proc. Natl. Acad. Sci. USA 100:8372–77 203. Marincola FM, Rivoltini L, Salgaller ML, Player M, Rosenberg SA. 1996. Differential anti-MART-1/MelanA CTL activity in peripheral blood of HLA-A2 melanoma patients www.annualreviews.org • Human T Cell Responses Against Melanoma

207

ANRV270-IY24-06

ARI

23 February 2006

15:18

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

in comparison to healthy donors: evidence of in vivo priming by tumor cells. J. Immunother. Emphasis Tumor Immunol. 19:266–77 204. Wang E, Miller LD, Ohnmacht GA, Mocellin S, Perez-Diez A, et al. 2002. Prospective molecular proﬁling of melanoma metastases suggests classiﬁers of immune responsiveness. Cancer Res. 62:3581–86

208

Boon et al.

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:175-208. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-07

ARI

13 February 2006

13:5

FOXP3: Of Mice and Men Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Steven F. Ziegler Immunology Program, Benaroya Research Institute and Department of Immunology, University of Washington School of Medicine, Seattle, Washington 98101; email: [email protected]

Annu. Rev. Immunol. 2006. 24:209–26 First published online as a Review in Advance on December 1, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090547 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0209$20.00

Key Words regulatory T cell, autoimmunity, forkhead, transcription factor

Abstract The immune system has evolved mechanisms to recognize and eliminate threats, as well as to protect against self-destruction. Tolerance to self-antigens is generated through two fundamental mechanisms: (a) elimination of self-reactive cells in the thymus during selection and (b) generation of a variety of peripheral regulatory cells to control self-reactive cells that escape the thymus. It is becoming increasing apparent that a population of thymically derived CD4+ regulatory T cells, exempliﬁed by the expression of the IL-2Rα chain, is essential for the maintenance of peripheral tolerance. Recent work has shown that the forkhead family transcription factor Foxp3 is critically important for the development and function of the regulatory T cells. Lack of Foxp3 leads to development of fatal autoimmune lymphoproliferative disease; furthermore, ectopic Foxp3 expression can phenotypically convert effector T cells to regulatory T cells. This review focuses on Foxp3 expression and function and highlights differences between humans and mice regarding Foxp3 regulation.

209

ANRV270-IY24-07

ARI

13 February 2006

13:5

INTRODUCTION Foxp3: forkhead box protein P3

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Tregs: regulatory T cells CD4+ CD25+ Treg: a subset of CD4+ T cells that is capable of suppressing the proliferation and cytokine production of naive or memory T cells Forkhead family: a large family of transcriptional regulators named after the founding member, which was found to be the gene responsible for the forkhead mutation in Drosophila. All members of the family have a closely conserved motif, known as a forkhead box, or Fox, that is involved in DNA binding. The family is further subdivided based on homology outside the forkhead box.

Immunological tolerance to self-antigens is a tightly regulated process. A primary mechanism for self-tolerance is deletion of selfreactive cells in the thymus. However, this mechanism is not perfect, and autoreactive clones do escape into the periphery. Tolerance is maintained in the periphery through a variety of mechanisms, including a population of regulatory T cells that actively suppress the function of autoreactive T cells. These T cells, identiﬁed by their expression of CD4, the IL-2Rα chain (CD25), and the forkhead family transcription factor Foxp3, are know as regulatory T cells, or Tregs. They have the ability to inhibit the development of autoimmunity when transferred into the appropriate host. Recent work has shown that the forkhead/winged-helix protein Foxp3 is expressed predominantly in Tregs and is both necessary and sufﬁcient for their development and function. Several excellent reviews have been written recently on the development and function of CD4+ CD25+ Tregs (1–5). Thus, this review does not directly address these issues. Instead, I review what is known concerning the expression and function of Foxp3, which is critical for the development and function of Tregs, especially in the mouse. Although the overall importance of Foxp3 is obvious from the phenotype of humans and mice that lack this protein, very little is actually known about how it functions and what controls its expression. This review highlights the similarities and differences in Foxp3 expression function in human and mouse, with an emphasis on the differences.

IDENTIFICATION AND CHARACTERIZATION OF MOUSE Foxp3 Characterization of Scurfy (sf ) Mutant Mice The scurfy (sf ) mutation arose spontaneously in the partially inbred MR strain in 1949 at the 210

Ziegler

Oak Ridge National Laboratory. Subsequent studies showed the gene to be X-linked, with only Xsf /Y males affected (here, these mice are referred to as sf/Y). Grossly, affected sf/Y males exhibit several external markers of disease, including ear thickening; scaling of the ears, eyelids, feet, and tail; and severe runting. Internally, the mice exhibit lymphadenopathy, splenomegaly, hepatomegaly, and massive lymphocytic inﬁltrates in the skin and liver. These lesions closely resemble those seen in a graft-versus-host reaction. Mice display anemia and have a positive direct Coombs’s test, but they lack evidence of autoantibodies against double-strand DNA or small ribonuclear protein antigens (6). Taken together, these gross assessments suggest that sf is a mutation that causes an autoimmune-like disease in affected animals. In this regard, sf/Y mice resemble mice bearing targeted mutations in the ctla-4 or tgf-β1 genes. These mice die at approximately three weeks of age from a massive lymphoproliferative disease, with peripheral lymphocyte levels up to 20-fold greater than normal mice (7, 8). Phenotypically, scurfy disease is most consistent with a diagnosis of autoimmune lymphoproliferative disease. This idea is supported by the ﬁnding that mice expressing a transgenic T cell receptor (TCR) survive signiﬁcantly longer in the absence of antigen stimulation (60 days compared with 20– 24 days) and that they live a normal lifespan and are free of sf disease when they are also rag-2-null (sf/ova/rag mice). Finally, T cells from sf/Y mice display reduced sensitivity to inhibitors of T cell activation, suggesting that TCR signaling is dysregulated in sf/Y mice. Taken together, these data suggest the sf disease results from an inability to properly regulate antigen-driven T cell activation. One of the in vitro hallmarks of sf disease is the spontaneous proliferation and cytokine production exhibited by T cells isolated from sf/Y mice. Initial studies showed that cytokine production in vitro by ConAstimulated sf/Y splenocytes was greatly elevated (9). More detailed studies, using puriﬁed

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

ARI

13 February 2006

13:5

CD4+ T cells from sf/Y and normal littermate control (NLC) mice, showed that these cells were capable of proliferation and cytokine production when examined directly ex vivo. However, to achieve maximal T cell stimulation, TCR engagement was required, suggesting that the defect in sf mice is not nonspeciﬁc T cell activation. CD4+ T cells from sf/Y mice were hyperresponsive to TCR-mediated signals, responding to TCR engagement with crosslinking antibodies at concentrations 10-fold lower than their wild-type counterparts (10). Consistent with the increased expression of costimulatory molecules in sf/Y mice, CD4+ T cells from these mice responded to CD28 crosslinking (10). Recent studies have shown that the transcription factors NFAT and AP-1 are constitutively activated in CD4+ T cells from sf/Y mice (11). Also, these same T cells have a marked decrease in their sensitivity to cyclosporin A, with an IC50 15-fold higher than NLCderived T cells (10). Both observations are indicative of an alteration in TCR-mediated signal transduction in CD4+ T cells from sf/Y mice and may account for the increased cytokine production exhibited by these cells.

Consequence of Foxp3 Expression on CD4+ T Cell Function The gene responsible for scurfy disease was molecularly cloned using a standard positional cloning approach (12). Sequence analysis of the cloned gene showed that it encoded a novel member of the forkhead family of transcriptional regulators, Foxp3 (see below for more details). To verify this gene as the gene responsible for the scurfy mutation, transgenic mice were generated using a cosmid clone containing mouse Foxp3. When bred to sf/Y mice, expression of the transgene was capable of rescuing the mice from scurfy disease and demonstrated that a mutation in Foxp3 was indeed responsible for the scurfy phenotype (12). The Foxp3 transgenic mice (referred to as Foxp3-Tg hereafter) also provided a model

system for examining the in vivo consequences of Foxp3 expression. Several individual Foxp3-Tg lines were established, each with differing levels of transgene expression (12, 13). Each of these lines, when bred into an otherwise wild-type background, resulted in a reduction in peripheral CD4+ and CD8+ T cell numbers, with the extent of the reduction directly correlated with transgene copy number and expression (12). However, thymic cellularity was unaffected, as was positive and negative selection. Thus, levels of Foxp3 determined the number of peripheral T cells, while having little effect on the number and differentiation of thymocytes. To examine the function of T cells from Foxp3-Tg mice, a single transgenic line was chosen. This line (2826) had approximately 16 copies of the transgene and had approximately ﬁvefold higher levels of Foxp3 expression (13, 13a). Although T cell development appears to be normal in these mice, peripheral T cell numbers and representation in secondary lymphoid organs are dramatically decreased. Using a variety of in vitro and in vivo assays, the function of T cells from the Foxp3Tg mice was shown to have severely decreased responses when activated through the TCR. For example, puriﬁed CD4+ T cells displayed dramatically reduced proliferative responses when stimulated in vitro (13). Similarly, these T cells produced virtually no IL-2 upon stimulation. In fact, by any measure, CD4+ T cells from Foxp3-Tg mice were functionally inert. Taken together, these data reﬂect a generalized deﬁcit in cellular activation of CD4+ T cells that express Foxp3. The failure of T cells from Foxp3-Tg mice to function in vitro is also reﬂected in the response of these mice to immunologic challenge in vivo. The Foxp3-Tg mice failed to mount an antigen-mediated contact sensitivity response (13). Also, recent data demonstrate that Foxp3-Tg mice have dramatically depressed responses to T-dependent antigens (14). The latter response appears to be due to a reduced ability to produce cytokines and to www.annualreviews.org • FOXP 3

NFAT: nuclear factor of activated T cells

211

ANRV270-IY24-07

ARI

13 February 2006

13:5

elaborate CD40L on their cell surface. Thus, overexpression of Foxp3 in CD4+ T cells in vivo results in lowered overall numbers of T cells, with impaired functionality of those T cells that remain.

Foxp3: Necessary and Sufficient for Mouse CD4+ CD25+ Treg Development and Function

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Although it is clear that expression of Foxp3 in CD4+ T cells has deleterious effects on their ability to respond normally to TCR-mediated signals (13, 14), the fact that a population of Foxp3+ T cells exists suggests a role for this protein in CD4+ T cell function. Clues for this function come from an initial analysis of the cells that express Foxp3 in the mouse. Several groups simultaneously reported that the predominant cell type that expressed Foxp3 was CD4+ CD25+ T cells (15–18), the population of CD4+ T cells that can suppress the proliferation and cytokine production of TCR-stimulated conventional CD4+ T cells (19, 20). The role of Foxp3 in these cells has been elucidated in mice through a combination of genetic and direct functional studies. Taken as a whole, these data demonstrate that Foxp3 expression in mouse CD4+ T cells is sufﬁcient to mark these T cells as regulatory. CD25, the IL-2Rα chain, had previously been shown to be the only reliable marker for these cells (19); however, it is also a marker of activated CD4+ T cells. Thus, the identiﬁcation of Foxp3 as a marker for this population of T cells was critical for the further analysis of these cells and their role in peripheral tolerance. In addition to the correlation of Foxp3 expression and CD4+ CD25+ Tregs, the ability of Foxp3 to reprogram CD4+ T cell function has recently been reported (15–17). Several lines of evidence demonstrate a role for Foxp3 in the development and/or function of CD4+ CD25+ Tregs. First, as described earlier, CD4+ CD25+ T cells display Foxp3 expression, whereas other T cell subsets in the mouse do not have detectable expression (15, 212

Ziegler

16). In fact, using a knockin allele of Foxp3 consisting of an in-frame fusion of green ﬂuorescent protein (GFP) and Foxp3, Fontenot et al. (21) showed that αβTCR+ CD4+ T cells are the predominant cell population that expresses Foxp3. They also found that the relative number of CD25+ and CD25− cells that were Foxp3+ varied according to the source of the T cells. The ability of the Foxp3+ T cells to act as suppressors in vitro was not dependent on expression of CD25. These data were conﬁrmed using a second knockin mouse containing an IRES-RFP (internal ribosome entry site–red ﬂuorescent protein) cassette in the 3 untranslated region of the Foxp3 gene (22). The second line of evidence that Foxp3 expression is necessary and sufﬁcient for mouse Treg development and function comes from studies using ectopic expression of Foxp3 in otherwise conventional T cells. Infection of CD4+ CD25− T cells with a retrovirus expressing Foxp3 converted those cells to a regulatory phenotype, with the infected cells capable of suppressing the proliferation of uninfected CD4+ CD25− T cells (16). The infected cells can also function as Tregs in vivo. In a wide variety of adoptive transfer models of autoimmune disease, cotransfer of CD4+ CD25+ T cells has been shown to protect the host from disease development in these model systems (23–26). Using naive CD4+ CD25− T cells expressing Foxp3 following retroviral gene transfer, Hori et al. (16) showed that cotransfer of cells infected with the Foxp3 retrovirus also protects host mice from autoimmune gastritis. Similarly, CD4+ CD25− T cells expressing Foxp3 also protects against colitis (15). These results have been conﬁrmed by several groups, demonstrating that expression of Foxp3 can convert cells to a Treg-like phenotype. Recent studies have suggested a potential therapeutic use of ectopic Foxp3 expression. BDC2.5 TCR transgenic T cells, transduced with a Foxp3-expressing retrovirus, were capable of ameliorating disease when transferred into nonobese diabetic (NOD) mice with recent

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

ARI

13 February 2006

13:5

onset disease (27). However, transfer of polyclonal NOD CD4+ T cells infected with the Foxp3 retrovirus did not protect, suggesting that the frequency of antigen-speciﬁc Tregs is a major factor in the ability of Tregs to control autoimmunity. Similar results have been seen with transfer of puriﬁed CD4+ CD25+ Tregs (25, 28). Complimentary to the experiments using retroviral-mediated delivery of Foxp3 to T cells, Treg function in Foxp3-Tg animals has also been studied. As described above, these mice contain a transgene consisting of cosmid clone containing the mouse Foxp3 gene (12), with approximately 80% of CD4+ T cells in these animals expressing Foxp3 (D.J. Kasprowicz & S.F. Ziegler, unpublished observation). An examination of the function of the CD4+ T cells from these mice showed that the entire population had in vitro regulatory activity (13a, 17). Thus, similar to what was seen with retroviral-mediated introduction of Foxp3, transgenic expression also converted conventional CD4+ T cells to a regulatory phenotype. Interestingly, CD8+ in these mice also displayed in vitro regulatory activity, demonstrating that Foxp3 expression in nonCD4+ T cells was also capable of phenotypic conversion. Just as ectopic Foxp3 expression has been shown to drive Treg function, lack of Foxp3 has been correlated with a lack of Treg cells. For example, mice that lack functional Foxp3, either via the scurfy mutation or a targeted mutation, lack Treg activity (15, 17, 21). Importantly, a conditional deletion of Foxp3 in CD4+ T cells also led to a lymphoproliferative disease indistinguishable from that seen in scurfy males (21). Further evidence for the critical role of Foxp3 in determining the Treg lineage comes from mixed bone marrow chimera experiments. Lethally irradiated mice were reconstituted with a mixture of bone marrow from wild-type and Foxp3− mice, congenically marked to allow the contribution of each to the reconstituted animal. Only bone marrow from the wild-type donor contributed to the CD4+ CD25+ compartment, demonstrat-

ing that Foxp3 is needed in the development of Tregs and that the role of Foxp3 is cell intrinsic (15). Taken as a whole, the data described in this section demonstrate the absolute need for Foxp3 in the development and function of Tregs. They also show that ectopic expression of Foxp3 can cause a phenotypic conversion of T cells to a regulatory phenotype. Thus, in the mouse, Foxp3 is both necessary and sufﬁcient for Treg development and function.

TGF-β: transforming growth factor-β Stat: signal transduction activator of transcription

Regulation of Foxp3 expression. Although obviously important to the understanding of Treg development, the factors that regulate Foxp3 expression remain elusive. The primary issue that has confounded these studies is that the factors that appear to affect Foxp3 expression also affect the survival and expansion of Tregs. However, some signaling pathways, including CD28, IL-2, and TGF-β, are emerging that appear to have an affect on the expression of Foxp3. For example, recent work has shown that CD28-mediated signals during thymic development are required for proper Foxp3 expression (29). However, other studies have implicated CD28 in Treg survival and expansion (30, 31), complicating an interpretation of the role of CD28 in Foxp3 expression. Similarly, mice lacking the IL-2 pathway owing to targeted mutations also display a deﬁcit in CD4+ CD25+ Tregs. This is true for mice with mutations in il-2, il-2rα, and il-2rβ, as well as for mice lacking the downstream mediator of IL-2 signaling, Stat5 (32–40). However, similar to the situation with CD28, distinguishing between a direct affect on transcription and survival is not possible. Recent studies have shown that IL-2 is absolutely required for Treg expansion in the periphery (41, 42). Furthermore, Malek et al. (39, 40) showed that thymus-restricted expression of IL-2Rβ-chain was sufﬁcient to rescue il-2rb−/− mice from fatal autoimmune disease, demonstrating a need for IL-2 signals during thymic development of Tregs. Finally, using the GFP-Foxp3 strain, Fontenot et al. (21) have shown that thymic expression of Foxp3 www.annualreviews.org • FOXP 3

213

ARI

13 February 2006

13:5

is absolutely dependent on TCR-MHC interactions and independent of commitment to either the CD4 or CD8 lineage. A more deﬁnitive assessment of the factors that control Foxp3 expression await an analysis of its cis-regulatory region. The availability of the knockin strains described above, as well as a characterization of the Foxp3 cis-regulatory region, will be invaluable in sorting out direct and indirect effects of these and other factors on Foxp3 expression. Perhaps the most controversial aspect of Foxp3 gene regulation is the role of TGF-β. Although TGF-β can have effects on Treg expansion and survival (43, 44), its actual role is not at all clear. Several groups have shown that in vitro culture of CD4+ CD25− T cells with a cocktail that includes TGF-β and IL2, in combination with TCR engagement and costimulation through CD28, can lead to induction of Foxp3 expression and acquisition of repressor activity (45–47). However, there are several issues with these experiments. The cultures used superphysiological concentrations of TGF-β, and investigators have shown that TGF-β can inhibit the proliferative affects of IL-2, at least in conventional T cells (48). Thus, one explanation for the in vitro results is that TGF-β inhibits the proliferation of CD4+ CD25− Foxp3− T cells in the cultures, allowing the expansion of contaminating Foxp3+ cells and an apparent increase in Foxp3 mRNA. Also, the culture systems used are very likely to produce a complex combination of signals in responding cells, making a deﬁnitive role for any of the signaling molecules difﬁcult to discern. In support of an indirect role for TGF-β in Foxp3 expression and function, Tregs from mice either lacking TGF-β1 or expressing a dominant-negative TGF-βRII are indistinguishable from their counterparts from wild-type mice in in vitro suppression assays (49). The data on the in vivo role of TGFβ are equally contradictory. In support of a role for TGF-β in Treg function and Foxp3 expression, several groups have shown that TGF-β is capable of expanding the in vivo

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

214

Ziegler

pool of Tregs (18). For example, T cell– speciﬁc expression of a dominant-active form of TGF-β leads to an elevation in the number of CD4+ CD25+ Tregs and increased Foxp3 expression (50, 51). However, these studies do not differentiate between expansion and de novo generation of Tregs. Using mice with an inducible TGF-β transgene, Peng et al. (44) showed that transient TGF-β expression speciﬁcally in the pancreatic islets was sufﬁcient to protect against onset of diabetes. They found that this transient expression increased the number of CD4+ T cells with a regulatory phenotype in the islets, consistent with a role for TGF-β in the expansion of Tregs in vivo. Studies using mice incapable either of producing TGF-β or of responding to it provide the most contradictory data. These mice have been used to demonstrate that TGF-β is either absolutely required for Treg expansion and function (50, 51) or completely dispensable (49). In addition, recent work has suggested that the effector T cell needs to be TGF-β responsive in order to be suppressed, at least in vivo (52). However, this appears not to be the case in vitro (49; K. Newton & S.F. Ziegler, manuscript in preparation). The recent development of Foxp3 reporter mice should help clarify this issue. In fact, using mice with a reporter for Foxp3 mRNA, Wan & Flavell (22) showed that TGF-β may directly inﬂuence Foxp3 expression.

FUNCTIONAL CHARACTERIZATION OF Foxp3 As described above, Foxp3 is a member of the forkhead/winged-helix family of transcriptional regulators. A common feature of members of this family is the FKH domain (Figure 1), which has been shown to be necessary and sufﬁcient for DNA binding (53, 54). DNA-binding analyses from a number of Fox family proteins have deﬁned a core DNA sequence (5 -A(A/T)TRTT(G/T)R-3 , where R = pyrimidine) surrounded by less conserved sequences (55). Experimentally, two

ANRV270-IY24-07

ARI

13 February 2006

13:5

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Figure 1 Schematic of human FOXP3 and FOXP3 with location of known IPEX mutations. Top schematic is FOXP3, with locations of missense mutations indicated by arrow. Bottom schematic is FOXP3, with location of mutations predicted to affect splicing or mRNA stability (red arrows) and to generated frameshifts (blue arrows) indicated. Exons are color coded according to the protein schematic, with pale blue indicating coding region of unknown function (FKH: forkhead).

canonical FKH-binding sequences [from the transthyretin promoter (TTR-S) and the immunoglobulin variable regions V1 promoter (V1P) (56)] have been used as templates to analyze the DNA-binding properties of Fox proteins. Similar to other members of the Fox family, Foxp3 can bind to each of these FKHbinding sites (57). Thus, as predicted from the sequence analysis, Foxp3 is a DNA-binding protein capable of speciﬁc binding to a canonical FKH DNA-binding site. Members of the Fox family are both transcriptional activators and transcriptional repressors. To assess the role of Foxp3 in transcriptional regulation, a reporter plasmid was constructed that contained three copies of the V1P FKH site linked to a minimal SV40 promoter and the ﬁreﬂy luciferase gene. Cotransfection of this reporter plasmid with a Foxp3 expression plasmid into Cos-7 cells resulted in a dramatic reduction of luciferase activity, suggesting that Foxp3 acts as a transcriptional repressor (57). The ability of Foxp3 to act as a transcriptional repressor required the presence of the FKH domain of Foxp3 and the V1P-binding site, showing that binding of Foxp3 to the FKH site is responsible for the observed effect of Foxp3 on the expression of the reporter. Several lines of evidence suggest that a major target of Foxp3-mediated transcriptional regulation is cytokine genes. For ex-

Other Forkhead Family Proteins and Immune Regulation In addition to Foxp3, several other members of the forkhead family have been implicated in regulating immune system development and function (90). For example, mutations in the Foxn1 gene are responsible for the phenotype seen in nude (nu) mice, rats, and humans (91, 92). These mice are characterized by abnormal development of the epidermis, lack of hair, and absence of a thymus. Mutations in Foxj1 lead to embryonic lethality, but fetal liver chimeras have been used to study its role in the immune system. These mice displayed multisystemic cellular autoimmunity, as well as T cell hyperproliferation and hyperactivity. CD4+ T cells from these animals were skewed toward a Th1 cytokine proﬁle, could be activated solely by IL-2 (CD3 and CD28 stimulation were not required), and exhibited decreased levels of the IκKβ regulatory subunit (93). These results are consistent with a role for Foxj1 in the negative regulation of NF-κB activity. Finally, Foxo3a has been implicated in cellular survival, where it is involved in the regulation of NF-κB (94, 95). Upon Aktdependent phosphorylation, Foxo3a is shuttled to cytoplasm and rendered inactive as a transcriptional regulator. Mice deﬁcient in Foxo3a develop T cell hyperactivity and multiorgan lymphocytic inﬁltrates with age. ample, ectopic expression of Foxp3 in Jurkat cells resulted in a marked reduction of IL2 production following stimulation (57). In addition, CD4+ T cells from mice expressing www.annualreviews.org • FOXP 3

215

ARI

13 February 2006

13:5

a Foxp3 transgene were unable to produce IL2, IL-4, or IFN-γ following TCR-mediated stimulation in vitro and had severely reduced ability to express cytokines in vivo following immunization (13, 14). Taken as whole, these data demonstrate a role for Foxp3 in the negative regulation of cytokine gene expression in CD4+ T cells. Inspection of composite NFAT/AP-1 sites upstream of several cytokine genes revealed a possible explanation for the ability of Foxp3 to regulate these genes. Either overlapping or directly adjacent to these DNA-binding sites was a potential FKH-binding site. Provocatively, oligonucleotide probes containing the composite sites from the human IL-2 and GM-CSF genes were capable of competing with the V1P site for Foxp3 binding in a gel shift assay (57). Conﬁrmation of the ability of Foxp3 to repress NFAT-mediated transcription comes from experiments using a reporter plasmid consisting of the –280 composite NFAT/AP-1 site from the mouse IL2 gene. NFAT activation was inhibited when the Foxp3 expression vector was cotransfected (57). Furthermore, work from our laboratory, using a bipartite reporter containing both NFAT and GAL4 DNA-binding sites, has shown that Foxp3 can inhibit NFAT-mediated transcription while binding to a more distant DNA site ( J. Lopes, T. Torgerson, L.A. Schubert, H. Ochs, and S.F. Ziegler, manuscript submitted). These experiments took advantage of fusion proteins consisting of the yeast GAL4 DNA-binding domain fused to the non-FKH sequences from FOXP3, which was found to be capable of inhibiting activation-induced expression of the NFAT reporter. Much effort is currently being used to deﬁne functional domains within FOXP3. As shown in Figure 1, FOXP3 contains at least three distinct structural domains: FKH, Leucine zipper, and C2H2 zinc ﬁnger. As described above, the FKH domain is critical for both DNA binding and for nuclear localization. Conﬁrmation of its role in nuclear localization comes from site-directed mutation of

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

216

Ziegler

two lysine residues (K415 and K416), at the carboxy end of the FKH domain, to glutamic acid. This mutant form of FOXP3, when expressed in cell lines, is localized to the cytoplasm ( J. Lopes, T. Torgerson, L.A. Schubert, H. Ochs, and S.F. Ziegler, manuscript submitted). The functional properties of the remainder of the protein are not well understood. However, studies on other members of the FOXP family may be relevant to the analysis of FOXP3 function. For example, the Leucine zipper domain of Foxp1 and Foxp2 is critical for homo- and heterodimer formation (58, 59). Additionally, deletion of the Leucine zipper domain from Foxp1 and Foxp2 abrogated their ability to act as transcriptional repressors. This domain of FOXP3 is likely also to be involved in dimerization (see below). The amino terminal half of FOXP3 contains no obvious identiﬁable functional domains, although it does contain a fairly high proportion of proline residues (Figure 1). However, we have shown that a fusion protein containing the DNA-binding domain of the yeast transcription factor GAL4 and the amino terminal half of human FOXP3 (amino acids 1–198) was functional as a transcriptional repressor, whereas a fusion of the zinc ﬁnger and Leucine zipper domains was nonfunctional (the GAL4 DNA-binding domains also contain nuclear localization and dimerization sequences). Using infection of primary T cells with FOXP3 deletion mutants, Bettelli et al. (11) showed that there were two functional domains within the amino terminal 200 amino acids of Foxp3, one between amino acids 1 and 150, and one between amino acids 150 and 200. Our lab has also deﬁned two functional domains with the amino terminus of Foxp3. One domain, between amino acids 67 and 132, is involved in general transcriptional repression by FOXP3. The second, between amino acids 135 and 198, is speciﬁcally required for repression of NFAT-mediated transcription. Bettelli et al. (11) also showed that Foxp3 can directly interact with NFAT

ANRV270-IY24-07

ARI

13 February 2006

13:5

and the NF-κB subunit P65. Although the region of interaction with Foxp3 was not identiﬁed, that data described above would predict that the region between amino acids 135 and 198 is involved.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

HUMAN FOXP3: NOT QUITE THE SAME AS MOUSE The identiﬁcation of Foxp3 as the gene mutated in scurfy mice led to an investigation of its possible role in a human syndrome with clinical features similar to those seen in scurfy animals. This syndrome, known as IPEX (immune dysfunction/polyendocrinopathy/ enteropathy/X-linked), was ﬁrst characterized in 1982 and subsequently mapped to the X chromosome (60, 61). It is characterized by watery, sometimes bloody, diarrhea in very early infancy. In addition, nearly all affected males develop dermatitis (usually eczematous in nature), and most cases develop autoimmune endocrinopathies, with type 1 diabetes and thyroiditis the most common. In general, affected individuals develop symptoms very early in infancy and usually die in the ﬁrst two years of life (61–64). The similarity of the phenotypes of scurfy mice and IPEX humans suggested a common cause for the two diseases. Indeed, mutations in FOXP3 have been identiﬁed in over 20 families with affected offspring (62, 65–69). Many of these are missense mutations in the FKH domain, but mutations have been found across the length of the gene (Figure 1). However, not all patients diagnosed with IPEX have FOXP3 mutations. Sequencing of the FOXP3 gene in a large cohort of individuals diagnosed with IPEX revealed identiﬁable coding region mutations in 60%, with an additional 10% having dramatically reduced mRNA levels (65). One interpretation of these data is that a mutation in FOXP3 leads to development of IPEX, and those IPEX patients without FOXP3 mutations may be mutations in genes whose products interact with FOXP3, either physically or functionally. Indeed, a recent study has found a patient with IPEX-like

symptoms for whom FOXP3 was ruled out as a candidate gene (66). As shown in Figure 1, FOXP3 mutations in IPEX patients have been identiﬁed throughout the coding region. We have begun to analyze the affect of these IPEX mutations on FOXP3 function. Accordingly, mutations in the FKH domain ablate the ability of FOXP3 to inhibit transcription of a reporter construct containing canonical forkhead-binding sites. Consistent with the prevailing view of the FKH domain, mutations in FOXP3 that are predicted to affect DNA binding fail to inhibit transcription of a FKH-dependent reporter ( J. Lopes, T. Torgerson, L. Schubert, H. Ochs, and S.F. Ziegler, manuscript submitted). Interestingly, two IPEX patients were shown to have single codon mutations in the Leucine zipper domain of FOXP3(K250 and E251) (70, 71). We have now found that introducing the E251 mutation into FOXP3 results in a failure to homodimerize and in reduced ability to repress transcription ( J. Lopes, T. Torgerson, L.A. Schubert, H. Ochs, and S.F. Ziegler, manuscript submitted). These data strongly suggest that the Leucine zipper domain of FOXP3 is critical for proper function. Additional mutations have been identiﬁed in the amino terminal domain identiﬁed as critical for FOXP3 function (see previous section); these mutations are now being analyzed to determine their impact on FOXP3 transcriptional repression.

IPEX: immune dysfunction/polyendocrinopathy/ enteropathy/ X-linked

Human FOXP3: Regulation of Expression and Multiple Isoforms A population of CD4+ CD25+ T cells has also been identiﬁed and characterized from human peripheral blood. These cells, when assayed in vitro, are anergic, have the ability to suppress the proliferation and cytokine production of cocultured CD4+ CD25− T cells, and express FOXP3 (18, 72–75). Thus, as has been demonstrated in the mouse, human CD4+ CD25+ T cells express FOXP3 and act as suppressors. www.annualreviews.org • FOXP 3

217

ARI

13 February 2006

13:5

However, further examination of FOXP3 expression showed several differences between humans and mice. When FOXP3 expression was monitored by Western blotting, it became apparent that the protein actually ran as a closely spaced doublet (18). RT-PCR analysis has shown that the upper isoform represents the ortholog to mouse Foxp3, whereas the lower isoform is encoded from an mRNA lacking exon 2 (amino acids 71–105) (76). This splicing variant has not been reported in mouse CD4+ CD25+ Tregs. Whether the same T cell expresses both isoforms simultaneously is currently not known, although all sources of FOXP3 in humans have both isoforms. In addition, it is unclear if there is a functional difference between the two isoforms. However, expression of the exon2 isoform in CD4+ CD25− FOXP3− human T cells leads to functional anergy, but not to the same extent as expression of fulllength FOXP3 (76). Human T cells expressing FOXP3exon2 have intermediate proliferative responses to TCR stimulation and make slightly more IL-2 than cells expressing only the full length. Interestingly, the sequences encoded by exon 2 fall within one of the functional domains of FOXP3 deﬁned above. Thus, it is possible that the lower isoform represents a natural dominant-negative form of FOXP3. Our lab has recently shown that FOXP3 and the orphan retinoic acid nuclear receptor RORα interact and that the region of FOXP3 involved in the interaction is encoded by exon 2 ( J. Du & S.F. Ziegler, manuscript in preparation). We are currently characterizing the molecular basis for this interaction. Another clue that the regulation of Foxp3 expression was not conserved between humans and mice comes from an analysis of FOXP3 expression in stimulated human CD4+ CD25− T cells. The starting population of T cells lacked detectable FOXP3 expression, but within 24 h of stimulation antiCD3+ CD28 FOXP3 expression was detected, peaking at 72 h (18). Similar experiments using mouse CD4+ CD25− T cells did not

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

218

Ziegler

demonstrate upregulation of Foxp3 (15, 16). Further analysis with human CD4+ T cells showed that the subset of T cells that upregulated CD25 also expressed FOXP3 (18). Thus, in humans, FOXP3 was behaving like an activation-induced gene in CD4+ cells. These initial ﬁndings have been conﬁrmed by other groups (76–78) and provide the ﬁrst clue that human and mouse Foxp3 are regulated differently. Consistent with the idea that FOXP3 expression in human CD4+ T cells is linked to TCR-mediated stimulation, CD4+ human T cell clones are also FOXP3+ (77–79; J.H. Buckner & S.F. Ziegler, unpublished observations). Mouse T cell clones have not been found to express Foxp3. As described earlier, ectopic expression of Foxp3 in conventional mouse CD4+ T cells converted those cells to Tregs (15, 16). Several groups have now attempted similar experiments with human FOXP3 using either retroviral or Lentiviral transduction of CD4+ CD25− T cells. As is seen in the mouse system, the transduced cells display anergy when stimulated through the TCR and are capable of suppressing responder CD4+ T cells (76, 80, 81). However, the level of suppression is not at the levels seen in the mouse experiments. Also, these experiments are somewhat confounded by the fact that the CD4+ T cells are stimulated prior to infection, which induces the expression of endogenous FOXP3 (18, 76, 77). As described above, stimulation of human CD4+ CD25− T cells leads to FOXP3 expression. An obvious question is what is the functional outcome for the CD4+ T cell that induces FOXP3 expression? Recent work from our group has shown that these cells develop the capability to act as Tregs (18). When the CD25+ cells are isolated following the 10-day culture period, they have the ability to suppress the proliferation of freshly isolated responder T cells. One possible explanation for this phenomenon is that a preexisting pool of CD4+ CD25− FOXP3+ cells preferentially expands in the cultures. However, this does not appear to be the case.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

ARI

13 February 2006

13:5

There is no detectable FOXP3 expression in the starting population of cells. In addition, when the CD4+ CD25− cells, isolated after the 10-day culture period, are restimulated, the same ratio of CD25+ FOXP3+ / CD25− FOXP3− cells results (M.R. Walker, P. Putheti, J.H. Buckner, and S.F. Ziegler, unpublished data). Further support comes from the ﬁnding that naive CD4+ T cells (CD45RA+ ) and CD4+ CD25− T cells from cord blood are also capable of inducing FOXP3 and becoming Tregs, suggesting that antigen experience is not required for this process (82). Taken as a whole, these data support a model in which the ability of a CD4+ T cell to induce FOXP3 expression is not a developmentally determined event. Rather, it is dependent on the conditions present at the time of antigen stimulation of a given T cell. The underlying molecular mechanism that governs this development remains to be determined. Another piece of evidence regarding the ability to generate Tregs in vitro using human CD4+ T cells comes from studies using speciﬁc antigens, rather than antibody crosslinking, as stimulus. For example, Walker et al. (82) showed that coculture of CD4+ CD25− T cells with allo-antigen-presenting cells (APCs) resulted in the development of Tregs speciﬁc for the stimulating allo-APCs. Although these experiments support a model of de novo Treg generation, the actual stimulating antigen is not known. However, they do demonstrate the speciﬁcity of the in vitro– generated Tregs in that they were unable to suppress responses stimulated by third-party APCs. The ﬁnal piece of evidence that induction of FOXP3 expression and de novo Treg generation can occur following antigen stimulation comes from studies using human MHC class II tetramers to isolate antigen-speciﬁc cells following in vitro stimulation. CD4+ CD25− T cells from DR ∗ 0401 (DR4) individuals were stimulated with the immunodominant peptide from inﬂuenza virus, HA(306–319), in the presence of ir-

radiated APCs. After 10 days of culture, tetramer+ CD25+ cells were isolated from the cultures. These cells expressed FOXP3 and inhibited the subsequent proliferation of freshly isolated responder CD4+ T cells from the same donor (82). However, these HA(306–319)+ CD25+ FOXP3+ T cells did not affect the proliferation of T cells when stimulated with an unrelated antigen, tetanus toxoid, unless they also were given the HA peptide. Thus, they required cognate antigen for activation, but once activated they were capable of suppressing responses in an antigennonspeciﬁc manner. The data outlined above suggest that there are at least two distinct populations of Tregs in humans. One population is generated in the thymus, is self-reactive, and is involved in protection from autoimmune responses. These Tregs are referred to as natural Tregs and would be analogous to those seen in the mouse that arise during thymic selection (83). The role of this population of Tregs is to protect against self-reactive effector T cells in the periphery. These cells comprise the CD4+ CD25+ T cell population in human peripheral blood and are equivalent to those Tregs described in the mouse. This population of Tregs is also likely to be most affected by the mutations in FOXP3 seen in IPEX patients. These mutations can be expected to result either in the development of self-reactive T cells that lack suppressor activity or in the complete lack of this population altogether. In either instance, the result would be a failure in the ability to control self-reactive effector T cells. In this respect, the IPEX patients resemble the scurfy mice. A second population of Tregs is represented by CD4+ CD25+ FOXP3+ cells that are generated upon in vitro stimulation of CD4+ CD25− FOXP3− cells. By all criteria measured, these cells are indistinguishable from natural Tregs (18, 82; J.H. Buckner & S.F. Ziegler, unpublished data). As described above, they can be generated following polyclonal or antigen-speciﬁc stimulation. Also, similar to natural Tregs, their ability to www.annualreviews.org • FOXP 3

219

ARI

13 February 2006

13:5

suppress is IL-10 and TGF-β independent, but is contact dependent (18). In addition, transcript array analysis has shown that the same set of genes regulated in natural Tregs appears to be similarly regulated in the in vitro–generated cells (P. Putheti & S.F. Ziegler, unpublished data). Taken together, the data demonstrate that CD4+ T cells with the same functional properties as thymically derived Tregs can be generated from naive conventional CD4+ T cells following TCR stimulation. Although the in vivo function of these cells is not known, we propose that they play a role in controlling immune responses, possibly serving to limit bystander activation at the site of the response. In addition, there are reports that CD4+ CD25+ Tregs can induce infectious tolerance through the local generation of classes of regulatory T cells (e.g., Tr1 cells) (84), further extending this paradigm. An unanswered question at this point is the fate of the induced Tregs. Because these cells appear to respond poorly to proliferative signals and are impaired in cytokine production, it seems likely that they will not persist once the response has ended and antigen is no longer available.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

FOXP3 AND THERAPEUTIC INTERVENTION FOR AUTOIMMUNITY As described above, following retroviralmediated introduction of Foxp3, conventional CD4+ T cells acquire a regulatory-like phenotype and are capable of suppressing immune responses both in vitro and in vivo (15–17). The in vivo experiments involved cotransfer of Foxp3-transduced T cells with pathogenic CD4+ T cells in a model of colitis. These studies suggest the possibility of using ectopic Foxp3 expression to convert T cells to take on a regulatory phenotype and then using these cells in adoptive cellular immunotherapy. Coupled with the ability to identify and isolate antigen-speciﬁc human CD4+ T cells using MHC class II tetramers (85–87), ex vivo generation of antigen-speciﬁc 220

Ziegler

Tregs has great promise. Indeed, some recent reports have suggested that this approach may be viable. For example, Jaeckel et al. (27) used Foxp3-transduced T cells from BDC2.5 TCR transgenic mice to treat NOD mice with recent onset diabetes. They found that the antigen-speciﬁc BDC2.5 T cells, expressing Foxp3, were capable of reversing disease. However, they also showed that transfer of Foxp3-transduced polyclonal T cells from NOD mice failed to rescue. These data suggest that the critical feature for this type of therapy may be getting enough antigenspeciﬁc Tregs to the correct site, and that the transduced polyclonal T cells did not contain sufﬁcient islet-antigen-speciﬁc cells to have a therapeutic effect. In contrast to that study, Loser et al. (88) have shown that Foxp3-transduced polyclonal naive T cells, when transferred to sensitized hosts, could inhibit contact hypersensitivity responses. These polyclonal T cells were also used to treat mice expressing a keratin 14-CD40L transgene. These mice develop chronic skin inﬂammation and systemic autoimmunity (89). Transfer of Foxp3transduced CD4+ T cells into these animals protected against both the local skin inﬂammation and the systemic autoimmunity (88). Transferring these types of therapies to humans presents unique problems. As described above, stimulation of human CD4+ CD25− T cells induces FOXP3 expression, and most methods for introducing foreign genes into T cells using viral vectors require some level of stimulation. Thus, the very cells one may want to transduce and convert to Tregs may be less likely to be infected owing to expression of FOXP3 and a lowered overall activation level. However, the work from Walker et al. (82) suggests that deriving antigen-speciﬁc CD4+ T cells in vitro with self-antigens, followed by isolation of CD25+ that are also MHC class II tetramer+ , may be a solution to this problem. The obvious limitation to this approach is the availability of the appropriate MHC class II tetramer for a given HLA haplotype and speciﬁc epitope.

ANRV270-IY24-07

ARI

13 February 2006

13:5

In summary, the reemergence of suppressor T cells and the discovery of FOXP3 as a critical player in their development and function hold great promise for the treatment of autoimmune diseases in humans. The differences between humans and mice may present

obstacles in the transfer from mouse models to actual human disease. Thus, a more concerted effort to develop better human in vitro and in vivo systems is required to understand better the potential therapeutic beneﬁts of manipulating FOXP3.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ACKNOWLEDGMENTS The author thanks Drs. Megan K. Levings (University of British Columbia) and Troy R. Torgerson (University of Washington) for sharing information and data prior to publication, and Drs. Jane H. Buckner, Daniel J. Campbell, and Gerald T. Nepom (Benaroya Research Institute) for critical reading of the manuscript prior to submission. The author is supported by grants from the NIH (AI48779, AI059926, DK068312), American Diabetes Association, Juvenile Diabetes Research Foundation, and DANA Foundation.

LITERATURE CITED 1. Baecher-Allan C, Viglietta V, Haﬂer DA. 2004. Human CD4+ CD25+ regulatory T cells. Semin. Immunol. 16:89–98 2. Piccirillo CA, Thornton AM. 2004. Cornerstone of peripheral tolerance: naturally occurring CD4+ CD25+ regulatory T cells. Trends Immunol. 25:374–80 3. Shevach EM. 2002. CD4+ CD25+ suppressor T cells: more questions than answers. Nat. Rev. Immunol. 2:389–400 4. Sakaguchi S. 2005. Naturally arising Foxp3-expressing CD25+ CD4+ regulatory T cells in immunological tolerance to self and non-self. Nat. Immunol. 6:345–52 5. Fontenot JD, Rudensky AY. 2005. A well adapted regulatory contrivance: regulatory T cell development and the forkhead family transcription factor Foxp3. Nat. Immunol. 6:331–37 6. Godfrey V, Rouse BT, Wilkinson JE. 1994. Transplantation of T cell-mediated lymphoreticular disease from the scurfy (sf ) mouse. Am. J. Pathol. 145:281–86 7. Waterhouse P, Penninger JM, Timms E, Wakeham A, Shahinian A, et al. 1995. Lymphoproliferative disorders with early lethality in mice deﬁcient in Ctla-4. Science 270:985–88 8. Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA, Sharpe AH. 1995. Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3:541–47 9. Kanangat S, Blair P, Reddy R, Deheshia M, Godfrey V, et al. 1996. Disease in the scurfy (sf ) mouse is associated with overexpression of cytokine genes. Eur. J. Immunol. 26:161– 65 10. Clark LB, Appleby MW, Brunkow ME, Wilkinson JE, Ziegler SF, Ramsdell F. 1999. Cellular and molecular characterization of the scurfy mouse mutant. J. Immunol. 162:2546–54 11. Bettelli E, Dastrange M, Oukka M. 2005. Foxp3 interacts with nuclear factor of activated T cells and NF-κB to repress cytokine gene expression and effector functions of T helper cells. Proc. Natl. Acad. Sci. USA 102:5138–43 12. Brunkow ME, Jeffery EW, Hjerrild KA, Paeper B, Clark LB, et al. 2001. Disruption of a new forkhead/winged-helix protein, scurfin, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat. Genet. 27:68–73 13. Khattri R, Kasprowicz DJ, Cox T, Yasayko S-A, Ziegler SF, Ramsdell F. 2001. The www.annualreviews.org • FOXP 3

12. This paper describes the identification of Foxp3 as the gene mutated in scurfy mice and describes the transgenic animals used to study Foxp3’s role in CD4 T cell function.

221

ANRV270-IY24-07

ARI

13 February 2006

13a.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

14. 15, 16, and 17. These seminal papers demonstrate conclusively Foxp3’s critical role in Treg development and function. 18. This was the first paper to show that, in humans, FOXP3 expression can be induced by TCR stimulation, that cells with regulatory capability can be generated de novo, and that regulation of human and mouse Foxp3 have key differences. 19. This paper “rediscovered” Tregs as a distinct population of cells. Much subsequent work on Tregs and control of peripheral tolerance derives from the work described here. 21 and 22. These two papers describe the Foxp3 reporter mice that are vital to the study of Tregs as, in the mouse, Foxp3 is the only marker that appears to be exclusive to these cells.

222

15. 16. 17. 18.

19.

20.

21.

22. 23.

24. 25.

26.

27. 28. 29.

Ziegler

13:5

amount of scurﬁn protein determines peripheral T cell number and responsiveness. J. Immunol. 167:6312–20 Kasprowicz DJ, Droin N, Ramsdell F, Green DR, Ziegler SF. 2006. Elevated FoxP3 expression leads to activation-induced cell death, while effector cell differentiation correlates with loss of FoxP3 expression. Eur J. Immunol. In press Kasprowicz DJ, Smallwood PS, Tyznik AJ, Ziegler SF. 2003. Scurﬁn (FoxP3) controls T-dependent immune responses in vivo through regulation of CD4+ T cell effector function. J. Immunol. 171:1216–23 Fontenot JD, Gavin MA, Rudensky AY. 2003. FoxP3 programs the development and function of CD4+ CD25+ regulatory T cells. Nat. Immunol. 4:330–36 Hori S, Nomura T, Sakaguchi S. 2003. Control of regulatory T cell development by the transcription factor FoxP3. Science 299:1057–61 Khattri R, Cox T, Yasayko S-A, Ramsdell F. 2003. An essential role for Scurfin in CD4+ CD25+ T regulatory cells. Nat. Immunol. 4:337–42 Walker MR, Kasprowicz DJ, Gersuk VH, Benard A, Van Landeghen M, et al. 2003. Induction of FoxP3 and acquisition of T regulatory activity by stimulated human CD4+ CD25− T cells. J. Clin. Invest. 112:1437–43 Sakaguchi S, Sakaguchi N, Asano M, Itoh M, Toda M. 1995. Immunologic selftolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J. Immunol. 155:1151–64 Suri-Payer E, Amar AZ, Thornton AM, Shevach EM. 1998. CD4+ CD25+ T cells inhibit both the induction and effector function of autoreactive T cells and represent a unique lineage of immunoregulatory cells. J. Immunol. 160:1212–18 Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, Rudensky AY. 2005. Regulatory T cell lineage specification by the forkhead transcription factor foxp3. Immunity 22:329–41 Wan YY, Flavell RA. 2005. Identifying Foxp3-expressing suppressor T cells with a bicistronic reporter. Proc. Natl. Acad. Sci. USA 102:5126–31 Kohm AP, Carpentier PA, Anger HA, Miller SD. 2002. Cutting edge: CD4+ CD25+ regulatory T cells suppress antigen-speciﬁc autoreactive immune responses and central nervous system inﬂammation during active experimental autoimmune encephalomyelitis. J. Immunol. 169:4712–16 Wraith DC, Nicolson KS, Whitely NT. 2004. Regulatory CD4+ T cells and the control of autoimmune disease. Curr. Opin. Immunol. 16:695–701 Tang Q, Henriksen KJ, Bi M, Finger EB, Szot G, et al. 2004. In vitro-expanded antigen-speciﬁc regulatory T cells suppress autoimmune diabetes. J. Exp. Med. 199:1455– 65 Chai JG, Xue SA, Coe D, Addey C, Bartok I, et al. 2005. Regulatory T cells, derived from naive CD4+ CD25− T cells by in vitro Foxp3 gene transfer, can induce transplantation tolerance. Transplantation 79:1310–16 Jaeckel E, von Boehmer H, Manns MP. 2005. Antigen-speciﬁc FoxP3-transduced T-cells can control established type 1 diabetes. Diabetes 54:306–10 Lundsgaard D, Holm TL, Hornum L, Markholst H. 2005. In vivo control of diabetogenic T-cells by regulatory CD4+ CD25+ T-cells expressing Foxp3. Diabetes 54:1040–47 Tai X, Cowan M, Feigenbaum L, Singer A. 2005. CD28 costimulation of developing thymocytes induces Foxp3 expression and regulatory T cell differentiation independently of interleukin 2. Nat. Immunol. 6:152–62

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

ARI

13 February 2006

13:5

30. Tang Q, Henriksen KJ, Boden EK, Tooley AJ, Ye J, et al. 2003. Cutting edge: CD28 controls peripheral homeostasis of CD4+ CD28+ regulatory T cells. J. Immunol. 171:3348–52 31. Boden E, Tang Q, Bour-Jordan H, Bluestone JA. 2003. The role of CD28 and CTLA4 in the function and homeostasis of CD4+ CD25+ regulatory T cells. Novartis Found. Symp. 252:55–63 ¨ ¨ 32. Scheffold A, Huhn J, Hofer T. 2005. Regulation of CD4+ CD25+ regulatory T cell activity: it takes (IL-)two to tango. Eur. J. Immunol. 35:1336–41 33. Nelson BH. 2004. IL-2, regulatory T cells, and tolerance. J. Immunol. 172:3983–88 34. Malek TR, Bayer AL. 2004. Tolerance, no immunity, critically depends on IL-2. Nat. Rev. Immunol. 4:665–74 35. Nelson BH. 2002. Interleukin-2 signaling and the maintenance of self-tolerance. Curr. Dir. Autoimmun. 5:92–112 36. Schimpl A, Berberich I, Kneitz B, Kramer S, Santner-Nanan B, et al. 2002. IL-2 and autoimmune disease. Cytokine Growth Factor Rev. 13:369–78 37. Snow JW, Abraham N, Ma MC, Herndier BG, Pastuszak AW, Goldsmith MA. 2003. Loss of tolerance and autoimmunity affecting multiple organs in STAT5A/5B-deﬁcient mice. J. Immunol. 171:5042–50 38. Burchill MA, Goetz CA, Prlic M, O’Neil JJ, Harmon IR, et al. 2003. Distinct effects of STAT5 activation on CD4+ and CD8+ T cell homeostasis: development of CD4+ CD25+ regulatory T cells versus CD8+ memory T cells. J. Immunol. 171:5853–64 39. Malek TR, Porter BO, Codias EK, Scibelli P, Yu A. 2000. Normal lymphoid homeostasis and lack of lethal autoimmunity in mice containing mature T cells with severely impaired IL-2 receptors. J. Immunol. 164:2905–14 40. Malek TR, Yu A, Vincek V, Scibelli P, Kong L. 2002. CD4 regulatory T cells prevent lethal autoimmunity in IL-2Rβ-deﬁcient mice: implications for the nonredundant function of IL-2. Immunity 17:167–78 41. Bayer AL, Yu A, Adeegbe D, Malek TR. 2005. Essential role for interleukin-2 for CD4+ CD25+ T regulatory cell development during the neonatal period. J. Exp. Med. 201:769–77 42. Setoguchi R, Hori S, Takahashi T, Sakaguchi S. 2005. Homeostatic maintenance of natural Foxp3+ CD25+ CD4+ regulatory T cells by interleukin (IL)-2 and induction of autoimmune disease by IL-2 neutralization. J. Exp. Med. 201:723–35 43. Marie JC, Letterio JJ, Gavin M, Rudensky AY. 2005. TGF-β1 maintains suppressor function and Foxp3 expression in CD4+ CD25+ regulatory T cells. J. Exp. Med. 201:1061– 67 44. Peng Y, Laouar Y, Li MO, Green EA, Flavell RA. 2004. TGF-β regulates in vivo expansion of Foxp3-expressing CD4+ CD25+ regulatory T cells responsible for protection against diabetes. Proc. Natl. Acad. Sci. USA 101:4572–77 45. Chen W, Jin W, Hardegen N, Lei KJ, Li L, et al. 2003. Conversion of peripheral CD4+ CD25− T cells to CD4+ CD25+ regulatory T cells by TGF-β induction of transcription factor FoxP3. J. Exp. Med. 198:1875–86 46. Fantini MC, Becker C, Monteleone G, Pallone F, Galle PR, Neurath MF. 2004. Cutting edge: TGF-β induces a regulatory phenotype in CD4+ CD25− T cells through FoxP3 induction and down-regulation of Smad7. J. Immunol. 172:5149–53 47. Fu S, Zhang N, Yopp AC, Chen D, Mao M, et al. 2004. TGF-β induces Foxp3 + Tregulatory cells from CD4+ CD25− precursors. Am. J. Transplant. 4:1614–27 48. Nelson BH, Martyak TP, Thompson LJ, Moon JJ, Wang T. 2003. Uncoupling of www.annualreviews.org • FOXP 3

223

ANRV270-IY24-07

ARI

13 February 2006

49.

50.

51.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

52.

53. 54. 55. 56. 57.

58.

59.

60. 61.

62.

63.

64.

65. 66. 224

Ziegler

13:5

promitogenic and antiapoptotic functions of IL-2 by Smad-dependent TGF-β signaling. J. Immunol. 170:5563–70 Piccirillo CA, Letterio JJ, Thornton AM, McHugh RS, Mamura M, et al. 2002. CD4+ CD25+ regulatory T cells can mediate suppressor function in the absence of transforming growth factor β1 production and responsiveness. J. Exp. Med. 196:237–46 Schramm C, Huber S, Protschka M, Czochra P, Burg J, et al. 2004. TGFβ regulates the CD4+ CD25+ T-cell pool and the expression of Foxp3 in vivo. Int. Immunol. 16:1241– 49 Huber S, Schramm C, Lehr HA, Mann A, Schmitt S, et al. 2004. Cutting edge: TGFβ signaling is required for the in vivo expansion and immunosuppressive capacity of regulatory CD4+ CD25+ T cells. J. Immunol. 173:6526–31 Fahlen L, Read S, Gorelik L, Hurst SD, Coffman R, et al. 2005. T cells that cannot respond to TGF-β escape control by CD4+ CD25+ regulatory T cells. J. Exp. Med. 201:737–46 ¨ Kaestner K, Knochel W, Martinez D. 2000. Uniﬁed nomenclature for the wingedhelix/forkhead transcription factors. Genes Dev. 14:142–46 ¨ Kaufmann E, Knochel W. 1996. Five years on the wings of fork head. Mech. Dev. 57:3–20 ¨ ¨ Kaufmann E, Muller D, Knochel W. 1995. DNA recognition site analysis of Xenopus winged helix proteins. J. Mol. Biol. 248:239–54 Li C, Tucker PW. 1993. DNA-binding properties and secondary structural model of the hepatocyte nuclear factor 3/fork head domain. Proc. Natl. Acad. Sci. USA 90:11583–87 Schubert LA, Jeffery EW, Zhang Y, Ramsdell F, Ziegler SF. 2001. Scurﬁn (FOXP3) acts as a repressor of transcription and regulates T cell activation. J. Biol. Chem. 276:37672– 79 Li S, Weidenfeld J, Morrisey EE. 2004. Transcriptional and DNA binding activity of the Foxp1/2/4 family is modulated by heterotypic and homotypic protein interactions. Mol. Cell. Biol. 24:809–22 Wang B, Lin D, Li C, Tucker PW. 2003. Multiple domains deﬁne the expression and regulatory properties of Foxp1 forkhead transcriptional repressors. J. Biol. Chem. 278:24259– 68 Powell BR, Buist N, Stenzel P. 1982. An X-linked syndrome of diarrhea, polyendocrinopathy, and fatal infection in infancy. J. Pediatr. 100:731–37 Bennett CL, Ochs HD. 2001. IPEX is a unique X-linked syndrome characterized by immune dysfunction, polyendocrinopathy, enteropathy, and a variety of autoimmune phenomena. Curr. Opin. Pediatr. 13:533–38 Gambineri E, Torgerson TR, Ochs HD. 2003. Immune dysregulation, polyendocrinopathy, enteropathy, and X-linked inheritance (IPEX), a syndrome of systemic autoimmunity caused by mutations of FOXP3, a critical regulator of T cell homeostasis. Curr. Opin. Rheumatol. 15:430–35 Torgerson TR, Ochs HD. 2002. Immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome: a model of immune dysregulation. Curr. Opin. Allergy Clin. Immunol. 2:481–87 Nieves DS, Phipps RP, Pollock SJ, Ochs HD, Zhu Q, et al. 2004. Dermatologic and immunologic ﬁndings in the immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome. Arch. Dermatol. 140:466–72 Ochs HD, Ziegler SF, Torgerson TR. 2005. FOXP3 acts as a rheostat of the immune response. Immunol. Rev. 203:156–64 Owen CJ, Jennings CE, Imrie H, Lachaux A, Bridges NA, et al. 2003. Mutational analysis

ANRV270-IY24-07

67.

68.

69.

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

70.

71.

72.

73.

74. 75.

76.

77.

78.

79. 80.

81.

82.

ARI

13 February 2006

13:5

of the FOXP3 gene and evidence for genetic heterogeneity in the immunodysregulation, polyendocrinopathy, enteropathy syndrome. J. Clin. Endocrinol. Metab. 88:6034–39 Kobayashi I, Shiari R, Yamada M, Kawamura N, Okano M, et al. 2001. Novel mutations of FOXP3 in two Japanese patients with immune dysregulation, polyendocrinopathy, enteropathy, X linked syndrome (IPEX). J. Med. Genet. 38:874–76 Wildin RS, Ramsdell F, Peake J, Faravelli F, Casanova J-L, et al. 2001. X-linked neonatal diabetes mellitus, enteropathy and endocrinopathy syndrome is the human equivalent of mouse scurfy. Nat. Genet. 27:18–20 Bennett CL, Christie J, Ramsdell F, Brunkow ME, Ferguson PF, et al. 2001. The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutation of FOXP3. Nat. Genet. 27:20–21 Chatila TA, Blaeser F, Ho N, Lederman HM, Voulgaropoulos C, et al. 2000. JM2, encoding a fork head-related protein, is mutated in X-linked autoimmunity-allergic disregulation syndrome. J. Clin. Invest. 106:R75–81 Wilden RS, Smyk-Pearson S, Filipovich AH. 2002. Clinical and molecular features of the immunodysregulation, polyendocrinopathy, enteropathy X linked (IPEX) syndrome. J. Med. Genet. 39:537–45 Jonuleit H, Schmitt E, Stassen M, Tuttenberg A, Knop J, Enk AH. 2001. Identiﬁcation and functional characterization of human CD4+ CD25+ T cells with regulatory properties isolated from peripheral blood. J. Exp. Med. 193:1285–94 Dieckmann D, Plottner H, Berchtold S, Berger T, Schuler G. 2001. Ex vivo isolation and characterization of CD4+ CD25+ T cells with regulatory properties from human blood. J. Exp. Med. 193:1303–10 Baecher-Allan C, Brown JA, Freeman GJ, Haﬂer DA. 2001. CD4+ CD25high regulatory cells in human peripheral blood. J. Immunol. 167:1245–53 Levings MK, Sangregorio R, Roncarolo MG. 2001. Human CD25+ CD4+ T regulatory cells suppress naive and memory T cell proliferation and can be expanded in vitro without loss of function. J. Exp. Med. 193:1295–302 Allan SE, Passerini L, Bacchetta R, Crellin N, Da M, et al. 2006. The role of FOXP3, and an isoform lacking exon 2, in the generation of human CD4+ T regulatory cells. J. Clin. Invest. In press Morgan ME, van Bilsen JH, Bakker AM, Heemskerk B, Schilham MW, et al. 2005. Expression of FOXP3 mRNA is not conﬁned to CD4+ CD25+ T regulatory cells in humans. Hum. Immunol. 66:13–20 Roncador G, Brown PJ, Maestre L, Hue S, Martinex-Torrecuadrada JL, et al. 2005. Analysis of FOXP3 protein expression in human CD4+ CD25+ regulatory T cells at the single-cell level. Eur. J. Immunol. 18:1681–91 Malone R, Kochik SA, Reijonen H, Carson BD, Ziegler SF, et al. 2006. Functional avidity directs T-cell fate in autoreactive CD4+ T-cells. Blood. In press Yagi H, Nomura T, Nakamura K, Yamazaki S, Kitawaki T, et al. 2004. Crucial role of FOXP3 in the development and function of human CD25+ CD4+ regulatory T cells. Int. Immunol. 16:1643–56 Oswald-Richter K, Grill SM, Shariat N, Leelawong M, Sundrud MS, et al. 2004. HIV infection of naturally occurring and genetically reprogrammed human regulatory T cells. PLoS Biol. 2:955–66 Walker MR, Carson BD, Nepom GT, Ziegler SF, Buckner JH. 2005. De novo generation of antigen-speciﬁc CD4+ CD25+ regulatory T cells from human CD4+ . Proc. Natl. Acad. Sci. USA 102:4103–8 www.annualreviews.org • FOXP 3

225

ARI

13 February 2006

13:5

83. Bluestone JA, Abbas AK. 2003. Natural versus adapted regulatory T cells. Nat. Rev. Immunol. 3:253–57 84. Moore KJ, de Waal MR, Coffman R, O’Garra A. 2001. Interleukin-10 and the interleukin10 receptor. Annu. Rev. Immunol. 19:683–765 85. Novak EJ, Masewics SA, Liu AW, Lernmark A, Kwok WW, Nepom GT. 2001. Activated human epitope-speciﬁc T cells identiﬁed by class II tetramers reside within a CD4high , proliferating subset. Int. Immunol. 13:799–806 86. Buckner JH, Holzer U, Novak EJ, Reijonen H, Kwok WW, Nepom GT. 2002. Deﬁning antigen-speciﬁc responses with human MHC class II tetramers. J. Allergy. Clin. Immunol. 110:199–208 87. Nepom GT, Buckner JH, Novak EJ, Reichstetter S, Reijonen H, et al. 2002. HLA class II tetramers. Tools for direct analysis of antigen-speciﬁc CD4+ T cells. Arthritis Rheum. 46:5–12 88. Loser K, Hansen W, Apelt J, Balkow S, Buer J, Beissert S. 2005. In vitro-generated regulatory T cells induced by FoxP3-retrovirus infection control murine contact allergy and systemic auroimmunity. Gene Ther. In press 89. Mehling A, Loser K, Varga G, Metze D, Luger TA, et al. 2001. Overexpression of CD40 ligand in murine epidermis results in chronic skin inﬂammation and systemic autoimmunity. J. Exp. Med. 194:615–28 90. Cofffer PJ, Burgering BM. 2004. Forkhead box transcription factors and their role in the immune system. Nat. Rev. Immunol. 4:889–99 91. Pignata C, Gaetaniello L, Masci AM, Frank J, Christiano A, et al. 2001. Human equivalent of the mouse Nude/SCID phenotype: long-term evaluation of immunologic reconstitution after bone marrow transplantation. Blood 97:880–85 92. Nehls M, Pfeifer D, Schorpp M, Hedrich H, Boehm T. 1994. New member of the winged-helix protein family disrupted in mouse and rat nude mutations. Nature 372:103– 7 93. Lin L, Spoor MS, Gerth AJ, Body SL, Peng SL. 2004. Modulation of Th1 activation and inﬂammation by the NF-κB repressor Foxj1. Science 303:1017–20 94. Lin L, Hron JD, Peng SL. 2004. Regulation of NF-κB, Th activation, and autoinﬂammation by the forkhead transcription factor Foxo3a. Immunity 21:203–13 95. Burgering BM, Kops GJ. 2002. Cell cycle and death control: long live Forkheads. Trends Biochem. Sci. 27:352–60

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-07

226

Ziegler

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:209-226. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-08

ARI

15 February 2006

18:20

HIV Vaccines Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Andrew J. McMichael MRC Human Immunology Unit, Weatherall Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DS UK; email: [email protected]

Annu. Rev. Immunol. 2006. 24:227–55 First published online as a Review in Advance on December 5, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090605 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0227$20.00

Key Words HIV-1, neutralizing antibody, T cell immunity, vaccination

Abstract A prophylactic vaccine for HIV-1 is badly needed. Despite 20 years of effort, it is still a long way off. However, considerable progress has been made in understanding the problem. The virus envelope has evolved to evade neutralizing antibodies in an extraordinary way, yet a vaccine that can stimulate such antibodies remains the best hope. Anti-HIV-1 T cell responses are evaded by continuous mutation of the virus. Vaccine strategies that concentrate on stimulating T cell immunity will at best generate broadly reactive and persisting T cell responses that can suppress virus without preventing infection, limiting or preventing the damage the virus causes. The SIV macaque models give encouragement that this is possible, but they need further understanding. Therapeutic vaccination should also be considered.

227

ANRV270-IY24-08

ARI

15 February 2006

18:20

INTRODUCTION SIV: simian immunodeﬁciency virus

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

GALT: gut-associated lymphoid tissue

The human immunodeﬁciency virus type 1 (HIV-1) pandemic shows no sign of slowing. WHO/UNAIDS estimates that 39.4 million people were living with HIV-1 infection at the end of 2004, with 5 million new infections each year. The introduction of potent antiretroviral drugs has lessened the impact of AIDS in developed countries, but their use in developing countries is patchy and is unlikely to slow the pandemic. Although the rates of infection appear to have fallen in Uganda and Senegal, possibly because of vigorous education campaigns, changes in social behavior, and increased condom use, the infection rates are still high in these countries. There is therefore a desperate need for a vaccine. Yet despite increasingly intense efforts, development of a vaccine is still a long way off. This review explores this difﬁcult problem, arguing that progress is being made and that investigators are still hopeful that a vaccine is possible.

THE NATURAL HISTORY OF HIV INFECTION The commonest route of transmission of HIV-1 is through heterosexual intercourse. Estimates of the rate of infection in very high risk sex worker cohorts show that ∼1 in 100 sexual contacts with an infected person results in infection (1–3); notably, 25% of adults in sub-Saharan Africa are in regular sexual contact with an infected partner in a stable monogamous relationship, and 5% become infected per year (4, 5). Studies of very early mucosal infection of rhesus macaques with simian immunodeﬁciency virus (SIV) have detailed what happens in early infection (6, 7). Despite deliberate insertion of virus at a high titer, infected cells cannot be detected in the mucosa for 1–3 days. The ﬁrst infected cells that become detectable are resting memory CD4 T cells that express CCR5, the virus coreceptor. Over the following 3 days, the number of infected cells increases. 228

McMichael

At ∼1 week, virus becomes detectable in local lymph nodes; Langerhans cells or dendritic cells (DCs) are involved in carrying virus to these sites. HIV can infect plasmacytoid DCs (pDCs) in peripheral tissue by binding to CD4 and mannose-binding proteins; after internalization of virus, expression of CCR7 is induced, which directs pDCs to lymph nodes (8, 9). HIV can also attach to myeloid DCs by binding to DC-SIGN, and this can target virus to HIV-speciﬁc CD4 T cells without infecting the DC (10, 11). Virus production may thus be ampliﬁed in draining lymph nodes when T cells come in contact with DC and are simultaneously activated and infected (12); HIV-speciﬁc CD4 T cells are particularly susceptible (13). By day 7, virus begins hematogenous spread and is seeded to other sites, particularly to gut-associated lymphoid tissue (GALT) where large numbers of CD4+ CCR5+ memory T cells reside. This seeding to other sites results in an exponential expansion of the infection, with up to 20% of gut-associated CD4 T cells infected and ∼80% destroyed, probably by Fas-mediated bystander killing (14). Virus appears in the blood, with viremia peaking at around day 21, often >107 virus particles per ml of plasma. Then the level of HIV-1 starts to fall, either because of saturation of the main target cell population (memory CD4+ CCR5+ T cells) (15) or because of the appearance of speciﬁc immune responses, or both (16, 17). Without therapeutic intervention, virus levels fall 10- to 100-fold and level out to a relatively stable set point between 2 and 6 months after infection. The virus set point is inversely related to the rate of progression to AIDS (18). After this acute phase, virus levels generally stay around the set point for many months. Blood CD4 T cell levels, which dip in acute infection, usually recover as set point is reached, but then gradually decline. However, this gradual decline masks what is really happening in the whole CD4+ T cell population. As blood CD4 count falls below 200– 400, virus levels rise as AIDS progresses.

ANRV270-IY24-08

ARI

15 February 2006

18:20

Virus load

10

6

CD8 T cell control Expansion in lymph node blood GALT

10

4

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

102

Infection d0

Maintenance

Delayed immune collapse

Early phase: infection of mucosal memory T cells CD4+/CCR5+ LC/DC 5

Neut Ab

10

20

NK CD8 T Neut Ab

40

CD8 T cells + CD4 T

Therapeutic vaccination

Prophylactic vaccination Prevention

Time

> 80

Abortive infection

Damage limitation

Figure 1 The natural history of HIV-1 infection and options for control by a vaccine-induced immune response. In the upper part of the ﬁgure, the change in virus load is shown over time as the virus spreads. Below is shown the time of action of different types of immune response and their effect on the virus. Abbreviations: LC, Langerhans cell; DC, dendritic cell; GALT, gut-associated lymphoid tissue; NK, natural killer cells; Neut Ab, neutralizing antibody.

IMMUNE RESPONSES DURING HIV-1 INFECTION These events, summarized in Figure 1, should be viewed in the context of the concomitant immune responses. In very early infection, innate immunity may impede virus spread, although that it does so is uncertain. It is clear that activated DC and NK cells can be found as the viremia becomes apparent (9, 19– 21); plasma IFN-α levels may rise (9). How these early events inﬂuence the early adaptive immune response and contribute to early control of virus needs further study. The ﬁrst adaptive immune responses that appear are the CD8 T cell response and a humoral antibody response that is non-

neutralizing (17). The HIV-speciﬁc CD8 T cell response appears and expands to ∼10% of all circulating CD8+ T cells as the viremia peaks, and this T cell response is probably responsible for much of the reduction in virus load (16, 22). In support of this hypothesis, depletion of CD8+ cells by anti-CD8 antibody infusion during primary SIV infection in macaques results in sustained high viremia that is only controlled when the antibody infusion ceases (23). However, a subpopulation of NK cells expresses CD8. Additional evidence that CD8+ T cells are responsible for virus control comes from cases in macaques and humans in which virus populations with escape mutations in immunodominant epitopes www.annualreviews.org • HIV Vaccines

229

ARI

15 February 2006

18:20

recognized by CD8+ T cells are selected during primary HIV infection (24–27). This ﬁnding implies strong selective pressure on the virus as the peak viremia is controlled. There is a CD4+ T cell response at the same time as the CD8+ T cell response (28, 29). As in other acute virus infections, the CD4+ T cell response is smaller than the CD8+ response (28, 30). The CD4+ T cell response has not been well studied in acute HIV-1 infection, but it may well be compromised by the susceptibility of CD4+ T cells to HIV-1 infection (13). Nonetheless, the CD8+ T cell response appears normal (31). Studies in mice have indicated that the quality of the CD8+ T cell response is inﬂuenced by T cell help. When CD4+ T cells were depleted by infusion of anti-CD4 antibody or were absent in CD4-null mice, CD8 T cell memory development was impaired (32–34); in the most extreme case, the acutely responding T cells apoptosed, and the clones were deleted (35). However, these studies involved complete deletion of CD4 T cells, including T regulatory cells, which is rather different from the partial impairment of CD4+ T cells in HIV-1 infection. Once virus set point is reached, the CD8+ T cell response continues to suppress virus. This suppression has been shown in macaques by using anti-CD8 antibody infusion to deplete CD8 T cells, which caused a rise in virus load for as long as the antibody-mediated CD8 T cell depletion was present (23, 36). The continuing selection of cytotoxic T lymphocyte escape mutants implies continuous selective pressure (37–40). The T cell response evolves over time, often with different patterns of immunodominance during the chronic phase compared with acute infection (41), which may in part reﬂect the selection of virus escape mutants and consequent new T cell responses. The phenotypic features of HIV-speciﬁc T cells, such as CD27, CD28, CCR7, CD57, and CD38 expression, differ between acute and chronic infection and differ from the phenotypes of CD8 T cells responding to other

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

230

McMichael

chronically infecting viruses (42). The differentiation state of HIV-speciﬁc T cells may be abnormally skewed (43) or may simply reﬂect subtle differences in antigen stimulation by different cell populations in different virus infections. Functional assays have also revealed heterogeneity in expression of perforin, granzymes A and B, IFN-γ, IL-2, MIP-1α, MIP-1β, and RANTES (43, 44). Expression of surface glycoproteins and secreted functional mediators can differ within single clones. It is not clear how these phenotypes and functions relate to the ability of CD8 T cells to suppress HIV-1 and whether impaired CD4 T cell help adversely affects the real function of CD8 T cells. Expression of IL-2 does have some role in slowing progression of HIV infection in both CD4+ and CD8+ T cells; antigen stimulation of IL-2 secretion correlates with slow progression to AIDS (45). Two to three months after primary infection, neutralizing antibody appears (17). However, the antibody fails to neutralize contemporary virus variants (46) and is always chasing the evolving virus so that it probably contributes little to control of the infection (47, 48). There must, however, be a strong enough antiviral force to select the escape variants, but the changes in envelope epitopes seem to have little cost to the virus, possibly because many of the epitopes are in ﬂexible parts of the envelope such as hypervariable loops of gp120 that have no other function. This contrasts with CD8 T cells, for which many escape variants come with a cost to the virus ﬁtness, so that virus escape may actually select less-virulent virus (49, 50).

WHAT COULD VACCINES DO? The problem with the natural immune response to HIV is that it is always chasing the persisting, variable virus. In these circumstances, the virus will always escape. Could this be prevented by ensuring that the immune response is there before the virus infects?

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

Vaccination has been a major cornerstone of disease prevention and control since 1798. However, vaccinologists often argue that for every successful vaccine, researchers knew that people who recovered from natural infection were subsequently protected against infection with the same pathogen. As no one has ever been known to clear HIV-1 infection, protection against this infection by an immune response may not be possible. Nonetheless, inability of those infected to clear the virus naturally is not unique to HIV1. For several persisting viruses, including Epstein Barr virus (EBV) and cytomegalovirus, natural clearance and subsequent protective immunity is unknown. This is also true of tuberculosis. However, in these infections lifelong immune control is regularly achieved, with no illness. Such a status for HIV-1 may be possible, particularly as it occurs naturally for HIV-2 in ∼80% of those infected (51). The best vaccine-induced immune response would be neutralizing antibody. If present at high enough titer in blood and mucosa at the time of ﬁrst infection, true sterilizing protection could be achieved (Figure 1). However, even this might fail if some part of the infection was caused by cell-cell spread of virus, for instance from macrophages in seminal ﬂuid to female lymphocytes, when epithelial integrity is breached. Once CD4+ CCR5+ memory T cells become infected, a vaccine-induced cellular immune response becomes important. These T cells could, in theory, destroy HIV-1infected cells before release of new virus particles (52) and act with the neutralizing antibody to limit early spread and abort infection (Figure 1). This is probably how many vaccines to other viruses actually work. But for HIV-1 there is no evidence that abortive infection occurs in humans, although it has been suggested experimentally in chimpanzees (53). There is, however, one line of evidence that suggests it could occur in humans. Highly HIV-1-exposed and -uninfected people have been identiﬁed, possibly comprising up to 5% of people at risk (54, 55).

Several studies have shown that they make HIV-1-speciﬁc CD8+ T cell and/or CD4+ T cell responses (56–59), but at variable levels (60). HIV-1-speciﬁc T cells have also been found in cervical mucosal lavage samples from exposed uninfected female sex workers (61). Some studies have claimed that mucosal IgA speciﬁc for HIV-1 is present, but this antibody in the female genital tract can be hard to distinguish from contamination by seminal ﬂuid IgA from infected partners (62, 63). The level of T cell response is often variable, possibly related to ﬂuctuating levels of HIV-1 exposure. One study of sex workers who had been exposed but remained uninfected found that later reduction in exposure to HIV actually increased susceptibility to infection (64). There are several possible explanations for these ﬁndings. The exposed uninfected people may actually be infected with HIV-1 at a low level, e.g., in GALT (65), and this infection may protect them from superinfection, by either immune mechanisms or retroviral interference [where there is target cell depletion and/or receptor blockade (66)]. Alternatively, the immune response in exposed uninfected people may simply reﬂect their undoubted exposure to HIV-1, but they are protected from infection by some other mechanism. Protective genes such as the ∂32 allele of CCR5, which abrogates expression of this virus coreceptor in homozygotes, have been excluded (57), but there could be others yet to be discovered. Finally, it is possible that these uninfected people really have been immunized, either by defective virus or by competent virus that has been cleared, and that this has given them a protective immune response. If such an immune response is protective, it must be remarkably efﬁcient because quantity of HIV-speciﬁc T cells in the blood is nearly always very low; however, the ﬁnding of HIV-1-speciﬁc T cells in the genital mucosa might mediate this protective effect by being in the right location (61). Given the lack of neutralizing antibody in such people, protection may result from abortion of very early infection by the CD8+ T cell response. www.annualreviews.org • HIV Vaccines

EBV: Epstein Barr virus

231

ANRV270-IY24-08

ARI

15 February 2006

SHIV: chimeric simian-human immunodeﬁciency virus

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

SCID: severe combined immunodeﬁciency

18:20

These results are therefore very intriguing, but proving that the immune responses actually protect remains difﬁcult. Once infection is established and spreading with high viremia, CD8+ T cell responses induced by vaccines may be the best means of damage limitation (Figure 1). Although they may be incapable of clearing infection, they may contain it for a long time. A vaccine that could exert this level of control and make HIV-1 more like HIV-2, and no more threatening than chronic EBV infection, would be a great gain. There is a widespread view that this might be the best that can be offered by a prophylactic T cell–stimulating vaccine and that this is achievable (see below); hence, a large number of vaccines and clinical trials are addressing this hypothesis. These trials include not only prophylactic vaccine trials but also therapeutic vaccination aimed at enhancing CD8+ T cell–mediated control of HIV-1, usually combining a course of antiretroviral drug therapy (ART) and vaccination (Figure 1, and see below).

NEUTRALIZING ANTIBODY VACCINES Stimulation of neutralizing antibodies with broad speciﬁcity for all HIV variants by a vaccine is undoubtedly the best approach, if it can be achieved. Passive administration of one or more broadly neutralizing monoclonal HIV-1-speciﬁc antibodies to macaques immediately before infection with SIV-HIV hybrid virus (SHIV) that has an HIV-1 envelope prevents infection (67). Similar results were obtained in mice with severe combined immunodeﬁciency (SCID) that were reconstituted with human lymphoid tissue and then infected with HIV-1 (68). Vaginal administration of such antibodies to female macaques also protected against vaginal challenge (69, 70). In contrast, administration of these antibodies after infection is too late; the established virus simply escapes by mutation (68). The message is clear: The presence of neutral232

McMichael

izing antibody, at high titer, before infection protects. When HIV was ﬁrst identiﬁed and sequenced, many thought it would be straightforward to make the virus envelope protein gp120 and create a vaccine. This idea has been thwarted by the intrinsic variability of the virus and the differing susceptibility to neutralization of virus cultured on cell lines in vitro compared with primary isolate virus (71). One gp120 vaccine, made by VAXGEN, was taken to two phase III efﬁcacy trials in the United States and Thailand, but there was no protection against HIV-1 infection (72, 73) (see Table 2, below). The problem lies in the structure of the envelope. Two seminal contributions have been the solutions of two crystal structures of HIV-1 gp120. The ﬁrst HIV-1 gp120 crystal was bound to CD4, as well as to a Fab antibody that is speciﬁc for the CCD5/CXCR4binding sites (74, 75). The second structure, SIV gp120, was on its own, not bound to any ligands (76). Given the functional and sequence homologies of HIV and SIV envelopes, these structures clearly explain much of the difﬁculty in ﬁnding neutralizing antibodies and show that there are serious obstacles ahead. Figure 2 shows the key features of the structure of gp120 and the conformational changes it undergoes on binding its ligands (77). Gp120 and the linked gp41 are homotrimers. Thus, much of the surface of monomeric gp120 is hidden and not accessible to antibody (75). The exposed surface is heavily glycosylated (76); the sugars are put there by host enzymes so that the carbohydrate moieties themselves are very unlikely to be immunogenic in humans. The importance of the sugars in protection against neutralization by antibodies is illustrated by an experiment in which macaques were infected with SIV that had mutations designed to remove glycosylation sites around the V1/V2 loops; the mutations were designed to require more than one mutation to repair. When macaques were infected with this virus, there was good

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

Figure 2 Changes in the structure of the HIV envelope on binding to target cells. (a) The heavily glycosylated envelope comprises gp120 and gp41 as a trimer. This structure was solved by Chen et al. (76) for SIV. (b) When the envelope binds to CD4, the gp120 undergoes a conformational change that forms and exposes the binding site (shown as red β sheet changing to green) for the coreceptor CCR5 or CXCR4. This structure was determined by Kwong et al. (74). (c) Subsequent to this event, the gp41 undergoes conformational changes to mediate fusion of the membranes so that the virus enters the target cell. [Figure from Kwong et al. (77), with permission of the author and the editors of Nature.]

initial control with potently neutralizing antibody, but after a few weeks virus escaped neutralization, and the sugars had been put back at the nearest site where this could be achieved by a point mutation (78). The CD4-binding site is deeply recessed (75) and is also guarded by the V1/V2 loops that can be varied in the virus without any ﬁtness penalty. The CCR5-binding site is completely absent in the unliganded structure and appears after a large conformational change in the structure after CD4 binding (76). In addition, this site is guarded by the V3 loop, which also can vary without cost to the virus, although the sequence does inﬂuence the choice of CCR5 or CXCR4 coreceptor. Thus, the virus has evolved in a remarkable way to be able to evade neutralizing antibody attack during the chronic phase of infection. Two potential Achilles heels have to be preserved, the CD4-binding site and the CCR5binding site, but both are well hidden. Most of the exposed surface where a binding antibody could sterically interfere with envelope function is protected by the sugars (75).

Many monoclonal antibodies speciﬁc for HIV envelope gp120 and gp41 have been made, both in mice and in humans, by immortalizing EBV-transformed B cells or by phage display. Out of these, only ﬁve have been found that are broadly neutralizing (79–83). These antibodies are summarized in Table 1. One antibody, b12, sees the CD4-binding site, having a long heavy chain CDR (complementarity-determining region) 3 loop. This good neutralizing antibody binds to gp120 with minimal entropic penalty, in contrast to poorly neutralizing antibodies to the same site that bind with an energetic handicap because of induced conformational changes (83a). Another antibody, X5, sees the CCR5/CXCR4-binding site, which is exposed only after CD4 binding, but this antibody only works as a Fab fragment. One extraordinary antibody, 2G12, sees a complex mannose determinant made up of mannans linked to the asparagines at positions 289, 332, and 396 (84). The spatial epitope dimensions are greater than normal for an antibody-binding site; this is explained by the www.annualreviews.org • HIV Vaccines

233

ANRV270-IY24-08

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Table 1

18:20

The five monoclonal antibodies that broadly neutralize HIV-1

Antibody

Specificity

Special features

Reference

b12

CD4-binding site

Long CDR3 region

(165)

X5

CD4 induced (coreceptor-binding site)

Fab only effective

(166)

2G12

Complex mannose

Antibody has VH domain swap

(80)

2F5

Conserved membrane proximal domain of gp41

IgG3 with long CDR3; cross-reacts with cardiolipin

(85, 167)

4E10

Conserved membrane

IgG3 with long CDR3; proximal domain of gp41 cross-reacts with cardiolipin

(79, 85)

remarkable ﬁnding that 2G12 has mutations in the heavy chain hinge region that cause a crossover of the variable regions to match the opposite light chains, making a single, large antigen-binding site (80). Failed attempts to ﬁnd other similar antibodies imply that this is a very rare if not unique antibody. Two broadly neutralizing human monoclonal antibodies that bind to the membrane proximal region of gp41 have been found; the two adjacent epitopes can be represented by linear peptide sequences. Both antibodies 2F5 and 4E10 are unusual in having long CDR3 regions. A structural analysis of 2F5 binding to the gp41 peptide indicated that the hydrophobic tips of the heavy chain CDR3 region interact with membrane lipid, extending the binding interactions (84a). Some autoantibodies have similar features, and when these monoclonal antibodies were tested against clinically relevant autoantigens, both were found to bind with high afﬁnity to cardiolipin (85). Using the peptides to scan for the presence of such antibodies in large numbers of infected patients has shown that these speciﬁcities are very rarely present. Self-tolerance of B cells likely makes generating antibodies with these speciﬁcities after infection or vaccination almost impossible (85). Thus, although these antibodies are very potent, their special features ensure that they probably cannot be reproduced by vaccination. The VAXGEN gp120 was the last of a series of early envelope vaccines that failed. Although some trials in progress still use gp120 234

McMichael

as a boost for a recombinant virus or DNA prime, nearly all current trials are of T cell– stimulating vaccines. At the same time, much effort is being put into design of immunogens that represent the intermediate conformational variants or the liganded gp120 structure (Figure 2). There is also a search for further broadly neutralizing determinants that might be more amenable to vaccine design; however, such determinants may not exist. Characterization of very early virus isolates, obtained before the immune response has a chance to mold the virus envelope, might reveal further sites that could be targets for neutralization, if the earliest infecting virus that grows unimpeded by antibodies has novel structural properties in its envelope that are normally rapidly selected out.

VACCINES THAT STIMULATE T CELL RESPONSES As described above, in natural HIV-1 infection and experimental SIV infection the primary CD8 T cell response clearly follows the virus, peaking just after virus load starts to decline (16, 17, 22, 24). The best hope for a T cell–inducing vaccine is to ensure a more rapid secondary T cell response targeted to mucosal sites so as to act at the early stages of infection and to eliminate the virus at this time (Figure 1). However, despite the puzzle of the exposed uninfected people, there are as yet no studies of SIV vaccination and challenge in macaques that show that this can

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

happen. A more likely scenario is that vaccinestimulated memory T cells will need (a) expansion by cell division, (b) differentiation to acquire the capacity to kill virus-infected cells, and (c) relocation to mucosal sites and draining lymph nodes. Thus, the window for being able to abort the initial infection may be too narrow. In this scenario, T cell–inducing vaccines will limit damage by controlling HIV-1 as effectively as EBV is controlled in most of us. Although there is good evidence that CD8 T cells suppress virus, how they do it and what functions will be required of vaccine-stimulated T cells is unclear. For long-term asymptomatic control of human T-lymphotropic virus type 1 (HTLV-1), T cell–mediated killing is important (86), but for hepatitis B virus, IFN-γ production is critical (87). For HIV-1, this information is badly needed, as different vaccine approaches, like different virus infections (42), are likely to stimulate T cells with different functional capabilities. In this context, we need to know what type of immune response to measure after vaccination. At present, all vaccine clinical trial teams use peptide antigenstimulated IFN-γ production as their main assays of vaccine immunogenicity, either using the ELISPOT method or intracellular cytokine production measured by ﬂow cytometry. However, this could be the wrong function to measure, although there is a chance that antigen-stimulated IFN-γ production could be a surrogate for the real protective functions. The critical information could come from studies of those already infected. Clearly, some T cell responses drive virus escape and must therefore be suppressive, whereas others appear to offer little protection, for example those that were not evaded by superinfecting virus (88). We do not know all T cell functions, and some of the less well deﬁned, such as the CD8+ cell antiviral (HIV-1) factor (CAF), could be important (89). We may not be able to design effective vaccines until we understand the most effective antiviral activities of CD8+ cells.

The CD8 T cell vaccine approach has been built on a series of inﬂuential studies in the rhesus macaque–SIV challenge model that demonstrate partial protection by vaccines that stimulate CD8 T cell immunity. Immunogens encoding or presenting internal virus proteins such as gag stimulate strong CD8 T cell responses. The vaccines include plasmid DNA encoding HIV genes, recombinant modiﬁed vaccinia virus Ankara (MVA), recombinant adenovirus-5, and recombinant vesicular stomatitis virus, as well as combinations of some of these (90–93). In all these experiments, the challenge virus was an aggressive SIV-HIV hybrid virus SHIV89.6P that expresses a CXCR4-speciﬁc HIV envelope (94) and causes very rapid loss of CD4+ T cells and onset of immunodeﬁciency. In no case did the vaccine give sterilizing immunity; all vaccinated animals were infected, but they did not lose CD4 T cells, and they remained healthy with virus loads 1000-times lower than the unvaccinated controls. However, there is a possible problem: Paradoxically, this virus appears to be easier to control by vaccination than some others, and the same vaccine approaches have been much less effective at controlling infection with the CCR5 using viruses SIVmac239 or SIVmac251 (95– 97), although temporary partial reductions in virus level have been seen (98). Furthermore, immunization of macaques with a phagedisplay peptide vaccine aimed at stimulating neutralizing antibody gave a similar level of protection against SHIV89.6P, suggesting that this aggressive virus may be relatively easy to control (94). Clearly, this issue needs further exploration, in particular to determine whether the apparent differences relate to the different coreceptors used by SHIV89.6P (CXCR4) and SIVmac239 (CCR5). The early evidence that the SHIV89.6P challenge virus is controlled by CD8 T cells was indirect; in some cases (92) the immunogens did not include envelope, and no neutralizing antibodies were stimulated. But now direct evidence exists (99): In an initially protected vaccinated animal that started to lose www.annualreviews.org • HIV Vaccines

MVA: modiﬁed vaccinia virus Ankara

235

ARI

15 February 2006

18:20

CD4 T cells with a rise in blood virus titer, a mutation was observed of a single amino acid in the immunodominant gag epitope, making it unable to bind the presenting MHC molecule (Figure 3). Thus, at least in this animal, the protection was likely mediated by T cells with this speciﬁcity. There are as yet no equivalent data in humans. In one clinical case, superinfection by a second B clade virus in an HIV-infected patient occurred despite a very strong pre-existing CD8+ T cell response, but the new virus was different in about half the epitopes, and the quality of the T cell response could have been impaired by poor CD4+ T cell function (88). Also, we do not know how often chronically HIV-1infected people resist superinfection. Stimulation of CD8+ T cell responses with vaccines to levels similar to those that protect macaques from SHIV is proving difﬁcult in humans. Vaccines that stimulate strong T cell responses in mice may fail completely in humans, and, although immune responses stimulated in primates should be closer to those seen in humans, they may not always be predictive. Key issues could be vaccine dose, different pattern speciﬁcity of Toll-like receptors (TLRs) in different species, degree of vector attenuation in different species, cross-reactivity with self-antigens or epitopes previously encountered, and unrepresentative T cell responses to highly immunodominant epitopes. For priming CD8 T cell responses,

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

vaccine antigen needs to enter the MHC class I–processing pathway. There are three routes: (a) direct entry by using intracellular pathogens as vectors; (b) transfection for DNA plasmids; or (c) cross-priming by DC, which can take up protein antigen from outside the cell. The last is increasingly recognized as an important route of priming and, unlike the direct route, favors antigen that is stable rather than rapidly degraded (100). Guided by these principles, several vaccine delivery systems are under development. The key candidates are shown in Table 2, which indicates the considerable activity in the past two years, with 35 trials in progress and at least another 5 completed but not yet formally reported. Most experience so far has been obtained with plasmid DNA, MVA, canarypox, fowlpox, and adenovirus-5. More recent additions are adeno-associated virus (AAV), Venezuelan equine encephalitis virus, and lipopeptides. Some of these vaccine vectors (MVA, fowlpox, adenoviruses) are also being used for stimulation of T cell responses to other pathogens, including malaria, mycobacterium tuberculosis, human papilloma virus, and tumor antigens. The HIV genes/proteins most commonly include gag, pol, nef, and env, but all genes are being used. The subtypes are most commonly B or C, but also include A, D, and E strains. Because these vaccine candidates are used in clinical trials that are very highly

−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 3 Escape of SHIV89.6P from T cell control in a vaccinated macaque shows that protection is mediated by CD8+ T cells. The top graph shows the time course of RNA virus load in the macaque that had been vaccinated by DNA-expressing gag and IL-2 and challenged with SHIV89.6P at week 0. Between weeks 10 and 20 there was good virus control with a low virus set point; the unvaccinated controls had virus loads ∼106 copies per ml (not shown here). After 20 weeks, virus load started to rise, and the CD4 T cell count fell dramatically (middle graph). At the same time, the strong CD8+ T cell response to the immunodominant gag epitope p11C declined, measured by tetramer staining of CD8 T cells (lower graph). These events after week 20 were associated with a selection of an escape mutant in the immunodominant epitope in gag p24 p11C 181–189 sequence shown at the top (single letter amino acid code). At week 20 and thereafter the threonine (T) at position 182, a residue that anchors the peptide to the presenting Mamu A∗ 01 MHC molecule, was mutated to isoleucine (I). This new peptide does not bind to Mamu A∗ 01 and is therefore an escape mutant that is represented in all virus sequences (indicated in parentheses as the number of variant sequences/number of sequences determined). [Figure adapted from Barouch et al. (99), reproduced with permission of the authors and editor of Nature.] 236

McMichael

ANRV270-IY24-08

ARI

15 February 2006

18:20

regulated, even a simple phase I trial can often take two years, and the clinical work may then be followed by more than a year for data analysis. This lengthy process means that there are to date very limited published data on

many of the HIV candidates. Earlier studies of vaccine-stimulated T cell immunity generally used less quantitative assays, such as T cell killing after culture in vitro or proliferation in vitro measured by thymidine incorporation

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Gag p11C (181–189) week 0 week 14 week 20 week 24 week 28 week 36 week 44

CTPYDINQM

––––––––– ––––––––– – I ––––––– – I ––––––– – I ––––––– – I ––––––– – I –––––––

(15/15) (8/8) (10/10) (11/11) (11/11) (11/11) (10/10)

7.0 6.0

RNA (copies / ml)

5.0 4.0 3.0 2.0 1200

800

+ CD4 T cells / µl 400

0 20 15

p11C tetramer (percent)

10 5 0

0

10

20

30

40

50

www.annualreviews.org • HIV Vaccines

237

ANRV270-IY24-08

ARI

15 February 2006

18:20

(101). Since 2002 the assays have been improved and validated by interlaboratory studies (e.g., 102), making them more reliable. Nevertheless, from the more recent studies, some general points have been reported at conferences, and some conclusions can also be drawn from studies with similar types of vaccine in other ﬁelds.

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

1. All the vaccine candidates are strongly immunogenic for CD8 T cell responses in mice. 2. Most, or all, vaccine candidates stimulate measurable T cell responses in macaques (e.g., 92). 3. Some candidates, including DNA given together with IL-2, DNA and MVA in a DNA prime with MVA boost combination, adenovirus-5, and vesicular stomatitis virus, partially protect macaques against SHIV-89.6P virus challenge (see above). However, protection against SIVmac239 has been much harder to achieve and has usually been marginal (95). 4. DNA stimulates weak T cell responses in humans that are predominately CD4 T cells (103, 104). 5. MVA, canarypox, and fowlpox stimulate weak primary T cell responses in humans (104, 105). 6. The DNA prime with an MVA boost gives stronger responses than either alone in humans; responses are mostly CD4+ T cells. Responses peak at day 7 after MVA and are short lived (104; N. Goonetilleke, S. Moore, N. Winston, A. Mahmood, S. Pinheirho, J. Roberts, T. Hanke, A. McMichael, unpublished data). 7. MVA can boost well-primed CD8+ T cell responses (106). 8. Recombinant replication defective adenovirus-5 gives fairly strong and sustained CD8+ T cell responses in humans (107). 9. Pre-existing antibodies to adenovirus5, which are present at high frequencies 238

McMichael

in many populations, reduce immunogenicity of recombinant adenovirus-5 vaccines (107, 108). DNA-induced responses are dose dependent in macaques, needing 2–5 mg to obtain responses (91). This dose is close to the maximum that is practical, and humans may need much larger doses to see equivalent T cell responses. Adding adjuvants to DNA vaccines may enhance their immunogenicity, and various combinations of cytokines, chemokines, and TLR ligands are under study. Also, delivery devices such as the gene guns that ﬁre particles of gold coated with the vaccine into the dermis/epidermis may enhance immunogenicity (109). The pox virus vectors are probably at a disadvantage because ∼200 pox virus proteins may compete to stimulate a T cell response, but this may be less of a problem when used to boost after priming with a different vector recombinant for the same vaccine antigen. MVA may be too attenuated compared with adenovirus-5, which may persist sufﬁciently to give a longer lasting T cell response. The responses to some vaccines may be underestimated by the IFN-γ ELISPOT assay; more sensitive assays for which the response is measured after several days of culture with the antigen might better reﬂect the presence of central memory T cells (104). This approach has revealed many more T cell responses than the conventional assays in vaccine recipients. The approaches above show what some of the vectors can do, and although somewhat disappointing they have established working methods and validated assays. Furthermore, for some vectors, MVA for instance, they have given an indication of what their role might be (stimulating CD4+ T cell responses after DNA priming and boosting well-primed CD4+ and CD8+ T cell responses) or, for adenovirus vectors, how they might be improved to avoid pre-existing immunity. There is little doubt that new immunogens, vectors, delivery systems, and vaccine combinations will be found that stimulate strong CD8+ and CD4+ responses to

ANRV270-IY24-08

ARI

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Table 2

15 February 2006

18:20

Current and recently completed (unpublished) HIV vaccine clinical trialsa

Delivery

Clade

Immunogen

Phaseb

DNA

A, B, C

e, g, p, n

I

DNA

C

g, p

I

DNA

C

e, g, p, n, t

I

DNA

A, B, C

g, p, n, e

I

DNA

A, B, C

e, g, p, n

I

DNA

A, B, C, D, E

epitopes

I

DNA

B

g

I

DNA

B

g, p

I

DNA

B

e, g, p, t, r, n

IIA

DNA

A, B, C

e, g, p, n

IIA

DNA+120

B

e

I

DNA+MVA

A

g

I×3

DNA+MVA

A

g

IIA × 2

DNA+FPV

B

e, g, p, r, t, v

IIA

DNA+Ad5

B

g

I

Lpep

B

g, p, n

I

Lpep

B

g, p, n

IIA

Lpep+CPV

B

g, p, n

IIA

MVA

A

g

I

MVA

C

e, g, p, n, t

I

MVA+FPV

B

e, g, p, t, r, n

I

CPV

B

e, g, p

I

CPV+120

B, E

e, g, p

III

Ad5

A, B, C

g, p, n

I

Ad5

B

g

I

Ad5+CPV

B

g, p

I

Ad5

B

g, p, n

IIB

AAV

C

g, p, n, t

I

VEE

C

g

I

Protein

D

e

I

Protein

B

t

I

Protein

B

e

I

120

B

e

III

120

E

e

III

VLP

B

e, g

I

VLP

C

e

I

a Abbreviations: DNA, plasmid DNA; MVA, modiﬁed vaccinia virus Ankara; 120, glycoprotein gp120; FPV, fowlpox virus; CPV, canarypox virus; Ad5, adenovirus-5; Lpep, lipopeptide; AAV, adeno-associated virus; VEE, Venezualan equine encephalitis virus; VLP, virus-like particle; e, HIV-1 envelope; g, HIV-1 gag; p, HIV-1 polymerase; n, HIV-1 nef; t, HIV-1 tat; v, HIV-1 vpr. b Phase I trials are in HIV-1-uninfected low-risk volunteers, n < 100; Phase IIA trials are in HIV-1-uninfected low-risk volunteers, n > 100; Phase IIB trials are proof-of-principle (efﬁcacy) trials in HIV-1-uninfected high-risk volunteers, n > 100; Phase III trials are powered efﬁcacy trials in HIV-1-uninfected high-risk volunteers, n > 5000. For more details of vaccines currently in trial, see http://www.iavireport.org/trialsdb/.

www.annualreviews.org • HIV Vaccines

239

ANRV270-IY24-08

ARI

15 February 2006

18:20

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Figure 4 Worldwide distribution of the major HIV-1 subtypes (clades) and HIV-2. The subtypes A–H are subdivisions of the M strain; the O and N strains are rare, but are very close to SIV strains found in chimpanzees.

recombinant antigens. Also, immunization protocols that optimize the size of response and include mucosal immune responses will be worked out. However, the process has to be done through clinical trials that are very highly regulated, and so progress is slow. Standing back to view the mass of current trials, one is excited that so many groups are working on HIV-1 vaccines but disappointed by the degree of repetition and by the extent to which current trials aim to test only a single hypothesis: that CD8 T cells induced by vaccination will protect against HIV-1 as well as they do against challenge viruses such as SHIV89.6P in macaques. However, if HIV-1 behaves more like SIVmac239, investigators in this ﬁeld will face serious difﬁculty and will need much more basic information about what type of T cell response is needed. In addition, virus variability is likely to be a serious problem.

VIRUS VARIABILITY AND VACCINE DESIGN HIV-1 is highly variable, and investigators have argued that this alone could compromise the success of any vaccine that goes into efﬁcacy trials (73). Phylogenetic tree analysis of virus RNA and protein sequences reveals considerable heterogeneity (110). There are three 240

McMichael

main types of HIV-1, M, N, and O, that are thought to represent separate introductions into humans from chimpanzees. The main M type is further divided into six principle subtypes, or clades, A, B, C, D, E, and G, that represent early branching of the virus in its evolution in humans (110). The clades differ by ∼20% of amino acid sequence and have distinct geographical distributions, as shown Figure 4. All the clades are found in central equatorial Africa, where the transfer of virus from chimpanzees to humans probably ﬁrst occurred. In most of the other countries, there is a single dominant clade, probably representing a founder effect. Within the clades there is further variability in up to 10% of the amino acid sequence. Despite this huge heterogeneity, some features of the sequences suggest conservation as well. There are clear consensus sequences for each clade that suggest ﬁtness costs as the virus varies so that there is a tendency toward conservation. Much of the intraclade sequence variability in the virus core and regulatory proteins is probably driven by immune escape from HLA class I–restricted CD8+ T cells. Moore et al. (40) have elegantly demonstrated that common HLA class I types in a population are associated with virus sequence variability in individual patients, the same HLA types selecting the same mutations

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

repeatedly (40). It is also notable that when escape mutations have been monitored closely, they tend to reiterate known variability in the very large virus sequence databases (111, 112). This tendency to mutate repeatedly in the same way suggests that there are constraints on virus sequence variability, with possible ﬁtness costs for some escape mutations. In at least two cases, a single escape mutation that affects epitope presentation is strongly associated with other sequence changes that are probably compensatory (38, 113). Also, when the virus is transmitted to partners or infants of different HLA type, the sequence may revert to consensus (111). This suggests a balance between immune control, virus escape, and viral ﬁtness. There are very few, possibly no, HIV-1 epitopes from which the virus cannot escape by changing one or more amino acids. Envelope variability is largely driven by neutralizing antibody. Much of the change occurs in the hypervariable loops that have little or no structural role and therefore no ﬁtness cost, although changes in the V3 loop contribute to coreceptor selection, with CCR5 speciﬁcity changing to CXCR4 speciﬁcity (114). The ﬁve broadly neutralizing monoclonal antibodies discussed above are less completely cross-neutralizing on non-B clade viruses, compatible with some structural variation, for instance in the distribution of glycosylation sites (82). Further analysis is needed to determine whether there might be different broadly neutralizing epitopes for the different clades. For the CD8 T cell response, vaccine designers need to know how sensitive T cell clones are to virus variability. There has been a widespread notion in the vaccine ﬁeld that T cells can cross-react quite broadly and therefore that a vaccine might not have to match the prevalent virus subtype (115). The idea of cross-reactive cytotoxic T cells may have come from very early studies on antiinﬂuenza-speciﬁc T cells that cross-reacted across inﬂuenza virus subtypes; however, this cross-reaction occurred because the T cells

respond to conserved internal virus proteins rather than to the variable surface glycoproteins (116). One study that emphasizes this point explored T cell recognition of all 171 single amino acid variants of an HIV gag nonamer epitope and found that only one third were recognized, even though most were bound by the presenting HLA A2 molecule (117). Similar results were found with different T cell clones. These data imply that most sequence variation in an epitope can generate escape mutants. This ﬁnding is compatible with many careful studies of individual escape mutations and the imprinting of sequence variation in the virus by HLA type (40). The implication for vaccine design is clear: Vaccine immunogen sequence must match the sequence of prevalent virus in the population to be vaccinated. At the very minimum, this means matching vaccine to clades. However, it also raises concerns that variability that already exists within HIV-1 subtypes could evade vaccine-induced T cell immunity quite easily. Furthermore, if the vaccine does not prevent infection but allows infection to occur, albeit better controlled than in unvaccinated people, there is a real risk of virus escape if the T cell response is too narrow. This risk was dramatically demonstrated in macaques when a single amino acid change in an immunodominant epitope enabled SIV to escape control in a vaccinated monkey (99) (Figure 3). The problem of virus variability has been raised as an almost insuperable obstacle to a successful HIV vaccine (73). However, there are ways to lessen the chances of virus escape. Having matched the vaccine and virus clades, vaccine developers will need to ensure that the T cell response is directed at several epitopes. The chances of escape are reduced if multiple mutations are needed, particularly if the T cell response can dominate the virus early on with a relatively low level of virus replication. The breadth in T cell response not only refers to number of epitopes seen but also to the clonality of the T cells. The phenomenon of immunodominance could be a serious obstacle www.annualreviews.org • HIV Vaccines

241

ANRV270-IY24-08

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ART: antiretroviral drug therapy

242

18:20

to vaccine design and must be understood if it is to be avoided. Immunodominance can focus the majority of the response to a few, or even one, of the epitopes and is usually apparent in T cell responses to HIV (39, 118). Indeed, immunodominance makes the T cell responses predictable according to HLA type of the infected person. These large T cell responses probably exert more selective pressure for virus escape. Possible mechanisms of immunodominance include the selection of epitopes during antigen processing, differing afﬁnity of different epitopes on the presenting HLA molecules, and half-life at the cell surface (119). Similarly, differing precursor frequencies of particular T cell clones, possibly inﬂuenced by cross-reactions to previous infections, can inﬂuence which T cell responses dominate (120). T cell receptor avidity for particular peptide-HLA complexes may also have an inﬂuence. This avidity is more complex than simply one-to-one afﬁnity and is not a matter of highest afﬁnity being best. Surface density of the receptor and coreceptors may be important. Competition between T cell clones may ultimately be a major determinant of T cell clone selection (120). Thus, some vaccine regimes aimed at stimulating high-level T cell responses may inadvertently favor narrow T cell responses. This possibility will have to be addressed by direct experimentation in humans, comparing different vaccination regimens. The use of complex mixtures of sequence variants in vaccines may favor more broadly cross-reactive T cell responses but could also run into trouble with T cell receptor antagonism, in which an altered epitope can inhibit the T cell response to the cognate epitope and therefore lessen rather than enhance the response (121). If variants are given sequentially, the response to the ﬁrst variant may dominate and impair the subsequent response to subsequent variants—a kind of original antigenic sin (122). The latter may arise because weak cross-reactivity to the variant gives a stronger signal to the weakly crossreacting memory T cells than to naive T cells McMichael

with more appropriate T cell receptors. This implies some suppression of the primary response by the memory response, possibly involving regulatory T cells (123). Thus, stimulation of inappropriate T cell responses by the vaccine might impair the subsequent response to the infecting virus, although this seems likely to be a more theoretical than real concern. All these issues need to be addressed. Studies in mice and macaques can only be of limited value because of differing MHC types and different rules for different epitopes and T cell receptors. There is a clear need for clinical studies that are small scale but intense in terms of the assays used and that can directly address these questions. To speed up the discovery process, these experimental medicine studies need to be less rigidly regulated than conventional phase I trials.

DAMAGE LIMITATION: A ROLE FOR THERAPEUTIC VACCINATION? If the most likely beneﬁcial effect of a T cell– inducing vaccine is better control of HIV-1 in those vaccinated, the question has to be asked whether therapeutic vaccination of those already infected might achieve this more efﬁciently (124, 125). The ﬁrst attempts at immunotherapy were disappointing (e.g., 126). Immunization with gp120, gag particles, or whole inactivated HIV, but without the use of effective antiretroviral drug therapy (ART), failed to reduce virus load, despite demonstrable increases in T cell immune responses (127–133). A more logical approach is to use a combination of effective (normally triple drug) ART and vaccination. Patients who start ART reduce virus load substantially, and, probably as a direct consequence, the level of T cell response declines (28, 134). When ART is interrupted, virus rebounds within 3–6 weeks to pretreatment levels (29), but although an increased T cell response follows, this does not result in better immune control (135). Some

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

investigators have argued that this rebound may be reduced by increasing the T cell response by vaccination to pretreatment levels (124, 125). The idea has been modeled in macaques infected with SIV. When animals that were chronically infected with SIVmac251 were treated with ART and then immunized with a recombinant attenuated vaccinia virus, NYVAC, expressing multiple SIV genes, temporary virus control was achieved after stopping the drugs (136). A similar result was achieved when animals were given ART and immunized with a NYVAC-SIV vaccine in primary infection (137). Control of virus rebound after stopping ART was also seen in SIV-infected monkeys after an intradermal immunization with DC transfected with DNA encoding SIV genes (138). These results are encouraging, although the effects are more modest than the surprisingly large reductions of virus load described by Lu et al. (139), who immunized untreated SIV-infected macaques with autologous DC pulsed with inactivated SIV. Vaccination of HIV-infected people who are on ART with subsequent drug treatment interruption has given more mixed results. A phase III trial of inactivated whole HIV, Remune, failed to show any effect (140). Similarly, when Remune was compared with a recombinant canarypox vaccine in patients who had started ART in acute infection, no effect was seen after 24 weeks of drug treatment interruption with either vaccination (141). Plasmid DNA vaccines encoding HIV genes have been tested for immunogenicity in ARTuntreated and -treated patients but have only stimulated weak T cell responses (101, 142), and drugs were not withdrawn. Several trials are being conducted of recombinant virus vaccines, canarypox, fowlpox, and MVA, with treatment interruptions. Initial reports at conferences indicate good boosting of T cell responses in patients on ART, and in some studies there have been some effects on virus load, e.g., delay in time to reach the virus load at which ART has to be restarted or a small reduction in virus load (143, 144). In another

trial, no reduction of virus load was seen on stopping ART after immunization with recombinant canarypox (145), but Harrer et al. (146), in an uncontrolled study, found a sustained reduction in viral load after immunization with an MVA-nef vaccine and subsequent interruption of ART. Finally, conﬂicting results have been obtained by immunizing patients, who were not on drug therapy, with HLA-matched or autologous DC pulsed with epitope peptides (147) or whole inactivated virus (148). In the former study, no effect on virus load was seen, whereas in the latter experiment, a substantial reduction in virus was seen in approximately half the patients studied. This surprising result, which is similar to a ﬁnding by the same investigators in macaques (139), implies correction of a major defect in DC function, and it needs urgent independent conﬁrmation. Despite rather equivocal results so far, therapeutic immunization is still worth pursuing. Care needs to be taken to deﬁne the quantity and quality of the T cell responses. Interruption of ART, which in effect gives a virus challenge in vivo, has been considered safe because it does not appear to select drug resistance and because virus control can be rapidly regained on treatment resumption (29). Therefore, this approach provides an opportunity to study correlates of protection, as virus control by the immune response is regained, and to modify and improve such responses. There is real scope for improving the quality of the anti-HIV-1 immune response (e.g., enhancing HIV-1-speciﬁc CD4 T cell activity or refocusing T cell responses on conserved epitopes) before treatment interruption. One could argue that positive effects of vaccination are likely to be seen more rapidly in this scenario than in the painfully slow efﬁcacy trials of prophylactic vaccines that are only just beginning. If a reliable way of gaining lasting immune control could be found, it would offer possibilities for therapy in developing countries at relatively low cost. It is also noteworthy that original antigenic sin could work in www.annualreviews.org • HIV Vaccines

243

ANRV270-IY24-08

ARI

15 February 2006

18:20

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

favor of the patient because if the vaccine did not match the virus perfectly it might nevertheless boost T cells highly speciﬁc for autologous virus—in contrast to prophylactic vaccine for which, in theory, vaccination that sets up responses to the wrong epitope variants could impair natural immune responses to a later infecting virus. Finally, such an approach would offer something for the millions who are already infected with little hope of access to effective ART and for whom prophylactic vaccination offers nothing.

ARE THERE OTHER APPROACHES TO HIV-1 VACCINATION? Three other approaches to protection against SIV appear to work in macaques but have not been fully explained. In the ﬁrst approach, Stott et al. (149) showed that macaques that had been vaccinated with inactivated SIV that had been grown on human cells were fully protected against challenge with SIV that had also been grown on humans cells, but not against SIV that had been grown on macaque cells. Further investigation implicated an immune response, probably antibody, to HLA antigens, which are incorporated into virions as they bud at cell surfaces (150). However, there has been no convincing evidence that allo- (rather than xeno-) antibodies to HLA antigens can protect against HIV-1. This approach might be worth revisiting now that there are better ways of characterizing both antigens and antibody immune responses. Induction of antibodies to self-antigens such as CCR5, although probably not to HLA, might be safe and feasible (151), given that people who are homozygous for the CCR5 ∂32 gene and who lack the protein are healthy and resistant to HIV-1 infection (152). In the second approach, Desrosiers and colleagues (153) showed that infection of macaques with attenuated SIV offered good protection against superinfection with the unattenuated parent strain of SIV. Re244

McMichael

searchers usually argue that the protection is due to an immune response, but sustained efforts to identify which components of the immune response are responsible have not given clear answers (73). In particular, the protection cannot be transferred by passive antibody (154), and protection could only be partly abrogated by depletion of CD8 T cells (155) or of all T cells (156). Furthermore, the level of measured immune response does not correlate with efﬁcacy (73). Protection against challenge with nonhomologous virus was less complete, with a median virus load reduction of 100-fold rather than 1000-fold that could reﬂect less efﬁcient immune recognition of variant epitopes (157); this is a key experiment that, although not perfect, may be the best evidence that the protection is mediated by a speciﬁc immune response. However, there could also be involvement of innate immune responses at the sites of virus infection. Innate antiviral function could be chronically activated by the persisting infection with the attenuated virus. It is also possible that a component of the protection is not immune at all. Retroviral interference in vitro or in vivo occurs when the ﬁrst virus depletes susceptible target cells or blockades the receptor so that the second virus cannot infect (66). This interference could explain why protection is seen before development of detectable immune responses (158). It could also account for the escape of virulent wild-type SIV in vivo in two macaques infected with an attenuated SIV with a 4 amino acid deletion in nef, which was repaired in the escaping virus (159). It is hard to explain this escape in the presence of an immune response that was able to protect against deliberate superinfection with a large dose of virulent virus in the same animals. Given the much greater susceptibility to infection of gut-associated memory T cells compared to other peripheral lymphoid tissue (160), it is possible (and testable) that attenuated SIV could infect and destroy the gut-associated T cells that are most susceptible to SIV infection. The difference

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

between the virulent and attenuated virus might be more related to the virus’ ability to activate and induce apoptosis in uninfected bystander T cells (14, 161). Thus, retroviral interference needs to be seriously reconsidered as one of the possible mechanisms behind the protection in animals infected with attenuated SIV. Whatever the mechanism of protection against attenuated SIV, this approach is very unlikely to be possible in humans, even though other attenuated virus vaccines have an excellent record in human vaccine successes. The possible use of attenuated HIV-1 would trigger concerns about long-term safety, heightened by instances of AIDS in macaques infected with attenuated SIV (162). The third approach was described by Lifson et al. (163), in which short-term treatment of macaques with ART within 48 h of SIV infection resulted in long-term control of the infection after withdrawal of the drugs and resistance to superinfection. Given the time frame of early virus spread (Figure 1), this type of protection is likely to be immune rather than retroviral interference because virus should not have disseminated at the time of ART introduction and virus suppression. This conclusion is largely conﬁrmed by the ﬁnding that infusion of anti-CD8 antibody resulted in large increases in viremia

of the infecting virus; however, the measured T cell immune responses were low (163), so the exact immune correlates of protection in this model need to be found.

CONCLUSIONS The design of an effective HIV-1 vaccine remains an enormous challenge. Nearly all previous vaccine successes have been built on traditional empirical approaches. So far, these have failed for HIV-1. Immunologists are therefore, for the ﬁrst time, attempting to design a real vaccine. This challenge asks penetrating questions about our understanding of the immune response. Although we have a good primate model of HIV infection, we ultimately have to test the vaccines in humans. Human testing is a very slow process because of the very complex safety and regulatory issues. There is now widespread recognition of the scale of the problem in the scientiﬁc and funding communities (164), and more coordinated approaches involving extensive international collaborations are essential. But this alone will not be enough. New ideas based on deeper understanding of all aspects of adaptive and innate, peripheral and mucosal immune responses are badly needed. The challenge, therefore, is to all of the immunological community.

DISCLOSURE STATEMENT A.J.M is cofounder and shareholder (<2.5%) of Oxxon Therapeutics.

ACKNOWLEDGMENTS The views expressed in this review are my own, but I have been greatly helped and inﬂuenced by extensive discussions with my Oxford colleagues, Tomas Hanke, Lucy Dorrell, Nilu Goonetilleke, and Seph Borrow, and with many other friends in the ﬁeld, notably Sarah Rowland-Jones, Peter Kwong, Bart Haynes, Norman Letvin, Dan Barouch, George Shaw, Beatrice Hahn, Joe Sodroski, Mike Cohen, David Monteﬁore, Wayne Koff, Jill Gilmour, Pat Fast, Brigitte Autran, Bruce Walker, and Ita Askonas.

LITERATURE CITED 1. Willerford DM, Bwayo JJ, Hensel M, Emonyi W, Plummer FA, et al. 1993. Human immunodeﬁciency virus infection among high-risk seronegative prostitutes in Nairobi. J. Infect. Dis. 167:1414–17 www.annualreviews.org • HIV Vaccines

245

ARI

15 February 2006

18:20

2. Gray RH, Wawer MJ, Brookmeyer R, Sewankambo NK, Serwadda D, et al. 2001. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV1-discordant couples in Rakai, Uganda. Lancet 357:1149–53 3. Vittinghoff E, Douglas J, Judson F, McKirnan D, MacQueen K, Buchbinder SP. 1999. Per-contact risk of human immunodeﬁciency virus transmission between male sexual partners. Am. J. Epidemiol. 150:306–11 4. Modjarrad K, Zulu I, Karita E, Kancheya N, Funkhouser E, Allen S. 2005. Predictors of HIV serostatus among HIV discordant couples in Lusaka, Zambia and female antenatal clinic attendants in Kigali, Rwanda. AIDS Res. Hum. Retrovir. 21:5–12 5. Fideli US, Allen SA, Musonda R, Trask S, Hahn BH, et al. 2001. Virologic and immunologic determinants of heterosexual transmission of human immunodeﬁciency virus type 1 in Africa. AIDS Res. Hum. Retrovir. 17:901–10 6. Li Q, Duan L, Estes JD, Ma ZM, Rourke T, et al. 2005. Peak SIV replication in resting memory CD4+ T cells depletes gut lamina propria CD4+ T cells. Nature 434:1148–52 7. Pope M, Haase AT. 2003. Transmission, acute HIV-1 infection and the quest for strategies to prevent infection. Nat. Med. 9:847–52 8. Fong L, Mengozzi M, Abbey NW, Herndier BG, Engleman EG. 2002. Productive infection of plasmacytoid dendritic cells with human immunodeﬁciency virus type 1 is triggered by CD40 ligation. J. Virol. 76:11033–41 9. Fonteneau JF, Larsson M, Beignon AS, McKenna K, Dasilva I, et al. 2004. Human immunodeﬁciency virus type 1 activates plasmacytoid dendritic cells and concomitantly induces the bystander maturation of myeloid dendritic cells. J. Virol. 78:5223–32 10. Geijtenbeek TB, Kwon DS, Torensma R, van Vliet SJ, van Duijnhoven GC, et al. 2000. DC-SIGN, a dendritic cell-speciﬁc HIV-1-binding protein that enhances trans-infection of T cells. Cell 100:587–97 11. Kwon DS, Gregorio G, Bitton N, Hendrickson WA, Littman DR. 2002. DC-SIGNmediated internalization of HIV is required for trans-enhancement of T cell infection. Immunity 16:135–44 12. Pope M, Betjes MG, Romani N, Hirmand H, Cameron PU, et al. 1994. Conjugates of dendritic cells and memory T lymphocytes from skin facilitate productive infection with HIV-1. Cell 78:389–98 13. Douek DC, Brenchley JM, Betts MR, Ambrozak DR, Hill BJ, et al. 2002. HIV preferentially infects HIV-speciﬁc CD4+ T cells. Nature 417:95–98 14. Mattapallil JJ, Douek DC, Hill B, Nishimura Y, Martin M, Roederer M. 2005. Massive infection and loss of memory CD4+ T cells in multiple tissues during acute SIV infection. Nature 434:1093–97 15. Phillips AN. 1996. Reduction of HIV concentration during acute infection: independence from a speciﬁc immune response. Science 271:497–99 16. Borrow P, Lewicki H, Hahn BH, Shaw GM, Oldstone MB. 1994. Virus-speciﬁc CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeﬁciency virus type 1 infection. J. Virol. 68:6103–10 17. Safrit JT, Andrews CA, Zhu T, Ho DD, Koup RA. 1994. Characterization of human immunodeﬁciency virus type 1-speciﬁc cytotoxic T lymphocyte clones isolated during acute seroconversion: recognition of autologous virus sequences within a conserved immunodominant epitope. J. Exp. Med. 179:463–72 18. Ioannidis JP, McQueen PG, Goedert JJ, Kaslow RA. 1998. Use of neural networks to model complex immunogenetic associations of disease: human leukocyte antigen impact on the progression of human immunodeﬁciency virus infection. Am. J. Epidemiol. 147:464–71

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

246

McMichael

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

19. Bhardwaj N. 2001. Processing and presentation of antigens by dendritic cells: implications for vaccines. Trends Mol. Med. 7:388–94 20. Pulendran B, Palucka K, Banchereau J. 2001. Sensing pathogens and tuning immune responses. Science 293:253–56 21. Alter G, Malenfant JM, Delabre RM, Burgett NC, Yu XG, et al. 2004. Increased natural killer cell activity in viremic HIV-1 infection. J. Immunol. 173:5305–11 22. Wilson JD, Ogg GS, Allen RL, Davis C, Shaunak S, et al. 2000. Direct visualization of HIV-1-speciﬁc cytotoxic T lymphocytes during primary infection. AIDS 14:225–33 23. Schmitz JE, Kuroda MJ, Santra S, Sasseville VG, Simon MA, et al. 1999. Control of viremia in simian immunodeﬁciency virus infection by CD8+ lymphocytes. Science 283:857–60 24. Evans DT, O’Connor DH, Jing P, Dzuris JL, Sidney J, et al. 1999. Virus-speciﬁc cytotoxic T-lymphocyte responses select for amino-acid variation in simian immunodeﬁciency virus Env and Nef. Nat. Med. 5:1270–76 25. Borrow P, Lewicki H, Wei X, Horwitz MS, Peffer N, et al. 1997. Antiviral pressure exerted by HIV-1-speciﬁc cytotoxic T lymphocytes (CTLs) during primary infection demonstrated by rapid selection of CTL escape virus. Nat. Med. 3:205–11 26. Jones NA, Wei X, Flower DR, Wong M, Michor F, et al. 2004. Determinants of human immunodeﬁciency virus type 1 escape from the primary CD8+ cytotoxic T lymphocyte response. J. Exp. Med. 200:1243–56 27. Price DA, Goulder PJ, Klenerman P, Sewell AK, Easterbrook PJ, et al. 1997. Positive selection of HIV-1 cytotoxic T lymphocyte escape variants during primary infection. Proc. Natl. Acad. Sci. USA 94:1890–95 28. Pitcher CJ, Quittner C, Peterson DM, Connors M, Koup RA, et al. 1999. HIV-1-speciﬁc CD4+ T cells are detectable in most individuals with active HIV-1 infection, but decline with prolonged viral suppression. Nat. Med. 5:518–25 29. Rosenberg ES, Altfeld M, Poon SH, Phillips MN, Wilkes BM, et al. 2000. Immune control of HIV-1 after early treatment of acute infection. Nature 407:523–26 30. Amyes E, Hatton C, Montamat-Sicotte D, Gudgeon N, Rickinson AB, et al. 2003. Characterization of the CD4+ T cell response to Epstein-Barr virus during primary and persistent infection. J. Exp. Med. 198:903–11 31. McMichael AJ, Rowland-Jones SL. 2001. Cellular immune responses to HIV. Nature 410:980–87 32. Shedlock DJ, Shen H. 2003. Requirement for CD4 T cell help in generating functional CD8 T cell memory. Science 300:337–39 33. Sun JC, Bevan MJ. 2003. Defective CD8 T cell memory following acute infection without CD4 T cell help. Science 300:339–42 34. Sun JC, Williams MA, Bevan MJ. 2004. CD4+ T cells are required for the maintenance, not programming, of memory CD8+ T cells after acute infection. Nat. Immunol. 5:927– 33 35. Zajac AJ, Blattman JN, Murali-Krishna K, Sourdive DJ, Suresh M, et al. 1998. Viral immune evasion due to persistence of activated T cells without effector function. J. Exp. Med. 188:2205–13 36. Jin X, Bauer DE, Tuttleton SE, Lewin S, Gettie A, et al. 1999. Dramatic rise in plasma viremia after CD8+ T cell depletion in simian immunodeﬁciency virus-infected macaques. J. Exp. Med. 189:991–98 37. Phillips RE, Rowland-Jones S, Nixon DF, Gotch FM, Edwards JP, et al. 1991. Human immunodeﬁciency virus genetic variation that can escape cytotoxic T cell recognition. Nature 354:453–59 www.annualreviews.org • HIV Vaccines

247

ARI

15 February 2006

18:20

38. Kelleher AD, Long C, Holmes EC, Allen RL, Wilson J, et al. 2001. Clustered mutations in HIV-1 gag are consistently required for escape from HLA-B27-restricted cytotoxic T lymphocyte responses. J. Exp. Med. 193:375–86 39. Goulder PJ, Phillips RE, Colbert RA, McAdam S, Ogg G, et al. 1997. Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nat. Med. 3:212–17 40. Moore CB, John M, James IR, Christiansen FT, Witt CS, Mallal SA. 2002. Evidence of HIV-1 adaptation to HLA-restricted immune responses at a population level. Science 296:1439–43 41. Goulder PJ, Altfeld MA, Rosenberg ES, Nguyen T, Tang Y, et al. 2001. Substantial differences in speciﬁcity of HIV-speciﬁc cytotoxic T cells in acute and chronic HIV infection. J. Exp. Med. 193:181–94 42. Appay V, Dunbar PR, Callan M, Klenerman P, Gillespie GM, et al. 2002. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections. Nat. Med. 8:379–85 43. Champagne P, Ogg GS, King AS, Knabenhans C, Ellefsen K, et al. 2001. Skewed maturation of memory HIV-speciﬁc CD8 T lymphocytes. Nature 410:106–11 44. Appay V, Nixon DF, Donahoe SM, Gillespie GM, Dong T, et al. 2000. HIV-speciﬁc CD8+ T cells produce antiviral cytokines but are impaired in cytolytic function. J. Exp. Med. 192:63–75 45. Younes SA, Yassine-Diab B, Dumont AR, Boulassel MR, Grossman Z, et al. 2003. HIV1 viremia prevents the establishment of interleukin 2-producing HIV-speciﬁc memory CD4+ T cells endowed with proliferative capacity. J. Exp. Med. 198:1909–22 46. Derdeyn CA, Decker JM, Bibollet-Ruche F, Mokili JL, Muldoon M, et al. 2004. Envelope-constrained neutralization-sensitive HIV-1 after heterosexual transmission. Science 303:2019–22 47. Wei X, Decker JM, Wang S, Hui H, Kappes JC, et al. 2003. Antibody neutralization and escape by HIV-1. Nature 422:307–12 48. Richman DD, Wrin T, Little SJ, Petropoulos CJ. 2003. Rapid evolution of the neutralizing antibody response to HIV type 1 infection. Proc. Natl. Acad. Sci. USA 100:4144 – 49 49. Friedrich TC, Dodds EJ, Yant LJ, Vojnov L, Rudersdorf R, et al. 2004. Reversion of CTL escape-variant immunodeﬁciency viruses in vivo. Nat. Med. 10:275–81 50. Goulder PJ, Watkins DI. 2004. HIV and SIV CTL escape: implications for vaccine design. Nat. Rev. Immunol. 4:630–40 51. Alabi AS, Jaffar S, Ariyoshi K, Blanchard T, van der Loeff MS, et al. 2003. Plasma viral load, CD4 cell percentage, HLA and survival of HIV-1, HIV-2, and dually infected Gambian patients. AIDS 17:1513–20 52. Van Baalen CA, Schutten M, Huisman RC, Boers PH, Gruters RA, Osterhaus AD. 1998. Kinetics of antiviral activity by human immunodeﬁciency virus type 1-speciﬁc cytotoxic T lymphocytes (CTL) and rapid selection of CTL escape virus in vitro. J. Virol. 72:6851–57 53. Girard M, Mahoney J, Wei Q, van der Ryst E, Muchmore E, et al. 1998. Genital infection of female chimpanzees with human immunodeﬁciency virus type 1. AIDS Res. Hum. Retrovir. 14:1357–67 54. Fowke KR, Nagelkerke NJ, Kimani J, Simonsen JN, Anzala AO, et al. 1996. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet 348:1347–51

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

248

McMichael

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

55. Langlade-Demoyen P, Ngo-Giang-Huong N, Ferchal F, Oksenhendler E. 1994. Human immunodeﬁciency virus (HIV) nef-speciﬁc cytotoxic T lymphocytes in noninfected heterosexual contact of HIV-infected patients. J. Clin. Invest. 93:1293–97 56. Rowland-Jones SL, McMichael A. 1995. Immune responses in HIV-exposed seronegatives: Have they repelled the virus? Curr. Opin. Immunol. 7:448–55 57. Rowland-Jones SL, Dong T, Fowke KR, Kimani J, Krausa P, et al. 1998. Cytotoxic T cell responses to multiple conserved HIV epitopes in HIV-resistant prostitutes in Nairobi. J. Clin. Invest. 102:1758–65 58. Goh WC, Markee J, Akridge RE, Meldorf M, Musey L, et al. 1999. Protection against human immunodeﬁciency virus type 1 infection in persons with repeated exposure: evidence for T cell immunity in the absence of inherited CCR5 coreceptor defects. J. Infect. Dis. 179:548–57 59. Clerici M, Levin JM, Kessler HA, Harris A, Berzofsky JA, et al. 1994. HIV-speciﬁc T-helper activity in seronegative health care workers exposed to contaminated blood. JAMA 271:42–46 60. Hladik F, Desbien A, Lang J, Wang L, Ding Y, et al. 2003. Most highly exposed seronegative men lack HIV-1-speciﬁc, IFN-γ-secreting T cells. J. Immunol. 171:2671–83 61. Kaul R, Plummer FA, Kimani J, Dong T, Kiama P, et al. 2000. HIV-1-speciﬁc mucosal CD8+ lymphocyte responses in the cervix of HIV-1-resistant prostitutes in Nairobi. J. Immunol. 164:1602–11 62. Mazzoli S, Trabattoni D, Lo Caputo S, Piconi S, Ble C, et al. 1997. HIV-speciﬁc mucosal and cellular immunity in HIV-seronegative partners of HIV-seropositive individuals. Nat. Med. 3:1250–57 63. Dorrell L, Hessell AJ, Wang M, Whittle H, Sabally S, et al. 2000. Absence of speciﬁc mucosal antibody responses in HIV-exposed uninfected sex workers from the Gambia. AIDS 14:1117–22 64. Kaul R, Rowland-Jones SL, Kimani J, Dong T, Yang HB, et al. 2001. Late seroconversion in HIV-resistant Nairobi prostitutes despite pre-existing HIV-speciﬁc CD8+ responses. J. Clin. Invest. 107:341–49 65. Zhu T, Corey L, Hwangbo Y, Lee JM, Learn GH, et al. 2003. Persistence of extraordinarily low levels of genetically homogeneous human immunodeﬁciency virus type 1 in exposed seronegative individuals. J. Virol. 77:6108–16 66. Mitchell T, Risser R. 1992. Interference established in mice by infection with Friend murine leukemia virus. J. Virol. 66:5696–702 67. Parren PW, Marx PA, Hessell AJ, Luckay A, Harouse J, et al. 2001. Antibody protects macaques against vaginal challenge with a pathogenic R5 simian/human immunodeﬁciency virus at serum levels giving complete neutralization in vitro. J. Virol. 75:8340–47 68. Gauduin MC, Parren PW, Weir R, Barbas CF, Burton DR, Koup RA. 1997. Passive immunization with a human monoclonal antibody protects hu-PBL-SCID mice against challenge by primary isolates of HIV-1. Nat. Med. 3:1389–93 69. Mascola JR, Stiegler G, VanCott TC, Katinger H, Carpenter CB, et al. 2000. Protection of macaques against vaginal transmission of a pathogenic HIV-1/SIV chimeric virus by passive infusion of neutralizing antibodies. Nat. Med. 6:207–10 70. Veazey RS, Shattock RJ, Pope M, Kirijan JC, Jones J, et al. 2003. Prevention of virus transmission to macaque monkeys by a vaginally applied monoclonal antibody to HIV-1 gp120. Nat. Med. 9:343–46 71. Moore JP, Cao Y, Qing L, Sattentau QJ, Pyati J, et al. 1995. Primary isolates of human immunodeﬁciency virus type 1 are relatively resistant to neutralization by monoclonal www.annualreviews.org • HIV Vaccines

249

ANRV270-IY24-08

ARI

15 February 2006

72.

73. 74.

75.

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

76. 77. 78. 79.

80.

81.

82.

83.

83a.

84.

84a.

85.

86.

250

18:20

antibodies to gp120, and their neutralization is not predicted by studies with monomeric gp120. J. Virol. 69:101–9 Pitisuttithum P, Berman PW, Phonrat B, Suntharasamai P, Raktham S, et al. 2004. Phase I/II study of a candidate vaccine designed against the B and E subtypes of HIV-1. J. Acquir. Immune Deﬁc. Syndr. 37:1160–65 Desrosiers RC. 2004. Prospects for an AIDS vaccine. Nat. Med. 10:221–23 Kwong PD, Wyatt R, Robinson J, Sweet RW, Sodroski J, Hendrickson WA. 1998. Structure of an HIV gp120 envelope glycoprotein in complex with the CD4 receptor and a neutralizing human antibody. Nature 393:648–59 Wyatt R, Kwong PD, Desjardins E, Sweet RW, Robinson J, et al. 1998. The antigenic structure of the HIV gp120 envelope glycoprotein. Nature 393:705–11 Chen B, Vogan EM, Gong H, Skehel JJ, Wiley DC, Harrison SC. 2005. Structure of an unliganded simian immunodeﬁciency virus gp120 core. Nature 433:834–41 Kwong PD. 2005. Human immunodeﬁciency virus: refolding the envelope. Nature 433:815–16 Reitter JN, Means RE, Desrosiers RC. 1998. A role for carbohydrates in immune evasion in AIDS. Nat. Med. 4:679–84 Cardoso RM, Zwick MB, Stanﬁeld RL, Kunert R, Binley JM, et al. 2005. Broadly neutralizing anti-HIV antibody 4E10 recognizes a helical conformation of a highly conserved fusion-associated motif in gp41. Immunity 22:163–73 Calarese DA, Scanlan CN, Zwick MB, Deechongkit S, Mimura Y, et al. 2003. Antibody domain exchange is an immunological solution to carbohydrate cluster recognition. Science 300:2065–71 Choe H, Li W, Wright PL, Vasilieva N, Venturi M, et al. 2003. Tyrosine sulfation of human antibodies contributes to recognition of the CCR5 binding region of HIV-1 gp120. Cell 114:161–70 Binley JM, Wrin T, Korber B, Zwick MB, Wang M, et al. 2004. Comprehensive crossclade neutralization analysis of a panel of anti-human immunodeﬁciency virus type 1 monoclonal antibodies. J. Virol. 78:13232–52 Moulard M, Phogat SK, Shu Y, Labrijn AF, Xiao X, et al. 2002. Broadly cross-reactive HIV-1-neutralizing human monoclonal Fab selected for binding to gp120-CD4-CCR5 complexes. Proc. Natl. Acad. Sci. USA 99:6913–18 Kwong PD, Doyle ML, Casper DJ, Cicala C, Leavitt SA, et al. 2002. HIV-1 evades antibody-mediated neutralization through conformational masking of receptor-binding sites. Nature 420:678–82 Scanlan CN, Pantophlet R, Wormald MR, Saphire EO, Stanﬁeld R, et al. 2002. The broadly neutralizing anti-human immunodeﬁciency virus type 1 antibody 2G12 recognizes a cluster of α1 ≥ 2 mannose residues on the outer face of gp120. J. Virol. 76:7306–21 Ofek G, Tang M, Sambor A, Katinger H, Mascola JR, et al. 2004. Structure and mechanistic analysis of the anti-human immunodeﬁciency virus type 1 antibody 2F5 in complex with its gp41 epitope. J. Virol. 78:10724–37 Haynes BF, Fleming J, St. Clair WE, Katinger H, Stiegler G, et al. 2005. Cardiolipin polyspeciﬁc autoreactivity in two broadly neutralizing HIV-1 antibodies. Science 308:1906–8 Vine AM, Heaps AG, Kaftantzi L, Mosley A, Asquith B, et al. 2004. The role of CTLs in persistent viral infection: cytolytic gene expression in CD8+ lymphocytes distinguishes between individuals with a high or low proviral load of human T cell lymphotropic virus type 1. J. Immunol. 173:5121–29

McMichael

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

87. Guidotti LG, Borrow P, Brown A, McClary H, Koch R, Chisari FV. 1999. Noncytopathic clearance of lymphocytic choriomeningitis virus from the hepatocyte. J. Exp. Med. 189:1555–64 88. Altfeld M, Allen TM, Yu XG, Johnston MN, Agrawal D, et al. 2002. HIV-1 superinfection despite broad CD8+ T-cell responses containing replication of the primary virus. Nature 420:434–39 89. Mackewicz CE, Blackbourn DJ, Levy JA. 1995. CD8+ T cells suppress human immunodeﬁciency virus replication by inhibiting viral transcription. Proc. Natl. Acad. Sci. USA 92:2308–12 90. Amara RR, Villinger F, Altman JD, Lydy SL, O’Neil SP, et al. 2001. Control of a mucosal challenge and prevention of AIDS by a multiprotein DNA/MVA vaccine. Science 292:69– 74 91. Barouch DH, Santra S, Schmitz JE, Kuroda MJ, Fu TM, et al. 2000. Control of viremia and prevention of clinical AIDS in rhesus monkeys by cytokine-augmented DNA vaccination. Science 290:486–92 92. Shiver JW, Fu TM, Chen L, Casimiro DR, Davies ME, et al. 2002. Replicationincompetent adenoviral vaccine vector elicits effective anti-immunodeﬁciency-virus immunity. Nature 415:331–35 93. Rose NF, Marx PA, Luckay A, Nixon DF, Moretto WJ, et al. 2001. An effective AIDS vaccine based on live attenuated vesicular stomatitis virus recombinants. Cell 106:539–49 94. Chen X, Scala G, Quinto I, Liu W, Chun TW, et al. 2001. Protection of rhesus macaques against disease progression from pathogenic SHIV-89.6PD by vaccination with phagedisplayed HIV-1 epitopes. Nat. Med. 7:1225–31 95. Horton H, Vogel TU, Carter DK, Vielhuber K, Fuller DH, et al. 2002. Immunization of rhesus macaques with a DNA prime/modiﬁed vaccinia virus Ankara boost regimen induces broad simian immunodeﬁciency virus (SIV)-speciﬁc T-cell responses and reduces initial viral replication but does not prevent disease progression following challenge with pathogenic SIVmac239. J. Virol. 76:7187–202 96. Allen TM, Mortara L, Mothe BR, Liebl M, Jing P, et al. 2002. Tat-vaccinated macaques do not control simian immunodeﬁciency virus SIVmac239 replication. J. Virol. 76:4108– 12 97. Pal R, Venzon D, Letvin NL, Santra S, Monteﬁori DC, et al. 2002. ALVAC-SIV-gag-polenv-based vaccination and macaque major histocompatibility complex class I (A∗ 01) delay simian immunodeﬁciency virus SIVmac -induced immunodeﬁciency. J. Virol. 76:292–302 98. Vogel TU, Reynolds MR, Fuller DH, Vielhuber K, Shipley T, et al. 2003. Multispeciﬁc vaccine-induced mucosal cytotoxic T lymphocytes reduce acute-phase viral replication but fail in long-term control of simian immunodeﬁciency virus SIVmac239. J. Virol. 77:13348–60 99. Barouch DH, Kunstman J, Kuroda MJ, Schmitz JE, Santra S, et al. 2002. Eventual AIDS vaccine failure in a rhesus monkey by viral escape from cytotoxic T lymphocytes. Nature 415:335–39 100. Norbury CC, Basta S, Donohue KB, Tscharke DC, Princiotta MF, et al. 2004. CD8+ T cell cross-priming via transfer of proteasome substrates. Science 304:1318–21 101. MacGregor RR, Boyer JD, Ugen KE, Lacy KE, Gluckman SJ, et al. 1998. First human trial of a DNA-based vaccine for treatment of human immunodeﬁciency virus type 1 infection: safety and host response. J. Infect. Dis. 178:92–100 102. Maecker HT, Rinfret A, D’Souza P, Darden J, Roig E, et al. 2005. Standardization of cytokine ﬂow cytometry assays. BMC Immunol. 6:13 www.annualreviews.org • HIV Vaccines

251

ARI

15 February 2006

18:20

103. MacGregor RR, Ginsberg R, Ugen KE, Baine Y, Kang CU, et al. 2002. T-cell responses induced in normal volunteers immunized with a DNA-based vaccine containing HIV-1 env and rev. AIDS 16:2137–43 104. Vuola JM, Keating S, Webster DP, Berthoud T, Dunachie S, et al. 2005. Differential immunogenicity of various heterologous prime-boost vaccine regimens using DNA and viral vectors in healthy volunteers. J. Immunol. 174:449–55 105. Edupuganti S, Weber D, Poole C. 2004. Cytotoxic T-lymphocyte responses to canarypox vector-based HIV vaccines in HIV-seronegative individuals: a meta-analysis of published studies. HIV Clin. Trials 5:259–68 106. McShane H, Pathan AA, Sander CR, Keating SM, Gilbert SC, et al. 2004. Recombinant modiﬁed vaccinia virus Ankara expressing antigen 85A boosts BCG-primed and naturally acquired antimycobacterial immunity in humans. Nat. Med. 10:1240–44 107. Harro C, Edupuganti S, Goepfert P, Priddya F, Lally M, et al. 2005. Safety and immunogenicity of Adenovirus Type 5 (Ad5) HIV-1 gag vaccines. 12th Conf. Retrovir. Oppor. Infect. Abstr. 504. http://www.retroconference.org/2005/ 108. Sumida SM, Truitt DM, Lemckert AA, Vogels R, Custers JH, et al. 2005. Neutralizing antibodies to adenovirus serotype 5 vaccine vectors are directed primarily against the adenovirus hexon protein. J. Immunol. 174:7179–85 109. Trimble C, Lin CT, Hung CF, Pai S, Juang J, et al. 2003. Comparison of the CD8+ T cell responses and antitumor effects generated by DNA vaccine administered through gene gun, biojector, and syringe. Vaccine 21:4036–42 110. Korber B, Muldoon M, Theiler J, Gao F, Gupta R, et al. 2000. Timing the ancestor of the HIV-1 pandemic strains. Science 288:1789–96 111. Leslie AJ, Pfafferott KJ, Chetty P, Draenert R, Addo MM, et al. 2004. HIV evolution: CTL escape mutation and reversion after transmission. Nat. Med. 10:282–89 112. Leslie A, Kavanagh D, Honeyborne I, Pfafferott K, Edwards C, et al. 2005. Transmission and accumulation of CTL escape variants drive negative associations between HIV polymorphisms and HLA. J. Exp. Med. 201:891–902 113. Peyerl FW, Bazick HS, Newberg MH, Barouch DH, Sodroski J, Letvin NL. 2004. Fitness costs limit viral escape from cytotoxic T lymphocytes at a structurally constrained epitope. J. Virol. 78:13901–10 114. Cocchi F, DeVico AL, Garzino-Demo A, Cara A, Gallo RC, Lusso P. 1996. The V3 domain of the HIV-1 gp120 envelope glycoprotein is critical for chemokine-mediated blockade of infection. Nat. Med. 2:1244–47 115. Graham BS. 2002. Clinical trials of HIV vaccines. Annu. Rev. Med. 53:207–21 116. Townsend AR, Rothbard J, Gotch FM, Bahadur G, Wraith D, McMichael AJ. 1986. The epitopes of inﬂuenza nucleoprotein recognized by cytotoxic T lymphocytes can be deﬁned with short synthetic peptides. Cell 44:959–68 117. Lee JK, Stewart-Jones G, Dong T, Harlos K, Di Gleria K, et al. 2004. T cell crossreactivity and conformational changes during TCR engagement. J. Exp. Med. 200:1455– 66 118. Yu XG, Addo MM, Rosenberg ES, Rodriguez WR, Lee PK, et al. 2002. Consistent patterns in the development and immunodominance of human immunodeﬁciency virus type 1 (HIV-1)-speciﬁc CD8+ T-cell responses following acute HIV-1 infection. J. Virol. 76:8690–701 119. Yewdell JW, Del Val M. 2004. Immunodominance in TCD8+ responses to viruses: cell biology, cellular immunology, and mathematical models. Immunity 21:149–53 120. Kedl RM, Kappler JW, Marrack P. 2003. Epitope dominance, competition and T cell afﬁnity maturation. Curr. Opin. Immunol. 15:120–27

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

252

McMichael

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

121. Klenerman P, Rowland-Jones S, McAdam S, Edwards J, Daenke S, et al. 1994. Cytotoxic T-cell activity antagonized by naturally occurring HIV-1 Gag variants. Nature 369:403–7 122. Mongkolsapaya J, Dejnirattisai W, Xu XN, Vasanawathana S, Tangthawornchaikul N, et al. 2003. Original antigenic sin and apoptosis in the pathogenesis of dengue hemorrhagic fever. Nat. Med. 9:921–27 123. Haeryfar SM, DiPaolo RJ, Tscharke DC, Bennink JR, Yewdell JW. 2005. Regulatory T cells suppress CD8+ T cell responses induced by direct priming and cross-priming and moderate immunodominance disparities. J. Immunol. 174:3344–51 124. Autran B, Carcelain G, Combadiere B, Debre P. 2004. Therapeutic vaccines for chronic infections. Science 305:205–8 125. Dorrell L. 2005. Therapeutic immunisation strategies for the control of HIV-1. Expert Rev. Vaccines 4:513–20 126. Churdboonchart V, Sakondhavat C, Kulpradist S, Ayudthya BIN, Chandeying V, et al. 2000. A double-blind, adjuvant-controlled trial of human immunodeﬁciency virus type 1 (HIV-1) immunogen (Remune) monotherapy in asymptomatic, HIV-1-infected Thai subjects with CD4-cell counts of >300. Clin. Diagn. Lab. Immunol. 7:728–33 127. Birx DL, Loomis-Price LD, Aronson N, Brundage J, Davis C, et al. 2000. Efﬁcacy testing of recombinant human immunodeﬁciency virus (HIV) gp160 as a therapeutic vaccine in early-stage HIV-1-infected volunteers. rgp160 Phase II Vaccine Investigators. J. Infect. Dis. 181:881–89 128. Kundu SK, Katzenstein D, Moses LE, Merigan TC. 1992. Enhancement of human immunodeﬁciency virus (HIV)-speciﬁc CD4+ and CD8+ cytotoxic T-lymphocyte activities in HIV-infected asymptomatic patients given recombinant gp160 vaccine. Proc. Natl. Acad. Sci. USA 89:11204–8 129. Goebel FD, Mannhalter JW, Belshe RB, Eibl MM, Grob PJ, et al. 1999. Recombinant gp160 as a therapeutic vaccine for HIV-infection: results of a large randomized, controlled trial. European Multinational IMMUNO AIDS Vaccine Study Group. AIDS 13:1461–68 130. Veenstra J, Williams IG, Colebunders R, Dorrell L, Tchamouroff SE, et al. 1996. Immunization with recombinant p17/p24:Ty virus-like particles in human immunodeﬁciency virus-infected persons. J. Infect. Dis. 174:862–66 131. Lindenburg CE, Stolte I, Langendam MW, Miedema F, Williams IG, et al. 2002. Longterm follow-up: no effect of therapeutic vaccination with HIV-1 p17/p24:Ty virus-like particles on HIV-1 disease progression. Vaccine 20:2343–47 132. Moss RB, Giermakowska WK, Savary JR, Theofan G, Daigle AE, et al. 1998. A primer on HIV type 1-speciﬁc immune function and REMUNE. AIDS Res. Hum. Retrovir. 14 (Suppl. 2):S167–75 133. Kelleher AD, Roggensack M, Jaramillo AB, Smith DE, Walker A, et al. 1998. Safety and immunogenicity of a candidate therapeutic vaccine, p24 virus-like particle, combined with zidovudine, in asymptomatic subjects. Community HIV Research Network Investigators. AIDS 12:175–82 134. Ogg GS, Jin X, Bonhoeffer S, Moss P, Nowak MA, et al. 1999. Decay kinetics of human immunodeﬁciency virus-speciﬁc effector cytotoxic T lymphocytes after combination antiretroviral therapy. J. Virol. 73:797–800 135. Oxenius A, Price DA, Gunthard HF, Dawson SJ, Fagard C, et al. 2002. Stimulation of HIV-speciﬁc cellular immunity by structured treatment interruption fails to enhance viral control in chronic HIV infection. Proc. Natl. Acad. Sci. USA 99:13747–52 136. Tryniszewska E, Nacsa J, Lewis MG, Silvera P, Monteﬁori D, et al. 2002. Vaccination of macaques with long-standing SIVmac251 infection lowers the viral set point after cessation of antiretroviral therapy. J. Immunol. 169:5347–57 www.annualreviews.org • HIV Vaccines

253

ARI

15 February 2006

18:20

137. Hel Z, Venzon D, Poudyal M, Tsai WP, Giuliani L, et al. 2000. Viremia control following antiretroviral treatment and therapeutic immunization during primary SIV251 infection of macaques. Nat. Med. 6:1140–46 138. Lisziewicz J, Trocio J, Xu J, Whitman L, Ryder A, et al. 2005. Control of viral rebound through therapeutic immunization with DermaVir. AIDS 19:35–43 139. Lu W, Wu X, Lu Y, Guo W, Andrieu JM. 2003. Therapeutic dendritic-cell vaccine for simian AIDS. Nat. Med. 9:27–32 140. Kahn JO, Cherng DW, Mayer K, Murray H, Lagakos S. 2000. Evaluation of HIV1 immunogen, an immunologic modiﬁer, administered to patients infected with HIV having 300 to 549 × 10(6)/L CD4 cell counts: a randomized controlled trial. JAMA 284:2193–202 141. Kinloch S, Perrin L, Hoen B, Lampe F, Phillips A, Goh L, et al. 2004. Evaluation of 2 therapeutic vaccination regimens in HAART-treated primary HIV infection subjects following analytical treatment interruption: QUEST PROB3005, a randomized, placebocontrolled study. 11th Conf. Retrovir. Oppor. Infect. Abstr. 168 142. Dorrell L, Yang H, Iversen AK, Conlon C, Suttill A, et al. 2005. Therapeutic immunization of highly active antiretroviral therapy-treated HIV-1 infected patients: safety and immunogenicity of an HIV-1 gag/polyepitope DNA vaccine. AIDS 19:1321–23 143. Levy Y, Gahery-Segard H, Durier C, Lascaux AS, Goujard C, et al. 2005. Immunological and virological efﬁcacy of a therapeutic immunization combined with interleukin-2 in chronically HIV-1 infected patients. AIDS 19:279–86 144. Kilby JM, Wang R, Mildvan D, Fischl M, Santana-Bagur J, et al. 2005. A phase II randomized partially blinded trial of antiretroviral therapy, HIV speciﬁc immunizations, and IL-2 cycles to promote efﬁcient control of viral replication (ACTG A5024). 12th Conf. Retrovir. Oppor. Infect. Abstr. 525 145. Markowitz M, Jin X, Hurley A, Simon V, Ramratnam B, et al. 2002. Discontinuation of antiretroviral therapy commenced early during the course of human immunodeﬁciency virus type 1 infection, with or without adjunctive vaccination. J. Infect. Dis. 186:634– 43 146. Harrer E, Bauerle M, Ferstl B, Chaplin P, Petzold B, et al. 2005. Therapeutic vaccination of HIV-1-infected patients on HAART with a recombinant HIV-1 nef-expressing MVA: safety, immunogenicity and inﬂuence on viral load during treatment interruption. Antivir. Ther. 10:285–300 147. Kundu SK, Engleman E, Benike C, Shapero MH, Dupuis M, et al. 1998. A pilot clinical trial of HIV antigen-pulsed allogeneic and autologous dendritic cell therapy in HIVinfected patients. AIDS Res. Hum. Retrovir. 14:551–60 148. Lu W, Arraes LC, Ferreira WT, Andrieu JM. 2004. Therapeutic dendritic-cell vaccine for chronic HIV-1 infection. Nat. Med. 10:1359–65 149. Stott J, Almond N. 1995. Assessing animal models of AIDS. Nat. Med. 1:295–97 150. Chan WL, Rodgers A, Grief C, Almond N, Ellis S, et al. 1995. Immunization with class I human histocompatibility leukocyte antigen can protect macaques against challenge infection with SIVmac-32H. AIDS 9:223–28 151. Bogers WM, Bergmeier LA, Ma J, Oostermeijer H, Wang Y, et al. 2004. A novel HIVCCR5 receptor vaccine strategy in the control of mucosal SIV/HIV infection. AIDS 18:25–36 152. Liu R, Paxton WA, Choe S, Ceradini D, Martin SR, et al. 1996. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV-1 infection. Cell 86:367–77

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

254

McMichael

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-08

ARI

15 February 2006

18:20

153. Daniel MD, Kirchhoff F, Czajak SC, Sehgal PK, Desrosiers RC. 1992. Protective effects of a live attenuated SIV vaccine with a deletion in the nef gene. Science 258:1938–41 154. Almond N, Rose J, Sangster R, Silvera P, Stebbings R, et al. 1997. Mechanisms of protection induced by attenuated simian immunodeﬁciency virus. I. Protection cannot be transferred with immune serum. J. Gen. Virol. 78(Pt. 8):1919–22 155. Schmitz JE, Johnson RP, McClure HM, Manson KH, Wyand MS, et al. 2005. Effect of CD8+ lymphocyte depletion on virus containment after simian immunodeﬁciency virus SIVmac251 challenge of live attenuated SIVmac2393-vaccinated rhesus macaques. J. Virol. 79:8131–41 156. Stebbings R, Stott J, Almond N, Hull R, Lines J, et al. 1998. Mechanisms of protection induced by attenuated simian immunodeﬁciency virus. II. Lymphocyte depletion does not abrogate protection. AIDS Res. Hum. Retrovir. 14:1187–98 157. Wyand MS, Manson K, Monteﬁori DC, Lifson JD, Johnson RP, Desrosiers RC. 1999. Protection by live, attenuated simian immunodeﬁciency virus against heterologous challenge. J. Virol. 73:8356–63 158. Stebbings R, Berry N, Stott J, Hull R, Walker B, et al. 2004. Vaccination with live attenuated simian immunodeﬁciency virus for 21 days protects against superinfection. Virology 330:249–60 159. Whatmore AM, Cook N, Hall GA, Sharpe S, Rud EW, Cranage MP. 1995. Repair and evolution of nef in vivo modulates simian immunodeﬁciency virus virulence. J. Virol. 69:5117–23 160. Brenchley JM, Schacker TW, Ruff LE, Price DA, Taylor JH, et al. 2004. CD4+ T cell depletion during all stages of HIV disease occurs predominantly in the gastrointestinal tract. J. Exp. Med. 200:749–59 161. Xu XN, Screaton GR, Gotch FM, Dong T, Tan R, et al. 1997. Evasion of cytotoxic T lymphocyte (CTL) responses by nef-dependent induction of Fas ligand (CD95L) expression on simian immunodeﬁciency virus-infected cells. J. Exp. Med. 186:7–16 162. Baba TW, Liska V, Khimani AH, Ray NB, Dailey PJ, et al. 1999. Live attenuated, multiply deleted simian immunodeﬁciency virus causes AIDS in infant and adult macaques. Nat. Med. 5:194–203 163. Lifson JD, Rossio JL, Piatak M Jr, Parks T, Li L, et al. 2001. Role of CD8+ lymphocytes in control of simian immunodeﬁciency virus infection and resistance to rechallenge after transient early antiretroviral treatment. J. Virol. 75:10187–99 164. Klausner RD, Fauci AS, Corey L, Nabel GJ, Gayle H, et al. 2003. Medicine. The need for a global HIV vaccine enterprise. Science 300:2036–39 165. Saphire EO, Parren PW, Pantophlet R, Zwick MB, Morris GM, et al. 2001. Crystal structure of a neutralizing human IGG against HIV-1: a template for vaccine design. Science 293:1155–59 166. Labrijn AF, Poignard P, Raja A, Zwick MB, Delgado K, et al. 2003. Access of antibody molecules to the conserved coreceptor binding site on glycoprotein gp120 is sterically restricted on primary human immunodeﬁciency virus type 1. J. Virol. 77:10557–65 167. Zwick MB, Komori HK, Stanﬁeld RL, Church S, Wang M, et al. 2004. The long third complementarity-determining region of the heavy chain is important in the activity of the broadly neutralizing anti-human immunodeﬁciency virus type 1 antibody 2F5. J. Virol. 78:3155–61

www.annualreviews.org • HIV Vaccines

255

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:227-255. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo Cytokines and Lymphoid Development Unit, INSERM Unit 668 Institut Pasteur, 75724 Paris, France; email: [email protected]

Annu. Rev. Immunol. 2006. 24:257–86 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090700 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0257$20.00

Key Words commitment, lymphocyte, hematopoietic precursor, transcription factor

Abstract NK cells sit at the crossroads of innate and adaptive immunity and help coordinate tumor immunosurveillance and the immune response against pathogens. Balancing signals to NK cell precursors is crucial for their early development, when transcription factors compete to specify the different lymphocyte subsets. Despite an elaborate schema for NK cell development and differentiation, several major issues remain to be addressed, such as identifying the sites for NK cell maturation and deﬁning the peripheral NK cell niche.

257

ANRV270-IY24-09

ARI

15 February 2006

17:51

INTRODUCTION

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

NKT cell: natural killer T cell Effector functions: the ability of NK cells to spontaneously kill target cell lines in vitro (natural cytotoxicity), to produce cytokines that prime the immune system and mediate antiviral and cytotoxic effects, and to produce chemokines that attract other hematopoietic cells

In the 1970s, several groups described a spontaneous cytotoxic antitumor activity in the spleens of unmanipulated mice and rats and in the peripheral blood of normal human subjects (1–5). Although initially received by the scientiﬁc community with a healthy dose of skepticism, the initial observations were rapidly conﬁrmed, and the term natural killer (NK) cell (4) durably entered the immunological lexicon. NK cells were distinguished from B and T cells (because they lack B and T cells’ characteristic cell surface antigens) and were initially referred to as null lymphocytes (reviewed in 6). The distinction of NK cells from adaptive lymphocytes was further underscored by the normal development and function of NK cells in mice and humans that were unable to rearrange their antigen receptor genes (7, 8). NK cells therefore belong to the innate arm of the immune system, whose cellular actors also include granulocytes, macrophages, and mast cells. NK cells represent one of the ﬁrst lines of immune defense; this ﬁts well with their putative role in tumor immunosurveillance and with their known roles in combating infection, in graft rejection, and in pregnancy (9–13). This review focuses on murine NK cells, although parallels to human NK cell biology are also made.

NK Cell Surface Markers Several antigens were initially described as NK speciﬁc, including the NK1.1 antigen (NKR-P1C, a C-type lectin) in C57BL/6 mice, the integrin α2 (CD49b) recognized by the DX5 antibody in most mice, the CD56 antigen in humans, and the asialoganglioN-tetraosylceramide (asialo-GM1) in most species. We now know that these cell surface antigens are also expressed by T cells, mast cells, and granulocytes. Moreover, murine NK cells also bear signatures of other hematopoietic cell types and express low levels of the dendritic cell (DC) marker CD11c 258

Di Santo

(14, 15), the B cell marker CD45R (B220) (16, 17), and the B and T cell CD2 antigen (16, 18). Some NK cell markers appear developmentally regulated: The integrin αM (CD11b or Mac-1, expressed by monocytes, macrophages, and DC) and the leukosialin CD43 are more highly expressed by splenic NK cells than bone marrow NK cells (19). Table 1 provides a nonexhaustive list of NK cell receptors expressed by mouse and human NK cells. A comprehensive review of all NK cell receptors is beyond the scope of this review, and interested readers are referred to recent reviews of the subject (20–22). A truly NK-speciﬁc antigen has not been identiﬁed in mice, although natural cytotoxicity receptors (NCRs) appear NK-restricted in humans (23). Murine NK cells should be analyzed using a combination of several antibodies detecting pan-NK cell determinants. Murine NK cells can be unambiguously identiﬁed as CD3− NK1.1+ CD122+ cells because (a) all NK cells express CD122, which is required for IL-15 responsiveness and is essential for NK cell generation and peripheral survival (reviewed in 24); (b) NK1.1 expression is attained early during NK cell maturation (18, 19) and appears stable throughout the lifespan of NK cells in vivo; and (c) CD3 allows exclusion of the only other hematopoietic cells that can potentially coexpress NK1.1 and CD122 [including natural killer T cells (NKT cells)]. In mouse strains that are NK1.1− , NK cells can be detected as CD3− DX5+ CD122+ cells, although some NK1.1+ NK cells that are DX5− will be missed (18, 25, 26). The inclusion of CD122 in combination with DX5 is important because many bone marrow and splenic cells have the DX5+ CD122− phenotype and are not bona ﬁde NK cells (C.A.J. Vosshenrich & J.P. Di Santo, unpublished observations).

NK Cell Functions NK cells are characterized by several important effector functions, including their capacity to spontaneously lyse susceptible

ANRV270-IY24-09

ARI

Table 1

15 February 2006

17:51

NK cell receptors in humans (H) and mice (M)

Receptor

Species

Class

Motif/Adaptor

Ligand

CD2 (LFA-2)

H, M

IgSF

Proline-rich domain

CD48, CD58 (LFA-3)

CD11a (LFA-1)

H, M

IgSF

Src family kinases, PI3K

CD54 (ICAM-1), CD102 (ICAM-2)

CD11b (Mac-1)

H, M

IgSF

?

CD54 (ICAM-1)

CD43 (sialoadhesin)

H, M

IgSF

?

?

CD44

H, M

IgSF

?

Hyaluronic acid

CD49b (DX5)

M

IgSF

?

?

CD56 (N-CAM)

H

IgSF

?

?

Lag3

H

IgSF

?

HLA class II

CD16 (FcγRIII)

H, M

IgSF

ITAM/FcγR

Immune complexes

CD25 (IL-2Rα)

H

CytoR

?

IL-2

CD27

H, M

TNFRSF

TRAF

CD70

CD28

H, M

IgSF

YXXM/PI3K

CD80, CD86

CD69

H, M

C-lectin

?

?

CD94/NKG2C, E

H, M

C-lectin

ITAM/DAP12

HLA-E (H), Qa-1b (M)

CD122 (IL-2Rβ)

H, M

CytoR

JAK1, 3/STAT5a, 5b

IL-2, IL-15

CD161

H, M

C-lectin

?

Clr-g (NKR-P1F)

CD226 (DNAM-1)

H

IgSF

?

CD112 (Nectin-2), CD155

CD244 (2B4)

H, M

SLAM

TXYXXV-I/SAP, Fyn

CD48

NKG2D

H, M

C-lectin

YINM/DAP10, PI3K

MICA, B, ULBs (H), Rae1s(H, M), H60 (M)

KIR2S, KIR3S

H

IgSF

ITAM/DAP12

HLA class I

Ly49D, H, P

M

C-lectin

ITAM/DAP12

H-2 class I, MCMV m157 (Ly49H)

NCR (NKp30, 44, 46)

H, M (NKp46)

IgSF

ITAM/FcγR, CD3ζ (H), DAP12

Viral hemagglutinins ( ?)

ILT-1 (Ig-like transcript 1)

H

IgSF

ITAM/FcγR, DAP12

?

IFN-α/βR

H, M

CytoR

JAK1, Tyk2/STAT1, 4

Type I interferons

gp49A

M

C-lectin

ITAM

?

CD85 (ILT-2)

H

IgSF

ITIM/SHP-1

HLA-A, -B, -G

CD94/NKG2A

H, M

C-lectin

ITIM/SHP-1, -2

HLA-E (H), Qa-1b (M)

CD161

M

C-lectin

ITIM/SHP-1, -2

Clr-b (NKR-P1D)

CD244 (2B4)

H, M

SLAM

TXYXXV-I/SAP, Fyn

CD48

KIR2DL, KIR3DL

H

IgSF

ITIM/SHP-1, -2

HLA class I

Ly49 A-C, E-G, I-O

M

C-lectin

ITIM/SHP-1, -2

H-2 class I

KLRG1

M

C-lectin

ITIM/SHP-1, -2

?

TGF-βR

H, M

CytoR

Smad2

TGF-β family

IL-10R

?

CytoR

JAK2, Tyk2/STAT3

IL-10

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Adhesion receptor

Activating receptor

Inhibitory Receptor

www.annualreviews.org • NK Cell Development

259

ANRV270-IY24-09

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Hematopoietic stem cell (HSC): deﬁned functionally as the one cell capable of self-renewing and generating all cell types in the blood Missing-self hypothesis: Klas K¨arre predicted the existence of inhibitory receptors on NK cells that would recognize self MHC molecules. The lack of the relevant MHC on target cells (missing-self) would consequently activate NK cells and result in target cell elimination KIR: killer cell immunoglobulinlike receptor SHP: SH2-containing protein-tyrosine phosphatase ITIM: immunoreceptor tyrosine-based inhibitory motifs NK cell repertoire: the collection of cell surface–expressed receptors on a population of NK cells. Individual NK cells can express combinations of activating and inhibitory NK cell receptors

260

17:51

targets. Natural killing involves exocytosis of perforin- and granzyme-containing cytoplasmic granules via a metabolically active process (reviewed in 27). However, NK cells are equipped with a variety of destructive arms and can also eliminate target cells via FasLand TRAIL-mediated pathways, sometimes in a developmentally related fashion (25, 28). An equally important function of NK cells involves their capacity to promptly produce cytokines [notably IFN-γ, TNF-α, and in some cases GM-CSF or IL-10 (29)] and chemokines [including MIP-1α and -β and RANTES (reviewed in 30, 31)]. Together, these functional activities place NK cells in a position to actively eliminate susceptible targets through multiple, nonredundant mechanisms and to recruit and amplify the inﬂammatory response. NK cell cytokine and chemokine production, as well as their lytic capacity, is stimulated by soluble factors (IL-12, IL-18, IFN-α/β) and is acquired during NK cell maturation. As is discussed in detail below, the mechanisms that control the acquisition of NK cell effector functions are being unraveled and require the expression of several transcription factors (TFs) in developing NK cells and interactions with soluble and membrane-bound factors present in the microenvironment. An outstanding question in NK cell biology involves the diversiﬁcation of NK cells. Do functionally distinct subsets of NK cells exist ? If so, how do they develop and what is their biological relevance? Human NK cells comprise several subsets that are phenotypically and functionally distinct (reviewed in 29). The CD56hi CD16− NK subset appears poised for cytokine secretion and, with a CCR7+ CD62L+ phenotype, can selectively trafﬁc to the lymph nodes (LNs) (32). In contrast, the CD56lo CD16hi subset has more potent cytotoxic activity and may be recruited into inﬂamed tissues (31). Until now, clear evidence for functional diversiﬁcation of NK cells has not been obtained in rodents or in other species. A developmental shift in cytokine production proﬁles has been Di Santo

observed for murine NKT cells in the thymus (33) and proposed for human NK cells that have been derived from hematopoietic stem cell (HSC) in vitro (34). The possibility that NK cells could modify their functional attributes depending on their environmental context could represent another form of NK cell functional plasticity.

NK Cell Receptors Several classes of receptors are employed in the regulation of NK cell effector functions. K¨arre’s initial observations that NK cells could lyse major histocompatibility complex (MHC)-negative targets (35) led to the formulation of the missing-self hypothesis (36) that predicted the existence of NK cell negative regulatory receptors that would interact with MHC ligands and thereby spare target cell destruction by NK cells. Inhibitory receptors (IR) for MHC were subsequently identiﬁed, cloned, and characterized (reviewed in 37, 38), thereby largely fulﬁlling K¨arre’s hypothesis. Multiple classes and alleles of IR for MHC exist in vertebrates [including the most Ly49 receptors in mice, the killer cell immunoglobulin-like receptor (KIR) L family in human, and the CD94/NKG2A complex in both species (reviewed in 37, 39, 40)]. Upon interaction with MHC ligands, these IR recruit SHP-1 and SHP-2 tyrosine phosphatases to immunoreceptor tyrosine-based inhibitory motifs (ITIM) in the cytoplasmic domain of these receptors, thereby blunting activation signals (reviewed in 20, 37, 39). IR for MHC are expressed as NK cell repertoires, with each being expressed by a subset of NK cells and most NK cells expressing at least one IR (reviewed in 39, 40), although there are exceptions (41). NK cells become functionally competent following interactions of IR with self-MHC molecules (41a). Non-MHCbinding IR also exist (Table 1) that broaden the deﬁnition of self (reviewed in 41b). Establishing a diverse IR repertoire on individual NK cells allows the global NK cell pool to carefully monitor self (and foreign) MHC

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

molecules and to react when pathological processes perturb MHC expression. IR do not control NK cell reactivity in isolation, but rather work in concert with diverse activating receptors (AR) (reviewed in 22). In recent years, a plethora of NK cell AR have been identiﬁed; these include a subset of Ly49 receptors in mice, the KIR S family in humans, and the CD94/NKG2C complex and the NKG2D receptor in both species (Table 1; reviewed in 22, 42–44). Signal transduction pathways used by AR couple through DAP10, DAP12, and SAP/Fyn adaptors (reviewed in 20–22, 45). Interestingly, diverse functional activities of NK cells appear hardwired through distinct AR-coupled adaptors (46–49). The concept that NK cell reactivity results from a delicate balance between activating and inhibitory signals is now well accepted (reviewed in 20, 22, 37). Moreover, NK cell triggering is not stereotyped, but can be due to imbalance of activating and inhibitory pathways (including not only MHC receptors, but also cytokines and costimulatory molecules, leading to missing, altered, stressed, or non-self situations; see Reference 50, for example). The mechanisms that control NK cell functional activity have provided new paradigms for understanding other types of self-reactive hematopoietic cells, including NKT cells, γδ T cells, mast cells, and B-1 B cells (51, 52).

NK Cells in Innate and Adaptive Immunity It was initially proposed that NK cells participate in tumor immunosurveillance, and there is evidence that NK cells (in concert with NKT cells) are essential actors that prevent tumor initiation and shape tumor evolution through their capacity to kill and secrete cytokines (reviewed in 53). Interestingly, although the lytic capacity of hematopoietic cells against NK-sensitive targets appears maintained in evolution, most signature features of modern NK cells (activating/ inhibitory MHC receptors and granule-

mediated lysis) are not (54). Recognition systems and effector molecules used to achieve target cell lysis vary depending on the species studied. Thus, NK cells, like their adaptive lymphocyte counterparts, are continually evolving. NK cells are also essential for protection against several types of viral infection, in particular herpesvirus in humans (55) and cytomegalovirus (CMV) in mice (56, 57). The role for NK cells in the antiviral response includes both lytic activity (destruction of virally infected cells) and cytokine secretion (IFN-γ as antiviral and chemokines as proinﬂammatory mediators), which may be dispensed in an organ-speciﬁc fashion (reviewed in 58). Nevertheless, NK cells alone are no match for a mutating virus (mutants emerge that can escape NK cell recognition), which will eventually overwhelm an immune system that lacks an adaptive response (59). Cytokine secretion by activated NK cells has been recently shown to reinforce Th1 differentiation of T cells in secondary lymphoid organs (14, 32, 60). These results demonstrate that NK cells can provide a link between innate and adaptive immunity and can thereby contribute to a concerted and effective immune response. Nevertheless, the cell:cell interactions that mediate the pro-Th1 effect of NK cells are undeﬁned. Knowing the precise physical localization of NK cells, DC, and T cells and their migration patterns before and after immunization could provide clues to the kinetics and directionality of the NK cell effect in the LN.

A MODEL FOR MURINE NK CELL DEVELOPMENT Technological advances have played a decisive role in allowing us to propose and re-formulate models for the NK cell differentiation process. Some of the most important developments have included (a) liquidand stroma-based culture systems for the generation of differentiated hematopoietic cells in vitro; (b) multiparameter cell surface immunophenotyping and puriﬁcation of www.annualreviews.org • NK Cell Development

261

ANRV270-IY24-09

ARI

15 February 2006

17:51

HSC

NKP

STAGE

A

iNK

B

mNK

C

D

E

F

CD122 NKG2D

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

NK1.1 CD94 TRAIL CD51(α α v) Ly49 CD49b (DX5, α 2) CD11b (Mac-1) CD43

Cytotoxicity IFN-γγ

Figure 1 Phenotypic markers of developing NK cells in the mouse. NK precursors (NKP, stage A) are characterized by CD122 and NKG2D expression, but they lack other NK cell markers. Immature NK cells (iNK) upregulate NK1.1 and CD94 and transiently express TRAIL and CD51 (stages B and C). Ly49 receptor repertoires and DX5 are relatively late markers of NK cell differentiation expressed by mature NK cells (mNK, stages D and E). CD11b and CD43 expression increase as NK cells differentiate, although effector functions do not depend on upregulation of these markers.

hematopoietic precursors; (c) gene-modiﬁed mouse models for assessing the roles of a given gene for hematopoietic development in vivo; and (d ) sensitive in vitro and in vivo assays of NK cell function. The pioneer research groups placed NK cells on the map of hematopoiesis by showing that they were transplantable cells from the bone marrow (61, 62) and required growth factors, cytokines, and stromal cell interactions to mature to phenotypically recognizable NK cells 262

Di Santo

with lytic capacity (17, 63–67). With the generation of additional antibodies against NK cell receptors, the map of NK cell development has been reﬁned and redeﬁned (Figure 1), reﬂecting a constant evolution from the previous models (17, 19, 24, 68). The scheme represented in Figure 1 is a composite map of immature and mature NK cells that have been identiﬁed and characterized in different organs (bone marrow, liver, spleen) at different ages in the mouse. It should

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

be emphasized that clear precursor-product relationships have been established between adjacent stages in some, but not all, cases (18, 19, 25, 69). NK cell development can be operationally divided into different stages (Figure 1), although this is obviously arbitrary, as the process of generating mature NK cells from HSC is a continuous process. Nevertheless, the major events in the life of the NK cell considered in this review include (a) commitment of HSC to the NK cell lineage (stage A), (b) phenotypic and functional NK cell maturation (stages B–E), and (c) peripheral NK cell homeostasis (stage F). For each of these steps, I summarize what is known about the differentiation process and highlight the recent key ﬁndings that have shed light on the governing molecular and cellular mechanisms.

NK CELL COMMITMENT—TAKING THE FIRST STEP The ﬁrst step down the road to becoming a mature NK cell involves making a commitment to becoming a natural killer. This point of no return means that other developmental options within the hematopoietic system are no longer available. Hematopoietic precursors having this proﬁle are deﬁned as NK cell precursors (NKP) and result from a sequential loss of pluripotency as HSC differentiate to more committed hematopoietic precursors. Several hematopoietic developmental intermediates have been identiﬁed in the mouse that retain lymphoid (including NK) potential, but have essentially lost the capacity to produce myeloid or erythroid lineages. These lymphoid progenitors include early lymphoid progenitors [ELP; deﬁned as lineage (Lin)negative = CD3− CD19− Ter119− Gr-1− and expressing c-kit and fms-like tyrosine kinase 3 (Flt3) (70)] and common lymphoid progenitors [CLP; deﬁned as Lin− ckit+ IL-7Rα+ (71)]. ELP and CLP are characterized by their ability to give rise to B, T, and NK

cells in culture or upon transfer to recipient mice. Although not solely committed to the NK cell lineage, the analysis of these progenitors has provided important clues about the transcriptional program that is active during lymphoid speciﬁcation (72). Human CLP have a CD34+ CD10+ phenotype [73; see also Blom & Spits (73a) in this volume]. Although these different hematopoietic intermediates have NK potential in vitro and in vivo (70, 71, 73), it is not clear whether developing NK cells must transit through these intermediates in order to mature. For example, bone marrow NKP and peripheral NK cell numbers are normal in c-kit-deﬁcient Vickid mice that completely lack CLP (74, 75). A bipotent T/NK progenitor (TNKP) has been clearly identiﬁed in the murine fetal liver and in the fetal thymus in mice and humans (66, 76, 77). TNKP in the thymus has the Lin− c-kit+ NK1.1+ phenotype in mice and the CD34+ CD7+ CD1a− phenotype in humans (reviewed in 78, 79). In the fetal thymus, TNKP cohabitate with more restricted NKP and T cell progenitors and are probably their immediate precursors. With the development of novel stromal cell–based culture systems expressing Notch ligands (80), the laboratories of Zuniger-Pﬂucker and Petrie have reassessed the T and NK potential of early murine thymocyte subsets (81, 82). The earliest thymocyte precursors (CD44+ CD25− ) harbor both T and NK potential. Interestingly, NK potential also exists (at low frequency) at the CD44+ CD25+ stage (82), when early thymocyte progenitors presumably have received ample Notch signals and already have clear molecular markers of T cell commitment (83). Although these observations could simply represent heterogeneity in the CD44+ CD25+ subset, an alternative interpretation is that NK potential remains possible even following Notch signaling of intrathymic precursors. Notch1 signaling is essential for T cell development in vivo (84), and mice with an inducible deletion in Notch1 lack the earliest T cell committed www.annualreviews.org • NK Cell Development

Homeostasis: a self-regulating process by which biological systems tend to maintain stability while adjusting to conditions that are optimal for survival. When applied to lymphocyte subsets, homeostasis refers to the mechanisms that regulate cell numbers in the peripheral pool Commitment: lineage commitment is a progressive process that results in the loss of the ability of pluripotent precursors to differentiate into multiple lineages NK cell precursors (NKP): deﬁned functionally as cells capable of differentiating into NK cells but not other hematopoietic lineages Developmental intermediates: phenotypically deﬁned cell types known to be part of a developmental pathway; in the hematopoietic system, CLP, ELP, and NKP are more committed lymphoid progenitors derived from HSC that may be intermediates in NK cell development TNKP: bipotent T/NK precursor

263

ARI

15 February 2006

17:51

precursors (A. Wilson, personal communication). Nevertheless, thymic NK cells are present in normal numbers under these conditions, suggesting that thymic NK cells can develop in the absence of TNKP. It seems reasonable to conclude that the thymic bipotent TNKP precursor is mainly involved with T cell development, and its NK potential may be elicited under certain conditions when T cell development fails or is inefﬁcient. Whether NK cells that develop outside the thymus transit via TNKP in the bone marrow, liver, or LN is unknown.

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

Natural Killer Cell Precursors (NKP) NKP have been identiﬁed and characterized in the fetal thymus and the bone marrow of adult mice with the following phenotype: Lin− CD122+ NK1.1− DX5− (18, 85). CD122 acquisition on Lin− hematopoietic precursors appears to correlate with their commitment to the NK lineage since murine NKP has no B, T, myeloid, or erythroid cell potential (18). CD122 expression is critical for allowing NKP to respond to IL-15 stimulation, and NKP likely represent the cellular substrate on which IL-15 acts to promote NK cell development in the bone marrow (17, 86, 87). Interestingly, the generation of NKP from HSC does not require IL-15 [or any common γ chain (γc )-dependent cytokine (87)], thereby deﬁning this ﬁrst stage of NK cell development as IL-15 independent. Upon stimulation by IL-15, NKP rapidly proliferate and differentiate into phenotypically mature NK cells capable of cytotoxicity and inducible cytokine synthesis (18). In mice, NKP are also found in the thymus, spleen, and LN (18, 85; C.A.J. Vosshenrich & J.P. Di Santo, unpublished observations), where they represent a relatively minor lymphocyte subset. Phenotypically characterized NKP are not a homogeneous population, yet they are enriched in NK potential [one in eight NKP give rise to NK cells in vitro (18)]. Freshly isolated NKP are not cytotoxic for YAC1 targets. Nevertheless, 10% 264

Di Santo

of NKP express NKG2D at the cell surface (87), suggesting that true NKP may have the Lin− CD122+ NKG2D+ NK1.1− DX5− phenotype. Whether NKG2D+ NKP have lytic potential remains to be tested. NKP in humans have not been rigorously characterized, but they are assumed to play an analogous role. Attempts have been made to identify human NKP after stimulation of hematopoietic precursors with IL-2 or IL-15 in liquid culture alone or in combination with stromal cells (64, 65, 88, 89). Still, this approach reveals not only NKP but also immature and mature NK cells. Human NKP should, like their murine counterparts, express CD122 but not other lineagespeciﬁc markers. Nevertheless, CD122 expression by human lymphoid progenitors (CD34+ CD38+ CD10+ ) has been difﬁcult to demonstrate, despite the fact that these precursors have NK cell potential and are highly responsive to IL-15 (73, 89). A recent report from the Caligiuri laboratory identiﬁed what may represent one type of human NKP (88). A novel CD34dim CD45RA+ integrin α4 β7 hi hematopoietic precursor was shown to differentiate on stromal cells in the presence of IL-2 or IL-15 into mature CD56hi NK cells with lytic capacity and cytokine production. These precursors were present in the bone marrow and peripheral blood and were highly enriched in LNs. LN homing is driven by CCR7 ligands and requires CD62L expression to interact with high endothelial venules. The α4 β7 hi NK precursors were CCR7+ CD62L+ , leading the authors to propose that these cells migrate from the bone marrow via the blood to reach the LN, where they are local precursors of LN-resident CD56hi NK cells (88). The α4 β7 hi NK precursors bear similarities to murine and human lymphoid tissue inducer cells that also have DC and NK cell potential (90, 91). It will be interesting to know whether the α4 β7 hi precursors identiﬁed by Freud and colleagues (88) are functional lymphoid inducer cells with DC potential.

ANRV270-IY24-09

ARI

15 February 2006

17:51

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Generating NKP: Soluble and Membrane Factors The signals that regulate the HSC to NKP transition are poorly understood but involve cell:cell interactions that license NKP to respond to soluble factors elaborated in the microenvironment. The latter quality depends on NKP expression of several distinct TFs. NKP generation likely involves interactions between HSC and stromal cells in vivo because highly puriﬁed human or murine HSC generate few NK cells in liquid culture (even with high levels of IL-15), whereas NK cell development is greatly enhanced when HSC are cultured on relevant stromal cell lines (63– 65, 89, 92, 93). In this way, the process of NK commitment bears similarities to early B cell commitment, for which the earliest stages have also been described as stromal cell dependent (94). The molecular signals delivered by stromal cells that induce NK commitment are unknown but must include those that control CD122 expression. Cytokine signals play a general role in lymphoid commitment, including those delivered via Flt3, c-kit, or γc -dependent receptors. These cytokines could inﬂuence NK commitment because NKP express these receptors (87), and culture of mouse or human HSC with Flt3L or stem cell factor was reported to induce CD122 responsiveness (89, 93). Nevertheless, absolute numbers of NKP are unaffected in c-kit mutants (75) and in γc -deﬁcient mice (87) and are only slightly reduced in the absence of Flt3 (C.A.J. Vosshenrich & J.P. Di Santo, unpublished observations). Individually, these pathways are therefore not essential for the ﬁrst step toward the NK lineage in vivo. This process may have some level of redundancy that could be revealed by studying the appropriate double-mutant mice. Interactions between lymphotoxin (LT) α1 β2 -expressing hematopoietic cells and LTβ receptor (LTβR)-expressing stromal cells play an important role in the generation of NK and NKT cells and in the formation of secondary lymphoid organ architecture (re-

viewed in 95). Mice deﬁcient in LTα or LTβR have slightly reduced absolute numbers of peripheral NK cells (96–98), which has been correlated with a decreased capacity of mutant bone marrow cells to generate NK cells upon culture in IL-15 (99). Whether this defect in NK cell production is accompanied by a decrease in NKP in LTα or LTβRdeﬁcient mice has not been analyzed. An intrinsic abnormality of the stromal cell compartment (which fails to support normal NK cell generation after transfer of HSC from wild-type mice) has been demonstrated (99). LTβR signals in thymic stromal cells result in a NIK (NF-κB-inducing kinase)-mediated activation of RelB, which is required for normal NKT cell development (100). A similar mechanism may be required for normal stromal cell function, a deﬁciency of which underlies the defective NK cell development in the absence of LTα/LTβR. Vitamin D3 upregulated protein 1 (VDUP-1) is a stress-response gene that was initially identiﬁed in vitamin D3–treated HL-60 leukemic cells (101). Recently, VDUP-1-deﬁcient mice were characterized and demonstrated a selective defect in NK and NKT cells (102). Bone marrow, spleen, and LN NK cells were substantially reduced in the absence of VDUP-1, which could be traced to a defect in CD122 expression. NKP could not be detected in the bone marrow of VDUP-1-mutant mice, although some mature NK cells were found in the spleen. Interestingly, vitamin D3 increased the yield of NK cells from IL-15-supplemented stromal cell/HSC cultures, which was correlated with increased CD122 expression. These results suggest a model in which regulation of VDUP-1 could be a determinant of NK cell commitment, through control of CD122 expression and IL-15 sensitivity.

Generating NKP: Transcription Factors TFs play determinant roles in hematopoietic lineage speciﬁcation, and distinct TFs have www.annualreviews.org • NK Cell Development

265

ARI

15 February 2006

17:51

been implicated in early lymphocyte development, including the Ets family members PU.1 and Ets-1 and members of the Ikaros zinc-ﬁnger family, including Ikaros, Helios, and Aiolos (reviewed in 103, 104). Deﬁciency in PU.1 affects all lymphocyte lineages [as well as certain myeloid cells (105)]. The lymphoid defect in PU.1-deﬁcient mice can be traced to deﬁciencies in expression of the IL7Rα chain and of the Flt3 receptor (106, 107), which are crucial for B and T cell development (reviewed in 108) but less so for NK cells (109). Transfer of PU.1-deﬁcient HSC generated markedly reduced numbers of NKP (perhaps secondary to reduced numbers of lymphoid progenitors) and a correspondingly diminished peripheral NK cell pool, although the phenotype and cytotoxic activity of residual PU.1-deﬁcient NK cells was largely conserved (110). These results suggest that PU.1 acts early in the NK cell lineage but may be dispensable once NK cells have fully matured, a notion supported by analysis of PU.1-GFPreporter mice (111). Ets-1 is critical for NK cell development (112). In its absence, NK cells are largely absent in the bone marrow, spleen, and LNs, with subsequent defects in NK cell cytotoxicity and cytokine production. Because IL-15 cannot rescue the NK cell defect in Ets-1-deﬁcient bone marrow cultures (112), NKP are likely severely reduced or absent in this setting. Concerning Ikaros, Helios, and Aiolos, so far only a deﬁciency in Ikaros appears to have a negative effect on NK cell development, which may be related to a diminished expression of Flt3 and CD122 (113). Whether Ikaros is required for the generation of NKP or at later stages of NK cell differentiation has not been reported. The inhibitors of DNA binding (Id proteins) are a family of four helix-loop-helix (HLH) proteins that can heterodimerize with the basic HLH E-box TF, including E2A (comprising E12 and E47), E2-2, and HEB, and thereby inhibit their transcriptional activity (reviewed in 114). Relative E-box and Id activities in developing lymphocyte precursors appear to determine cell fate decisions:

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

266

Di Santo

E2A-mutant mice lack B cells (115), and overexpression of Id2 or Id3 can inhibit B and T cell development from HSC (116), whereas Id2-mutant mice lack NK cells (117). The mechanism by which E-box/Id TF ratios control NK cell development is unclear. Thymic NKP appear to be severely reduced in Id2deﬁcient mice (118), suggesting that an excess of E-box transcriptional activity inhibits NK cell generation at least at the level of the NKP. Because E-box TF can interact with histoneacetyltransferase (HAT) complexes (119), the choice to become a B cell (or a T cell) may involve an active chromatin remodeling event with transcriptional activation through E-box interactions with HAT. In this case, NK cells would be generated when E-box TF levels are insufﬁcient to drive early lymphocyte precursors to the B or T cell lineage (Figure 2). This hypothesis is consistent with the lack of appreciable NK cell phenotype in mice deﬁcient in E2A (120) and with the reduced expression of E-box transcripts in thymic NK cells (121). The process of generating NKP from HSC appears tightly controlled and results from a careful balance of extracellular signals derived from stromal cells that cooperate with optimal levels of critical TFs. The identiﬁcation and characterization of murine and putative human NKP provide a means to understand more fully this key step in the process of NK cell development.

FROM NKP TO MATURE NK CELL—ARE WE THERE YET? Once committed to the NK cell lineage, NKP must acquire the phenotypic and functional qualities that characterize mature peripheral NK cells. The already perplexing process of NK cell maturation has come to seem more complex in recent years as investigators have identiﬁed several new markers that subdivide differentiating NK cells (Figure 1) and characterized several mutant mice that manifest partial defects in NK cell differentiation (Table 2). How do we deﬁne an immature NK cell (iNK)? The broadest deﬁnition

ANRV270-IY24-09

ARI

15 February 2006

17:51

E2A

E2A

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B

pro-T

T

Notch, HEB

ELP

Ikaros

HSC

CLP

pro-B

PU.1 Id2

Id2

NKP

NKP

Id2

NKP

NK Figure 2 A balance of E-box and Id proteins determines NK cell fate. HSC commit to the lymphocyte lineage and generate early lymphoid progenitors (ELP) and common lymphoid progenitors (CLP) under the inﬂuence of the TFs Ikaros, PU.1, and E2A. B and T cell development is associated with increased expression of E-box TFs (E2A, HEB), whereas NK cell development requires Id2, which counterbalances E-box TF activity. Notch activation subsequently promotes T cell development at the expense of B cells.

includes any NK cell that lacks the complete phenotypic and functional attributes of mature peripheral NK cells (having uniform expression of CD122, NK1.1, DX5, NKG2D, and CD16, with variable expression of Ly49 receptors and CD94/NKG2A/C and high expression of CD11b and CD43; stage F). Using these criteria, iNK cells have been identiﬁed in the liver that express NK1.1 but not DX5, Ly49 receptors, or CD11b (25). Conspicuously, liver iNK cells express TRAIL on their cell surface and can use this effector molecule to lyse susceptible targets. These results in the mouse recall earlier studies on human NK cells differentiated in vitro, in which CD161+ CD56− iNK cells were characterized that could kill targets in a TRAIL-dependent, perforin-independent fashion (67). TRAIL+ iNK cells could further develop into TRAIL−

progeny that express Ly49s and DX5 and produce cytokines. Whether TRAIL+ iNK cells in the liver develop in situ or derive from bone marrow NKP is not known. One possibility is that the liver represents an important site for iNK cells to ﬁnish their maturation, a process that could depend on transcription factors that specify hepatic-homing potential (109). Kim and colleagues (19) identiﬁed several iNK cells in the bone marrow and liver. Some of these iNK cells lack Ly49 receptors and DX5 and co-express CD51 (corresponding to Figure 1, stages B and C). Others have a more mature NK cell phenotype (DX5+ Ly49+ ) but are CD11blo CD43lo (Figure 1, stages D and E). Although these cells show lower cytotoxicity and cytokine production compared with their CD11bhi CD43hi counterparts, they nevertheless do exert NK effector function, so www.annualreviews.org • NK Cell Development

iNK cell: immature NK cell

267

ANRV270-IY24-09

Table 2

ARI

15 February 2006

17:51

Genetic mutations affecting the maturation of NK cells

Genes

Phenotype

Repertoire

Stage

Cytotoxicity

Cytokines

Reference

IL-15

CD11blo , CD43lo

Normal

E

Reduced

Normal

87

LTα

?

Dec Ly49A, C/1, D, G2

?

Reduced

?

97

LTβR

?

Normal

?

?

?

98

VDUP-1

?

Absent

?

Reduced

?

102

Syk/ZAP70

Normal

Normal

F

Normal (except ITAM)

Normal (except ITAM)

16

DAP12

Normal

Normal

F

Normal (except Ly49D)

?

138

DAP10

Normal

Normal

F

Reduced (NKG2D)

?

136

SAP/Fyn

Normal

Normal

F

Normal (except 2B4)

Normal (except 2B4)

135

Tyro 3 family

CD11blo , CD43lo

Dec Ly49D

C

Reduced

Reduced

143

PLCγ2

Normal

Red Ly49C/I, G2

E

Absent

Normal

48, 141b

Vav family

Normal

Normal

F

Reduced

Reduced

47

Soluble Factors

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Signaling

Transcription Factors PU.1

Normal

Dec Ly49A, D

A

Normal

Normal

110

Ets-1

?

?

A

Reduced

Reduced

112

Id2

?

?

A

Reduced

Reduced

117

Gata-3

CD11blo , CD43lo

Dec Ly49C/I, D,

C

Normal

Reduced

109

IRF-2

CD11blo , CD43lo

Dec Ly49D, C/I, G2

D

Normal

Reduced

129

T-bet

CD11blo , CD43lo

Normal

E

Normal

Normal

145

MEF, MITF, CEBP-γ

Normal

Normal

F

Reduced

Normal

24, 68

they could simply represent a subset of mature NK cells with a different functional capacity. Care should be taken when applying the label immature to NK cells based solely on lower CD11b and CD43 expression levels. One way to address the functional consequences of NK cell heterogeneity is to isolate and study the cells that have nonuniform expression. For NK cells, this includes cells bearing the CD11b, CD43, CD45R, CD69, CD90, and c-kit markers. Although this can be achieved (with some difﬁculty), an alternative solution involves culture systems that expand and maintain iNK cells. Two recent reports suggest that this may be possible for human and mouse NK cell precursors (122, 268

Di Santo

123). Both rely on the use of IL-7 to maintain cell viability and generate poorly lytic NK cells that robustly respond to IL-15 exposure with additional proliferation, differentiation, and cytotoxic potential. Using these new culture systems, questions relating to the roles of TFs, soluble and membrane factors, and precursor-product relationships between different subsets of developing NK cells may be answered.

NK Maturation: Soluble and Membrane Factors IL-15 is the essential fuel for NK cell development and for maintenance of NK cells in

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

the periphery (124–126). IL-15 is presented by the IL-15Rα subunit (127, and see below) and is produced by hematopoietic and nonhematopoietic stromal cells in an IFN regulatory factor (IRF)-1-dependent fashion (128). The analysis of the few residual NK cells in IL-15-deﬁcient mice revealed that some degree of NK maturation could proceed in the absence of this cytokine (87, 97), although the published results were not always in complete agreement. Vosshenrich et al. (87) found that NK cells in Rag2-deﬁcient x IL-15-deﬁcient mice express Ly49A, D, G2, and CD94 at near normal levels and have lytic- and cytokine-producing potential, despite a CD11blo CD43lo phenotype, a result conﬁrmed independently (129). In contrast, the Kumar laboratory (97, 130) found abnormalities in Ly49 receptor expression in IL-15deﬁcient mice. Exogenous IL-15 administration can correct the defect in NK cell numbers (86) and the Ly49 receptor repertoire in IL-15-deﬁcient mice (97), implying that IL-15 might control Ly49 receptor acquisition. Still, it is not clear at what stage IL-15 is acting under these conditions (NKP, iNK, etc.) or whether IL-15 is acting directly or indirectly to inﬂuence the overall numbers of Ly49+ NK cells. The mechanism by which IL-15 affects NK cells (via survival, expansion, or differentiation) remains enigmatic and may be different for NK cells and their precursors. A report from the Raulet laboratory (131) demonstrated that mature NK cells could expand normally in irradiated IL-15-deﬁcient hosts and suggested that IL-15 was not a major proliferative factor for NK cells. Whether this holds true for iNK cells is not known. The residual iNK cells in IL-15-deﬁcient mice have normal Bcl-2 levels and mitochondrial membrane potential (87) and therefore show no overt signs of ongoing apoptosis; this contrasts with mature NK cells that lose Bcl-2 expression when transferred to an IL15-deprived environment (126, 132). Thus, IL-15 responses of immature and mature NK cells are clearly different.

Beyond the role for LTα/LTβR in the generation of NKP (described above), conﬂicting results exist concerning the role for LTα/LTβR in the formation of the NK cell Ly49 receptor repertoire (97, 98). Lian and colleagues (97) studied LTα-deﬁcient mice and found alterations in Ly49 expression (two- to threefold decreases in percentages of Ly49A-, C-, D-, and G-expressing NK cells). Stevenaert and colleagues (98) studied LTβR-deﬁcient mice and found no evidence for perturbations in the Ly49 repertoire. A trivial explanation for these differences involves technical approaches (the use of DX5 versus NK1.1 to study NK cells, and the absence of co-staining with CD122). Stevenaert raises the alternative explanation that another LTβR ligand—LIGHT—may play a negative regulatory role in this process. Further studies will be required to decipher the relative roles of LIGHT versus LTα signals not only in generating the NK cell Ly49 receptor repertoire, but also in understanding how this pathway promotes the full phenotypic and functional maturation of NK cells (133). Soluble factors also drive the maturation of human NK cells. The relationship between the two well-deﬁned human NK cell subsets (CD56hi CD16− , CD56dim CD16+ ) remains enigmatic. A recent report from Ferlazzo and colleagues (134) demonstrated that IL-2 can promote the acquisition of CD16 and KIR on CD56hi CD16− NK cells. Because the latter are selectively enriched in the LN, one possibility is that T cell–derived IL-2 is involved in the conversion of CD56hi CD16− NK cells to the CD56dim CD16+ phenotype in vivo (88, 134). This would represent a clear example of adaptive immunity inﬂuencing innate immune development.

IRF: IFN regulatory factor

NK Maturation: Signaling Pathways Most activating receptors on NK cells (Table 1) couple to one of three primary signaling pathways: DAP12/Syk-ZAP70, DAP10/PI-3K, or SAP/Fyn. Phospholipase Cγ (PLCγ)-1 and PLCγ2 and Vav family www.annualreviews.org • NK Cell Development

269

ANRV270-IY24-09

ARI

15 February 2006

17:51

members are subsequently activated to transmit these membrane events to NK cell effector functions (reviewed in 20, 21, 45). Recent reports provide evidence for selective targeting of NK cell cytotoxicity through some of these biochemical pathways (Table 2). NK cells develop normally in the absence of Syk-ZAP70, DAP10, or SAP/Fyn (16, 135, 136), demonstrating that individually these signaling pathways are redundant for normal NK cell generation. This result is in marked contrast with B and T cell development, which require signaling through immunoreceptor tyrosine-based activating motif (ITAM)-containing antigen receptors (16). In contrast, cytotoxicity and cytokine production are completely inhibited when NK cells are triggered via CD16 or Ly49D in Syk-ZAP70-deﬁcient mice, via NKG2D in DAP10-deﬁcient mice, or via 2B4 in SAP/Fyn-deﬁcient mice (16, 135, 136). Further analyses of NK cells in SykZAP70-, DAP10-, or SAP/Fyn-mutant mice have provided new insights into the receptors responsible for natural cytotoxicity. SykZAP70-deﬁcient NK cells kill YAC1 targets normally, demonstrating that ITAM-bearing receptors are not required for lysis (16); YAC1 killing can be inhibited by blocking NKG2D (137), and YAC1 killing is compromised in the absence of DAP10 (136), suggesting that NKG2D/DAP10 signaling (138) is the major receptor involved in YAC1 recognition and lysis. Interestingly, Syk-ZAP70-deﬁcient NK cells could lyse targets that do not express NKG2D ligands (16), thereby suggesting additional pathways that are involved in natural cytotoxicity. The cytosolic protein adaptor SAP is necessary to couple activating signals from several NK cell receptors (2B4, NTB-A; reviewed in 21). Recent reports from the Latour and Kumar laboratories demonstrate that 2B4:CD48 interactions that signal through a SAP/Fyn cascade operate independently of NKG2D activation (135, 139). Whether 2B4/SAP/Fyn activation provides an ITAM- and NKG2D-independent path-

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ITAM: immunoreceptor tyrosine-based activation motifs

270

Di Santo

way for NK cell–mediated cytotoxicity remains to be tested. PLCγ is a central regulator of intracellular calcium mobilization and protein kinase C activation that is engaged following triggering of different lymphocyte receptors (reviewed in 140). Previous studies have demonstrated a nonredundant role for PLCγ2 in NK cytotoxicity (141); whether this molecule played other roles in NK cell development was not fully appreciated. Work from the Colonna and Colucci laboratories shows that PLCγ2 controls not only NK cell cytotoxicity but also inﬂuences the formation of the Ly49 receptor repertoire (48, 141a). How PLCγ2 controls cytotoxicity is not known, but evidence points to a defect in the process of granule polarization and secretion. These NK abnormalities have biological consequences in vivo, as PLCγ2-deﬁcient mice fail to control experimental CMV infection and to reject tumors. PLCγ2 therefore plays an essential role in establishing a diverse NK cell pool and in maintaining distinct NK cell functions in mature NK cells. Work from the Leibson laboratory indicates that ITAM-containing receptors in human NK cells can use either PLCγ1 or PLCγ2, whereas NKG2D only activates PLCγ2, demonstrating that certain receptor pathways may be hardwired (49, 142). This observation provides a potential means to selectively target NKG2D function in cytotoxic cells. Similar observations were made in mice deﬁcient in the three different members of the Vav family of guanine nucleotide exchange factors (47). In this case Vav1 was selectively coupled to NKG2D, whereas Vav2 and 3 were required for ITAM-mediated signaling. Interestingly, mice lacking all three Vav family members developed normal numbers of phenotypically mature NK cells (47). A recent NK meeting report suggests that receptors of the Axl protein tyrosine kinase (PTK) family are involved in NK cell maturation in the bone marrow (143). A differential transcript analysis identiﬁed Axl in functionally distinct subsets of NK cells. Axl is a

ANRV270-IY24-09

ARI

15 February 2006

17:51

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

member of the Tyro3 family of PTKs (which includes Tyro3, Mer, and Axl) that control diverse biological processes, including spermatogenesis, inﬂammation, and possibly malignant transformation (reviewed in 144). Ligands for Axl include Gas6 and Protein S, which are expressed by stromal cells. Interestingly, NK cells require Axl PTKs for their functional maturation and for their acquisition of Ly49 receptor repertoire (143). These results identify a new signaling axis that is crucial for proper NK cell differentiation in mice.

NK Maturation: Transcription Factors Several TFs (see Table 2) guide the process of NK cell maturation, and their absence results in incomplete or aberrant NK cell differentiation (reviewed in 24, 68). Closer examination of the phenotype and function of NK cells in TF mutant mice reveals that certain TFs appear to exert their roles during the iNK stage in the bone marrow (Gata-3, IRF-2, T-bet), whereas others inﬂuence the effector functions of mature NK cells [myeloid ELF1-like factor (MEF), the microphthalmia-associated TF (MITF), and the CCAAT/enhancer-binding protein-γ (CEBP-γ); Figure 3]. There are some striking similarities in the phenotype of bone marrow NK cells from some of these different TF mutant mice, suggesting that certain aspects of the NK cell differentiation process may be linked by TF networks (68). Gata-3 and the T-box TF T-bet have opposing roles in T helper cell differentiation, yet these TFs are co-expressed in developing NK cells at the population (145) and singlecell level (J.P. Di Santo, unpublished observations). The comparison of Gata-3-deﬁcient and T-bet-deﬁcient NK cells suggests a complex interaction between these two TFs during NK cell development. Previous studies on NK cells from chimeras generated with Gata-3-deﬁcient HSC demonstrated that this TF is required for proper expression of CD11b, CD43, and

TRANSCRIPTION FACTORS

NKP

iNK

mNK

Ets-1

Gata-3

CEBP-γ

Id2

IRF-2

MEF

Ikaros

T-bet

MITF

PU.1 Figure 3 TFs that condition NK cell development, maturation, and effector functions. Several TFs have been identiﬁed that play an important role in the generation of NKP (Ets-1, Id2, Ikaros, PU.1), the further maturation of iNK (Gata-3, IRF-2, T-bet), or the functional differentiation of mNK (CEBP-γ, MEF, MITF). The interplay between these TFs remains to be fully elucidated.

Ly49 receptors on developing NK cells (109). Although abnormal in phenotype, Gata3-deﬁcient NK cells were fully cytotoxic, demonstrating that the acquisition of some effector functions could be achieved in the absence of a mature NK cell phenotype. Gata3-deﬁcient NK cells were poor producers of IFN-γ, and this defect was associated with reduced levels of T-bet and Hlx (109). Thus, Gata-3 promotes several different aspects of the NK cell differentiation program (reviewed in 68), and its expression is linked to T-bet expression. In comparison, T-bet-deﬁcient NK cells showed a similar phenotypic abnormality (CD11blo CD43lo ) and were normally cytotoxic. Interestingly, NK cells that develop in the absence of T-bet were unable to sustain IFN-γ production after stimulation (145). The residual IFN-γ production may result from the compensatory expression of another T-box TF, Eomesodermin, in T-bet-deﬁcient NK cells (145). It would be interesting to know whether Eomesodermin expression is modiﬁed in Gata-3-deﬁcient NK cells. A recent study reported a more complete analysis of NK cell differentiation in mice deﬁcient in IRF-2 (129). Previous work had shown that IRF-2 is required in a www.annualreviews.org • NK Cell Development

271

ARI

15 February 2006

17:51

cell-intrinsic fashion for normal development of NK cells (146). Taki and colleagues (129) found that IRF-2-deﬁcient NK cells are poor IFN-γ producers, which recalls previous reports of the defect in Th1 development seen in these mice (146). They further demonstrated that IRF-2 plays an important role in NK cell development in the bone marrow and is required for normal CD11b and CD43 expression. NK cells in the bone marrow of IRF-2-deﬁcient mice have an immature DX5+ Ly49+ CD11blo CD43lo phenotype (Figure 1, stage D), although they appear normally cytotoxic against sensitive targets (129). In contrast, splenic NK cells are strongly reduced in number, have an even less mature DX5− phenotype, and lack Ly49 receptor expression altogether (Figure 1, stage C). Fully mature NK cells are basically absent in the bone marrow and spleen of IRF-2deﬁcient mice. This result suggests that IRF-2 plays a major role in maintaining developing NK cells in the bone marrow after stage D. The presence of the few immature stage C NK cells in the spleen of IRF-2-deﬁcient mice suggests that NK cells can export the bone marrow before their differentiation is complete. TFs likely act in a sequential fashion to promote NK cell maturation such that different TFs specify the transcriptional proﬁles that could be associated with unique functional outcomes (Figure 3). For example, early stages of NK cell development in the bone marrow and liver are characterized by TRAIL expression (Figure 1, stage B), which does not persist as NK cells mature and take up residence in the spleen. One may speculate that certain TFs encode TRAIL expression at this stage, and transcriptional proﬁling of developing NK may provide an answer. Nevertheless, the interplay between Gata-3, IRF-2, and T-bet appears complex. Some hierarchical organization can be proposed with a sequential (Gata-3 to IRF-2 to T-bet) TF activation cascade. Gata-3 may control CD11b and CD43 expression through IRF-2 and/or T-bet, since NK cells from mice

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

272

Di Santo

lacking these TFs share the same phenotype (129, 145). Moreover, Gata-3 expression in IRF-2-deﬁcient NK cells and IRF-2 expression in T-bet-deﬁcient NK cells are normal (129, 145). The effector functions in phenotypically mature NK cells are insured by a group of TFs that include the MEF, MITF, and CEBP-γ (Figure 3). Eomesodermin may also belong in this group, although the phenotype of Eomesodermin-deﬁcient NK cells has not yet been reported. NK cells deﬁcient in MEF, MITF, or CEBP-γ develop normally but have reduced cytotoxicity and cytokine production. The molecular mechanisms include decreased perforin or granzyme expression, providing an example of how TFs can promote speciﬁc functional outcomes in fully differentiated NK cells. One conclusion from the analysis of TF mutants is that phenotypically immature NK cells can attain functional cytolytic competence. This may reﬂect a degree of functional heterogeneity within the NK cell pool: NK cells may be capable of eliciting independent or linked effector functions. T-bet and Gata-3 co-expression in NK cells could be an example of a mechanism to maximize options for programming diverse effector functions. NK cells lacking classical functions (cytotoxicity, IFN-γ secretion) are not necessarily anergic but may nonetheless contribute to immune responses in other ways (for example, as regulatory NK cells secreting chemotactic factors).

HOMEOSTASIS OF THE PERIPHERAL NK CELL POOL Mature NK cells are found in the blood, bone marrow, spleen, liver, LNs, lung, and omentum and in the uterus during gestation. The steady-state levels of these peripheral NK cells are tightly controlled through a process referred to as homeostasis. NK cell homeostasis is controlled by cell-intrinsic mechanisms (via TFs) and by extrinsic signals; the latter include the cytokine IL-15 and the transforming growth factor (TGF)-β (14, 124,

ANRV270-IY24-09

ARI

15 February 2006

17:51

Stroma IL-15Rα IL-15

mNK

1 MΦ / DC Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

2

mNK

mNK

6 mNK

3

5 CD8mem

4 NKT mNK Figure 4 The peripheral NK cell niche. Newly generated (or recirculating) NK cells enter peripheral tissues (1) and require IL-15 signals for survival (2). Upon stimulation (during infection or stress), NK cells may proliferate (3) or further differentiate (4). The peripheral NK cell niche may be the site of cellular competition for IL-15 between IL-15-dependent cells, such as CD8mem and NKT cells (5). NK cells may leave the peripheral niche to recirculate to other sites (6).

126, 131). Once again, a balance must be achieved between survival, proliferation, and cell death to maintain an optimal pool of NK cells that can participate in immune defense (Figure 4). IL-15 plays a major role in peripheral NK cell homeostasis. Transfer studies have shown that IL-15 maintains the survival and promotes the proliferation of mature NK cells in vivo (124, 126, 131). This effect is mediated through maintenance of the antiapoptotic factor bcl-2, which can protect mature NK cells

from death after cytokine withdrawal (132). As alluded to above, IRF-2 and T-bet are also involved in protecting mature NK cells from apoptosis. Deﬁciencies in either of these two TFs result in a reduced ability to maintain the peripheral NK cell pool (129, 145). The mechanisms that account for the cell death observed in IRF-2- or T-bet-deﬁcient NK cells are unknown, but the enhanced apoptosis does not appear to be due to an inability to respond to IL-15 (129, 145). One possibility is that an activation-induced signal is involved www.annualreviews.org • NK Cell Development

273

ARI

15 February 2006

17:51

because IRF-2-deﬁcient NK cells express elevated levels of CD69 (129). Substantial progress has been made in deciphering the cellular context in which IL-15 promotes NK cell homeostasis. Studies from the Waldmann laboratory provided the ﬁrst evidence for a novel mode of IL-15 transpresentation, involving a cell surface IL-15Rα– IL-15 complex that could activate responding cells via a complex comprising CD122 (IL2Rβ) and γc (127). In vivo HSC transfer experiments have clearly validated this model by demonstrating that both hematopoietic and parenchymal (nonhematopoietic) cells can transpresent IL-15, and that IL-15 and IL-15Rα need to be expressed by the same cell to do so (147–149). Once in the periphery, NK cells reside in IL-15-dependent niches, where they may simply survive, proliferate, or further differentiate (69, 126, 131, 150). Competition between NK cells and other IL-15dependent lymphocytes has been documented after adoptive transfer (151), but whether these cells actually compete in vivo under physiological conditions is not known. Collectively, these results provide a ﬁrst glimpse into the NK cell niche that can be further dissected by visualizing the cells that produce the IL-15Rα–IL-15 complex. Negative regulation also appears to operate under steady-state conditions to maintain the homeostasis of the NK cell pool. TGF-β is a potent antiproliferative factor that can suppress the proliferation of different lymphocyte subsets in vivo (reviewed in 152). In this way, tumor-derived TGF-β may contribute to tumor escape from immunosurveillance. The absence of TGF-β results in a multiorgan inﬂammatory syndrome with prominent lymphocytic inﬁltration (153). A recent report from the Flavell laboratory showed that TGF-β also regulates NK cell homeostasis in the mouse (14). A dominantnegative truncated form of the TGF-βRII controlled by the CD11c promoter was expressed in NK cells and resulted in the expansion and activation of the splenic NK cell pool. NK cell maturation was normal, but the

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

274

Di Santo

peripheral NK cell pool was established more rapidly. Because cycling peripheral NK cells were not increased, NK cell survival was likely augmented in the absence of TGF-β signaling. An alternative explanation is that bone marrow NK cell development was quantitatively enhanced. Thus, TGF-β signaling acts as a negative regulator of NK cell homeostasis. It is not clear at which stage of NK cell development TGF-β exerts its effects or the nature of the cell type that provides TGF-β. Human NK cells treated with TGF-β modulate their expression of NKG2D and NCR, providing a potential mechanism to account for the suppressive effects of TGF-β on NK cell cytotoxicity (154). It is tempting to speculate that regulatory T cell subsets might also be involved because these cells can use a TGF-β-dependent effector mechanism to suppress T cell activation and thereby maintain tolerance (155).

ORGAN-SPECIFIC FUNCTIONS FOR NK CELLS? The fact that NK cells are derived from HSC is unquestioned. In contrast, the deﬁnitive site(s) for NK cell development can only be inferred from where we ﬁnd immature and mature NK cells. NK cells present in different organs may subserve different functions. A tissue hierarchy is suggested when one examines the presence of NK cells at different developmental stages in various organs (Figure 5).

NK Cell Development: Beyond the Bone Marrow The bone marrow is considered the primary site of NK cell generation because it harbors the cellular substrates and soluble factors required for NK cells to mature. Bone marrow ablation (after treatment with bone-seeking isotopes or following estrogen-inducing osteopetrosis) affects peripheral NK cell numbers and their functional capacities (17, 156, 157). Although often cited as evidence that the

ANRV270-IY24-09

ARI

15 February 2006

17:51

ELP

Bone marrow ?

Thymus

Figure 5

NKP

?

? Liver

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

iNK

Spleen

mNK

mNK

mNK ?

? Uterus

bone marrow is required for complete NK cell maturation, these treatments may affect the capacity of NK precursors to respond to maturation signals that they might see elsewhere. Whether the bone marrow alone is sufﬁcient to assure complete NK cell development is still unknown. The bone marrow may only be involved in the initial steps of NK cell differentiation, with other sites required for ﬁnal maturation. Could the bone marrow represent the generative site for NKP that ﬁnish their maturation elsewhere, perhaps in the liver, LN, or spleen? The identiﬁcation of a LN-resident NKP in humans (88) that apparently trafﬁcs from the bone marrow via the blood to the LN would be consistent with this possibility. The identiﬁcation of NKP and immature TRAIL+ NK cells in the liver and spleen (19, 25, 109; C.A.J. Vosshenrich & J.P. Di Santo, unpublished observations) provide additional evidence. Sev-

Lymph nodes

Generation and recirculation of NK cells. ELP, NKP, and iNK are found in the bone marrow. Whether these developmental intermediates leave the bone marrow to complete their differentiation elsewhere is unknown, although NKP and iNK can be found in the liver and spleen. Whether NK cells generated in the thymus have speciﬁc homing potential is not known.

eral alternatives to the bone marrow for completing some aspects of NK cell differentiation exist and should be considered. The bone marrow environment may also provide a niche for mature NK cells. It is accepted that long-lived, terminally differentiated plasma cells reside in the bone marrow (158), and recent evidence demonstrates that memory CD8+ T cells ﬁnd long-term refuge in the bone marrow as well (159). Could a long-lived subset of NK cells exist that follows a similar pattern?

NK Cells in the Thymus Small numbers of NK cells are found in the thymus that are phenotypically mature, are cytotoxic, and can produce cytokines (78, 160). The thymus can export mature NK cells as evidenced by transplants of Ragdeﬁcient thymuses in alymphoid recipients www.annualreviews.org • NK Cell Development

275

ARI

15 February 2006

17:51

(Rag2-deﬁcient x γc -deﬁcient mice) that show a small but detectable population of donorderived NK cells in the spleen ( J.P. Di Santo, unpublished observations). An open question remains as to whether the mature NK cells found in the thymus develop and differentiate in situ in a fashion similar to that observed in the bone marrow. NK cells present in Ragdeﬁcient thymuses resemble their wild-type counterparts, suggesting that neither more mature thymocytes nor a fully developed thymic medullary stromal compartment is required for thymic NK cell homeostasis. What could be the function of NK cells within the thymus? Speciﬁc roles could include modulating thymopoiesis, maintaining thymic architecture, or contributing to defense against thymocyte transformation. Previous reports demonstrate that NK cells could speciﬁcally lyse early thymocytes (161) such that thymic NK cells could be involved in immunosurveillance of rapidly dividing thymic precursors.

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

NK Cell Recirculation Few data are available on the recirculation of NK cells between various organs that harbor them: Adoptive transfer experiments have shown that splenic NK cells can repopulate the spleen, liver, and bone marrow of recipient mice (69, 124, 126). Whether this tissue repartition is uniform (all NK cells migrate equally well to all tissues) or selective (subsets

of NK cells migrate to speciﬁc tissues) has not been extensively studied. NK cells are rare or absent in most tissues, including the digestive tract, muscle, brain, and skin. NK cells are not present in uterine tissue in the steady-state in the mouse (but are in humans), and are massively increased during pregnancy in both species (reviewed in 162). There is no evidence that NK cells develop at these sites. Nevertheless, NK cells are rapidly recruited to tissues after infection or inﬂammation. Chemokines play a major role in this process, and NK cells express a panoply of chemokine receptors (including CCR2, CCR5, CX3CR1, and CXCR3) that allow them to participate in diverse types of inﬂammatory reactions (reviewed in 30, 31). Whether chemokine receptors also dictate NK cell export from the bone marrow or thymus or allow speciﬁc homing to the liver, spleen, or LN is probable, although no published data exist in the mouse. Little is known about the precise physical localization of mature NK or their precursor cells in the bone marrow, spleen, or LN under normal or pathological conditions. Liver NK cells patrol the sinusoids and likely act as sensors for hepatic parenchymal inﬂammation and stress. The topological organization of developing NK cells within generative organs (bone marrow, thymus) under steadystate conditions and following stress of infection or inﬂammation could provide important clues to the mechanisms that control NK cell development.

SUMMARY POINTS 1. NK precursors and immature NK cells have been identiﬁed in mice and humans. The developmental intermediates in NK cell differentiation mature through the concerted action of transcription factors that specify their phenotype and effector functions. The wealth of new information on developing NK cells has been integrated into a coherent model for NK cell development from hematopoietic precursors. Nevertheless, the precise sites where NK cell maturation take place are not yet fully deﬁned.

276

Di Santo

ANRV270-IY24-09

ARI

15 February 2006

17:51

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

2. Clear evidence for functional NK cell subsets exists in humans. Whether this “division of labor” applies to NK cells from other species is unknown, but could provide a means to allow NK cells to respond under diverse conditions. The presence of NK cells in the thymus and immature NK cells in the liver suggests speciﬁc functions for NK cells in these tissues. 3. New assays for NK cell development should help investigators to understand the process. Novel stromal cell–based systems and humanized mouse models will undoubtedly provide the impetus for re-examining the models of human NK cell differentiation and may provide answers to the long-standing questions regarding the origins of different NK cell subsets and their biological roles in normal and pathophysiological processes.

ACKNOWLEDGMENTS J.P.D. is supported by grants from the Institut Pasteur, the Institut National de la Sant´e et de la Recherche M´edicale (INSERM), and the Ligue Nationale contre le Cancer. I thank past and present members of my laboratory for their enthusiastic collaboration, and M. Colonna, F. Colucci, C.A.J. Vosshenrich, and A. Wilson for sharing unpublished results.

LITERATURE CITED 1. Herberman RB, Nunn ME, Holden HT, Lavrin DH. 1975. Natural cytotoxic reactivity of mouse lymphoid cells against syngeneic and allogeneic tumors. II. Characterization of effector cells. Int. J. Cancer 16:230–39 2. Herberman RB, Nunn ME, Lavrin DH. 1975. Natural cytotoxic reactivity of mouse lymphoid cells against syngeneic acid allogeneic tumors. I. Distribution of reactivity and speciﬁcity. Int. J. Cancer 16:216–29 3. Kiessling R, Klein E, Pross H, Wigzell H. 1975. “Natural” killer cells in the mouse. II. Cytotoxic cells with speciﬁcity for mouse Moloney leukemia cells. Characteristics of the killer cell. Eur. J. Immunol. 5:117–21 4. Kiessling R, Klein E, Wigzell H. 1975. “Natural” killer cells in the mouse. I. Cytotoxic cells with speciﬁcity for mouse Moloney leukemia cells. Speciﬁcity and distribution according to genotype. Eur. J. Immunol. 5:112–17 5. Rosenberg EB, McCoy JL, Green SS, Donnelly FC, Siwarski DF, et al. 1974. Destruction of human lymphoid tissue-culture cell lines by human peripheral lymphocytes in 51Cr-release cellular cytotoxicity assays. J. Natl. Cancer Inst. 52:345–52 6. Herberman RB, Ortaldo JR. 1981. Natural killer cells: their roles in defenses against disease. Science 214:24–30 7. de Villartay JP, Fischer A, Durandy A. 2003. The mechanisms of immune diversiﬁcation and their disorders. Nat. Rev. Immunol. 3:962–72 8. Shinkai Y, Rathbun G, Lam KP, Oltz EM, Stewart V, et al. 1992. RAG-2-deﬁcient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68:855–67 9. Trinchieri G. 1989. Biology of natural killer cells. Adv. Immunol. 47:187–376 10. Yu YY, Kumar V, Bennett M. 1992. Murine natural killer cells and marrow graft rejection. Annu. Rev. Immunol. 10:189–213 www.annualreviews.org • NK Cell Development

277

ARI

15 February 2006

17:51

11. Moffett-King A. 2002. Natural killer cells and pregnancy. Nat. Rev. Immunol. 2:656–63 12. Orange JS. 2002. Human natural killer cell deﬁciencies and susceptibility to infection. Microbes Infect. 4:1545–58 13. French AR, Yokoyama WM. 2003. Natural killer cells and viral infections. Curr. Opin. Immunol. 15:45–51 14. Laouar Y, Sutterwala FS, Gorelik L, Flavell RA. 2005. Transforming growth factor-β controls T helper type 1 cell development through regulation of natural killer cell interferon-γ. Nat. Immunol. 6:600–7 15. Pelayo R, Hirose J, Huang J, Garrett KP, Delogu A, et al. 2005. Derivation of 2 categories of plasmacytoid dendritic cells in murine bone marrow. Blood 105:4407–15 16. Colucci F, Schweighoffer E, Tomasello E, Turner M, Ortaldo JR, et al. 2002. Natural cytotoxicity uncoupled from the Syk and ZAP-70 intracellular kinases. Nat. Immunol. 3:288–94 17. Puzanov IJ, Bennett M, Kumar V. 1996. IL-15 can substitute for the marrow microenvironment in the differentiation of natural killer cells. J. Immunol. 157:4282–85 18. Rosmaraki EE, Douagi I, Roth C, Colucci F, Cumano A, Di Santo JP. 2001. Identiﬁcation of committed NK cell progenitors in adult murine bone marrow. Eur. J. Immunol. 31:1900–9 19. Kim S, Iizuka K, Kang HS, Dokun A, French AR, et al. 2002. In vivo developmental stages in murine natural killer cell maturation. Nat. Immunol. 3:523–28 20. Colucci F, Di Santo JP, Leibson PJ. 2002. Natural killer cell activation in mice and men: different triggers for similar weapons? Nat. Immunol. 3:807–13 21. Latour S, Veillette A. 2004. The SAP family of adaptors in immune regulation. Semin. Immunol. 16:409–19 22. Lanier LL. 2005. NK cell recognition. Annu. Rev. Immunol. 23:225–74 23. Moretta L, Moretta A. 2004. Unravelling natural killer cell function: triggering and inhibitory human NK receptors. EMBO J. 23:255–59 24. Colucci F, Caligiuri MA, Di Santo JP. 2003. What does it take to make a natural killer? Nat. Rev. Immunol. 3:413–25 25. Takeda K, Cretney E, Hayakawa Y, Ota T, Akiba H, et al. 2005. TRAIL identiﬁes immature natural killer cells in newborn mice and adult mouse liver. Blood 105:2082– 89 26. Arase H, Saito T, Phillips JH, Lanier LL. 2001. Cutting edge: the mouse NK cellassociated antigen recognized by DX5 monoclonal antibody is CD49b (α2 integrin, very late antigen-2). J. Immunol. 167:1141–44 27. Smyth MJ, Cretney E, Kelly JM, Westwood JA, Street SE, et al. 2005. Activation of NK cell cytotoxicity. Mol. Immunol. 42:501–10 28. Screpanti V, Wallin RP, Ljunggren HG, Grandien A. 2001. A central role for death receptor-mediated apoptosis in the rejection of tumors by NK cells. J. Immunol. 167:2068–73 29. Cooper MA, Fehniger TA, Caligiuri MA. 2001. The biology of human natural killer-cell subsets. Trends Immunol. 22:633–40 30. Robertson MJ. 2002. Role of chemokines in the biology of natural killer cells. J. Leukoc. Biol. 71:173–83 31. Maghazachi AA. 2003. G protein-coupled receptors in natural killer cells. J. Leukoc. Biol. 74:16–24 32. Fehniger TA, Cooper MA, Nuovo GJ, Cella M, Facchetti F, et al. 2003. CD56bright natural killer cells are present in human lymph nodes and are activated by T cell-derived IL-2: a potential new link between adaptive and innate immunity. Blood 101:3052–57

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

278

Di Santo

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

33. Benlagha K, Kyin T, Beavis A, Teyton L, Bendelac A. 2002. A thymic precursor to the NK T cell lineage. Science 296:553–55 34. Loza MJ, Perussia B. 2001. Final steps of natural killer cell maturation: a model for type 1-type 2 differentiation? Nat. Immunol. 2:917–24 35. Karre K, Ljunggren HG, Piontek G, Kiessling R. 1986. Selective rejection of H2-deﬁcient lymphoma variants suggests alternative immune defence strategy. Nature 319:675–78 36. Ljunggren HG, Karre K. 1990. In search of the ‘missing self’: MHC molecules and NK cell recognition. Immunol. Today 11:237–44 37. Long EO. 1999. Regulation of immune responses through inhibitory receptors. Annu. Rev. Immunol. 17:875–904 38. Yokoyama WM. 1993. Recognition structures on natural killer cells. Curr. Opin. Immunol. 5:67–73 39. Lanier LL. 1998. NK cell receptors. Annu. Rev. Immunol. 16:359–93 40. Raulet DH, Vance RE, McMahon CW. 2001. Regulation of the natural killer cell receptor repertoire. Annu. Rev. Immunol. 19:291–330 41. Fernandez NC, Treiner E, Vance RE, Jamieson AM, Lemieux S, Raulet DH. 2005. A subset of natural killer cells achieves self-tolerance without expressing inhibitory receptors speciﬁc for self-MHC molecules. Blood 105:4416–23 41a. Kim S, Poursine-Laurent J, Truscott SM, Lybarger L, Song YJ, et al. 2005. Licensing of natural killer cells by host major histocompatibility complex class I molecules. Nature 436:709–13 41b. Kumar V, McNerney ME. 2005. A new self:MHC-class-I-independent natural-killercell self-tolerance. Nat. Rev. Immunol. 5:363–74 42. Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, et al. 2001. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19:197–223 43. Raulet DH. 2003. Roles of the NKG2D immunoreceptor and its ligands. Nat. Rev. Immunol. 3:781–90 44. Yokoyama WM, Plougastel BF. 2003. Immune functions encoded by the natural killer gene complex. Nat. Rev. Immunol. 3:304–16 45. Vivier E, Nunes JA, Vely F. 2004. Natural killer cell signaling pathways. Science 306:1517–19 46. Colucci F, Rosmaraki E, Bregenholt S, Samson SI, Di Bartolo V, et al. 2001. Functional dichotomy in natural killer cell signaling: Vav-1-dependent and -independent mechanisms. J. Exp. Med. 193:1–13 47. Cella M, Fujikawa K, Tassi I, Kim S, Latinis K, et al. 2004. Differential requirements for Vav proteins in DAP10- and ITAM-mediated NK cell cytotoxicity. J. Exp. Med. 200:817–23 48. Tassi I, Presti R, Kim S, Yokoyama WM, Gilﬁllan S, Colonna M. 2005. PLCγ2 is a critical signaling mediator for murine NK cell activating receptors. J. Immunol. 175:749– 54 49. Upshaw JL, Schoon RA, Dick CJ, Billadeau DD, Leibson PJ. 2005. The isoforms of phospholipase C-γ are differentially used by distinct human NK activating receptors. J. Immunol. 175:213–18 50. Carayannopoulos LN, Yokoyama WM. 2004. Recognition of infected cells by natural killer cells. Curr. Opin. Immunol. 16:26–33 www.annualreviews.org • NK Cell Development

279

ARI

15 February 2006

17:51

51. Colonna M, Nakajima H, Navarro F, Lopez-Botet M. 1999. A novel family of Ig-like receptors for HLA class I molecules that modulate function of lymphoid and myeloid cells. J. Leukoc. Biol. 66:375–81 52. Bendelac A, Bonneville M, Kearney JF. 2001. Autoreactivity by design: innate B and T lymphocytes. Nat. Rev. Immunol. 1:177–86 53. Smyth MJ, Hayakawa Y, Takeda K, Yagita H. 2002. New aspects of natural-killer-cell surveillance and therapy of cancer. Nat. Rev. Cancer 2:850–61 54. Du Pasquier L, Zucchetti I, De Santis R. 2004. Immunoglobulin superfamily receptors in protochordates: before RAG time. Immunol. Rev. 198:233–48 55. Biron CA, Byron KS, Sullivan JL. 1989. Severe herpesvirus infections in an adolescent without natural killer cells. N. Engl. J. Med. 320:1731–35 56. Brown MG, Dokun AO, Heusel JW, Smith HR, Beckman DL, et al. 2001. Vital involvement of a natural killer cell activation receptor in resistance to viral infection. Science 292:934–37 57. Lee SH, Girard S, Macina D, Busa M, Zafer A, et al. 2001. Susceptibility to mouse cytomegalovirus is associated with deletion of an activating natural killer cell receptor of the C-type lectin superfamily. Nat. Genet. 28:42–45 58. Biron CA, Nguyen KB, Pien GC, Cousens LP, Salazar-Mather TP. 1999. Natural killer cells in antiviral defense: function and regulation by innate cytokines. Annu. Rev. Immunol. 17:189–220 59. French AR, Pingel JT, Wagner M, Bubic I, Yang L, et al. 2004. Escape of mutant double-stranded DNA virus from innate immune control. Immunity 20:747–56 60. Martin-Fontecha A, Thomsen LL, Brett S, Gerard C, Lipp M, et al. 2004. Induced recruitment of NK cells to lymph nodes provides IFN-γ for TH 1 priming. Nat. Immunol. 5:1260–55 61. Haller O, Kiessling R, Orn A, Wigzell H. 1977. Generation of natural killer cells: an autonomous function of the bone marrow. J. Exp. Med. 145:1411–16 62. Hackett J Jr, Bosma GC, Bosma MJ, Bennett M, Kumar V. 1986. Transplantable progenitors of natural killer cells are distinct from those of T and B lymphocytes. Proc. Natl. Acad. Sci. USA 83:3427–31 63. Shibuya A, Nagayoshi K, Nakamura K, Nakauchi H. 1995. Lymphokine requirement for the generation of natural killer cells from CD34+ hematopoietic progenitor cells. Blood 85:3538–46 64. Miller JS, Alley KA, McGlave P. 1994. Differentiation of natural killer (NK) cells from human primitive marrow progenitors in a stroma-based long-term culture system: identiﬁcation of a CD34+ 7+ NK progenitor. Blood 83:2594–601 65. Mrozek E, Anderson P, Caligiuri MA. 1996. Role of interleukin-15 in the development of human CD56+ natural killer cells from CD34+ hematopoietic progenitor cells. Blood 87:2632–40 66. Jaleco AC, Blom B, Res P, Weijer K, Lanier LL, et al. 1997. Fetal liver contains committed NK progenitors, but is not a site for development of CD34+ cells into T cells. J. Immunol. 159:694–702 67. Zamai L, Ahmad M, Bennett IM, Azzoni L, Alnemri ES, Perussia B. 1998. Natural killer (NK) cell-mediated cytotoxicity: differential use of TRAIL and Fas ligand by immature and mature primary human NK cells. J. Exp. Med. 188:2375–80 68. Vosshenrich CA, Samson-Villeger SI, Di Santo JP. 2005. Distinguishing features of developing natural killer cells. Curr. Opin. Immunol. 17:151–58 69. Prlic M, Blazar BR, Farrar MA, Jameson SC. 2003. In vivo survival and homeostatic proliferation of natural killer cells. J. Exp. Med. 197:967–76

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

280

Di Santo

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

70. Igarashi H, Gregory SC, Yokota T, Sakaguchi N, Kincade PW. 2002. Transcription from the RAG1 locus marks the earliest lymphocyte progenitors in bone marrow. Immunity 17:117–30 71. Kondo M, Weissman IL, Akashi K. 1997. Identiﬁcation of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91:661–72 72. Terskikh AV, Miyamoto T, Chang C, Diatchenko L, Weissman IL. 2003. Gene expression analysis of puriﬁed hematopoietic stem cells and committed progenitors. Blood 102:94–101 73. Galy A, Travis M, Cen D, Chen B. 1995. Human T, B, natural killer, and dendritic cells arise from a common bone marrow progenitor cell subset. Immunity 3:459–73 73a. Blom B, Spits H. 2006. Development of human lymphoid cells. Annu. Rev. Immunol. 24:287–320 74. Waskow C, Paul S, Haller C, Gassmann M, Rodewald H. 2002. Viable c-KitW/W mutants reveal pivotal role for c-kit in the maintenance of lymphopoiesis. Immunity 17:277–88 75. Waskow C, Rodewald HR. 2002. Lymphocyte development in neonatal and adult cKit-deﬁcient (c-KitW/W ) mice. Adv. Exp. Med. Biol. 512:1–10 76. Michie AM, Carlyle JR, Schmitt TM, Ljutic B, Cho SK, et al. 2000. Clonal characterization of a bipotent T cell and NK cell progenitor in the mouse fetal thymus. J. Immunol. 164:1730–33 77. Douagi I, Colucci F, Di Santo JP, Cumano A. 2002. Identiﬁcation of the earliest prethymic bipotent T/NK progenitor in murine fetal liver. Blood 99:463–71 78. Carlyle JR, Michie AM, Cho SK, Zuniga-Pﬂucker JC. 1998. Natural killer cell development and function precede αβ T cell differentiation in mouse fetal thymic ontogeny. J. Immunol. 160:744–53 79. Spits H, Lanier LL, Phillips JH. 1995. Development of human T and natural killer cells. Blood 85:2654–70 80. Schmitt TM, Zuniga-Pﬂucker JC. 2002. Induction of T cell development from hematopoietic progenitor cells by delta-like-1 in vitro. Immunity 17:749–56 81. Porritt HE, Rumfelt LL, Tabrizifard S, Schmitt TM, Zuniga-Pﬂucker JC, Petrie HT. 2004. Heterogeneity among DN1 prothymocytes reveals multiple progenitors with different capacities to generate T cell and non-T cell lineages. Immunity 20:735–45 82. Schmitt TM, Ciofani M, Petrie HT, Zuniga-Pﬂucker JC. 2004. Maintenance of T cell speciﬁcation and differentiation requires recurrent Notch receptor-ligand interactions. J. Exp. Med. 200:469–79 83. Taghon TN, David ES, Zuniga-Pﬂucker JC, Rothenberg EV. 2005. Delayed, asynchronous, and reversible T-lineage speciﬁcation induced by Notch/Delta signaling. Genes Dev. 19:965–78 84. Radtke F, Wilson A, Stark G, Bauer M, van Meerwijk J, et al. 1999. Deﬁcient T cell fate speciﬁcation in mice with an induced inactivation of Notch1. Immunity 10:547–58 85. Ikawa T, Fujimoto S, Kawamoto H, Katsura Y, Yokota Y. 2001. Commitment to natural killer cells requires the helix-loop-helix inhibitor Id2. Proc. Natl. Acad. Sci. USA 98:5164– 69 86. Kennedy MK, Glaccum M, Brown SN, Butz EA, Viney JL, et al. 2000. Reversible defects in natural killer and memory CD8 T cell lineages in interleukin 15-deﬁcient mice. J. Exp. Med. 191:771–80 87. Vosshenrich CA, Ranson T, Samson SI, Corcuff E, Colucci F, et al. 2005. Roles for common cytokine receptor γ-chain-dependent cytokines in the generation, differentiation, and maturation of NK cell precursors and peripheral NK cells in vivo. J. Immunol. 174:1213–21 www.annualreviews.org • NK Cell Development

281

ARI

15 February 2006

17:51

88. Freud AG, Becknell B, Roychowdhury S, Mao HC, Ferketich AK, et al. 2005. A human CD34+ subset resides in lymph nodes and differentiates into CD56bright natural killer cells. Immunity 22:295–304 89. Yu H, Fehniger TA, Fuchshuber P, Thiel KS, Vivier E, et al. 1998. Flt3 ligand promotes the generation of a distinct CD34+ human natural killer cell progenitor that responds to interleukin-15. Blood 92:3647–57 90. Mebius RE, Rennert P, Weissman IL. 1997. Developing lymph nodes collect CD4+ CD3− LTβ+ cells that can differentiate to APC, NK cells, and follicular cells but not T or B cells. Immunity 7:493–504 91. Yoshida H, Kawamoto H, Santee SM, Hashi H, Honda K, et al. 2001. Expression of α4 β7 integrin deﬁnes a distinct pathway of lymphoid progenitors committed to T cells, fetal intestinal lymphotoxin producer, NK, and dendritic cells. J. Immunol. 167:2511–21 92. Williams NS, Klem J, Puzanov IJ, Sivakumar PV, Bennett M, Kumar V. 1999. Differentiation of NK1.1+ , Ly49+ NK cells from ﬂt3+ multipotent marrow progenitor cells. J. Immunol. 163:2648–56 93. Williams NS, Moore TA, Schatzle JD, Puzanov IJ, Sivakumar PV, et al. 1997. Generation of lytic natural killer 1.1+ , Ly-49− cells from multipotential murine bone marrow progenitors in a stroma-free culture: deﬁnition of cytokine requirements and developmental intermediates. J. Exp. Med. 186:1609–14 94. Melchers F, Haasner D, Streb M, Rolink A. 1992. B-lymphocyte lineage-committed, IL-7 and stroma cell-reactive progenitors and precursors, and their differentiation to B cells. Adv. Exp. Med. Biol. 323:111–17 95. Ware CF. 2005. Network communications: lymphotoxins, LIGHT, and TNF. Annu. Rev. Immunol. 23:787–819 96. Iizuka K, Chaplin DD, Wang Y, Wu Q, Pegg LE, et al. 1999. Requirement for membrane lymphotoxin in natural killer cell development. Proc. Natl. Acad. Sci. USA 96:6336–40 97. Lian RH, Chin RK, Nemeth HE, Libby SL, Fu YX, Kumar V. 2004. A role for lymphotoxin in the acquisition of Ly49 receptors during NK cell development. Eur. J. Immunol. 34:2699–707 98. Stevenaert F, Van Beneden K, De Colvenaer V, Franki AS, Debacker V, et al. 2005. Ly49 and CD94/NKG2 receptor acquisition by NK cells does not require lymphotoxin-β receptor expression. Blood 106:956–62 99. Wu Q, Sun Y, Wang J, Lin X, Wang Y, et al. 2001. Signal via lymphotoxin-βR on bone marrow stromal cells is required for an early checkpoint of NK cell development. J. Immunol. 166:1684–89 100. Sivakumar V, Hammond KJ, Howells N, Pfeffer K, Weih F. 2003. Differential requirement for Rel/nuclear factor κB family members in natural killer T cell development. J. Exp. Med. 197:1613–21 101. Chen KS, DeLuca HF. 1994. Isolation and characterization of a novel cDNA from HL60 cells treated with 1,25-dihydroxyvitamin D-3. Biochim. Biophys. Acta 1219:26–32 102. Lee KN, Kang HS, Jeon JH, Kim EM, Yoon SR, et al. 2005. VDUP1 is required for the development of natural killer cells. Immunity 22:195–208 103. Rothenberg EV, Taghon T. 2005. Molecular genetics of T cell development. Annu. Rev. Immunol. 23:601–49 104. Singh H, Medina KL, Pongubala JM. 2005. Contingent gene regulatory networks and B cell fate speciﬁcation. Proc. Natl. Acad. Sci. USA 102:4949–53 105. Scott EW, Simon MC, Anastasi J, Singh H. 1994. Requirement of transcription factor PU.1 in the development of multiple hematopoietic lineages. Science 265:1573–77

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

282

Di Santo

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

106. DeKoter RP, Lee HJ, Singh H. 2002. PU.1 regulates expression of the interleukin-7 receptor in lymphoid progenitors. Immunity 16:297–309 107. Medina KL, Pongubala JM, Reddy KL, Lancki DW, Dekoter R, et al. 2004. Assembling a gene regulatory network for speciﬁcation of the B cell fate. Dev. Cell 7:607–17 108. Maeurer MJ, Lotze MT. 1998. Interleukin-7 (IL-7) knockout mice. Implications for lymphopoiesis and organ-speciﬁc immunity. Int. Rev. Immunol. 16:309–22 109. Samson SI, Richard O, Tavian M, Ranson T, Vosshenrich CA, et al. 2003. GATA-3 promotes maturation, IFN-γ production, and liver-speciﬁc homing of NK cells. Immunity 19:701–11 110. Colucci F, Samson SI, DeKoter RP, Lantz O, Singh H, Di Santo JP. 2001. Differential requirement for the transcription factor PU.1 in the generation of natural killer cells versus B and T cells. Blood 97:2625–32 111. Nutt SL, Metcalf D, D’Amico A, Polli M, Wu L. 2005. Dynamic regulation of PU.1 expression in multipotent hematopoietic progenitors. J. Exp. Med. 201:221–31 112. Barton K, Muthusamy N, Fischer C, Ting CN, Walunas TL, et al. 1998. The Ets-1 transcription factor is required for the development of natural killer cells in mice. Immunity 9:555–63 113. Boggs SS, Trevisan M, Patrene K, Geogopoulos K. 1998. Lack of natural killer cell precursors in fetal liver of Ikaros knockout mutant mice. Nat. Immunol. 16:137–45 114. Engel I, Murre C. 2001. The function of E- and Id proteins in lymphocyte development. Nat. Rev. Immunol. 1:193–99 115. Bain G, Maandag EC, Izon DJ, Amsen D, Kruisbeek AM, et al. 1994. E2A proteins are required for proper B cell development and initiation of immunoglobulin gene rearrangements. Cell 79:885–92 116. Heemskerk MH, Blom B, Nolan G, Stegmann AP, Bakker AQ, et al. 1997. Inhibition of T cell and promotion of natural killer cell development by the dominant negative helix loop helix factor Id3. J. Exp. Med. 186:1597–602 117. Yokota Y, Mansouri A, Mori S, Sugawara S, Adachi S, et al. 1999. Development of peripheral lymphoid organs and natural killer cells depends on the helix-loop-helix inhibitor Id2. Nature 397:702–6 118. Ikawa T, Kawamoto H, Fujimoto S, Katsura Y. 1999. Commitment of common T/natural killer (NK) progenitors to unipotent T and NK progenitors in the murine fetal thymus revealed by a single progenitor assay. J. Exp. Med. 190:1617–26 119. Massari ME, Grant PA, Pray-Grant MG, Berger SL, Workman JL, Murre C. 1999. A conserved motif present in a class of helix-loop-helix proteins activates transcription by direct recruitment of the SAGA complex. Mol. Cell 4:63–73 120. Ikawa T, Kawamoto H, Wright LY, Murre C. 2004. Long-term cultured E2A-deﬁcient hematopoietic progenitor cells are pluripotent. Immunity 20:349–60 121. Anderson MK, Hernandez-Hoyos G, Diamond RA, Rothenberg EV. 1999. Precise developmental regulation of Ets family transcription factors during speciﬁcation and commitment to the T cell lineage. Development 126:3131–48 122. Barao I, Hudig D, Ascensao JL. 2003. IL-15-mediated induction of LFA-1 is a late step required for cytotoxic differentiation of human NK cells from CD34+ Lin- bone marrow cells. J. Immunol. 171:683–90 123. DeHart SL, Heikens MJ, Tsai S. 2005. Jagged2 promotes the development of natural killer cells and the establishment of functional natural killer cell lines. Blood 105:3521–27 124. Cooper MA, Bush JE, Fehniger TA, VanDeusen JB, Waite RE, et al. 2002. In vivo evidence for a dependence on interleukin 15 for survival of natural killer cells. Blood 100:3633–38 www.annualreviews.org • NK Cell Development

283

ARI

15 February 2006

17:51

125. Kennedy MK, Glaccum M, Brown SN, Butz EA, Viney JL, et al. 2000. Reversible defects in natural killer and memory CD8 T cell lineages in interleukin 15-deﬁcient mice. J. Exp. Med. 191:771–80 126. Ranson T, Vosshenrich CA, Corcuff E, Richard O, Muller W, Di Santo JP. 2003. IL-15 is an essential mediator of peripheral NK cell homeostasis. Blood 101:4887–93 127. Dubois S, Mariner J, Waldmann TA, Tagaya Y. 2002. IL-15Rα recycles and presents IL-15 in trans to neighboring cells. Immunity 17:537–47 128. Ogasawara K, Hida S, Azimi N, Tagaya Y, Sato T, et al. 1998. Requirement for IRF-1 in the microenvironment supporting development of natural killer cells. Nature 391:700–3. Erratum. 1998. Nature 392(6678):843 129. Taki S, Nakajima S, Ichikawa E, Saito T, Hida S. 2005. IFN regulatory factor-2 deﬁciency revealed a novel checkpoint critical for the generation of peripheral NK cells. J. Immunol. 174:6005–12 130. Kawamura T, Koka R, Ma A, Kumar V. 2003. Differential roles for IL-15R α-chain in NK cell development and Ly-49 induction. J. Immunol. 171:5085–90 131. Jamieson AM, Isnard P, Dorfman JR, Coles MC, Raulet DH. 2004. Turnover and proliferation of NK cells in steady state and lymphopenic conditions. J. Immunol. 172:864–70 132. Carson WE, Fehniger TA, Haldar S, Eckhert K, Lindemann MJ, et al. 1997. A potential role for interleukin-15 in the regulation of human natural killer cell survival. J. Clin. Invest. 99:937–43 133. Iizuka K, Chaplin DD, Wang Y, Wu Q, Pegg LE, et al. 1999. Requirement for membrane lymphotoxin in natural killer cell development. Proc. Natl. Acad. Sci. USA 96:6336–40 134. Ferlazzo G, Thomas D, Lin SL, Goodman K, Morandi B, et al. 2004. The abundant NK cells in human secondary lymphoid tissues require activation to express killer cell Ig-like receptors and become cytolytic. J. Immunol. 172:1455–62 135. Bloch-Queyrat C, Fondaneche MC, Chen R, Yin L, Relouzat F, et al. 2005. Regulation of natural cytotoxicity by the adaptor SAP and the Src-related kinase Fyn. J. Exp. Med. 202:181–92 136. Gilﬁllan S, Ho EL, Cella M, Yokoyama WM, Colonna M. 2002. NKG2D recruits two distinct adapters to trigger NK cell activation and costimulation. Nat. Immunol. 3:1150–55 137. Ogasawara K, Hamerman JA, Hsin H, Chikuma S, Bour-Jordan H, et al. 2003. Impairment of NK cell function by NKG2D modulation in NOD mice. Immunity 18:41–51 138. Wu J, Song Y, Bakker AB, Bauer S, Spies T, et al. 1999. An activating immunoreceptor complex formed by NKG2D and DAP10. Science 285:730–32 139. Lee KM, Forman JP, McNerney ME, Stepp S, Kuppireddi S, et al. 2005. Requirement of homotypic NK cell interactions through 2B4(CD244)/CD48 in the generation of NK effector functions. Blood. In press 140. Wilde JI, Watson SP. 2001. Regulation of phospholipase C γ isoforms in haematopoietic cells: why one, not the other? Cell. Signal. 13:691–701 141. Wang D, Feng J, Wen R, Marine JC, Sangster MY, et al. 2000. Phospholipase Cγ2 is essential in the functions of B cell and several Fc receptors. Immunity 13:25–35 141a. Caraux A, Kim N, Bell SE, Zompi S, Ranson T, et al. 2005. Phospholipase C-γ2 is essential for NK cell cytotoxicity and innate immunity to malignant and virally infected cells. Blood. In press 142. Ting AT, Karnitz LM, Schoon RA, Abraham RT, Leibson PJ. 1992. Fcγ receptor activation induces the tyrosine phosphorylation of both phospholipase C (PLC)-γ 1 and PLC-γ 2 in natural killer cells. J. Exp. Med. 176:1751–55

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

284

Di Santo

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-09

ARI

15 February 2006

17:51

143. Caraux A, Lu Q, Fernandez N, Di Santo JP, Raulet DH, Lemke G, et al. 2004. Receptor Tyrosine kinases of the Tyro 3 family play a critical role in natural killer cell differentiation. Presented at 8th Meet. Soc. Nat. Immun., The Netherlands 144. Crosier KE, Crosier PS. 1997. New insights into the control of cell growth; the role of the AxI family. Pathology 29:131–35 145. Townsend MJ, Weinmann AS, Matsuda JL, Salomon R, Farnham PJ, et al. 2004. Tbet regulates the terminal maturation and homeostasis of NK and Vα14i NKT cells. Immunity 20:477–94 146. Lohoff M, Duncan GS, Ferrick D, Mittrucker HW, Bischof S, et al. 2000. Deﬁciency in the transcription factor interferon regulatory factor (IRF)-2 leads to severely compromised development of natural killer and T helper type 1 cells. J. Exp. Med. 192:325–36 147. Burkett PR, Koka R, Chien M, Chai S, Boone DL, Ma A. 2004. Coordinate expression and trans presentation of interleukin (IL)-15Rα and IL-15 supports natural killer cell and memory CD8+ T cell homeostasis. J. Exp. Med. 200:825–34 148. Koka R, Burkett PR, Chien M, Chai S, Chan F, et al. 2003. Interleukin (IL)-15Rαdeﬁcient natural killer cells survive in normal but not IL-15Rα-deﬁcient mice. J. Exp. Med. 197:977–84 149. Schluns KS, Nowak EC, Cabrera-Hernandez A, Puddington L, Lefrancois L, Aguila HL. 2004. Distinct cell types control lymphoid subset development by means of IL-15 and IL-15 receptor α expression. Proc. Natl. Acad. Sci. USA 101:5616–21 150. Brady J, Hayakawa Y, Smyth MJ, Nutt SL. 2004. IL-21 induces the functional maturation of murine NK cells. J. Immunol. 172:2048–58 151. Ranson T, Vosshenrich CA, Corcuff E, Richard O, Laloux V, et al. 2003. IL-15 availability conditions homeostasis of peripheral natural killer T cells. Proc. Natl. Acad. Sci. USA 100:2663–68 152. Schmidt-Weber CB, Blaser K. 2004. Regulation and role of transforming growth factor-β in immune tolerance induction and inﬂammation. Curr. Opin. Immunol. 16:709– 16 153. Kulkarni AB, Huh CG, Becker D, Geiser A, Lyght M, et al. 1993. Transforming growth factor β 1 null mutation in mice causes excessive inﬂammatory response and early death. Proc. Natl. Acad. Sci. USA 90:770–74 154. Castriconi R, Cantoni C, Della Chiesa M, Vitale M, Marcenaro E, et al. 2003. Transforming growth factor β 1 inhibits expression of NKp30 and NKG2D receptors: consequences for the NK-mediated killing of dendritic cells. Proc. Natl. Acad. Sci. USA 100:4120–25 155. von Boehmer H. 2005. Mechanisms of suppression by suppressor T cells. Nat. Immunol. 6:338–44 156. Haller O, Wigzell H. 1977. Suppression of natural killer cell activity with radioactive strontium: effector cells are marrow dependent. J. Immunol. 118:1503–6 157. Seaman WE, Gindhart TD, Greenspan JS, Blackman MA, Talal N. 1979. Natural killer cells, bone, and the bone marrow: studies in estrogen-treated mice and in congenitally osteopetrotic (mi/mi) mice. J. Immunol. 122:2541–47 158. Manz RA, Thiel A, Radbruch A. 1997. Lifetime of plasma cells in the bone marrow. Nature 388:133–34 159. Mazo IB, Honczarenko M, Leung H, Cavanagh LL, Bonasio R, et al. 2005. Bone marrow is a major reservoir and site of recruitment for central memory CD8+ T cells. Immunity 22:259–70 www.annualreviews.org • NK Cell Development

285

ANRV270-IY24-09

ARI

15 February 2006

17:51

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

160. Garni-Wagner BA, Witte PL, Tutt MM, Kuziel WA, Tucker PW, et al. 1990. Natural killer cells in the thymus. Studies in mice with severe combined immune deﬁciency. J. Immunol. 144:796–803 161. Hansson M, Karre K, Kiessling R, Roder J, Andersson B, Hayry P. 1979. Natural NK-cell targets in the mouse thymus: characteristics of the sensitive cell population. J. Immunol. 123:765–71 162. Moffet-King A. 2002. Natural killer cells and pregnancy. Nat. Rev. Immunol. 2:656–63

286

Di Santo

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:257-286. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-10

ARI

15 February 2006

21:28

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Development of Human Lymphoid Cells Bianca Blom and Hergen Spits Department of Cell Biology and Histology, Academic Medical Center, University of Amsterdam, Amsterdam 1105 AZ, The Netherlands; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:287–320 First published online as a Review in Advance on December 12, 2005 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090612 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0287$20.00

Key Words human lymphocytes, T cells, B cells, NK cells, DC

Abstract The lymphocytes, T, B, and NK cells, and a proportion of dendritic cells (DCs) have a common developmental origin. Lymphocytes develop from hematopoietic stem cells via common lymphocyte and various lineage-restricted precursors. This review discusses the current knowledge of human lymphocyte development and the phenotypes and functions of the rare intermediate populations that together form the pathways of development into T, B, and NK cells and DCs. Clearly, development of hematopoietic cells is supported by cytokines. The studies of patients with genetic deﬁciencies in cytokine receptors that are discussed here have illuminated the importance of cytokines in lymphoid development. Lineage decisions are under control of transcription factors, and studies performed in the past decade have provided insight into transcriptional control of human lymphoid development, the results of which are summarized and discussed in this review.

287

ANRV270-IY24-10

ARI

15 February 2006

21:28

INTRODUCTION CLP: common lymphoid precursor

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TdT: terminal deoxynucleotidyl transferase

288

All hematopoietic cells are derived from pluripotent hematopoietic stem cells (HSCs). These cells differentiate into mature hematopoietic cells through various intermediate cell types that are deﬁned by expression of cell surface antigens. Traditionally, investigators have assumed that the ﬁrst step in hematopoietic development is differentiation of HSCs into myeloid and lymphocyte precursors. Myeloid precursors differentiate into erythroid, megakaryocytic, and granulocytic/monocytic (GM) lineages, whereas lymphoid precursors develop into natural killer (NK), T, and B cells. A wealth of data support this concept. However, a recent study has put this model into question with the identiﬁcation of a precursor that possesses a combined GM and lymphoid precursor potential but is unable to develop into an erythrocyte or megakaryocyte (1). Most researchers agree that the lymphocyte T, B, and NK cells originate from a common precursor, generally referred to as a common lymphoid precursor (CLP), although some researchers maintain that T and B cells are not derived from a common precursor without myeloid precursor activity but instead that there are T/macrophage and B/macrophage precursors (2). The developmental status of dendritic cells (DCs) is more complicated. The DC lineages consist of several often not clearly deﬁned populations that seem to have a myeloid, lymphoid, or mixed lymphoid/myeloid origin. Most of our knowledge about lymphoid development stems from studies with mice. The obvious reason for the reliance on mice studies is the possibility of performing in vivo experiments to test precursor activities of certain cell populations. In addition, genetically modiﬁed mice have been instrumental not only in illuminating developmental pathways but also in elucidating mechanisms behind certain developmental cellular transitions. By contrast, most information on human hematopoietic development is derived

Blom

·

Spits

from in vitro studies, although certain genetic abnormalities have greatly contributed to our understanding of some underlying mechanisms of human hematopoietic cell development. Studies on human hematopoiesis seem to be generally consistent with principles outlined in experimental models, but the cell surface phenotypes of human transitional cell populations are often different from those in the mouse. In this review, we discuss the current knowledge regarding human lymphoid cell development, in particular of T, B, and NK cells and DCs. We do not exhaustively review current information from mouse models, as excellent reviews have appeared recently (2a–d, 94, 160), but this information is used as reference to compare human and mouse lymphocyte development.

HUMAN LYMPHOID PRECURSORS The idea that hematopoiesis progresses by gradually limiting the developmental potential of precursors predicted that precursor cells exist with limited lymphocyte-restricted precursor potential. Indeed, precursor cells that are restricted to lymphoid lineages have been identiﬁed in both mice and humans. That all hematopoietic precursors in humans are present within a population of cells that express CD34 is well established (reviewed in 3), and this marker is useful in elucidating pathways in the development of particular hematopoietic lineages. In searching CD34+ human bone marrow cells for the presence of cells that express lymphocyte-restricted antigens, several research groups have found CD34+ cells expressing the lymphocyte markers terminal deoxynucleotidyl transferase (TdT) and CD10, and they speculate that these cells represent precursors of T and B cells (4, 5). Galy and coworkers (6) obtained experimental support for this idea by demonstrating that CD34+ CD10+ CD45RA+ cells obtained from fetal and adult bone marrow develop into CD33+ and CD1a+ CD33+ DCs and CD19+

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

B cells, although these cells were unable to develop into CD14+ monocytes. Importantly, investigators found that 14% of single-sorted fetal bone marrow CD34+ CD10+ cells produced B cells, NK cells, and DCs. In addition, they showed that a limited number of CD34+ CD10+ cells injected into fragments of fetal thymus were able to develop into T cells, but technical limitations prevented testing T cell development on a clonal level (6). Mouse bone marrow CLP, as described for the ﬁrst time by Kondo et al. (7), express the IL-7Rα chain. Indeed, two groups reported that the IL-7Rα chain is expressed on a large proportion of human bone marrow CD34+ CD10+ cells, which were CD45RA and CD43 positive but negative for CD24 (8, 9). These cells expressed transcripts characteristic for the B cell lineage, such as Pax-5 and Igβ, and the T cell–associated transcripts GATA3 and pTα, and were able to differentiate into B cells and NK cells (8, 9). These characteristics, and the similarity of these cells with those deﬁned by Galy et al. (6), suggested that the CD10+ IL-7Rα + cells represent CLP. However, the number of NK cells recovered after four weeks culture of these cells with Sys-1 cells, a murine stromal cell line in the presence of stem cell factor (SCF), Flt3-L, IL-7, and IL-2, was much lower than that of B cells (9), and the T cell precursor activity of the CD34+ CD10+ IL-7Rα + bone marrow precursors was not tested, leaving open the possibility that these IL-7Rα + precursor cells were biased to the B cell lineage. This notion is supported by identiﬁcation of a very similar cell type in cord blood, which was shown to be B cell biased (10) (see below). Many studies have used CD7 as a marker for lymphoid progenitors because this marker is not expressed or is only weakly expressed on myeloid cells (11). An early study demonstrated that CD34+ CD7+ bone marrow cells contain NK cell precursors, but no information was provided about T cell, B cell, and DC developmental potential (12). More recently, a careful phenotypic and functional analysis of CD34+ bone marrow

cells revealed the presence of CD34+ lin− , CD7+ CD10− (2%), CD7− CD10+ (10%), and CD7+ CD10+ (0.3%) cells (13). Both the bone marrow CD7+ and CD10+ populations had NK and B cell precursor activities, but the B cell potential of CD34+ lin− CD10+ was much higher than that of the CD7+ cells. Because of the scarcity of the CD7+ CD10+ populations, the functional activity of those cells was not tested (13). Polymerase chain reaction (PCR) analysis of B cell– and NK cell–speciﬁc transcripts of these populations strongly suggested that the CD34+ lin− , CD7+ CD10− and CD7− CD10+ , are NK and B cell biased, respectively (13). The chemokine receptor CXCR4 is expressed on cells with restricted lymphoid precursor activities, and the CD34+ CXCR4+ cells can be further subdivided into IL-7Rα + and IL-7Rα − cells. Although the developmental potential of CD34+ CXCR4+ IL-7Rα − bone marrow cells has yet to be tested (14), these cells may be precursors of recently identiﬁed lymphoid precursors in neonatal cord blood, which were found to be IL7Rα − (10, 15). Hao et al. (15) identiﬁed CD34+ CD38lo CD45RA+ CD7+ CD10+ IL7Rα − cells in cord blood able to differentiate into NK cells, B cells, and CD1a+ DCs, and although the T cell potential of these cells was not tested, Hao et al. suggested that these were lymphoid precursors. Two other groups conﬁrmed that CD34+ CD7+ CD45RA+ cord blood cells have lymphoid potential but found that CD10 and CD7 tended not to be coexpressed on CD34+ CD45RA+ cells (10, 13). Interestingly, the CD34+ CD45RA+ CD7− CD10+ population expressed IL-7Rα, whereas IL7Rα was absent on CD34+ CD45RA+ CD7+ cells (10). Both CD34+ CD7+ IL-7Rα − and CD34+ CD10+ IL-7Rα + populations had lymphoid potential, but the CD34+ CD45RA+ CD7+ cells preferentially differentiated into T and NK cells, whereas the CD34+ CD45RA+ CD10+ population was biased to develop into B cells (10, 13). Although differentiation assays revealed

www.annualreviews.org • Development of Human Lymphoid Cells

289

ANRV270-IY24-10

ARI

15 February 2006

pDC: plasmacytoid dendritic cell TSP: thymus seeding progenitor

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ETP: early thymic progenitor

290

21:28

that both CD34+ CD45RA+ CD7+ and CD34+ CD45RA+ CD10+ populations are biased to lymphoid lineages and completely lack the ability to develop into erythrocytes and megakaryocytes, they possess some GM precursor potential (10). It is not clear whether all cells within the CD34+ CD45RA+ CD7+ population are multipotent or whether this population contains both GM and lymphoid precursors. More recently, Chicha et al. (16) reported that either CD34+ CD38+ CD10+ or CD34+ CD38+ CD7+ cord blood precursors can develop into B cells and plasmacytoid DCs (pDCs), but that they fail to develop into GM cells. This study did not show whether the CD34+ CD7+ cord blood cell precursors coexpress CD10 or not, and these authors did not use anti-CD45RA in separating their precursors (16), making it difﬁcult to compare their results with those of others (10, 15). The phenotype of lymphoid-restricted progenitors from cord blood and fetal and adult bone marrow is not yet ﬁrmly established. An important reason for this lies in the use by different groups of dissimilar antibodies against the same differentiation antigen or the same antibodies tagged with different ﬂuorochromes. For example, anti-CD7FITC (ﬂuorescein isothiocyanate) used by Haddad et al. (10) is much less sensitive than anti-CD7-PE (phyco erythrin) or Tricolor, and the use of CD7-FITC misses cells that do express CD7 but at lower levels (11). Another reason the phenotype has not been ﬁrmly established is the controversy concerning expression of CD38, which was absent on lymphoid precursors in one study (17) but expressed in two others (10, 16); this difference could be due to the use of anti-CD38 antibodies with different afﬁnities. With these caveats in mind, there is consensus that CD7 and CD10 deﬁne CD34+ CD45RA+ precursors biased to, but probably not fully committed to, lymphoid lineages (Figure 1). In addition, within the CD34+ CD45RA+ population both in bone marrow and cord blood, there are CD7+ CD10+ , Blom

·

Spits

CD7+ CD10− , and CD7− CD10+ cells, although the existence and function of the very rare CD7+ CD10+ cells have yet to be conﬁrmed. It is tempting to speculate that CD34+ lin− CD45RA+ CD7+ CD10+ are the precursors of the CD7+ CD10− IL-7Rα − T/NK cell–biased and CD7− CD10+ IL7Rα + B cell–biased populations (10, 13), but this has to be proven by using various antibody preparations combined with functional, preferentially clonal assays.

DEVELOPMENT OF T CELLS Human Thymus Seeding Progenitor (TSP) and Early Thymic Progenitor (ETP) The transient developmental potential of thymic precursors demands that a continuous source of thymus seeding progenitors (TSP) can enter the thymus. Because the thymus remains active probably through an advanced age, TSP should be present in adult blood as well. As discussed in a previous section, CD34+ CD45RA+ CD7+ cells that have T, B, and NK and some GM precursor activities have been found in cord blood (10, 15). These cells resemble CD34+ CD38lo ETP with T cell, NK cell, and DC precursor activities (18) (Figure 2). An analysis of the T cell receptor (TCR) rearrangement status conﬁrms that the CD34+ CD38lo cells form the most immature population in the thymus (19), and it is tempting to speculate that CD34+ CD38lo thymocytes, which mostly coexpress CD10, are the direct progeny of the CD34+ CD45RA+ CD7+ CD10+ CD38lo cord blood cells (15). The presence of multipotential precursors in the human thymus indicates that T cell commitment takes place within this organ in line with data in the mouse (20). Recently, however, investigators argued that the murine thymus is seeded not only by multipotent precursors but also by precursor cells that are lineage restricted (21, 22). If that is also the case in humans, such cells should be present in cord

ANRV270-IY24-10

ARI

15 February 2006

21:28

HSC

CD34+CD45RA+ (CD7+CD10+)

MEP

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CLP

CMP

CD34+CD45RA–

CD34+CD45RA+ CD10 + IL-7Rα α+

CD7+

T

NK

Figure 1

pDC

GMP CD34+CD45RA+CD123+

B

cDC

blood. One study failed to ﬁnd evidence for the presence of TCR rearrangements in cord blood CD34+ cells (23), but another study reported the presence of complete TCRδ and partial TCRβ (Dβ-Jβ) rearrangements in CD34+ CD7+ cord blood precursors (24). This issue should therefore be readdressed by single-cell PCR analysis of the recently identiﬁed rare CD34+ CD45RA+ CD7+ cord blood population (10, 15). T cell–restricted precursors were convincingly identiﬁed in human bone marrow (25), but whether these cells can migrate to the thymus is unclear. CD34+ CD19+ B cell precursors are also present in the human thymus (26). Recently, it was reported that although CD34+ lin− cord blood cells could develop into B cells in a fetal thymic organ culture (FTOC) in the presence of Notch inhibitors (26a), CD34+ CD1a− thymocytes failed to do so, suggesting that the

pDC

cDC

mo

Gr

Model for development of CLP and GMP. CLP develop into pDCs and cDCs, but whether they derive from precursors that are CD7+ , CD10+ , or CD7+ CD10+ is not completely clear (Reference 16) (MEP, megakaryocyte/erythrocyte progenitor).

latter cells lack B cell precursor potential. Thus, the thymic B lineage cells (26) may be derived from CD34+ CD10+ IL-7Rα + cord blood progenitors that express a B cell transcript signature and are developmentally biased to the B cell lineage (10). In summary, at least a proportion of the precursors that seed the human thymus are multipotential. Whether or not some of the precursors that migrate into the thymus are lineage restricted before entrance is unclear.

Cellular Stages in the Development of ETP into CD4+ CD8+ Double-Positive Immature T Cells Over the past few decades, the various transitional stages of T cell development in the human thymus have been characterized

www.annualreviews.org • Development of Human Lymphoid Cells

291

ANRV270-IY24-10

ARI

15 February 2006

21:28

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Figure 2 Model for the earliest stages of development in the thymus. Evidence from in vitro experiments suggests that NK cells can develop not only from T/NK precursors but also from NK/DC myeloid precursors. On the basis of expression of pTα, CD2, CD5, and CD7, thymic pDCs may develop from ETP via a lymphoid pathway, but it is possible that the CD34+ CD44hi myeloid precursors can also develop into pDCs in the human thymus.

CD4 ISP: CD4 immature single-positive cell

292

CD34hi CD33lo CD38lo CD44+ IL-7Rα+

CD34– CD38hi CD44hi CD33– IL3Rαhi BDCA2+

CD34hi CD38+ CD1a– IL-7Rα+

pDC

pre-T/NK

ETP

NK

pre- cDC

CD34lo CD3+ CD44hi CD33+ GM-CSFR+

regarding phenotype and status of the TCR gene rearrangements (reviewed in 27, 28) (Figure 3). As discussed in the previous section, ETP are enclosed within the CD34+ CD1a− population. The downstream CD34+ CD1a+ population is committed to the T cell lineage because they are unable to develop into non-T cells (27, 29–31). A recent study has shown that the subsequent CD4 immature single-positive (CD4 ISP) (32) population can be divided into two subgroups: a CD4+ CD1a+ (CD4 ISPlo ) and a CD4hi CD1ahi (CD4 ISPhi ) population (S. Ligthart, Y. Yasuda, H. Spits, and B. Blom, manuscript submitted). Downstream of the CD4 ISP subset are CD4+ CD8α + β− (early double-positive) and CD4+ CD8α + β+ populations (29, 32), the latter cells being Blom

·

Spits

CD34+ CD38+ CD1a+ IL-7Rα+ pre-T

CD34– CD38hi CD56+

cDC

THYMUS

the precursors of double-positive TCRαβ+ cells. During the early stages in T cell development, the TCR loci undergo rearrangement in a sequence TCRδ > γ > β > α (19, 33, 34); however, there is disagreement with respect to the cell type in which rearrangement of particular loci occurs, probably owing to differences in sensitivity of the methods used to analyze TCR rearrangements. For example, by using a PCR-based and GeneScanning analysis, some productive TCRβ V-DJ rearrangements were found in the CD34+ CD1a+ cells, although the sensitivity of this assay raises the possibility that a small contamination was responsible for this signal (19). A less-sensitive southern blot analysis detected the ﬁrst TCRβ V-DJ rearrangements in the CD4 ISP cells (33).

ANRV270-IY24-10

ARI

15 February 2006

21:28

icTCRβ+ TSP

ETP

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

BLOOD

small CD4 ISP

pre-T

large CD4 ISP

EDP

CD3lo DP

icTCRβ–

THYMUS

CD34

+

+

+

+

-

-

-

CD1a

-

-

+

+

+

+

+

icTCRβ

-

-

-

-

+

+/-

+

icCD3ε

?

+/-

+

+

+

+

+

CD2

-

+/-

+

+

+

+

+

CD5

-

+/-

+

+

+

+

+

CD7

+

+

+

+

+

+

+

CD4

-

-

-

+

+

+

+

CD8α

-

-

-

-

-

+

+

CD8β

-

-

-

-

-

-

+

CD45RA

+

+

-

-

-

-

-

CD45RO

-

-

-

-

+

+/-

+

IL-7Rα

-

+

+

+

+

+

+

DJβ

-

+

+

+

+

+

+

V-DJβ

-

-

-

+

+

+

+

V-Jα

-

-

-

-

-

-

+

Several lines of evidence indicate that selection for productive, in-frame TCRβ rearrangements and production of a TCRβ protein, a process referred to as β-selection, can occur in distinct populations of cells that differ in CD4 and CD8 expression. A very low frequency of productive TCRβ V-DJ rearrangements was detected in the CD34+ CD1a+ CD4− population, but expression of TCRβ protein was not analyzed (19).

Figure 3 Intermediate stages in development of early thymocyte subsets into double-positive cells. The TCR rearrangement status of the various subpopulations are based on data presented in References 33 and 34. However, it should be noted that a recent study (19) positions the onset of these rearrangements at earlier stages. This is discussed in the text. (Abbreviation: EDP, early double-positive.)

Nonetheless, these results suggest that a few cells may undergo β-selection before CD4 is expressed. However, not all populations downstream of the CD34+ CD1a+ CD4− cells are post-β-selection cells because intracytoplasmic (ic) TCRβ− populations have been found in the CD4 ISP cells (33; S. Ligthart, Y. Yasuda, H. Spits, and B. Blom, manuscript submitted) and in the CD4+ CD8α + β− early double-positive cells (33, 35). Ten to twenty

www.annualreviews.org • Development of Human Lymphoid Cells

293

ARI

15 February 2006

21:28

percent of the CD4 ISP are icTCRβ+ . In contrast to the TCRβ− CD4 ISP, the icTCRβ+ CD4 ISP were in cycle; expressed elevated levels of CD1a, CD4, CD28, CD45RO and CD71; and differentiated into TCRαβ+ T cells in a FTOC with faster kinetics than the icTCRβ− CD4 ISP (S. Ligthart, Y. Yasuda, H. Spits, and B. Blom, manuscript submitted). These data indicate that β-selection can occur within the CD4 ISP population. On the basis of the results of our studies and those of others, we propose that expression of a TCRβ protein and the ensuing β-selection occur within a certain developmental window and are not tightly coupled to regulation of CD4, CD8α, and CD8β expression. Thus, a few cells already undergo β-selection before CD4 is expressed (19). A larger proportion is β-selected after upregulation of CD4 (33; S. Ligthart, Y. Yasuda, H. Spits, and B. Blom, manuscript submitted), and a third group of the pre-T cells upregulate CD4 and CD8α before initiating and completing TCR rearrangements (35, 36).

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

The Role of Cytokines in Early T Cell Development The important role of IL-7 in T cell development is well documented. The IL-7 receptor consists of two chains, IL-7Rα and gamma common (γc), which is also part of the receptors for IL-2, IL-4, IL-9, IL-15, and IL-21. Genetic defects in the genes encoding for γc (37, 38), IL-7Rα (39, 40), or the Janus kinase Jak3, a component of the IL-7-induced signal transduction pathway (41, 42), account for the majority of severe combined immune deﬁciencies (SCID) characterized by strongly reduced numbers of T cells. The most frequent form of SCID is caused by mutations in the γc-encoding gene (reviewed in 43). In these patients, T and NK cells are absent, whereas in contrast to what is observed in γc-deﬁcient mice, B cell development is normal (37, 38, 43). IL-7Rα-deﬁcient patients also display a profound T cell deﬁciency and have near normal B cell numbers. In contrast to γc-deﬁcient 294

Blom

·

Spits

patients, IL-7Rα-deﬁcient patients have normal frequencies of NK cells in the periphery (39, 40). These manifestations argue against an important role of IL-7 in survival and proliferation of lymphoid precursors. Rather, they indicate that human T cell development speciﬁcally and critically depends on IL-7. The precise function of IL-7 in human T cell development is not fully understood. Inhibition of IL-7R signaling by blocking anti-IL-7 and anti-IL-7R antibodies prevents expansion and differentiation of developing T cells in a FTOC (44, 45), indicating an important role of IL-7 in mediating survival and proliferation of human T cell precursors. This effect is mediated by IL-7-induced PI3K activation through one tyrosine residue at position 449 in the cytoplasmic tail of the IL-7Rα chain (45). In the mouse, IL-7 does not appear to be critical for differentiation of TCRαβ cells in the thymus, as near normal distributions of subsets were observed in the thymuses of IL-7Rα-deﬁcient mice (46). There is evidence, however, that IL-7 is important for differentiation of human T cells. First, differentiation of CD34+ precursors in a FTOC in the presence of anti-IL-7R antibody is almost completely blocked at the transition of CD34+ CD1a+ cells into CD4 ISP (44). Second, according to one report, thymocytes from γc-deﬁcient infants possess TCRβ D-J but lack V-DJβ rearrangements (47). Although a function of IL-7 in TCRβ rearrangements in human pre-T cells has yet to be conﬁrmed, this ﬁnding may explain the complete lack of T cells in many γc- and IL7Rα-deﬁcient patients (40, 48). On the other hand, IL-7 is required for peripheral T cell homeostasis in humans (49), raising the alternative possibility that the absolute T cell deﬁciency in γc-deﬁcient patients is caused by a combination of defects in early T cell development and in T cell homeostasis. Taken together, the data indicate that IL7 is indispensable for human T cell development. The data also suggest that IL-7 is more critical for human than for mouse T cell development, which may be attributed to a

ANRV270-IY24-10

ARI

15 February 2006

21:28

role for IL-7 in TCRβ V-DJ rearrangements, but more data are required to conﬁrm this speculation.

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Transcription Factors Involved in Human T Cell Development Recently, it has become clear that Notch receptor ligand interactions play crucial roles in T cell development. Notch receptors control differentiation and proliferation in response to ligands on neighboring cells (50). Interaction of Notch receptors (Notch1, 2, 3, 4) with one of their ligands (DL1, 2, 4, and Jagged-1, -2) results in proteolytical cleavage of its cytoplasmic portion (icNotch), which migrates to the nucleus and interacts with C promoter binding factor (CBF)-1/recombination signal–binding protein (RBP)-J to mediate transcription of target genes. In the absence of icNotch, RBP-J represses transcription by interacting with various corepressors. When icNotch binds to RBP-J, it recruits the coactivator mastermind-like (MAML)-1, which binds to icNotch in the icNotch-CSLDNA complex, thereby converting the CSL (CBF-1/suppressor of Hairless/Lag1) complex into a transcriptional activator. Radtke et al. (51) were the ﬁrst to demonstrate that conditional deletion of Notch1 resulted in a switch in development in that, in the thymus, T cell development was inhibited and B cell development stimulated (51). More recently, Jaleco et al. (52) demonstrated that coculture of human cord blood CD34+ cells with the murine stromal cell line S17 expressing the Notch ligand DL1 resulted in generation of CD7+ cells with strong NK cell precursor activity while inhibiting B cell development (52). The cells generated in this system expressed high levels of CD7 and cytoplasmic CD3ε and may represent T/NK precursors (52). These data were conﬁrmed and extended by Zuniga-Pﬂucker and collaborators (53), who documented that cord blood and bone marrow CD34+ cells cocultured with another DL1-expressing bone marrow stro-

mal cell line, OP9-DL1, IL-7, and Flt3-L, mediated full T cell development (53, 54), similar to murine HSCs (55). Recently, the function of Notch in development of T and non-T cells in a FTOC system (26a) and in the OP9-DL1 system (55a) was investigated using the γ-secretase inhibitor 7 (N-[N-(3,5-diﬂuorophenyl)-L-alanyl]-Sphenylglycinet-butyl ester) (DAPT), which prevents proteolytic cleavage of Notch, thereby inhibiting Notch signaling. As expected, DAPT strongly inhibited T cell development, whereas development of non-T cells was stimulated. As expected, inhibition of TCRαβ development by DAPT was rescued by forced expression of icNotch1 by retrovirus-mediated gene transfer (55a). TCRβ V-DJ but not D-Jβ rearrangements were strongly inhibited in cells that developed from CD34+ CD1a− thymocytes in a FTOC in the presence of DAPT (26a), as in the mouse (56). This is not, however, the only mechanism of inhibition of T cell development by Notch1 because introduction of TCRαβ-encoding cDNA by retrovirus-mediated gene transfer into CD34+ CD1a− fails to overcome the requirement of Notch for development of these precursors into T cells (N. Legrand & H. Spits, unpublished observations). These ﬁndings indicate that Notch is required for induction and maintenance of T cell speciﬁcation, as found previously in the mouse (57). We and others also observed that, in the presence of DAPT, thymic CD34+ CD1a− cells generate more NK cells, monocytic/DCs (26a), and pDCs (55a), strongly suggesting that Notch drives lineage decisions in thymus of humans, similar to what occurs in the mouse thymus (58). Two research groups have documented that ectopic expression of icNotch1 in CD34+ cells affected T cell development in an in vitro FTOC (59, 60). icNotch1 favored development of human TCRγδ above TCRαβ cells when expressed in precursor cells before the TCRβ selection checkpoint, suggesting that a

www.annualreviews.org • Development of Human Lymphoid Cells

γc: gamma common chain of the IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21 receptor DL1: Delta-like 1

295

ANRV270-IY24-10

ARI

15 February 2006

21:28

HEB

DP TCRαβ αβ

Notch1 Notch1 CLP

GATA3

Notch1 T/NK

pre-T

E2A

Notch1

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Ets-1

Id2

γδ TCRγδ

NK

Figure 4 Transcription factors involved in T and NK cell development. The model assumes that NK cells develop from common T/NK cell precursors. It is possible, as indicated in Figure 2, that NK cells can also develop from NK/DC precursors that are unable to develop into T cells. E2A (i.e., E47) blocks NK cell development. Because Id2 sequesters E47, ablation of Id2 leads to inhibition of NK cell development, which can be neutralized by ablation of E47 (B. Kee, personal communication).

strong Notch1 signal may promote TCRγδ at the expense of TCRαβ development (59, 60). These data contrast with earlier ﬁndings in the mouse system that indicate the reverse, that diminished Notch1 signaling favors TCRγδ above TCRαβ development (61). However, De Smedt et al. (59) found that CD34+ expressing icNotch1 developed into TCRαβ cells in the bone marrow, similarly to the effect found in the mouse, indicating that in an in vivo setting there is no preferential TCRγδ development from icNotch1 overexpressing CD34+ precursor cells. Thus, the role of Notch1 in TCRαβ and γδ diversiﬁcation in humans remains to be established. Other transcription factors that have been implicated in human T cell development are members of the basic helix-loop-helix (bHLH) subfamily of E-box binding (E) proteins (Figure 4). There are four E proteins, E12, E47, HEB, and E2-2, all of which are involved in both T and/or B cell development in the mouse (62). Ectopic expression

296

Blom

·

Spits

of natural antagonists of the E proteins, the Id (inhibitors of DNA binding)-2 and Id3, in human CD34+ cells inhibits T cell development and promotes NK cell development in a FTOC (63, 64). Stimulation of NK cells by Id2 and Id3 overexpression is consistent with the fact that Id2−/− mice have few NK cells (65), and together these data indicate that the balance of Id and E proteins plays a role in T/NK lineage diversiﬁcation both in humans and mice. Interestingly, introduction of Id3 in CD4 ISP cells results in their inhibition of TCRαβ development, but not of TCRγδ development, in a FTOC (64), which could be reversed by coexpressing HEB (R. Schotte, Y. Yasuda, and H. Spits, manuscript submitted). Studies in the mouse have shown that HEB controls pTα expression (66, 67). These data, combined with the demonstration in mice that β-selection commits precursor cells to the TCRαβ lineage (68), suggest that HEB stimulates TCRαβ development by control of pTα expression (Figure 4).

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

Another transcription factor that has been associated with various stages of T cell development is GATA3 (69). Taghon et al. (70) observed that overexpression of GATA3 in human CD34+ thymic precursors resulted initially in stimulation of development into CD4+ CD8+ cells in a FTOC, consistent with a role for GATA3 in early development. However, further development of GATA3overexpressing cells was compromised, as the absolute numbers of TCRαβ cells were severely reduced at later time points. Taghon et al. (70) also found a strongly reduced expression of TCRβ protein by GATA3 overexpression and suggested that deregulated expression of GATA3 interferes with appropriate TCRβ rearrangement or translation. However, these ﬁndings are difﬁcult to reconcile with observations of conditional GATA3deﬁcient mice that suggested that GATA3 is required for TCRβ expression and pre-TCR signaling (69). Experiments, yet to be performed, to test the effect of GATA3 knockdown on T cell development should reveal the role of this transcription factor in human T cell development. In summary, the role of Notch1 in human T cell development has been established. In addition, it is clear that E proteins are required for early T cell development. It remains to be investigated how Notch1 and E proteins are linked in the control of human T cell development. Also, it has yet to be established that GATA3 is essential for early T cell development in humans.

DEVELOPMENT OF NATURAL KILLER (NK) CELLS Upon their discovery, NK cells were operationally deﬁned as leukocytes able to kill transformed cells without prior sensitization. Today, NK cells are known to play essential roles in the innate immune response, as well as in the generation of an adaptive immune response. Initially, investigators believed that NK cells were related to other cells of the

innate immune system, such as monocytes; however, work in the 1990s established that NK cells are more closely related to T cells, with whom they can share a common precursor. NK cells are functionally similar in particular to CTL, as both types of cells have the capacity to kill other cells and to produce cytokines such as interferon (IFN)-γ. NK cells probably coevolved with other cell types of the lymphoid system, in particular with T cells, because both of these lymphocytes recognize conventional and nonconventional major histocompatibility complex (MHC) molecules. Functional MHC molecules are present in cartilaginous ﬁsh, but not in more primitive species. Similarly, NK cells, as they are currently deﬁned, have not been identiﬁed in species lower than ﬁsh (71). In humans, there are two major subsets of NK cells: one expressing high levels of CD56 and low or no CD16 (CD56hi CD16+/− ), and a second that is CD56+ CD16hi (72, 73). The distribution of the cells and the functional properties of these populations are different in that CD56hi CD16+/− cells have relatively lower cytolytic activity and produce more cytokines than the CD56+ CD16hi cells (72, 73). As discussed below, the developmental relationship between these populations has yet to be resolved.

Sites of NK Cell Development NK cells can develop at multiple sites. In the fetus, NK cell precursors have been found in the bone marrow (6), liver (74), thymus (75), spleen, lymph nodes (T. Cupedo & H. Spits, manuscript in preparation), and possibly in the intestine (76). However, identiﬁcation of an NK precursor does not prove that NK cells actively develop in those organs. Such proof would require in situ identiﬁcation of committed NK cell precursors and other downstream intermediate stages. Indirect evidence suggests that the population of fetal liver CD34+ CD38+ cells contains committed NK cell precursors

www.annualreviews.org • Development of Human Lymphoid Cells

297

ARI

15 February 2006

21:28

because these cells have robust NK cell precursor activity but are unable to develop into T cells (74). The fetal thymus contains an immature CD34− CD5− CD56− population with clonogenic NK cell potential (75), as well as bipotential T/NK precursors (77). Together, these data suggest that the fetal liver and thymus are sites for NK cell development. The biological function of fetal NK cells is currently unknown. The fact that patients deﬁcient in γc, most of whom lack NK cells at birth, are born at full term suggests that these cells are not essential for the fetus during pregnancy. Whether the thymus is required for the development of human fetal NK cells is also questionable because infants with DiGeorge syndrome have normal numbers of NK cells (78). The presence of NK cells in the thymus may be a consequence of the fact that some bipotential T/NK precursors, which form obligatory intermediates in the development of multipotent precursors of T cells, by chance develop into NK cells and that these cells have no function in the thymus. The general consensus is that the bone marrow is the site for NK cell development in children and adults. One study in the mouse has identiﬁed a committed NK cell precursor, unable to develop into other lymphoid cells, in the mouse bone marrow (79), but the exact phenotype of the human bone marrow equivalent has yet to be determined. Whether NK cells fully develop and mature in the bone marrow before entering the circulation is unclear. A recent study suggests that human NK cells can develop and mature via another route. A CD34lo CD45RAhi cell that expresses high levels of the integrin α4β7 was identiﬁed in adult peripheral blood and in lymph nodes (80). These cells express c-kit, CD2, low levels of CD7, L-selectin, and IL-2Rα, but were CD10− and were able to develop into CD56hi CD16dim/− NK cells with high efﬁciency in vitro. Because of their location in the lymph nodes, these cells likely represent committed NK cell precursors, which

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

298

Blom

·

Spits

implies that CD34lo CD45RAhi cells lack T cell-, pDC-, or B cell–differentiating activities, although this was not directly veriﬁed in this study. Therefore, proof that these cells are NK lineage–restricted is lacking (80). In the lymph nodes, CD34lo CD45RAhi α4β7+ cells are located within T cell–rich regions, where CD56hi cells also reside (81). Because CD34lo CD45RAhi lymph node cells could develop into CD56hi NK cells when cultured in IL-2, it was speculated that in vivo T cell activation in the lymph nodes may promote the development of new NK cells from these precursors (80). The study by Freud et al. (80) left unresolved the question of where the CD56+ CD16hi NK cells develop because the committed NK precursors in the lymph nodes did not give rise to CD56+ CD16hi precursors. However, results of another study suggest that highly cytolytic CD56+ CD16+ lymph node NK cells arise from IL-2-activated noncytolytic CD56hi CD16+/− cells (82). In addition, two research groups have reported that IL-21, in concert with Flt3-L, SCF, and IL-15, stimulated the generation of CD56+ CD16+ NK cells in vitro (83, 84), but whether this is a reﬂection of what happens in vivo is unknown.

Early Stages in NK Cell Development Experiments both in mouse and human systems have established that T and NK cells share a common precursor (75, 85–89; reviewed in 27, 90). Despite the strong evidence for a common origin of T and NK cells, there are data suggesting that NK cells can be descendants of myeloid precursors. Marquez et al. (91) demonstrated that CD34+ CD33lo thymocyte precursors, which are probably identical to ETP, have the capacity of developing into CD34+ CD44hi CD33+ myeloid precursors. These cells were unable to develop into T cells but contained bipotent NK/DC precursors (91). Furthermore, Perez et al. (92) recently reported that a

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

CD14+ CD11b+ CD13+ CD33+ cell isolated from cord blood can develop into cytolytic CD56+ NK cells after incubation with Flt3L and IL-15. The frequency of NK precursors within the CD14+ cord blood population was low, however (1/50) (92), raising the possibility that the CD14+ cells were contaminated with CD56− NK precursors. The idea that NK cells can derive from myeloid precursors is not supported by studies that examined the lymphoid and myeloid precursor potential of CLP, common myeloid precursors (CMP), and granulocyte/monocyte precursors (GMP) both in humans and in mice. Rather, these studies concluded that NK cells develop predominantly from lymphoid precursors (7, 9, 16, 93). However, as putative human lymphoid precursors retain some GM potential and lack erythrocyte precursor activity (10), the issue of whether in humans there are bipotential precursors able to develop into NK cells and monocytes and/or DC, but not into T cells, remains to be settled. Although early cellular stages in human T and B cell development have been reasonably well deﬁned, our knowledge about early stages of human NK development is very limited. Roughly, three stages can be deﬁned

in NK cell development: lineage commitment, NK receptor repertoire selection, and functional maturation (94, 95). Only very recently has a putative committed NK precursor within the CD34+ cell compartment been identiﬁed (80). These CD34lo CD45RA+ α4β7hi CD7+/− CD10− NK precursors that give rise to CD56hi CD16− NK cells in vitro may be downstream from CD34+ CD45RA+ CD7+ CD10− (10) or CD34+ CD45RA+ CD7+ CD10+ (15) lymphocyte precursors. The immature NK cells developing from committed NK cell precursors are deﬁned by expression of CD161 (NKR-P1) (96). These cells do not express CD56 or CD16. Immature NK cells can be induced to express these markers as well as the activating and inhibitory receptors, CD94NKG2A and killer inhibitory receptors (KIR), upon culture with stromal cells and cytokines such as IL-15 (97) or Flt3-L (98). Figure 5 shows a hypothetical model of NK cell development that is deduced from phenotypic and functional analyses of small populations of immature NK cells isolated from different sources and from in vitro experiments (74, 98, 99; reviewed in 94). As discussed in the previous paragraph, the model, which is adapted

COMMITMENT

T/NK

CLP

precursor

CD34+ CD45RA+ CD7+ (CD10+)

CD34+ CD45RA+ CD7+ CD10 –

CMP: common myeloid precursor GMP: granulocyte/ monocyte precursor

MATURATION

NKp

NKi

NK

CD34lo CD45RA+ CD7lo CD10– CD122lo α4β7hi

CD34CD161+ (NKR-P1) CD16– CD56– CD122+

CD161+ CD56hi CD16 – ?

?

?

CD161+ CD56+ CD16+

www.annualreviews.org • Development of Human Lymphoid Cells

Figure 5 Model of NK cell development modiﬁed from Reference 94. Committed NK precursors develop into CD56hi CD16− NK cells (80), but the immediate precursor of the CD56+ CD16+ NK cells is not yet known. The dashed lines indicate the various possibilities. 299

ANRV270-IY24-10

ARI

15 February 2006

21:28

from Reference 94, does not resolve the developmental pathway of CD56hi CD16−/+ and CD56+ CD16hi subsets. There is no consensus whether CD56hi NK cells are precursors of CD56+ CD16hi NK cells. These subsets may arise from different precursors, but so far no evidence supports this speculation.

Cytokines Involved in Human NK Cell Development Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

A number of cytokines have been identiﬁed that can support development of human NK cells from CD34+ cells in vitro, notably SCF, Flt3-L, IL-7, IL-2, and IL-15. Data from mouse experiments indicate that SCF and Flt3-L may act on early lymphoid precursors, causing their development and thereby promoting NK cell development (reviewed in 94). However, the exact roles of SCF and Flt3-L in human NK cell development have not been determined. IL-2 and IL15 were previously proposed to be the most relevant factors for human NK cell development. There is now consensus that IL-15, and not IL-2, is the critical NK cell factor, a ﬁnding that is supported by the strongly reduced NK cell numbers in mice deﬁcient in IL-15 or in its receptor components, IL15Rα, β, and γc. SCID patients suffering γc deﬁciencies lack NK cells (43), whereas IL2-deﬁcient (100) and IL-2Rα-deﬁcient patients have normal numbers of NK cells (94, 101). One SCID patient, who presented with an absence of NK cells and reduced T cell numbers, was reported to have a strongly reduced expression of the IL-2/15Rβ chain and a marked decrease of signaling through this receptor (102). Data in the mouse suggest that IL-15 is not a differentiation factor but serves to maintain the viability and support proliferation of developing NK cells because NK cell precursors are present in normal numbers in mice deﬁcient for IL-15 or its receptor components, IL-15Rα, β, and γc (103). It is unknown whether NK precursors are present in γc-deﬁcient patients, partly be300

Blom

·

Spits

cause our knowledge of early NK cell precursors in humans is limited. However, the limited availability of the patient’s tissues and the scarcity of NK precursors are more important roadblocks to resolving these questions. Another cytokine implicated in NK cell development and function is IL-21 (83), but this cytokine acts only in concert with others, including SCF, Flt3-L, and IL-15, and may play a role in later stages of NK development (84). Mice lacking IL-21 develop normal numbers of NK cells (104), which is consistent with the notion that this cytokine may be affecting the function, not development, of NK cells.

Transcription Factors Involved in NK Cell Development Our knowledge of transcription factors that are speciﬁcally involved in NK cell development is very limited compared with what we know about their roles in T and B cell development. Overexpression of Id2 and Id3 in human CD34+ cells blocks their development into T cells, B cells, and pDCs but by contrast promotes NK cell development (Figure 4). As discussed above, Id2 and Id3 are natural antagonists of E proteins. Id2 also controls NK development in mice: Id2deﬁcient mice lack NK cells and NK cell precursors (105). However, Id2 is not a transcription factor but rather sequesters E proteins, thereby preventing their activity (62). Thus, the ﬁndings that Id2 overexpression stimulates NK cell development and that Id2 ablation prevents this indicate that one or more of the E proteins inhibit NK cell development. Indeed, overexpression of E12 and E47 inhibits NK cell development in vitro (R. Schotte & H. Spits, manuscript in preparation). Moreover, introduction of E47 deﬁciency into Id2−/− mice rescues NK cell development (B. Kee, personal communication), indicating that Id2 stimulates NK cell development by sequestering the inhibitory factor E47. E47 is an example of a factor that is required for the development of one lineage

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

(B cells) but that inhibits alternative lineage choices. Another example is Spi-B, which is required for pDC development yet inhibits the development of NK cells (106). The role of Notch1, which drives T cell commitment, is complicated in NK cell development. Triggering Notch by DL1 drives human CD34+ cord blood cells to T/NK cell precursors (52), but inhibition of Notch signaling in a FTOC system that allows for both T and NK cell development inhibits T and stimulates NK cell development (26a). A similar observation has been made in experiments that examined the effect of γ-secretase inhibitor on rat thymus development (107). Thus, although Notch signaling may be required for the early steps in the development of HSCs into NK cells, it may not be required in later stages. A master transcription factor for NK cells has not been identiﬁed. One possible candidate is Ets-1, which is predominantly expressed in lymphoid cells (108). Ets-1deﬁcient mice have reduced numbers of NK cells, and the few NK cells present in these mice are nonfunctional (109). No follow-up studies have been reported, and there is no information about the stage in which NK development is inhibited. Moreover, the role of Ets-1 in human NK cell development has not been investigated.

DEVELOPMENT OF DENDRITIC CELLS Several types of DCs, including interstitial DCs, Langerhans DCs, blood DCs, and pDCs, have been identiﬁed that, depending on their local microenvironment, mediate different types of immune responses (110, 111). The remarkable versatility and ﬂexibility of DCs, instructed by the priming signals from microbial and tissue-derived factors, distinguish them from other cell types in the immune system that exhibit more limited functions and suggests that DCs may play a central role in integrating the innate and adaptive aspects of various immune responses. In the mouse, there are

three main DC subsets: CD8α − DCs, CD8α + DCs, and B220+ Ly6C+ pDCs. The relationship between the CD8α + and CD8α − subsets is unclear, but for the sake of simplicity these cells are often categorized as conventional DCs (cDCs). The B220+ Ly6C+ pDCs, also referred to as IFN-producing cells (IPCs), are clearly a distinct subset. Human cDCs are HLA-DR+ cells that express high levels of CD11c and consist of a major BDCA3− and a minor BDCA3+ population (112). Human HLA-DR+ pDCs are deﬁned by absence of CD11c expression and by high levels of CD123 (the IL-3Rα chain) and BDCA2 (111). The CD11c+ HLADR+ BDCA3− population can be further subdivided into CD16+ and CD16− populations (112). The developmental and functional relationships of the multiple HLA-DR+ DC populations remain unclear. For the remainder of this review, we use the terms cDC and pDC to denote populations expressing CD11c and CD123, respectively.

cDC: conventional dendritic cell

Developmental Origin of cDCs and pDCs Galy and coworkers (6) showed that a fraction of bone marrow CD34+ CD45RA+ CD10+ lymphoid progenitors could develop into CD1a+ CD14− DCs. Later work of the Galy group showed that the in vitro conditions that induced development of CD1a+ CD14− DCs failed to induce differentiation of myeloid cells into cDCs, suggesting that development of DCs from lymphoid or myeloid precursors has different signaling requirements (113). The underlying mechanisms have not been elucidated. Other groups have shown that lymphoid-restricted progenitors from cord blood (CD34+ CD38− CD45RA+ CD7+ ) when cultured in standard conditions (SCF/Flt3-L/GM-CSF/TNF-α) (114) or lymphoid conditions (S17 with IL-3, IL-6, c-kit, Flt3-L) (15) were able to give rise to a similar type of DC (CD1a+ CD14− ). Because some of these CD1a+ DCs expressed S100 and low levels of CD207/Langerin,

www.annualreviews.org • Development of Human Lymphoid Cells

301

ARI

15 February 2006

21:28

Canque et al. (114) suggested that they are Langerhans DCs. More recently, Manz and coworkers (16) tested cord blood CLP, CMP (CD34+ CD38+ CD123+ CD45RA− ), and GMP (CD34+ CD38+ CD123+ CD45RA+ ) for cDC precursor activity and conﬁrmed that cDCs (CD11c+ CD123lo ) can arise from CLP, CMP, and GMP. However, they also noted that the frequency of cDC precursors in CMP and GMP was higher than in CLP (16). As is discussed below, this conclusion is valid only if one assumes that the tested populations are homogeneous. Information about the early stages in development of human HSCs into pDCs is accumulating. We have postulated that the direct progenitor of pDCs is contained within the CD34lo compartment of cord blood, fetal liver, and bone marrow. These progenitors coexpress CD45RA, CD4, and high levels of CD123, and so we have denoted these cells pro-pDCs (for progenitor of pDCs) based on their phenotypic and functional similarities with pDCs (115). The pro-pDCs may be restricted to the pDC lineage, but whether the pro-pDCs lack potential to develop into other lineages has not formally been tested. In early studies, several research groups found that pDCs express lymphoid-related genes, including CD2, CD5, CD7, pTα, Spi-B, and λ5; contain IgH D-J rearrangements (116– 120); and lack myeloid-associated markers, including CD11c, CD11b, CD13, CD33, and Mannose receptor (121). On the basis of these ﬁndings, it was postulated that pDCs derive from lymphoid progenitors. This idea was tested recently by Chicha et al. (16). They observed that single cells among CD34+ CD7+ and/or CD34+ CD10+ cells from cord blood considered to be lymphoid precursors were able to develop into pDCs, cDCs (CD11c+ CD123lo ), and B cells (16). Cord blood–derived CMP and GMP (or CD34+ CD38+ CD123+ CMP/GMP together when tested in parallel with CLP) also gave rise to both pDCs and cDCs but not to B cells (16). By using a limiting dilution assay, Chicha et al. observed that sin-

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

302

Blom

·

Spits

gle cells with pDC and cDC potential are more frequent within CMP/GMP (16) than within CLP. These results were taken as evidence that human pDCs are predominantly myeloid derived (16). However, in this study (16) lymphoid precursors were not puriﬁed on the basis of CD45RA expression, ignoring that a proportion of CD34+ CD7+ cord blood cells lacks expression of CD45RA and is unable to develop into NK cells (114). Therefore, these CD34+ CD7+ CD45RA− cells cannot qualify as lymphoid precursors. Nonetheless, it seems likely that, akin to cDCs, human pDCs can develop from both lymphoid- and myeloid-restricted precursors, which is consistent with what was observed earlier in mouse models (9, 16, 122–124). Whether lymphoid-derived cDCs and pDCs are identical to their myeloid-derived counterparts remains unclear. Relative differences in levels of expression of pDC surface antigens (114) and pTα and Spi-B transcripts (16) have been observed, but none of these markers was discriminating. There are data indicating that different subsets of pDCs and cDCs exist in vivo. Differences between thymic and peripheral pDCs have been noted in expression of CD2, CD5, and CD7 (125, 126). Mutually exclusive pDC subsets were identiﬁed in Flt3-L-treated healthy volunteers (127). In their blood, CD7+ CD56− and CD7− CD56− pDCs express pTα, whereas CD56+ CD7− pDCs lack pTα expression (127). In addition, several distinct cDC subsets have been described (112). Whether the various cDC and pDC subsets, most of which are poorly characterized, have different developmental origins is, however, unclear.

Growth Factors Involved in pDC Development It has now been well established that Flt3-L is important for human, mouse, and rhesus monkey DC development in vitro (115, 128, 129) and in vivo (130–134). Mouse pDCs can be generated only from Flt3+ CLP or Flt3+ CMP (135, 136). In humans, Flt3 mRNA has

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

been observed in both CLP and CMP populations (9, 16). Moreover, CD34+ CD45RA− early progenitors are responsive to Flt3-L alone, without feeder cells to support pDC development, conﬁrming Flt3 cell surface expression on at least some of these cells (115). The observation that Flt3-L-deﬁcient mice were severely depleted of all DC subsets underscores the signiﬁcance of Flt3-L in DC development (137). Moreover, mice deﬁcient for STAT3, which mediates the checkpoint of Flt3-L but not GM-CSF-induced DC development, and mice lacking Gﬁ1, a zinc ﬁnger repressor molecule that regulates levels of STAT3, demonstrate a severe block in DC development (138, 139). Development of cDCs and pDCs can be differentially regulated by various cytokines. For example, cytokines that promote cDC development, including GM-CSF and TNF-α, have a strong inhibitory effect on the Flt3-L-induced generation of both human and mouse pDCs (115, 128). Thrombopoietin, which alone does not induce pDC development, enhances the effect of Flt3-L on human pDC development (140). Another cytokine, G-CSF, signiﬁcantly increased the number of blood pDCs when injected into healthy volunteers (134). Because G-CSF alone did not induce pDC development from HSCs (115), these data suggest that it increases pDC levels by stimulating mobilization of pDC or its progenitors from the bone marrow.

Transcriptional Control of Human DC Development Several transcription factors involved in differentiation of DCs have now been identiﬁed. These include Ikaros (141) and members of the Rel family of transcription factors (142). Information about transcription factors in development of cDCs in humans is scarce. Wu and collaborators (141) observed that mice homozygous for an Ikaros dominantnegative mutation that lack all cells of lymphoid origin, including T, B, and NK cells, also show a defect in development of a sub-

set of cDCs, presumably lymphoid-derived cDCs. Consistent with these observations, Galy et al. (113) reported that ectopic expression of this dominant-negative form of Ikaros interfered with development of lymphoid but not of myeloid cDCs. These ﬁndings suggest that Ikaros is involved in development of lymphoid-derived cDCs in both humans and mice. More is known about the role of transcription factors in human pDC development (Figure 6). CD34+ hematopoietic precursor cells ectopically expressing either Id2 or Id3 are blocked in their development into pDCs but not into myeloid cells (126). These ﬁndings are in line with observations in mice lacking Id2, which have increased numbers of pDCs (143). As discussed above, Id proteins are not transcription factors but act by sequestering E proteins. It remains to be elucidated, however, which member(s) of the bHLH family is (are) required for pDC development. We recently found that the Ets family member Spi-B is expressed in pDCs and its mature progeny but not in cDCs (119, 144), a ﬁnding that has been conﬁrmed in the mouse (118). Moreover Spi-B transcripts have been observed in in vitro CLP-/CMPderived pDC, but not in other DC subsets (16). Spi-B is involved in development of human pDCs because downregulation of Spi-B by RNA interference in human CD34+ progenitor cells compromises their development into pDCs in vitro and in vivo (145). In contrast, knockdown of Spi-B stimulates myeloid development (145). RNA interference experiments further indicate that another Ets family transcription factor, PU.1, which is highly homologous to Spi-B, is needed for development of both pDCs and CD14+ myeloid cells (145). These data suggest that PU.1 acts at an earlier stage of pDC development than does Spi-B, which is consistent with data in the mouse that show that PU.1 is required for development and survival of myeloid and lymphoid precursors (146) (Figure 6). Studies in mice have shown that pDC development also depends on members of the

www.annualreviews.org • Development of Human Lymphoid Cells

303

ANRV270-IY24-10

ARI

15 February 2006

21:28

T/NK Notch1

PU.1 HSC

E protein CLP

Spi-B

pDC

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

E2A EBF Pax-5 B

Figure 6 Transcription factors involved in pDC development. The Ets transcription factor Spi-B is involved in development of human pDCs, as shown by the fact that its ablation results in inhibition of human pDCs both in vivo and in vitro (145). Ectopic expression of Spi-B results in inhibition of T, B, and NK cell development. Knocking down Spi-B leads to an increase in B cell and monocyte development. These ﬁndings suggest that Spi-B can also play a role in regulation of pDC/monocyte diversiﬁcation. Based on inhibition of pDC development by Id2 or Id3 (126), an E protein should be involved in pDC development, but the nature of this E protein is unknown. (Abbreviation: EBF, early B cell factor.)

IRF: IFN regulatory factor

304

IFN regulatory factor (IRF) family, including IRF-8 [or IFN consensus sequence-binding protein (ICSBP)] and to a lesser degree IRF-4 (147, 148). To date no studies have been performed to determine their role in human pDC development. Interestingly, IRF-8, in addition to binding to other members of the IRF family, can also bind to Spi-B and to members of the bHLH family to form transcriptional complexes apparently critical for the regulation of the immune system (149–151). Therefore, it is tempting to speculate that pDC development is driven by a complex of IRF-8, Spi-B, and a bHLH protein. Recently, the role of Notch signaling in development of pDCs was investigated. In agreement with an observation by Kincade’s group that interaction of DL1 with Notch inhibited pDC development from murine bone marrow precursor cells in vitro (118), we Blom

·

Spits

found that DL1 inhibited pDC development both from thymic and fetal liver CD34+ cell precursors (55a). Inhibition of pDC development of thymic precursors by Notch signaling could be overcome by overexpressing SpiB. Thus, Notch, in concert with Spi-B, may regulate T/pDC lineage diversiﬁcation in the thymus.

DEVELOPMENT OF B CELLS Most hematopoietic cell lineages, including B cells, are believed to originate in the bone marrow. As in mice, all cellular stages of early B cell development in humans can be detected in fetal bone marrow as well as in bone marrow of young and aged adults (see below), indicating that B cell development proceeds throughout life (152). Early in ontogeny, B cell development may take place

ANRV270-IY24-10

ARI

15 February 2006

21:28

in the omentum (153) and fetal liver. The presence of early B cell–biased precursors in cord blood was recently demonstrated (10, 154, 155), and, even after birth, B cells may develop at sites other than the bone marrow. Notably, pre-B cells and mature B cells have been shown to be present in the thymus of young children (26).

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Early Stages of B Cell Development In the early 1990s, Hardy et al. (156) introduced a differentiation pathway for B220+ progenitors in mouse bone marrow. They showed, on the basis of differential surface expression of CD43, BP-1, and HSA, that B220+ progenitors develop via pre-pro-B, early pro-B, late pro-B, and pre-B into B cells. Functionally similar populations of early, pro-, and pre-B cells have been identiﬁed in humans, but the transition of common lymphocytes into B cell–restricted or –committed B cell precursors is as yet poorly deﬁned. The current consensus is that human B lineage–restricted cells pass through a CD34+ CD19− CD10+ early B, CD34+ CD19+ CD10+ pro-B, large CD34+ CD19+ CD10+ pre-BI, large CD34− CD19+ CD10+ pre-BII, small CD34− CD19+ CD10+ pre-BII cell development pathway (Figure 7). As discussed above, human CLP appear to be present in CD34+ CD45RA+ cells that express CXCR4 and that also express CD10 or CD7, or both (6, 10, 13–15). The CD34+ CD45RA+ CD10+ CD7− cells express IL-7Rα, and although these cells, found in bone marrow as well as cord blood, contain precursors able to develop into other lymphoid lineages, they are biased to develop into B cells (8, 10). Other studies have documented that early B cells, which express CD34, CD10, and IL-7Rα but lack CD19, have initiated DJH rearrangements (157) and express cytoplasmic CD79a (158) and Vpre-B (159). Although these studies suggest the initiation of B cell commitment in a CD34+ CD45RA+ CD10+ IL-7Rα + cell

population, other studies have found cells with B cell characteristics in a small fraction of bone marrow and cord blood, CD34+ cells that lacked CD10 and CD19 but that expressed CD79a transcripts (3.3%) and protein (0.4%) (154, 155). In addition, upon in vitro culture of CD34+ CD10− CD19− cord blood progenitor cells with mouse stromal cells and cytokines (SCF, IL-2, IL-15), these cells ﬁrst acquire CD79a and IL-7Rα before CD10 and CD19 (154). These CD34+ CD79a+ IL-7Rα + CD10− CD19− cells had DJH but no VDJH rearrangements and faintly expressed Pax-5 transcripts, which is considered the hallmark of B cell commitment in the mouse (160). Although these ﬁndings suggest that CD79a+ IL-7Rα + cells are committed to the B cell lineage, in vitro–generated CD79a+ IL-7Rα + pro-B cells were not restricted to the B cell lineage but could still differentiate into macrophages, NK cells, and some T cells in vitro (154). The developmental potential of freshly puriﬁed CD34+ CD79a+ CD10− CD19− cord blood cells remains to be assessed. However, considering that the possibly more differentiated CD34+ CD19− CD10+ cord blood cells still have B/NK potential and even retained some myeloid-erythroid potential (15), it seems fair to assume that CD34+ CD79a+ CD10− CD19− cells are multipotential. A developmental pathway of human preB cell subpopulations in bone marrow was proposed by Ghia et al. (161), who used single-cell PCR analyses to determine the expression of TdT and RAG (recombination activating gene)-1 transcripts, Vpre-B protein, the cycling status, and the conﬁguration of the IgH and IgL chain alleles in human B lineage subpopulations (reviewed in 162) (Figure 7). Human pro-B cells in the bone marrow are a well-characterized population expressing CD34, CD10, and CD19 (163). The vast majority of pro-B cells express TdT (161, 163, 164), some express Vpre-B at the cell surface (159, 161), and V-DJH rearrangements are easily detected (157, 165). Only a

www.annualreviews.org • Development of Human Lymphoid Cells

305

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

CLP

early B

pro-B

pre-BI

large pre-BII

CD34

+

+

+

-

-

-

-

-

CD10

+

+

+

+

+

+

+

-

IL-7Rα

+

+

+

-

-

-

-

-

CD19

-

-

+

+

+

+

+

+

CD79a

-

+

+

+

+

+

+

+

TdT

-

-

+

-

-

-

-

-

RAG

-

-

+

+

-

+

+

-

Vpre-B

-

+

+

+

+

-

-

-

μH

-

-

+/-

+

+

+

+

+

pre-BCR

-

-

-

-

+

-

-

-

lgH

GL

DJH

VHDJH

VHDJH

VHDJH

VHDJH

VHDJH

VHDJH

κL

GL

GL

GL

GL

GL

VL J L

VLJL

VL JL

cycling

-

-

-

+

+

-

-

-

Pax-5

-

-

+

+

+

+

+

+

slgM

-

-

-

-

-

-

+

+

slgD

-

-

-

-

-

-

-

+

Figure 7 Model of early stages of human B cell development. The model is adapted from a model presented in Reference 162.

21:28

fraction of the pro-B cells express cytoplasmic μHC (161, 166, 167). Recently, a subpopulation (20%) of CD34+ CD19+ bone marrow cells was described that lack CXCR4 (168), which is the receptor for the chemokine CXCL12 (SDF-1). CD34+ CD19+ CXCR4− cells rapidly express CXCR4 and differentiate into mature κ or λ positive B cells in vitro (PHA-stimulated PBL–conditioned medium, SDF-1, IL-6). Surprisingly, however, a fraction of these cells (12%) were also able to form myeloid colonies containing granulocytes, macrophages, or erythrocytes (168). 306

Blom

·

Spits

small immature mature pre-BII B B

If we assume that human CD19 expression, like in the mouse, is dependent on Pax-5 (169, 170), which restricts the developmental options of early progenitors to the B cell pathway (160), these ﬁndings suggest that lineagerestricted pro-B cells retain a certain plasticity and may develop into either B cells or myeloid cells, depending on the organism’s requirements and subsequent bone marrow microenvironmental cues. Alternatively, the observed myeloid differentiation potential of CD34+ CD10+ CD19+ bone marrow cells in the study by Hou et al. (168) may derive from

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

cells that lack Pax-5 expression. In line with this, Sanz et al. (155) showed that in cord blood 35% of CD34+ CD10+ CD19+ cells do not express Pax-5. Although this remains to be conﬁrmed in bone marrow cells, it will be interesting to determine how CD19 expression is regulated in the absence of Pax-5 in these cells. Pro-B cell differentiation into pre-BI cells is characterized by loss of CD34 and TdT and by acquisition of cytoplasmic μHC in more than 95% of the cells (161, 163, 164). Based on cell-cycle analysis, human pre-B cells can be generally subdivided into large proliferating pre-BI cells, large proliferating pre-BII cells, and small resting pre-BII cells (161). Pre-BI cells that bear in-frame productive (V-DJ)H rearrangements transport the μH chains to the membrane associated with L (Vpre-B and λ5) and CD79 to form the pre-B cell receptor, which signals several rounds of division. The large cycling pre-BI cells that express functional μH downregulate CD34, TdT, RAG-1, and RAG-2 and maintain expression of L, CD10, and CD19 genes. To allow for Ig κ or λ L gene recombination, RAG-1 and RAG-2 expression are reinduced in small resting CD34− CD19+ CD10+ pre-BII cells.

Growth Factors Involved in B Cell Development In mice, the c-kit, Flt3, and IL-7R signaling systems together account for the generation of all B lymphocytes in the bone marrow of adult mice (160, 171). In conspicuous contrast to what is known about the role of cytokines in mouse B cell development, the essential growth factors required for the growth of normal human B cell precursors remain elusive, although in vitro studies do suggest some function of c-kit, Flt3, and IL-7 in proliferation and survival of early pro-B cells (167, 172; reviewed in 162). However, in contrast to the well-documented requirement of IL-7 for mouse B cell development, this cytokine is not needed for human B cell development

in vitro (173) or in vivo (39, 40). Data obtained from SCID patients lacking expression of either IL-7Rα (39, 40), γc (37), or Jak3 (42, 48, 174), which are all components involved in the IL-7R signaling pathway, have normal or even increased levels of circulating B cells, although their function is greatly impaired (175). The presence of B cells in IL-7Rαdeﬁcient patients also excludes an important role of TSLP (thymic stromal-derived lymphopoietin) in human B cell development because its receptor consists of the IL-7Rα chain, in addition to a speciﬁc γc-homolog, TSLP-R. Moreover, the addition of TSLP has no effect on in vitro development of human CD34+ cells cocultured with the murine stromal cell line S17 (B. Blom, unpublished data).

Transcriptional Control of Human B Cell Development From mice studies, a wealth of information is available on the transcriptional control of early B cell commitment and development, which has been extensively reviewed by Busslinger (160). Development of HSCs via lymphoid progenitors into early B cells requires the concerted actions of multiple transcription factors, including Ikaros, PU.1, E2A, early B cell factor (EBF), and Pax-5. PU.1 and Ikaros act in parallel pathways to control transition of HSCs into lymphoid precursors. E2A, EBF, and Pax-5 regulate development of CLP into early B cells. E2A proteins are assumed to control EBF, which in turn regulates Pax-5. Not surprisingly, the limited information about expression of transcription factors in precursor cell populations and transcriptional control of human B cell development suggests that data obtained in murine systems can be largely extrapolated to the human situation. CD34+ Lin− CD10− CD7− cord blood cells, like mouse progenitor cells (146, 176), express the Ets family member PU.1 (154). Knockdown of PU.1 in human CD34+ CD38− fetal liver progenitors

www.annualreviews.org • Development of Human Lymphoid Cells

307

ARI

15 February 2006

21:28

results in inhibition of B cell, monocyte, and pDC development, indicating a role for this transcription factor in development of both myeloid and lymphoid precursors (145). Multipotent CD34+ lin− CD10− CD7− cord blood cells lack expression of Pax5 and EBF (154, 177), whereas early B cells (CD34+ lin− CD10+ CD19− CD7− ) from adult bone marrow were reported to have markedly upregulated expression of Pax-5 relative to the more primitive CD34+ lin− CD10− CD7− subset or CD34+ lin− CD10− CD7+ NK-/T-biased precursors (13) and pro-B cells (CD34+ CD19+ ) in cord blood (155). These ﬁndings correlate well with observations in the mouse, where CLP already express transcripts of various lineage-speciﬁc transcription factors, including Pax-5. Indirect evidence for a role for EBF in human B cell development comes from ﬁndings that early hematopoietic zinc ﬁnger (EHZF, the mouse homolog of Evi3), which inhibits the transcriptional activity of EBF, is highly expressed in CD34+ cells but absent in CD19+ B cells (178). A study by Jaleco et al. (179) strongly suggests that E2A proteins are required for human B cell development because forced expression of one of their antagonists, Id3, strongly inhibits B cell development in vitro. Interestingly, following ectopic expression of Id3 in CD34+ CD38− CD10− fetal liver cells, their development into CD10+ IL-7Rα + cells was almost completely inhibited (179). These data indicate that E2A proteins are required for generation and/or survival of CD10+ IL-7Rα + B cell–biased precursors (8, 10). These ﬁndings are in accord with the observation that expression of another member of the Id family, Id1, is strongly

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

reduced in CD34+ CD10+ lin− Pax-5+ EBF+ cord blood progenitor cells (13).

CONCLUDING REMARKS Our understanding of human lymphocyte development has increased signiﬁcantly over the past 20 years. In particular, our understanding of human T and B cell development has improved. Nonetheless, there are many gaps in our knowledge, particularly regarding the early stages of development of HSCs into lymphoid-restricted precursors. Much remains to be learned about the phenotype of human CLP and downstream lineage-biased and -restricted precursors. Despite enormous progress in our understanding of the function of NK cells and the mechanisms underlying their activities, our knowledge about the developmental pathways and the mechanisms of development of human NK cells is still surprisingly limited. The same holds true for our knowledge of DC development. One problem is that the different subsets of cDCs and pDCs are still poorly deﬁned. Future studies in the ﬁeld of human lymphocyte development will undoubtedly make use of recently described mouse strains that support development of all human lymphocyte classes (180–182). The possibility of stably expressing siRNA-encoding DNA fragments in human precursor cells, which is transferred to the mature progeny in an in vivo setting (181), allows for mechanistic studies in the ﬁeld of human hematopoietic development that were not possible until recently. We may also expect that future investigations in patients with deﬁned deﬁciencies in lymphocyte development will yield valuable information about the underlying mechanisms.

SUMMARY POINTS 1. CLPs coexpress CD34, CD45RA, and CD7 but are still poorly deﬁned. 2. T cell development in the human thymus occurs via a series of well-deﬁned intermediate stages.

308

Blom

·

Spits

ANRV270-IY24-10

ARI

15 February 2006

21:28

3. T cells require IL-7 and NK cells require IL-15 for their development. The importance of IL-7 in human T cell development may be more stringent than it is in mice. 4. Our understanding of early stages in NK cell development is improving, but the identity of the earliest NK-restricted precursor is not yet ﬁrmly established. The developmental relationship of the two major human NK cell subsets, CD56hi CD16− and CD56+ CD16hi , remains unresolved.

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

5. cDCs and pDCs can develop from both lymphoid and myeloid precursors. There are multiple subsets of cDCs and pDCs, but the developmental relationship of those subsets is as yet unknown. 6. Development of lymphocytes is driven by transcription factors. The essential role of Notch1 in human T cell development has been demonstrated. bHLH factors belonging to the E protein family control T cell, B cell, and pDC development and negatively regulate NK cell development. Spi-B, probably in concert with other factors, such as IRF-8 and members of the bHLH E protein family, is indispensable for pDC development. 7. We have a clear understanding of early stages of human B cell development. It is completely unknown, however, which cytokine(s) is (are) driving human B cell development. Although IL-7 is essential for B cell development in mice, it is not required for B cell development in humans.

FUTURE ISSUES TO BE RESOLVED 1. We need to develop a uniform deﬁnition of human pluripotent stem cells and CLPs and to determine their precursor potential on a clonal level. 2. Improved mouse models are needed to test the developmental potential of human lymphocyte precursors in an vivo setting. 3. We need improved methods that completely knockdown genes in primary hematopoietic precursors to elucidate the roles of key genes in human lymphoid development. 4. The developmental origin and function of various NK cell and DC subsets need to be elucidated.

LITERATURE CITED 1. Adolfsson J, Mansson R, Buza-Vidas N, Hultquist A, Liuba K, et al. 2005. Identiﬁcation of Flt3+ lympho-myeloid stem cells lacking erythro-megakaryocytic potential: a revised road map for adult blood lineage commitment. Cell 121:295–306 2. Katsura Y. 2002. Redeﬁnition of lymphoid progenitors. Nat. Rev. Immunol. 2:127–32 2a. Maillard I, Fang T, Pear WS. 2005. Regulation of lymphoid development, differentiation, and function by the Notch pathway. Annu. Rev. Immunol. 23:945–74 2b. Murre C. 2005. Helix-loop-helix proteins and lymphocyte development. Nat. Immunol. 6(11):1079–86 www.annualreviews.org • Development of Human Lymphoid Cells

309

ARI

15 February 2006

21:28

2c. Radtke F, Wilson A, Mancini SJ, MacDonald HR. 2004. Notch regulation of lymphocyte development and function. Nat. Immunol. 5(3):247–53 2d. Yokoyama WM, Kim S, French AR. 2004. The dynamic life of natural killer cells. Annu. Rev. Immunol. 22:405–29 3. Payne KJ, Crooks GM. 2002. Human hematopoietic lineage commitment. Immunol. Rev. 187:48–64 4. Gore SD, Kastan MB, Civin CI. 1991. Normal human bone marrow precursors that express terminal deoxynucleotidyl transferase include T-cell precursors and possible lymphoid stem cells. Blood 77:1681–90 5. Terstappen LW, Huang S, Picker LJ. 1992. Flow cytometric assessment of human T-cell differentiation in thymus and bone marrow. Blood 79:666–77 6. Galy A, Travis M, Cen D, Chen B. 1995. Human T, B, natural killer, and dendritic cells arise from a common bone marrow progenitor cell subset. Immunity 3:459–73 7. Kondo M, Weissman IL, Akashi K. 1997. Identiﬁcation of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91:661–72 8. Ryan DH, Nuccie BL, Ritterman I, Liesveld JL, Abboud CN, Insel RA. 1997. Expression of interleukin-7 receptor by lineage-negative human bone marrow progenitors with enhanced lymphoid proliferative potential and B-lineage differentiation capacity. Blood 89:929–40 9. Manz MG, Miyamoto T, Akashi K, Weissman IL. 2002. Prospective isolation of human clonogenic common myeloid progenitors. Proc. Natl. Acad. Sci. USA 99:11872–77 10. Haddad R, Guardiola P, Izac B, Thibault C, Radich J, et al. 2004. Molecular characterization of early human T/NK and B-lymphoid progenitor cells in umbilical cord blood. Blood 104:3918–26 11. Barcena A, Muench MO, Galy AH, Cupp J, Roncarolo MG, et al. 1993. Phenotypic and functional analysis of T-cell precursors in the human fetal liver and thymus: CD7 expression in the early stages of T- and myeloid-cell development. Blood 82:3401– 14 12. Miller JS, Alley KA, McGlave P. 1994. Differentiation of natural killer (NK) cells from human primitive marrow progenitors in a stroma-based long-term culture system: identiﬁcation of a CD34+ 7+ NK progenitor. Blood 83:2594–601 13. Rossi MI, Yokota T, Medina KL, Garrett KP, Comp PC, et al. 2003. B lymphopoiesis is active throughout human life, but there are developmental age-related changes. Blood 101:576–84 14. Ishii T, Nishihara M, Ma F, Ebihara Y, Tsuji K, et al. 1999. Expression of stromal cellderived factor-1/pre-B cell growth-stimulating factor receptor, CXC chemokine receptor 4, on CD34+ human bone marrow cells is a phenotypic alteration for committed lymphoid progenitors. J. Immunol. 163:3612–20 15. Hao Q-L, Zhu J, Price MA, Payne KJ, Barsky LW, Crooks GM. 2001. Identiﬁcation of a novel, human multilymphoid progenitor in cord blood. Blood 97:3683–90 16. Chicha L, Jarrossay D, Manz MG. 2004. Clonal type I interferon-producing and dendritic cell precursors are contained in both human lymphoid and myeloid progenitor populations. J. Exp. Med. 200:1519–24 17. Hao QL, Zhu J, Price MA, Payne KJ, Barsky LW, Crooks GM. 2001. Identiﬁcation of a novel, human multilymphoid progenitor in cord blood. Blood 97:3683–90 18. Res P, Martinez-Caceres E, Cristina Jaleco A, Staal F, Noteboom E, et al. 1996. CD34+ CD38dim cells in the human thymus can differentiate into T, natural killer, and dendritic cells but are distinct from pluripotent stem cells. Blood 87:5196–206

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

310

Blom

·

Spits

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

19. Dik WA, Pike-Overzet K, Weerkamp F, de Ridder D, de Haas EFE, et al. 2005. New insights on human T cell development by quantitative T cell receptor gene rearrangement studies and gene expression proﬁling. J. Exp. Med. 201:1715–23 20. Bhandoola A, Sambandam A, Allman D, Meraz A, Schwarz B. 2003. Early T lineage progenitors: new insights, but old questions remain. J. Immunol. 171:5653–58 21. Porritt HE, Rumfelt LL, Tabrizifard S, Schmitt TM, Zuniga-Pﬂucker JC, Petrie HT. 2004. Heterogeneity among DN1 prothymocytes reveals multiple progenitors with different capacities to generate T cell and non-T cell lineages. Immunity 20:735–45 22. Petrie HT, Kincade PW. 2005. Many roads, one destination for T cell progenitors. J. Exp. Med. 202:11–13 23. Blom B, Res P, Noteboom E, Weijer K, Spits H. 1997. Prethymic CD34+ progenitors capable of developing into T cells are not committed to the T cell lineage. J. Immunol. 158:3571–77 24. Ktorza S, Blanc C, Laurent C, Sarun S, Verpilleux MP, et al. 1996. Complete TCR-δ rearrangements and partial (D-J) recombination of the TCR-β locus in CD34+ 7+ precursors from human cord blood. J. Immunol. 156:4120–27 25. Klein F, Feldhahn N, Lee S, Wang H, Ciufﬁ F, et al. 2003. T lymphoid differentiation in human bone marrow. Proc. Natl. Acad. Sci. USA 100:6747–52 26. Weerkamp F, de Haas EF, Naber BA, Comans-Bitter WM, Bogers AJ, et al. 2005. Agerelated changes in the cellular composition of the thymus in children. J. Allergy Clin. Immunol. 115:834–40 26a. De Smedt M, Hoebeke I, Reynvoet K, Leclercq G, Plum J. 2005. Different thresholds of Notch signaling bias human precursor cells towards B, NK, monocytic/dendritic cell or T cell lineage in thymus microenvironment. Blood. In press 27. Spits H, Blom B, Jaleco AC, Weijer K, Verschuren MC, et al. 1998. Early stages in the development of human T, natural killer and thymic dendritic cells. Immunol. Rev. 165:75–86 28. Spits H. 2002. Development of αβ T cells in the human thymus. Nat. Rev. Immunol. 2:760–72 29. Galy A, Verma S, Barcena A, Spits H. 1993. Precursors of CD3+ CD4+ CD8+ cells in the human thymus are deﬁned by expression of CD34. Delineation of early events in human thymic development. J. Exp. Med. 178:391–401 30. Dalloul AH, Patry C, Salamero J, Canque B, Grassi F, Schmitt C. 1999. Functional and phenotypic analysis of thymic CD34+ CD1a− progenitor-derived dendritic cells: predominance of CD1a+ differentiation pathway. J. Immunol. 162:5821–28 31. Res P, Spits H. 1999. Developmental stages in the human thymus. Semin. Immunol. 11:39–46 32. Hori T, Cupp J, Wrighton N, Lee F, Spits H. 1991. Identiﬁcation of a novel human thymocyte subset with a phenotype of CD3− CD4+ CD8α + β-1. Possible progeny of the CD3− CD4− CD8− subset. J. Immunol. 146:4078–84 33. Blom B, Verschuren MC, Heemskerk MH, Bakker AQ, van Gastel-Mol EJ, et al. 1999. TCR gene rearrangements and expression of the pre-T cell receptor complex during human T-cell differentiation. Blood 93:3033–43 34. Verschuren MC, Blom B, Bogers AJ, Spits H, van Dongen JJ. 1998. PJA-BP expression and TCR δ deletion during human T cell differentiation. Int. Immunol. 10:1873–80 35. Carrasco YR, Trigueros C, Ramiro AR, de Yebenes VG, Toribio ML. 1999. β-selection is associated with the onset of CD8β chain expression on CD4+ CD8αα + pre-T cells during human intrathymic development. Blood 94:3491–98 www.annualreviews.org • Development of Human Lymphoid Cells

311

ARI

15 February 2006

21:28

36. Trigueros C, Ramiro AR, Carrasco YR, de Yebenes VG, Albar JP, Toribio ML. 1998. Identiﬁcation of a late stage of small noncycling pTα − pre-T cells as immediate precursors of T cell receptor α/β+ thymocytes. J. Exp. Med. 188:1401–12 37. Noguchi M, Yi H, Rosenblatt HM, Filipovich AH, Adelstein S, et al. 1993. Interleukin-2 receptor gamma chain mutation results in X-linked severe combined immunodeﬁciency in humans. Cell 73:147–57 38. Russell SM, Keegan AD, Harada N, Nakamura Y, Noguchi M, et al. 1993. Interleukin-2 receptor γ chain: A functional component of the interleukin-4 receptor. Science 262:1880– 83 39. Puel A, Ziegler SF, Buckley RH, Leonard WJ. 1998. Defective IL7R expression in T− B+ NK+ severe combined immunodeﬁciency. Nat. Genet. 20:394–97 40. Giliani S, Mori L, de Saint Basile G, Le Deist F, Rodriguez-Perez C, et al. 2005. Interleukin-7 receptor α (IL-7Rα) deﬁciency: cellular and molecular bases. Analysis of clinical, immunological, and molecular features in 16 novel patients. Immunol. Rev. 203:110–26 41. Macchi P, Villa A, Gillani S, Sacco MG, Frattini A, et al. 1995. Mutations of Jak-3 gene in patients with autosomal severe combined immunodeﬁciency. Nature 377:65– 68 42. Russell SM, Tayebi N, Nakajima H, Riedy MC, Roberts JL, et al. 1995. Mutation of Jak3 in a patient with SCID: essential role of Jak3 in lymphoid development. Science 270:797–800 43. Fischer A, Le Deist F, Hacein-Bey-Abina S, Andre-Schmutz I, Basile GD, et al. 2005. Severe combined immunodeﬁciency. A model disease for molecular immunology and therapy. Immunol. Rev. 203:98–109 44. Plum J, De Smedt M, Leclercq G, Verhasselt B, Vandekerckhove B. 1996. Interleukin-7 is a critical growth factor in early human T-cell development. Blood 88:4239–45 45. Pallard C, Stegmann AP, van Kleffens T, Smart F, Venkitaraman A, Spits H. 1999. Distinct roles of the phosphatidylinositol 3-kinase and STAT5 pathways in IL-7-mediated development of human thymocyte precursors. Immunity 10:525–35 46. Peschon JJ, Morrissey PJ, Grabstein KH, Ramsdell FJ, Maraskovsky E, et al. 1994. Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deﬁcient mice. J. Exp. Med. 180:1955–60 47. Sleasman JW, Harville TO, White GB, George JF, Barrett DJ, Goodenow MM. 1994. Arrested rearrangement of TCR V beta genes in thymocytes from children with X-linked severe combined immunodeﬁciency disease. J. Immunol. 153:442–48 48. Pesu M, Candotti F, Husa M, Hofmann SR, Notarangelo LD, O’Shea JJ. 2005. Jak3, severe combined immunodeﬁciency, and a new class of immunosuppressive drugs. Immunol. Rev. 203:127–42 49. Napolitano LA, Grant RM, Deeks SG, Schmidt D, De Rosa SC, et al. 2001. Increased production of IL-7 accompanies HIV-1-mediated T-cell depletion: implications for T-cell homeostasis. Nat. Med. 7:73–79 50. Artavanis-Tsakonas S, Rand MD, Lake RJ. 1999. Notch signaling: cell fate control and signal integration in development. Science 284:770–76 51. Radtke F, Wilson A, Stark G, Bauer M, van Meerwijk J, et al. 1999. Deﬁcient T cell fate speciﬁcation in mice with an induced inactivation of Notch1. Immunity 10:547–58 52. Jaleco AC, Neves H, Hooijberg E, Gameiro P, Clode N, et al. 2001. Differential effects of Notch ligands Delta-1 and Jagged-1 in human lymphoid differentiation. J. Exp. Med. 194:991–1002

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

312

Blom

·

Spits

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

53. La Motte-Mohs RN, Herer E, Zuniga-Pﬂucker JC. 2005. Induction of T-cell development from human cord blood hematopoietic stem cells by Delta-like 1 in vitro. Blood 105:1431–39 54. De Smedt M, Hoebeke I, Plum J. 2004. Human bone marrow CD34+ progenitor cells mature to T cells on OP9-DL1 stromal cell line without thymus microenvironment. Blood Cells Mol. Dis. 33:227–32 55. Schmitt TM, Zuniga-Pﬂucker JC. 2002. Induction of T cell development from hematopoietic progenitor cells by Delta-like-1 in vitro. Immunity 17:749–56 55a. Dontje W, Schotte R, Cupedo T, Nagasawa M, Scheeren F, et al. 2006. DeltaLike1 induced Notch1 signalling regulates the human plasmacytoid dendritic cell versus T cell lineage decision through control of GATA-3 and Spi-B. Blood. In press 56. Wolfer A, Wilson A, Nemir M, MacDonald HR, Radtke F. 2002. Inactivation of Notch2 impairs VDJβ rearrangement and allows pre-TCR-independent survival of early αβ lineage thymocytes. Immunity 16:869–79 57. Schmitt TM, Ciofani M, Petrie HT, Zuniga-Pﬂucker JC. 2004. Maintenance of T cell speciﬁcation and differentiation requires recurrent Notch receptor-ligand interactions. J. Exp. Med. 200:469–79 58. Radtke F, Ferrero I, Wilson A, Lees R, Aguet M, MacDonald HR. 2000. Notch1 deﬁciency dissociates the intrathymic development of dendritic cells and T cells. J. Exp. Med. 191:1085–94 59. De Smedt M, Reynvoet K, Kerre T, Taghon T, Verhasselt B, et al. 2002. Active form of Notch imposes T cell fate in human progenitor cells. J. Immunol. 169:3021–29 60. Garcia-Peydro M, de Yebenes VG, Toribio ML. 2003. Sustained Notch1 signaling instructs the earliest human intrathymic precursors to adopt a γδ T-cell fate in fetal thymus organ culture. Blood 102:2444–51 61. Washburn T, Schweighoffer E, Gridley T, Chang D, Fowlkes BJ, et al. 1997. Notch activity inﬂuences the αβ versus γδ T cell lineage decision. Cell 88:833–43 62. Engel I, Murre C. 2001. The function of E- and Id proteins in lymphocyte development. Nat. Rev. Immunol. 1:193–99 63. Heemskerk MH, Blom B, Nolan G, Stegmann AP, Bakker AQ, et al. 1997. Inhibition of T cell and promotion of natural killer cell development by the dominant negative helix loop helix factor Id3. J. Exp. Med. 186:1597–602 64. Blom B, Heemskerk MH, Verschuren MC, van Dongen JJ, Stegmann AP, et al. 1999. Disruption of αβ but not of γδ T cell development by overexpression of the helix-loophelix protein Id3 in committed T cell progenitors. EMBO J. 18:2793–802 65. Yokota Y, Mansouri A, Mori S, Sugawara S, Adachi S, et al. 1999. Development of peripheral lymphoid organs and natural killer cells depends on the helix-loop-helix inhibitor Id2. Nature 397:702–6 66. Herblot S, Steff AM, Hugo P, Aplan PD, Hoang T. 2000. SCL and LMO1 alter thymocyte differentiation: inhibition of E2A-HEB function and pre-Tα chain expression. Nat. Immunol. 1:138–44 67. Reizis B, Leder P. 2001. The upstream enhancer is necessary and sufﬁcient for the expression of the pre-T cell receptor α gene in immature T lymphocytes. J. Exp. Med. 194:979–90 68. Aifantis I, Azogui O, Feinberg J, Saint Ruf C, Buer J, von Boehmer H. 1998. On the role of the pre-T cell receptor in αβ versus γδ T lineage commitment. Immunity 9:649–55 69. Pai SY, Truitt ML, Ting CN, Leiden JM, Glimcher LH, Ho IC. 2003. Critical roles for transcription factor GATA-3 in thymocyte development. Immunity 19:863–75 www.annualreviews.org • Development of Human Lymphoid Cells

313

ARI

15 February 2006

21:28

70. Taghon T, De Smedt M, Stolz F, Cnockaert M, Plum J, Leclercq G. 2001. Enforced expression of GATA-3 severely reduces human thymic cellularity. J. Immunol. 167:4468– 75 71. Lanier LL. 2005. NK cell recognition. Annu. Rev. Immunol. 23:225–74 72. Nagler A, Lanier LL, Cwirla S, Phillips JH. 1989. Comparative studies of human FcRIIIpositive and negative natural killer cells. J. Immunol. 143:3183–91 73. Cooper MA, Fehniger TA, Turner SC, Chen KS, Ghaheri BA, et al. 2001. Human natural killer cells: a unique innate immunoregulatory role for the CD56bright subset. Blood 97:3146–51 74. Jaleco AC, Blom B, Res P, Weijer K, Lanier LL, et al. 1997. Fetal liver contains committed NK progenitors, but is not a site for development of CD34+ cells into T cells. J. Immunol. 159:694–702 75. S´anchez MJ, Spits H, Lanier LL, Philips JH. 1993. Human natural killer cell committed thymocytes and their relationship to the T cell lineage. J. Exp. Med. 178:1857–66 76. Gunther U, Holloway JA, Gordon JG, Knight A, Chance V, et al. 2005. Phenotypic characterization of CD3− 7+ cells in developing human intestine and an analysis of their ability to differentiate into T cells. J. Immunol. 174:5414–22 77. S´anchez MJ, Muench MO, Roncarolo MG, Lanier LL, Phillips JH. 1994. Identiﬁcation of a common T/natural killer cell progenitor in human fetal thymus. J. Exp. Med. 180:569– 76 78. Sirianni MC, Businco L, Seminara R, Aiuti F. 1983. Severe combined immunodeﬁciencies, primary T-cell defects and DiGeorge syndrome in humans: characterization by monoclonal antibodies and natural killer cell activity. Clin. Immunol. Immunopathol. 28:361–70 79. Rosmaraki EE, Douagi I, Roth C, Colucci F, Cumano A, Di Santo JP. 2001. Identiﬁcation of committed NK cell progenitors in adult murine bone marrow. Eur. J. Immunol. 31:1900–9 80. Freud AG, Becknell B, Roychowdhury S, Mao HC, Ferketich AK, et al. 2005. A human CD34+ subset resides in lymph nodes and differentiates into CD56bright natural killer cells. Immunity 22:295–304 81. Fehniger TA, Cooper MA, Nuovo GJ, Cella M, Facchetti F, et al. 2003. CD56bright natural killer cells are present in human lymph nodes and are activated by T cell-derived IL-2: a potential new link between adaptive and innate immunity. Blood 101:3052–57 82. Ferlazzo G, Thomas D, Lin SL, Goodman K, Morandi B, et al. 2004. The abundant NK cells in human secondary lymphoid tissues require activation to express killer cell Ig-like receptors and become cytolytic. J. Immunol. 172:1455–62 83. Parrish-Novak J, Dillon SR, Nelson A, Hammond A, Sprecher C, et al. 2000. Interleukin 21 and its receptor are involved in NK cell expansion and regulation of lymphocyte function. Nature 408:57–63 84. Sivori S, Cantoni C, Parolini S, Marcenaro E, Conte R, et al. 2003. IL-21 induces both rapid maturation of human CD34+ cell precursors towards NK cells and acquisition of surface killer Ig-like receptors. Eur. J. Immunol. 33:3439–47 85. Rodewald HR, Moingeon P, Lucich JL, Dosiou C, Lopez P, Reinherz EL. 1992. A population of early fetal thymocytes expressing FcγRII/III contains precursors of T lymphocytes and natural killer cells. Cell 69:139–50 86. Phillips JH, Hori T, Nagler A, Bhat N, Spits H, Lanier LL. 1992. Ontogeny of human natural killer (NK) cells: Fetal NK cells mediate cytolytic function and express cytoplasmic CD3δ proteins. J. Exp. Med. 175:1055–66

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

314

Blom

·

Spits

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

87. S´anchez MJ, Muench MO, Roncarolo MG, Lanier LL, Phillips JH. 1994. Identiﬁcation of a common T/NK cell progenitor in human fetal thymus. J. Exp. Med. 180:569– 76 88. Carlyle JR, Michie AM, Furlonger C, Nakano T, Lenardo MJ, et al. Identiﬁcation of a novel developmental stage marking lineage commitment of progenitor thymocytes. J. Exp. Med. 186:173–82 89. Ikawa T, Kawamoto H, Fujimoto S, Katsura Y. 1999. Commitment of common T/natural killer (NK) progenitors to unipotent T and NK progenitors in the murine fetal thymus revealed by a single progenitor assay. J. Exp. Med. 190:1617–26 90. Spits H, Lanier LL, Phillips JH. 1995. Development of human T and natural killer cells. Blood 85:2654–70 91. Marquez C, Trigueros C, Franco JM, Ramiro AR, Carrasco YR, et al. 1998. Identiﬁcation of a common developmental pathway for thymic natural killer cells and dendritic cells. Blood 91:2760–71 92. Perez SA, Sotiropoulou PA, Gkika DG, Mahaira LG, Niarchos DK, et al. 2003. A novel myeloid-like NK cell progenitor in human umbilical cord blood. Blood 101:3444– 50 93. Akashi K, Traver D, Miyamoto T, Weissman IL. 2000. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature 404:193–97 94. Colucci F, Caligiuri MA, Di Santo JP. 2003. What does it take to make a natural killer? Nat. Rev. Immunol. 3:413–25 95. Vosshenrich CA, Samson-Villeger SI, Di Santo JP. 2005. Distinguishing features of developing natural killer cells. Curr. Opin. Immunol. 17:151–58 96. Bennett IM, Zatsepina O, Zamai L, Azzoni L, Mikheeva T, Perussia B. 1996. Deﬁnition of a natural killer NKR− P1A+ /CD56− /CD16− functionally immature human NK cell subset that differentiates in vitro in the presence of interleukin 12. J. Exp. Med. 184:1845– 56 97. Miller JS, McCullar V. 2001. Human natural killer cells with polyclonal lectin and immunoglobulinlike receptors develop from single hematopoietic stem cells with preferential expression of NKG2A and KIR2DL2/L3/S2. Blood 98:705–13 98. Sivori S, Falco M, Marcenaro E, Parolini S, Biassoni R, et al. 2002. Early expression of triggering receptors and regulatory role of 2B4 in human natural killer cell precursors undergoing in vitro differentiation. Proc. Natl. Acad. Sci. USA 99:4526–31 99. Zamai L, Ahmad M, Bennett IM, Azzoni L, Alnemri ES, Perussia B. 1998. Natural killer (NK) cell-mediated cytotoxicity: differential use of TRAIL and Fas ligand by immature and mature primary human NK cells. J. Exp. Med. 188:2375–80 100. Di Santo JP, Keever CA, Small T, Nichols GL, O’Reilly RJ, Flomenberg N. 1990. Absence of interleukin 2 production in a severe combined immunodeﬁciency disease syndrome with T cells. J. Exp. Med. 171:1697–704 101. Roifman CM. 2005. Studies of patients’ thymi aid in the discovery and characterization of immunodeﬁciency in humans. Immunol. Rev. 203:143–55 102. Gilmour KC, Fujii H, Cranston T, Davies EG, Kinnon C, Gaspar HB. 2001. Defective expression of the interleukin-2/interleukin-15 receptor β subunit leads to a natural killer cell-deﬁcient form of severe combined immunodeﬁciency. Blood 98:877–79 103. Vosshenrich CAJ, Ranson T, Samson SI, Corcuff E, Colucci F, et al. 2005. Roles for common cytokine receptor γ-chain-dependent cytokines in the generation, differentiation, and maturation of NK cell precursors and peripheral NK cells in vivo. J. Immunol. 174:1213–21 www.annualreviews.org • Development of Human Lymphoid Cells

315

ARI

15 February 2006

21:28

104. Kasaian MT, Whitters MJ, Carter LL, Lowe LD, Jussif JM, et al. 2002. IL-21 limits NK cell responses and promotes antigen-speciﬁc T cell activation: a mediator of the transition from innate to adaptive immunity. Immunity 16:559–69 105. Ikawa T, Fujimoto S, Kawamoto H, Katsura Y, Yokota Y. 2001. Commitment to natural killer cells requires the helix-loop-helix inhibitor Id2. Proc. Natl. Acad. Sci. USA 98:5164– 69 106. Schotte R, Rissoan MC, Bendriss-Vermare N, Bridon JM, Duhen T, et al. 2003. The transcription factor Spi-B is expressed in plasmacytoid DC precursors and inhibits T-, B-, and NK-cell development. Blood 101:1015–23 107. van den Brandt J, Voss K, Schott M, Hunig T, Wolfe MS, Reichardt HM. 2004. Inhibition of Notch signaling biases rat thymocyte development towards the NK cell lineage. Eur. J. Immunol. 34:1405–13 108. Bories JC, Willerford DM, Grevin D, Davidson L, Camus A, et al. 1995. Increased T-cell apoptosis and terminal B-cell differentiation induced by inactivation of the Ets-1 proto-oncogene. Nature 377:635–38 109. Barton K, Muthusamy N, Fischer C, Ting CN, Walunas TL, et al. 1998. The Ets1 transcription factor is required for the development of natural killer cells in mice. Immunity 9:555–63 110. Banchereau J, Briere F, Caux C, Davoust J, Lebecque S, et al. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18:767–811 111. Liu Y-J. 2005. IPC: professional type 1 interferon-producing cells and plasmacytoid dendritic cell precursors. Annu. Rev. Immunol. 23:275–306 112. MacDonald KPA, Munster DJ, Clark GJ, Dzionek A, Schmitz J, Hart DNJ. 2002. Characterization of human blood dendritic cell subsets. Blood 100:4512–20 113. Galy A, Christopherson I, Ferlazzo G, Liu G, Spits H, Georgopoulos K. 2000. Distinct signals control the hematopoiesis of lymphoid-related dendritic cells. Blood 95:128–37 114. Canque B, Camus S, Dalloul A, Kahn E, Yagello M, et al. 2000. Characterization of dendritic cell differentiation pathways from cord blood CD34+ CD7+ CD45RA+ hematopoietic progenitor cells. Blood 96:3748–56 115. Blom B, Ho S, Antonenko S, Liu YJ. 2000. Generation of interferon α-producing predendritic cell (pre-DC)2 from human CD34+ hematopoietic stem cells. J. Exp. Med. 192:1785–96 116. Bendriss-Vermare N, Barthelemy C, Durand I, Bruand C, Dezutter-Dambuyant C, et al. 2001. Human thymus contains IFN-α-producing CD11c− , myeloid CD11c+ , and mature interdigitating dendritic cells. J. Clin. Invest. 107:835–44 117. Res PC, Couwenberg F, Vyth-Dreese FA, Spits H. 1999. Expression of pTα mRNA in a committed dendritic cell precursor in the human thymus. Blood 94:2647–57 118. Pelayo R, Hirose J, Huang J, Garrett KP, Delogu A, et al. 2005. Derivation of 2 categories of plasmacytoid dendritic cells in murine bone marrow. Blood 105:4407–15 119. Rissoan MC, Duhen T, Bridon JM, Bendriss-Vermare N, Peronne C, et al. 2002. Subtractive hybridization reveals the expression of immunoglobulin-like transcript 7, EphB1, granzyme B, and 3 novel transcripts in human plasmacytoid dendritic cells. Blood 100:3295–303 120. Corcoran L, Ferrero I, Vremec D, Lucas K, Waithman J, et al. 2003. The lymphoid past of mouse plasmacytoid cells and thymic dendritic cells. J. Immunol. 170:4926–32 121. Grouard G, Rissoan MC, Filgueira L, Durand I, Banchereau J, Liu YJ. 1997. The enigmatic plasmacytoid T cells develop into dendritic cells with interleukin (IL)-3 and CD40ligand. J. Exp. Med. 185:1101–11

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

316

Blom

·

Spits

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

122. Shigematsu H, Reizis B, Iwasaki H, Mizuno S, Hu D, et al. 2004. Plasmacytoid dendritic cells activate lymphoid-speciﬁc genetic programs irrespective of their cellular origin. Immunity 21:43–53 123. Traver D, Akashi K, Manz M, Merad M, Miyamoto T, et al. 2000. Development of CD8α-positive dendritic cells from a common myeloid progenitor. Science 290:2152– 54 124. Martin CH, Aifantis I, Scimone ML, von Andrian UH, Reizis B, et al. 2003. Efﬁcient thymic immigration of B220+ lymphoid-restricted bone marrow cells with T precursor potential. Nat. Immunol. 4:866–73 125. Blom B, Ligthart SJ, Schotte R, Spits H. 2002. Developmental origin of pre-DC2. Hum. Immunol. 63:1072–80 126. Spits H, Couwenberg F, Bakker AQ, Weijer K, Uittenbogaart CH. 2000. Id2 and Id3 inhibit development of CD34+ stem cells into pre-dendritic cell (pre-DC)2 but not into pre-DC1. Evidence for a lymphoid origin of pre-DC2. J. Exp. Med. 192:1775–84 127. Comeau MR, de Vries ARV, Maliszewski CR, Galibert L. 2002. CD123bright plasmacytoid predendritic cells: progenitors undergoing cell fate conversion? J. Immunol. 169:75– 83 128. Gilliet M, Boonstra A, Paturel C, Antonenko S, Xu XL, et al. 2002. The development of murine plasmacytoid dendritic cell precursors is differentially regulated by FLT3-ligand and granulocyte/macrophage colony-stimulating factor. J. Exp. Med. 195:953–58 129. Naik SH, Proietto AI, Wilson NS, Dakic A, Schnorrer P, et al. 2005. Cutting edge: generation of splenic CD8+ and CD8− dendritic cell equivalents in Fms-like tyrosine kinase 3 ligand bone marrow cultures. J. Immunol. 174:6592–97 130. Bjorck P. 2001. Isolation and characterization of plasmacytoid dendritic cells from Flt3 ligand and granulocyte-macrophage colony-stimulating factor-treated mice. Blood 98:3520–26 131. Coates PT, Barratt-Boyes SM, Zhang L, Donnenberg VS, O’Connell PJ, et al. 2003. Dendritic cell subsets in blood and lymphoid tissue of rhesus monkeys and their mobilization with Flt3 ligand. Blood 102:2513–21 132. Manfra DJ, Chen SC, Jensen KK, Fine JS, Wiekowski MT, Lira SA. 2003. Conditional expression of murine Flt3 ligand leads to expansion of multiple dendritic cell subsets in peripheral blood and tissues of transgenic mice. J. Immunol. 170:2843–52 133. Maraskovsky E, Daro E, Roux E, Teepe M, Maliszewski CR, et al. 2000. In vivo generation of human dendritic cell subsets by Flt3 ligand. Blood 96:878–84 134. Pulendran B, Banchereau J, Burkeholder S, Kraus E, Guinet E, et al. 2000. Flt3-ligand and granulocyte colony-stimulating factor mobilize distinct human dendritic cell subsets in vivo. J. Immunol. 165:566–72 135. D’Amico A, Wu L. 2003. The early progenitors of mouse dendritic cells and plasmacytoid predendritic cells are within the bone marrow hemopoietic precursors expressing Flt3. J. Exp. Med. 198:293–303 136. Karsunky H, Merad M, Cozzio A, Weissman IL, Manz MG. 2003. Flt3 ligand regulates dendritic cell development from Flt3+ lymphoid and myeloid-committed progenitors to Flt3+ dendritic cells in vivo. J. Exp. Med. 198:305–13 137. McKenna HJ, Stocking KL, Miller RE, Brasel K, De Smedt T, et al. 2000. Mice lacking ﬂt3 ligand have deﬁcient hematopoiesis affecting hematopoietic progenitor cells, dendritic cells, and natural killer cells. Blood 95:3489–97 138. Laouar Y, Welte T, Fu XY, Flavell RA. 2003. STAT3 is required for Flt3L-dependent dendritic cell differentiation. Immunity 19:903–12 www.annualreviews.org • Development of Human Lymphoid Cells

317

ARI

15 February 2006

21:28

139. Rathinam C, Geffers R, Yucel R, Buer J, Welte K, et al. 2005. The transcriptional repressor Gﬁ1 controls STAT3-dependent dendritic cell development and function. Immunity 22:717–28 140. Chen W, Antonenko S, Sederstrom JM, Liang X, Chan AS, et al. 2004. Thrombopoietin cooperates with FLT3-ligand in the generation of plasmacytoid dendritic cell precursors from human hematopoietic progenitors. Blood 103:2547–53 141. Wu L, Nichogiannopoulou A, Shortman K, Georgopoulos K. 1997. Cell-autonomous defects in dendritic cell populations of Ikaros mutant mice point to a developmental relationship with the lymphoid lineage. Immunity 7:483–92 142. Wu L, D’Amico A, Winkel KD, Suter M, Lo D, Shortman K. 1998. RelB is essential for the development of myeloid-related CD8α − dendritic cells but not of lymphoid-related CD8α + dendritic cells. Immunity 9:839–47 143. Hacker C, Kirsch RD, Ju XS, Hieronymus T, Gust TC, et al. 2003. Transcriptional proﬁling identiﬁes Id2 function in dendritic cell development. Nat. Immunol. 4:380–86 144. Schotte R, Rissoan MC, Bendriss-Vermare N, Bridon JM, Duhen T, et al. 2002. The transcription factor Spi-B is expressed in plasmacytoid DC precursors and inhibits T, B, and NK cell development. Blood 101:1015–23 145. Schotte R, Nagasawa M, Weijer K, Spits H, Blom B. 2004. The ETS transcription factor Spi-B is required for human plasmacytoid dendritic cell development. J. Exp. Med. 200:1503–9 146. Iwasaki H, Somoza C, Shigematsu H, Duprez EA, Iwasaki-Arai J, et al. 2005. Distinctive and indispensable roles of PU.1 in maintenance of hematopoietic stem cells and their differentiation. Blood 106:1590–600 147. Tamura T, Tailor P, Yamaoka K, Kong HJ, Tsujimura H, et al. 2005. IFN regulatory factor-4 and -8 govern dendritic cell subset development and their functional diversity. J. Immunol. 174:2573–81 148. Tsujimura H, Tamura T, Ozato K. 2003. Cutting edge: IFN consensus sequence binding protein/IFN regulatory factor 8 drives the development of type I IFN-producing plasmacytoid dendritic cells. J. Immunol. 170:1131–35 149. Brass AL, Kehrli E, Eisenbeis CF, Storb U, Singh H. 1996. Pip, a lymphoid-restricted IRF, contains a regulatory domain that is important for autoinhibition and ternary complex formation with the Ets factor PU.1. Genes Dev. 10:2335–47 150. Nakano N, Nishiyama C, Masuoka N, Nishiyama M, Yamane H, et al. 2005. Analysis of PU.1/ICSBP (IRF-8) complex formation with various PU.1 mutants: molecular cloning of rat Icsbp (Irf-8) cDNA. Immunogenetics 56:871–77 151. van der Stoep N, Quinten E, Rezende MM, van den Elsen PJ. 2004. E47, IRF-4, and PU.1 synergize to induce B-cell-speciﬁc activation of the class II transactivator promoter III (CIITA-PIII). Blood 104:2849–57 152. Nunez C, Nishimoto N, Gartland GL, Billips LG, Burrows PD, et al. 1996. B cells are generated throughout life in humans. J. Immunol. 156:866–72 153. Solvason N, Kearney JF. 1992. The human fetal omentum: a site of B cell generation. J. Exp. Med. 175:397–404 154. Reynaud D, Lefort N, Manie E, Coulombel L, Levy Y. 2003. In vitro identiﬁcation of human pro-B cells that give rise to macrophages, natural killer cells, and T cells. Blood 101:4313–21 155. Sanz E, Alvarez-Mon M, Martinez AC, de la Hera A. 2003. Human cord blood CD34+ Pax-5+ B-cell progenitors: single-cell analyses of their gene expression proﬁles. Blood 101:3424–30

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

318

Blom

·

Spits

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

ARI

15 February 2006

21:28

156. Hardy RR, Carmack CE, Shinton SA, Kemp JD, Hayakawa K. 1991. Resolution and characterization of pro-B and pre-pro-B cell stages in normal mouse bone marrow. J. Exp. Med. 173:1213–25 157. Bertrand FE 3rd, Billips LG, Burrows PD, Gartland GL, Kubagawa H, et al. 1997. Ig DH gene segment transcription and rearrangement before surface expression of the pan-B-cell marker CD19 in normal human bone marrow. Blood 90:736–44 158. Dworzak MN, Fritsch G, Froschl G, Printz D, Gadner H. 1998. Four-color ﬂow cytometric investigation of terminal deoxynucleotidyl transferase-positive lymphoid precursors in pediatric bone marrow: CD79a expression precedes CD19 in early B-cell ontogeny. Blood 92:3203–9 159. Wang YH, Nomura J, Faye-Petersen OM, Cooper MD. 1998. Surrogate light chain production during B cell differentiation: differential intracellular versus cell surface expression. J. Immunol. 161:1132–39 160. Busslinger M. 2004. Transcriptional control of early B cell development. Annu. Rev. Immunol. 22:55–79 161. Ghia P, ten Boekel E, Sanz E, de la Hera A, Rolink A, Melchers F. 1996. Ordering of human bone marrow B lymphocyte precursors by single-cell polymerase chain reaction analyses of the rearrangement status of the immunoglobulin H and L chain gene loci. J. Exp. Med. 184:2217–29 162. LeBien TW. 2000. Fates of human B-cell precursors. Blood 96:9–23 163. Loken MR, Shah VO, Dattilio KL, Civin CI. 1987. Flow cytometric analysis of human bone marrow. II. Normal B lymphocyte development. Blood 70:1316–24 164. LeBien TW, Wormann B, Villablanca JG, Law CL, Steinberg LM, et al. 1990. Multiparameter ﬂow cytometric analysis of human fetal bone marrow B cells. Leukemia 4:354–58 165. Davi F, Faili A, Gritti C, Blanc C, Laurent C, Sutton L, et al. 1997. Early onset of immunoglobulin heavy chain gene rearrangements in normal human bone marrow CD34+ cells. Blood 90:4014–21 166. Lemmers B, Gauthier L, Guelpa-Fonlupt V, Fougereau M, Schiff C. 1999. The human (L+ μ− ) proB complex: cell surface expression and biochemical structure of a putative transducing receptor. Blood 93:4336–46 167. Dittel BN, LeBien TW. 1995. The growth response to IL-7 during normal human B cell ontogeny is restricted to B-lineage cells expressing CD34. J. Immunol. 154:58–67 168. Hou YH, Srour EF, Ramsey H, Dahl R, Broxmeyer HE, Hromas R. 2005. Identiﬁcation of a human B-cell/myeloid common progenitor by the absence of CXCR4. Blood 105:3488–92 169. Nutt SL, Morrison AM, Dorﬂer P, Rolink A, Busslinger M. 1998. Identiﬁcation of BSAP (Pax-5) target genes in early B-cell development by loss- and gain-of-function experiments. EMBO J. 17:2319–33 170. Nutt SL, Urbanek P, Rolink A, Busslinger M. 1997. Essential functions of Pax5 (BSAP) in pro-B cell development: difference between fetal and adult B lymphopoiesis and reduced V-to-DJ recombination at the IgH locus. Genes Dev. 11:476–91 171. Milne CD, Fleming HE, Zhang Y, Paige CJ. 2004. Mechanisms of selection mediated by interleukin-7, the preBCR, and hemokinin-1 during B-cell development. Immunol. Rev. 197:75–88 172. Namikawa R, Muench MO, de Vries JE, Roncarolo MG. 1996. The FLK2/FLT3 ligand synergizes with interleukin-7 in promoting stromal-cell-independent expansion and differentiation of human fetal pro-B cells in vitro. Blood 87:1881–90

www.annualreviews.org • Development of Human Lymphoid Cells

319

ARI

15 February 2006

21:28

173. Prieyl JA, LeBien TW. 1996. Interleukin 7 independent development of human B cells. Proc. Natl. Acad. Sci. USA 93:10348–53 174. Macchi P, Villa A, Giliani S, Sacco MG, Frattini A, et al. 1995. Mutations of Jak-3 gene in patients with autosomal severe combined immune deﬁciency (SCID). Nature 377:65–68 175. Kovanen PE, Leonard WJ. 2004. Cytokines and immunodeﬁciency diseases: critical roles of the γc -dependent cytokines interleukins 2, 4, 7, 9, 15, and 21, and their signaling pathways. Immunol. Rev. 202:67–83 176. Nutt SL, Metcalf D, D’Amico A, Polli M, Wu L. 2005. Dynamic regulation of PU.1 expression in multipotent hematopoietic progenitors. J. Exp. Med. 201:221–31 177. Berardi AC, Meffre E, Pﬂumio F, Katz A, Vainchenker W, et al. 1997. Individual CD34+ CD38low CD19− CD10− progenitor cells from human cord blood generate B lymphocytes and granulocytes. Blood 89:3554–64 178. Bond HM, Mesuraca M, Carbone E, Bonelli P, Agosti V, et al. 2004. Early hematopoietic zinc ﬁnger protein (EHZF), the human homolog to mouse Evi3, is highly expressed in primitive human hematopoietic cells. Blood 103:2062–70 179. Jaleco AC, Stegmann AP, Heemskerk MH, Couwenberg F, Bakker AQ, et al. 1999. Genetic modiﬁcation of human B-cell development: B-cell development is inhibited by the dominant negative helix loop helix factor Id3. Blood 94:2637–46 180. Traggiai E, Chicha L, Mazzucchelli L, Bronz L, Piffaretti JC, et al. 2004. Development of a human adaptive immune system in cord blood cell-transplanted mice. Science 304:104–7 181. Gimeno R, Weijer K, Voordouw A, Uittenbogaart CH, Legrand N, et al. 2004. Monitoring the effect of gene silencing by RNA interference in human CD34+ cells injected into newborn RAG2−/− γc−/− mice: functional inactivation of p53 in developing T cells. Blood 104:3886–93 182. Ishikawa F, Yasukawa M, Lyons B, Yoshida S, Miyamoto T, et al. 2005. Development of functional human blood and immune systems in NOD/SCID/IL2 receptor γ chainnull mice. Blood 106:1565–73

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-10

320

Blom

·

Spits

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:287-320. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

Genetic Disorders of Programmed Cell Death in the Immune System∗ Nicolas Bid`ere,† Helen C. Su,† and Michael J. Lenardo Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892; email: [email protected]

Annu. Rev. Immunol. 2006. 24:321–52

Key Words

First published online as a Review in Advance on December 12, 2005

cytokine, Fas/APO-1/CD95, caspase, tumor necrosis factor receptor, autoimmune lymphoproliferative syndrome (ALPS), caspase-eight deﬁciency state (CEDS)

The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090513 c 2006 by Copyright Annual Reviews. All rights reserved ∗ The U.S. Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper. †

These authors contributed equally to this work.

Abstract Human genetics offers new possibilities for understanding physiological regulatory mechanisms and disorders of the immune system. Genetic abnormalities of lymphocyte cell death programs have provided insights into mechanisms of receptor biology and principles of immune homeostasis and tolerance. Thus far, there are two major diseases of programmed cell death associated with inherited human mutations: the autoimmune lymphoproliferative syndrome and the caspase-eight deﬁciency state. We describe the details of their molecular pathogenesis and discuss how these diseases illustrate important concepts in immune regulation and genetics.

0732-0582/06/0423-0321$20.00

321

ANRV270-IY24-11

ARI

23 February 2006

15:22

INTRODUCTION

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ALPS: autoimmune lymphoproliferative syndrome Apoptosis: programmed cell death characterized by cell shrinkage, nuclear condensation with DNA fragmentation, membrane blebbing, and breakdown into rapidly phagocytosed apoptotic bodies Caspases: a family of highly conserved intracellular cysteine proteases involved in apoptosis, development, cytokine maturation, and normal cell function

322

Human genetics has entered a new era. Technological advances allow investigators to probe inherited diseases with unprecedented molecular precision and thus shed light on normal physiological pathways. The complete DNA sequence of the human genome, new techniques to introduce DNA into primary and transformed cells, and RNA interference for selectively inhibiting gene expression now allow experimentalists to examine human cells as powerfully as cells from model organisms. The major advantage of human research has always been the widespread detection and documentation of unusual genetic conditions. The marriage of new technology and disease investigation brings humans to the proscenium and may make experimental animal systems supporting actors in future genetic research. How the immune system maintains homeostasis and tolerance remains a question of fundamental importance. On the low side, lymphocyte depletion and immunodeﬁciency can result. On the high side, a failure of regulatory mechanisms can cause an overabundance of lymphocytes and possibly cancer, as well as autoimmune and allergic reactions. The autoimmune lymphoproliferative syndrome (ALPS) is one of the ﬁrst human inherited disorders of apoptosis to be described (1). Investigations of this disease led to the ﬁrst descriptions of inherited mutations in genes encoding caspase proteins in humans (2, 3). It is clear that deﬁciencies in the death receptor Fas/APO-1/CD95, Fas ligand (FasL or CD95L), caspase-8, and caspase-10 primarily affect the immune system. In fact, it has been difﬁcult to ﬁnd abnormalities in any other organ system associated with mutations in these genes in humans. In this context, these mutations reveal the central role of apoptosis in immune homeostasis. Recent work also may point to a key role for these molecules and an associated form of nonapoptotic death in immune regulation (N. Bid`ere & M. Lenardo, unpublished observations). One key obser-

Bid`ere

· · Su

Lenardo

vation is that dominant-negative mutations in genes that encode death factors predominate in ALPS. This type of mutation predominates because the transduction of a death signal involves speciﬁc stoichiometric protein complexes that are susceptible to interference by the inclusion of mutant signaling components. Given the widespread involvement of noncovalent protein complexes in cellular processes, dominant interference is likely to be a common mode for genetic abnormalities in humans, who are typically not consanguineous. Thus, disease mechanisms due to heterozygous mutations may be more prominent than those attributable to homozygous mutations. Finally, another conceptual advance stemming from these diseases is that an overabundance of lymphocytes does not necessarily equate with increased function. Caspase-8 deﬁciency causes, for example, lymphocytosis due to impaired apoptosis together with immunodeﬁciency due to defective antigen receptor signaling (2, 4).

IMMUNE REGULATION BY PROGRAMMED CELL DEATH A simpliﬁed view of lymphocyte regulation by apoptosis is useful for thinking about the molecular mechanisms involved (Figure 1). The basic principles of intrinsic (lymphokine withdrawal) and extrinsic (death receptor) death pathways have been reviewed previously (5, 6). In essence, high antigen load stimulates the extrinsic death pathway as a means to restrain the immune response and prevent immunotoxicity. On the other hand, the clearance of antigen leads to decreased trophic cytokine production and cell death by the intrinsic pathway. This event rids the body of excess, unnecessary lymphocytes and restores homeostasis. Genetic analysis in mice shows that an imbalance of either pathway can lead to a loss of lymphocyte control and tolerance (7). Antigen-speciﬁc cells that escape both pathways and reenter a resting state contribute to the pool of memory cells. The memory pool can be later called upon for anamnestic

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

Figure 1 The propriocidal death pathway in lymphocytes. Antigen stimulation of resting T lymphocytes leads to their activation and proliferation. Activated T cells express Fas and its cognate ligand, FasL, which regulates Fas-expressing B cells and APC. Similar effects might be mediated by other death receptors. When T cells reencounter high antigen levels, the extrinsic pathway of cell death (active apoptosis) is triggered. After cessation of antigen and cytokine production, passive apoptosis ensues through the intrinsic pathway. The remaining T cells constitute the memory population.

responses. The key concept is that mature T and B lymphocytes can be programmed to die in an antigen-speciﬁc fashion. This antigencontrolled cell death affords dynamic control of both lymphocyte cell number and immunological tolerance in the peripheral immune system. Because variation in antigen levels stimulates T cells to kill either themselves or B cells, this constitutes a self-regulatory or propriocidal pathway of death. Recently, it has been proposed that an individual lineage of regulatory T cells can carry out suppressive functions in immune responses (8). We and others have observed that the suppressive effects of such cells is due to the induction of target cell death (9; P. Pandiyan & M. Lenardo, unpublished results). However, little has been done to examine possible immune protective functions of death induced by this variety of T cells. Therefore, it is difﬁcult to know whether the suppressive effects contribute to self-tolerance by deletion or through bystander effects of an immunological response involving cell killing. Although it was initially believed that the extrinsic pathway caused apoptosis mediated by death receptors in the tumor necrosis factor

(TNF) receptor superfamily, subsequent data has revealed a more complex situation. Antigen stimulation of T cells causes the induction of FasL and TNF, which can trigger the corresponding death domain (DD)-containing receptors (10–13). Moreover, the triggering of FasL on T cells can induce the death of B cells that have been activated and express Fas (Figure 1). The requirement for antigen stimulation to provoke the pathway—though often confused with a means to remove lymphocytes at the end of an immune response— indicates its importance for negative feedback regulation during a proliferative lymphocyte response to antigen. One complexity is that the blockade of death receptors does not completely eliminate the ability of antigen receptor stimulation to kill T cells (14). This suggests that alternative death mechanisms exist that have not yet been clearly identiﬁed. A second complexity is that the blockade of Fas-induced apoptosis does not prevent death. In fact, the addition of the pan-caspase inhibitor zVAD to Fas stimulation leads to increased alternative cell death in activated human peripheral blood T cells (N. Bid`ere & M. Lenardo, unpublished observations). There www.annualreviews.org • Genetic Disorders and Cell Death

323

ANRV270-IY24-11

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TNFR: tumor necrosis factor receptor Necrosis: nonapoptotic cell death characterized by cell swelling, early membrane damage, nuclear lysis without condensation, and a strong inﬂammatory response DD: death domain FADD: Fas-associated death domain DISC: death-inducing signaling complex DED: death effector domain

324

15:22

have been several observations that triggering either Fas or TNF receptor 1 (TNFR1) can lead to programs of necrosis (15, 16). Presumably, this provides an important alternative in cases in which viruses or other infectious agents can impede apoptosis. However, a more thorough understanding of these complexities will be an important focus of further work in this area.

GENETIC INSIGHTS INTO PROGRAMMED DEATH BY FAS Physiological elimination of activated T cells during an immune response can occur via an active extrinsic pathway triggered by the engagement of specialized transmembrane receptors called death receptors. These death receptors belong to the TNFR superfamily and include Fas (also called CD95, TNFRSF6, or APO-1), TNFR1 (also known as p60 or TNFRSF1A), death receptor 3 (also known as DR3, or TNFRSF25), death receptor 4 [also known as DR4, TNFrelated apoptosis-inducing ligand receptor 1 (TRAIL-R1), or TNFRSF10A], death receptor 5 (DR5, or TRAIL-R2 or TNFRSF10B), and death receptor 6 (DR6 or TNFRSF21). The TNFR superfamily is characterized by arrays of two to ﬁve extracellular cysteine-rich domains (CRDs) (17). Members of the death receptor subgroup of TNFRs share a moderately conserved region of 80 amino acids in the cytoplasmic portion termed the death domain (DD), as it is required for death signaling. Fas, the prototypical death receptor, plays a critical role in lymphocyte regulation by apoptosis and, potentially, nonapoptotic death. Fas is activated by its cognate ligand, FasL, a type 2 transmembrane protein that can be released in soluble form by metalloproteinases (18). Both the ligand and the receptor can be potently upregulated by antigenic stimulation of T cells, especially if the cells are activated and cultured in interleukin (IL)-2 (19). Stimulation of B cells can induce death receptors and TNFα, but not expression of FasL. Rather, B cells appear to be killed Bid`ere

· · Su

Lenardo

by conjugation to T cells that express FasL (20) (Figure 1). The fundamental mode of signaling by death receptors is protein complex formation at two levels. First, the receptor assembly is triggered by ligand to form stoichiometric complexes with cytoplasmic signaling proteins. Second, higher order aggregates of receptors, ligands, and signaling proteins immediately follow and are crucial for signal transmission. As an example, we review Fas signaling. Within seconds after FasL stimulation, Fas recruits the adapter molecule FADD and the initiators caspases-8 and -10 to form a large complex termed death-inducing signaling complex (DISC) (21). DISC formation is crucial for generating caspase activity to initiate the apoptotic process. Both Fas and FasL operate as homotrimers. Their interaction triggers changes in conformation and/or orientation of the receptor trimer that allow homotypic interactions between the DDs in Fas and FADD. In addition to its COOHterminal DD, FADD also possesses a NH2 terminal protein-protein-interaction domain of approximately 80 residues called the death effector domain (DED), which is also found in the initiator (or apical) caspases-8 and -10. Remarkably, the DD and DED are both hexahelical bundles with a similar overall topology that is also shared by the caspase-recruitment domain (CARD) (22). Hence, these domains may be specializations of a common ancestor. It is generally the case, however, that each domain interacts in a homotypic manner, i.e., DD to DD, but do not cross bind one another (DD to DED, etc.). Therefore, as illustrated in Figure 2, FADD is brought to the DISC by its DD association with Fas and then recruits caspases-8 and -10 to the complex through DED interactions (23, 24). Within this multiprotein DISC aggregate, the increased local concentration of caspase8/10 zymogens presumably promotes an altered conformation that leads to their autoprocessing and activation. Caspases-8 and -10 each contain three parts: a double-DEDcontaining prodomain, a large enzymatic

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

Figure 2 Caspase-8 is recruited to the plasma membrane and integrates two different signaling complexes, with two different outcomes. As described in the text, Fas triggers DISC formation with the release of highly active, processed caspase-8 into the cytosol that leads to the demise of the cells. By contrast, antigen receptor stimulation triggers activation-receptor induced signalosome (ARIS) formation. In that case, active caspase-8 remains unprocessed and allows the recruitment of IκB kinase (IKK) complex to promote NF-κB activation.

subunit, and a small enzymatic subunit. Each part is ﬂanked by aspartate residues that form caspase cleavage sites especially sensitive to its own activity. Initially, it was believed that autoprocessing and release of the two enzymatic subunits were essential for death signaling (25). However, recent data has suggested that apical caspases can, under certain conditions, transmit signals without cleavage by relying only on dimerization or other rearrangements of inactive unprocessed monomers (26). Nevertheless, processing of apical caspases stabilizes the catalytic activity as heterotetramers (two each of the large and small subunits) and allows its release into the cytosol (26). The released, highly active caspases cleave multiple

downstream targets, including effector caspases that initiate a caspase ampliﬁcation loop leading to the cell’s demise. In the so-called type I cells, including thymocytes and peripheral T cells, a high quantity of active caspase8/10 is generated at the DISC that directly activates effector caspases, i.e., caspases-3, -6, and -7. By contrast, in type II cells, including tumor cell lines such as Jurkat and CEM, the DISC is weaker, and caspase activation occurs more slowly and to a lesser extent than in type I cells. In type II cells, robust apoptosis is achieved by caspase-8 processing of Bid, a proapoptotic BH3-only Bcl-2 family member. The truncated active form of Bid causes cytochrome c release from the mitochondria www.annualreviews.org • Genetic Disorders and Cell Death

325

ANRV270-IY24-11

ARI

23 February 2006

SPOTS: signaling protein oligomerization transduction structures

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

c-FLIP: cellular FLICE inhibitory protein DNT cells: double-negative T cells

326

15:22

and activation of the intrinsic apoptosis pathway (27). Because of mitochondrial involvement, death receptor–mediated death in type II cells, but not type I cells, can be inhibited by anti-apoptotic Bcl-2 family members (28). DISC formation is followed by the formation of Fas clusters recently designated as signaling protein oligomerization transduction structures (SPOTS) (29, 30). SPOTS are composed of perhaps thousands of individual receptor complexes (making them visible microscopically) that further increase the local concentration of caspase-8 and promote its autoprocessing and activation (30). This receptor oligomerization depends on the Fas DD and FADD but does not require caspase activity. Once SPOTS are formed, a further caspase-dependent clustering and capping of Fas complexes is followed by receptor internalization in vesicles bearing endosomal markers. It is unclear whether internalization downregulates Fas-induced death or is a critical step for death signaling, as recently shown for the TNF receptosomes (31). Alternatively, since Fas internalization depends on caspase activation, it may be a post-lethal event that has no regulatory role at all. One potentially important regulator of Fas at the DISC level is the cellular FLICE inhibitory protein (c-FLIP) that is present in two isoforms: c-FLIPS and c-FLIPL (32). The c-FLIPL protein is very similar to caspases-8 and -10, and the genes for all three are found in a gene cluster on chromosome 2q33-34; this observation suggests that the genes may have arisen by tandem duplication of one ancestral gene (33–35). Although a double DED is intact in c-FLIPL , the part of the gene that encodes the enzymatic subunits contains numerous mutations that prevent caspase activity (34). c-FLIPS is closer in structure to a viral counterpart, v-FLIP, which consists only of the double DED and a short COOH-terminal addition (34). Both c-FLIPL and c-FLIPS can be recruited to the DISC, where they typically exert an inhibitory effect on caspase-8 activation and cell death. c-FLIPL has been observed to have different, sometimes conBid`ere

· · Su

Lenardo

tradictory functions. c-FLIPL in the DISC is processed by caspase-8 but cannot process caspase-8 in return, so that the latter is kept in an uncleaved form at the receptor level (36). B cell receptor stimulation can induce c-FLIPL expression and block Fas- and TRAIL-mediated killing; this observation indicates a potential role in B cell selection and tolerance (37). On the other hand, small amounts of c-FLIPL can cause caspase-8 activation and cell death and have also been proposed to promote Fas-induced lymphocyte proliferation (32).

THE AUTOIMMUNE LYMPHOPROLIFERATIVE SYNDROME (ALPS) Clinical Presentation ALPS is a rare human disorder deﬁned by lymphoproliferation, peripheral expansion of double-negative αβ T (DNT) cells, and impaired lymphocyte apoptosis. Autoimmune disease and an increased risk of lymphoma are also observed. The clinical presentation of ALPS in humans, which has been summarized in previous excellent reviews, reﬂects impaired lymphocyte homeostasis, particularly of the DNT cells (38, 39). The diagnostic criteria are listed in Table 1. Lymphoproliferation apparently results from the gradual accumulation of lymphocytes that have not undergone normal programmed cell death. This proliferation leads to chronic enlargement of the lymph nodes, thymus, liver, and/or spleen, beginning in early childhood. Both B cells, including CD5+ B cells, and T cells are elevated. Not all lymphocyte subsets are affected, as CD4+ CD25+ absolute numbers are reduced and CD4+ CD25− absolute numbers are not increased. By contrast, expansion of an unusual population of peripheral CD4− CD8− T cells is striking. These peripheral DNT cells express αβ T cell receptor (TCR) chains as well as the CD45R isoform B220 (typically expressed on B cells); this expression distinguishes them from

ANRV270-IY24-11

ARI

23 February 2006

15:22

Table 1 Diagnostic criteria for autoimmune lymphoproliferative syndrome (adapted from Reference 39)

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Criteria

Description

Required features

Chronic nonmalignant lymphadenopathy and/or splenomegaly Peripheral expansion of CD4− CD8− TCRα/β T lymphocytes (DNT cells) Impaired lymphocyte apoptosis

Supporting features

Family history of ALPS Characteristic histopathology Autoimmune disease Germline mutations in Fas, Fas ligand, or caspase-10 genes Somatic mutation in Fas gene in DNT cells

peripheral γδ T cells, which naturally lack CD4 and CD8 coreceptors (40). DNTs also differ from double-negative thymocytes, an immature stage of T cells in which TCR genes have not yet completely rearranged (41). The DNT cells could be either previously activated, mature T cells that have lost CD8 or CD4 coreceptor expression, or a special minor cell lineage (42). In the blood, DNT cells constitute less than 1% in normal individuals, but can reach up to 40% in patients with ALPS (40). The DNT expansion is also evident histologically by nearly pathognomic abnormalities of the architecture of secondary lymphoid tissues, with paracortical and follicular hyperplasia (39). DNT cells appear to play a crucial role in disease development. In certain ALPS patients lacking germline mutations in the death receptor Fas, Holzelova et al. (43) found a population of DNT cells harboring somatic dominantly interfering Fas mutations. Thus, unregulated and excessive DNT cells are seemingly sufﬁcient to cause disease. Two previously described characteristics of DNT cells may explain how. First, the DNT cells are primary producers of strikingly elevated IL10 observed in ALPS patients. By contrast, DNT cells from normal individuals do not produce IL-10. Moreover, healthy relatives of ALPS patients bearing both Fas mutations and in vitro apoptotic defects have only modestly elevated DNTs and IL-10 (44). Besides elevated IL-10, ALPS patients also exhibit in-

creased IL-4 and IL-5, but decreased IL-2 and IFN-γ (44, 45). This cytokine proﬁle is characteristic of T helper type 2 (Th2) cells, which inhibit cell-mediated immunity and promote humoral immune responses. Thus, the overall cytokine environment may favor autoantibody production in ALPS. Second, DNT cells exhibit an unusual phenotype including B220 expression and altered cell surface O-glycans (40). This phenotype could change the trafﬁcking pattern and/or potential interactions of DNT cells with other cell types. There are some important and unresolved paradoxes regarding DNTs. First, they are very difﬁcult to culture in vitro and almost immediately die despite an ostensible resistance to apoptosis in vivo. Their lifespan is not extended by IL10. Second, although DNTs are believed to play a role in hyperactive immune responses, they are generally unresponsive to proliferative and activating stimuli (46). Finally, the antigen speciﬁcities recognized by the TCRs on DNTs have not been deﬁned, but the general assumption is that they will include selfantigens. Autoimmunity accompanies lymphoproliferation in ALPS. Most patients have elevated serum immunoglobulin levels and autoantibodies. Anti-cardiolipin antibodies are frequent but not usually associated with thromboembolic disease (39). Many patients exhibit Coombs’ positive hemolytic anemia and/or immune thrombocytopenia (39). These two features are also observed in

www.annualreviews.org • Genetic Disorders and Cell Death

327

ARI

23 February 2006

15:22

Evans syndrome, a diagnosis that can be confused with ALPS (47). Less common autoimmune manifestations in ALPS include anti-nuclear antibodies, rheumatoid factor, autoimmune neutropenia, glomerulonephritis, uveitis, autoimmune hepatitis, primary biliary cirrhosis, Guillain-Barr´e, vasculitis, linear IgA dermopathy, and anti-Factor VIII antibodies with coagulopathy (39, 48). It is not known why, in contrast to other autoimmune diseases, antibody-mediated autoimmune disease is primarily directed against the hematopoietic system in ALPS patients. However, autoimmune phenotyping is imprecise, as illustrated by the discovery of a patient, initially diagnosed with systemic lupus erythematosus (SLE), who was later found to have a Fas ligand mutation with defective apoptosis (49). An interesting question is whether the reduced CD4+ CD25+ cells seen in ALPS patients might represent a defect in regulatory cells that contributes to autoimmune disease. Although lymphoproliferation is initially nonmalignant, ALPS patients have a marked propensity to develop B or T cell malignancies. To date, we observed lymphomas in type I ALPS patients bearing dominant Fas DD mutations that cause severe apoptosis defects. We found a 14-fold and 51-fold increased incidence of non-Hodgkin lymphoma and Hodgkin lymphoma, respectively (50). The lymphomas were found anywhere between 15 to 48 years after onset of ALPS symptoms, with a median age of 21 for non-Hodgkin lymphoma and 11 for Hodgkin lymphoma (50; J. Dale & S. Straus, personal communication). Analysis of lymphoma tissue from ALPS patients showed no loss of Fas heterozygosity or increased apoptosis resistance. Interestingly, the responsible mutations impair apoptosis, but not Fas-induced NF-κB and mitogenactivated protein kinase (MAPK) signaling (51). Hence, the contribution of Fas mutations to lymphomagenesis may involve blocking death and releasing growth-promoting effects. It is also possible that the increased number of lymphocytes in general

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

328

Bid`ere

· · Su

Lenardo

may provide a larger pool for tumorigenesis (52). In an early clinical description of ALPS, Sneller and colleagues (46) recognized that ALPS closely resembled the lpr/lpr (lpr) and gld/gld (gld) mouse strains. On certain inbred genetic backgrounds, homozygosity of these alleles causes mice to develop profound lymphoproliferation involving DNT cells, hypergammaglobulinemia, autoantibodies, autoimmune disease, and lymphomas (52, 53). Not surprisingly, these naturally arising mouse strains have molecular defects in some of the same genes as ALPS patients. The lpr and gld mice have autosomal recessive mutations in Fas and FasL, respectively, resulting in deﬁcient expression, whereas the lprcg variant has a point mutation in Fas that renders it nonfunctional (54, 55). Although the Fas locus is the primary genetic determinant, there is a strong effect of the inbred mouse strain background on the autoimmune phenotype. Severe disease occurs on the MRL background, whereas disease is substantially reduced on a B6 or BALB/c background. These mice do not develop the autoimmune hemolytic anemia and thrombocytopenia seen in human patients with ALPS. Instead, they display certain features resembling SLE (anti-dsDNA, anti-Sm, anti-immunoglobulin autoantibodies) with a predilection for developing glomerulonephritis, polyarteritis, sialoadenitis, and, with lesser frequency, arthritis or primary biliary cirrhosis (56, 57). Using intercrosses and backcrosses of diseasesusceptible and disease-resistant lpr strains, several groups have performed genome-wide linkage analysis (56). Mapping of genetic modiﬁers of disease revealed loci for lymphoproliferation and autoantibody production (i.e., Lmb1, Lmb2, and Lmb3), sialoadenitis (i.e., Asm and Asm2), glomerulonephritis (i.e., Agnm1, Agnm2, and Agnm3), and arthritis (i.e., Paam1 and Paam2) (56, 58). The Agnm3 locus associated with glomerulonephritis corresponds to allelic polymorphisms of the osteopontin gene. Polymorphisms of this gene, which is highly expressed in

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

T cells, correspond to functional differences in activating macrophages and B cell antibody production (59). Interestingly, allelic polymorphisms of osteopontin may inﬂuence disease penetrance in humans who have a variant of ALPS termed ALD (described below) (60). In all but the most severely affected patients, the clinical course of ALPS waxes and wanes over time and often improves with age. Thymic involution may decrease the output of T cells, and it is known that neonatal thymectomy in lpr mice prevents disease (61, 62). Clinical complications relate mainly to the severity of autoimmune disease. Infections may occur after splenectomy or secondary to immunosuppressive treatment. Newer agents, such as mycophenolate mofetil, may bypass the need for prolonged courses of corticosteroids or splenectomy (39). Notably, pyrimethamine plus sulphadoxine has also ameliorated symptoms in some ALPS patients, possibly through a mechanism involving induction of lymphocyte apoptosis via an alternative mitochondrial pathway (63). There are published reports of curative bone marrow transplant in patients with Fas mutations on both alleles and intractable lymphoproliferative and autoimmune disease (64). However, conservative treatment generally remains sufﬁcient to control disease in most patients. Finally, minor clinical variants of ALPS may reﬂect different and uncharacterized molecular defects in lymphocyte homeostasis mechanisms. A signiﬁcant fraction of ALPS or ALPS-like patients has no mutations in the genes encoding proteins involved in the Fas signaling. Approximately half of these patients have no apoptosis defect following direct challenge with an agonist anti-Fas antibody (H. Su & M. Lenardo, unpublished results). Although the ALPS diagnosis requires documentation of impaired apoptosis of mature lymphocytes, most laboratories rely on assessment of apoptosis induced through the Fas death receptor. Defects in Fas-mediated apoptosis usually coincide with defects in TCR-restimulation-induced death;

however, an ALPS patient with normal Fasmediated apoptosis but impaired apoptosis following stimulation with phytohemagglutin plus IL-2 has been described (65). ALPS could also result from defects involving non– death receptor–induced intrinsic pathways of apoptosis. We have recently characterized an ALPS patient who displays the usual clinical features, but has cytokine withdrawal apoptotic defects with intact Fas-mediated death ( J.B. Oliveira & M. Lenardo, unpublished results). Patients can also exhibit the main clinical features as ALPS, with lymphoproliferation, apoptosis defects, autoimmune disease, and cancer, but lack DNT cell expansion. This ALPS variation has been termed autoimmune lymphoproliferative disease (ALD) (66). Although the responsible molecular defect(s) is (are) yet undiscovered, ALD patients likely share with ALPS patients a common affected pathway downstream of the Fas receptor. In summary, there are number of variant and overlapping clinical subtypes that will be extremely interesting to explore at the molecular level.

ALD: autoimmune lymphoproliferative disease

Molecular Diagnosis The molecular diagnosis of ALPS encompasses the following three, possibly four, major categories as well as some selected molecular subtypes (summarized in Table 2 and on the web at http://research.nhgri.nih. gov/ALPS/). For type Ia, the heterozygous mutations may be found in any part Table 2 Molecular classification of autoimmune lymphoproliferative syndrome ALPS classification

Gene mutated

Ia

Fas (TNFRSF6)

Ib

FasL (TNFSF6)

Ic or Im

Fas somatic mutation in DNT cells

II

Caspase-10

III

Molecularly undeﬁned

www.annualreviews.org • Genetic Disorders and Cell Death

329

ANRV270-IY24-11

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

PLAD: pre-ligand assembly domain

330

15:22

of the Fas coding sequence, but over half are found in exon 9, which encodes the intracytoplasmic DD (see http://research.nhgri.nih.gov/ALPS/dist fas mutations.shtml). In general, the mutations permit the expression of a pre-ligand assembly domain (PLAD)-containing abnormal version of the Fas protein that has a dominant-interfering effect (see below; for review of the PLAD domain, see References 67 and 68). Nevertheless, there are exceptions to this rule. For example, some early exon 2 mutations essentially eliminate protein expression; this observation raises the question of whether haploinsufﬁciency may cause disease in certain circumstances. Recently, evidence supporting this hypothesis has been put forward (69). Type Ib involving FasL is much less common, but several cases are known (38, 70; S. Straus & M. Lenardo, unpublished observations). Molecular analysis shows that point mutations generate mutant versions of the FasL protein that are dominant interfering. Type Ic (or Im) is ALPS that results from somatic mutations in DNT cells. The possibility of somatic mutation in DNT should be investigated in any ALPS-like patient who does not seem to have a Fas, FasL, or caspase-10 mutation, especially those that do not have a clear-cut apoptosis defect in normal lymphocytes. Type II ALPS involves mutations in the caspase-10 gene (see http://research. nhgri.nih.gov/ALPS/alpsII mut.shtml and Reference 3). Originally, two cases were reported, although reinvestigation of the V410I mutation has revealed that it is widely distributed in Western European populations and brought into question its disease-causing potential (71). However, at least two more caspase-10 mutations have been detected in ALPS patients lacking Fas or FasL mutations ( J. Puck, S. Straus & M. Lenardo, unpublished observations). Originally, there was a suggestion that caspase-8 mutations may also cause type II ALPS. However, in the single family identiﬁed so far with caspase-8 mutations, the disease features constitute a distinct Bid`ere

· · Su

Lenardo

entity, as no dramatic increases in DNTs are observed; this fact may account for the lack of many hallmarks of ALPS. Additional families will be required to determine whether caspase8 mutations can cause a true ALPS phenotype. Until now, type III has been used to describe ALPS with an unknown molecular pathogenesis. This terminology is likely to change rapidly as new molecular abnormalities are uncovered. Hence, the cases with an unknown molecular pathogenesis will become type IV, etc. These undiagnosed cases are likely to harbor a wealth of interesting molecular insights into lymphocyte regulation and autoimmunity.

ALPS Type Ia Unveils Dominant Interference, Fas Preassociation, and Cooperativity ALPS Ia patients have heterozygous Fas mutations that impair apoptosis. More than 200 ALPS Ia patients have been described so far. The vast majority of them harbor mutations located in the region of the gene that encodes the intracellular domain of Fas, with a preponderance (60%) in the DD (38). These mutations are mostly missense or nonsense mutations that generate mutant proteins, as opposed to full loss-of-expression mutations. The mutant proteins have been shown to greatly reduce FADD binding and consequently impair recruitment and activation of caspases even in the presence of the wildtype protein (72). Why do dominant mutations have this effect? It does not appear to be due to selective expression of the mutant chains. Rather, the mutant chains enter complexes with the good chains, and these mixed complexes do not transduce the death signal effectively. In genetic terms, this constitutes dominant interference (73). It was presumed that the association of good and bad chains was due to the recruitment or crosslinking of the chains into mixed trimeric complexes by ligand binding. Mutations in the region of the gene that encodes the Fas extracellular domain also

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

occur in ALPS type Ia. Roughly half of these mutations lead to truncated forms of Fas, and the other half result in an amino acid sequence modiﬁcation. Again, most of these mutations proved to be dominant interfering despite the puzzling fact that they failed to bind ligand (74, 75). These unanticipated observations led to the theory that Fas receptor chains formed trimers in a ligand-independent fashion (75, 76). It has now been directly demonstrated that Fas, TNFR1, and TNFR2 selfassociate as oligomers, independently of any ligand, through the PLAD, an approximately 50 amino acid N-terminal domain (75, 76). For Fas, the PLAD and ligand binding site are distinct. This fact explains why ALPS mutant chains with impaired FasL binding can form mixed trimers with wild-type Fas chains. The PLAD was found to be intact in all ALPS type Ia patients with dominant-interfering mutations. Moreover, Fas preassembly is necessary for FasL binding and correct DISC formation (75, 76). How is Fas preassociation linked to downstream signaling, and why does preassembled Fas not form a DISC in the absence of FasL? One can imagine two major allosteric conformations (or chain orientations): inactive and active. Binding of FasL would mediate an activating change that would propagate to the intracellular portion of Fas to induce a stoichiometric 3:3 Fas:FADD complex. Evidence suggests that this complex is cooperatively formed and highly stable (H. Wu & M. Lenardo, unpublished observations). The preassociation model has important implications in addition to explaining dominant interference in ALPS. It explains how distinct receptor chains that bind the same ligand, such as TNFR1 and TNFR2, can form homotypic complexes, as the PLAD from each receptor appears to be speciﬁc for self binding. Conformational change of preassembled dimeric receptors upon stimulation has been described for the erythropoietin receptor and other surface receptors (77). The fact that preassembly was also observed for TNFR1 and TNFR2 suggests that other members of the TNFR superfamily may have PLADs and are there-

fore susceptible to dominant-interfering mutations (76). Besides the dynamics of protein interactions in the DISC, higher order interactions such as those leading to SPOTS may also participate in ALPS signaling defects. Higher order complex formation may require cooperative interactions between individual receptor complexes in addition to cooperative recruitment of signaling proteins. For example, transient transfection of increasing amounts of Fas mutant alleles in Jurkat cells exerts a powerful transdominant interfering effect in preventing FADD and caspase association (74). The disruption of cooperative interactions may explain why dominant mutations—some of which only partially inhibit FADD:Fas binding—can markedly decrease apoptosis signaling. A DD mutation was found in an ALPS patient that permitted DISC formation but impaired SPOTS and subsequent caspase-8 cleavage. Cooperativity may serve to make signaling an all or none response to the level of receptor stimulation by ligand. Holler et al. (78) recently showed that two trimers of FasL are required to trigger a potent DISC formation. However, supercluster formation is also dependent on signals transduced from the intracellular portion of Fas to the extracellular domain (“inside-out” signaling) as reduced FADD and caspase-8 recruitment in ALPS Ia patients also abolishes SPOTS formation (30). Thus, cooperativity makes signaling efﬁcient but also vulnerable to dominant interference.

Caspase-10 as a Major Player in Death Receptor–Induced Apoptosis The recent report of two ALPS patients who bear caspase-10 mutations has unveiled the pivotal role of this protease during death receptor–induced apoptosis (3). The caspase-10 mutants show signiﬁcantly reduced enzymatic and autoprocessing activities, and lymphocytes from these patients exhibit an impaired Fas-, CD3-, and www.annualreviews.org • Genetic Disorders and Cell Death

331

ARI

23 February 2006

15:22

TRAIL-induced death. In line with these observations, caspases-8 and -10 are recruited and activated at the DISC level in FasLtreated PBL or Jurkat cells (79, 80). It is likely that these caspases cooperate, as mutants for caspases-8 and -10 could cross-inhibit one another and can enter the same receptor complex (80, 81). However, the relationships between these two very close caspases remain unclear. Although one study suggests that caspase-10 cannot substitute for caspase-8 in Fas- or TRAIL-induced death (82), some reports demonstrate that the two enzymes can trigger apoptosis independently from each other (79, 80). Two different roles for these proteases are conceivable, as their regions involved in substrate recognition and binding clearly differ (33). Moreover, caspase-10 as well as caspase-8 triggers cell death via Bid processing and caspase activation, but caspase-10 does not cleave receptor interacting protein (RIP), a canonical substrate for caspase-8 (80). Interestingly, one patient with a deleterious caspase-10 mutation exhibited an unprecedented defect in dendritic cell apoptosis in response to TRAIL that is associated with an overaccumulation of dendritic cells in the T cell areas of the lymph nodes (3). Other apoptosis-inactivating caspase-10 mutations were also described in non-Hodgkin lymphomas as well as in gastric cancers (83, 84).

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

Type III and Variant ALPS There are two important categories of patients within the spectrum of ALPS that do not easily fall into the aforementioned groups. The ﬁrst group is referred to as type III ALPS. These patients meet ALPS clinical criteria and exhibit a mild resistance to Fasinduced apoptosis but do not carry any mutations in the genes that encode Fas, FasL, FADD, caspase-8, or caspase-10. In some of these patients, FADD and caspase-8 are normally recruited to the DISC following Fas crosslinking. However, DISC formation is abortive, as caspase-8 remains largely un332

Bid`ere

· · Su

Lenardo

processed at the receptor level (N. Bid`ere, J.B. Oliveira & M. Lenardo, unpublished results). This observation suggests that the processing and release of active caspase-8 from the DISC is tightly regulated, although the complete mechanism is unknown. We hypothesize that mutations in such regulators might also lead to ALPS development. Among possible regulators, c-FLIP is an appealing candidate as it blocks caspase-8 activation at the receptor level. However, transgenic mice that overexpress FLIPL in the T cell compartment do not exhibit T cell accumulation in secondary lymphoid organs as in MLR/lpr mice (85). It is also plausible that defects in the molecules controlling the conformational change of Fas or its aggregation and internalization might lead to the disease. Other less-characterized cytoplasmic partners of Fas, such as Daxx, RIP, or FAF1, might be involved. The DD-containing serine/threonine-containing kinase RIP has been described as a component of the Fas DISC (86). RIP seems to play a critical role in the nonapoptotic death following Fas stimulation when caspases are inhibited (16). Another category of patients, called variant ALPS, exhibits the phenotypic characteristics of ALPS despite their lymphocytes’ normal sensitivity to Fas killing.

Genetic Modifiers of ALPS Disease ALPS exhibits variable disease penetrance and severity. Over two thirds of patients have heterozygous mutations in Fas and exhibit a range of phenotypes from no symptoms at all to lifethreatening autoimmunity. A major factor in disease phenotype appears to be the nature of the Fas mutation and how it affects apoptosis signaling. One patient was found to be a compound heterozygote of two Fas mutations and presented a severe clinical phenotype (87). Among patients with heterozygous Fas mutations, the location of the mutation affects penetrance and severity. For instance, dominantinterfering alleles encoding DD alterations that severely affect apoptosis signaling are

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

associated with a greater penetrance and disease severity (75, 88). By contrast, mutations in the region of the gene that encodes the extracellular domain, even if dominant interfering, are less penetrant and cause milder disease. A minority of patients who have haploinsufﬁcient mutations in the region of the gene that encodes the extracellular domain tends also to have milder disease. Nonetheless, even within these broad groups, there is signiﬁcant heterogeneity. For any given patient, it is not uncommon that multiple family members possessing the same mutation manifest impaired apoptosis in vitro but do not have ALPS disease. This situation has been mimicked by generating mice transgenic for human Fas DD mutations that were backcrossed onto FVB/N or MRL strains (89). Compared to transgenic FVB/N or MRL strains, FVB/N X MRL mice transgenic for identical Fas mutations had incompletely penetrant autoimmune disease that correlated with degree of apoptotic defect (89). The existence of healthy family members of ALPS patients and the inﬂuence of the inbred genetic background on the phenotype of lpr and gld mice strongly suggests the existence of modiﬁer genes. A modiﬁer gene could be deﬁned as a gene encoded by a locus unlinked to the primary ALPS mutation that determines the ultimate disease phenotype. Several modiﬁer or susceptibility genes have been linked to other autoimmune diseases and are therefore candidates for ALPS modiﬁers. Perhaps the most common is the association of speciﬁc human leukocyte antigen (HLA) alleles with autoimmune diseases (90). However, analyses of large numbers of ALPS patients and their relatives have not yet yielded any speciﬁc HLA associations with disease exacerbation, although HLA-B44 may have a protective effect (M. Vacek & J. Puck, unpublished observations). This result is surprising as HLA-B44 had been previously linked to Crohn’s disease (91). Certain inherited forms of caspase-10 may also act as genetic modiﬁers of disease in ALPS patients (S. Zhu, L. Zheng, M. Lenardo & J. Puck, unpublished

results). In addition to caspase-10, other cellkilling molecules may modulate disease in the presence of Fas mutations. A patient was recently described with a compound heterozygous Fas splicing mutation and a perforin mutation, which were inherited separately from healthy parents (92). However, it was unclear if the Perforin alteration represented a polymorphism or functional mutation. In a separate family, the identical Perforin alteration was found together with a Fas mutation, yet that individual did not have ALPS (93). Cytotoxic T lymphocyte–associated antigen4 (CTLA-4) polymorphisms have been associated with autoimmune thyroid disease and type 1 diabetes mellitus (94). However, analysis of a small number of ALPS patients revealed no relationship to CTLA-4 polymorphisms (L. Wicker, personal communication). Besides modifying ALPS disease penetrance, other genes may affect the clinical presentation of disease. Of 151 ALPS patients currently followed at the National Institutes of Health, several have other superimposed inﬂammatory conditions. Periodic fevers suggestive of TNF-receptor-associated periodic syndrome (TRAPS) were observed in six ALPS patients, and two of these individuals were found to have a R92Q functional polymorphism in the p55 TNF-R1 gene (I. Aksentijevich, R. Siegel, S. Straus & D. Kastner, personal communication). The carrier frequency of this polymorphism is higher in TRAPS and other disorders such as rheumatoid arthritis when compared with the general population (95). The biochemical mechanism by which the R92Q polymorphism contributes to disease is unclear. A similar situation exists for another ALPS patient, who also has Familial Mediterranean Fever with an associated functional polymorphism of the pyrin gene (E148Q) (96). Moreover, one of the ALPS patients with the R92Q polymorphism for p55 TNF-R1 has a caspase10 gene alteration; this fact suggests that, by modifying coexisting inﬂammatory conditions, these genes may inﬂuence ALPS disease penetrance and severity. www.annualreviews.org • Genetic Disorders and Cell Death

333

ANRV270-IY24-11

ARI

23 February 2006

15:22

CASPASE-8-DEFICIENCY STATE (CEDS)

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CEDS: caspase-8-deﬁciency state

334

Clinical Presentation The study of a wide spectrum of patients exhibiting some features of ALPS has revealed new and unexpected biological roles for molecules that function in programmed cell death. We found two siblings with several ALPS features who surprisingly also fulﬁlled diagnostic criteria for common variable immunodeﬁciency (CVID) (2). Remarkably, these children were found to have homozygous inactivating mutations in caspase-8 (2). It was surprising that the caspase-8-deﬁciency state (CEDS) was compatible with embryonic development in humans as the caspase-8 gene knockout caused embryonic lethality in mice (97). One explanation could be the presence in humans, but not in mice, of the highly related caspase-10. The CVID-affected children had recurrent sinopulmonary infections and mucocutaneous herpesvirus infections. Despite elevated lymphocyte numbers, serum immunoglobulin levels were reduced, and immunization with polysaccharide antigen yielded poor antibody responses. Moreover, activation of B and T cells (through antigen receptors) or natural killer (NK) cells (through Fc - or NK-activating receptors) were all impaired. Although immunodeﬁciency was their most striking clinical feature, the patients also had secondary lymphoid enlargement, impaired lymphocyte apoptosis in vitro, and occasional borderline elevations in DNT cells (2). Although caspase-8 mutations explained abnormal apoptosis and lymphocyte homeostasis, the clinical ﬁndings of immunodeﬁciency also pointed to an unappreciated role for caspase-8 in lymphocyte activation. By using a novel transfection technology to introduce caspase-8-speciﬁc small-interferingRNAs into primary human lymphocytes, we discovered that caspase-8 was indeed required in normal lymphocytes for activation through the T and B cell antigen receptors, the Fc and 2B4 receptors of NK cells, and TollBid`ere

· · Su

Lenardo

like receptor 4 (TLR4) (2, 4). Our observations explained the puzzling ﬁnding that pharmacologic caspase-inhibitors could interfere with T lymphocyte activation in vitro (98). These ﬁndings have also been essentially conﬁrmed in mice in which T cells or B cells were conditionally rendered caspase-8 deﬁcient by homologous recombination (99; R. Hakem, personal communication). The T cell–knockout mice eventually succumb to an unusual autoimmune disease characterized by lymphocytic inﬁltration of lung, kidney, and liver parenchyma (L. Salmena & R. Hakem, personal communication). Clinical follow-up of the caspase-8-deﬁcient patients has revealed new bronchiectasis, as well as autoantibodies or a compensated autoimmune hemolytic anemia (H. Su, S. Straus & K. Rao, unpublished observations). Whether these new symptoms reﬂect complications of CVID, which in itself can be associated with autoimmune disease, and/or a superimposed forme fruste of ALPS is presently unclear. The resulting clinical syndrome shares certain features with ALPS, but clearly has unique features. Its striking clinical hallmark is a primary combined lymphoid immunodeﬁciency with susceptibility to infection. Importantly, despite an extensive literature on caspase-8 and Fas-induced programmed cell death in nonlymphoid tissues, so far CEDS patients have not manifested primary abnormalities in other organ systems. In summary, these ﬁndings underscore the importance of examining critical pathways in humans rather than relying upon mice or other model organisms. Clearly, there are important limitations to the use of studies in mice for the elucidation of human disease.

Molecular Effects of Caspase-8 Deficiency The possible involvement of caspase-8 in antigen receptor–induced lymphocyte activation was puzzling at ﬁrst glance because the known functions of caspases were limited to cytokine maturation and mediation of programmed

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

cell death (25). The pivotal role of caspase8 in both activation and death of lymphocytes was ﬁnally clearly demonstrated (2, 4, 99). These observations extended the functions of caspases to include roles as intracellular signaling mediators during cell activation and proliferation. The implications of the evolution of these dual roles for caspases and other death molecules have been previously discussed (23). These observations also raised the question of the molecular pathway on which caspase-8 exerts its effect during lymphocyte activation. Recently, we conclusively demonstrated that caspase-8 is essential for antigen receptor–induced NF-κB activation (4). During the antigenic stimulation of T cells, the transcription factor NF-κB induces genes that regulate cellular proliferation and cell survival, such as IL-2 and CD25 (highafﬁnity IL-2 receptor α-chain) (100). The DNA-binding components of NF-κB are constitutively present in the cytosol bound to speciﬁc inhibitory proteins called inhibitors of κB (IκBs). NF-κB activation involves the phosphorylation and ubiquitin-mediated degradation of IκBs. The degradation of IκB enables NF-κB to translocate to the nucleus and govern the transcription of a large number of genes by binding to speciﬁc DNA sites in gene regulatory regions. The phosphorylation of IκBs is carried out by the IκB kinase (IKK) complex, which is composed of two catalytically active kinases (IKKα and IKKβ) and a catalytically inactive regulatory subunit (IKKγ, also known as NEMO) (101). How the signal is conveyed from antigen receptor to IKK complex has not been fully elucidated. However, the recent discovery of three critical proteins, CARMA1, BCL10, and MALT1, upstream of IKK has shed new light on the NF-κB activation pathway (102). It is now accepted that the Ca2+ -independent protein kinase C (PKC) θ is rapidly recruited to membrane lipid rafts at the immunological synapse following antigen receptor stimulation (Figure 2). PKCθ then triggers IKK acti-

vation by initiating the formation of a large complex that includes CARMA1, BCL10, and MALT1, termed the CBM complex. Genetic and biochemical approaches have established the pivotal role of CARMA1 upstream of a preassociated BCL10 and MALT1 complex (102). Recently, Lee et al. (103) have shown that the 3-phosphoinositide-dependent kinase 1 (PDK1) is a bridge linking the IKK and CBM complexes to PKCθ. How the CBM complex activates IKK is still unclear. It is thought that IKK activation requires the noncanonical K63-linked ubiquitylation of NEMO (102). This could be achieved by the binding of the ubiquitin-conjugating E2 complex (composed of UBC13 and MMS2) to MALT1 after its oligomerization or, alternatively, through the binding of the RING domain ubiquitin ligase TRAF6. This modiﬁcation of NEMO would induce the phosphorylation and activation of IKK. Our evidence now shows that caspase-8 appears to be an essential nexus between the CBM and IKK complexes (Figure 2). This protease physically interacts with both the CBM and IKK, and its absence precludes the entire formation of the activation-receptor induced signalosome (ARIS) and prevents de facto IKK activation (4). This unexpected role for caspase-8 in lymphocyte activation explains the characteristic immunodeﬁciency in CEDS patients. Defective induction of NFκB during the activation of T, B, and NK cells will severely incapacitate critical gene expression events. How caspase-8 protein assembles the ARIS and what comprises the ﬁnal signaling complex is still unknown. Mutation of the active site cysteine inhibits the NF-κB activating function of caspase-8; this result clearly implies that the complex includes a crucial proteolytic substrate. The identiﬁcation of this substrate will provide considerable insight into how the pathway works. The adaptor FADD is also involved, although there is no evidence that it is cleaved. FADD is transiently recruited to the CBM, and whether it plays a key role in the assembly of the ARIS needs to be further explored. www.annualreviews.org • Genetic Disorders and Cell Death

IKK: IκB kinase CBM complex: CARMA1-BCL10MALT1 complex ARIS: activation-receptor induced signalosome

335

ARI

23 February 2006

15:22

Genetic deﬁciency of FADD or transgenic expression of a dominant-interfering mutant of FADD have been found to impair both death receptor–induced and mitogen- and antigeninduced T cell activation (24, 104). Future work should also be aimed at determining whether the caspase-8 regulator FLIP is part of ARIS and whether it regulates caspase-8 activation or downstream signaling events, as has been previously suggested (32). Interestingly, enzymatic activity but not autoprocessing of caspase-8 is required during antigen receptor–induced NF-κB activation (4). It is likely that caspase-8 activity generated within the ARIS is weaker than during DISC formation, as the enzyme remains unprocessed. Furthermore, evidence suggests that there is no release of the active caspase-8 from the ARIS, as compared with DISC formation, where the active enzyme is proteolytically released from the prodomain. The local and restricted caspase-8 activity in the ARIS likely triggers pro-survival events and avoids release of the protease into the cytosol, where it might ﬁnd apoptosis substrates. Thus, depending on its location, its concentration, and its partners, caspase-8 controls the fragile balance between life and death in lymphocytes during immune responses.

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

RELATED DISORDERS OF LYMPHOCYTE HOMEOSTASIS The use of knockout mice has enabled the identiﬁcation of several other genes that regulate programmed cell death during normal lymphocyte homeostasis. These genes may provide clues to human diseases that have not yet been uncovered. Table 3 lists mouse strains that exhibit lymphocyte apoptotic defects with lymphoproliferation or other characteristics suggestive of ALPS in humans. Because approximately 20% of ALPS (type III) patients have no identiﬁed gene defect, the mouse studies may provide candidates as to potential defects in these individuals. Several themes emerge, which we discuss below. 336

Bid`ere

· · Su

Lenardo

Abnormal IL-2 signaling results in an ALPS-like picture in mice. Mice genetically deﬁcient in IL-2 or the IL-2 receptor subunits IL-2Rα (CD25) and IL-2Rβ (CD122) develop autoimmune hemolytic anemia, colitis, lymphoproliferation, and early death (105– 107). The one reported human with a deleterious IL-2Rα mutation had severe multisystem autoimmune disease (108). Despite its growth-promoting properties, IL-2 is also required for programmed death of T cells after antigen receptor restimulation. Defective death following TCR restimulation was directly shown when T cells were initially treated with limiting doses of IL-2 (109) and was conﬁrmed using cells from IL-2−/− or IL2Rα−/− mice (19, 110). Defective death mediated through the Fas death receptor was also observed in T cells from IL-2−/− or IL-2Rα−/− mice (112, 113). Apoptotic defects occurred in vivo, as administration of anti-IL-2 neutralizing antibodies with antigen resulted in persistence of adoptively transferred antigen-speciﬁc TCR-transgenic T cells (110), and impaired peripheral deletion of Vβ8+ T cells was seen after administration of the bacterial superantigen Staphylococcus enterotoxin B (SEB) in mice treated with anti-IL-2Rα blocking antibodies (109). These results were conﬁrmed in similar mouse models (107, 112, 114). IL-2 regulates numerous genes known to affect apoptosis (19, 115), and slight differences between these mouse models may reﬂect variations in the molecular targets perturbed by impaired signaling downstream of IL-2. Subsequent studies seemed to challenge the idea that an intrinsic inability of T cells to die in the absence of proper IL-2 signaling contributes to disease, but they failed to consider the possibility of compensatory mechanisms for cell death. T cell hyperactivation and hematopoietic abnormalities in IL-2Rβ −/− mice can be corrected by adoptive transfer of highly puriﬁed CD8+ CD122+ cells expressing TNF, Perforin, and Granzyme B but not requiring Fas or FasL (116, 117). By contrast, adoptive

ANRV270-IY24-11

Table 3

ARI

15:22

Mutant mouse strains with phenotypes resembling human autoimmune lymphoproliferative syndrome

Mouse strain

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

23 February 2006

Molecular target affected

Lymphoproliferation

Apoptotic defect

Other features, including those suggestive of ALPS

References

lpr/lpr, lprcg

Fas

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation or Fas stimulation

Autoantibodies Hypergammaglobulinemia Glomerulonephritis Sialitis Arthritis

(144–146)

gld/gld

FasL

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation Impaired apoptosis after stimulation with FasL from gld mice

Autoantibodies Hypergammaglobulinemia Glomerulonephritis

(144, 147, 148)

IL-2−/−

Cytokine

Splenomegaly Lymphadenopathy

Impaired Fas-mediated apoptosis or peripheral deletion after SEB Persistence after TCR stimulation Increased c-Flip expression

Autoimmune hemolytic anemia Colitis Hypergammaglobulinemia

(105, 110, 112)

IL-2Rα−/−

IL-2 receptor subunit

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation Impaired peripheral deletion after SEB administration Impaired Fas-mediated apoptosis (in IL-2Rα−/− X DO.11)

Autoimmune hemolytic anemia Colitis Hypergammaglobulinemia

(19, 107, 113)

IL-2Rβ−/−

IL-2/IL-15 receptor subunit

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation

Autoimmune hemolytic anemia Colitis Hypergammaglobulinemia Lack of NK cells

(106)

p65PI3K transgene under T cellspeciﬁc p56 lck promoter

Constitutively active phosphoinositide 3-kinase truncation mutation of p85α regulatory subunit

Splenomegaly Lymphadenopathy

Decreased Annexin V and subG1 staining of expanded CD4 cells

Autoantibodies Hypergammaglobulinemia Glomerulonephritis Lymphoma

(124)

(Continued )

www.annualreviews.org • Genetic Disorders and Cell Death

337

ANRV270-IY24-11

Table 3

ARI

15:22

(Continued )

Mouse strain

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

23 February 2006

Molecular target affected

Lymphoproliferation

Apoptotic defect

Other features, including those suggestive of ALPS

References

PKB transgene under T cellspeciﬁc CD2 or lck promoter

Constitutively active myristolated protein kinase

Splenomegaly Lymphadenopathy

Increased viability after γ irradiation Increased viability after dexamethasone Increased viability in culture Impaired Fas-mediated death

Hypergammaglobulinemia Autoantibodies Glomerulonephritis

(125–127)

PTEN+/− , PTEN−/− , − PTENﬂox/ (in T cells), bPTENﬂox/ﬂox

Phosphatase

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation or Fas stimulation Decreased baseline B cell Annexin V staining Impaired apoptosis after Fas stimulation Impaired apoptosis after γ or UV irradiation Impaired apoptosis after IL-2 (±serum) withdrawal Impaired apoptosis after anti-IgM stimulation (in conditionally deleted mice)

Autoantibodies Glomerulonephritis Hypergammaglobulinemia Heterogeneous tumors including lymphomas

(128–132)

caspase-3−/− reconstituted into RAG2−/−

Effector caspase

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation Impaired Fas-mediated apoptosis

Central nervous system defects in caspase-3−/− mice B cell hyperproliferation due to dysregulated cell cycle in caspase-3−/− mice

(140, 141)

tcaspase-8ﬂ/ﬂ

Initiator caspase

Splenomegaly Lymphadenopathy

Impaired Fas-mediated apoptosis

T cell immunodeﬁciency with impaired NF-κB activation

(4, 99)

Bim−/−

Proapoptotic Bcl-2 family member

Splenomegaly Lymphadenopathy

Impaired apoptosis after IL-2 withdrawal Impaired peripheral deletion after SEB Impaired BCR-induced death

Autoantibodies Glomerulonephritis Hypergammaglobulinemia

(133, 150, 151)

(Continued ) 338

Bid`ere

· · Su

Lenardo

ANRV270-IY24-11

ARI

Table 3

15:22

(Continued )

Mouse strain

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

23 February 2006

Molecular target affected

Lymphoproliferation

Apoptotic defect

Other features, including those suggestive of ALPS

References

Bax−/−

Proapoptotic Bcl-2 family member

Splenomegaly

Impaired apoptosis after anti-IgM stimulation Improved viability of Bax−/− T cells (without mitogens)

Aspermatogenesis Increased atrophic granulosa cells in ovaries

(134, 135)

Bax−/− X Bak−/−

Proapoptotic Bcl-2 family members

Splenomegaly Lymphadenopathy

Impaired death in culture (without mitogens)

Persistence of interdigital webs Imperforate vaginal introitis Accumulation of cells in central nervous system

(136)

Bcl-2 transgene under B cell-speciﬁc immunoglobulin promoter

Anti-apoptotic Bcl-2 family member

Splenomegaly Lymphadenopathy

Increased viability in culture

Hypergammaglobulinemia Autoantibodies Glomerulonephritis Lymphoma

(137, 138)

Mutant cytochrome c K72A/K72A, reconstituted into RAG1−/−

Mitochondrial component of the apoptosome

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation Impaired apoptosis after γ or UV irradiation

Central nervous system defects in cytochrome c K72A/K72A mice

(139)

Stra13−/−

Transcription factor affecting IL-2

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation Decreased FasL expression

Autoantibodies Glomerulonephritis

(152)

T cell-speciﬁc adapter protein (TSAd)−/−

Adaptor-like molecule in T cells, transcriptional regulator affecting IL-2

Splenomegaly Lymphadenopathy

Impaired peripheral deletion after SEB

Autoantibodies Glomerulonephritis Hypergammaglobulinemia

(153)

IRF-4−/−

Transcription factor

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation or peripheral deletion after SEB

T cell functional defect Hypogammaglobulinemia

(154)

(Continued )

www.annualreviews.org • Genetic Disorders and Cell Death

339

ANRV270-IY24-11

Table 3

ARI

15:22

(Continued ) Molecular target affected

Lymphoproliferation

Apoptotic defect

E2F-1−/−

Transcription factor

Splenomegaly Lymphadenopathy

Impaired apoptosis after TCR restimulation

Lymphoma and other tumors Testicular atrophy Exocrine gland dysplasia Anemia Glomerulonephritis

(155–157)

Par4−/−

Atypical protein kinase C inhibitor Inhibits NF-κB activation as well

Splenomegaly

Impaired apoptosis after TCR restimulation

T cell hyperproliferation and Th2 differentiation

(143)

Mutant LAT Y136F/Y136F

Adaptor protein in T cells

Splenomegaly Lymphadenopathy

Impaired Fas-mediated apoptosis Decreased Fas and FasL expression Decreased IL-2 production

Lymphoproliferation differs from the T cell functional defect in LAT−/− mice

(158)

NFAT1−/− X NFAT4−/−

Transcription factors

Splenomegaly Lymphadenopathy (greater than in NFAT1−/− mice)

Impaired apoptosis after TCR restimulation Decreased FasL expression also seen in NFAT1−/− mice or NFAT1−/− X NFAT2−/−

Allergic blepharitis Pneumonitis Hypergammaglobulinemia

(159–161)

Act1−/−

Adaptor molecule

Splenomegaly Lymphadenopathy

Decreased Annexin V staining of B cells after CD40L stimulation

Hypergammaglobulinemia Autoantibodies

(162)

Mouse strain

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

23 February 2006

transfer of highly puriﬁed CD8+ CD122− cells exacerbated disease. Since in vivo antibody depletion of CD4 T cells had previously been shown to reduce autoantibody production and hemolytic anemia (106), these results suggest that CD8+ CD122+ cells constitute a potent regulatory population equipped to kill disease-causing T cells. IL-2Rβ −/− mice lack CD4+ CD25+ cells. Although adoptive transfer of populations enriched in CD4+ CD25+ cells corrected disease in IL-2Rβ −/− mice (118), highly puriﬁed CD4+ CD25+ cells 340

Bid`ere

· · Su

Lenardo

Other features, including those suggestive of ALPS

References

were clearly less capable than CD8+ CD122+ cells (116). Recent studies have shown that CD4+ CD25+ regulatory T cells can exert suppressive effects through Granzyme B or Perforin (8, 9); these studies raise the possibility that these cells may act by killing disease-causing T cells. IL-2−/− mice also lack CD4+ CD25+ cells, and adoptive transfer of wild-type CD4 cells seemed equally or slightly more efﬁcacious than CD8 cells in ameliorating disease (119). However, this study was limited by small sample sizes and large

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

mouse-to-mouse variability, as well as by the failure to separate CD122+ and CD122− cells in the CD8 preparation used for adoptive transfer. Hence, these studies did not refute the importance of IL-2 for induction of apoptosis in T cells. The role of regulatory cells as a potential alternative mode of lymphocyte deletion and tolerance will be an interesting topic for future investigation. These results collectively lead to the prediction that related cytokines using the common gamma chain (γc ) receptor component, which also participates in IL-2 signaling, promote apoptosis in T cells. Although several groups initially reported that other cytokines could not substitute for IL-2, we showed that given sufﬁciently high concentrations to drive T cells into cell cycle, prior exposure of IL2Rα−/− cells to IL-4 or IL-7 restored death after TCR restimulation (19). Other investigators have also shown that IL-2, IL-4, IL7, and IL-15 could promote death of IL-2−/− cells and that, as expected, IL-4 and IL-7 but not IL-2 or IL-15 could promote death of IL-2Rβ −/− cells (IL-2Rβ also is IL-15Rβ and serves as a component of IL-15R) (120). Mice or humans genetically deﬁcient in γc display severe immunodeﬁciency and generally die of infections early in life unless curative bone marrow transplantation is undertaken. However, aging mice do develop splenomegaly with CD4 T cell accumulation, as well as inﬂammatory bowel disease lesions (121). Similarly, there is a case report of an incompletely immunodeﬁcient human patient with a splicing mutation in the γc chain who had peripheral lymphocytes and developed severe colitis (122). Mice deﬁcient in γc were found to have impaired peripheral deletion after SEB administration, with decreased Bcl-2 and FasL expression (although they maintained Fas sensitivity) (123). Interestingly, mice deﬁcient in IL-4, IL-4Rα, IL-7, IL-7Rα, IL-15, or IL-15Rα have not been reported to develop lymphoproliferative or autoimmune disease. Taken together, these ﬁndings underscore the critical importance of IL-2 signaling, in particular, for lymphocyte homeostasis in vivo.

Although studies in ALPS patients have established the physiologic importance of molecules involved in the extrinsic pathway of apoptosis controlled by death receptors, no comparable studies exist in humans for molecules participating in the intrinsic mitochondrial pathway of apoptosis. Mice rendered genetically deﬁcient in many of these candidate molecules have been generated, and several display phenotypes that suggest these genes may contribute to uncharacterized forms of ALPS in humans. For instance, following growth factor receptor stimulation, phosphoinositide-3 kinase (PI3K) generates the lipid second messenger PIP3 , which activates the serine/threonine protein kinase B (PKB)/Akt. PKB/Akt in turn promotes cell survival through multiple effects, including inhibition of translocation of the pro-death Bcl-2 family members Bax and Bak to the mitochondria. Bax and Bak activate the intrinsic apoptosis pathway by permeabilizing the mitochondria and thus releasing cytochrome c to associate with Apaf-1 and caspase-9, which leads to caspase activation. Other Bcl-2 family members such as Bim or Bid can regulate the activity of Bax and Bak. Bid is particularly interesting because it can feed the extrinsic pathway of death into the intrinsic pathway through the ability of caspase-8 to activate Bid. In this general framework, mice transgenic for T cell–speciﬁc, constitutively active mutant forms of PI3K or PKB/Akt developed lymphoproliferation, autoimmune disease, lymphomas, and defective apoptosis (124–127). Phosphatase and tensin homolog (PTEN) limits the available pool of PIP3 and thus has proapoptotic effects. Mice in which PTEN was heterozygously or conditionally deleted in lymphocytes developed a similar ALPS-like phenotype (128–132). Molecules acting further downstream in the intrinsic pathway have been evaluated by rendering mice genetically deﬁcient in various proapoptotic Bcl-2 family members. Mice deﬁcient in Bax or Bim develop lymphoproliferative disease, and, in the case of Bim, autoimmune disease as well (133–135). Impaired www.annualreviews.org • Genetic Disorders and Cell Death

341

ARI

23 February 2006

15:22

lymphocyte apoptosis could be demonstrated when Bax-deﬁcient mice were intercrossed with Bak-deﬁcient mice (136). Transgenic expression of the anti-apoptotic Bcl-2 in B cells also caused an ALPS-like disease (137, 138). For genes that cause embryonic lethality, a role in mature lymphocyte homeostasis was assessed by reconstitution into recombinase activating gene (RAG)-deﬁcient mice, which lack endogenous T and B cells. Using this approach, adoptive transfer of mutant cytochrome c knockin cells gave rise to lymphoproliferative disease and impaired apoptosis following TCR restimulation, gamma-, or UV-irradiation (139). Adoptive transfer of caspase-3-deﬁcient cells also resulted in lymphoproliferative disease and impaired apoptosis to TCR restimulation or Fas (140, 141), as would be expected for an effector caspase that participates in both the intrinsic and extrinsic pathways of apoptosis. Finally, other knockout mice have been generated in which the role of certain transcription factors and other regulators for programmed cell death have been less well characterized. Based on the discussion above, these proteins may regulate IL-2 or other various proapoptotic and anti-apoptotic molecules functioning in extrinsic or intrinsic apoptosis pathways. In this light, it is interesting that signaling through the Fas death receptor or through TCR results in NF-κB activation, which normally antagonizes simultaneous death effects (142). This mechanism may contribute to lymphomagenesis in ALPS patients (51). Thus, gene mutations that affect NF-κB activation can perturb lymphocyte homeostasis through their effects on apoptosis. For example, mice rendered genetically deﬁcient in Par-4, an atypical PKC inhibitor that normally inhibits NF-κB activation, exhibit hyperproliferation (143). In summary, mice rendered genetically deﬁcient by homologous recombination have been useful in identifying potential genes that govern lymphocyte homeostasis and tol-

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

342

Bid`ere

· · Su

Lenardo

erance. In particular, genes involved in IL2 signaling, genes involved in the intrinsic and extrinsic pathways for apoptosis, and certain transcription factors, including those that regulate NF-κB activation, are likely to be responsible for some cases of ALPS type III.

CONCLUSION The study of humans with rare genetic defects has been made possible by new technological breakthroughs that have enabled unprecedented molecular manipulation and rigorous scientiﬁc analyses. These studies are complemented by knowledge of disease pathogenesis gleaned from medical studies of individual patients and patient populations. Two such examples of this powerful new approach are discussed in this review. Studies in patients with ALPS and CEDS have shed considerable light into the normal signal transduction events that lead to programmed cell death, activation, and lymphocyte homeostasis in humans. These studies have led to a number of insights not previously appreciated from work done in model organisms. ALPS patients show that the consequences of defective apoptosis are lymphoproliferative disease, autoimmunity, and lymphoma. Although identiﬁcation of genetic disease modiﬁers remains unresolved, CD4− CD8− αβ T cells clearly play an important role for this disease pathogenesis. An understanding of the biochemical events leading to apoptosis explains how dominant mutations cause disease in most ALPS patients. At the molecular level, we now know that Fas death receptor preassociation and its higher order interactions are essential for apoptosis induction. Caspase-10, which is found in humans but not in mice, is also a critical component. By contrast, its related molecule, caspase-8, though contributing to death, serves a more important role in the activation of lymphocytes through antigen receptors and other immunoreceptors

ANRV270-IY24-11

ARI

23 February 2006

15:22

through its effects on NF-κB. The study of other ALPS patients with as yet unidentiﬁed genetic defects and patients with defects in other pathways of programmed cell

death may prove to be equally illuminating for understanding the molecular underpinnings of this basic biological process in humans.

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ACKNOWLEDGMENTS We thank our long-term collaborators Stephen Straus, Janet Dale, Jennifer Puck, Joie Davis, Elaine Jaffe, Thomas Fleisher, Faith Dugan, Koneti Rao, and Roxanne Fischer. We thank Richard Siegel for helpful suggestions and discussions and Lixin Zheng, Joao Bosco Oliveira, and Keiko Sakai for critically reading the manuscript. We apologize to many investigators whose important work was not cited owing to editorial constraints. This work was supported by the Intramural Research Program of the NIH, NIAID, as well as the Cancer Research Institute (HS).

LITERATURE CITED 1. Lenardo MJ. 2003. Molecular regulation of T lymphocyte homeostasis in the healthy and diseased immune system. Immunol. Res. 27:387–98 2. Chun HJ, Zheng LX, Ahmad M, Wang J, Speirs CK, et al. 2002. Pleiotropic defects in lymphocyte activation caused by caspase-8 mutations lead to human immunodeﬁciency. Nature 419:395–99 3. Wang J, Zheng L, Lobito A, Chan FK, Dale J, et al. 1999. Inherited human caspase 10 mutations underlie defective lymphocyte and dendritic cell apoptosis in autoimmune lymphoproliferative syndrome type II. Cell 98:47–58 4. Su H, Bid`ere N, Zheng LX, Cubre A, Sakai K, et al. 2005. Requirement for caspase-8 in NF-κB activation by antigen receptor. Science 307:1465–68 5. Lenardo M, Chan FK-M, Hornung F, McFarland H, Siegel R, et al. 1999. Mature T lymphocyte apoptosis–immune regulation in a dynamic and unpredictable antigenic environment. Annu. Rev. Immunol. 17:221–53 6. Lenardo MJ. 1996. Fas and the art of lymphocyte maintenance. J. Exp. Med. 183:721–24 7. Strasser A, Pellegrini M. 2004. T-lymphocyte death during shutdown of an immune response. Trends Immunol. 25:610–15 8. Grossman WJ, Verbsky JW, Barchet W, Colonna M, Atkinson JP, et al. 2004. Human T regulatory cells can use the perforin pathway to cause autologous target cell death. Immunity 21:589–601 9. Gondek DC, Lu LF, Quezada SA, Sakaguchi S, Noelle RJ. 2005. Cutting edge: contactmediated suppression by CD4+ CD25+ regulatory cells involves a granzyme B-dependent, perforin-independent mechanism. J. Immunol. 174:1783–86 10. Brunner T, Mogil RJ, LaFace D, Yoo NJ, Mahboubi A, et al. 1995. Cell-autonomous Fas (CD95)/Fas-ligand interaction mediates activation-induced apoptosis in T-cell hybridomas. Nature 373:441–44 11. Dhein J, Walczak H, Baumler C, Debatin KM, Krammer PH. 1995. Autocrine T-cell suicide mediated by APO-1/(Fas/CD95). Nature 373:438–41 12. Ju ST, Panka DJ, Cui H, Ettinger R, el-Khatib M, et al. 1995. Fas(CD95)/FasL interactions required for programmed cell death after T-cell activation. Nature 373:444–48 13. Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, et al. 1995. Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377:348–51 www.annualreviews.org • Genetic Disorders and Cell Death

343

ARI

23 February 2006

15:22

14. Muppidi JR, Siegel RM. 2004. Ligand-independent redistribution of Fas (CD95) into lipid rafts mediates clonotypic T cell death. Nat. Immunol. 5:182–89 15. Chan FK-M, Shisler J, Bixby JG, Felices M, Zheng LX, Lenardo MJ. 2003. A role for tumor necrosis factor receptor-2 and receptor-interacting protein in programmed necrosis and antiviral responses. J. Biol. Chem. 278:51613–21 16. Holler N, Zaru R, Micheau O, Thome M, Attinger A, et al. 2000. Fas triggers an alternative, caspase-8-independent cell death pathway using the kinase RIP as effector molecule. Nat. Immunol. 1:489–95 17. Locksley RM, Killeen N, Lenardo MJ. 2001. The TNF and TNF receptor superfamilies: integrating mammalian biology. Cell 104:487–501 18. Suda T, Takahashi T, Golstein P, Nagata S. 1993. Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family. Cell 75:1169–78 19. Zheng L, Trageser CL, Willerford DM, Lenardo MJ. 1998. T cell growth cytokines cause the superinduction of molecules mediating antigen-induced T lymphocyte death. J. Immunol. 160:763–69 20. Wang J, Lenardo MJ. 1997. Essential lymphocyte function associated 1 (LFA-1): intercellular adhesion molecule interactions for T cell-mediated B cell apoptosis by Fas/APO1/CD95. J. Exp. Med. 186:1171–76 21. Kischkel FC, Hellbardt S, Behrmann I, Germer M, Pawlita M, et al. 1995. Cytotoxicitydependent APO-1 (Fas/CD95)-associated proteins form a death-inducing signaling complex (DISC) with the receptor. EMBO J. 14:5579–88 22. Fesik SW. 2000. Insights into programmed cell death through structural biology. Cell 103:273–82 23. Tibbetts MD, Zheng LX, Lenardo MJ. 2003. The death effector domain protein family: regulators of cellular homeostasis. Nat. Immunol. 4:404–9 24. Zhang J, Cado D, Chen A, Kabra NH, Winoto A. 1998. Fas-mediated apoptosis and activation-induced T-cell proliferation are defective in mice lacking FADD/Mort1. Nature 392:296–300 25. Nicholson DW, Thornberry NA. 1997. Caspases: killer proteases. Trends Biochem. Sci. 22:299–306 26. Boatright KM, Salvesen GS. 2003. Mechanisms of caspase activation. Curr. Opin. Cell Biol. 15:725–31 27. Jiang X, Wang X. 2004. Cytochrome C-mediated apoptosis. Annu. Rev. Biochem. 73:87– 106 28. Krueger A, Fas SC, Baumann S, Krammer PH. 2003. The role of CD95 in the regulation of peripheral T-cell apoptosis. Immunol. Rev. 193:58–69 29. Algeciras-Schimnich A, Shen L, Barnhart BC, Murmann AE, Burkhardt JK, et al. 2002. Molecular ordering of the initial signaling events of CD95. Mol. Cell. Biol. 22:207–20 30. Siegel RM, Muppidi JR, Sarker M, Lobito A, Jen M, et al. 2004. SPOTS: signaling protein oligomeric transduction structures are early mediators of death receptor-induced apoptosis at the plasma membrane. J. Cell Biol. 167:735–44 31. Schneider-Brachert W, Tchikov V, Neumeyer J, Jakob M, Winoto-Morbach S, et al. 2004. Compartmentalization of TNF receptor 1 signaling: internalized TNF receptosomes as death signaling vesicles. Immunity 21:415–28 32. Thome M, Tschopp J. 2001. Regulation of lymphocyte proliferation and death by FLIP. Nat. Rev. Immunol. 1:50–58 33. Fernandes-Alnemri T, Armstrong RC, Krebs J, Srinivasula SM, Wang L, et al. 1996. In vitro activation of CPP32 and Mch3 by Mch4, a novel human apoptotic cysteine

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

344

Bid`ere

· · Su

Lenardo

ANRV270-IY24-11

34. 35.

36.

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

37.

38. 39. 40.

41. 42.

43.

44.

45.

46.

47.

48.

49.

ARI

23 February 2006

15:22

protease containing two FADD-like domains. Proc. Natl. Acad. Sci. USA 93:7464– 69 Irmler M, Thome M, Hahne M, Schneider P, Hofmann K, et al. 1997. Inhibition of death receptor signals by cellular FLIP. Nature 388:190–95 Rasper DM, Vaillancourt JP, Hadano S, Houtzager VM, Seiden I, et al. 1998. Cell death attenuation by ‘Usurpin,’ a mammalian DED-caspase homologue that precludes caspase8 recruitment and activation by the CD-95 (Fas, APO-1) receptor complex. Cell Death Differ. 5:271–88 Krueger A, Schmitz I, Baumann S, Krammer PH, Kirchhoff S. 2001. Cellular FLICEinhibitory protein splice variants inhibit different steps of caspase-8 activation at the CD95 death-inducing signaling complex. J. Biol. Chem. 276:20633–40 Wang J, Lobito AA, Shen F, Hornung F, Winoto A, Lenardo MJ. 2000. Inhibition of Fas-mediated apoptosis by the B cell antigen receptor through c-FLIP. Eur. J. Immunol. 30:155–63 Rieux-Laucat F, Fischer A, Deist FL. 2003. Cell-death signaling and human disease. Curr. Opin. Immunol. 15:325–31 Sneller MC, Dale JK, Straus SE. 2003. Autoimmune lymphoproliferative syndrome. Curr. Opin. Immunol. 15:417–21 Bleesing JJ, Brown MR, Dale JK, Straus SE, Lenardo MJ, et al. 2001. TcR-α/β+ CD4− CD8− T cells in humans with the autoimmune lymphoproliferative syndrome express a novel CD45 isoform that is analogous to murine B220 and represents a marker of altered O-glycan biosynthesis. Clin. Immunol. 100:314–24 Zuniga-Pﬂucker JC, Lenardo MJ. 1996. Regulation of thymocyte development from immature progenitors. Curr. Opin. Immunol. 8:215–24 Bleesing JJ, Brown MR, Novicio C, Guarraia D, Dale JK, et al. 2002. A composite picture of TcR α/β+ CD4− CD8− T cells (α/β-DNTCs) in humans with autoimmune lymphoproliferative syndrome. Clin. Immunol. 104:21–30 Holzelova E, Vonarbourg C, Stolzenberg MC, Arkwright PD, Selz F, et al. 2004. Autoimmune lymphoproliferative syndrome with somatic Fas mutations. N. Engl. J. Med. 351:1409–18 Lopatin U, Yao X, Williams RK, Bleesing JJ, Dale JK, et al. 2001. Increases in circulating and lymphoid tissue interleukin-10 in autoimmune lymphoproliferative syndrome are associated with disease expression. Blood 97:3161–70 Fuss IJ, Strober W, Dale JK, Fritz S, Pearlstein GR, et al. 1997. Characteristic T helper 2 T cell cytokine abnormalities in autoimmune lymphoproliferative syndrome, a syndrome marked by defective apoptosis and humoral autoimmunity. J. Immunol. 158:1912–18 Sneller MC, Straus SE, Jaffe ES, Jaffe JS, Fleisher TA, et al. 1992. A novel lymphoproliferative/autoimmune syndrome resembling murine lpr/gld disease. J. Clin. Invest. 90:334–41 Teachey DT, Manno CS, Axsom KM, Andrews T, Choi JK, et al. 2004. Unmasking Evans Syndrome: T cell phenotype and apoptotic response reveal autoimmune lymphoproliferative syndrome (ALPS). Blood 105:2443–48 Wong CS, Arkwright PD, Rieux-Laucat F, Cant AJ, Stevens RF, et al. 2004. Childhood linear IgA disease in association with autoimmune lymphoproliferative syndrome. Br. J. Dermatol. 150:578–80 Wu J, Wilson J, He J, Xiang L, Schur PH, Mountz JD. 1996. Fas ligand mutation in a patient with systemic lupus erythematosus and lymphoproliferative disease. J. Clin. Invest. 98:1107–13 www.annualreviews.org • Genetic Disorders and Cell Death

345

ARI

23 February 2006

15:22

50. Straus SE, Jaffe ES, Puck JM, Dale JK, Elkon KB, et al. 2001. The development of lymphomas in families with autoimmune lymphoproliferative syndrome with germline Fas mutations and defective lymphocyte apoptosis. Blood 98:194–200 51. Legembre P, Barnhart BC, Zheng LX, Vijayan S, Straus SE, et al. 2004. Induction of apoptosis and activation of NF-κB by CD95 require different signalling thresholds. EMBO Rep. 5:1084–89 52. Davidson WF, Giese T, Fredrickson TN. 1998. Spontaneous development of plasmacytoid tumors in mice with defective Fas-Fas ligand interactions. J. Exp. Med. 187:1825–38 53. Cohen PL, Eisenberg RA. 1991. Lpr and gld: single gene models of systemic autoimmunity and lymphoproliferative disease. Annu. Rev. Immunol. 9:243–69 54. Kimura M, Matsuzawa A. 1994. Autoimmunity in mice bearing lprcg: a novel mutant gene. Int. Rev. Immunol. 11:193–210 55. Nagata S. 1998. Human autoimmune lymphoproliferative syndrome, a defect in the apoptosis-inducing Fas receptor: a lesson from the mouse model. J. Hum. Genet. 43:2– 8 56. Nose M, Nishihara M, Kamogawa J, Terada M, Nakatsuru S. 2000. Genetic basis of autoimmune disease in MRL/lpr mice: dissection of the complex pathological manifestations and their susceptibility loci. Rev. Immunogenet. 2:154–64 57. Tsuneyama K, Nose M, Nisihara M, Katayanagi K, Harada K, et al. 2001. Spontaneous occurrence of chronic non-suppurative destructive cholangitis and antimitochondrial autoantibodies in MRL/lpr mice: possible animal model for primary biliary cirrhosis. Pathol. Int. 51:418–24 58. Kamogawa J, Terada M, Mizuki S, Nishihara M, Yamamoto H, et al. 2002. Arthritis in MRL/lpr mice is under the control of multiple gene loci with an allelic combination derived from the original inbred strains. Arthritis Rheum. 46:1067–74 59. Miyazaki T, Ono M, Qu WM, Zhang MC, Mori S, et al. 2005. Implication of allelic polymorphism of osteopontin in the development of lupus nephritis in MRL/lpr mice. Eur. J. Immunol. 35:1510–20 60. Chiocchetti A, Indelicato M, Bensi T, Mesturini R, Giordano M, et al. 2004. High levels of osteopontin associated with polymorphisms in its gene are a risk factor for development of autoimmunity/lymphoproliferation. Blood 103:1376–82 61. Hang L, Theoﬁlopoulos AN, Balderas RS, Francis SJ, Dixon FJ. 1984. The effect of thymectomy on lupus-prone mice. J. Immunol. 132:1809–13 62. Wofsy D, Ledbetter JA, Roubinian JR, Seaman WE, Talal N. 1982. Thymic inﬂuences on autoimmunity in MRL-lpr mice. Scand. J. Immunol. 16:51–58 63. van der Werff Ten Bosch J, Schotte P, Ferster A, Azzi N, Boehler T, et al. 2002. Reversion of autoimmune lymphoproliferative syndrome with an antimalarial drug: preliminary results of a clinical cohort study and molecular observations. Br. J. Dermatol. 117:176–88 64. Benkerrou M, Le Deist F, de Villartay JP, Caillat-Zucman S, Rieux-Laucat F, et al. 1997. Correction of Fas (CD95) deﬁciency by haploidentical bone marrow transplantation. Eur. J. Immunol. 27:2043–47 65. Hundt M, Posovszky C, Schmidt RE. 2002. A new disorder of lymphocyte apoptosis: combination of autoimmunity, infectious lymphadenopathy, double negative T cells, and impaired activation-induced cell death. Immunobiology 206:514–18 66. Ramenghi U, Bonissoni S, Migliaretti G, DeFranco S, Bottarel F, et al. 2000. Deﬁciency of the Fas apoptosis pathway without Fas gene mutations is a familial trait predisposing to development of autoimmune diseases and cancer. Blood 95:3176–82

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

346

Bid`ere

· · Su

Lenardo

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

67. Chan FK-M, Siegel RM, Lenardo MJ. 2000. Signaling by the TNF receptor superfamily and T cell homeostasis. Immunity 13:419–22 68. Siegel RM, Chan FK-M, Chun HJ, Lenardo MJ. 2000. The multifaceted role of Fas signaling in immune cell homeostasis and autoimmunity. Nat. Immunol. 1:469–74 69. Roesler J, Izquierdo JM, Ryser M, Rosen-Wolff A, Gahr M, et al. 2005. Haploinsufﬁciency rather than the effect of an excessive production of soluble CD95 (CD95 Delta TM) is the basis for ALPS Ia in a family with duplicated 3 splice site AG in CD95 intron 5 on one allele. Blood 106:1652–59 70. Rieux-Laucat F, Le Deist F, Fischer A. 2003. Autoimmune lymphoproliferative syndromes: genetic defects of apoptosis pathways. Cell Death Differ. 10:124–33 71. Gronbaek K, Dalby T, Zeuthen J, Ralfkiaer E, Guidberg P. 2000. The V410I (G1228A) variant of the caspase-10 gene is a common polymorphism of the Danish population. Blood 95:2184–85 72. Fisher GH, Rosenberg FJ, Straus SE, Dale JK, Middleton LA, et al. 1995. Dominant interfering Fas gene mutations impair apoptosis in a human autoimmune lymphoproliferative syndrome. Cell 81:935–46 73. Herskowitz I. 1987. Functional inactivation of genes by dominant negative mutations. Nature 329:219–22 74. Martin DA, Zheng L, Siegel RM, Huang B, Fisher GH, et al. 1999. Defective CD95/APO-1/Fas signal complex formation in the human autoimmune lymphoproliferative syndrome, type Ia. Proc. Natl. Acad. Sci. USA 96:4552–57 75. Siegel RM, Frederiksen JK, Zacharias DA, Chan FK-M, Johnson M, et al. 2000. Fas preassociation required for apoptosis signaling and dominant inhibition by pathogenic mutations. Science 288:2354–57 76. Chan FK-M, Chun HJ, Zheng LX, Siegel RM, Bui KL, Lenardo MJ. 2000. A domain in TNF receptors that mediates ligand-independent receptor assembly and signaling. Science 288:2351–54 77. Remy I, Wilson IA, Michnick SW. 1999. Erythropoietin receptor activation by a ligandinduced conformation change. Science 283:990–93 78. Holler N, Tardivel A, Kovacsovics-Bankowski M, Hertig S, Gaide O, et al. 2003. Two adjacent trimeric Fas ligands are required for Fas signaling and formation of a deathinducing signaling complex. Mol. Cell. Biol. 23:1428–40 79. Kischkel FC, Lawrence DA, Tinel A, LeBlanc H, Virmani A, et al. 2001. Death receptor recruitment of endogenous caspase-10 and apoptosis initiation in the absence of caspase8. J. Biol. Chem. 276:46639–46 80. Wang J, Chun HJ, Wong W, Spencer DM, Lenardo MJ. 2001. Caspase-10 is an initiator caspase in death receptor signaling. Proc. Natl. Acad. Sci. USA 98:13884–88 81. Siegel RM, Chan FK-M, Zacharias DA, Swofford R, Holmes KL, et al. 2000. Measurement of molecular interactions in living cells by ﬂuorescence resonance energy transfer between variants of the green ﬂuorescent protein. Science STKE 2000:PL1 82. Sprick MR, Rieser E, Stahl H, Grosse-Wilde A, Weigand MA, et al. 2002. Caspase-10 is recruited to and activated at the native TRAIL and CD95 death-inducing signalling complexes in a FADD-dependent manner but can not functionally substitute caspase-8. EMBO J. 21:4520–30 83. Park WS, Lee JH, Shin MS, Park JY, Kim HS, et al. 2002. Inactivating mutations of the caspase-10 gene in gastric cancer. Oncogene 21:2919–25 84. Shin MS, Kim HS, Kang CS, Park WS, Kim SY, et al. 2002. Inactivating mutations of CASP10 gene in non-Hodgkin lymphomas. Blood 99:4094–99 www.annualreviews.org • Genetic Disorders and Cell Death

347

ARI

23 February 2006

15:22

85. Lens SM, Kataoka T, Fortner KA, Tinel A, Ferrero I, et al. 2002. The caspase 8 inhibitor c-FLIP(L) modulates T-cell receptor-induced proliferation but not activation-induced cell death of lymphocytes. Mol. Cell. Biol. 22:5419–33 86. Stanger BZ, Leder P, Lee TH, Kim E, Seed B. 1995. RIP: a novel protein containing a death domain that interacts with Fas/APO-1 (CD95) in yeast and causes cell death. Cell 81:513–23 87. Rieux-Laucat F, Le Deist F, Hivroz C, Roberts IA, Debatin KM, et al. 1995. Mutations in Fas associated with human lymphoproliferative syndrome and autoimmunity. Science 268:1347–49 88. Jackson CE, Fischer RE, Hsu AP, Anderson SM, Choi Y, et al. 1999. Autoimmune lymphoproliferative syndrome with defective Fas: genotype inﬂuences penetrance. Am. J. Hum. Genet. 64:1002–14 89. Choi Y, Ramnath VR, Eaton AS, Chen A, Simon-Stoos KL, et al. 1999. Expression in transgenic mice of dominant interfering Fas mutations: a model for human autoimmune lymphoproliferative syndrome. Clin. Immunol. 93:34–45 90. Klein J, Sato A. 2000. The HLA system. Second of two parts. N. Engl. J. Med. 343:782–86 91. Purrmann J, Bertrams J, Knapp M, Cleveland S, Hengels KJ, et al. 1990. Gene and haplotype frequencies of HLA antigens in 269 patients with Crohn’s disease. Scand. J. Gastroenterol. 25:981–85 92. Clementi R, Dagna L, Dianzani U, Dupre L, Dianzani I, et al. 2004. Inherited perforin and Fas mutations in a patient with autoimmune lymphoproliferative syndrome and lymphoma. N. Engl. J. Med. 351:1419–24 93. Rieux-Laucat F, Le Deist F, de Saint Basile G. 2005. Autoimmune lymphoproliferative syndrome and perforin. N. Engl. J. Med. 352:306 94. Ueda H, Howson JM, Esposito L, Heward J, Snook H, et al. 2003. Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423:506–11 95. Hull KM, Drewe E, Aksentijevich I, Singh HK, Wong K, et al. 2002. The TNF receptorassociated periodic syndrome (TRAPS): emerging concepts of an autoinﬂammatory disorder. Medicine 81:349–68 96. Touitou I. 2001. The spectrum of Familial Mediterranean Fever (FMF) mutations. Eur. J. Hum. Genet. 9:473–83 97. Varfolomeev EE, Schuchmann M, Luria V, Chiannilkulchai N, Beckmann JS, et al. 1998. Targeted disruption of the mouse caspase 8 gene ablates cell death induction by the TNF receptors, Fas/Apo1, and DR3 and is lethal prenatally. Immunity 9:267–76 98. Kennedy NJ, Kataoka T, Tschopp J, Budd RC. 1999. Caspase activation is required for T cell proliferation. J. Exp. Med. 190:1891–96 99. Salmena L, Lemmers B, Hakem A, Matysiak-Zablocki E, Murakami K, et al. 2003. Essential role for caspase 8 in T-cell homeostasis and T-cell-mediated immunity. Genes Dev. 17:883–95 100. Li Q, Verma IM. 2002. NF-κB regulation in the immune system. Nat. Rev. Immunol. 2:725–34 101. Karin M, Delhase M. 2000. The IκB kinase (IKK) and NF-κB: key elements of proinﬂammatory signalling. Semin. Immunol. 12:85–98 102. Thome M. 2004. CARMA1, BCL-10 and MALT1 in lymphocyte development and activation. Nat. Rev. Immunol. 4:348–59 103. Lee KY, D’Acquisto F, Hayden MS, Shim JH, Ghosh S. 2005. PDK1 nucleates T cell receptor-induced signaling complex for NF-κB activation. Science 308:114–18

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

348

Bid`ere

· · Su

Lenardo

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

104. Newton K, Harris AW, Bath ML, Smith KG, Strasser A. 1998. A dominant interfering mutant of FADD/MORT1 enhances deletion of autoreactive thymocytes and inhibits proliferation of mature T lymphocytes. EMBO J. 17:706–18 105. Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, et al. 1993. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75:253–61 106. Suzuki H, Kundig TM, Furlonger C, Wakeham A, Timms E, et al. 1995. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor β. Science 268:1472–76 107. Willerford DM, Chen J, Ferry JA, Davidson L, Ma A, et al. 1995. Interleukin-2 receptor alpha chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3:521–30 108. Sharfe N, Dadi HK, Shahar M, Roifman CM. 1997. Human immune disorder arising from mutation of the alpha chain of the interleukin-2 receptor. Proc. Natl. Acad. Sci. USA 94:3168–71 109. Lenardo MJ. 1991. Interleukin-2 programs mouse αβ T lymphocytes for apoptosis. Nature 353:858–61 110. Khoruts A, Mondino A, Pape KA, Reiner SL, Jenkins MK. 1998. A natural immunological adjuvant enhances T cell clonal expansion through a CD28-dependent, interleukin (IL)2-independent mechanism. J. Exp. Med. 187:225–36 111. Deleted in proof 112. Kneitz B, Herrmann T, Yonehara S, Schimpl A. 1995. Normal clonal expansion but impaired Fas-mediated cell death and anergy induction in interleukin-2-deﬁcient mice. Eur. J. Immunol. 25:2572–77 113. Van Parijs L, Biuckians A, Ibragimov A, Alt FW, Willerford DM, et al. 1997. Functional responses and apoptosis of CD25 (IL-2R α)-deﬁcient T cells expressing a transgenic antigen receptor. J. Immunol. 158:3738–45 114. Suzuki H, Hayakawa A, Bouchard D, Nakashima I, Mak TW, et al. 1997. Normal thymic selection, superantigen-induced deletion and Fas-mediated apoptosis of T cells in IL-2 receptor beta chain-deﬁcient mice. Int. Immunol. 9:1367–74 115. Kovanen PE, Young L, Al-Shami A, Rovella V, Pise-Masison CA, et al. 2005. Global analysis of IL-2 target genes: identiﬁcation of chromosomal clusters of expressed genes. Int. Immunol. 17:1009–21 116. Rifa’i M, Kawamoto Y, Nakashima I, Suzuki H. 2004. Essential roles of CD8+CD122+ regulatory T cells in the maintenance of T cell homeostasis. J. Exp. Med. 200:1123–34 117. Suzuki H, Zhou YW, Kato M, Mak TW, Nakashima I. 1999. Normal regulatory α/β T cells effectively eliminate abnormally activated T cells lacking the interleukin 2 receptor β in vivo. J. Exp. Med. 190:1561–72 118. Malek TR, Yu A, Vincek V, Scibelli P, Kong L. 2002. CD4 regulatory T cells prevent lethal autoimmunity in IL-2Rβ-deﬁcient mice. Implications for the nonredundant function of IL-2. Immunity 17:167–78 119. Wolf M, Schimpl A, Hunig T. 2001. Control of T cell hyperactivation in IL-2-deﬁcient mice by CD4+ CD25− and CD4+ CD25+ T cells: evidence for two distinct regulatory mechanisms. Eur. J. Immunol. 31:1637–45 120. Kung JT, Beller D, Ju ST. 1998. Lymphokine regulation of activation-induced apoptosis in T cells of IL-2 and IL-2R β knockout mice. Cell. Immunol. 185:158–63 121. Cao X, Shores EW, Hu-Li J, Anver MR, Kelsall BL, et al. 1995. Defective lymphoid development in mice lacking expression of the common cytokine receptor gamma chain. Immunity 2:223–38 www.annualreviews.org • Genetic Disorders and Cell Death

349

ARI

23 February 2006

15:22

122. DiSanto JP, Rieux-Laucat F, Dautry-Varsat A, Fischer A, de Saint Basile G, et al. 1994. Defective human interleukin 2 receptor gamma chain in an atypical X chromosomelinked severe combined immunodeﬁciency with peripheral T cells. Proc. Natl. Acad. Sci. USA 91:9466–70 123. Nakajima H, Leonard WJ. 1997. Impaired peripheral deletion of activated T cells in mice lacking the common cytokine receptor gamma-chain: defective Fas ligand expression in gamma-chain-deﬁcient mice. J. Immunol. 159:4737–44 124. Borlado LR, Redondo C, Alvarez B, Jimenez C, Criado LM, et al. 2000. Increased phosphoinositide 3-kinase activity induces a lymphoproliferative disorder and contributes to tumor generation in vivo. FASEB J. 14:895–903 125. Jones RG, Parsons M, Bonnard M, Chan VS, Yeh WC, et al. 2000. Protein kinase B regulates T lymphocyte survival, nuclear factor κB activation, and Bcl-X(L) levels in vivo. J. Exp. Med. 191:1721–34 126. Parsons MJ, Jones RG, Tsao MS, Odermatt B, Ohashi PS, et al. 2001. Expression of active protein kinase B in T cells perturbs both T and B cell homeostasis and promotes inﬂammation. J. Immunol. 167:42–88 127. Rathmell JC, Elstrom RL, Cinalli RM, Thompson CB. 2003. Activated Akt promotes increased resting T cell size, CD28-independent T cell growth, and development of autoimmunity and lymphoma. Eur. J. Immunol. 33:2223–32 128. Di Cristofano A, Kotsi P, Peng YF, Cordon-Cardo C, Elkon KB, Pandolﬁ PP. 1999. Impaired Fas response and autoimmunity in Pten+ -mice. Science 285:2122–25 129. Podsypanina K, Ellenson LH, Nemes A, Gu J, Tamura M, et al. 1999. Mutation of Pten/Mmac1 in mice causes neoplasia in multiple organ systems. Proc. Natl. Acad. Sci. USA 96:1563–68 130. Suzuki A, de la Pompa JL, Stambolic V, Elia AJ, Sasaki T, et al. 1998. High cancer susceptibility and embryonic lethality associated with mutation of the PTEN tumor suppressor gene in mice. Curr. Biol. 8:1169–78 131. Suzuki A, Kaisho T, Ohishi M, Tsukio-Yamaguchi M, Tsubata T, et al. 2003. Critical roles of Pten in B cell homeostasis and immunoglobulin class switch recombination. J. Exp. Med. 197:657–67 132. Suzuki A, Yamaguchi MT, Ohteki T, Sasaki T, Kaisho T, et al. 2001. T cell-speciﬁc loss of Pten leads to defects in central and peripheral tolerance. Immunity 14:523–34 133. Bouillet P, Metcalf D, Huang DC, Tarlinton DM, Kay TW, et al. 1999. Proapoptotic Bcl-2 relative Bim required for certain apoptotic responses, leukocyte homeostasis, and to preclude autoimmunity. Science 286:1735–38 134. Knudson CM, Korsmeyer SJ. 1997. Bcl-2 and Bax function independently to regulate cell death. Nat. Genet. 16:358–63 135. Knudson CM, Tung KS, Tourtellotte WG, Brown GA, Korsmeyer SJ. 1995. Bax-deﬁcient mice with lymphoid hyperplasia and male germ cell death. Science 270:96–99 136. Lindsten T, Ross AJ, King A, Zong WX, Rathmell JC, et al. 2000. The combined functions of proapoptotic Bcl-2 family members bak and bax are essential for normal development of multiple tissues. Mol. Cell 6:1389–99 137. Strasser A, Harris AW, Cory S. 1993. E mu-bcl-2 transgene facilitates spontaneous transformation of early pre-B and immunoglobulin-secreting cells but not T cells. Oncogene 8:1–9 138. Strasser A, Whittingham S, Vaux DL, Bath ML, Adams JM, et al. 1991. Enforced BCL2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease. Proc. Natl. Acad. Sci. USA 88:8661–65

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

350

Bid`ere

· · Su

Lenardo

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-11

ARI

23 February 2006

15:22

139. Hao Z, Duncan GS, Chang CC, Elia A, Fang M, et al. 2005. Speciﬁc ablation of the apoptotic functions of cytochrome C reveals a differential requirement for cytochrome C and Apaf-1 in apoptosis. Cell 121:579–91 140. Woo M, Hakem R, Furlonger C, Hakem A, Duncan GS, et al. 2003. Caspase-3 regulates cell cycle in B cells: a consequence of substrate speciﬁcity. Nat. Immunol. 4:1016–22 141. Woo M, Hakem R, Soengas MS, Duncan GS, Shahinian A, et al. 1998. Essential contribution of caspase 3/CPP32 to apoptosis and its associated nuclear changes. Genes Dev. 12:806–19 142. Dudley E, Hornung F, Zheng L, Scherer D, Ballard D, Lenardo M. 1999. NF-κB regulates Fas/APO-1/CD95- and TCR-mediated apoptosis of T lymphocytes. Eur. J. Immunol. 29:878–86 143. Lafuente MJ, Martin P, Garcia-Cao I, Diaz-Meco MT, Serrano M, Moscat J. 2003. Regulation of mature T lymphocyte proliferation and differentiation by Par-4. EMBO J. 22:4689–98 144. Gillette-Ferguson I, Sidman CL. 1994. A speciﬁc intercellular pathway of apoptotic cell death is defective in the mature peripheral T cells of autoimmune lpr and gld mice. Eur. J. Immunol. 24:1181–85 145. Matsuzawa A, Moriyama T, Kaneko T, Tanaka M, Kimura M, et al. 1990. A new allele of the lpr locus, lprcg, that complements the gld gene in induction of lymphadenopathy in the mouse. J. Exp. Med. 171:519–31 146. Watanabe-Fukunaga R, Brannan CI, Copeland NG, Jenkins NA, Nagata S, et al. 1992. Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356:314–17 147. Roths JB, Murphy ED, Eicher EM. 1984. A new mutation, gld, that produces lymphoproliferation and autoimmunity in C3H/HeJ mice. J. Exp. Med. 159:1–20 148. Takahashi T, Tanaka M, Brannan CI, Jenkins NA, Copeland NG, et al. 1994. Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76:969–76 149. Deleted in proof 150. Enders A, Bouillet P, Puthalakath H, Xu Y, Tarlinton DM, Strasser A. 2003. Loss of the pro-apoptotic BH3-only Bcl-2 family member Bim inhibits BCR stimulation-induced apoptosis and deletion of autoreactive B cells. J. Exp. Med. 198:1119–26 151. Hildeman DA, Zhu Y, Mitchell TC, Bouillet P, Strasser A, et al. 2002. Activated T cell death in vivo mediated by proapoptotic bcl-2 family member bim. Immunity 16:759– 67 152. Sun H, Lu B, Li RQ, Flavell RA, Taneja R. 2001. Defective T cell activation and autoimmune disorder in Stra13-deﬁcient mice. Nat. Immunol. 2:1040–47 153. Drappa J, Kamen LA, Chan E, Georgiev M, Ashany D, et al. 2003. Impaired T cell death and lupus-like autoimmunity in T cell-speciﬁc adapter protein-deﬁcient mice. J. Exp. Med. 198:809–21 154. Fanzo JC, Hu CM, Jang SY, Pernis AB. 2003. Regulation of lymphocyte apoptosis by interferon regulatory factor 4 (IRF-4). J. Exp. Med. 197:303–14 155. Field SJ, Tsai FY, Kuo F, Zubiaga AM, Kaelin WG Jr, et al. 1996. E2F-1 functions in mice to promote apoptosis and suppress proliferation. Cell 85:549–61 156. Lissy NA, Davis PK, Irwin M, Kaelin WG, Dowdy SF. 2000. A common E2F-1 and p73 pathway mediates cell death induced by TCR activation. Nature 407:642–45 157. Yamasaki L, Jacks T, Bronson R, Goillot E, Harlow E, Dyson NJ. 1996. Tumor induction and tissue atrophy in mice lacking E2F-1. Cell 85:537–48 www.annualreviews.org • Genetic Disorders and Cell Death

351

ANRV270-IY24-11

ARI

23 February 2006

15:22

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

158. Sommers CL, Park CS, Lee J, Feng C, Fuller CL, et al. 2002. A LAT mutation that inhibits T cell development yet induces lymphoproliferation. Science 296:2040–43 159. Hodge MR, Ranger AM, Charles de la Brousse F, Hoey T, Grusby MJ, et al. 1996. Hyperproliferation and dysregulation of IL-4 expression in NF-ATp-deﬁcient mice. Immunity 4:397–405 160. Peng SL, Gerth AJ, Ranger AM, Glimcher LH. 2001. NFATc1 and NFATc2 together control both T and B cell activation and differentiation. Immunity 14:13–20 161. Ranger AM, Oukka M, Rengarajan J, Glimcher LH. 1998. Inhibitory function of two NFAT family members in lymphoid homeostasis and Th2 development. Immunity 9:627– 35 162. Qian YC, Qin JZ, Cui G, Naramura M, Snow EC, et al. 2004. Act1, a negative regulator in CD40- and BAFF-mediated B cell survival. Immunity 21:575–87

352

Bid`ere

· · Su

Lenardo

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:321-352. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe Department of Immunology, Scripps Research Institute, La Jolla, California 92037; email: [email protected]

Annu. Rev. Immunol. 2006. 24:353–89 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090552 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0353$20.00

Key Words mutagenesis, Mendelian genetics, infection, innate immunity

Abstract Classical genetic methods, driven by phenotype rather than hypotheses, generally permit the identiﬁcation of all proteins that serve nonredundant functions in a deﬁned biological process. Long before this goal is achieved, and sometimes at the very outset, genetics may cut to the heart of a biological puzzle. So it was in the ﬁeld of mammalian innate immunity. The positional cloning of a spontaneous mutation that caused lipopolysaccharide resistance and susceptibility to Gram-negative infection led directly to the understanding that Toll-like receptors (TLRs) are essential sensors of microbial infection. Other mutations, induced by the random germ line mutagen ENU (N-ethyl-N-nitrosourea), have disclosed key molecules in the TLR signaling pathways and helped us to construct a reasonably sophisticated portrait of the afferent innate immune response. A still broader genetic screen—one that detects all mutations that compromise survival during infection—is permitting fresh insight into the number and types of proteins that mammals use to defend themselves against microbes.

353

ANRV270-IY24-12

ARI

15 February 2006

4:15

INTRODUCTION

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Forward genetics: the genetic approach that begins with phenotype of unknown origin and ends with the identiﬁcation of culpable mutation(s)

Phenomena and Phenotypes: The Geneticist’s View of Resistance to Infectious Disease Few immunologists are content to forsake hypotheses and follow a classical genetic strategy. But perhaps more of them should. The forward genetic approach to biological problems is an unbiased one. It does not begin with broad speculation about how a biological system might operate, or even with more restricted hypotheses, such as guesses about the function of a particular protein or how cells interact. On the contrary, genetic exploration begins with no preconceptions at all. Hypotheses are the outcome of genetic investigation rather than the starting point (Figure 1). In genetics, as in biology at large, phenomena come ﬁrst. They are the ultimate source of all curiosity and all inquiry. Some phenom-

Figure 1 Summary of the classical (“forward”) genetic approach. Hypotheses (and experiments to test them) are the result of genetic inquiry rather than the starting point. 354

Beutler et al.

ena (the shapes of leaves, the presence of spots on the wings of moths, and the fact that dogs beget dogs whereas humans beget humans) are either so subtle or so commonplace that most people hardly think about them at all, although their importance may belie their banality. Other phenomena (aging, consciousness, and cancer) are uppermost in the popular mind and have impelled investigation on a broad scale, using all the tools technology can offer. However prosaic or glamorous the phenomena might seem, their analysis by classical genetic methods entails the identiﬁcation of phenotypes: alternative forms of phenomena. Where such phenotypes do not exist, they must be created. Once a monogenic phenotype is at hand, existing technology invariably permits identiﬁcation of the DNA sequence difference that is behind it. Ultimately, all proteins with nonredundant functions in a particular biological phenomenon may be identiﬁed in this manner. Even if it falls short of this idealized goal, classical genetic analysis creates a substrate upon which other methods may operate. The central phenomenon at issue in immunology is resistance to infection: the fact that we do not passively succumb to microbes when inoculated with them. It is a phenomenon that declares itself every day, and it would certainly fall into the “commonplace” category if not for the impressive consequences of its failure. Resistance to infection usually goes unnoticed, but it is essential to the survival of individuals and species. Genetic reasoning immediately tells us that the preponderance of resistance to infection is inherited rather than acquired, despite the quintessentially environmental status of microbes themselves. To argue in the most general way, different mammalian species show great variation in susceptibility to speciﬁc infectious agents. In fact, it is safe to say that there are no universal pathogens, i.e., microbes to which all multicellular organisms (or even all vertebrates) are susceptible. Never yet has a human died of mouse

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

cytomegalovirus (MCMV) infection, at least as far as we know. Never yet has a mouse died of human immunodeﬁciency virus (HIV) infection. Admittedly, a vast number of mutational differences have set humans and mice forever apart. But a discrete and ultimately deﬁnable collection of mutations must specify ability to resist infection by these organisms and all others. Although far fewer mutational events distinguish individuals belonging to a single species, formal measurement of heritability reveals that here, too, susceptibility to infection is largely determined by genetic makeup. Premature death from infection in a biological parent confers a high probability of premature death in a child separated from the parent by adoption (1). Moreover, variation in susceptibility to a given infectious agent is frequently demonstrable among individuals of a single species and usually has a genetic basis. Among mice, polymorphisms at speciﬁc loci inﬂuence susceptibility to MCMV (discussed further below) (2–5), and among humans, polymorphisms at speciﬁc loci inﬂuence susceptibility to HIV (6, 7). What types of immunity are there, and which type confers the heritability of resistance to infection? The term “adaptive” is used to describe a particular type of immunity, represented only in vertebrates as far as we know and markedly inﬂuenced by the environment, i.e., by previous exposure to an infectious agent or antigen, which impels the expansion of lymphoid clones that are directed toward speciﬁc recognition of the same target. The term “innate” is used to describe all other forms of immunity. There is a tendency to regard innate immunity as heritable and adaptive immunity, with its remarkable plasticity, as acquired. But, of course, the fact that adaptive immunity can be eliminated by a well-placed mutation reveals that it, too, is inherited. To the geneticist, all heritable effects on immunity are of interest, and the distinction between innate and adaptive systems is of secondary importance, although often easy to make.

Adaptive immunity evolved at least twice (8). The exact selective pressures that drove the development of our own adaptive immune system are unknown and can only be imagined. Perhaps an anticipatory immune system offered species a means of coping with frequent and devastating plagues caused by microbes that escaped containment by the relatively static innate immune response. Whatever the reason, it may be assumed that the adaptive immune system arose in the context of an advanced and highly functional innate immune system, and it remains largely dependent on the innate system today. Subsequent to the rise of adaptive immunity, certain innate immune functions must have become redundant and as a consequence were lost; hence, innate immunity alone offers suboptimal protection to present-day vertebrates. The innate system has come to require certain molecules of adaptive immune origin (for example, IFNγ) to function properly. For these reasons, many of the mutations that affect innate immune function also affect adaptive immune function and vice versa.

MCMV: mouse cytomegalovirus Resistome: the set of all genes encoding proteins with nonredundant functions in host resistance

The Concept of the Resistome The resistome is deﬁned as the set of genes encoding proteins with nonredundant function in resistance to infection (9). It is possible to speak of a universal resistome, as well as of the resistome for speciﬁc organisms. Many components of the resistome are entirely conditional. If infectious organisms are not present, there is no need for these genes to exist. In effect, they are fully dedicated to immune function. Other components of the resistome fulﬁll separate and essential biological functions, having recently been appropriated by evolution to create resistance, or conversely, to serve a function unrelated to resistance. In some instances, it is difﬁcult to decide on the primary function of a protein that is known to participate in resistance. Toll, a transmembrane receptor with a dual function in host resistance (10) and in embryonic development www.annualreviews.org • Genetic Analysis of Host Resistance

355

ANRV270-IY24-12

ARI

15 February 2006

4:15

in Drosophila (11), probably was inherited as a mediator of immunity and co-opted to serve a developmental function, based on the use of the TIR (Toll/IL-1R/Resistance) domain for immunity in most species (12). But the premise is debatable, and some believe that the primordial function of Toll was developmental. On occasion, a protein that clearly evolved to serve a function that has nothing to do with immunity may offer strong resistance to infection in one allelic isoform, but it may fail to do so in another form (e.g., sickle hemoglobin, which confers resistance to Plasmodium falciparum, versus hemoglobin A, which does not). Examples of this sort may be viewed as evolutionary works in progress, and sufﬁcient selective pressure might drive the mutant isoform to homozygosity, whereupon, like Toll, the hemoglobin protein would necessarily be viewed as bifunctional, with a role in both oxygen transport and in immunity. Some resistome components (certainly most of them) confer resistance to a broad range of pathogens. In this sense, the resistome is degenerate. Where the innate immune system is concerned, low speciﬁcity is the general rule, to allow for effective resistance to many microbes within the constraints imposed by limitations in genomic size. Some components of the resistome, however, are highly speciﬁc in their effects, and although they serve to recognize or overcome a very restricted repertoire of microbes, they may be essential to survival of the species. Irrespective of functional category, a concerted and determined genetic approach will ultimately reveal every protein that plays a nonredundant role in resistance and, in principle, will even identify those with latent resistance functions, given that such functions do not yet exist. Such a genetic approach will do so even if such functions of the proteins in question are beyond the imagination, are without precedent, and are unpredictable by any exercise of logic. Therein lies the appeal of the genetic approach and the source of its power.

TIR: Toll/IL1R/Resistance

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TLR: Toll-like receptor

356

Beutler et al.

The classical genetic approach is beholden to keen observation and is empowered by experimental designs that reveal phenotypes inapparent in the absence of special conditions. But the classical genetic approach does not depend on guesswork and may ultimately succeed where hypotheses fail. The classical genetic approach led to the identiﬁcation of the Toll-like receptors (TLRs) as the key proteins that allow us to recognize infection and, in large part, to the decipherment of signaling pathways that are activated by the TLRs. Genetic reasoning has also pointed to the existence of alternative pathways leading to activation of both the innate and adaptive immune systems.

THE PHENOMENON: HOST RESISTANCE Cell-Autonomous Immunity and the Evolution of Specialized Systems for Host Resistance Resistance to microbes must be measured with reference to a control rather than in absolute terms because deﬁning a state in which no resistance exists is presently impossible. For example, in the absence of the type I IFN receptor, mice are markedly compromised in their ability to resist MCMV infection (13). If the type II IFN receptor is also eliminated, susceptibility is still greater (13). Given that IFN-independent resistance mechanisms also may exist, susceptibility might be further enhanced by additional mutations. For the present, resistance must be assayed by measuring mean lethal inoculum, mean cytolytic inoculum, or the proliferation of the agent that is tested. Even in the absence of a functioning immune system, cultured cells offer resistance to the growth of microbes. For example, many cells, including ﬁbroblasts, are capable of mounting a type I IFN response. Cellautonomous immunity presumably operates in vivo as well as in vitro. However, advanced multicellular organisms have evolved

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

additional defensive strategies that make use of specialized cells designed to sense microbes, alert the host, and destroy microbes. The innate immune system depends heavily on macrophages, several classes of dendritic cells (DCs), natural killer (NK) cells, NKT cells, and granulocytes for host defense. Interestingly, although lymphoid cells provide the cellular basis for adaptive immunity, they have also been used for purely innate immune functions. The responses of macrophages and myeloid DCs have been studied in greatest detail; NK cells have also received considerable attention.

Sensing, Response, and Self-Tolerance: The Strategy of the Innate Immune System Immune responses are potentially injurious to the host and are also energy intensive. Although antimicrobial peptides are constitutively produced in some compartments, most responses are induced by infection. This requires that a system for recognition of microbes must exist, and in almost all biological systems recognition implies the presence of speciﬁc receptors. These receptors and the signals they elicit are part of the sensing apparatus of the host. In adaptive immune responses, recombination and clonal selection permit the generation of highly diverse, clonally expressed receptors that recognize microbial determinants, but not determinants of the host. However, this system is not called into play until several days after inoculation has occurred, and the host must initially rely on innate immune responses to contain an infection. The mammalian innate immune sensing apparatus has evolved to recognize broadly conserved molecules that are not found in the host but are represented in many microbial taxa. These include molecules such as lipopolysaccharide (LPS), bacterial lipopeptides, double-stranded RNA (dsRNA), and DNA bearing unmethylated CpG motifs. But occasionally, rather narrowly expressed deter-

minants are sensed as well. As discussed below, the G-glycoprotein of VSV is detected via the CD14/TLR4 apparatus of the host, a sensor that is principally devoted to LPS detection in mammals. Moreover, some vertebrates (ﬁsh, amphibians, reptiles) do not exhibit any response at all to LPS, and in some vertebrates (adult birds) the response is minimal. Yet representative birds and ﬁsh retain TLR4 orthologs. The simple model of receptors that are highly degenerate in their speciﬁcity may therefore need to be modiﬁed with time. An immune system must also create an environment that is inhospitable or lethal to microbes. This implies an effector arm of immune function. Processes that we associate with inﬂammation, including the generation of toxic radicals within phagocytic cells, the elaboration of hydrolytic enzymes, and mechanisms for containment of microbes (e.g., coagulation, granuloma formation), are components of the effector arm. It is reasonable to think that inﬂammation evolved chieﬂy to counter infection. Yet sterile inﬂammation, when it occurs, is a major cause of morbidity. In all likelihood, it depends on the same biochemical pathways as those used by the innate immune system. The ﬁnal requirement of the immune system is self-tolerance. A part of self-tolerance is self-/nonself-discrimination. This is accomplished by the innate immune system through an evolutionary process: The principal receptors of the innate immune system fail to recognize molecules of the host, at least under normal circumstances. There may be exceptions in that TLR3, -7, -8, and -9 recognize nucleic acids and, at least under some conditions, are able to recognize mammalian nucleic acids (14). Hence, one may speak of innate autoimmunity and adaptive autoimmunity (15) rather than autoimmunity per se. In addition to self-/nonself-discrimination, self-tolerance assumes that bystander effects should be minimized. As already mentioned, inﬂammation often damages healthy www.annualreviews.org • Genetic Analysis of Host Resistance

VSV: vesicular stomatitis virus LPS: lipopolysaccharide

357

ANRV270-IY24-12

ARI

15 February 2006

4:15

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

tissues. When infections are widespread, the process may be lethal. This appears to be a necessary imperfection in a system designed to enforce containment of small infections before they grow out of control. Small inocula occur far more commonly than inocula that are so large as to provoke life-threatening responses; hence, an aggressive system for dealing with microscale infections would be favored by evolution, even if this system is harmful when infections involve a substantial volume of tissue.

Figure 2 The Lps mutation held the key to understanding innate immune sensing. All LPS-induced phenomena were abolished by the mutation, which was widely assumed to affect the LPS receptor.

358

Beutler et al.

Microbial Sensing, LPS, and the Link Between Pathology and Protection Molecules detected by the innate immune system began to be identiﬁed in the early part of the twentieth century, and structural analysis of some such molecules was completed more than 30 years ago (16), leaving open the question of which receptors must be responsible for their recognition. The prototypic microbial inducer, sensed strongly by mammals although not by all vertebrates (17), is LPS. The toxicity of LPS also depends on mononuclear phagocytic cells (18); most other cells of the host are essentially unresponsive to it. Ultimately, toxicity depends on speciﬁc cytokines elaborated by these cells, notably tumor necrosis factor (TNF) (19) and type I IFN (20). Widely used to induce local and systemic inﬂammation and known to reproduce most of the pathologic features of sepsis, LPS elicits responses at picomolar concentrations, a fact that spoke strongly in favor of the existence of a speciﬁc receptor. The dependence of LPS signaling on a single protein was revealed by the identiﬁcation, in 1965, of a remarkable phenotype in mice (21) seemingly caused by a spontaneous mutation and later traced to a single locus, termed Lps (22–24). The phenotype was ﬁrst observed in mice of the C3H/HeJ strain and was marked by a very speciﬁc and profound insensitivity to LPS. Neither the lethal effect of LPS nor any of the cellular effects of LPS occurred in these mice. This included the well-known adjuvant effect of LPS (25), i.e., its ability to provoke an adaptive immune response to coadministered protein antigens (26). Later, it was shown that mice of the C57BL/10ScCr strain had an allelic defect (27). The Lps mutation was widely believed to affect the LPS receptor or an essential component of the receptor, but formal proof awaited positional cloning data (Figure 2). Long in advance of the positional cloning of Lps, researchers showed that failure to sense LPS lowered the mean lethal inoculum

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

of Salmonella typhimurium organisms by approximately four orders of magnitude (28, 29). Similar observations were later made for other Gram-negative microbes administered by other routes (30, 31). These studies provided solid proof that sensing LPS was important to an effective innate immune response to Gram-negative microbes or, conversely, that LPS was one of the key molecules recognized by the innate immune system. Finding the Lps locus assumed great practical importance because it seemed to offer the key to understanding innate immune sensing. Hidden in these studies was an important principle already mentioned above: The innate response evolved to be highly aggressive on a microscopic scale. The same response, when disseminated, can be lethal to the host, as in the example of endotoxin-induced shock. The innate response has evolved to balance the need for containment of microscale microbial infections, which are very common, against the risk of sepsis. Sepsis occurs as the result of a large infusion of microbes (an uncommon event) or because early containment of an infection has been unsuccessful. Sepsis may be seen, in most cases, as the result of innate immune failure.

The Basis of Innate Immune Sensing as Revealed by a Phenotype The positional cloning of the Lps locus was achieved by following a phenotype (failure of macrophages to produce TNF in response to LPS). This was deemed a biologically relevant phenotype because TNF had earlier been shown to be one of the central executors of the toxicity of LPS (19). TNF production was well known to depend on nuclear translocation of NF-κB (32). However, the primary signals elicited by LPS were unknown at that time. The mutation responsible for LPS unresponsiveness was found to be a missense error (P712H) altering the cytoplasmic domain of TLR4 (33). TLR4 was then known only as one of ﬁve mammalian paralogs of

Toll (34–37), a Drosophila protein with a dual role in development and in innate immunity. Speciﬁcally, ﬂies with mutations affecting Toll or components of the Toll signaling pathway failed to mount an adequate antimicrobial peptide response to fungi (10) [and, as was later shown, to Gram-positive bacteria as well (38)]. The Toll superfamily also included the IL-1 receptor (IL-1R) and IL-18R, which bear cytoplasmic domain homology to Toll, provoke inﬂammatory responses, and activate NF-κB (39–41). Because IL-1R and IL18R served immune functions, investigators speculated that TLR proteins in mammals might have either developmental functions (35) or immune function (36), or perhaps both. In fact, investigators had shown that ligation of TLR4, enforced by the creation of a chimeric protein in which a sequence encoding CD4 was substituted for the native TLR4 ectodomain-encoding sequence, could activate NF-κB in transfected cells (36), an observation consistent with the fact that Toll and IL-1R could activate NF-κB. However, this observation in itself did not actually address the function of the TLRs because NFκB activation has many consequences, some related to immunity and some not. Moreover, the experiment did not give any insight into the natural ligand for TLR4, nor did it reveal whether the ligand was endogenous (as in the case of Toll, IL-1R, and IL-18R) or exogenous. The fact that a point mutation in TLR4 entirely prevented mammalian responses to LPS and to Gram-negative bacteria, yet had no developmental consequences, strongly favored the interpretation that the TLRs had an immunologic function and speciﬁcally suggested that the TLRs acted as the long-sought sensors of molecules made by microbes. Although LPS was sensed speciﬁcally by TLR4, investigators immediately postulated that the other TLRs might sense other inﬂammatory molecules (e.g., dsRNA, unmethylated DNA, glucans, and lipopeptides), and that collectively the mammalian TLR paralogs www.annualreviews.org • Genetic Analysis of Host Resistance

359

ANRV270-IY24-12

ARI

15 February 2006

4:15

might alert the host to infection, triggering an immune response.

The Structure of the TLRs

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

All TLRs are single-spanning transmembrane proteins with ectodomains largely composed of leucine-rich repeats (LRRs), and with a cytoplasmic domain largely composed of a TIR domain (Figure 3).

Figure 3 Schematic rendering of TLRs, based on the structure of the TLR3 ectodomain and the TLR2 TIR motif. TLR ectodomains are all likely to be dimerized, horseshoe-shaped, curved solenoids composed of numerous repeating LRRs. The cytoplasmic domains are compact and globular, consisting mostly of a TIR motif. In some cases (e.g., TLR2, TLR4), multiple accessory molecules may be required for ligand engagement. Adapter proteins, four of which are known to carry TLR signals, also have TIR domain structures. It is believed that ligands elicit a conformational change, allowing recruitment of speciﬁc adapter(s). The critical proline in the BB loop and the critical valine at the Pococurante site are rendered at the atomic level and colored red and yellow, respectively. These residues are essential for most, but not all, TIR motif interactions, as discussed in text. Ligands are presumed to engage TLRs near the point of interface between LRR subunits. (Rendering of ﬁgure was performed with PyMOL, 2005 DeLano Scientiﬁc LLC.) 360

Beutler et al.

The structure of a representative TLR (TLR3) has recently been solved through Xray crystallography, revealing that TLR3 is a dimeric protein composed of two horseshoeshaped subunits that stack together side by side. A large quantity of carbohydrate projects from each horseshoe in one direction, so that only the carbohydrate-free surfaces can remain in contact with one another. At the point of interaction between subunits, a positively charged cluster of residues marks the point of presumed interaction between TLR3 and its ligand, dsRNA (42). The cytoplasmic domains of TLR1 and TLR2 have also been crystallized and their structures solved (43). Each is compact and globular, and there is no evidence of interaction between subunits in the unit cell. A surface structure known as the BB loop is present in each TIR domain and contains the residue corresponding to residue that is altered in the TLR4 protein of C3H/HeJ mice (P712H for TLR4; P681H for TLR2). The mutation does not prevent normal folding of the TIR domain, as shown by crystallization studies (43). However, it does abolish signal transduction from most of the TLRs (with some key exceptions, noted below), indicating that the residue in question is, in those cases, part of the signaling interface for interaction with adapter proteins, as discussed below. The similarity between TIR domains is such that threading programs can be used to model each TIR domain on those that have been directly analyzed by X-ray crystallography (44). It is widely believed that all the TLRs are dimeric, with some homodimeric and others (TLR2/TLR6 and TLR2/TLR1) heterodimeric. Dimeric structure is assumed because enforced dimerization triggers a response (36) and because membraneproximal modiﬁcations of cysteine residues in Toll cause dorsalization of the embryo in Drosophila (45). Moreover, IL-1R and IL-18R are heterodimers and signal via TIR domains, although the extracellular domain of each receptor contains immunoglobulin superfamily repeats. Dimeric structure must be taken

ANRV270-IY24-12

ARI

15 February 2006

4:15

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

into account in considering precisely how the TLRs signal. Some TLRs are clearly intracellular, residing predominantly or entirely within the endosomes. These include TLR3, -7, -8, and -9, which sense nucleic acids. TLR1, -2, -4, and -6 are at least largely expressed on the surface, although their presence within the phagosome and later components of the endocytic pathway is not excluded.

Other Components of the LPS Receptor At least some TLRs depend on other proteins to signal effectively (Figure 4). In the case of the TLR2/TLR6 heterodimer, CD36, a double-spanning plasma membrane protein of the class B scavenger receptor family, is known to participate in signaling events, as shown by the phenotype of mutant mice identiﬁed in a TLR signaling screen (46), described in detail below. The TLR2/TLR6 heterodimer also depends in part on CD14 to signal effectively, and mutations that eliminate CD14 partly (∼50%) impair sensing of all TLR2/TLR6 ligands.

The LPS receptor TLR4 depends even more strongly on CD14 than does the TLR2/TLR6 heterodimer, and without CD14, TLR4 cannot mobilize all of the adapter proteins that it requires for full signaling activity. This dependence suggests that TLR4 is organized into a supramolecular complex through interaction with CD14 and with its ligand, LPS. In fact, CD14 was ﬁrst shown to enhance LPS signals in 1990 and was the ﬁrst component of the LPS receptor complex to be identiﬁed (47). Only more recently was it shown, through a forward genetic approach, that CD14 is speciﬁcally required for the detection of smooth LPS (LPS with abundant O-glycosylation) rather than rough LPS or lipid A, and for signaling via the TRIF/TRAM pathway. CD14 recognizes other TLR4 ligands as well, such as glycoprotein G of VSV (P. Georgel, Z. Jiang, S. Kunz, K. Hoebe, E. Janssen, M. Oldstone, and B. Beutler, manuscript submitted). In addition, a small protein known as MD-2 associates with the TLR4 ectodomain and is required for LPS signal transduction. Mutations in MD-2 entirely prevent LPS signaling (48), matching the effect of mutations in TLR4.

Figure 4 Shared and unique components of TLR complexes. Germ line mutations have proven the participation of CD14, CD36, and MD-2 in signaling by TLR2 and TLR4 complexes. Coreceptors broaden the speciﬁcity of the receptor complexes and in some cases inﬂuence the choice of adapters that are recruited. MALP-2 and LTA require CD36 for full signaling efﬁcacy; zymosan and PAM2 CSK4 do not. CD14 is partly required by all TLR2/TLRX and TLR2/TLR6 agonists, is fully required by smooth LPS, and is partly required by rough LPS. TLRX indicates uncertainty concerning the partner for TLR2 in the PAM2 CSK4 receptor complex; it is most likely a second TLR2 subunit. (Abbreviations: VSV-G, vesicular stomatitis virus glycoprotein G; MMTV-G, mouse mammary tumor virus surface glycoprotein.) www.annualreviews.org • Genetic Analysis of Host Resistance

361

ANRV270-IY24-12

ARI

15 February 2006

4:15

The Sensing Function of Other TLRs TRIF: Toll/IL-1R domain-containing adapter inducing IFN-β

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TRAM: TRIF-related adapter molecule PAM3 CSK4 : triacylated lipopeptide (N-Palmitoyl-S[2,3bis(palmitoyloxy)(2RS)-propy1]-[R]cysteinyl-[S]-seryl[S]-lysyl-[S]-lysyl[S]-lysyl-[S]-lysine x 3 HCI) MALP-2: macrophage activating lipopeptide 2 LTA: lipoteichoic acid

Targeted gene deletion has revealed that TLR2/TLR1 heterodimers are required for detection of triacylated bacterial lipopeptides or synthetic peptides such as PAM3 CSK4 (49). Some diacylated lipopeptides, such as the Mycoplasma-derived macrophage activating lipopeptide 2 (MALP-2), signal via the TLR2/TLR6 heterodimer, and they do so in a stereospeciﬁc manner: Only the R enantiomer uses this signaling complex (50). The TLR2/TLR6 complex reportedly also recognizes lipoteichoic acid (LTA) and β-glucans such as zymosan, although some caution is needed in interpreting cellular signaling events induced by microbial fractions such as these, given that they are not always entirely pure. Other diacylated peptides (e.g., PAM2 CSK4 ), although TLR2 dependent, require neither TLR1 nor TLR6 (51). Either a TLR2 homodimer or TLR2 associated with TLRX (candidates include TLR11, -12, and -13) might serve recognition of these molecular species. TLR5 recognizes ﬂagellin, a protein represented in both Gram-positive and Gram-

negative microbes (52). TLR7 (in mice and humans) and TLR8 (in humans only) recognize nucleoside analogs such as resiquimod or imiquimod (53), drugs with antineoplastic and antiviral potential that are now believed to mimic the natural ligand ssRNA (54– 56). TLR9 senses DNA bearing unmethylated CpG-containing motifs (57) and does so in a species-speciﬁc fashion (58). Humans (but not mice) express TLR10. Mice (but not humans) express TLR11, -12, and -13. There is some confusion of nomenclature in that TLR11 has been called TLR12 by some authors, and vice versa (59, 60). The molecular speciﬁcity of TLR10 remains unknown, although because of its structural similarity to TLR1 and TLR6, TLR10 likely senses lipopeptides. Its absence in mice has prevented examination of the knockout phenotype. In mice, TLR11 recognizes a proﬁlin-like component of Toxoplasma gondii (61) (evidently distinguishable from the host molecule). The functions of mouse TLR12 and TLR13 have not yet been determined (Table 1). Particularly where nucleotide ligands are concerned, speciﬁcity is not absolute, and

Table 1 TLR ligand specificities, adapters used either alone or in combination with one another, and species representationa Ligands

TLR

Adapters

Species

PAM3 CSK4

1,2

MyD88, MAL

Human, Mouse

PAM2 CSK4

2,X

MyD88, MAL

Human, Mouse

MALP-2, LTA, Zym

2,6

MyD88, MAL

Human, Mouse

dsRNA

3

TRIF

Human, Mouse

LPS, VSV-G, MMTV-G

4

MyD88, MAL, TRIF, TRAM

Human, Mouse

Flagellin

5

MyD88

Human, Mouse

ssRNA, IAQ

7

MyD88

Human, Mouse

ssRNA, IAQ

8

MyD88

Humanb

CpG-ODN

9

MyD88

Human, Mouse

Unknown

10

Unknown

Human

Proﬁlin

11

MyD88

Mouse

Unknown

12

Unknown

Mouse

Unknown

13

Unknown

Mouse

a

IAQ, imidazoquinolines, including resiquimod and imiquimod; MAL, MyD88 adapter–like; CpG-ODN, synthetic oligodeoxynucleotides containing CpG motifs. Other abbreviations as described in text. b TLR8 is encoded in the mouse genome, but activating ligands, if any, are unknown. 362

Beutler et al.

ANRV270-IY24-12

ARI

15 February 2006

4:15

host molecules can also trigger inﬂammatory responses, although perhaps less intense responses than those initiated by microbes. TLR3, -7, -8, and -9 are largely (although not necessarily entirely) dedicated to the detection of viral infection. However, other TLRs, most notably TLR4, also play an important role in viral sensing.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

The Net Importance of TLRs in Resistance to Microbes TLRs signal via four adapter proteins (MyD88, Mal, TRIF, and TRAM), discussed in greater detail below. Without the adapter protein MyD88, which is required for all signal transduction by many of the TLRs and for full signaling by TLR4, mice frequently succumb to infections derived from their own oral ﬂora [mixed infections of the submandibular nodes with α-hemolytic streptococci and Pasteurella pneumotropica have been observed (Z. Jiang & P. Georgel, unpublished observations)]. When both MyD88 and the adapter TRIF are absent, mice are even more susceptible to infection, and very few survive to weaning age (B. Beutler & K. Hoebe, unpublished observation). Individual TLR mutations are also compromising. For example, TLR4 mutations cause enhanced susceptibility to Salmonella typhimurium (28) and Escherichia coli (29) infections; TLR2 deletion causes enhanced susceptibility to Staphylococcus aureus (62) and other organisms. TLR3 deﬁciency (59) and to an even greater extent TLR9 deﬁciency (59, 63) cause susceptibility to MCMV infection, a highly prevalent pathogen in wild mice. By implication, the same mutations may enhance susceptibility to β-herpesviruses in other species. These observations suggest that the TLRs make an essential contribution to the recognition of infectious organisms.

The Special Case of Viruses Viruses were the last pathogens found to be recognized by TLRs (59, 63–68), and at the

time, it came as something of a surprise to see that they were so recognized. Viruses are primarily, although not entirely, sensed by TLRs dedicated to the detection of unmethylated DNA, dsRNA, and ssRNA. As such, the challenge of self-/nonself-discrimination is greater because host nucleic acids are only slightly different than virally encoded nucleic acids. ssRNA is, of course, very abundant in host cells, and most RNA species have at least some dsRNA regions. Some CpG dinucleotides within the mammalian genome are unmethylated and should therefore be recognized by TLR9. These facts have led to the suggestion that autoimmune diseases are fueled in part by TLR signaling, stimulated by host nucleoprotein complexes (14). At the very least, viruses test the limits of self-/nonself-discrimination, and why TLRs are so effective in recognizing them is still not entirely clear. TLRs 3, 7, 8, and 9 not only detect structural differences between viral nucleic acids and host nucleic acids, but also must engage the nucleic acids within endosomes, a compartment from which host nucleic acids are normally excluded (69). There are numerous examples of viral protein detection by TLRs as well. The envelope glycoprotein of mouse mammary tumor virus (MMTV) (65, 66) and glycoprotein G of VSV (P. Georgel, manuscript in preparation) are both recognized via TLR4. It has also been reported that the F-glycoprotein of respiratory syncitial virus activates TLR4 (64), although this assertion has recently been challenged (70). The ability of TLR4 to detect viruses is interesting for several reasons. Like TLR3, TLR4 uses the adapter protein TRIF and is capable of stimulating type I IFN responses. But TLR4 is normally thought of as the LPS receptor, and it is surprising to ﬁnd that entirely different molecules are engaged by it. The fact that they are raises structural and evolutionary questions. The ability of receptors to recognize structurally disparate molecules is not without precedent. For example, opiates are plant alkaloid agonists that www.annualreviews.org • Genetic Analysis of Host Resistance

MyD88: myeloid differentiation factor 88 MMTV: mouse mammary tumor virus

363

ANRV270-IY24-12

ARI

15 February 2006

4:15

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TIR-DEPENDENT AND TIR-INDEPENDENT ADAPTIVE IMMUNE ACTIVATION TLRs mediate adjuvant effects induced by microbes, known since the time of Freund. The adjuvant signals initiated by the TLRs are largely directed toward a CD4 response. But they are not required for adaptive immune activation per se. In the absence of TLR adapter proteins MyD88 and TRIF, TLR signaling is at least greatly diminished. However, normal lymphoid development, IgG production, and allograft rejection are observed. These ﬁndings point to the existence of TIR-independent pathways for adaptive immune activation. Certain types of cell death, including death induced by UVor γ-irradiation or by Fas ligation, are capable of triggering a strong immunoadjuvant pathway that is chieﬂy (although not exclusively) directed toward CTL activation, and entirely TLR independent. This pathway is known to depend on the integrity of the type I IFN receptor and UNC-93B (the 12spanning ER protein affected by the 3d mutation). To the extent that the pathway can (weakly) activate CD4 cells, it is additionally dependent on CD36 (CD4 activation is selectively impaired by the oblivious mutation). However, CD36 plays no part in the internalization of antigen molecules. The TIR-independent immunoadjuvant pathway, which presumably serves the recognition of cell death triggered by certain viral infections, is currently being analyzed using ENU mutagenesis.

IRF: IFN regulatory factor IRAK: IL-1R-associated kinase

364

trigger a response from endorphin receptors. However, it is a different task for receptors to maintain contact with multiple targets that are driven to evade detection. In the case of VSV glycoprotein G, mutations might nullify recognition, permitting the virus to evade detection. In the case of the MMTV envelope glycoprotein, investigators believe that the virus uses TLR4 for a subversive purpose, stimulating IL-10 production rather than the full spectrum of TLR signals, and minimizing the immune response (66).

tity long known as something made by microbes that was sensed by the host (33). TLRs are the key sensors of microbial infections of many and perhaps all types. It follows that they are ultimately responsible for most (although not necessarily all) infection-related phenomena, no matter how complex those phenomena may be (15). One of these infection-related phenomena is the adjuvant effect of microbes. The gold standard of adjuvants is Freund’s complete adjuvant, which is made using mycobacteria suspended in an oil-in-water emulsion. Molecules of microbial origin must have adjuvant properties. In 1955, investigators showed that LPS is endowed with adjuvant properties (26), and in subsequent years other molecules of microbial origin were also shown to have adjuvant effects (71–76). Among the key molecular events leading to an adaptive immune response is upregulation of costimulatory molecules (including but not limited to CD80 and CD86), which must occur on the surface of antigen-presenting cells (APCs) to activate T cell mitogenesis (77). LPS elicits upregulation of these molecules on APCs. Medzhitov and colleagues (36) once speculated that this occurs through the ability of TLRs to activate NF-κB. However, it is now clear that the key event in upregulation of costimulatory molecules is the activation of a type I IFN response (78, 79). This, in turn, depends on LPS activation of the MyD88independent pathway and on activation of the adapters TRIF (67, 80), TRAM (81), and IRF3, discussed in more detail below. Other microbial adjuvants activate type I IFN synthesis through activation of the adapter MyD88, via IRAK-1 and IRF-7 (82).

The Connection Between Innate and Adaptive Immunity

TIR-Dependent and TIR-Independent Pathways: A Death-Driven Pathway for CTL Activation

As described so far, the function of TLRs in mammals was originally established by ﬁnding the receptor for LPS, a molecular en-

The fact that TLRs mediate microbial adjuvanticity led to early suggestions that they might mediate adaptive immune activation at

Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

Figure 5 An immunoadjuvant pathway driven by programmed cell death. The pathway does not depend on TLR signaling, and the triggering ligands induced by programmed cell death are unknown. The 3d mutation prevents the APC from presenting exogenous antigen, so that both priming and cross-priming are inhibited. The oblivious mutation prevents priming, but not cross-priming. The process also depends on the production of type I IFN and on its signaling via IFNAR on the APC.

large. However, in mice that lack two of the adapters required for TLR signaling (MyD88 and TRIF), there is little or no response to most microbial ligands. Although we have already noted that such mice are drastically immunocompromised (K. Hoebe & B. Beutler, unpublished observation), they exhibit normal concentrations of circulating antibodies, have normal lymphoid architecture, and can reject allografts. Although Freund’s adjuvant does not promote an immune response in such mice, alternative pathways for adaptive immune activation clearly exist and are presumably activated by nonmicrobial stimuli. The danger model of costimulation holds that cell death or injury may provide the essential signal for an adaptive immune response (83). It appears that some, but not all, forms of cell death trigger a strong and TIR-independent signal that leads to the induction of type I IFN synthesis, as well as to priming and cross-priming of T cells with speciﬁcity for antigens expressed by the dy-

ing cell (Figure 5). Programmed cell death induced by γ- or UV-irradiation or by Fas ligation will trigger such a response (E. Janssen, K. Tabeta, M. Barnes, S. McBride, S. Schoenberger, A. Theoﬁlopoulos, B. Beutler, and K. Hoebe, manuscript submitted). The cells that are capable of sensing programmed cell death are B220− and CD8low lymphoid elements, and they can be differentiated from bone marrow precursors using Flt-3 ligand. They acquire antigen by nibbling. Notably, the system is not represented in myeloid DCs or in macrophages; the macrophages probably dispose of many or most of the cells that die through senescence or other processes of attrition in vivo. The system for adaptive immune activation may be called into service when normal routes for removal of apoptotic cells are overwhelmed. Moreover, cell death that is induced by pathologic stimuli (viruses, irradiation) is likely qualitatively different than that occurring as a result of constitutive apoptosis, and it www.annualreviews.org • Genetic Analysis of Host Resistance

365

ANRV270-IY24-12

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ENU: N-ethyl-Nnitrosourea

4:15

leads to the expression of recognizable surface molecules that are as yet unknown. The death-driven pathway evokes a far more powerful cytolytic T lymphocyte (CTL) response than Freund’s adjuvant. Antigens are usually detected by the system when injected into mice in picogram to nanogram quantities, expressed by syngeneic cells that have been induced to undergo apoptosis. The system presumably evolved to detect cell death induced by viral infections and comprises a route to rapid adaptive immune activation in response to this stimulus. Because it is TIR independent, it would be effective when TLR sensing is not. This system of recognition conceivably participates in diverse immunological events, including allograft rejection and autoimmunity.

THE INTENTIONAL CREATION OF PHENOTYPE Although a spontaneous mutation led to the concept that TLRs detect molecules of microbial origin, investigators have had to induce mutations, either deliberately or at random, to dissect the TLR signaling pathways and to identify participating molecules that are structurally dissimilar from those that are known. The induction of mutations that cause phenotypes of interest, and subsequent identiﬁcation of those mutations, is the essence of the classical or forward genetic approach. The germ line mutagen used for this purpose is N-ethyl-N-nitrosourea (ENU).

ENU and Its Efficiency as a Mutagen In using a germ line mutagen, one would ideally like to create as many mutations as possible, consistent with the production of monogenic (as opposed to polygenic) phenotype, because only monogenic phenotype is readily mapped and positionally cloned. Practically speaking, ENU dosing is limited by the fact that male mice become sterile if administered too high a dose. ENU has been tested in many different strains of mice, and the standard pro366

Beutler et al.

tocol that has emerged, in C57BL/6 mice, is administration of 100 mg/kg three times at weekly intervals. A 12- to 15-week period of infertility follows. During this interval, the testis is repopulated by spermatogonia, arising from approximately 10 to 100 precursors. After this, mice may be bred to transmit germ line mutations to the G1 population. For studies of immunity, C57BL/6 females are best used in breeding to maintain a pure genetic background. Among G1 mice, dominant phenotype may be identiﬁed. To produce recessive phenotype, G1 mice are crossed again to C57BL/6 females, and then to their own daughters. Among G3 mice, recessive phenotype may be identiﬁed. The potency of a germ line mutagen is ultimately limited by the mutational load compatible with life. Even if ENU did not induce sterility, it clearly induces nearly the maximum number of mutations tolerated in the G3 population. Kile et al. (84) have analyzed the frequency of lethal hits induced by ENU within a restricted area of the genome using a balancer chromosome, a dominantly marked, inverted chromosome that carries a homozygous lethal mutation. Balancer chromosomes can be used to capture lethal mutations because only carriers of the balancer are evident in the G3 population, whereas homozygotes for the mutagenized chromosome and homozygotes for the balancer itself are absent. Approximately 1/13 G1 mice born to mutagenized sires carried lethal hits within the balancer region, with lethality deﬁned as a failure to survive to the age of weaning. Because the inverted region encompassed only about 5% of all genes in the genome, and because lethal hits occur elsewhere as well, it may be surmised that each G1 mouse is, on average, heterozygous for about 1.5 recessive lethal hits genome-wide. If a G1 mouse bears a lethal hit, the probability of its transmission to homozygosity in each G3 mouse is 1/8 (0.125), and the likelihood of nontransmission to homozygosity in each G3 mouse is 1–0.125, or 0.875. Hence, 1–0.8751.5 , or about 18%, of the G3 population is lost as

ANRV270-IY24-12

ARI

15 February 2006

4:15

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

a result of homozygosity for ENU-induced mutation. If the efﬁciency of ENU were tenfold higher (15 lethal hits per G1 mouse), 86% of the G3 population would be lost, an amount of attrition that might be considered undesirable because multiple litters would be required to produce a very small number of mice for screening. In effect, ENU provides mutagenic power approaching the maximum that can be tolerated.

The Concept of a Genomic Footprint For any phenotype that the investigator deﬁnes, a certain and very exact set of nucleotide changes across the genome are capable of producing that phenotype. These nucleotides comprise the genomic footprint of the phenotype in question. Estimates from direct sequencing of sentinel regions of G1 genomic DNA reveal that ENU changes about 1 base pair (bp) per million bp in the haploid genome (85), and 1/8 of these changes are transmitted to homozygosity in each G3 mouse. If we assume that all bp within a genomic footprint are equally at risk from ENU, it is possible to calculate the size of the genomic footprint based on the frequency of observed phenotypic change. If 100 bp are at risk, and if pedigree construction is such that 50% of G1 mutations are captured in the homozygous state (e.g., two G2 mice per G1 sire and three G3 mice per G2 dam), one would expect to see the phenotype once, on average, for every 120,000 G3 mice examined. If 100,000 bp are at risk, the phenotype should be apparent once in every 120 G3 mice examined. Probably the phenotype with the largest genomic footprint of all is lethality, which, as noted earlier, is realized in about 18% of G3 mice, suggesting a genomic footprint of about 1.6 million bp. The genomic footprint of a phenotype may be small because it is masked by other phenotypes (typically lethality). For example, a particular immunodeﬁciency phenotype might rarely be observed because mice do not survive

to the age required by the assay. Alternatively, the genes that support a particular immune function might also be required for development in utero, and only the exceptional nucleotide change will produce the phenotype yet be compatible with viability to term. Although viable hypomorphic alleles probably exist for all genes, the nucleotide substitutions that produce them are sometimes scarce. Rather, it is more common that a mutational change has no measurable effect on the function of a protein at all, or that it utterly destroys the protein. On other occasions, the genomic footprint of a phenotype may be small because the phenomenon under analysis is served by redundant pathways. Few genes are truly redundant because locus duplication most commonly leads to the formation of a degenerate pseudogene. Most commonly, the claim of redundancy belies a screen that is not sufﬁciently powerful to resolve genes with overlapping function, although the functions of the two genes are actually distinguishable. But some recently duplicated genes do exist in the genome, and their functions, although important, might not be revealed by ENU mutagenesis.

Genomic footprint (of a phenotype): the set of all nucleotides that can cause the phenotype in question when mutated

How the Footprint Is Parceled into Genes The genomic footprint of a phenotype is scattered among genes that support the phenomenon under study. The mathematical distribution of the target nucleotide population among genes has not been described and is not known from experiment because genome-wide saturation of a phenotype has not been approached. Gene size presumably inﬂuences the distribution; perhaps more inﬂuential still is variation in the resilience of encoded proteins to mutagenic change. In addition, it is possible (though uncertain) that nucleotides differ regionally in their susceptibility to ENU. The distribution of the genomic footprint among genes is a matter of importance in www.annualreviews.org • Genetic Analysis of Host Resistance

367

ARI

15 February 2006

4:15

considering the approach to saturation. Starting with the assumption that all nucleotides are equally vulnerable to change induced by ENU [which is assuredly untrue because the great majority of base changes are known to be A → T transversions or A → G transitions (86)], and with the fact that 1/8 of 1 millionth of the genome is altered in homozygous form in each G3 mouse, we might imagine that 5.9% of all nucleotides would be struck if one million G3 mice were examined, following the same pedigree parameters as those outlined above. One might thus assume that all genes relevant to the phenotype would be revealed. This assumption, however, depends entirely on the distribution of the target nucleotide population. A given phenotypic footprint might have 10,000 nucleotides (a typical number) parceled into 50 genes (perhaps also a typical number). The average gene would have 200 target nucleotides and would indeed be revealed. But a gene with 10 vulnerable nucleotides would have only a 46% chance of being revealed. And the footprint might include many such genes. At any rate, no ENU screen has yet approached a depth of one million mice. According to our best estimate, using a protocol in which half of all mutations are transmitted to homozygosity, ∼10% phenotypic saturation is achieved with the analysis of ∼10,000 G3 mice.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

Some Summary Statistics Almost all ENU-induced phenotypes result from coding change (missense errors, nonsense errors, splicing errors, or, more rarely, single base pair deletions), although exceptions have been reported (87, 88). The coding region of the mouse genome encompasses approximately 42 Mb of DNA (1.5% of the genome as a whole). It may therefore be calculated that if ENU creates 1 bp change per million bp, about 42 mutations fall within parts of the genome where a phenotypic effect can be exerted. About 76% of nucleotide changes 368

Beutler et al.

that fall within a coding sequence create coding change. Therefore, 32 coding changes occur in each G1 mouse born to a mutagenized sire. About four of these changes are transmitted to homozygosity in each G3 mouse. About 18% of mice die as a result of four random homozygous changes in coding sequence, which suggests that each individual change has approximately a 4.8% chance of delivering the lethal blow. The number of genes that are targets in this process (i.e., the number of genes with recessive lethal alleles) remains unknown. But if we assume that 1/3 of all genes have lethal alleles (a low-end estimate), we could conclude that, on average, about 16% of random coding changes within genes that can cause lethality do cause lethality. If 2/3 of all genes have lethal alleles (a high-end estimate), the ﬁgure would be 8%. This reveals the average resilience of proteins with respect to a deﬁned qualitative phenotype.

The Design of Screens A good screen reﬂects the designer’s appreciation of a biological phenomenon and curiosity about its cause. It avoids phenomena that are well understood, avoids redundancy, and, where possible, probes a large genomic footprint. The screening assay must be sufﬁciently robust to avoid even rare false positives. It is best if the assay is qualitative or, better still, binomial. But whatever the screen, phenotypes selected for analysis must be strong enough and penetrant enough to permit meiotic mapping, without which no progress can be made.

THE GENETIC DISSECTION OF TLR SIGNALING PATHWAYS What Was Known from the Start Analysis of TLR signaling pathways began before the mammalian TLRs were recognized, with the discovery that IL-1R signaled by way of the adapter protein MyD88 (89) and

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

Figure 6 Proteins known prior to beginning ENU mutagenesis to participate in signaling from seven of the TLRs to the level of TNF activity.

with recognition that the serine kinase IRAK transduces IL-1 signals (90, 91), depending on TRAF-6 (92) and leading to the activation of NF-κB. Similarities between the IL-1 signaling pathway and the Toll developmental pathway were noted early on (93), as was the similarity of the IL-1R cytoplasmic domain to the cytoplasmic domain of Toll (94). Early and rapid progress in understanding TLR signal transduction rested on this similarity. Homology searches led to the identiﬁcation of all the TLRs that are presently known, and gene targeting established the speciﬁcity of most of them once the LPS sensing function of TLR4 was determined by positional cloning (33). Homology searches have also disclosed the existence of at least four cytoplasmic TIR adapter proteins. Whereas MyD88 and Mal (also known as Tirap) were easily detected by using the Basic Linear Alignment Search Tool (BLAST), TRIF and TRAM were far more distant and were found belatedly using the Hidden Markov Model search algorithm (HMMER).

However, the participation of proteins structurally unrelated to the TLRs was much more difﬁcult to ascertain, and in that research a pure genetic approach has also played a valuable role. By 2000, a total of 22 proteins were known to participate in TLR signaling from the level of seven of the TLRs to the level of TNF bioactivity, and these proteins could be designated as potential targets (Figure 6). A genetic screen was applied to identify additional components of the pathway. The screen was performed by stimulating peritoneal macrophages with TLR agonists and then measuring TNF production by biological assay (killing of L-929 cells). To date, a total of 11 mutations have been identiﬁed by screening approximately 20,000 mice. These mutations alter a total of 10 genes. Five of the genes were known at the outset of screening. The other ﬁve, however, were unknown components of the TLR signaling apparatus. Three of these have been positionally cloned to date. www.annualreviews.org • Genetic Analysis of Host Resistance

TRAF-6: TNF receptor–associated factor 6

369

ANRV270-IY24-12

ARI

15 February 2006

4:15

Calculation of Pathway Complexity

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

A forward genetic approach permits unbiased estimation of the total number of protein components that are required for signaling from TLRs to TNF because each serves a nonredundant function. Given that 5 of the genes identiﬁed fell within the category of known genes (22 of which exist in all) whereas 5 did not, it might be guessed that 44 targets exist in total. Taking an independent approach to estimate the number of nonredundant proteins, 23% (5/22) of the known protein population was identiﬁed in the screen, corresponding to the degree of saturation achieved in a study of 20,000 mice. If 23% saturation nets 10 genes, 43 targets may be estimated to exist in all.

Figure 7 The position of Lps2 on the signaling map. The phenotype imparted by the mutation was one in which TLR3 and TLR4 were both unable to activate IRF-3. Production of TNF was also minimal, presumably because the normal protein also contributes to the activation of NF-κB. 370

Beutler et al.

Independently again, if one assumes that all targets have uniform size (each subsumes an equal fraction of the genomic footprint), the identiﬁcation of one instance of allelism among 11 mutations would suggest the presence of approximately 53 targets. Although these estimates are tentative and deeper saturation must be pursued to determine the number of targets with conﬁdence, one may conclude that a relatively small number of targets exist and that by this time most of them are probably known.

Lps2 and the Identification of TRIF MyD88-deﬁcient mice showed evidence of a MyD88-independent pathway, whereby LPS, acting via TLR4, could partially stimulate NF-κB translocation after a brief delay and activate type I IFN production without impediment (95). With the targeted deletion of both MyD88 and Mal genes, the same phenomenon was observed (96). TLR3 signaling also had no requirement for either adapter. The molecular basis of the MyD88independent pathway remained unclear. The hallmark of MyD88-independent signaling was not NF-κB activation, but rather IRF-3 activation and subsequent type I IFN synthesis. However, the MyD88-independent pathway is also capable of activating NF-κB and presumably does so by activating a component of the cascade that is distal to MyD88 itself, thereby bypassing the lesion imposed by mutations in MyD88 and/or Mal (Figure 7). Researchers now believe that the point of intersection between MyD88-independent and MyD88-dependent pathways is TRAF-6 (97, 98), a protein that coordinates the activation of numerous kinases, including TAK-1 [transforming growth factor (TGF)-β-activated kinase 1], which is required for the phosphorylation of IKKβ (99, 100), leading to NF-κB activation. A mutation termed Lps2 ﬁrst revealed the molecular basis of MyD88-independent signaling. The Lps2 mutation prevented IRF-3 phosphodimer formation and IFN synthesis

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

in response to either LPS or dsRNA, indicating that it affected a protein required for signaling from both TLR4 and TLR3. Moreover, Lps2 rendered mice susceptible to infection by MCMV (101). Positional cloning of the Lps2 mutation revealed a frameshift error in a new adapter protein, independently identiﬁed by homology searching (102) and by two-hybrid system analysis (103), and respectively known as TRIF or TICAM (TIR-containing adapter molecule)-1. On the basis of transfection studies, investigators concluded that this protein stimulates type I IFN production; however, investigators initially disagreed as to whether the protein served all TLRs (102) or only TLR3 (103). The phenotype of the mutation revealed a speciﬁc role in signaling via TLR4 and TLR3 (104), a conclusion conﬁrmed by gene targeting (80). Compound mutants lacking MyD88 and homozygous for Lps2 show no detectable responses to LPS and, indeed, show no responses at all to most TLR ligands (104). However, researchers have recently observed that TLR4 can indeed signal by way of TRAM, even in the absence of both MyD88 and TRIF, when activated by the VSV glycoprotein G (P. Georgel, Z. Jiang, S. Kunz, K. Hoebe, E. Janssen, M. Oldstone, and B. Beutler, manuscript submitted).

Inferences Concerning the Function of TRAM BLAST analysis using the TRIF protein sequence as a query disclosed a fourth adapter, termed adapter X (104) and known elsewhere as TICAM-2 (105) or TRAM (81). A speciﬁc role for TRAM in TLR4 signaling was suggested by the ﬁnding that in cells homozygous for the TrifLps2 allele, a fraction of macrophages remained capable of producing TNF in response to LPS. This TRIFindependent population of cells must depend on another adapter, and that adapter could not be MyD88 because all cells (and not merely some of them) were MyD88-

dependent; hence, MyD88 must be expressed and used in all cells (67). Hoebe et al. proposed that adapter X was TRAM, and that TRAM must serve MyD88-independent LPS signaling alongside TRIF, but not dsRNA signaling (67). Gene targeting, performed independently (81), substantiated this proposal (Figures 8 and 9). It is now clear that TRAM (but not TRIF) mediates signals initiated by the VSV glycoprotein G, which depends on CD14 and TLR4. Hence, TRAM can function entirely by itself and is not necessarily codependent on TRIF (P. Georgel, Z. Jiang, S. Kunz, K. Hoebe, E. Janssen, M. Oldstone, and B. Beutler, manuscript submitted).

Heedless Gives Fresh Insight into the Nature of the TLR4 Signaling Complex The Heedless mutation was identiﬁed because it prevented TNF production in response to smooth LPS, that is, LPS with abundant Oglycosylation. Oddly, however, unlike TLR4 mutants (which show no responses to LPS of any kind), Heedless mutants were able to make almost normal quantities of TNF in response to rough LPS or synthetic lipid A (106). Remarkably, lipid A could not activate IRF-3 phosphodimerization, nor could it trigger type I IFN production, suggesting a defect of MyD88-independent signaling (106). Macrophages from Heedless mice also showed diminished responses to all TLR2/TLR6 ligands and were less capable of coping with infection by VSV, as were mice with the P712H mutation of TLR4 (106). The positional cloning of Heedless revealed a premature stop codon in CD14, previously regarded as a coreceptor for LPS signaling in general. The new interpretation of CD14 function is therefore as follows: 1. Rough LPS or lipid A can directly engage TLR4/MD-2 complexes, causing recruitment and activation of www.annualreviews.org • Genetic Analysis of Host Resistance

371

ANRV270-IY24-12

ARI

15 February 2006

2.

3.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

4.

5.

4:15

MyD88, whereas smooth LPS cannot (Figure 10a, left); CD14 nulliﬁes the distinction between rough and smooth isoforms, permitting both types of LPS to activate TLR4 (Figure 10a, right); CD14 is absolutely required for LPSinduced recruitment of TRIF and TRAM, which leads to the activation of IRF-3; CD14 is additionally required for full activation of the TLR2/TLR6 complex (which, as noted below, also requires CD36 for sensing of some ligands) (Figures 8, 9). CD14 and TLR4 mediate sensing of VSV (Figures 8, 9).

Because CD14 is a glycosylphosphoinositol-tethered protein that projects from the surface of the cell, it would seem incapable of directly interacting with the cytoplasmic domains of TLRs and must inﬂuence signaling through interaction with the ectodomains. We may conclude that CD14 coordinates the interaction of ectodomains so as to permit the recruitment of adapter molecules that would otherwise be excluded from that activated complex (Figure 10b). These observations establish that TLR4 is a switch with several possible “on” positions. It can signal in different modes depending on the inducing ligand and on the proﬁle of coreceptors and/or adapters present in the responding cell. Macrophage production of IFN-β, induced by UV-inactivated VSV, is abolished by the Heedless mutation, as well as by mutation of TLR4 or TRAM (or, to a much lesser extent, TRIF), and by the Feckless mutation (described

below). The inducing molecule present in the virion is glycoprotein G rather than a nucleic acid (P. Georgel, Z. Jiang, S. Kunz, K. Hoebe, E. Janssen, M. Oldstone, and B. Beutler, manuscript submitted). Glycoprotein G of VSV is a conserved molecule, broadly represented in Rhabdoviridae, encompassing vertebrate pathogens such as VSV, rabies, and hemorrhagic viral septicemia virus of ﬁsh. As already noted, LPS is chieﬂy a stimulus to mammalian cells. Yet TLR4 is represented in ﬁsh, which do not respond to LPS. We may speculate that its primordial function entailed the detection of viruses, which would explain its connection to the type I IFN induction pathway. This function has been retained in mammals as well, but only in mammals has the ability to sense LPS also been acquired.

CpG1 and the Role of TLRs in Resistance to Viral Infection TLR3, -7, -8, and -9 are at least mostly intracellular molecules, synthesized in the ER, and are eventually transferred to the endosomes, possibly via transit over the cell surface and endocytosis (107). Several mutations affected signaling via these TLRs. The ﬁrst of these to be identiﬁed was one that blocked CpG sensing (CpG1). It was traced to a missense error in the ectodomain of TLR9. The Lps2 mutation had earlier suggested the importance of the TLR3 → TRIF axis in sensing MCMV infection and responding to it (67). The CpG1 mutation was associated with even more severe susceptibility to MCMV (59), as was a MyD88 mutation (59), indicating that the TLR9 → MyD88 and TLR3 → TRIF

−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 8 An overview of the pathways discussed in text. All TLRs, whether located at the cell surface (TLR1, -2, -4, -5, and -6) or within endosomes [TLR3, -7 (or -8), and -9], converge on four adapter proteins, which act in discrete combinations with one another. These adapters lead to activation of protein kinases, including IRAK, IRAK4, TBK1, and IKKi. These kinases ultimately lead to activation of transcription factors, including NF-κB, IRF-3, and IRF-7, which mediate many of the inﬂammatory effects of microbial inducer molecules. Among the thousands of genes modulated by TLR signaling, TNF and IFN-β may be taken as landmarks of the response and are themselves known to be of key importance in subsequent activation of innate and adaptive immune responses. 372

Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12 ARI 15 February 2006 4:15

www.annualreviews.org • Genetic Analysis of Host Resistance

373

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

4:15

axes played nonredundant, additive roles in MCMV resistance. Although both axes contribute to type I IFN production, they may have type I IFN–independent effects as well, some of which may be nonoverlapping and may contribute to resistance (e.g., the activation of IL-12 production, which leads to IFN-γ production). The importance of TLR9 and MyD88 in MCMV resistance was independently analyzed using knockout mice, with similar conclusions (63). Moreover, TLR9 was shown to contribute to the response to herpes simplex virus I (HSV-I) (68).

The Poc mutation causes pronounced susceptibility to MCMV. However, mice are spared the severe immunodeﬁciency disease caused by a MyD88-knockout mutation and are demonstrably more resistant to at least some Gram-positive infections, suggesting that limited signaling through a single TLR complex (TLR2/TLR6) is sufﬁcient to offer fairly strong protection against a wide variety of microbes. The Lkd mutation has little or no effect on susceptibility to MCMV, probably because some signaling via TLR9 is retained in Lkd homozygotes, compared with the situation in CpG1 homozygotes.

Pococurante and Lackadaisical, MyD88 Alleles with Receptor-Selective Function, and Their Interpretation

Pococurante and Insouciant Reveal Something Special about MyD88 Signaling from the TLR2/TLR6 Complex

The Pococurante (Poc) phenotype is one in which there is an absence of MyD88 signaling, with the exception of signals that emanate from the TLR2/TLR6 heterodimer (Z. Jiang, P. Georgel, C. Li, J. Choe, K. Crozat, S. Rutschmann, X. Du, T. Bigby, S. Mudd, S. Sovath, I. Wilson, A. Olson, and B. Beutler, manuscript in preparation). The Lackadaisical (Lkd) phenotype shows normal MyD88 signaling, except for signals from TLR7 and TLR9, which are markedly diminished. Both mutations are missense errors in MyD88, each distinguishable from the knockout allele (Figure 11). The Poc mutation (I179N) affects a surface residue within the TIR domain of MyD88, whereas the Lkd mutation (Y116C) affects a portion of the polypeptide chain between the death domain and the TIR domain. Because the modiﬁed adapter protein is, in each case, capable of transmitting signals from some of the TLRs, it may be inferred that the mutations do not destroy the protein. Rather, they must affect the signaling interfaces that unite the adapters and receptors.

Poc is a particularly interesting mutation because it does disrupt signaling initiated by some TLR2/TLR6 ligands, notably zymosan and LTA. However, it does not disrupt signaling by MALP-2 or PAM2 CSK4 , both diacylated lipopeptides that signal via the TLR2/TLR6 heterodimer and via TLR2 (alone or in conjunction with a still unknown TLR), respectively. Insouciant (Int), a point mutation in the ectodomain of TLR6 (V327A), abolishes zymosan, LTA, and MALP-2 signaling but not PAM2 CSK4 signaling. When engrafted onto TLR4 or TLR9, the Poc mutation abolishes MyD88-dependent signaling. However, when engrafted onto TLR2 or TLR6, the mutation does not affect signaling by MALP-2 or PAM2 CSK4 . Moreover, the BB loop mutation, if engrafted onto MyD88 (P200H), does not prevent downstream signal transduction. These observations suggest that certain TLR2/TLR6 and TLR2/TLRX ligands

−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 9 An overview of the pathways discussed in text, with mutants shown with red X. Mutations shown were induced with ENU, and all but two (Spacey and Feckless) have been identiﬁed positionally. 374

Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12 ARI 15 February 2006 4:15

www.annualreviews.org • Genetic Analysis of Host Resistance

375

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

(MALP-2 and PAM2 CSK4 ) cause a conformational change in the receptor that favors a unique mode of MyD88 association with the receptor complex (Figure 12). Although most TLRs (and for some ligands, TLR2/TLR6 as well) associate with MyD88 in a manner that causes both the BB loop and the Poc site to be included in the signaling interface, diacylated lipopeptides trigger a different form of receptor activation, requiring the participation of neither residue. The TIR domains of MyD88 and the TLR2/TLR6 receptor complex likely engage one another in a reciprocal manner in both cases. However, MyD88 clearly has two different ways of propagating a signal.

Oblivious: A Co-Factor for TLR2/TLR6 Signaling

Figure 10 (a) The Heedless mutation reveals that TLR4 can signal in two modes. Without CD14, Lipid A activates TLR4, causing recruitment of MyD88 and MAL (MyD88 adapter–like) only, whereas smooth LPS (long polysaccharide chain) cannot activate TLR4 at all. In the presence of CD14 (bent solenoid), both smooth and rough LPS can recruit both MyD88/MAL and TRIF/TRAM pathways. (b) Surface view of TLR4 dimer, with or without MyD88. A supramolecular effect on TLR4 aggregation is envisioned, so that with CD14, the geometry of interaction between subunits is correct for recruitment of all adapters. Without CD14, smooth LPS is excluded from the activation complex and Lipid A is only able to stimulate MyD88/MAL recruitment.

376

Beutler et al.

Still another point of complexity concerning TLR2/TLR6 signaling is that the diacylated lipopeptide MALP-2 (but not PAM2 CSK4 ) and LTA are largely dependent on CD36 for signaling, a fact revealed by a premature stop codon in CD36 that caused a phenotype called oblivious (46). Other TLR2/TLR6 ligands (zymosan, for example) are not. Hence, the TLR2/TLR6 complex uses CD36 as a coreceptor for some of its ligands. Perhaps for this reason, oblivious mice show exaggerated susceptibility to Gram-positive infections (46). The oblivious mutation also prevents CD4 priming by cells exposed to antigens in the context of apoptosis. Thus, oblivious impairs one endpoint of the deathdriven immunoadjuvant pathway described earlier in this review (E. Janssen, K. Tabeta, M. Barnes, S. McBride, S. Schoenberger, A. Theoﬁlopoulos, B. Beutler, and K. Hoebe, manuscript submitted). This impairment suggests that CD36 has other immunological functions yet to be found. CD36 is a double-spanning plasma membrane protein and is one of three members of the class B scavenger receptor family (46, 108). It is known as a receptor for thrombospondin (109), for a role in the translocation

ANRV270-IY24-12

ARI

15 February 2006

4:15

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Figure 11 The Pococurante and Lackadaisical phenotypes. MyD88 is required for signaling via most of the TLRs, but not TLR3, and is only partly required for signaling via TLR4. Pococurante mimics a null allele except insofar as it permits full signaling by some of the ligands that use the TLR2/TLR6 and TLR2/TLRX complexes. Lackadaisical is similar to wild-type MyD88, except for diminished signaling via TLR7 and TLR9.

of fatty acids (110, 111), and for its ability to bind oxidized LDL (112). Some endogenous molecules with inﬂammatory effects may stimulate TLR2/TLR6 signaling by way of CD36, although speciﬁc examples of this have not yet been demonstrated.

PanR1: A Point Mutation in TNF on a Pure C57BL/6 Background The codominant PanR1 phenotype was so named for pan resistance to all TLR lig-

ands, where TNF activity was the endpoint of response. The mutation was mapped to the TNF locus itself, and PanR1 was found to be a missense allele (P138T) with dominant inhibitory effects, resulting from the fact that a single mutant subunit prevents engagement of the TNF trimer by the p55 TNF receptor (Figure 13). Little if any TNF activity is evident in homozygotes, which consequently show exaggerated susceptibility to Listeria monocytogenes (S. Rutschmann, K. Hoebe, J. Zalevsky, X. Du,

Figure 12 The Pococurante mutation (Poc) reveals that MyD88 can interact with TLR2 complexes in two conformationally distinct ways. Poc site and the BB loop site mutations inactivate most TLRs, and also inactivate MyD88, preventing interactions when engrafted onto either molecule. The TLR2/TLR6 heterodimer and the TLR2/TLRX heterodimer are unique in that they are capable of interacting with MyD88 in two ways: one that is disrupted by Poc or BB loop mutations in either molecule, and one that is not (special signaling mode). The latter mode of interaction is induced only by diacylated lipopeptides. www.annualreviews.org • Genetic Analysis of Host Resistance

377

ANRV270-IY24-12

ARI

15 February 2006

4:15

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

3d: A Defect of Signaling via TLR3, -7, and -9, Coupled with Defective Exogenous Antigen Presentation

Figure 13 The location of the PanR1 mutation, which affects a surface residue (P138T) on the TNF homotrimer (yellow). Subunits are each shown in different colors.

N. Mann, P. Steed, and B. Beutler, manuscript submitted). Mice with the mutation, unlike mice with a knockout allele of TNF, have entirely normal lymphoid development, including Peyer’s patches and marginal zone B cells in the spleen and lymph nodes (S. Rutschmann, K. Hoebe, J. Zalevsky, X. Du, N. Mann, P. Steed, and B. Beutler, manuscript submitted). This may suggest that the Tnf gene knockout affects expression of the neighboring lymphotoxin (LTa) gene, which is known to be required for lymphoid development, and that TNF itself has no developmental role. Alternatively, an exceedingly low level of TNF activity—present in the homozygous PanR1 mutants but not in homozygous knockout mice—may be sufﬁcient for lymphoid development.

378

Beutler et al.

A combined defect of nucleic acid sensing via TLR3, -7, and -9 was designated Triple D (3d) to denote a triple defect of signaling (113). Probably because nucleic acid sensing via both TLR3 and TLR9 is impaired, the mice are highly susceptible to MCMV infection. However, they also show susceptibility to other organisms, such as Staphylococcus aureus, which is not effectively cleared and sometimes causes metastatic infections when inoculated intradermally. The 3d mutation abolishes exogenous antigen presentation, preventing both priming and cross-priming of T cells in response to ovalbumin, administered either as a particulate antigen (expressed by cells that were induced to undergo apoptosis) or as a soluble protein. However, normal levels of class I and class II MHC proteins are expressed on the surface of 3d APCs (113). The phenotype was shown to result from a missense error affecting the 12-spanning, ER-resident membrane protein UNC-93B, which had no previously recognized function in mammals. The effect of the mutation implies communication between the ER and the endosomal/lysosomal pathway. We speculate that 3d is required for the trafﬁcking of certain membrane proteins within the cell, including proteins required for signaling via TLR3, -7, and -9 and other proteins required for exogenous antigen presentation (Figure 14). A named homolog of UNC-93B (UNC-93A) is also encoded in the mouse genome, and still more distant homologs, as yet unnamed, are identiﬁable as well. These homologs are related to a Caenorhabditis elegans protein, known to be a component of a two-pore potassium channel (114). However, UNC-93B apparently does not fulﬁll this function in mammals (113).

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

Figure 14 Possible mode of action of UNC-93B. The trafﬁc of speciﬁc membrane proteins required for signaling by TLR3, -7, -8, and -9 is postulated to depend on the ER-resident protein UNC-93B. These co-factors for TLR activity, and probably other proteins required for presentation and cross-presentation of exogenous antigens, cannot traverse the normal route from ER to endosome. Note that the TLR-independent cell death–induced activation pathway is also endocytic and, like the TLR pathway, leads to the induction of IFN-β. However, IFN-β induction does not depend on UNC-93B.

Feckless: A Protein Required for TLR3 to Activate NF-κB and for the Response to VSV As previously noted, TRIF activates IRF-3, yielding activation of the IFN-β gene, but also activates NF-κB via an interaction with TRAF-6 (97, 98). The Feckless phenotype was characterized by failure to activate NF-κB in response to dsRNA, although dsRNA could initiate IFN-β production (Z. Jiang & B. Beutler, unpublished data). Macrophages from Feckless mice were, moreover, found to be highly susceptible to VSV, and UV-inactivated VSV was incapable of stimulating a type I IFN response in Feckless cells. The G-glycoprotein of VSV is known to activate a type I IFN response via CD14, TLR4, TRAM, and IRF-7 (P. Georgel, Z. Jiang, S. Kunz, K. Hoebe, E. Janssen, M. Oldstone, and B. Beutler, manuscript submitted), but does not require TRIF to do so. Clearly required to carry a signal from TLR3 and TRIF to activate NF-κB, Feckless may also in-

teract with TRAM in an entirely independent pathway, leading to activation of IRF-7.

Spacey Spacey homozygotes are runted animals with low fertility and were identiﬁed because of diminished ability to respond to TLR9. The mutation has been mapped to chromosome 10 and is unlinked to the TLR9 locus itself. No genes known to be required for TLR signaling reside within the Spacey critical region. However, the mutation remains to be found.

Reverse genetics: the genetic approach that begins with a gene of unknown function and ends with a phenotype TBK1: TANK-binding kinase 1 IKKi: IκB kinase i

REVERSE GENETIC INVESTIGATIONS TBK1 and IKKi, IRF-7 and IRF-5 Intercurrent investigations using reverse genetic methods revealed that TBK1 and IKKi are the protein kinases responsible for activation of IRF-3 (98, 116–118), which is chieﬂy www.annualreviews.org • Genetic Analysis of Host Resistance

379

ARI

15 February 2006

4:15

responsible for induction of IFN-β gene expression caused by LPS or dsRNA. TBK1 appears to be the more important of the two (119). IRF-7, which exists at minimal concentrations in resting cells, is induced (presumably autoinduced) to high concentrations in response to TLR7 or TLR9 stimulation. These TLRs do not signal by way of TRIF or TRAM, and they activate type I IFN production via IRF-7 rather than IRF-3 (120). IRF-5 is required for MyD88-dependent signaling initiated by diverse MyD88-dependent TLRs (121). Investigators have suggested, although not formally proved, that IRF-5 is activated by IRAK-4. Alternatively, IRF-5 may be required for the transcription of genes encoding certain components of the pathway itself.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

Inhibitors of Signaling IRAK-M (122) and SOCS (suppressor of cytokine signaling)-1 (123) function as inhibitors of signaling via TLRs, the former probably inhibiting IRAK/IRAK-2/IRAK-4 signaling and the latter preventing STAT (signal transducer and activator of transcription)1 signaling. The orphan TIR domain receptors SIGIRR (single Ig repeat–related protein) (124) and ST2 (125) exert inhibitory inﬂuences as well, probably by virtue of direct interactions with TLRs. Upregulation of the phosphatase SHIP (Src homology inducible phosphatase)-1, which depends on autocrine stimulation by TGF-β, also inhibits LPS signal transduction, possibly blocking both MyD88-dependent and TRIFdependent pathways (126). The phenomenon of endotoxin tolerance is thus partially explained (127) and appears to operate at several different levels.

Saturation: How Many Genes Serve the Pathways? We can now point to 30 genes that make positive (as opposed to inhibitory) contributions to TLR signal transduction. As outlined above, roughly 40 to 50 genes are expected 380

Beutler et al.

to make nonredundant contributions to signaling, subject to fairly broad limits of error. These ﬁgures might be taken to indicate that 10 to 20 genes remain to be found. We should emphasize that these ﬁgures include only those genes that are required for the production of TNF activity in response to ligands that depend on 7 of the 12 mouse TLRs. If the screen for responses were less focused, and if all the TLRs were stimulated, more targets might be detected.

A BROADER LOOK: THE MCMV RESISTOME AND ITS SIZE Because 11 transmissible mutations were identiﬁed by screening about 20,000 G3 mice, an authentic phenotype is picked up in about 1 of every 1800 mice examined. A substantially higher hit rate is observed in a screen that tests the integrity of all essential systems for the containment of MCMV infection. Approximately 1 in 400 G3 mice score positive in this screen, in which 105 PFU of virus (an inoculum insufﬁcient to affect normal C57BL/6 animals) is administered by an intraperitoneal route, and sickness or death is recorded within the ﬁrst seven days to assure that only innate immunodeﬁciency phenotypes are detected. This corresponds to 1 recessive mutation in every 33 pedigrees examined, and to a genomic footprint that encompasses about 45,000 nucleotides, parceled among about 300 genes. The innate resistome for MCMV thus corresponds to about 1% of the genome. Some of the mutations identiﬁed in a screen for MCMV susceptibility cause cosusceptibility to other viral pathogens. For example, macrophage susceptibility to VSV is commonly observed. But some mutations have a speciﬁc effect on NK cell function, or rather broad effects on vesicle transport, leading to hypopigmentation coupled with immunodeﬁciency. As of this writing, approximately 30 MCMV mutations have been identiﬁed in screening, and 10 mutations are displayed in

ANRV270-IY24-12

ARI

Table 2

15 February 2006

Nine MCMV susceptibility mutations and their phenotypic characteristics

Name

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

4:15

Outcome of MCMV infection

VSV Susceptible?a

IFN-α/β Rescue?b

NK cell defect?

Cloned?

Domino (recessive)

death on day 4 Necrotic spleen

Yes

No

Yes

Yes (STAT1)

Goodnight (recessive)

death on day 2–3 Necrotic liver

n.d.

n.a.

n.d.

No

Solitaire (recessive)

death on day 2–3

n.d.

n.a.

n.d.

No

Slumber (recessive)

death on day 2–3

n.d.

n.a.

n.d.

No

Jinx (recessive)

Elevated viral load in spleen

Yes

Yes

Yes

No

Warmﬂash (co-dominant)

Elevated viral load in spleen

Yes

Yes

No

No

May Day (recessive)

death on day 3

n.d.

n.a.

n.d.

No

Paris (recessive)

Exanthem

n.d.

n.a.

n.d.

No

Moneypenny (recessive)

death on day 6

Yes

Yes

n.d.

No

Havelock (recessive)

Elevated viral load in spleen

No

n.a.

No

No

a

Refers to the susceptibility of macrophages ex vivo. Refers to rescue of the VSV phenotype. n.a., not applicable; n.d., not determined.

b

Table 2. One among these has been positionally cloned. This mutation, Domino, is a missense allele of STAT-1, modifying the DNA-binding domain of the protein and also preventing normal phosphorylation and expression of the protein in cells (K. Crozat, P. Georgel, S. Rutschmann, N. Mann, X. Du, K. Hoebe, and B. Beutler, manuscript submitted). Remarkably, many of the mutations identiﬁed to date cause far greater susceptibility to MCMV than the Domino mutation, which effectively disrupts both type I and type II IFN signaling, as well as IL-27 signaling. We must therefore conclude that some proteins are far more important to the host than the IFNs, given the circumstance of MCMV infection. Clearly, much remains to be learned about antiviral defense.

CONCLUSIONS A classical genetic inquiry led directly to our present understanding of how the immune system detects infection. But beyond pointing to the function of the TLRs and helping

to elucidate their signaling pathways, the genetic approach has done much more as well. Intracellular sensors of microbes also exist. The nucleotide-binding oligomerization domain (NOD)-1 and NOD2 proteins, for example, are believed to function as cytoplasmic sensors of peptidoglycans (128–130). Their general role in immunity was originally discovered by a pure genetic investigation, undertaken in humans rather than in mice (131, 132). The NK receptor Ly49H, which detects the MCMV-encoded protein m157 (133), was identiﬁed by positional cloning as well (2, 3). So, too, was the key transcription factor required for the development of regulatory T cells (134). Mutagenesis is being directly applied to the problem of autoimmunity and has already yielded impressive results (135, 136). What other large-scale phenomena interest us in immunology? This is for the reader to ponder. Sometimes, asking the question—or in this case, seeing the phenomenon, cleverly concealed as it is—may be the most difﬁcult part of the process. www.annualreviews.org • Genetic Analysis of Host Resistance

381

ANRV270-IY24-12

ARI

15 February 2006

4:15

APPENDIX: GLOSSARY OF MUTATIONS Domino (Dom): STAT-1 (missense; null) Feckless (fks): unknown (failure to activate NF-κB with preserved IFN production in response to polyI:C) Heedless (hdl): CD14 (nonsense; null) Insouciant (Int): TLR6 (missense; null) Lackadaisical (Lkd): MyD88 (missense; diminished signaling via TLRs 7 and 9) Lps: TLR4 (missense; null) Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Lps2: TRIF (frameshift; null; codominant or haploinsufﬁcient) Oblivious (Obl): CD36 (nonsense; null) Pococurante (Poc): MyD88 (missense; retains signaling only via TLR2/TLR6 and TLR2/TLRX) PanR1: TNF (missense; extreme hypomorph; dominant) Triple D (3d): unc-93b (missense; null) Spacey (Spy): unknown (diminished TLR9 signaling)

LITERATURE CITED 1. This classic study of the heritability of premature death due to many causes, including infection, formally established that premature death due to infection is more heritable than premature death of any other cause.

382

1. Sorensen TI, Nielsen GG, Andersen PK, Teasdale TW. 1988. Genetic and environmental influences on premature death in adult adoptees. N. Engl. J. Med. 318:727–32 2. Lee SH, Girard S, Macina D, Busa M, Zafer A, et al. 2001. Susceptibility to mouse cytomegalovirus is associated with deletion of an activating natural killer cell receptor of the C-type lectin superfamily. Nat. Genet. 28:42–45 3. Brown MG, Dokun AO, Heusel JW, Smith HR, Beckman DL, et al. 2001. Vital involvement of a natural killer cell activation receptor in resistance to viral infection. Science 292:934–37 4. Desrosiers MP, Kielczewska A, Loredo-Osti JC, Adam SG, Makrigiannis AP, et al. 2005. Epistasis between mouse Klra and major histocompatibility complex class I loci is associated with a new mechanism of natural killer cell-mediated innate resistance to cytomegalovirus infection. Nat. Genet. 37:593–99 5. Lee SH, Zafer A, De Repentigny Y, Kothary R, Tremblay ML, et al. 2003. Transgenic expression of the activating natural killer receptor Ly49H confers resistance to cytomegalovirus in genetically susceptible mice. J. Exp. Med. 197:515–26 6. Samson M, Libert F, Doranz BJ, Rucker J, Liesnard C, et al. 1996. Resistance to HIV1 infection in caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Nature 382:722–25 7. Martinson JJ, Chapman NH, Rees DC, Liu YT, Clegg JB. 1997. Global distribution of the CCR5 gene 32-basepair deletion. Nat. Genet. 16:100–3 8. Pancer Z, Amemiya CT, Ehrhardt GR, Ceitlin J, Gartland GL, Cooper MD. 2004. Somatic diversiﬁcation of variable lymphocyte receptors in the agnathan sea lamprey. Nature 430:174–80 Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

9. Beutler B, Crozat K, Koziol JA, Georgel P. 2005. Genetic dissection of innate immunity to infection: the mouse cytomegalovirus model. Curr. Opin. Immunol. 17:36–43 10. Lemaitre B, Nicolas E, Michaut L, Reichhart JM, Hoffmann JA. 1996. The dorsoventral regulatory gene cassette spatzle/Toll/cactus controls the potent antifungal response in Drosophila adults. Cell 86:973–83 11. Anderson KV, Bokla L, Nusslein-Volhard C. 1985. Establishment of dorsal-ventral polarity in the Drosophila embryo: the induction of polarity by the Toll gene product. Cell 42:791–98 12. Beutler B. 2004. Innate immunity: an overview. Mol. Immunol. 40:845–59 13. Presti RM, Pollock JL, Dal Canto AJ, O’Guin AK, Virgin HW. 1998. Interferon γ regulates acute and latent murine cytomegalovirus infection and chronic disease of the great vessels. J. Exp. Med. 188:577–88 14. Leadbetter EA, Rifkin IR, Hohlbaum AM, Beaudette BC, Shlomchik MJ, MarshakRothstein A. 2002. Chromatin-IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors. Nature 416:603–7 15. Beutler B. 2004. Inferences, questions and possibilities in Toll-like receptor signalling. Nature 430:257–63 16. Luderitz O, Galanos C, Lehmann V, Nurminen M, Rietschel ET, et al. 1973. Lipid A: chemical structure and biological activity. J. Infect. Dis. 128:29 17. Berczi I, Bertok L, Bereznai T. 1966. Comparative studies on the toxicity of Escherichia coli lipopolysaccharide endotoxin in various animal species. Can. J. Microbiol. 12:1070– 71 18. Michalek SM, Moore RN, McGhee JR, Rosenstreich DL, Mergenhagen SE. 1980. The primary role of lymphoreticular cells in the mediation of host responses to bacterial endotoxin. J. Infec. Dis. 141:55–63 19. Beutler B, Milsark IW, Cerami A. 1985. Passive immunization against cachectin/tumor necrosis factor (TNF) protects mice from the lethal effect of endotoxin. Science 229:869– 71 20. Karaghiosoff M, Steinborn R, Kovarik P, Kriegshauser G, Baccarini M, et al. 2003. Central role for type I interferons and Tyk2 in lipopolysaccharide-induced endotoxin shock. Nat. Immunol. 4:471–77 21. Heppner G, Weiss DW. 1965. High susceptibility of strain A mice to endotoxin and endotoxin-red blood cell mixtures. J. Bacteriol. 90:696–703 22. Watson J, Riblet R. 1974. Genetic control of responses to bacterial lipopolysaccharides in mice. I. Evidence for a single gene that inﬂuences mitogenic and immunogenic respones to lipopolysaccharides. J. Exp. Med. 140:1147–61 23. Watson J, Riblet R. 1975. Genetic control of responses to bacterial lipopolysaccharides in mice. II. A gene that inﬂuences a membrane component involved in the activation of bone marrow-derived lymphocytes by lipipolysaccharides. J. Immunol. 114:1462–68 24. Watson J, Riblet R, Taylor BA. 1977. The response of recombinant inbred strains of mice to bacterial lipopolysaccharides. J. Immunol. 118:2088–93 25. Skidmore BJ, Chiller JM, Weigle WO, Riblet R, Watson J. 1976. Immunologic properties of bacterial lipopolysaccharide (LPS). III. Genetic linkage between the in vitro mitogenic and in vivo adjuvant properties of LPS. J. Exp. Med. 143:143–50 26. Condie RM, Zak SJ, Good RA. 1955. Effect of meningococcal endotoxin on the immune response. Proc. Soc. Exp. Biol. Med. 90:355–60 27. Coutinho A, Meo T. 1978. Genetic basis for unresponsiveness to lipopolysaccharide in C57BL/10Cr mice. Immunogenetics 7:17–24 www.annualreviews.org • Genetic Analysis of Host Resistance

10. Demonstration that the Toll protein of Drosophila, and upstream and downstream signaling components, are essential for activation of Drosomycin synthesis.

21. The first demonstration that mice of the C3H/HeJ substrain are highly resistant to the lethal effect of LPS.

383

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

33. This work first revealed that mammalian TLRs act as specific sensors of microbial molecules and are required for recognition and response to infection.

384

4:15

28. O’Brien AD, Rosenstreich DL, Scher I, Campbell GH, MacDermott RP, Formal SB. 1980. Genetic control of susceptibility to Salmonella typhimurium in mice: role of the LPS gene. J. Immunol. 124:20–4 29. Hagberg L, Hull R, Hull S, McGhee JR, Michalek SM, Svanborg Eden C. 1984. Difference in susceptibility to Gram-negative urinary tract infection between C3H/HeJ and C3H/HeN mice. Infect. Immun. 46:839–44 30. Woods JP, Frelinger JA, Warrack G, Cannon JG. 1988. Mouse genetic locus Lps inﬂuences susceptibility to Neisseria meningitidis infection. Infect. Immun. 56:1950–55 31. Macela A, Stulik J, Hernychova L, Kroca M, Krocova Z, Kovarova H. 1996. The immune response against Francisella tularensis live vaccine strain in Lpsn and Lpsd mice. FEMS Immunol. Med. Microbiol. 13:235–38 32. Shakhov AN, Collart MA, Vassalli P, Nedospasov SA, Jongeneel CV. 1990. κB-type enhancers are involved in lipopolysaccharide-mediated transcriptional activation of the tumor necrosis factor α gene in primary macrophages. J. Exp. Med. 171:35–47 33. Poltorak A, He X, Smirnova I, Liu M-Y, Van Huffel C, et al. 1998. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282:2085–88 34. Nomura N, Miyajima N, Sazuka T, Tanaka A, Kawarabayasi Y, et al. 1994. Prediction of the coding sequences of unidentiﬁed human genes. I. The coding sequences of 40 new genes (KIAA0001-KIAA0040) deduced by analysis of randomly sampled cDNA clones from human immature myeloid cell line KG-1. DNA Res. 1:27–35 35. Taguchi T, Mitcham JL, Dower SK, Sims JE, Testa JR. 1996. Chromosomal localization of TIL, a gene encoding a protein related to the Drosophila transmembrane receptor Toll, to human chromosome 4p14. Genome 32:486–88 36. Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388:394–97 37. Rock FL, Hardiman G, Timans JC, Kastelein RA, Bazan JF. 1998. A family of human receptors structurally related to Drosophila Toll. Proc. Natl. Acad. Sci. USA 95:588– 93 38. Rutschmann S, Kilinc A, Ferrandon D. 2002. Cutting edge: the Toll pathway is required for resistance to Gram-positive bacterial infections in Drosophila. J. Immunol. 168:1542– 46 39. Chaturvedi MM, Higuchi M, Aggarwal BB. 1994. Effect of tumor necrosis factors, interferons, interleukins, and growth factors on the activation of NF-κB: evidence for lack of correlation with cell proliferation. Lymphokine Cytokine Res. 13:309–13 40. Matsumoto S, Tsuji-Takayama K, Aizawa Y, Koide K, Takeuchi M, et al. 1997. Interleukin-18 activates NF-κB in murine T helper type 1 cells. Biochem. Biophys. Res. Commun. 234:454–57 41. Robinson D, Shibuya K, Mui A, Zonin F, Murphy E, et al. 1997. IGIF does not drive Th1 development but synergizes with IL-12 for interferon-γ production and activates IRAK and NFκB. Immunity 7:571–81 42. Choe J, Kelker MS, Wilson IA. 2005. Crystal structure of human Toll-like receptor 3 (TLR3) ectodomain. Science 309:581–85 43. Xu Y, Tao X, Shen B, Horng T, Medzhitov R, et al. 2000. Structural basis for signal transduction by the Toll/interleukin-1 receptor domains. Nature 408:111–15 44. Dunne A, Ejdeback M, Ludidi PL, O’Neill LA, Gay NJ. 2003. Structural complementarity of Toll/interleukin-1 receptor identity regions in Toll-like receptors and the adaptors Mal and MyD88. J. Biol. Chem. 278:41443–51 Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

45. Schneider DS, Hudson KL, Lin TY, Anderson KV. 1991. Dominant and recessive mutations deﬁne functional domains of Toll, a transmembrane protein required for dorsalventral polarity in the Drosophila embryo. Genes Dev. 5:797–807 46. Hoebe K, Georgel P, Rutschmann S, Du X, Mudd S, et al. 2005. CD36 is a sensor of diacylglycerides. Nature 433:523–27 47. Wright SD, Ramos RA, Tobias PS, Ulevitch RJ, Mathison JC. 1990. CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 249:1431–33 48. Nagai Y, Akashi S, Nagafuku M, Ogata M, Iwakura Y, et al. 2002. Essential role of MD-2 in LPS responsiveness and TLR4 distribution. Nat. Immunol. 3:667–72 49. Takeuchi O, Sato S, Horiuchi T, Hoshino K, Takeda K, et al. 2002. Cutting edge: role of Toll-like receptor 1 in mediating immune response to microbial lipoproteins. J. Immunol. 169:10–14 50. Takeuchi O, Kawai T, Muhlradt PF, Morr M, Radolf JD, et al. 2001. Discrimination of bacterial lipoproteins by Toll-like receptor 6. Int. Immunol. 13:933–40 51. Buwitt-Beckmann U, Heine H, Wiesmuller KH, Jung G, Brock R, et al. 2005. Toll-like receptor 6-independent signaling by diacylated lipopeptides. Eur. J. Immunol. 35:282– 89 52. Hayashi F, Smith KD, Ozinsky A, Hawn TR, Yi EC, et al. 2001. The innate immune response to bacterial ﬂagellin is mediated by Toll-like receptor 5. Nature 410:1099–103 53. Hemmi H, Kaisho T, Takeuchi O, Sato S, Sanjo H, et al. 2002. Small anti-viral compounds activate immune cells via the TLR7 MyD88-dependent signaling pathway. Nat. Immunol. 3:196–200 54. Heil F, Hemmi H, Hochrein H, Ampenberger F, Kirschning C, et al. 2004. Speciesspeciﬁc recognition of single-stranded RNA via Toll-like receptor 7 and 8. Science 303:1526–29 55. Diebold SS, Kaisho T, Hemmi H, Akira S, Reis e Sousa C. 2004. Innate antiviral responses by means of TLR7-mediated recognition of single-stranded RNA. Science 303:1529–31 56. Lund JM, Alexopoulou L, Sato A, Karow M, Adams NC, et al. 2004. Recognition of single-stranded RNA viruses by Toll-like receptor 7. Proc. Natl. Acad. Sci. USA 101:5598– 603 57. Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, et al. 2000. A Toll-like receptor recognizes bacterial DNA. Nature 408:740–45 58. Bauer S, Kirschning CJ, Hacker H, Redecke V, Hausmann S, et al. 2001. Human TLR9 confers responsiveness to bacterial DNA via species-speciﬁc CpG motif recognition. Proc. Natl. Acad. Sci. USA 98:9237–42 59. Tabeta K, Georgel P, Janssen E, Du X, Hoebe K, et al. 2004. Toll-like receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection 3. Proc. Natl. Acad. Sci. USA 101:3516–21 60. Zhang D, Zhang G, Hayden MS, Greenblatt MB, Bussey C, et al. 2004. A Toll-like receptor that prevents infection by uropathogenic bacteria. Science 303:1522–26 61. Yarovinsky F, Zhang D, Andersen JF, Bannenberg GL, Serhan CN, et al. 2005. TLR11 activation of dendritic cells by a protozoan proﬁlin-like protein. Science 308:1626–29 62. Takeuchi O, Hoshino K, Akira S. 2000. Cutting edge: TLR2-deﬁcient and MyD88deﬁcient mice are highly susceptible to staphylococcus aureus infection. J. Immunol. 165:5392–96 63. Krug A, French AR, Barchet W, Fischer JA, Dzionek A, et al. 2004. TLR9-dependent recognition of MCMV by IPC and DC generates coordinated cytokine responses that activate antiviral NK cell function. Immunity 21:107–19 www.annualreviews.org • Genetic Analysis of Host Resistance

46. The forward genetic approach revealed an essential function of CD36, showing that it acts as a coreceptor for specific ligands of the TLR2/TLR6 complex.

385

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

67. An ENU-induced mutation (Lps2) was the first to reveal the function of the adapter protein TRIF, later confirmed by gene knockout.

79. The long-known adjuvant effect of LPS, biochemically defined by the upregulation of costimulatory molecules, was ascribed to the MyD88independent pathway.

386

4:15

64. Kurt-Jones EA, Popova L, Kwinn L, Haynes LM, Jones LP, et al. 2000. Pattern recognition receptors TLR4 and CD14 mediate response to respiratory syncytial virus. Nat. Immunol. 1:398–401 65. Rassa JC, Meyers JL, Zhang Y, Kudaravalli R, Ross SR. 2002. Murine retroviruses activate B cells via interaction with Toll-like receptor 4. Proc. Natl. Acad. Sci. USA 99:2281– 86 66. Jude BA, Pobezinskaya Y, Bishop J, Parke S, Medzhitov RM, et al. 2003. Subversion of the innate immune system by a retrovirus. Nat. Immunol. 4:573–78 67. Hoebe K, Du X, Georgel P, Janssen E, Tabeta K, et al. 2003. Identification of Lps2 as a key transducer of MyD88-independent TIR signaling. Nature 424:743–48 68. Lund J, Sato A, Akira S, Medzhitov R, Iwasaki A. 2003. Toll-like receptor 9-mediated recognition of herpes simplex virus-2 by plasmacytoid dendritic cells. J. Exp. Med. 198:513–20 69. Crozat K, Beutler B. 2004. TLR7: A new sensor of viral infection. Proc. Natl. Acad. Sci. USA 101:6835–36 70. Ehl S, Bischoff R, Ostler T, Vallbracht S, Schulte-Monting J, et al. 2004. The role of Toll-like receptor 4 versus interleukin-12 in immunity to respiratory syncytial virus. Eur. J. Immunol. 34:1146–53 71. Youmans AS, Youmans GP. 1969. Factors affecting immunogenic activity of mycobacterial ribosomal and ribonucleic acid preparations. J. Bacteriol. 99:42–50 72. Youmans GP, Youmans AS. 1969. Allergenicity of mycobacterial ribosomal and ribonucleic acid preparations in mice and guinea pigs. J. Bacteriol. 97:134–39 73. Schmidtke JR, Johnson AG. 1971. Regulation of the immune system by synthetic polynucleotides. I. Characteristics of adjuvant action on antibody synthesis. J. Immunol. 106:1191–200 74. Capanna SL, Kong YM. 1974. Further studies on the prevention of tolerance induction by poly A:U. Immunology 27:647–53 75. Nauciel C, Fleck J, Martin JP, Mock M, Nguyen-Huy H. 1974. Adjuvant activity of bacterial peptidoglycans on the production of delayed hypersensitivity and on antibody response. Eur. J. Immunol. 4:352–56 76. Bultmann B, Finger H, Heymer B, Schachenmayr W, Hof H, Haferkamp O. 1975. Adjuvancy of streptococcal nucleic acids Z. Immunitatsforsch. Exp. Klin. Immunol. 148:425– 30 77. Borriello F, Sethna MP, Boyd SD, Schweitzer AN, Tivol EA, et al. 1997. B7–1 and B7–2 have overlapping, critical roles in immunoglobulin class switching and germinal center formation. Immunity 6:303–13 78. Kaisho T, Takeuchi O, Kawai T, Hoshino K, Akira S. 2001. Endotoxin-induced maturation of MyD88-deﬁcient dendritic cells. J. Immunol. 166:5688–94 79. Hoebe K, Janssen EM, Kim SO, Alexopoulou L, Flavell RA, et al. 2003. Upregulation of costimulatory molecules induced by lipopolysaccharide and doublestranded RNA occurs by Trif-dependent and Trif-independent pathways. Nat. Immunol. 4:1223–29 80. Yamamoto M, Sato S, Hemmi H, Hoshino K, Kaisho T, et al. 2003. Role of adapter TRIF in the MyD88-independent Toll-like receptor signaling pathway. Science 301:640– 43 81. Yamamoto M, Sato S, Hemmi H, Uematsu S, Hoshino K, et al. 2003. TRAM is speciﬁcally involved in the Toll-like receptor 4-mediated MyD88-independent signaling pathway. Nat. Immunol. 4:1144–50 Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

82. Uematsu S, Sato S, Yamamoto M, Hirotani T, Kato H, et al. 2005. Interleukin-1 receptorassociated kinase-1 plays an essential role for Toll-like receptor (TLR)7- and TLR9mediated interferon-α induction. J. Exp. Med. 201:915–23 83. Gallucci S, Lolkema M, Matzinger P. 1999. Natural adjuvants: endogenous activators of dendritic cells. Nat. Med. 5:1249–55 84. Kile BT, Hentges KE, Clark AT, Nakamura H, Salinger AP, et al. 2003. Functional genetic analysis of mouse chromosome 11. Nature 425:81–86 85. Concepcion D, Seburn KL, Wen G, Frankel WN, Hamilton BA. 2004. Mutation rate and predicted phenotypic target sizes in ethylnitrosourea-treated mice. Genetics 168:953–59 86. Justice MJ, Noveroske JK, Weber JS, Zheng B, Bradley A. 1999. Mouse ENU mutagenesis. Hum. Mol. Genet. 8:1955–63 87. Sagai T, Hosoya M, Mizushina Y, Tamura M, Shiroishi T. 2005. Elimination of a longrange cis-regulatory module causes complete loss of limb-speciﬁc Shh expression and truncation of the mouse limb. Development 132:797–803 88. Sagai T, Masuya H, Tamura M, Shimizu K, Yada Y, et al. 2004. Phylogenetic conservation of a limb-speciﬁc, cis-acting regulator of Sonic hedgehog (Shh). Mamm. Genome 15:23–34 89. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, et al. 1998. Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function. Immunity 9:143– 50 90. Croston GE, Cao Z, Goeddel DV. 1995. NF-κB activation by interleukin-1 (IL-1) requires an IL-1 receptor-associated protein kinase activity. J. Biol. Chem. 270:16514–17 91. Cao Z, Henzel WJ, Gao X. 1996. IRAK: a kinase associated with the interleukin-1 receptor. Science 271:1128–31 92. Cao Z, Xiong J, Takeuchi M, Kurama T, Goeddel DV. 1996. TRAF6 is a signal transducer for interleukin-1. Nature 383:443–46 93. Wasserman SA. 1993. A conserved signal transduction pathway regulating the activity of the rel-like proteins dorsal and NF-κB. Mol. Biol. Cell 4:767–71 94. Gay NJ, Keith FJ. 1991. Drosophila Toll and IL-1 receptor. Nature 351:355–56 95. Kawai T, Adachi O, Ogawa T, Takeda K, Akira S. 1999. Unresponsiveness of MyD88deﬁcient mice to endotoxin. Immunity 11:115–22 96. Yamamoto M, Sato S, Hemmi H, Sanjo H, Uematsu S, et al. 2002. Essential role for TIRAP in activation of the signalling cascade shared by TLR2 and TLR4. Nature 420:324–29 97. Jiang Z, Mak TW, Sen G, Li X. 2004. Toll-like receptor 3-mediated activation of NF-κB and IRF-3 diverges at Toll-IL-1 receptor domain-containing adapter inducing IFN-β. Proc. Natl. Acad. Sci. USA 101:3533–38 98. Sato S, Sugiyama M, Yamamoto M, Watanabe Y, Kawai T, et al. 2003. Toll/IL-1 receptor domain-containing adaptor inducing IFN-β (TRIF) associates with TNF receptorassociated factor 6 and TANK-binding kinase 1, and activates two distinct transcription factors, NF-κB and IFN-regulatory factor-3, in the Toll-like receptor signaling. J. Immunol. 171:4304–10 99. Sakurai H, Miyoshi H, Toriumi W, Sugita T. 1999. Functional interactions of transforming growth factor β-activated kinase 1 with IκB kinases to stimulate NF-κB activation. J. Biol. Chem. 274:10641–48 100. Takaesu G, Surabhi RM, Park KJ, Ninomiya-Tsuji J, Matsumoto K, Gaynor RB. 2003. TAK1 is critical for IκB kinase-mediated activation of the NF-κB pathway. J. Mol. Biol. 326:105–15 www.annualreviews.org • Genetic Analysis of Host Resistance

84. A fundamental study of ENU mutagenesis in a balancer chromosome system, designed to trap lethal alleles of genes within a restricted genomic interval.

85. A highly informative study defining the ENU-induced mutation rate and relating it to the production of phenotype.

387

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

113. The forward genetic method discloses that UNC-93B, a protein with no previously known function, is required for diverse immunologic events and suggests communication between the ER and endosomes.

388

4:15

101. Hoebe K, Du X, Goode J, Mann N, Beutler B. 2003. Lps2: a new locus required for responses to lipopolysaccharide, revealed by germline mutagenesis and phenotypic screening. J. Endotoxin. Res. 9:250–55 102. Yamamoto M, Sato S, Mori K, Hoshino K, Takeuchi O, et al. 2002. Cutting edge: a novel Toll/IL-1 receptor domain-containing adapter that preferentially activates the IFN-β promoter in the Toll-like receptor signaling. J. Immunol. 169:6668–72 103. Oshiumi H, Matsumoto M, Funami K, Akazawa T, Seya T. 2003. TICAM-1, an adaptor molecule that participates in Toll-like receptor 3-mediated interferon-β induction. Nat. Immunol. 4:161–71 104. Park JM, Greten FR, Li ZW, Karin M. 2002. Macrophage apoptosis by anthrax lethal factor through p38 MAP kinase inhibition. Science 297:2048–51 105. Oshiumi H, Sasai M, Shida K, Fujita T, Matsumoto M, Seya T. 2003. TIR-containing adapter molecule (TICAM)-2, a bridging adapter recruiting to Toll-like receptor 4 TICAM-1 that induces interferon-β. J. Biol. Chem. 278:49751–62 106. Jiang Z, Georgel P, Du X, Shamel L, Sovath S, et al. 2005. CD14 is required for MyD88independent LPS signaling. Nat. Immunol. 6:565–70 107. Latz E, Schoenemeyer A, Visintin A, Fitzgerald KA, Monks BG, et al. 2004. TLR9 signals after translocating from the ER to CpG DNA in the lysosome. Nat. Immunol. 5:190–98 108. Calvo D, Dopazo J, Vega MA. 1995. The CD36, CLA-1 (CD36L1), and LIMPII (CD36L2) gene family: cellular distribution, chromosomal location, and genetic evolution. Genomics 25:100–6 109. Asch AS, Barnwell J, Silverstein RL, Nachman RL. 1987. Isolation of the thrombospondin membrane receptor 3. J. Clin. Invest. 79:1054–61 110. Abumrad NA, el Maghrabi MR, Amri EZ, Lopez E, Grimaldi PA. 1993. Cloning of a rat adipocyte membrane protein implicated in binding or transport of long-chain fatty acids that is induced during preadipocyte differentiation. Homology with human CD36 1. J. Biol. Chem. 268:17665–68 111. Aitman TJ, Glazier AM, Wallace CA, Cooper LD, Norsworthy PJ, et al. 1999. Identiﬁcation of Cd36 (Fat) as an insulin-resistance gene causing defective fatty acid and glucose metabolism in hypertensive rats. Nat. Genet. 21:76–83 112. Endemann G, Stanton LW, Madden KS, Bryant CM, White RT, Protter AA. 1993. CD36 is a receptor for oxidized low density lipoprotein 2. J. Biol. Chem. 268:11811–16 113. Tabeta K, Hoebe K, Janssen EM, Du X, Georgel P, et al. 2006. 3d, a missense mutation that disrupts exogenous antigen presentation and signaling via Toll-like receptors 3, 7, and 9. Nat. Immunol. In press 114. de la Cruz IP, Levin JZ, Cummins C, Anderson P, Horvitz HR. 2003. sup-9, sup-10, and unc-93 may encode components of a two-pore K+ channel that coordinates muscle contraction in Caenorhabditis elegans. J. Neurosci. 23:9133–45 115. Meylan E, Burns K, Hofmann K, Blancheteau V, Martinon F, et al. 2004. RIP1 is an essential mediator of Toll-like receptor 3-induced NF-κB activation. Nat. Immunol. 5:503–7 116. Fitzgerald KA, McWhirter SM, Faia KL, Rowe DC, Latz E, et al. 2003. IKKε and TBK1 are essential components of the IRF-3 signaling pathway. Nat. Immunol. 4:491–96 117. McWhirter SM, Fitzgerald KA, Rosains J, Rowe DC, Golenbock DT, Maniatis T. 2004. IFN-regulatory factor 3-dependent gene expression is defective in Tbk1-deﬁcient mouse embryonic ﬁbroblasts. Proc. Natl. Acad. Sci. USA 101:233–38 118. tenOever BR, Sharma S, Zou W, Sun Q, Grandvaux N, et al. 2004. Activation of TBK1 and IKKε kinases by vesicular stomatitis virus infection and the role of viral ribonucleoprotein in the development of interferon antiviral immunity. J. Virol. 78:10636–49 Beutler et al.

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-12

ARI

15 February 2006

4:15

119. Hemmi H, Takeuchi O, Sato S, Yamamoto M, Kaisho T, et al. 2004. The roles of two IκB kinase-related kinases in lipopolysaccharide and double stranded RNA signaling and viral infection. J. Exp. Med. 199:1641–50 120. Kawai T, Sato S, Ishii KJ, Coban C, Hemmi H, et al. 2004. Interferon-α induction through Toll-like receptors involves a direct interaction of IRF-7 with MyD88 and TRAF6. Nat. Immunol. 5:1061–68 121. Takaoka A, Yanai H, Kondo S, Duncan G, Negishi H, et al. 2005. Integral role of IRF-5 in the gene induction programme activated by Toll-like receptors. Nature 434:243–49 122. Kobayashi K, Hernandez LD, Galan JE, Janeway CA Jr, Medzhitov R, Flavell RA. 2002. IRAK-M is a negative regulator of Toll-like receptor signaling. Cell 110:191–202 123. Kinjyo I, Hanada T, Inagaki-Ohara K, Mori H, Aki D, et al. 2002. SOCS1/JAB is a negative regulator of LPS-induced macrophage activation. Immunity 17:583–91 124. Wald D, Qin J, Zhao Z, Qian Y, Naramura M, et al. 2003. SIGIRR, a negative regulator of Toll-like receptor-interleukin 1 receptor signaling. Nat. Immunol. 4:920–27 125. Brint EK, Xu D, Liu H, Dunne A, McKenzie AN, et al. 2004. ST2 is an inhibitor of interleukin 1 receptor and Toll-like receptor 4 signaling and maintains endotoxin tolerance. Nat. Immunol. 5:373–79 126. Sly LM, Rauh MJ, Kalesnikoff J, Song CH, Krystal G. 2004. LPS-induced upregulation of SHIP is essential for endotoxin tolerance. Immunity 21:227–39 127. Beutler B. 2004. SHIP, TGF-β, and endotoxin tolerance. Immunity 21:134–35 128. Girardin SE, Boneca IG, Carneiro LA, Antignac A, Jehanno M, et al. 2003. Nod1 detects a unique muropeptide from Gram-negative bacterial peptidoglycan. Science 300:1584–87 129. Girardin SE, Boneca IG, Viala J, Chamaillard M, Labigne A, et al. 2003. Nod2 is a general sensor of peptidoglycan through muramyl dipeptide (MDP) detection. J. Biol. Chem. 278:8869–72 130. Tanabe T, Chamaillard M, Ogura Y, Zhu L, Qiu S, et al. 2004. Regulatory regions and critical residues of NOD2 involved in muramyl dipeptide recognition. EMBO J. 23:1587–97 131. Hugot JP, Laurent-Puig P, Gower-Rousseau C, Olson JM, Lee JC, et al. 1996. Mapping of a susceptibility locus for Crohn’s disease on chromosome 16. Nature 379:821–23 132. Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, et al. 2001. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature 411:599– 603 133. Smith HR, Heusel JW, Mehta IK, Kim S, Dorner BG, et al. 2002. Recognition of a virus-encoded ligand by a natural killer cell activation receptor. Proc. Natl. Acad. Sci. USA 99:8826–31 134. Brunkow ME, Jeffery EW, Hjerrild KA, Paeper B, Clark LB, et al. 2001. Disruption of a new forkhead/winged-helix protein, scurﬁn, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat. Genet. 27:68–73 135. Vinuesa CG, Cook MC, Angelucci C, Athanasopoulos V, Rui L, et al. 2005. A RINGtype ubiquitin ligase family member required to repress follicular helper T cells and autoimmunity. Nature 435:452–58 136. Goodnow CC, Sprent J, de St Groth BF, Vinuesa CG. 2005. Cellular and genetic mechanisms of self tolerance and autoimmunity. Nature 435:590–97

www.annualreviews.org • Genetic Analysis of Host Resistance

389

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:353-389. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni,1,3,∗ Steven M. Chan,1,∗ Michael Kattah,1,∗ Jessica D. Tenenbaum,1,2,∗ Atul J. Butte,2,4 and Paul J. Utz1,2 1

Department of Medicine, Division of Immunology and Rheumatology; 2 Department of Medicine, Stanford Medical Informatics; 3 Department of Pediatrics, Division of Rheumatology; 4 Department of Pediatrics, Division of Endocrinology; Stanford University School of Medicine, Stanford, California 94305; email: [email protected]

Annu. Rev. Immunol. 2006. 24:391–418 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090709 c 2006 by Copyright Annual Reviews. All rights reserved ∗ The ﬁrst four authors (I.B., S.M.C., M.K., and J.D.T.) contributed equally to this manuscript and are listed in alphabetical order.

0732-0582/06/0423-0391$20.00

Key Words autoimmunity, proteomics, autoantibodies, cytokines, signaling

Abstract Several proteomics platforms have emerged in the past decade that show great promise for ﬁlling in the many gaps that remain from earlier studies of the genome and from the sequencing of the human genome itself. This review describes applications of proteomics technologies to the study of autoimmune diseases. We focus largely on biased technology platforms that are capable of analyzing a large panel of known analytes, as opposed to techniques such as twodimensional gel electrophoresis (2DIGE) or mass spectroscopy that represent unbiased approaches (as reviewed in 1). At present, the main analytes that can be systematically studied in autoimmunity include autoantibodies, cytokines and chemokines, components of signaling pathways, and cell-surface receptors. We review the most commonly used platforms for such studies, citing important discoveries and limitations that exist. We conclude by reviewing advances in biomedical informatics that will eventually allow the human proteome to be deciphered.

391

ANRV270-IY24-13

ARI

15 February 2006

0:17

INTRODUCTION TO PROTEIN MICROARRAYS

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Protein array/microarray: any platform in which a large number of proteins are immobilized on a solid support in a spatially (planar arrays) or spectrally (bead-based arrays) addressable manner ELISA: enzyme-linked immunosorbent assay

Table 1

Protein microarrays represent a validated platform for proﬁling protein levels and their post-translational modiﬁcations at a scale that is beyond what traditional techniques such as Western immunoblotting or enzyme-linked immunosorbent assays (ELISA) can realistically achieve. Protein microarrays have the potential to impact not only the broadly deﬁned ﬁeld of proteomics, but also other more deﬁned biological disciplines, including immunology. As we describe here, the use of protein microarrays in immunology has largely been limited to the proﬁling of two main classes of proteins: (a) secreted factors (e.g., cytokines, chemokines, or growth factors) and (b) autoantibodies. Multiplexed analysis of in-

tracellular proteins in a microarray format has proven to be a much more challenging task. The reverse-phase lysate (RPL) microarray platform stands out from other microarray platforms as the one with the greatest potential to achieve that goal. In the following sections, we review the work that has been published using each of the above mentioned technologies (summarized in Table 1) and speculate on their potential impact in the study of autoimmunity.

ANTIGEN MICROARRAYS The relative frequency of autoreactive T and B cells in circulation is low, estimated at less than 1:10,000 lymphocytes (2–5), making the isolation and study of individual autoreactive

Proteomics technologies for autoimmune disease research

Method

Format

Advantages

Autoantigen microarrays

Print peptides, puriﬁed/ recombinant autoantigens, lipids, carbohydrates, DNA; probe with serum/ﬂuid, detect autoantibodies

Proﬁle many markers on small amounts of sample

Biased, currently limited to known autoantigens, controversial correlation with pathogenesis

Antigen-speciﬁc therapy, disease classiﬁcation, monitor response to therapy

(8, 33)

Reverse-phase (lysate) protein microarrays

Print whole-cell lysates; proﬁle with speciﬁc antibodies

Proﬁle multiple samples with ≥100 antibodies, need only small amounts of lysate, monitor signal transduction events

Antibody cross-reactivity, multiple slides

Signaling defects in autoimmune cells (Tregs, autoreactive lymphocytes)

(1–3, 88–90)

Forwardphase (lysate) protein microarrays

Print/coat beads with antibodies against intracellular target, probe with lysate

Proﬁle one or two samples simultaneously with many antibodies

Poor performance; denaturation of antibody; detection difﬁcult (modify sample or generate two antibodies for every target); antibody cross-reactivity

Analysis of intracellular markers during apoptosis, could be used to proﬁle intracellular disease markers

(135)

392

Balboni et al.

Limitations

Applications (Potential or Reported)

References

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

cells problematic. However, a hallmark of many autoimmune diseases is the production of high-titer, high-speciﬁcity autoantibodies directed against a variety of evolutionarily conserved molecules (6). These autoantibodies are readily detectable with various immunoassays, including ELISA, Western blotting, and immunoprecipitation. Although these assays do not directly provide data regarding the speciﬁcity of autoreactive T cells, there is a high degree of concordance between autoreactive B cell and T cell responses (7), and therefore the speciﬁcity of the autoantibody response is likely representative of the overall autoimmune response. Although such immunoassays are readily available and relatively easy to perform for a given autoantigen, analysis of multiple autoantibodies is costly, is time and labor intensive, and requires signiﬁcant amounts of sera. Proteomic technologies including antigen array platforms enable the large-scale characterization of immune responses against foreign and selfantigens that may be involved in the development and progression of autoimmune disease. Antigen microarrays allow the comprehensive analysis of autoantibodies directed against hundreds to thousands of antigens, including proteins, peptides, nucleic acids, macromolecular complexes (8–10), and, more recently, lipids (11). Probing these arrays requires microliter volumes of sera. Joos et al. (12) were the ﬁrst group to describe a large-scale antigen microarray-based assay to detect serum autoantibodies. Eighteen autoantigens known to be serologic markers for several autoimmune diseases, including systemic lupus erythematosus (SLE), mixed con¨ nective tissue disease (MCTD), Sjogren’s syndrome, scleroderma, and polymyositis were deposited onto nitrocellulose membranes or derivatized slides using a robotic microarrayer. Arrays were incubated with autoimmune sera followed by horseradish peroxidase (HRP)-conjugated secondary antibody. Chemiluminescent measurements were obtained following application of luminol substrate. These studies demonstrated that the

microarray could be used to determine multiple autoantibody titers simultaneously in a single assay, and this assay was both sensitive and speciﬁc for autoantigen recognition (12). In our laboratory, Robinson et al. (8) adapted the methods of others (13, 14) to design a 1152-feature connective tissue disease (CTD) array. The CTD array included 196 putative autoantigens targeted in several autoimmune diseases, including ¨ SLE, Sjogren’s syndrome, rheumatoid arthritis (RA), polymyositis, scleroderma, and primary biliary cirrhosis. Antigens were spotted onto coated glass slides in an ordered array using a robotic arrayer. Arrays were incubated with either monoclonal antibodies or highly characterized autoimmune serum samples, washed, and probed with ﬂuorescently labeled secondary antibodies. Arrays were scanned and ﬂuorescence intensity measured. These studies demonstrated speciﬁc autoantibody binding that was linear over a 1000-fold range and was four to eight times more sensitive than ELISAs. To date, hundreds of highly characterized autoimmune serum samples have been analyzed and disease-speciﬁc autoantibody patterns detected. Using immunoglobulin G (IgG) subclass-speciﬁc secondary antibodies, the authors also demonstrated that the arrays could be used to characterize autoantibody subclasses that may be important in disease pathogenesis (8). The potential applications of antigen microarrays include (a) improved diagnosis of autoimmune and other diseases, (b) identiﬁcation of autoantibody signatures that may represent subgroups of disease or have prognostic value, (c) monitoring of disease progression or response to therapy, (d ) development of antigen-speciﬁc therapy, and (e) discovery of novel autoantigens or epitopes.

Proteomics: the use of techniques in molecular biology, biochemistry, and genetics to analyze the abundance, modiﬁcations, and interactions of a large number of proteins Cytokines: secreted and soluble proteins that mediate communication between immune cells and their surrounding cells Autoantibodies: any immunoglobulin that binds to self-antigens RPL: reverse-phase lysate Autoimmunity: a state in which an organism’s immune system mounts a response to self-antigens, causing inﬂammation and damage to tissues and organs CTD array: connective tissue disease array

Diagnosis Several autoimmune diseases are characterized by speciﬁc autoantibodies that are important in diagnosis. These include anti-doublestranded DNA (dsDNA) antibodies and www.annualreviews.org • Protein Arrays to Study Autoimmunity

393

ANRV270-IY24-13

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

RAST: radioallergosorbent test

394

0:17

anti-Smith antibodies in SLE, anti-U1snRNP antibodies in MCTD, antiacetylcholine receptor antibodies in myasthenia gravis, and antithyroid stimulating hormone receptor antibodies in Graves’ disease. Measurement of serum antibodies speciﬁc for multiple epitopes of a particular antigen have proven useful in identifying dominant linear epitopes in SLE antigens (18–21) and in increasing the sensitivity of detection of anticitrulline antibodies in RA by combining the frequencies of reactivity to a panel of citrulline-containing peptides (22, 23). Studies using antigen arrays allow for the simultaneous detection of hundreds of autoantibodies on one array with 1 μl or less of patient serum and therefore are ideal for proﬁling autoantibodies in SLE and other autoimmune diseases. In addition to diagnosis of autoimmune disease, antigen microarrays have been designed to determine and monitor IgE reactivity proﬁles in patients with seasonal allergies. Hiller et al. (24) developed a microarray containing 94 puriﬁed allergens. They demonstrated that the array results were consistent with patients’ known sensitization proﬁles based on skin testing or radioallergosorbent test (RAST)-based assays (24). Additional studies showed that these microarrays had a dynamic range comparable to RAST assays, that the sensitivity was similar to ELISA and exceeded that of RAST, and that no signiﬁcant cross-reactivity was observed (25). Antigen arrays have also been applied for serodiagnosis of infectious diseases. Mezzasoma et al. (26) designed arrays with antigens from multiple perinatal pathogens, including Toxoplasma gondii, cytomegalovirus, herpes simplex virus type 1 and 2, and rubella virus. Using a panel of characterized human sera, they validated the arrays, demonstrating a detection limit of 0.5 pg of antibody and sensitivity similar to ELISA (26). The arrays also included internal calibration curves for IgM and IgG, allowing quantiﬁcation of individual immune responses (26). A comprehensive array for simian human immunodeBalboni et al.

ﬁciency virus has also been used to dissect the B cell response in monkeys enrolled in an antigen-speciﬁc DNA vaccine trial (27). With each of these microarray systems, the advantages are that the time to run the assay, the cost, and the amount of serum remain the same regardless of the number of analytes. In contrast, with ELISA each analyte must be assayed individually, increasing time, cost, and amount of serum required. In fact, clinical laboratories typically request at least 0.5 mL of serum for autoantibody studies, with a bare minimum of 0.15 mL per assay. Therefore, the amount of serum needed to test all the antigens on an array could be prohibitive, especially in seriously ill or pediatric patients.

Classification of Autoantibody Biosignatures and Prognostication For many autoimmune diseases, autoantibody proﬁles not only provide diagnostic information but prognostic information as well, allowing patients to be divided into subgroups based on organ system involvement or disease severity. For example, in SLE anti-dsDNA antibodies are associated with active disease and nephritis; antiribosomal P antibodies are associated with neuropsychiatric lupus; and antiRo antibodies are associated with cutaneous lupus, photosensitivity, and the neonatal lupus syndrome (6, 28). Analogous examples can be found in scleroderma, polymyositis, and other diseases (6). Clinicians often use these limited proﬁles to determine prognosis and risk for disease ﬂares. Some reports suggest that serum from patients obtained prior to disease onset contained antibodies predictive of future disease. Analysis of serum samples from 130 people in the Department of Defense Serum Repository who eventually developed SLE revealed that 88% of patients had at least one autoantibody present in serum prior to the diagnosis of SLE. Certain autoantibodies, including antinuclear antibodies, antiphospholipid, antiRo, and anti-La antibodies, developed earlier than others (29). In addition, type 1 diabetes

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

(T1D) autoantibody proﬁles can be used to diagnose and predict future development of T1D. The presence of two or more autoantibodies directed against islet antigens, including insulin, glutamic acid decarboxylase, and tyrosine phosphatase-like protein IA-2, correlates with disease (30, 31). Our laboratory, the Robinson laboratory, and others have developed several diseasespeciﬁc antigen arrays containing putative antigens from the tissues affected by these diseases. These include the previously described CTD array, with almost 200 autoantigens from several rheumatic diseases (8); a synovial proteome array, with ∼650 candidate RA autoantigens (32); a myelin proteome array, with ∼500 myelin peptides and proteins (33); a vasculitis array (G. Alemi and P.J. Utz, unpublished data); and an islet cell proteome array, with pancreatic islet-derived autoantigens (34). Arrays of other biomolecules such as carbohydrates and lipids have also been created (11, 35). These arrays are being used to screen large cohorts of sera from patients and from animal models of disease to predict organ system involvement and to determine prognosis. Such arrays and other proteomic technologies will likely allow the identiﬁcation of autoantibody biosignatures associated with a multitude of other autoimmune diseases in the future. In addition, these assays have the potential to enable better standardization and interpretation of clinical trials by allowing researchers to better classify patients prior to enrollment. Relative levels of speciﬁc autoantibody isotypes may be important in the development and progression of SLE and other autoimmune diseases. Serum derived from patients with lupus nephritis often contains high-afﬁnity IgG antibodies directed against dsDNA, although certain IgM antibodies against dsDNA have been associated with nephritis (36). Analysis of the antiribosomal P antibody response in two SLE patients showed that patients were well when their peak antiribosomal P response was of the IgM isotype and that the development of disease

ﬂares coincided with a switch to high-titer IgG antiribosomal P antibodies (37). Similarly, IgG subclasses may play a role in autoimmunity. Increases in IgG4 responses to desmoglein-1 are associated with onset of clinical disease in pemphigus foliaceus (38). Antigen arrays can be used to determine autoantibody isotype and subclass by probing the arrays with isotype-speciﬁc secondary antibodies differentially labeled with spectrally resolvable ﬂuorophores (8). A preliminary study in our laboratory using antigen arrays showed that the presence of IgG2a antibodies against certain lupus autoantigens correlated with more severe nephritis on histopathologic evaluation in a murine model of lupus (K.L. Graham, unpublished observation). Using antigen arrays to characterize autoantibody isotypes may aid in the identiﬁcation of important antigens for which pathogenic responses are generated and help discriminate other reactivities that do not impact disease progression or activity.

Monitoring Disease Progression and Response to Antigen-Specific Therapeutic Interventions Following the initial stimulation of the immune system, autoantibody diversiﬁcation occurs by the process of epitope spreading (7). Epitope spreading has been demonstrated in several autoimmune disease models (21, 39– 44) and in patients with SLE (45–47) and T1D (48, 49). Investigators believe that the accrual of reactivities to different epitopes over time plays a role in the pathogenesis of these diseases, and epitope spreading has been associated with disease progression in lupus (46) and T1D (48). Antigen arrays printed with overlapping peptides of known autoantigens provide a format for efﬁciently studying epitope spreading. Such studies could be useful in monitoring an individual patient’s response to therapy and in determining which patients are at risk for progression of disease. Robinson et al. (33) have used myelin proteome arrays to study www.annualreviews.org • Protein Arrays to Study Autoimmunity

395

ARI

15 February 2006

0:17

experimental autoimmune encephalomyelitis (EAE), a murine model of multiple sclerosis (MS), and they demonstrated that the frequency of disease relapse was increased in mice with evidence of epitope spreading. Current treatments for most autoimmune diseases are nonspeciﬁc and use agents that globally suppress the immune system. Such treatments have signiﬁcant risks, including systemic infections and secondary malignancies, and there are many diseases for which response to such treatment is marginal at best. By determining the speciﬁc epitopes driving the autoimmune response, therapies that target only those cells that are reactive to these epitopes could be designed, leaving the rest of the immune system intact to function in its role of defense and surveillance. Robinson et al. (33) identiﬁed the early dominant epitopes targeted in EAE and used this information to develop tolerizing DNA vaccines encoding these epitopes. The DNA vaccines encoding myelin sheath components prevented epitope spreading and reduced relapses in mice, particularly when the vaccine was coadministered with a plasmid encoding interleukin-4 (IL-4) (33). Whether similar results can be obtained in human patients treated with antigen-speciﬁc or other immunomodulatory therapies remains to be determined.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

Autoantigen Discovery For many autoimmune diseases, including juvenile idiopathic arthritis, inﬂammatory bowel disease, psoriatic arthritis, and several vasculitides, the inciting autoantigen(s) is (are) currently unknown. cDNA expression libraries, peptide libraries, or arrayed fractions of tissues can be used to screen serum from autoimmune disease patients or animal models to discover novel autoantigens in these diseases. Once autoantigens are identiﬁed, further analyses can be performed to determine the sensitivity and speciﬁcity of the autoantibodies for a given disease. Arrays composed of bacterially expressed proteins have recently 396

Balboni et al.

been used to identify candidate antigens in alopecia areata, an autoimmune disease leading to baldness (50). Newer methodology that uses in situ synthesis of proteins by mammalian ribosomes from cDNAs deposited on the surface of glass microscope slides may also enable such studies (51).

Limitations and Future Directions Given the complex nature of proteins, optimal conditions for antigen arrays have not been established, and variation is seen using different slide surfaces and printing conditions (52, 53). In addition, following attachment of antigens to a planar surface, epitopes may be altered, resulting in lack of detection of autoantigens on the array (8, 13). Our laboratory has evaluated multiple (more than 25) commercially available slide surfaces and assay parameters to optimize arraying conditions for printing whole antigens and linear peptides. We have determined that FAST slides (manufactured by Whatman Schleicher & Scheull) consistently have lower coefﬁcient of variance than other surfaces we analyzed, generally less than 25% and, under certain conditions, as low as 6%–8% (I. Balboni, unpublished observations). We are currently investigating other methods that will allow better internal control of the arrays. Internal control is particularly important for assays that will be used to monitor changes within patients over time, which may be quite subtle. Finally, arrays will need to be validated with many well-characterized serum samples before this platform will be ready for use in the clinical setting.

CYTOKINE ARRAYS Proteomics platforms hold particular promise for the study of cytokines in autoimmunity. This is an important application because therapies that either increase or decrease cytokine levels have proven useful in the treatment of many autoimmune diseases (54–56). The therapeutic beneﬁt of interferon-β in the treatment of relapsing-remitting MS is well

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

established (57), and therapies that inhibit IL-1, IL-6, IL-15, and tumor necrosis factor α (TNF-α) have been shown to improve RA (54, 55, 58–60). In addition to guiding the design of novel pharmacologic strategies, knowledge of cytokine levels adds to our understanding of disease pathogenesis and aids in identifying markers of disease. For example, in SLE, interferon signatures have been identiﬁed at the genetic level (61–63), and the increase in type I interferons has been veriﬁed at the protein level (64). In other cases, the role of cytokines is more complex. The controversial role of TNF-α as an inﬂammatory or immunomodulatory cytokine in MS is one example (65). In these situations in particular, the global analysis of multiple cytokines could provide a more complete understanding of complicated cytokine biology. Antibody arrays are a logical proteomics approach to studying serum proteins such as cytokines and chemokines in autoimmune disease. Some of the technological variations on this theme are highlighted here, as are their implications for the study of disease susceptibility. Planar antibody arrays for analyzing cytokines in biological ﬂuids were developed as a logical extension of the ELISA platform. The most common format involves a variation of the sandwich ELISA (66–68). This method requires two high-afﬁnity cytokine binders, typically commercially available monoclonal or polyclonal antibodies. The capture reagent is spotted on a slide surface, a sample is applied and washed, and then a cocktail of labeled secondary reagents is added to bind available epitopes on the captured species. Detection typically involves ﬂuorescent- or chemiluminescent-based methods (67–69). Some initial reports employing these assays were disappointing (70), but others have demonstrated reliable detection of multiple cytokines on planar arrays (71). Because sensitivity is a concern in cytokine detection, rolling-circle ampliﬁcation (RCA) has emerged as a well-suited detection method. RCA provides a means of linear signal ampliﬁcation that remains local-

ized to the microarray feature (72, 73). Arrays using this method operate exactly like other multiplex sandwich ELISAs, with the exception that the ﬁnal step involves an oligonucleotide elongation reaction from a circular template. The ﬁrst application of RCA to cytokine arrays was impressive. Schweitzer et al. (72) reported simultaneous detection of 75 human cytokines, femtomolar sensitivity, and a time-course analysis of cytokine secretion by mature dendritic cells (DCs) that conﬁrmed and more importantly extended previously described results. Although RCA is slightly more elaborate than other detection methods, investigators have been able to simultaneously measure up to 180 cytokines using planar RCA arrays (74) and have developed two-color methods for analyzing two serum samples on the same array (74, 75). Early studies established immuno-RCA as a powerful detection system, and a subsequent study demonstrated the utility of this microarray technology for investigating autoimmune disease in human patients (76). Kader and colleagues (76) investigated the cytokine proﬁles of pediatric patients with both Crohn’s disease and ulcerative colitis, comparing clinical remission versus active disease. Although the authors anticipated increases in proinﬂammatory cytokines in patients with active disease, instead they detected an increase in socalled regulatory cytokines during remission. Placental growth factor, transforming growth factor-β1, IL-7, and IL-12p40 were all upregulated in serum from patients who were in clinical remission (76). These unexpected ﬁndings implicated a role for immune regulation in suppressing disease activity. Furthermore, the authors reported that a 10-cytokine classiﬁer was better than a 4-cytokine classiﬁer, suggesting that increased multiplexing of cytokine levels could better discriminate between disease states. With any developing technology, veriﬁcation of important ﬁndings by an alternative measure is essential, and in this study data from conventional ELISA could have strengthened the ﬁndings. Nevertheless, this study exempliﬁes the advantages www.annualreviews.org • Protein Arrays to Study Autoimmunity

RCA: rolling-circle ampliﬁcation

397

ARI

15 February 2006

0:17

of cytokine microarrays in studying autoimmune disease. While the planar array format identiﬁes capture antibodies by a coordinate on the slide, optically encoded microspheres can also serve as unique identiﬁers of the capture antibodies (bead arrays, or bead-based assays). Although adding additional features to suspension arrays is less trivial than for planar arrays, current bead array technologies can easily handle 100 analytes simultaneously and could be developed to handle thousands or tens of thousands of probes (77). The main advantages of this technology include easier quality control because beads are coated in bulk, higher density of array features, faster analysis of each sample, and higher throughput for multiple samples when compared with planar arrays (77). These advantages have made bead-based arrays useful for the study of cytokines in autoimmune disease. A study by Szodoray ¨ and colleagues (78) investigated Sjogren’s syndrome using cytokine bead-based arrays. ¨ Serum was collected from Sjogren’s syndrome patients, along with clinical and laboratory data. Differences among patients and controls were observed, including levels of IL-12p40, TNF-α, IL-6, and TNF receptor I (TNF-RI) and TNF-RII, some of which had not been previously linked to ¨ Sjogren’s syndrome. Furthermore, subsets of ¨ Sjogren’s syndrome patients showed differential cytokine levels. Levels of IL-2, epidermal growth factor (EGF), macrophage inﬂammatory protein-1α (MIP-1α), and TNF-RI were associated with elevations in erythrocyte sedimentation rate, but only IL-12p40 differed according to extraglandular manifestations. These studies demonstrate that quantitative, multiplexed analysis of serum cytokines in human autoimmune disease is indeed possible. Determining the scientiﬁc and clinical impact of these observations will be the next major step for cytokine arrays. Planar arrays and bead-based suspension arrays identify capture antibodies by the use of encoded solid supports, whether that sup-

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

398

Balboni et al.

port is a microarray slide or a microsphere. An entirely separate approach to multiplex cytokine assays is the proximity assay, in which all reagents remain free in solution. Proximity assays have the advantages that no wash steps are necessary and no immobilization of capture reagents is required. Monogram BioSciences’ eTAGTM system is an example of a proximity assay involving cleavable tags, although oligonucleotides are also useful for proximity-dependent cytokine detection (79). Fredriksson and colleagues (80) demonstrated the power of this oligonucleotide technology in highly sensitive cytokine detection using a pair of aptamers. Aptamers are probes made entirely of oligonucleotides that are selected through rounds of in vitro evolution for the ability to bind a target with high afﬁnity. The aptamers used by Fredriksson et al. (80) were speciﬁc for two separate epitopes on platelet-derived growth factor. When the aptamers bound their target, the local concentration of the complementary aptamer ends increased by orders of magnitude, allowing detection of ligation by quantitative real-time polymerase chain reaction. After extensive optimization, zeptomole (10−21 ) concentrations could be detected, which is far more sensitive than most cytokine assays. Multiplexing an assay of this kind is possible, although more difﬁcult than planar or bead-based arrays. Subsequent application of proximity ligation employed oligonucleotide-conjugated antiIL-2 and anti-IL-4 mono- and polyclonal antibodies, afﬁrming the generalizability of this proteomics approach (81). Proximity assays require more sophisticated reagents and assay conditions than planar or bead-based arrays, but they can be developed for exquisite sensitivity, speciﬁcity, and throughput. A variety of technological advances have made quantitative, multiplex cytokine proﬁling possible. The main advantages of these techniques over existing single analyte ELISA include small sample requirement, highthroughput capability, and reliable data on many markers simultaneously. The multiplexing ability of planar, bead-based, and

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

proximity assays will most likely not be the limiting factor in the development of these assays. One thing all the above technologies have in common, however, is absolute dependence on highly speciﬁc, high-afﬁnity binders, sometimes relying on pairs of binders. Intense commercial and academic effort will be necessary to generate and evaluate all these reagents for every cytokine of interest. As these proteomics platforms evolve, studying cytokines should prove invaluable for investigating autoimmunity. These data could provide new therapeutic targets, a better understanding of pathogenesis, improved disease classiﬁcation, and novel markers for measuring response to therapies. Because many therapies are beginning to target cytokines, and because cytokine levels can now be studied in a high-throughput fashion, in the future cytokine arrays may be able to guide patient-tailored cytokine therapy.

REVERSE-PHASE LYSATE (RPL) MICROARRAYS The concept of RPL microarrays was originally described by Paweletz et al. (82) in a paper showing activation of the prosurvival signaling protein Akt and suppression of extracellular regulated kinases (ERK)1/2 phosphorylation in the transition from normal prostate epithelium to prostate intraepithelial neoplasia, and then to invasive prostate cancer. Although only a small number of patient samples and signaling proteins were analyzed, the potential to scale up this microarray platform for high-throughput analysis of intracellular proteins was clear. The ability to proﬁle the abundance and activation state of a large panel of signaling proteins is particularly exciting for immunology research because most, if not all, immunological decisions depend on the signaling pathways that are turned on and their degree of activation. Before we focus on the technical aspects of RPL microarrays, it is helpful to ﬁrst distinguish between forward-phase and reverse-

phase approaches and to discuss why RPL arrays have unique advantages over other array formats for protein proﬁling, especially for the study of intracellular signaling networks. In the forward-phase approach, analytes of interest are captured from solution phase by an array of immobilized antibodies. The bound analytes, usually proteins, are subsequently detected and quantiﬁed. A major challenge to the development of forward-phase arrays is the lack of antibodies that function in this format. In one study involving 115 antibody/antigen pairs, less than 20% of the antibodies provided speciﬁc and accurate measurements (13). Notably, in this study a 3% nonfat dry milk solution spiked with puriﬁed antigens was used to probe the arrays. Other groups have reported that only about 5% of intracellular antibodies are suitable for forward-phase arrays when the arrays are probed with a complex solution such as a cell lysate (83, 84). The forwardphase approach is further complicated by the issue of detection. Direct labeling of the protein sample with ﬂuorescent tags is one approach and has been tried by several groups for protein microarray studies (13, 84, 85). However, the tags can sterically interfere with antibody binding if they are located within binding epitopes (83). Furthermore, the efﬁciency of protein labeling reactions is notoriously variable, making quantitative comparisons between samples difﬁcult (13). Currently, the “sandwich immunoassay” approach appears to be the best option for forward-phase microarrays. However, antibody pairs are not available for many intracellular proteins. The reverse-phase approach immobilizes lysate samples as distinct microspots on the array surface instead of antibodies (86) (Figure 1). RPL arrays are then probed with highly speciﬁc antibodies that are either phosphorylation state dependent for detecting activation states or independent for measuring abundance. Bound antibodies can then be detected using a secondary antibody that is directly conjugated to a ﬂuorophore www.annualreviews.org • Protein Arrays to Study Autoimmunity

399

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

0:17

Figure 1 Reverse-phase lysate (RPL) microarrays. A schematic diagram illustrating the steps involved in sample collection, array fabrication, array processing, and data analysis. (Adapted from Reference 86 with permission.)

[e.g., quantum dots (87)] or enzymes (e.g., HRP) for signal ampliﬁcation. In our experience, enzyme-mediated signal ampliﬁcation is necessary for detecting low abundance proteins and their phosphorylated epitopes. We employ a tyramide-based ampliﬁcation technique that deposits biotin molecules adjacent to HRP-conjugated secondary antibodies for signal ampliﬁcation (88). A ﬂuorescent signal is then generated by incubating the arrays with ﬂuorophore-conjugated streptavidin. The ﬂuorescent intensity of each spot correlates with the abundance or level of phosphorylation of the analyte in question. The main advantage of this technique is that thousands of samples can be analyzed simultaneously on the same platform, greatly increasing throughput and simplifying quantitative comparisons between samples. Furthermore, an exceedingly small amount of sample (a few micrograms of protein) is required for printing hundreds of RPL arrays, thus permitting the comprehensive analysis of rare cell types and valuable patient samples. Importantly, most if not all commercially available antibodies should work in this for400

Balboni et al.

mat, provided they are not cross-reactive with other proteins. A clear limitation to this approach is that only one analyte can be measured on a single array. To partially overcome this limitation, arrays can be printed in a multisector format (arrays of arrays), allowing up to 16 analytes to be analyzed on a single slide. Although this number is still orders of magnitude smaller than the number of probes on a DNA microarray, it is not as limiting as one might think if the analysis is focused on one functional class of proteins (e.g., signal transduction proteins or apoptosis-related proteins). A particularly successful application of RPL microarrays has been the proﬁling of phosphorylation states of signaling proteins with the use of phosphorylationspeciﬁc antibodies (88). This application has only been possible within the past few years with the commercial availability of an expanding collection of well-characterized phosphorylation-speciﬁc antibodies. The majority of work that has been done using RPL microarrays is in the ﬁeld of oncology. This body of work, mostly contributed by Lance Liotta’s and Emanuel Petricoin’s research groups, involves the use of laser capture

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

microdissection (LCM) to isolate relevant portions of tumor specimens. The isolated cells are then lysed and spotted onto nitrocellulose-coated slides. Each spot contains the entire complement of proteins from a single specimen. RPL microarrays have been used to conﬁrm changes in protein expression levels found using other techniques [e.g., DNA microarrays (89, 90), 2DIGE (91), immunohistochemistry (92, 93), and Western blot analysis (92)]. For instance, RPL microarrays were used to validate the ﬁnding that the regulatory signaling protein Rho G-protein dissociation inhibitor is selectively overexpressed in invasive human ovarian cancer versus low malignant potential ovarian cancer (91). More sophisticated studies involving phosphorylation-speciﬁc antibodies demonstrate dysregulated signaling in cancer cells (93). A study examining the phosphorylation states of ERK1/2 and Akt in ovarian cancer samples showed a trend toward increasing ERK1/2 phosphorylation with disease stage independent of histological type, but this same trend was not observed with phospho-Akt (94). Discovery of novel and unexpected signaling defects in diseased states (e.g., autoimmunity) is a powerful application of RPL microarrays, as these defects can potentially point to new therapeutic targets. This application is demonstrated in a prostate cancer study in which activation of PKC-α was found to be downregulated in tumor cells compared with normal epithelium by screening a panel of six phosphorylation antibodies using RPL arrays (95). Several studies have also reported early attempts to classify cancer cells into subtypes on the basis of their phosphorylation proﬁles. In a recent study comparing the phosphorylation states of 26 signaling proteins between primary and matched metastatic ovarian carcinomas, the level of c-Kit phosphorylation alone was found to be sufﬁcient for categorizing samples as being of either primary or secondary origin (96). Interestingly, the phosphorylation-signaling signature appears to change dramatically when a primary tumor transforms into metastatic can-

cer (96, 97). In another study involving breast cancer specimens from 54 patients, RPL microarrays were used to proﬁle both the activation state and abundance of 11 signaling proteins (97, 98). Hierarchical clustering of the data revealed four distinct tumor subtypes. Although the study did not report whether these subtypes correlated with disease progression and/or response to therapy, we predict that phosphorylation state proﬁling of clinical specimens will have clinical utility in classifying not only cancer patients, but also autoimmune patients into prognostic subgroups. LCM can be used to isolate relevant cells, including inﬁltrating lymphocytes, antigen-presenting cells, or cells that are targeted in the immune response (e.g., kidney cells in SLE, beta cells in T1D, or brain tissue in MS). All the above mentioned work has been done using tumor specimens that are frozen immediately after resection. Thus, the specimen represents a snapshot of the tumor’s signaling status at the time of resection. Evidence shows that dysregulation of signaling in cancer cells may only be revealed upon triggering with environmental stimuli (98a). We believe that a similar situation occurs in immune-related cells isolated from autoimmune patients. Therefore, we have focused our attention on stimulating living cells and monitoring their response to stimulation using RPL microarrays. We have successfully applied RPL microarrays to the study of signaling kinetics and pathway delineation in Jurkat T lymphocytes stimulated with phorbol myristate acetate (PMA) and with surface receptor antibodies to CD3 and CD28 (88). Furthermore, by monitoring changes in the phosphorylation state of 62 signaling proteins, we discovered a previously unrecognized link between CD3 crosslinking and dephosphorylation of Raf-1. Because only small amounts of sample are required for printing arrays, RPL microarrays have the potential to analyze rare primary cell populations. As a feasibility study, we proﬁled the phosphorylation state of 23 signaling proteins www.annualreviews.org • Protein Arrays to Study Autoimmunity

LCM: laser capture microdissection 2DIGE: two-dimensional gel electrophoresis

401

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

0:17

Figure 2 Proﬁling the phosphorylation state of 23 signaling proteins in Tregs in response to stimulation. (a) Naive CD4+ CD25− T cells (T) or CD4+ CD25+ T cells (Treg) were freshly isolated and stimulated with PMA and ionomycin. (b) Day 3 T cell blasts (D3B) and freshly isolated CD4+ CD25+ T cells were puriﬁed and stimulated with IL-2. Yellow intensities indicate increases in phosphorylation. Blue intensities indicate decreases in phosphorylation.

in CD4+ CD25+ regulatory T cells (Tregs) isolated from naive mice. We previously reported differential STAT protein phosphorylation in these cells in response to IL-2 stimulation (88). Unlike naive CD4+ CD25− T cells, Tregs do not normally proliferate in response to T cell receptor (TCR) and CD28 stimulation in the absence of high concentrations of exogenous IL-2 (2000 U/ml) (99). However, some signals must be transmitted through the TCR because stimulation through the TCR is required for Tregs to exert suppression (100). The biochemical mechanism behind this nonproliferative state is not clear. Using RPL microarrays, we found that Tregs responded almost identically to naive CD4+ CD25− T cells when stimulated with PMA and ionomycin (Figure 2a). This response suggests that the signaling pathways downstream of PMA and ionomycin are intact 402

Balboni et al.

in Tregs. Alternatively, the pathways that become activated with IL-2 stimulation may be different between Tregs and activated T cell blasts that also express CD25. RPL microarray analysis revealed that the phosphorylation proﬁles of IL-2-stimulated Tregs and T cell blasts were dramatically different (Figure 2b). Some signaling proteins such as STAT-5 appear to be activated equivalently, but in general the response in Tregs was clearly decreased relative to activated T cell blasts. In particular, Akt was not activated in response to IL-2 stimulation, which may in part explain the lack of proliferative response in Tregs. In summary, we believe that RPL microarrays represent a new platform that will ﬁnd numerous applications in the study of immunology. The proper functions of most immune-related cells depend on signals that are transmitted from the cell surface to the

ANRV270-IY24-13

ARI

15 February 2006

0:17

nucleus. In autoimmune states, signal transduction pathways are often dysregulated, leading to inappropriate responses. The advent of RPL microarrays should provide a more complete picture of the dynamic signaling networks that occur in normal and disease states.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

EMERGING TECHNOLOGIES Fluorescence-Activated Cell Sorting (FACS) and FACS Signaling This review has focused on protein array– based approaches for proﬁling autoantibody reactivity and intracellular signaling pathways and for measuring cytokine levels. Many of the processes involved in autoimmunity, however, involve interactions between cell-surface molecules. A growing number of therapies target cell-surface molecules in autoimmune disease, including altered peptide ligands, antibodies against adhesion or costimulatory molecules, and antibody depletion. Flow cytometry has been the traditional means of studying cell-surface markers in immunology. Currently, ﬂow cytometric analysis can accommodate 17 ﬂuorescent parameters and 2 scatter parameters (101). By interfacing the cytometer with autosampling instruments (102), the speed of analysis has rapidly increased, whereas the number of cells required has dramatically decreased. This technology will undoubtedly continue to evolve as a proteomics tool, but the complexities in instrumentation, reagent preparation, and ﬂuorescence compensation are formidable obstacles (101). The advent of phospho FACS for studying signaling pathways in heterogeneous cell populations is certain to be a dominant technology platform over the next ﬁve years (98a,b) (see FACS Signaling: Emerging Technology). Recent work has combined planar arrays with analysis of whole cells and cell-surface markers. These novel arrays could complement ﬂow cytometry as proteomics tools for the study of autoimmune disease.

FACS SIGNALING: EMERGING TECHNOLOGY Planar protein arrays and ﬂow cytometry assays have unique sets of advantages and disadvantages, allowing both platforms to contribute in different ways to the study of autoimmune disease. In addition to analyzing cell-surface markers, ﬂow cytometry is now being used to study intracellular signaling events (98a,b). A comparison of RPL microarrays and phospho-ﬂow highlights the complementary nature of the two technologies. The advantages of phospho-ﬂow include the ability to handle heterogeneous cell populations, to perform single cell analysis, and to generate multiparameter data. RPL microarrays, however, have the advantages that more antibodies recognize their epitope in this format and fewer cells and smaller amounts of reagent are needed for parallel experiments than in phospho-ﬂow cytometry. Combining the two technologies to study autoimmunity could circumvent the limitations of either approach, allowing the interpretation of more meaningful data. A similar approach could take advantage of both cell-surface marker arrays and conventional ﬂow cytometry.

Peptide-Major Histocompatibility Complex (MHC) Arrays Until recently, phenotypic analysis of antigenspeciﬁc T lymphocytes was limited to a ﬂow cytometric approach involving single peptideMHC tetramer staining. To overcome the limitations of this approach, Soen et al. (103) developed peptide-MHC tetramer arrays. This work has been reviewed in depth elsewhere (86). To brieﬂy summarize, various peptide-MHC complexes were spotted on a slide, and the slide was probed with ﬂuorescently labeled T cell populations. Stone et al. (104) advanced this tetramer microarray technology by incorporating more functional assays on the array. The authors successfully characterized the activation status and cytokine production of captured T cells. As proof of principle, the authors spotted anti-IL-2 capture antibodies with HLA-DR1 presenting an inﬂuenza hemagglutinin peptide (HA), as well as various other HLA-DR1-peptide complexes. After www.annualreviews.org • Protein Arrays to Study Autoimmunity

FACS: ﬂuorescenceactivated cell sorting

403

ARI

15 February 2006

0:17

incubation with an inﬂuenza-speciﬁc T cell hybridoma, IL-2 secretion was detectable only on the feature where HLA-DR1-HA had been immobilized, demonstrating the speciﬁcity of this technology and the ability to perform a miniaturized ELISPOT on the array. Antibodies against costimulatory molecules were also spotted with the peptide-tetramer complexes and were shown to enhance cytokine production. Similar to the study by Soen et al. (103), Stone and colleagues (104) reported a sensitivity of ∼0.1%, putting their technology within the biological range of autoimmune T cell responses. Stone et al. (104) went on to report simultaneous detection of multiple cytokines and staining of bound T cells for activation markers. With this type of technology, investigators might consider proﬁling and characterizing the autoreactive T cell repertoire in autoimmune disease, in much the same way that array-based autoantibody proﬁling monitors the humoral immune response. Although less sophisticated than the reports involving peptide-MHC arrays, arrays composed of antibodies directed against cellsurface markers have been used for immunophenotyping populations of cells. The experimental design involves spotting antibodies directed against cell-surface markers and incubating populations of cells on the array (105, 106). Each 500 μm feature can bind roughly 1500 cells, and the use of a planar surface is believed to enhance the avidity of interactions. The difﬁculty with interpreting these data is that cell immobilization depends on a poorly characterized function of antibody afﬁnity, molecule copy number, and cell abundance. This developing technology suffers from a number of limitations, including the need for homogeneous populations, such as clonal leukemias and lymphomas. Despite these problems, the main advantage of planar arrays is that a few million cells can be proﬁled simultaneously for up to 90 surface markers (106). This multiplex screening could help guide

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

404

Balboni et al.

more focused ﬂow cytometry and functional studies.

INFORMATICS CHALLENGES IN PROTEOMICS Informatics, generally deﬁned, is the collection, classiﬁcation, storage, and analysis of data. Recent developments in experimental methods have changed the nature of each one of these tasks in the context of proteomics. Classiﬁcation and storage, although important, are relatively straightforward hurdles to scale. What is truly novel about postgenome era proteomics is found in the areas of data collection and analysis. One major change is that, increasingly, a number of methods in proteomics are as much about computationally collecting data as they are about generating data. Once the data have been generated or collected, the sheer scale necessitates new approaches to analysis. In this section, we address current challenges in the pursuits enumerated above, some solutions to these challenges, and computational methods in some commonly used experimental techniques. Our focus on relatively mature techniques should not be interpreted to mean that the experimental methods described above do not present their own set of informatics challenges. Rather, the techniques themselves are so recent that the informatics side of the problem has not yet had a chance to catch up. Storage and analysis of the large data sets generated by techniques like ﬂow cytometry and protein microarrays are areas of active research, sure to yield innovative new approaches in the near future. We conclude by describing some exciting research areas and opportunities that are enabled by these new informatics developments.

Data Classification and Storage Innovative proteomic techniques, and even some relatively mature ones, have led to the generation of novel types of data sets, unique both in character and in scale. As

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

both the quantity and variety of proteomic data have increased with recent developments in high-throughput experimental methods, so too have the challenges involved in managing these data. Whereas historically results have been stored predominantly in lab notebooks, gel ﬁlms, and free text publications, recent years have seen a paradigm shift. The cause for this change is multifold: (a) Desktop machines have grown in their capacity for data storage and processing, (b) the World Wide Web has facilitated easy data exchange, and (c) experimental techniques have become high throughput in nature and as such produce larger quantities of data. In addition, new methods have been devised for automatically and efﬁciently converting what previously had been stored as qualitative analog data, e.g., 2DIGE images, into digital information, as we discuss below. Interestingly, although the sheer quantity of proteomic data being produced is novel, these data are still largely of a qualitative nature. Most numeric values reported in proteomics are in somewhat arbitrary, or at least relative, units. Protein microarrays measure ﬂuorescence intensity, but the actual numeric values are determined as much by the laser voltage level at which the slide is scanned as they are by the number of ﬂuorophores present in a given spot. Mass spectrometry measures peak intensity values of different molecular fragments, but the units of intensity are relative to the maximum peak in a given experiment. Qualitative proteomic data may also take the form of protein-protein interaction data, protein presence and posttranslational modiﬁcation state, amino acid sequence, and three-dimensional structure. Beyond the data themselves, far more challenging to informaticians are issues surrounding metadata, or data about data. For example, along with peak intensity in mass spectroscopy, or ﬂuorescence level in FACS or protein arrays, a researcher is likely to record experimental conditions: sample tissue type, species, method of stimulation, etc. Such metadata must be stored in a structured way

for them to be useful at some future date to another scientist trying to retrieve information speciﬁc to her/his own areas of interest. The term “structured” in this context can mean many things. At a minimum, it means that the metadata summarizing the experiment are recorded digitally and not in one long entry of prose. Rather, different ﬁelds are used to store attributes such as tissue type, stimulation method, duration, etc. Ideally, there is a predeﬁned format or a controlled vocabulary for what may be entered in these ﬁelds so that, e.g., a search for all T cell experiments does not miss those described as CTL, CD8+ , Th, etc. Notably, these issues are only magniﬁed as one moves toward a clinical setting, where standard practice is free text, paperbased patient records. In the clinic, as in the wet lab, implementation of electronic data entry systems is a constant tug-of-war between structuring data for ease of retrieval and exchange and ﬂexibility to express ideas not easily summed up by, for example, selecting an option from a list.

Data Generation As mentioned above, collection of data in proteomics often assumes the form of computational methods for generation of digital data from analog media. A prime example is that of image processing for analysis of 2DIGE. A plethora of packages exist for detecting spots on 2D gels, including MELANIETM , PDQuestTM , Z3 and Z4000, PhoretixTM , and ProgenesisTM (107–109). Each package has its own algorithm, but the general method is to scan and digitize a picture, ﬁlter the noise from the digital image, and identify individual spots on the gel. Spots are often then manually extracted and analyzed using mass spectrometry to identify the protein of which that spot is composed (see below). Once these gel images have been processed and the information stored in a structured fashion, this information can be deposited into a data repository to be used by other researchers. The ExPASy (Expert www.annualreviews.org • Protein Arrays to Study Autoimmunity

405

ARI

15 February 2006

0:17

Protein Analysis System) proteomics server, developed by the Swiss Institute of Bioinformatics, hosts the Swiss-2DPAGE database in which a researcher can search for images based on various criteria (110). Once an image has been selected, the researcher can click on a labeled spot and ﬁnd annotated information about that protein as determined by the original contributor. Although not discussed at length in this review, mass spectrometry is another popular technique in which computational methods are needed to generate and analyze results. Data from mass spectrometry experiments originate as a stream of relative intensities of molecules displaying a variety of massto-charge ratios. These data may be used to identify components of a sample, to compare protein quantities in multiple samples, to sequence a protein de novo, or even to detect post-translational modiﬁcations such as phosphorylation or ubiquitination (111). Tens of thousands of spectra may be produced from a single experiment, and output ﬁles can be up to megabytes in size, depending on which speciﬁc technology is used. Different types of mass spectrometry vary by how the sample is separated prior to analysis, by how it is ionized, and by the method for mass analysis (112). An approach like liquid chromatography mass spectrometry will produce chromatograph data in addition to peak intensities. Tandem mass spectrometry breaks peptides recursively into smaller fragments, generating more peaks with each step. For protein identiﬁcation, the mass spectrometry results are compared against a database containing peak information for known proteins that have been fragmented with the same method as the sample. Advanced database search techniques are used to search these existing databases of massto-charge ratios of known proteins to identify a sample (113, 114). The success of this method relies on the availability and comprehensiveness of the database being searched.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

406

Balboni et al.

To sequence a protein de novo, a combinatorial algorithm is used to piece together multiple peptide fragments (115–118) in much the same way stretches of sequenced DNA are assembled in the process of whole genome sequencing. Differences between peaks can correspond to the differences in mass of a single amino acid in the predigestion step. A sequence of differences may then be used to make predictions regarding the sequence of amino acids. Finally, mass spectrometry may also be used to compare both protein abundance and phosphorylation state in two different samples by using labeled isotopes (119). For protein abundance, peaks for fragments of each sample will be only as far apart as the difference in weight of the labeled isotope (multiplied by the number of atoms of the isotope in the fragment). If the protein is puriﬁed before being analyzed, then the relative intensities of these closely spaced peaks will reﬂect the relative abundance of the proteins from which they are derived. Similarly, if comparing the phosphorylation state of two puriﬁed peptides, one can compare the relative intensities of the peaks for the phosphorylated and nonphosphorylated amino acid, which will be separated by the weight of the phosphate group.

Standards for Data Storage and Exchange Once proteomic data have been generated, they must be stored in such a way as to facilitate information retrieval, analysis, and exchange. As the open source model of scientiﬁc data grows in popularity and databases like the Database of Interacting Proteins and Biomolecular Interaction Network Database (BIND) grow and ﬂourish, standards must be established as to what format the shared data will assume (120, 121). A number of collaborating organizations have been founded toward this end that focus on different aspects of proteomics standards. The Proteomics Standards Initiative (PSI) is a working group within the

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

Human Proteome Organization (http:// www.hupo.org/) that focuses on data storage standards. Whereas DNA microarrays have MIAME (minimum information about a microarray experiment), which researchers are required to follow to submit DNA microarray data to various journals and repositories (121), PSI has put forth MIAPE (minimum information about a proteomics experiment), based in part on the Proteomics Experiment Data Repository model, or PEDRo, as well as on an XML-based molecular interaction standard (122–124). With respect to data exchange, the systems biology markup language, or SBML, is an XML-based language that allows researchers to share quantitative information relating to biochemical reaction networks, including cell signaling and gene regulation (125). BioPAX (http://www.biopax.org/) was founded in 2002, also for the purpose of creating an exchange format for biological pathway data. It deﬁnes an ontology for pathway data, that is, a set of terms and their relationships to each other, and is of a more qualitative nature than is SBML and less geared toward simulation.

Making Sense of the Results So now that the proteomics data have been generated and stored, what is next? Inspecting a spreadsheet of ﬂuorescence intensity values seldom gives insight into what new knowledge the data hold. Even a detailed set of reaction-based differential equations will generally fail to convey an intuition for what is taking place in the cell. A number of products exist, both commercial and academic, to enable visualization of molecular pathways and interactions. These have been commonly applied to the results of analysis using RNA expression microarrays, but they can be applied to proteomics results as well. At the simplest level, websites such as KEGG (http:// www.genome.jp/kegg/pathway.html) and Biocarta (http://www.biocarta.com) have static images depicting a number of different

pathways. More advanced software applications allow the user to create or infer pathways from complex data sets and also to map time series data onto pathway images (126– 128). Additional software can ﬁnd and sort pathways in order of potential interest based statistically on the genes or proteins involved (129, 130). The underlying motivation for storing data in an organized fashion, in addition to facilitating human comprehension of those data, is to allow computers to operate on these large data sets. That is, computers are able to integrate thousands of data points at one time into a model of the underlying mechanisms far better than can the human mind. This modeling may be done at a high level of abstraction, for example to determine correlation or causality between phosphorylation events in autoreactive lymphocytes, or at a low level, for example using differential equations to model enzymatic reactions in a pathway. In a previous study (131), the authors developed a mathematical model consisting of ordinary differential equations to model 94 different compounds known to be involved in signaling downstream of EGF receptor stimulation. The kinetic parameters used were taken from the literature, determined experimentally, or computed from published time-dependent quantitative observations. Model results were in good agreement with experimental observations. Clearly, this type of modeling could not be done without fairly detailed knowledge of known pathways, and so one may wonder what knowledge is gained from making such a model. First, researchers would have learned something new had the experimental results not reﬂected the model’s predictions; this would have indicated that there was something amiss in what is currently recognized as truth in this system. Second, as the development and use of such models mature, researchers will be able to perform in silico experiments before doing them at the bench. This will save time and resources by avoiding experiments that the www.annualreviews.org • Protein Arrays to Study Autoimmunity

407

ANRV270-IY24-13

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Bayesian network: a graphical representation of a system in which nodes represent the entities and the edges represent probabilistic dependence relations between the nodes

408

0:17

model shows will not give good results, and it could enable scientists to do virtual experiments that would otherwise be impossible owing to cost, technology, or ethical considerations. But what can informatics techniques offer in areas where not much is known? Sachs et al. (132) used Bayesian networks, a subﬁeld of graph theory, and ﬂow cytometry data to elucidate known and novel relationships in signaling pathways in human T cells. In computer science, a graph represents some number of nodes, and between these nodes relationships are depicted using edges, or lines. In the context of proteomics, graph nodes represent proteins and edges represent relationships between the proteins. The graph may be further enriched by storing information about these nodes and edges. For example, edges may be labeled as “binds to,” “inhibits,” “phosphorylates,” etc. Bayesian networks are a subset of graph models where edges have a speciﬁc directionality to them, and each node reﬂects an event with a probability distribution determined by its parent nodes. So, an arrow from the “MEK is phosphorylated” node to the “ERK is phosphorylated” node strictly indicates a belief that ERK phosphorylation state is dependent on the MEK phosphorylation state and suggests that MEK phosphorylates ERK. Sachs et al. (132) used data on the phosphorylation state of 11 different molecules in each of thousands of individual cells under various conditions. Clearly, this goes beyond the capacity of the human mind to process, but creating a Bayesian network from data requires very large data sets. Using these data, the authors predicted 15 known relationships in human T cell signaling as well as two novel connections not previously reported in these types of cells. These results were subsequently conﬁrmed experimentally. This represents an exciting new approach toward elucidation of cell signaling

Balboni et al.

in those pathways about which not much is known.

Challenges Ahead One of the most exciting informatics challenges facing proteomics researchers is that of integration with other types of highthroughput data in molecular biology. As structured data are created, stored, and shared, both in proteomics and in other areas of biology experiencing similar growth, an opportunity arises to combine these different types of data in ways never before possible. In doing so, we enable exciting new systems approaches that allow researchers to combine proteomic information with other types of biological data, thus enabling new discoveries that rely on these multiple inputs. For example, Ideker et al. (133) used the protein-protein interaction data from the BIND database in conjunction with gene expression data for the proteins in that network. By analyzing which genes are coexpressed and whose protein products are known to interact, they determined coherent functional modules that could not be identiﬁed by using either data set alone. As another illustration of this integrative approach, Mootha et al. (134) identiﬁed the gene involved in Leigh syndrome by combining DNA sequence, RNA expression, and mass spectrometry protein expression data. Having previously narrowed the search to a speciﬁc region of chromosome 2, and knowing that the disease involved mitochondrial function, they combined these different data sets to identify one gene as the site of mutation in this disease (134). With these types of scenarios as a promising start, it is clear that this data-driven, integrative approach will be key in understanding how molecular mechanisms function at all levels in the cell, which in turn is a key step to understanding why they fail and contribute to the development of autoimmunity (133, 134).

ANRV270-IY24-13

ARI

15 February 2006

0:17

SUMMARY POINTS 1. Proteomics technologies fall into two major categories: unbiased technologies such as mass spectrometry that identify thousands of different proteins or peptides, and biased technology platforms such as FACS, antigen array, and antibody array platforms that use existing capture agents such as puriﬁed proteins, peptides, or antibodies. 2. The choice of technology platform is dictated by the hypothesis that is being tested and the availability of speciﬁc reagents.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

3. Antibody proﬁles may identify autoimmune disease subsets or predict response to therapy. 4. Cytokine proﬁling can be employed to identify targets for therapeutic monoclonal antibody development. 5. A major challenge facing protein microarrays is the development of highly speciﬁc reagents. 6. Identiﬁcation of biomarkers and surrogate markers of autoimmunity and tolerance represents a holy grail of clinical immunology. 7. Flow cytometry and emerging planar array technologies have complementary sets of advantages and disadvantages. 8. Various factors, including computing power, the Internet, and the quantity of experimental data being produced, motivate and necessitate novel approaches to data storage, analysis, and visualization. 9. The storage of experimental data in a structured format enables exciting new possibilities for automated and integrative biological research.

FUTURE ISSUES TO BE RESOLVED 1. As promising as protein arrays have become for studying autoimmunity, it is unclear if successful pilot studies in animal models will translate into similar ﬁndings in an outbred human population. 2. Which technology platform is correct for a given application is unclear; the current gold standard method for measurement of an analyte (e.g., cytokine or autoantibody measurements by ELISA) often does not correlate with results obtained by other platforms (e.g., bead-based or array-based assays), and it is not clear which data set to believe. 3. Major efforts should be organized to standardize reagents, assays, data storage, and normalization techniques. 4. The most important area for future development in proteomics is not the assay format or instrumentation, but rather creation of computational and statistical tools for analysis of disparate data sets such as transcript proﬁles, protein proﬁles, and cell-surface phenotypes.

www.annualreviews.org • Protein Arrays to Study Autoimmunity

409

ANRV270-IY24-13

ARI

15 February 2006

0:17

DISCLOSURE STATEMENT P.J.U. states the following conﬂict of interest disclosures, which are all directly relevant to the proteomics technology described in this manuscript: In the past three years he has served as a consultant to Becton Dickinson Biosciences (San Jose, CA) and Genentech, Inc. (South San Francisco, CA), is a member of the Scientiﬁc Advisory Board of Monogram Biosciences (formerly ACLARA Biosciences/Virologic) (South San Francisco, CA) and XDx, Inc. (South San Francisco, CA), and is a cofounder and consultant at Bayhill Therapeutics (Palo Alto, CA).

ACKNOWLEDGMENTS Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

We thank members of the Utz lab and the laboratories of Bill Robinson and Larry Steinman for development of proteomics technologies, some of which are described in this review. I.B. is funded by the Arthritis Foundation Postdoctoral Fellowship. S.M.C. and M.K. are funded by the Stanford MSTP and the Floren Family Foundation. J.D.T. is funded by a fellowship from the National Library of Medicine. A.J.B. is funded by the Lucille Packard Foundation for Children’s Health and NIH grants K22LM008261, T32DK063702, and R01DK062948. P.J.U. is the recipient of a Donald E. and Delia B. Baxter Foundation Career Development Award and was supported by the Dana Foundation, the Northern California Chapter of the Arthritis Foundation, the Stanford Program in Molecular and Genetic Medicine, NIH Grants P30 DK56339 (Stanford Digestive Disease Center), DK61934, AI0151614, AI50854, AI50865, AI40093, AR49328, and NHLBI Proteomics Contract N01-HV-28183 during the period when many of these technologies were developed.

LITERATURE CITED 1. Fathman CG, Soares L, Chan S, Utz PJ. 2005. An array of possibilities for understanding the genome and proteome in autoimmunity. Nature 435:605–11 2. Anderson AC, Nicholson LB, Legge KL, Turchin V, Zaghouani H, Kuchroo VK. 2000. High frequency of autoreactive myelin proteolipid protein-speciﬁc T cells in the periphery of naive mice: mechanisms of selection of the self-reactive repertoire. J. Exp. Med. 191:761–70 3. Steinman L. 1996. Multiple sclerosis: a coordinated immunological attack against myelin in the central nervous system. Cell 85:299–302 4. Steinman L. 1996. A few autoreactive cells in an autoimmune inﬁltrate control a vast population of nonspeciﬁc cells: a tale of smart bombs and the infantry. Proc. Natl. Acad. Sci. USA 93:2253–56 5. Targoni OS, Baus J, Hofstetter HH, Hesse MD, Karulin AY, et al. 2001. Frequencies of neuroantigen-speciﬁc T cells in the central nervous system versus the immune periphery during the course of experimental allergic encephalomyelitis. J. Immunol. 166:4757– 64 ¨ 6. von Muhlen CA, Tan EM. 1995. Autoantibodies in the diagnosis of systemic rheumatic diseases. Semin. Arthritis Rheum. 24:323–58 7. Craft J, Fatenejad S. 1997. Self antigens and epitope spreading in systemic autoimmunity. Arthritis Rheum. 40:1374–82 8. Robinson WH, DiGennaro C, Hueber W, Haab BB, Kamachi M, et al. 2002. Autoantigen microarrays for multiplex characterization of autoantibody responses. Nat. Med. 8:295– 301 410

Balboni et al.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

9. Robinson WH, Garren H, Utz PJ, Steinman L. 2002. Millennium Award. Proteomics for the development of DNA tolerizing vaccines to treat autoimmune disease. Clin. Immunol. 103:7–12 10. Robinson WH, Steinman L, Utz PJ. 2002. Proteomics technologies for the study of autoimmune disease. Arthritis Rheum. 46:885–93 11. Kanter J, Narayana S, Ho P, Catz I, Warren K, et al. 2006. Lipid microarrays identify key mediators of autoimmune brain inﬂammation. Nat. Med. In press 12. Joos TO, Schrenk M, Hopﬂ P, Kroger K, Chowdhury U, et al. 2000. A microarray enzyme-linked immunosorbent assay for autoimmune diagnostics. Electrophoresis 21:2641–50 13. Haab BB, Dunham MJ, Brown PO. 2001. Protein microarrays for highly parallel detection and quantitation of speciﬁc proteins and antibodies in complex solutions. Genome Biol. 2:RESEARCH0004 14. MacBeath G, Schreiber S. 2000. Printing proteins as microarrays for high-throughput function determination. Science 289:1760–63 15. Deleted in proof 16. Deleted in proof 17. Deleted in proof 18. James JA, Harley JB. 1996. Human lupus anti-spliceosome A protein autoantibodies bind contiguous surface structures and segregate into two sequential epitope binding patterns. J. Immunol. 156:4018–26 19. James JA, Harley JB. 1998. B-cell epitope spreading in autoimmunity. Immunol. Rev. 164:185–200 20. James JA, Mamula MJ, Harley JB. 1994. Sequential autoantigenic determinants of the small nuclear ribonucleoprotein Sm D shared by human lupus autoantibodies and MRL lpr/lpr antibodies. Clin. Exp. Immunol. 98:419–26 21. James JA, Gross T, Scoﬁeld RH, Harley JB. 1995. Immunoglobulin epitope spreading and autoimmune disease after peptide immunization: Sm B/B’-derived PPPGMRPP and PPPGIRGP induce spliceosome autoimmunity. J. Exp. Med. 181:453–61 22. Hueber W, Kidd B, Toomoka B, Lee B-J, Bruce B, et al. 2005. Antigen microarray proﬁling of autoantibodies in rheumatoid arthritis. Arthritis Rheum. 52:2645–55 23. Schellekens GA, de Jong BA, van den Hoogen FH, van de Putte LB, van Venrooij WJ. 1998. Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritis-speciﬁc autoantibodies. J. Clin. Invest. 101:273–81 24. Hiller R, Laffer S, Harwanegg C, Huber M, Schmidt WM, et al. 2002. Microarrayed allergen molecules: diagnostic gatekeepers for allergy treatment. FASEB J. 16:414–16 25. Jahn-Schmid B, Harwanegg C, Hiller R, Bohle B, Ebner C, et al. 2003. Allergen microarray: comparison of microarray using recombinant allergens with conventional diagnostic methods to detect allergen-speciﬁc serum immunoglobulin E. Clin. Exp. Allergy 33:1443– 49 26. Mezzasoma L, Bacarese-Hamilton T, Di Cristina M, Rossi R, Bistoni F, Crisanti A. 2002. Antigen microarrays for serodiagnosis of infectious diseases. Clin. Chem. 48:121–30 27. de Vegvar HEN, Amara RR, Steinman L, Utz PJ, Robinson HL, Robinson WH. 2003. Microarray proﬁling of antibody responses against simian-human immunodeﬁciency virus: postchallenge convergence of reactivities independent of host histocompatibility type and vaccine regimen. J. Virol. 77:11125–38 28. Routsias JG, Tzioufas AG, Moutsopoulos HM. 2004. The clinical value of intracellular autoantigens B-cell epitopes in systemic rheumatic diseases. Clin. Chim. Acta 340:1–25 www.annualreviews.org • Protein Arrays to Study Autoimmunity

411

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

33. Describes a series of elegant experiments that demonstrate the use of arrays to identify antigens to include in a therapeutic vaccine in an animal model of MS.

412

0:17

29. Arbuckle MR, McClain MT, Rubertone MV, Scoﬁeld RH, Dennis GJ, et al. 2003. Development of autoantibodies before the clinical onset of systemic lupus erythematosus. N. Engl. J. Med. 349:1526–33 30. Pietropaolo M, Eisenbarth GS. 2001. Autoantibodies in human diabetes. Curr. Dir. Autoimmun. 4:252–82 31. Verge CF, Gianani R, Kawasaki E, Yu L, Pietropaolo M, et al. 1996. Prediction of type I diabetes in ﬁrst-degree relatives using a combination of insulin, GAD, and ICA512bdc/IA-2 autoantibodies. Diabetes 45:926–33 32. Hueber W, Utz PJ, Steinman L, Robinson WH. 2002. Autoantibody proﬁling for the study and treatment of autoimmune disease. Arthritis Res. 4:290–95 33. Robinson WH, Fontoura P, Lee BJ, de Vegvar HE, Tom J, et al. 2003. Protein microarrays guide tolerizing DNA vaccine treatment of autoimmune encephalomyelitis. Nat. Biotechnol. 21:1033–39 34. Quintana FJ, Hagedorn PH, Elizur G, Merbl Y, Domany E, Cohen IR. 2004. Functional immunomics: microarray analysis of IgG autoantibody repertoires predicts the future response of mice to induced diabetes. Proc. Natl. Acad. Sci. USA 101:14615– 21 35. Wang DN, Liu SY, Trummer BJ, Deng C, Wang AL. 2002. Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20:275–81 36. Hahn BH. 1998. Antibodies to DNA. N. Engl. J. Med. 338:1359–68 37. Stafford HA, Anderson CJ, Blalock DB, Reichlin M. 1998. Development of the antiribosomal P autoantibody response. Clin. Exp. Rheumatol. 16:119–24 38. Warren SJ, Arteaga LA, Rivitti EA, Aoki V, Hans-Filho G, et al. 2003. The role of subclass switching in the pathogenesis of endemic pemphigus foliaceus. J. Invest. Dermatol. 120:104–8 39. Kaufman DL, Clare-Salzler M, Tian J, Forsthuber T, Ting GS, et al. 1993. Spontaneous loss of T-cell tolerance to glutamic acid decarboxylase in murine insulin-dependent diabetes. Nature 366:69–72 40. Keech CL, Gordon TP, McCluskey J. 1996. The immune response to 52-kDa Ro and 60-kDa Ro is linked in experimental autoimmunity. J. Immunol. 157:3694–99 41. Lehmann PV, Forsthuber T, Miller A, Sercarz EE. 1992. Spreading of T-cell autoimmunity to cryptic determinants of an autoantigen. Nature 358:155–57 42. McCarron RM, Fallis RJ, McFarlin DE. 1990. Alterations in T cell antigen speciﬁcity and class II restriction during the course of chronic relapsing experimental allergic encephalomyelitis. J. Neuroimmunol. 29:73–79 43. Tisch R, Yang XD, Singer SM, Liblau RS, Fugger L, McDevitt HO. 1993. Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice. Nature 366:72–75 44. Topfer F, Gordon T, McCluskey J. 1995. Intra- and intermolecular spreading of autoimmunity involving the nuclear self-antigens La (SS-B) and Ro (SS-A). Proc. Natl. Acad. Sci. USA 92:875–79 45. Arbuckle MR, Reichlin M, Harley JB, James JA. 1999. Shared early autoantibody recognition events in the development of anti-Sm B/B’ in human lupus. Scand. J. Immunol. 50:447–55 46. Orth T, Mayet WJ, Deister H, Chang S, Schmitz M, Bachmann M. 1998. Analysis of epitope spreading over an eleven-year period in a patient with systemic lupus erythematosus. Scand. J. Rheumatol. 27:461–64 Balboni et al.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

47. Scoﬁeld RH, Zhang F, Kurien BT, Anderson CJ, Reichlin M, et al. 1996. Development of the anti-Ro autoantibody response in a patient with systemic lupus erythematosus. Arthritis Rheum. 39:1664–68 48. Naserke HE, Ziegler AG, Lampasona V, Bonifacio E. 1998. Early development and spreading of autoantibodies to epitopes of IA-2 and their association with progression to type 1 diabetes. J. Immunol. 161:6963–69 49. Sohnlein P, Muller M, Syren K, Hartmann U, Bohm BO, et al. 2000. Epitope spreading and a varying but not disease-speciﬁc GAD65 antibody response in Type I diabetes. The Childhood Diabetes in Finland Study Group. Diabetologia 43:210–17 50. Lueking A, Huber O, Wirths C, Schulte K, Stieler K, et al. 2005. Proﬁling of alopecia areata autoantigens based on protein microarray technology. Mol. Cell Proteomics 4:1382– 90 51. Ramachandran N, Hainsworth E, Bhullar B, Eisenstein S, Rosen B, et al. 2004. Selfassembling protein microarrays. Science 305:86–90 52. Angenendt P, Glokler J, Murphy D, Lehrach H, Cahill DJ. 2002. Toward optimized antibody microarrays: a comparison of current microarray support materials. Anal. Biochem. 309:253–60 53. Angenendt P, Glokler J, Sobek J, Lehrach H, Cahill DJ. 2003. Next generation of protein microarray support materials: evaluation for protein and antibody microarray applications. J. Chromatogr. A 1009:97–104 54. Feldmann M. 2002. Development of anti-TNF therapy for rheumatoid arthritis. Nat. Rev. Immunol. 2:364–71 55. Feldmann M, Steinman L. 2005. Design of effective immunotherapy for human autoimmunity. Nature 435:612–19 56. Vilcek J, Feldmann M. 2004. Historical review: cytokines as therapeutics and targets of therapeutics. Trends Pharmacol. Sci. 25:201–9 57. Revel M. 2003. Interferon-β in the treatment of relapsing-remitting multiple sclerosis. Pharmacol. Ther. 100:49–62 58. Bresnihan B, Alvaro-Gracia JM, Cobby M, Doherty M, Domljan Z, et al. 1998. Treatment of rheumatoid arthritis with recombinant human interleukin-1 receptor antagonist. Arthritis Rheum. 41:2196–204 59. McInnes IB, Gracie JA. 2004. Interleukin-15: a new cytokine target for the treatment of inﬂammatory diseases. Curr. Opin. Pharmacol. 4:392–97 60. Nishimoto N, Yoshizaki K, Miyasaka N, Yamamoto K, Kawai S, et al. 2004. Treatment of rheumatoid arthritis with humanized anti-interleukin-6 receptor antibody: a multicenter, double-blind, placebo-controlled trial. Arthritis Rheum. 50:1761–69 61. Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, et al. 2003. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc. Natl. Acad. Sci. USA 100:2610–15 62. Bennett L, Palucka AK, Arce E, Cantrell V, Borvak J, et al. 2003. Interferon and granulopoiesis signatures in systemic lupus erythematosus blood. J. Exp. Med. 197:711– 23 63. Kirou KA, Lee C, George S, Louca K, Peterson MG, Crow MK. 2005. Activation of the interferon-α pathway identiﬁes a subgroup of systemic lupus erythematosus patients with distinct serologic features and active disease. Arthritis Rheum. 52:1491–503 64. Bengtsson AA, Sturfelt G, Truedsson L, Blomberg J, Alm G, et al. 2000. Activation of type I interferon system in systemic lupus erythematosus correlates with disease activity but not with antiretroviral antibodies. Lupus 9:664–71 www.annualreviews.org • Protein Arrays to Study Autoimmunity

413

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

76. Investigated cytokines and growth factors in inflammatory bowel disease using cytokine arrays; one of the best demonstrations of the clinical applications of this technology.

78. Analyzed healthy individuals and patients with ¨ primary Sjogren’s syndrome using bead-based arrays; found certain cytokine profiles correlate with specific laboratory and clinical parameters.

414

0:17

65. Magnano MD, Robinson WH, Genovese MC. 2004. Demyelination and inhibition of tumor necrosis factor (TNF). Clin. Exp. Rheumatol. 22:S134–40 66. Huang RP. 2004. Cytokine protein arrays. Methods Mol. Biol. 264:215–31 67. Tam SW, Wiese R, Lee S, Gilmore J, Kumble KD. 2002. Simultaneous analysis of eight human Th1/Th2 cytokines using microarrays. J. Immunol. Methods 261:157–65 68. Wang CC, Huang RP, Sommer M, Lisoukov H, Huang R, et al. 2002. Array-based multiplexed screening and quantitation of human cytokines and chemokines. J. Proteome Res. 1:337–43 69. Moody MD, Van Arsdell SW, Murphy KP, Orencole SF, Burns C. 2001. Array-based ELISAs for high-throughput analysis of human cytokines. BioTechniques 31:186–90, 192– 94 70. Copeland S, Siddiqui J, Remick D. 2004. Direct comparison of traditional ELISAs and membrane protein arrays for detection and quantiﬁcation of human cytokines. J. Immunol. Methods 284:99–106 71. Nagorsen D, Marincola FM, Panelli MC. 2004. Cytokine and chemokine expression proﬁles of maturing dendritic cells using multiprotein platform arrays. Cytokine 25:31– 35 72. Schweitzer B, Roberts S, Grimwade B, Shao WP, Wang MJ, et al. 2002. Multiplexed protein proﬁling on microarrays by rolling-circle ampliﬁcation. Nat. Biotechnol. 20:359– 65 73. Schweitzer B, Wiltshire S, Lambert J, O’Malley S, Kukanskis K, et al. 2000. Inaugural article: immunoassays with rolling circle DNA ampliﬁcation: a versatile platform for ultrasensitive antigen detection. Proc. Natl. Acad. Sci. USA 97:10113–19 74. Shao WP, Zhou ZM, Laroche I, Lu H, Zong QL, et al. 2003. Optimization of rollingcircle ampliﬁed protein microarrays for multiplexed protein proﬁling. J. Biomed. Biotechnol. 2003:299–307 75. Zhou H, Bouwman K, Schotanus M, Verweij C, Marrero JA, et al. 2004. Two-color, rolling-circle ampliﬁcation on antibody microarrays for sensitive, multiplexed serumprotein measurements. Genome Biol. 5:R28 76. Kader HA, Tchernev VT, Satyaraj E, Lejnine S, Kotler G, et al. 2005. Protein microarray analysis of disease activity in pediatric inflammatory bowel disease demonstrates elevated serum PLGF, IL-7, TGF-β1, and IL-12p40 levels in Crohn’s disease and ulcerative colitis patients in remission versus active disease. Am. J. Gastroenterol. 100:414–23 77. Nolan JP, Sklar LA. 2002. Suspension array technology: evolution of the ﬂat-array paradigm. Trends Biotechnol. 20:9–12 78. Szodoray P, Alex P, Brun JG, Centola M, Jonsson R. 2004. Circulating cytokines ¨ in primary Sjogren’s syndrome determined by a multiplex cytokine array system. Scand. J. Immunol. 59:592–99 79. Chan-Hui P-Y, Stephens K, Warnock RA, Singh S. 2003. Applications of eTag(R)(TM) assay platform to systems biology approaches to clinical immunology studies. Clin. Immunol. 111:162–74 80. Fredriksson S, Gullberg M, Jarvius J, Olsson C, Pietras K, et al. 2002. Protein detection using proximity-dependent DNA ligation assays. Nat. Biotechnol. 20:473– 77 81. Gullberg M, Gustafsdottir SM, Schallmeiner E, Jarvius J, Bjarnegard M, et al. 2004. Cytokine detection by antibody-based proximity ligation. Proc. Natl. Acad. Sci. USA 101:8420–24 Balboni et al.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

82. Paweletz CP, Charboneau L, Bichsel VE, Simone NL, Chen T, et al. 2001. Reverse phase protein microarrays which capture disease progression show activation of pro-survival pathways at the cancer invasion front. Oncogene 20:1981–89 83. Macbeath G. 2002. Protein microarrays and proteomics. Nat. Genet. 32:526–32 84. Nielsen UB, Cardone MH, Sinskey AJ, MacBeath G, Sorger PK. 2003. Proﬁling receptor tyrosine kinase activation by using Ab microarrays. Proc. Natl. Acad. Sci. USA 100:9330– 35 85. Sreekumar A, Nyati MK, Varambally S, Barrette TR, Ghosh D, et al. 2001. Proﬁling of cancer cells using protein microarrays: discovery of novel radiation-regulated proteins. Cancer Res. 61:7585–93 86. Utz PJ. 2005. Protein arrays for studying blood cells and their secreted products. Immunol. Rev. 204:264–82 87. Geho D, Lahar N, Gurnani P, Huebschman M, Herrmann P, et al. 2005. Pegylated, steptavidin-conjugated quantum dots are effective detection elements for reverse-phase protein microarrays. Bioconjug. Chem. 16:559–66 88. Chan SM, Ermann J, Su L, Fathman CG, Utz PJ. 2004. Protein microarrays for multiplex analysis of signal transduction pathways. Nat. Med. 10:1390–96 89. Nishizuka S, Charboneau L, Young L, Major S, Reinhold WC, et al. 2003. Proteomic proﬁling of the NCI-60 cancer cell lines using new high-density reverse-phase lysate microarrays. Proc. Natl. Acad. Sci. USA 100:14229–34 90. Nishizuka S, Chen ST, Gwadry FG, Alexander J, Major SM, et al. 2003. Diagnostic markers that distinguish colon and ovarian adenocarcinomas: identiﬁcation by genomic, proteomic, and tissue array proﬁling. Cancer Res. 63:5243–50 91. Jones MB, Krutzsch H, Shu H, Zhao Y, Liotta LA, et al. 2002. Proteomic analysis and identiﬁcation of new biomarkers and therapeutic targets for invasive ovarian cancer. Proteomics 2:76–84 92. Herrmann PC, Gillespie JW, Charboneau L, Bichsel VE, Paweletz CP, et al. 2003. Mitochondrial proteome: altered cytochrome c oxidase subunit levels in prostate cancer. Proteomics 3:1801–10 93. Zha H, Raffeld M, Charboneau L, Pittaluga S, Kwak LW, et al. 2004. Similarities of prosurvival signals in Bcl-2-positive and Bcl-2-negative follicular lymphomas identiﬁed by reverse phase protein microarray. Lab. Invest. 84:235–44 94. Wulfkuhle JD, Aquino JA, Calvert VS, Fishman DA, Coukos G, et al. 2003. Signal pathway proﬁling of ovarian cancer from human tissue specimens using reverse-phase protein microarrays. Proteomics 3:2085–90 95. Grubb RL, Calvert VS, Wulkuhle JD, Paweletz CP, Linehan WM, et al. 2003. Signal pathway proﬁling of prostate cancer using reverse phase protein arrays. Proteomics 3:2142– 46 96. Sheehan KM, Calvert VS, Kay EW, Lu Y, Fishman D, et al. 2005. Use of reverse phase protein microarrays and reference standard development for molecular network analysis of metastatic ovarian carcinoma. Mol. Cell Proteomics 4:346–55 97. Petricoin EF 3rd, Bichsel VE, Calvert VS, Espina V, Winters M, et al. 2005. Mapping molecular networks using proteomics: a vision for patient-tailored combination therapy. J. Clin. Oncol. 23:3614–21 98. Petricoin E, Zoon K, Kohn E, Barrett J, Liotta L. 2002. Clinical proteomics: translating benchside promise into bedside reality. Nat. Rev. 1:683–95 98a. Irish JM, Hovland R, Krutzik PO, Perez OD, Bruserud O, et al. 2004. Single cell proﬁling of potentiated phospho-protein networks in cancer cells. Cell 118:217–28 www.annualreviews.org • Protein Arrays to Study Autoimmunity

82. The original description of the RPL microarray technology in combination with LCM.

88. Employed RPL microarray technology to study dynamic changes in the phosphorylation state of numerous signaling proteins in T cells upon stimulation.

415

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

104. Describes planar microarrays of recombinant MHC-peptide complexes, costimulatory molecules, and anti-cytokine capture antibodies.

416

0:17

98b. Perez OD, Nolan GP. 2002. Simultaneous measurement of multiple active kinase states using polychromatic ﬂow cytometry. Nat. Biotechnol. 20:155–62 99. Tang Q, Henriksen KJ, Bi M, Finger EB, Szot G, et al. 2004. In vitro-expanded antigen-speciﬁc regulatory T cells suppress autoimmune diabetes. J. Exp. Med. 199:1455– 65 100. Sakaguchi S. 2004. Naturally arising CD4+ regulatory T cells for immunologic selftolerance and negative control of immune responses. Annu. Rev. Immunol. 22:531– 62 101. Perfetto SP, Chattopadhyay PK, Roederer M. 2004. Seventeen-colour ﬂow cytometry: unravelling the immune system. Nat. Rev. Immunol. 4:648–55 102. Edwards BS, Oprea T, Prossnitz ER, Sklar LA. 2004. Flow cytometry for highthroughput, high-content screening. Curr. Opin. Chem. Biol. 8:392–98 103. Soen Y, Chen DS, Kraft DL, Davis MM, Brown PO. 2003. Detection and characterization of cellular immune responses using peptide-MHC microarrays. PLoS Biol. 1:E65 104. Stone JD, Demkowicz WE Jr, Stern LJ. 2005. HLA-restricted epitope identification and detection of functional T cell responses by using MHC-peptide and costimulatory microarrays. Proc. Natl. Acad. Sci. USA 102:3744–49 105. Belov L, de la Vega O, dos Remedios CG, Mulligan SP, Christopherson RI. 2001. Immunophenotyping of leukemias using a cluster of differentiation antibody microarray. Cancer Res. 61:4483–89 106. Belov L, Huang P, Barber N, Mulligan SP, Christopherson RI. 2003. Identiﬁcation of repertoires of surface antigens on leukemias using an antibody microarray. Proteomics 3:2147–54 107. Appel RD, Hochstrasser DF, Funk M, Vargas JR, Pellegrini C, et al. 1991. The MELANIE project: from a biopsy to automatic protein map interpretation by computer. Electrophoresis 12:722–35 108. Appel RD, Vargas JR, Palagi PM, Walther D, Hochstrasser DF. 1997. Melanie II—a third-generation software package for analysis of two-dimensional electrophoresis images: II. Algorithms. Electrophoresis 18:2735–48 109. Marengo E, Robotti E, Antonucci F, Cecconi D, Campostrini N, Righetti PG. 2005. Numerical approaches for quantitative analysis of two-dimensional maps: a review of commercial software and home-made systems. Proteomics 5:654–66 110. Sanchez J, Chiappe D, Converset V, Hoogland C, Binz P, et al. 2001. The mouse SWISS2D PAGE database: a tool for proteomics study of diabetes and obesity. Proteomics 1:136– 63 111. Gygi S, Corthals G, Zhang Y, Rochon Y, Aebersold R. 2000. Evaluation of twodimensional gel electrophoresis-based proteome analysis technology. Proc. Natl. Acad. Sci. USA 97:9390–95 112. Aebersold R, Mann M. 2003. Mass spectrometry-based proteomics. Nature 422:198–207 113. Gevaert K, Vandekerckhove J. 2000. Protein identiﬁcation methods in proteomics. Electrophoresis 21:1145–54 114. Perkins DN, Pappin DJ, Creasy DM, Cottrell JS. 1999. Probability-based protein identiﬁcation by searching sequence databases using mass spectrometry data. Electrophoresis 20:3551–67 115. Bruni R, Gianfranceschi G, Koch G. 2005. On peptide de novo sequencing: a new approach. J. Pept. Sci. 11:225–34 116. Lin T, Glish GL. 1998. C-terminal peptide sequencing via multistage mass spectrometry. Anal. Chem. 70:5162–65 Balboni et al.

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-13

ARI

15 February 2006

0:17

117. Stults JT. 1990. Peptide sequencing by mass spectrometry. Methods Biochem. Anal. 34: 145–201 118. Taka H, Kaga N, Fujimura T, Mineki R, Imaizumi M, et al. 2000. Rapid determination of parvalbumin amino acid sequence from Rana catesbeiana (pI 4.78) by combination of ESI mass spectrometry, protein sequencing, and amino acid analysis. J. Biochem. 127:723– 29 119. Oda Y, Owa T, Sato T, Boucher B, Daniels S, et al. 2003. Quantitative chemical proteomics for identifying candidate drug targets. Anal. Chem. 75:2159–65 120. Barrett T, Suzek TO, Troup DB, Wilhite SE, Ngau WC, et al. 2005. NCBI GEO: mining millions of expression proﬁles—database and tools. Nucleic Acids Res. 33:D562– 66 121. Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, et al. 2001. Minimum information about a microarray experiment (MIAME)—toward standards for microarray data. Nat. Genet. 29:365–71 122. Hermjakob H, Montecchi-Palazzi L, Bader G, Wojcik J, Salwinski L, et al. 2004. The HUPO PSI’s molecular interaction format—a community standard for the representation of protein interaction data. Nat. Biotechnol. 22:177–83 123. Orchard S, Hermjakob H, Julian RK Jr, Runte K, Sherman D, et al. 2004. Common interchange standards for proteomics data: public availability of tools and schema. Proteomics 4:490–91 124. Taylor CF, Paton NW, Garwood KL, Kirby PD, Stead DA, et al. 2003. A systematic approach to modeling, capturing, and disseminating proteomics experimental data. Nat. Biotechnol. 21:247–54 125. Hucka M, Finney A, Sauro HM, Bolouri H, Doyle JC, et al. 2003. The systems biology markup language (SBML): a medium for representation and exchange of biochemical network models. Bioinformatics 19:524–31 126. Demir E, Babur O, Dogrusoz U, Gursoy A, Nisanci G, et al. 2002. PATIKA: an integrated visual environment for collaborative construction and analysis of cellular pathways. Bioinformatics 18:996–1003 127. Nikitin A, Egorov S, Daraselia N, Mazo I. 2003. Pathway studio—the analysis and navigation of molecular networks. Bioinformatics 19:2155–57 128. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, et al. 2003. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 13:2498–504 129. Doniger SW, Salomonis N, Dahlquist KD, Vranizan K, Lawlor SC, Conklin BR. 2003. MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression proﬁle from microarray data. Genome Biol. 4:R7 130. Draghici S, Khatri P, Bhavsar P, Shah A, Krawetz SA, Tainsky MA. 2003. Onto-Tools, the toolkit of the modern biologist: Onto-Express, Onto-Compare, Onto-Design and Onto-Translate. Nucleic Acids Res. 31:3775–81 131. Schoeberl B, Eichler-Jonsson C, Gilles ED, Muller G. 2002. Computational modeling of the dynamics of the MAP kinase cascade activated by surface and internalized EGF receptors. Nat. Biotechnol. 20:370–75 132. Sachs K, Perez O, Pe’er D, Lauffenburger DA, Nolan GP. 2005. Causal proteinsignaling networks derived from multiparameter single-cell data. Science 308:523– 29 133. Ideker T, Ozier O, Schwikowski B, Siegel AF. 2002. Discovering regulatory and signalling circuits in molecular interaction networks. Bioinformatics 18(Suppl. 1):S233–40 www.annualreviews.org • Protein Arrays to Study Autoimmunity

121. Put forth standards that have been adopted by numerous journals and repositories as a required format for sharing microarray data.

131. Demonstrated that detailed modeling of a biological system can be accurate and robust as determined by experimental findings and sensitivity analysis.

132. Set a precedent for using a data set with the thousands of replicates commonly desired in Bayesian analysis; led to two experimentally verifiable biological discoveries in a relatively well-known system.

417

ANRV270-IY24-13

ARI

15 February 2006

0:17

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

134. Mootha VK, Lepage P, Miller K, Bunkenborg J, Reich M, et al. 2003. Identiﬁcation of a gene causing human cytochrome c oxidase deﬁciency by integrative genomics. Proc. Natl. Acad. Sci. USA 100:605–10 135. Rhyne PW, Scull JD, Stiles LM, Eisinger DP. 2003. Analysis of apoptotic cells using Beadlyte suspension arrays. Biotechniques 35:624–29

418

Balboni et al.

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:391-418. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-14

ARI

15 February 2006

0:51

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph,1 Robyn L. Stanﬁeld,2 and Ian A. Wilson2,3 1

¨ Department of Molecular Structural Biology, University of Gottingen, ¨ 37077 Gottingen, Germany; email: [email protected]

2

Department of Molecular Biology, The Scripps Research Institute, and 3 The Skaggs Institute for Chemical Biology, La Jolla, California 92037; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:419–466 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.23.021704.115658 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0419$20.00

Key Words T cell receptor, major histocompatibility complex, protein-protein interaction, crystal structure, immunological synapse

Abstract Since the ﬁrst crystal structure determinations of αβ T cell receptors (TCRs) bound to class I MHC-peptide (pMHC) antigens in 1996, a sizable database of 24 class I and class II TCR/pMHC complexes has been accumulated that now deﬁnes a substantial degree of structural variability in TCR/pMHC recognition. Recent determination of free and bound γδ TCR structures has enabled comparisons of the modes of antigen recognition by αβ and γδ T cells and antibodies. Crystal structures of TCR accessory (CD4, CD8) and coreceptor molecules (CD3εδ, CD3εγ) have further advanced our structural understanding of most of the components that constitute the TCR signaling complex. Despite all these efforts, the structural basis for MHC restriction and signaling remains elusive as no structural features that deﬁne a common binding mode or signaling mechanism have yet been gleaned from the current set of TCR/pMHC complexes. Notwithstanding, the impressive array of self, foreign (microbial), and autoimmune TCR complexes have uncovered the diverse ways in which antigens can be speciﬁcally recognized by TCRs.

419

ANRV270-IY24-14

ARI

15 February 2006

0:51

INTRODUCTION

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Humoral (antibody-mediated) and cellular (T cell–mediated) immunity are the two main lines of defense that higher organisms rely on for combating microbial pathogens. While antibodies recognize intact antigens, T cells distinguish foreign material from self through presentation of fragments of the antigen by the MHC cell surface receptors. Only if an MHC molecule presents an appropriate antigenic peptide will a cellular immune response be triggered. The orchestration of recognition and signaling events, from the initial recognition of antigenic peptides to the lysis of the target cell, is performed in a localized environment on the T cell, called the immunological synapse, and requires the coordinated activities of several TCR-associated molecules, including coreceptors CD3 and CD8 or CD4, and other costimulatory receptors. Insights into TCR structure have come from crystallized TCR fragments and individual chains (1–6), intact TCR ectodomains (7– 10), and TCR/pMHC complexes (7, 11–31) (Figure 1). Analysis of the current database of 24 TCR/pMHC complexes has resolved many pressing questions in cellular immunity, but other issues have not yet been clariﬁed, particularly in regard to what constitutes the structural basis of MHC restriction and its implications for positive and negative selection. Further, how do TCRs distinguish between agonist, partial agonist, and antagonist ligands in order to produce different signaling outcomes? One serious obstacle remaining is the generation of sufﬁcient quantities of soluble TCR/pMHC complexes for crystallographic structure determination. Despite the presence of multiple disulﬁde bonds in these heterodimeric complexes, many TCRs and MHCs have been produced and refolded from Escherichia coli inclusion bodies (Table 1). Some MHCs have been produced in insect cells, such as Drosophila melanogaster (K-2Kb , HLA-DR1, HLA-DR4, I-Au ) or Spodoptera frugiperda (HLA-DR2), and TCRs have been

420

Rudolph

·

Stanfield

·

Wilson

produced in D. melanogaster (2C), Trichoplusia ni (γδ TCR), and S. frugiperda (Ob.1A12) systems (Table 1). Mammalian myeloma cells enabled production of the scBM3.3 and scKB5-C20 TCRs. To increase peptide afﬁnities and to reduce the unfavorable change in entropy during complex formation, stable complexes have also been engineered by covalently attaching the antigenic peptide to either the N terminus of the β-chain of class II MHC (15) or the N terminus of the TCR β-chain (18, 27, 29). In this review, we discuss the recent advances in our understanding of TCR/pMHC recognition and signaling (via associated coreceptors) and outline some important questions that remain unanswered. For other notable previous reviews on this topic, see References 32–38.

ARCHITECTURE OF MHCs AND TCRs Structural Variation and Functional Promiscuity of the MHC Fold In the cellular immune response, antigens, generally peptides, are displayed to αβ T cells in complex with class I or class II MHC molecules. Both classes of MHC are heterodimers with similar architectures and are composed of three domains, one α-helix/ β-sheet (αβ) superdomain that forms the peptide-binding site and two Ig-like domains. In class I MHC molecules, the peptide-binding site (called the α1 α2 domain) is constructed from the heavy chain only, and an additional light chain subunit, β2 -microglobulin (β2 m), associates with α3 of the heavy chain. In contrast, the class II MHC peptide-binding site is assembled from two heavy chains (α1 β1 ). Notwithstanding, in both MHC classes, the overall architecture is the same where a seven-stranded β-sheet represents the ﬂoor of the binding groove, and the sides are formed by two long α-helices (or continuous α-helical segments in the α2 - or β1 -helices) that straddle the

Cumulative number of structures

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

MHC structures

200

TCR structures TCR/pMHC structures

150

100

50

0

1988

1990

1992

1994

1986

1998

2000

2002

2004

Year Figure 1 Cumulative number of pMHC [unliganded or with ligand (peptide, superantigen, etc.)], TCR (unliganded or with ligand other than pMHC), and TCR/pMHC complex crystal structures. The number of structures is plotted as a function of their deposition year in the Protein Data Bank (PDB) (151). The plot does not contain structures that were superseded by redetermination at higher resolution. However, MHC and TCR complexes with other molecules, such as superantigens or antibodies, were included. For the TCRs, all fragments and constructs (such as single chains), which were determined by either X-ray diffraction or NMR spectroscopy, are included. The ﬁrst MHC crystal structure was determined in 1987 (152), and after an approximately ﬁve-year lag, the number of MHC structures increased drastically, with 39 structures added to the PDB in 2004. Since the ﬁrst TCR and TCR/pMHC structures in 1996, no such dramatic increase has yet been seen in the annual output of new TCR or TCR/pMHC structures.

β-sheet (Figure 2a,b). Polymorphic residues cluster within and around the binding groove in order to provide the required variation in shape and chemical properties that accounts for the speciﬁc peptide-binding motifs identiﬁed for each MHC allele (39–41). Class I MHC molecules usually bind peptides of 8–10 residues length (on average 9-mers, P1–P9) (Figure 3) in an extended conformation with the termini and the socalled anchor residues buried in speciﬁcity pockets that differ from allele to allele (42, 43). This binding mode leaves the upwardpointing peptide side chains available for direct interaction with the TCR (Figure 3).

Longer peptides can either bind by extension at the C terminus (44) or, due to the ﬁxing of their termini, bulge out of the binding groove, providing additional surface area for TCR recognition (22, 45). In class II MHC, the groove is open at either end, and the peptide termini are not ﬁxed so that bound peptides are usually signiﬁcantly longer than in MHC class I (Figure 3). The peptide backbone in class II MHC is conﬁned mainly to a polyproline type II conformation (44) and resides slightly deeper in the binding groove. Thus, the bound peptide (P1–P9) is more accessible for TCR inspection in MHC class I due to its ability to bulge out of the groove, even for www.annualreviews.org • MHC/ TCR Interactions

421

ANRV270-IY24-14

Table 1

ARI

15 February 2006

0:51

Overview of TCR/pMHC complex structures, 1996–2005

Complex

Peptide activity

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MHC class I 2C/H-2Kb /dEV8

Weak agonist

2C/H-2Kb /SIYR 2C/H-2Kbm3 /dEV8 scBM3.3/H-2Kb /pBM1

Superagonist Weak agonist Agonist (naturally processed)

scBM3.3/H-2Kb /VSV8 scKB5-C20/H-2Kb /pKB1

Agonist Agonist (naturally processed)

B7/HLA-A2/Tax A6/HLA-A2/Tax A6/HLA-A2/TaxP6A A6/HLA-A2/TaxV7R A6/HLA-A2/TaxY8A JM22/HLA-A2/MP

Strong agonist Strong agonist Weak antagonist Weak agonist Weak antagonist Agonist

1G4/HLA-A2/ESO9V 1G4/HLA-A2/ESO9C AHIII12.2/HLA-A2.1/p1049 SB27/HLA-B3508/EBV LC13/HLA-B8/FLR

Agonist Agonist Agonist (xenoreactive) Agonist Agonist (immuno-dominant)

MHC class II scD10/I-Ak /CA

Agonist

HA1.7/HLA-DR1/HA

Agonist

HA1.7/HLA-DR4/HA Ob.1A12/HLA-DR2b/MBP

Agonist Agonist (autoreactive self-peptide)

sc172.10/I-Au /MBP

Agonist (autoreactive self-peptide) Agonist (autoreactive self-peptide)

3A6/HLA-DR2a/MBP

γδ TCR/H2-T22

—

Constructs and expression systems

Reference

D. melanogaster, acidic/basic leucine zipper for speciﬁc TCR chain-pairing As above As above Myeloma cells for TCR, E. coli for MHC (refolded from inclusion bodies) As above Myeloma cells for TCR, E. coli for MHC (refolded from inclusion bodies) E. coli, refolded from inclusion bodies E. coli, refolded from inclusion bodies As above As above As above E. coli, refolded from inclusion bodies. C-terminal extension of TCR chains coding for a cysteine to promote disulﬁde formation E. coli, refolded from inclusion bodies As above E. coli, refolded from inclusion bodies E. coli, refolded from inclusion bodies E. coli, refolded from inclusion bodies

(12)

E. coli for TCR, refolded from inclusion bodies; CHO cells for MHC. Peptide covalently connected to the MHC. E. coli for TCR, refolded from inclusion bodies; D. melanogaster for MHC. Peptide covalently connected to the TCR β-chain. As above Baculovirus-infected S. frugiperda cells for both HLA-DR2 and TCR. Peptide covalently attached to the N terminus of the TCR β-chain. Jun/Fos leucine zipper for speciﬁc TCR chain-pairing. E. coli periplasm for TCR, D. melanogaster for MHC E. coli, refolded from inclusion bodies for MHC and TCR; peptide covalently connected to the N terminus of the TCR β-chain.

(15)

Baculovirus-infected Trichoplusia ni cells, acidic/basic leucine zipper for speciﬁc TCR chain-pairing

(102)

(17) (19) (16) (30) (31) (13) (11) (14) (14) (14) (21)

(25) (25) (20) (23) (24)

(18)

(153) (27)

(28) (29)

αβ class I, class II, and γδ TCR complexes are separated by horizontal lines. (Abbreviations: sc, single chain Fv fragment of the TCR.)

422

Rudolph

·

Stanfield

·

Wilson

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

Figure 2 Architecture of MHC-like molecules. The top panel shows the domain organization of the MHC(-like) molecules and the lower panel focuses on the ligand and/or receptor binding sites. (a) Class I molecules consist of a heavy chain (blue) and a light β2 m chain (orange). The peptide-binding site is formed exclusively by elements of the heavy chain, whereas in class II molecules (b), it is assembled from both subunits. (c) The nonclassical MHC-like molecule MICA, which is a ligand for the natural killer (NK) cell receptor NKG2D, is structurally analogous to a class I molecule but lacks the β2 m subunit. (d ) The NKG2D ligand Rae-1β is formed solely by the α1 α2 platform, so that the α3 domain is expendable. (e) A view from the TCR perspective onto the class I peptide-binding site with the peptide drawn as a stick model and atoms colored according to atom type. The α1 - and α2 -helices close off the ends of the groove, ﬁxing the N and C termini of the peptide in the A and F pockets, respectively. ( f ) In class II molecules, the helices bordering the peptide are shorter and less curved, allowing the peptide to protrude from the ends of the groove. ( g) Closer proximity of the helices and a hydrophobic binding groove are the hallmarks of the CD1 binding pocket for binding lipids, glycolipids, and lipopeptides. (h) In the nonclassical MHC molecule T22, which is a γδ T cell ligand, part of the α2 -helix has unwound, exposing one end of the underlying β-sheet. The newly acquired loop region (shown in dark gray) apparently is ﬂexible as judged by the very high B values of the structure in this region. (i ) No small molecule ligand can be bound by Rae-1β as the distance between the helices is minimal, which permits formation of an interhelical disulﬁde bond.

9-mer peptides; however, in MHC class II, the termini, particularly the N-terminal extension (P-4 to P-1), can play a major role in the TCR interaction. Apart from displaying peptides to TCRs, the MHC fold has garnered many other functions during evolution that impact its domain organization and ﬂexibility, as well as its substrate speciﬁcity. For instance, in the nonclassical MHC molecule CD1, the ligandbinding groove is deeper, narrower, and more hydrophobic than in class I MHCs, such that lipid tails of glycolipids and lipopeptides are

bound in the groove and their polar moieties presented to T cells (46–55) (Figure 2g). Other MHC-like molecules do not seem to present any antigen, such as γδ TCR ligands T10 (56) and T22 (57). In these structures, a 13-residue sequence deletion results in the partial unfolding of the α2 -helix and a concomitant exposure of the β-sheet ﬂoor of the α1 α2 domain (Figure 2h). This “rupture” of the ligand-binding site appears to account for the loss of peptide or other small molecule ligand-binding capability, although, initially, questions arose whether this disordered loop www.annualreviews.org • MHC/ TCR Interactions

423

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

0:51

Figure 3 Comparison of peptide conformations as observed in class I (top) and class II (bottom) TCR/pMHC complexes. The Cα traces of the bound peptides (removed from their respective MHCs) are drawn as tubes with the TCR-contacting side chains (see Table 3) as stick representations. Class I–bound peptides of 8, 9, and 13 residues are colored yellow, red, and green, respectively. Peptides from class II complexes are colored yellow. The peptides are oriented with their TCR-contacting residues pointing upward. The β-sheet ﬂoors of the peptide-binding sites of the MHC molecules were superimposed to align the peptides. TCR interaction with the central P1–P9 residues is common to both class I and class II, but the bound peptides adopt substantially different conformations.

region would fold back into an α-helix upon TCR binding. Yet another class of nonclassical MHC molecules that apparently lacks any afﬁnity for small molecule antigens comprises ligands for the NK cell activating receptor NKG2D (58– 61), i.e., human MICA, MICB, ULBP, and murine Rae1 and H60 (62). These cell surface receptors serve as general stress sensors and, as they do not present peptide antigens, are independent of transporter associated with antigen processing (TAPs) (63). Their expression levels are low and the receptors are displayed on ﬁbroblast, epithelial, dendritic, and endothelial cells only in response to stress such as heat shock, oxidative stress, bacterial infection, and tumor growth (64, 65). Crystal structures of MICA (66) and Rae-1β (67) indicated that the loss of peptide, or any 424

Rudolph

·

Stanfield

·

Wilson

other ligand, binding was due to elimination of any binding groove because of the reduced distance between the α1 - and α2 -helices (Figure 2i ). In Rae-1β, these helices come close enough to permit formation of a noncanonical disulﬁde bond with a leucine-rich interface ﬁlling the former ligand-binding cavity. Thus, natural evolution of the MHC fold has taken nonclassical MHCs even further from the canonical MHC fold. In contrast to class I MHCs, MIC proteins do not associate with β2 m, and H60 and Rae-1β are even simpler modules as they dispense with an α3 domain and exist only as an isolated α1 α2 platform (Figure 2). Receptor binding to MHCs is complemented by additional interaction events prior to T cell or killer cell activation. Coreceptors CD4 and CD8 bind not only to the underside

ANRV270-IY24-14

ARI

15 February 2006

0:51

of the α1 α2 platform and α3 domain of pMHCs, but also to nonclassical MHCs, such as the thymic leukemia tumor antigen TL. TL modulates T cell activation through a moderate afﬁnity (Kd = 10 μM) interaction with the CD8 coreceptor, but also does not serve as an antigen-presenting molecule, because its binding site is also occluded by close packing of the α1 - and α2 -helices (68).

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

αβ and γδ TCRs TCRs are cell surface heterodimers consisting of either disulﬁde-linked α- and β- or γ- and δ-chains. Sequence analyses correctly predicted that TCRs would share a domain organization and binding mode similar to those of antibody Fab fragments (69, 70). Each TCR chain is composed of variable and constant Iglike domains, followed by a transmembrane domain and a short cytoplasmic tail. The αβ TCRs bind pMHC with relatively low afﬁnity (∼1–100 μM) through complementaritydetermining regions (CDRs) present in their variable domains. Compared with αβ TCRs, where a variety of structures have been determined since 1996, much less is known about γδ TCRs. The only structure available until recently was that of a Vδ domain (71). This lack of structural information was paralleled by the ill-deﬁned biological function of γδ T cells. γδ T cells appear to respond to bacterial and parasitic infections (72) and primarily recognize phosphate-containing antigens (phosphoantigens) from mycobacteria by an unknown mechanism (72, 73). Other identiﬁed ligands (74) for γδ T cells are few, with the exception of nonclassical MHC class Ib molecules T10 and T22, mouse MHC class II I-Ek , herpes simplex virus glycoprotein gI, and CD1 (75). However, the mechanism of engagement of the γδ TCR with these ligands was not understood until recently. The crystal structure of the G115 Vγ9Vδ2 TCR has addressed some of these issues (76). As expected, the overall architecture

of the γδ TCR closely resembles that of αβ TCRs and antibodies (Figure 4). The most striking observation is an acute Vγ/Cγ interdomain angle of 42◦ , which deﬁnes an unusually small elbow angle of 110◦ . Whether this is indeed a general feature of all γδ TCRs or represents an extreme example must await further determination of γδ TCR structures. The corresponding elbow angles of αβ TCRs have so far been restricted to a slightly narrower range (140◦ –210◦ ) than those seen for antibodies (125◦ –225◦ ), presumably due to the smaller database of αβ TCR structures. The requirement of αβ and γδ TCRs to interact with the common CD3 components might restrict ﬂexibility for the V-C domain, but no structural data are available for any TCR/CD3 complexes to elucidate this requirement. The γδ TCR structure also raises further questions about CD3 recognition in the TCR complex. Comparison of the C domain surfaces of both γδ and αβ TCRs revealed no apparent similarities (76) that could explain the dual binding speciﬁcity of CD3 for these different classes of TCRs; only a few solvent-exposed residues are structurally conserved. The striking distinctions between the exposed surfaces of γδ and αβ TCRs are corroborated by the large differences of the respective proposed CD3ε-binding FG loops of Cβ and Cγ, and the very different secondary structural features of Cα and Cδ. Cα shows a secondary structure unlike the normal Ig-fold in the outer β-sheet, as opposed to Cδ, which has the regular, canonical three-stranded β-sheet. Thus, the possibility of two very distinct TCR/CD3 signaling complexes exists, the biological signiﬁcance of which is unclear. Alternatively, the main driving force for TCR/CD3 complex formation may not come from speciﬁc interaction of the extracellular domains, but may stem, at least in the primary stages of complex formation, from ionic interactions with the TCR stalk regions or through their transmembrane segments.

www.annualreviews.org • MHC/ TCR Interactions

425

ARI

15 February 2006

0:51

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

Figure 4 Overall comparison of the anatomy of complexes formed between MHC(-like) proteins and αβ TCR (a), Fab (b), and γδ TCR (c). The bottom panel is rotated 90◦ around the horizontal axis. Only one representative structure is shown for each type. The Cα trace of the TCR or Fab is on top in light gray with colored CDR loops and the MHC in dark gray below. The peptides in the αβ and Fab complexes are drawn as red ball-and-stick representations, while the CDR loops are colored as follows: CDR1α(24−31) : dark blue, CDR2α(48−55) : magenta, CDR3α(93−104) : green, CDR1β(26−31) : cyan, CDR2β(48−55) : pink, CDR3β(95−107) : yellow, and HV4(69−74) : orange. This color scheme is continued through Figures 5 and 7. 426

Rudolph

·

Stanfield

·

Wilson

ANRV270-IY24-14

ARI

15 February 2006

0:51

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Structures of αβ TCR and Peptide-MHC Complexes Clonotypic αβ TCRs recognize peptides presented by either class I or class II MHCs. Class II MHCs present peptides that originate from proteolysis of extracellular antigens in endosomal-type compartments, whereas class I MHCs present peptides primarily derived from intracellular degradation of proteins in the cytosol. TCRs that recognize these MHCs are found on two distinct cytotoxic and T-helper cell lineages, depending on the class of the MHC to which they are restricted. A current debate in class I MHC antigen presentation is over “cross-priming” of T cells for activation of CD8 T cells by transfer of peptide antigen or other substrates from a donor cell containing viral or tumor antigens to an acceptor cell (77–79). Peptide transfer is achieved by either (a) uptake of cell-derived proteins by receptor-mediated (LOX-1, CD91, and Toll-like receptors) endocytosis or (b) fusion of phagocytotic vesicles that contain material from apoptotic or necrotic cells with endoplasmic reticulum membranes. The proteasome is assumed to further degrade the proteins to peptides, which are then bound to chaperones, such as glycoprotein 69, HSP90, HSP70, and calreticulin before they are transferred to their new MHC hosts by an as yet unknown mechanism (78). Class I and class II MHCs both present peptides in an extended conformation in a vice-like groove, with two ﬂanking α-helices and a ﬂoor composed of antiparallel β-strands (Figure 2). Although the ends of the peptidebinding groove are occluded in class I MHC molecules, they are open in class II molecules; therefore, class II MHCs can accommodate peptides signiﬁcantly longer than can those of class I MHCs. The ﬁrst two turns of the class I MHC α1 -helix are replaced in class II MHCs by a β-strand. Whereas both classes of MHCs are composed of two noncovalently linked, polypeptide chains, in class I MHCs, the peptide-binding site is formed by the heavy chain only and, in class II MHCs, the

α- and β-chains assemble into a similar fold that constitutes the peptide-binding groove. Given the biological and structural divergence between these two MHC classes, it is of interest to compare and contrast the interactions with their cognate TCRs (Figure 5). Tables 2–4 provide a detailed analysis of the TCR/pMHC interface that includes buried surface areas, relative contributions of CDR loops, shape complementarities, hydrogen bonds, salt bridges, and van der Waals’ contacts. These tables have been updated from our previous analyses (36, 37) to include all complexes determined from 1996 to 2005. In the following section, we focus mainly on structures determined since 2002 and on new insights gained from this substantially increased database of TCR/pMHC complexes.

TCR/pMHC BINDING GEOMETRY Several techniques have been used in various laboratories to deﬁne the relative TCR binding orientation on pMHC. These diverse methods of calculation often result in dissimilar values for this “crossing angle,” making comparison of structures from multiple laboratories difﬁcult and confusing. Hence, we outline simple, reproducible, and easy-touse methods to describe the TCR-to-pMHC binding orientation and to calculate buried surface area. We do not suggest that other proposed methods are incorrect, but only that TCR complexes should be compared using a standard method. It is the relative rather than the absolute values in these calculations that are important here for deﬁning general principles of TCR/pMHC recognition. The MHCTCR crossing angle has been described previously in our laboratory by the angle between the MHC binding platform helices or MHC-bound peptide and an axis drawn approximately through the center of the α- and β-chain TCR CDR loops. Unsatisﬁed with the general applicability of this method, we www.annualreviews.org • MHC/ TCR Interactions

427

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

0:51

Figure 5 Footprints of TCR/pMHC complexes. The top surface of the MHC is colored in gray where it is not contacted by the TCR. Surface areas buried by TCR are colored by their contributions from each CDR loop, as in Figure 3. The red line represents a vector between the conserved disulﬁdes in the α/β (or γδ or light/heavy) chains, which has been translated to the center of gravity of the CDR loops, and indicates the relative orientation of the TCR onto the MHC. At this level of analysis, substantial variation is seen in the ﬁne speciﬁcity of the TCR on the pMHC. Class I and II complexes are labeled in black and green, respectively. A corresponding view of the γδ TCR/T22 complex (red label ) and the Fab/HLA-A1 (blue label ) complex is shown on the bottom row, with the δ, γ, light, and heavy chain CDRs colored correspondingly.

have experimented with several other ways to calculate this angle and now conclude that the method described here is optimal. We have now recalculated the crossing angles for all TCR/pMHC complexes, as listed in Tables 2 and 3, and the vectors representing the TCR are shown in Figure 5. We encourage other labs to adopt this method so that crossing angles for different complexes will be more easily comparable. 428

Rudolph

·

Stanfield

·

Wilson

In our current algorithm, the vector along the MHC helix axis is calculated as the bestﬁt straight line through the Cα atoms from the two MHC helices. For class I MHC, we use Cα atoms A50–A86 and A140–A176, for class II MHC A46–78, B54–64, and B67–91, and for the nonclassical MHC T22 (which has only one ordered helix) Cα atoms A61– A82. The vector describing the long axis of the TCR binding site is calculated by

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

Figure 5 (Continued)

constructing a vector between the centroids of the conserved Ig disulﬁde-forming sulfur atoms in the light and heavy chains (Sγ atoms from L22–L90 and H23–H92 for αβ TCRs and L22–L88 and H21–H94 for δγTCRs). The angle between the MHC and TCR vectors is then the dot product of the two vectors. For graphic visualization, the vector between the disulﬁdes is translated to the center of gravity of the TCR CDR loops. Buried surface areas (see Tables 2 and 3) are calculated using the molecular surface from the program MS (80) with a 1.7 A˚ probe radius (a value historically used for most antibody-antigen analyses). The use of a solvent accessible, rather than molecular surface,

for estimating the buried surface will give erroneous results, especially for more concave/ convex binding regions. Generally, the TCR heterodimer is oriented approximately diagonally relative to the long axis of the MHC peptide-binding groove (7, 11). The Vα domain is poised above the Nterminal half of the peptide, whereas the Vβ domain is located over the C-terminal portion of the peptide (Figures 4, 5). Peptide contacts are made primarily through the CDR3 loops, which exhibit the greatest degree of genetic variability. The preponderance of generally conserved contacts with the MHC α-helices are mediated through CDRs 1 and 2 (32), particularly for Vα, with the CDR3 loops www.annualreviews.org • MHC/ TCR Interactions

429

ANRV270-IY24-14

ARI

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Table 2

15 February 2006

0:51

Analysis of TCR/pMHC class I complexes

TCR

2C

2C

2C

scBM3.3

scBM3.3

scKB5-C20

B7

MHC

H-2Kb

H-2Kb

H-2Kbm3

H-2Kb

H-2Kb

H-2Kb

HLA-A2

Peptide

dEV8

SIYR

dEV8

pBM1

VSV8

pKB1

Tax

Resolution

3.0/32.2

2.8/32.7

2.4/31.3

2.5/27.6

2.7/29.8

2.7/27.8

2.5/31.2

PDB ID

2ckb

1g6r

1mwa

1fo0

1nam

1kj2

1bd2

Reference

(12)

(17)

(19)

(16)

(30)

(31)

(13)

TCR/pMHC crossing angle ( ◦ )

22

23

23

41

40

31

48

Buried surface areaa /Kd (μM) ˚ 2) TCR/pMHC (A

1842/83

1795/54

1878/56.5

1239/2.6

1444/114

1678/>100

1697/-

907/935

847/948

900/978

597/642

675/769

825/853

813/884

MHC/peptide (%) ˚ 2 ) / (%) Vα (A

76/24

76/24

75/25

79/21

81/19

79/21

69/31

490/54

438/52

469/52

221/37

348/52

371/45

552/68

CDR1/CDR2/ CDR3 (%) ˚ 2 ) / (%) Vβ (A

23/13/16/1

18/16/16/1

20/14/17/1

14/17/6

17/13/21

9/9/25/1

27/13/22/2

417/46

409/48

431/48

376/63

327/48

454/55

261/32

CDR1/CDR2/ CDR3/HV4 (%)

16/19/10/1

15/19/12/3

19/18/11/1

10/14/39/1

6/14/27/2

18/12/24/0

1/10/21/0

Sc b

0.41

0.49

0.62

0.61

0.60

0.55

0.64

HB/salt links/vdW contactsc

4/1/80

5/0/69

6/0/107

8/0/83

3/0/78

5/3/84

4/3/99

Vα

4/1/63

4/0/36

4/0/46

1/0/27

3/0/49

3/2/47

3/3/66

CDR1(24−31)

2/0/21

2/0/17

1/0/13

1/0/10

1/0/11

0/2/3

1/0/24

CDR2(48−55)

0/0/17

0/0/1

1/0/12

0/0/15

0/0/10

0/0/5

1/0/17

CDR3(93−104)

2/0/24

2/0/18

2/0/21

0/0/2

2/0/28

3/0/39

1/3/25

HV4(68−74)

0/1/1

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

Vβ

0/0/17

1/0/33

2/0/61

7/0/56

0/0/29

2/1/37

1/0/33

CDR1(26−31)

0/0/7

1/0/15

2/0/35

0/0/1

0/0/3

0/0/11

0/0/0

CDR2(48−55)

0/0/6

0/0/3

0/0/12

1/0/8

0/0/7

1/1/2

0/0/3

CDR3(95−107)

0/0/4

0/0/15

0/0/14

6/0/47

0/0/19

1/0/23

1/0/30

HV4(69−74)

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

MHC

2/1/59

2/0/37

3/0/69

3/0/54

3/0/60

3/2/73

1/3/42

Peptide

2/0/21

3/0/32

3/0/38

5/0/29

0/0/18

2/1/11

3/0/57

Calculated with MS (80) using 1.7A˚ probe radius. Calculated with Sc (154) using a 1.7A˚ probe radius. c Number of hydrogen bonds (HB), salt links and van der Waals (vdW) interactions calculated with HBPLUS (155) and CONTACSYM (156). Only the ﬁrst molecule in the asymmetric unit in all complexes was analyzed. a

b

430

Rudolph

·

Stanfield

·

Wilson

ANRV270-IY24-14

ARI

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Table 2

15 February 2006

0:51

(Continued )

A6

A6

A6

A6

JM22

LC13

1G4

1G4

AHIII 12.2

SB27

HLAA2

HLA-A2

HLA-A2

HLA-A2

HLA-A2

HLA-B8

HLA-A2

HLA-A2

HLA-A2.1

HLAB3508

Tax

TaxP6A

TaxV7R

TaxY8A

MP58– 66

EBV

ESO 9C

ESO 9V

p1049

EBV

2.6/32.0

2.8/27.3

2.8/29.0

2.8/28.6

1.4/23.1

2.5/28.8

1.9/26.0

1.7/25.3

2.0/25.3

2.5/27.9

1ao7

1qrn

1qse

1qsf

1oga

1mi5

2bnr

2bnq

1lp9

2ak4

(13)

(14)

(14)

(14)

(21)

(24)

(25)

(25)

(20)

(23)

34

32

36

34

62

42

69

69

67

70

1816/0.9

1768/-

1752/7.2

1666/-

1471/5.6

2020/10

1916/13.3

1924/5.7

1838/11.3

1752/9.9

908/908

851/917

838/914

810/856

738/733

1008/1012 979/936

979/945

943/895

827/925

66/34

67/33

66/34

73/27

72/28

80/20

65/35

64/36

73/27

60/40

587/65

561/66

536/64

598/74

241/33

573/57

465/47

470/48

550/58

474/57

24/10/ 24/5

23/13/ 25/6

23/10/ 26/5

29/12/ 27/5

11/6/13/3

17/18/ 22/0

8/12/27/0

8/13/27/0

16/13/27/1

10/17/ 31/0

321/35

290/34

302/36

211/26

496/67

435/43

515/53

509/52

393/42

354/43

2/1/33/0

2/1/31/0

2/0/34/0

0/0/26/0

5/34/27/1

3/17/22/0

9/20/19/5

9/19/20/4

6/16/20/0

19/5/18/1

0.64

0.61

0.66

0.61

0.63

0.61

0.72

0.75

0.70

0.73

11/4/105

10/1/120

7/2/136

6/1/102

8/0/92

8/1/122

10/0/178

10/0/184

5/1/147

11/0/106

7/3/60

7/1/86

4/2/81

6/1/78

0/0/29

3/0/62

5/0/100

5/0/109

4/0/99

5/0/58

3/0/21

3/0/20

2/0/19

2/0/21

0/0/9

2/0/18

0/0/41

0/0/40

0/0/27

1/0/7

0/0/3

0/0/4

0/0/8

1/0/8

0/0/10

0/0/12

1/0/7

2/0/18

0/0/23

1/0/6

3/2/33

3/1/53

2/2/49

2/1/43

0/0/10

1/0/32

4/0/52

3/0/51

4/0/49

3/0/45

1/1/2

1/0/9

0/0/5

1/0/6

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

4/1/45

3/0/34

3/0/55

0/0/24

8/0/63

5/1/60

5/0/78

5/0/75

1/0/48

6/0/48

1/1/2

1/0/3

1/0/3

0/0/0

1/0/4

0/0/1

1/0/16

1/0/14

1/0/2

5/0/32

0/0/0

0/0/0

0/0/0

0/0/0

2/0/23

1/1/15

1/0/23

1/0/21

0/0/11

0/0/1

3/0/43

2/0/31

2/0/52

0/0/24

5/0/33

4/0/44

2/0/33

2/0/37

0/0/31

0/0/12

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

1/0/3

1/0/2

0/0/0

0/0/0

4/4/65

3/1/67

1/2/84

3/1/75

4/0/67

6/1/96

5/0/73

6/0/73

2/1/94

2/0/40

7/0/40

7/0/53

6/0/52

3/0/27

4/0/25

2/0/26

5/0/105

4/0/111

3/0/53

9/0/66

www.annualreviews.org • MHC/ TCR Interactions

431

ARI

15 February 2006

0:51

contributing fewer conserved MHC contacts. The ﬁrst crystal structures of TCRs with class I molecules led to proposals that the TCR orientation is approximately diagonal with a mean around 35◦ (36). By contrast, in the ﬁrst class II complexes, the orientation was described as being closer to perpendicular (15, 18), suggesting a different binding mode between the MHC classes (81). However, we calculated this angle to be 50◦ , which is still roughly diagonal. Furthermore, the recent crystal structure determination of class I HLA-A2 in complex with the xenoreactive AHIII 12.2 TCR (20) showed a TCR/pMHC binding orientation of 67◦ , arguing against any real differences in receptor orientation between class I and class II complexes. In very low-afﬁnity interactions (<10 μM), such as in the TCR/pMHC system, there is always a danger of stabilizing nonproductive conformations during the crystallization process that are not representative of those most populated in solution, or those that represent only one snapshot of the possible complex orientations. Nevertheless, the key questions are still how many ways can the TCR dock on the pMHC and what controls the docking, the TCR or the pMHC. To partially address these questions, an antibody Fab/pMHC crystal structure was determined whereby the Fab Hyb3 serves as a TCR surrogate for binding to its antigen HLA-A1 complexed with a melanomaassociated human leukocyte peptide (82). In this complex, the Fab adopts a diagonal orientation of 41◦ , close to the range found in class I TCR/pMHC complexes (21◦ –70◦ ), but the binding is shifted toward the C-terminal half of the peptide-binding site. The Fab binds with its heavy chain over the central part and the light chain over the C-terminal part of the peptide, respectively, which suggests that, although the structurally equivalent antibody can dock in a similar binding orientation on the pMHC, antibodies can still display a much more promiscuous binding mode toward their antigens, as they are not required to signal nor read out the peptide content, but only to bind

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

432

Rudolph

·

Stanfield

·

Wilson

with high afﬁnity. Why then do TCRs generally dock on the pMHC in a generally diagonal orientation over the center of the binding groove? The contacts of the individual TCR CDR loops with the pMHC are quite diverse and still do not allow deﬁnitive conclusions as to their contributions in determining the docking angle. The current database of TCR/pMHC crystal structures supports a scenario in which the TCR approaches the pMHC in a roughly diagonal manner, driven by either long-range electrostatic steering or through a low-afﬁnity binding event, and then uses the conformational plasticity inherent in the CDR loops to maximally mold to and contact the pMHC, which then determines the ﬁnal docking outcome (Figure 6). However, the rotational freedom of the TCR has been somewhat limited in that no 180◦ ﬂip of the TCR has been observed to date that would poise the Vα (rather than Vβ) domain over the C-terminal half of the peptide or vice versa, although such a scenario has been predicted (83). Because the docking angle dictated by the TCR/pMHC interface inﬂuences the disposition of the TCR constant domains relative to other components of the TCR signaling complex, such as CD4 or CD8, it is likely also to inﬂuence T cell signaling. However, in the absence of a ternary TCR/pMHC/CD3 or TCR/pMHC/CD8 complex structure, no conclusions can yet be drawn as to the precise role or restrictions that the TCR/pMHC docking geometry plays in coreceptor binding and downstream signaling. What is known is that ligand engagement of the TCR/CD3 complex induces a conformational change in CD3ε, which exposes a C-terminal prolinerich sequence in its cytoplasmic tail that recruits the adapter molecule Nck for downstream signaling via binding of its SH3 domain (84). In addition, comparison of the crystal structures of free and pMHC-bound LC13 TCR has revealed a conformational change in the AB-loop of the TCRα constant domain, which is predicted to be close to the CD3ε binding site (24). Furthermore, a

ANRV270-IY24-14

ARI

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Table 3

15 February 2006

0:51

Analysis of TCR/pMHC class II and nonclassical complexes

TCR

scD10

HA1.7

HA1.7

Ob.1A12

172.10

3A6

G8

MHC

I-Ak

HLA-DR1

HLA-DR4

HLADR2b

I-Au

HLADR2a

T22

Peptide

CA

HA

HA

MBP

MBP

MBP

—

Resolution/Rfree

3.2/29.3

2.6/25.5

2.4/24.6

3.5/31.8

2.42/27.4

2.8/32.9

3.40/33.0

PDB ID

1d9k

1fyt

1j8h

1ymm

1u3h

1zgl

1ypz

Reference

(15)

(18)

(26)

(27)

(28)

(29)

(102)

TCR/pMHC crossing angle (◦ )

53

47

49

84

43

40

87

Buried surface area/ Kd (μM)

1734/1-2

1945/-

1934/-

1916/-

1697/8.8

1984/≤500

1750/-

TCR/pMHC (A2 )

868/866

975/970

975/959

968/948

866/831

953/1032

845/905

MHC/peptide (%) ˚ 2 )/(%) Vα (A

77/23

67/33

68/32

60/40

76/24

66/34

100/0

530/61

456/47

471/48

473/49

459/53

394/41

755/89

CDR1/CDR2/ CDR3 (%) ˚ 2 )/(%) Vβ (A

22/15/22/2

15/8/23/0

15/9/23/0

6/19/24/0

14/14/25/1

17/4/19/0

15/5/61/4

338/39

519/53

504/52

495/51

408/47

558/59

90/11

CDR1/CDR2/ CDR3/HV4 (%)

3/20/16/0

8/22/22/1

8/19/23/1

5/11/32/2

5/19/23/0

6/24/28/0

0/0/11/0

Sc

0.71

0.56

0.56

0.52

0.62

0.63

0.66

HB/salt links/vdW contacts

6/3/119

4/4/104

2/5/101

4/0/116

5/0/110

8/1/127

3/0/116

Vα/Vδ

1/2/64

0/1/41

1/1/45

2/0/65

1/0/37

6/0/60

3/0/113

CDR1(24−31)

1/1/24

0/0/13

0/0/16

0/0/9

0/0/6

1/0/39

0/0/30

CDR2(48−55)

0/0/16

0/0/2

1/0/4

0/0/18

0/0/7

0/0/0

0/0/2

CDR3(93−104)

0/0/24

0/1/26

0/1/25

2/0/38

1/0/24

5/0/21

3/0/68

HV4 (68−74)

0/1/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/0

0/0/4

Vβ/Vγ

5/1/55

3/3/63

1/4/56

2/0/51

4/0/73

2/1/67

0/0/3

CDR1(26−31)

0/0/0

0/2/10

0/2/8

1/0/6

1/0/5

1/0/11

0/0/0

CDR2(48−55)

2/0/29

1/1/25

0/1/15

0/0/0

1/0/28

0/0/29

0/0/0

CDR3(95−107)

3/0/21

2/0/20

1/0/24

1/0/40

2/0/34

1/1/23

0/0/3

HV4(69−74)

0/0/0

0/0/0

0/0/1

0/0/0

0/0/0

0/0/0

0/0/0

MHC

5/3/86

2/1/79

1/2/77

2/0/69

3/0/89

4/1/86

3/0/116

Peptide

1/0/33

2/3/25

1/3/24

2/0/47

2/0/21

4/0/41

n.a.

n.a., not applicable.

similar AB-loop conformation has been found in the B7/HLA-A2/Tax complex structure (13), which, as for LC13, is free of crystal contacts in this region. A similar conformational change has not been observed for the 2C system, but here crystal contacts may have reduced the conformational freedom of the AB-loop. However, these data are still not compelling, and the necessity and extent of any conformational changes in the TCR/CD3

complex required for signaling must await a TCR/CD3 complex crystal structure or analysis by other biophysical methods, such as FRET. One pertinent analysis of these TCR/ pMHC docking orientations has resulted in grouping of the available TCR/pMHC complexes according to the positioning of their Vα domains with respect to the MHC-bound peptide (20). Four TCRs with their Vα www.annualreviews.org • MHC/ TCR Interactions

433

434

Rudolph

·

Stanfield

·

Wilson 12 13 13 14 14 16

172.10/I-Au /MBP (1u3h)

HA1.7/HLA-DR4/HA (1j8h)

HA1.7/HLA-DR1/HA (1fyt)

Ob.1A12/HLA-DR2b/MBP (1ymm)

3A6-HLA-DR2a-MBP (1zgl)

scD10/I-Ak /CA (1d9k)

2

0

6

0

0

0

2

0

0

1

3

9

6

4

2

0

0

0

0

0

0

13

SB27-HLA-B3508-EBV (2ak4)

0

9

1G4-HLA-A2-ESO9V (2bnq)

P1

9

1G4-HLA-A2-ESO9C (2bnr)

P-1

9

A6-HLA-A2-TaxY8A (1qsf)

P-2

9

A6-HLA-A2-TaxV7R (1qse)

P-3

9

A6-HLA-A2-TaxP6A (1qrn)

P-4

9

LC13/HLA-B8/EBV (1mi5)

Human/mouse class II

9

JM22/HLA-A2/MP58–66 (1oga)

11

10

7

2

2

0

P2

0

0

0

10

4

7

0

P3

0

0 P2

0

0

0

0

0

0

0

0

P1

0

4

1

0

0

0

0

9

AHIII12.2/HLA-A2.1/p1049 (1lp9)

0

0

P2

P1 1

0

0

0

0

0

0

P2

0

4

8

BM3.3-H-2K-VSV8 (1nam)

0

9

8

2C-H-2Kbm3 -dEV8 (1mwa)

0

9

8

2C-H-2Kb -SIYR (1g6r)

0

A6/HLA-A2/Tax (1ao7)

8

BM3.3/H-2Kb /pBM1 (1fo0)

0

1

P1

0

1

0

0

0

0

P4

0

P3

0

0

0

0

0

0

0

4

0

0

P3

0

0

0

0

0

0

P3

8

P4

9

2

9

7

3

8

P5

32

0

0

0

0

0

0

P6

0

2 P10

60

2

0

0

0

3

3

12

0

1

P6

0

0

0

0

0

0

P5

P52

56

241 24

15

19

26

0

5

13

29

16

P5

0

0

0

0

0

0

0

0

P4

22

16

15

1

19

7

P4

4

9

0

0

0

5

P7

0

P11

9

8

1

12

5

17

2

2

4

5

P7

7

24

22

24

4

1

P6

7

0

0

9

9

1

P8

0

P12

10

10

0

14

12

0

6

6

10

11

P8

0

1

0

6

2

5

P7

Peptide residue (P) and # contacts per residue

B7/HLA-A2/Tax (1bd2)

8

2C/H-2Kb /dEV8 (2ckb)

8

# peptide residues

0

0

0

0

0

0

P9

0

P13

0

0

0

0

0

0

0

0

0

0

P9

0

0

0

0

0

0

P8

0

1

0

0

0

0

P10

0

0

0

0

P11

0

P12

P13

32

32

39

28

25

16

40

105

100

20

46

43

20

16

37

47

34

29

41

37

31

25

14

# Total contacts

15 February 2006

KB5-C20/H-2Kb /pKB1 (1kj2)

Human/mouse class I

Peptide side-chain contacts with TCR

TCR/MHC/peptide

Interactions of the peptide component of the pMHC with the TCR

ARI

Table 4

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14 0:51

9 9 9 9 9 9

LC13/HLA-B8/EBV (1mi5)

A6-HLA-A2-TaxP6A (1qrn)

A6-HLA-A2-TaxV7R (1qse)

A6-HLA-A2-TaxY8A (1qsf)

1G4-HLA-A2-ESO9C (2bnr)

1G4-HLA-A2-ESO9V (2bnq)

13 13 14 14 16

HA1.7/HLA-DR4/HA (1j8h)

HA1.7/HLA-DR1/HA (1fyt)

Ob.1A12/HLA-DR2b/MBP (1ymm)

3A6-HLA-DR2a-MBP (1zgl)

scD10/I-Ak /CA (1d9k)

3

P-4

0

8

0

P-3

0

2

3

0

0

0

P-2

0

3

1

0

0

0

P-1

0

0

0

0

0

0

P1

0

0

0

0

0

8

P2

0

0

3

0

0

2

P3

0

P2

P1 0

0

0

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

P4

0

P3

0

0

0

0

0

0

0

0

0

0

P3

0

0

0

0

0

1

P3

0

P4

5

5

9

10

11

0

10

15

5

10

P4

0

0

0

1

0

0

P5

1

0

0

3

1

0

P6

P10

P53 39

6

5

0

0

2

7

2

1

1

1

P6

0

0

0

1

0

0

P5

2

2

0

1

0

0

2

3

0

0

P5

0

0

0

0

0

0

P4

1

0

0

1

2

1

P7

P11

0

0

0

0

1

0

0

2

4

1

P7

0

0

0

1

0

0

P6

1

1

0

3

4

0

P8

P12

0

0

0

1

1

1

0

0

3

1

P8

0

2

0

5

0

0

P7

0

0

0

0

0

0

P9

P13

0

0

0

0

0

0

0

0

0

0

P9

0

0

0

0

0

0

P8

0

0

0

0

0

0

P10

0

0

0

0

P11

0

P12

P13

3

14

10

8

7

11

39

13

12

10

13

16

8

14

20

13

14

0

2

0

7

0

2

is a Met and P5 is a Trp in the 1G4 structures; P4 is a Gly in the other 9-mer structures, and P5 is a Tyr, Phe, or Arg in the other 9-mer structures. This explains the high number of contacts at this position for the 1G4 structures. 2 P5 is a P5, P6, P7, P8, P9 insert, with 3,0,16,0,13 contacts for these residues. 3 P5 is a P5, P6, P7, P8, P9 insert, with 8,4,8,16,3 contacts for these residues.

1 P4

12

172.10/I-Au /MBP (1u3h)

Human/mouse class II

13

9

JM22/HLA-A2/MP58–66 (1oga)

0

0

1

P2

P1 0

0

0

0

0

0

0

0

0

0

0

1

P2

15 February 2006

SB27-HLA-B3508-EBV (2ak4)

9

8

BM3.3-H-2Kb -VSV8 (1nam)

AHIII12.2/HLA-A2.1/p1049 (1lp9)

8

2C-H-2Kbm3 -dEV8 (1mwa)

9

8

2C-H-2Kb -SIYR (1g6r)

9

8

BM3.3/H-2Kb /pBM1 (1fo0)

A6/HLA-A2/Tax (1ao7)

8

2C/H-2Kb /dEV8 (2ckb)

0

P1

ARI

B7/HLA-A2/Tax (1bd2)

8

KB5-C20/H-2Kb /pKB1 (1kj2)

Human/mouse class I

Peptide main-chain contacts with TCR

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14 0:51

www.annualreviews.org • MHC/ TCR Interactions

435

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

0:51

Figure 6 Variation in the tilt and roll of the TCR on top of the MHC. The left and right views are related by a 90◦ rotation about a horizontal axis. The MHC peptide backbones and the MHC helices are shown as gray tubes. The orientation axes are colored individually for each TCR. For 15 individual TCRs, the pseudo-twofold axes that relate the Vα and Vβ domains of the TCRs to each other are shown, giving a good estimate of the inclination (roll, tilt) of the TCR on top of the MHC. The TCR twofold axes tend to cluster around P4-P6 at the center of the interface. Labels are placed at the top of each axis. The ﬁgure also indicates any shifts of the TCR along the peptide where the Ob.1A12 and LC13 TCRs mark the extremes, centered around P1 and P6, respectively. 3A6 and SB27 also are outliers at present where they are centered on one half of the peptide.

domains located closer to the N terminus of the peptide exhibited CD8-dependent signaling, whereas another four TCRs in which their Vα domains were closer to the C terminus of the peptide acted independently of CD8. A geometric model was put forward to explain this correlation of Vα domain positioning with the CD8-dependence: A diagonal orientation of the TCR with the Vα domain over the N terminus of the peptide would allow efﬁcient recruitment of CD8, whereas the TCR/pMHC docking mode with the Vα domain closer to the C terminus of the peptide would require high TCR/pMHC afﬁnity to initiate CD8independent signaling. Thus, it was speculated that CD8-independent TCRs would 436

Rudolph

·

Stanfield

·

Wilson

generally exhibit a higher afﬁnity for pMHCs, which in turn raised the question why these TCRs survived negative selection in the thymus that would be biased against high-afﬁnity self-recognition. To reconcile this apparent discrepancy, a very different docking orientation was proposed during TCR selection compared to T cell/APC engagement, in contrast to other views that dispute any such global rearrangement of TCRs once they have engaged pMHC (16, 85, 86). Furthermore, in the H2Kb system, the BM3.3 TCR requires CD8 for signaling when engaging H-2Kb /VSV8, but can signal independently of CD8 when bound to a different peptide (pBM1) in the context of the same MHC (87), yet crystal structures of the two complexes do not show

ANRV270-IY24-14

ARI

15 February 2006

0:51

any signiﬁcant differences in their docking geometries (16, 30).

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TCR-INDUCED FIT Insight into the structural changes that accompany TCR/antigen engagement (i.e., induced ﬁt) must include crystal structures of the same TCR in its free and bound forms or of the same TCR bound to different pMHCs. Until recently, only two well-studied systems, the 2C and A6 TCRs, fulﬁlled these requirements. The 2C system allowed comparison of the free 2C TCR (7) with an agonist (12) and a superagonist peptide (17) in complex with the same H-2Kb MHC (Figure 7a). This comparison disclosed a functional hotspot between the CDR3 loops in the 2C TCR that ﬁnely discriminated between side chains and conformations of centrally located peptide residues through increased complementarity and additional hydrogen bonds. In the A6 system (13, 14), altered peptide ligands (APLs), i.e., peptides of slightly different sequence than the natural ligand, induced only subtle conformational changes in the TCR (Figure 7b). In both the 2C and A6 systems, conformational changes are restricted mainly to the CDR3 loop regions, and the largest conformational differences were observed when comparing free versus bound TCR (36). Recently, two crystal structures of the BM3.3 TCR bound to different peptides (pBM1 and VSV8) in complex with the class I MHC H-2Kb (30, 31) provided another system for study of conformational changes (Figure 7c). The BM3.3 TCR not only recognizes the naturally processed, allogeneic pBM1 peptide presented by H-2Kb , but also cross-reacts with an H-2Kb -bound peptide from vesicular stomatitis virus (VSV8). The BM3.3 TCR rotates 5◦ and shifts by 1.2 A˚ when contacting the VSV8 peptide compared to the pBM1 peptide, which is comparatively small given the complete absence of any sequence homology between the two peptides. The α1 α2 -helices move slightly, in

synchronization with peptide conformational changes, but similar changes have already been seen within unliganded pMHC complexes (88) and, hence, are not necessarily attributable to TCR binding. Large differences between the two complexes are, however, seen in the CDR3 conformations, which allow the BM3.3 TCR to adapt to different peptides bound to the same H-2Kb MHC. In the allogeneic BM3.3/H-2Kb /pBM1 pMHC complex, the CDR3α loop ﬂares away from the peptide, leaving a large, water-ﬁlled cavity between the pMHC and the TCR. In the BM3.3/H-2Kb /VSV8 complex, the CDR3α loop adopts a very different conformation with a maximum displacement of >5 A˚ for the Tyr97 Cα atom (Figure 7c). This movement results in burial of a larger surface area (∼16%) for this complex due to closer proximity of CDR3α to the pMHC interface (Table 2). This altered CDR3α conformation can explain the cross-reactive properties of this TCR, but also raises the question of how reasonable speciﬁcity is maintained given the large loop movements in CDR3α. The afﬁnity of the self BM3.3/H-2Kb /pBM1 complex is 44-fold higher than for the VSV8 complex (Kd at 298 K of 2.6 μM and 114 μM, respectively) despite the buried surface area being smaller. The TCR conformational changes led to increased Vβ interaction (56 versus 29 contacts), but decreased Vα contacts (27 versus 49) in the allogeneic complex compared with the self-syngeneic complex, although slightly more overall TCR-peptide contacts (83 versus 78) are made in the self complex (Table 2). Although this afﬁnity difference amounts only to 9.4 kJ/mol at 298 K, which is equivalent to a single hydrogen bond, it is signiﬁcant, and the nonconformity with the size of the buried surface area is somewhat unexpected. The antidromic behavior of afﬁnity and buried surface may result from unfavorable entropic contributions due to conformational changes of CDR3α during BM3.3 binding to H-2Kb /pBM1. The KB5-C20 TCR has also been determined in its free form (8) and bound to the www.annualreviews.org • MHC/ TCR Interactions

437

ANRV270-IY24-14

ARI

15 February 2006

0:51

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

pKB1 peptide presented by H-2Kb(31) . In the free form, the unusually long CDR3β loop of 13 residues is packed tightly against the CDR3α loop, leaving no pocket for binding of pMHC. Thus, this unliganded KB5-

438

Rudolph

·

Stanfield

·

Wilson

C20 structure indicated that a large conformational change of at least CDR3β must accompany pMHC engagement, which was indeed found in the KB5-C20/H-2Kb /pKB1 complex (31). Although the other CDRs displayed

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

only minor hinge movements upon pMHC complex formation, the apex of CDR3β un˚ concomitant with derwent a large shift of 15 A, a complete reorganization of its loop structure (Figure 7d ). In these four examples described above, the majority of CDR conformational changes were limited to either CDR3α or CDR3β, but in a recent complex between the “public” LC13 TCR and its immunodominant Epstein Barr virus (EBV) peptide antigen in complex with HLA-B8, both CDR3 loops, as well as CDR1α and CDR2α, underwent large conformational changes when free and bound TCRs were compared (10, 24) (Figure 7e). Apparently, the dominant LC13 TCR response represents the optimal immunological answer to persistent EBV infection, as it is selected by unrelated individuals, and thus termed public. One hallmark of this TCR is a ∼7–10 A˚ translational shift (calculated after overlapping the HLA-A2 MHC molecules in the B7, A6, JM22, and 1G4 complexes) toward the EBV peptide C terminus (Figure 5), contacting peptide residues P6 and P7 rather than the more common peptide contact residue P5 (Table 4). However, the marked lateral shift of the LC13 TCR

is peculiar to this complex and not a general characteristic of HLA-B MHC complexes as the SB27/HLAB3508/EBV complex (23) does not show this feature (Figure 5). Of the three C-terminal peptide residues, the TyrP7 side chain (17 contacts plus two watermediated hydrogen bonds to LC13 residues His33α and His48α) dominates the TCR contact area. Mutation of this tyrosine to phenylalanine reduces CTL recognition by a factor of 10 (which would translate into a Kd value of ∼100 μM) (24), similar to a functional hotspot described for the 2C system, where mutation of a lysine to an arginine residue in the dEV8 peptide converts an agonist into a superagonist with ∼1000-fold increase in cytotoxicity (17). Comparison of free and bound LC13 TCRs reveals a 2.5 A˚ rigid body shift and rotation by 38◦ of CDR3β, which displaces individual loop residues (Gln98β, Ala99β, Tyr100β) by up to 5 A˚ and maximizes the peptide readout by increasing the shape complementarity (Sc ). More drastic changes of up to 8 A˚ are found for CDR3α, which switches from a mobile, extended structure in the unliganded LC13 to a crumpled structure (89) that makes extensive contacts with the HLA-B8 α1 -helix. Pro93 appears to

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 7 Conformational variation and induced ﬁt in the TCR CDR loops. The TCR Vα- and Vβ-chains are shown in light gray looking down onto their antigen-binding site (MHC view). Their CDR loops are colored as in Figure 3. The central CDR3 loops are the most structurally diverse and recognize mainly peptide, whereas the CDR1 and CDR2 loops recognize the mostly conserved helical structural features on the MHC. (a) Overlay of the unliganded 2C TCR with three pMHC-liganded structures. The unliganded 2C TCR structure shows signiﬁcant conformational differences of its CDR3α (red) and CDR1α loops (dark blue). (b) Overlay of four liganded A6 TCR structures. The only A6 CDR loop showing conformational variability in response to the different Tax peptide mutants in the HLA-A2 complexes is CDR3β (red for the wild-type Tax complex, yellow for all others). (c) Comparison of the BM3.3 conformation when bound to H-2Kb carrying either the pBM1 or the VSV8 peptide. The CDR3α loop ﬂares away from the peptide in the pBM1 complex (red), interacting with the MHC α1 -helix, while it is closer to the peptide in the VSV8 complex, where it also buries a larger surface area of the pMHC compared to the pBM1 complex. (d ) Comparison of the unliganded KB5-C20 TCR and its structure in the H-2Kb /pKB1 complex. The large conformational change of the CDR3β loop ( yellow in the unliganded form) is highlighted in red for the H-2Kb /pKB1 complex. (e) Comparison of the unliganded class II LC13 TCR and its structure in the HLA-B8/EBV complex. The large conformational changes of the CDR3α and CDR3β loops are highlighted in red for the complex. ( f ) Overlay of all TCRs, free and bound, to show the degree of variation in the CDR loop regions. The Cα atoms of the variable domains were used to generate the alignment. (g) Comparison of the G8 (102) and the Vγ9/Vδ2 γδ (76) TCRs. Both molecules in the asymmetric unit of the G8 receptor are shown. www.annualreviews.org • MHC/ TCR Interactions

439

ARI

15 February 2006

0:51

mediate this radical change in CDR3α conformation, as it is present in six different HLA-B8-restricted CTL clones (90) and is encoded by an N-region addition, indicating that somatically derived TCR residues may be important for specifying cognate interactions. When complexed to HLA-B8/EBV, CDR1α and CDR2α deviate strongly from the canonical conformations (91) that they adopt in the unbound state. Both rigid body shifts and structural crumpling lead to maximum displacements of up to 7 A˚ in each loop. Such large changes were not apparent in the 2C and KB5-C20 systems, where nonrigid body conformational changes were conﬁned to the CDR3 loops (Figure 7).

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

FROM ALLOREACTIVITY TO XENOREACTIVITY An impressive 1% to 10% of mature T cells recognize and respond to nonself MHC (92), a phenomenon termed alloreactivity, which is the molecular reason for organ and skin graft rejection and, in immunocompromised individuals, graft-versus-host disease. As an exact tissue typing between donor and acceptor is not always possible, graft rejection poses a major obstacle for long-term stability of organ transplants in patients. Crystal structures of alloreactive TCR/pMHC complexes and comparison with their syngeneic counterparts have recently begun to shed light on the structural basis of alloreactivity. The alloreactive BM3.3/H-2Kb /pBM1 (16) and 2C/H-2Kbm3 /dEV8 (19) complexes showed an increased propensity for TCR Vβ interactions with the pMHC (Table 2). Although the syngeneic complex is still unavailable for BM3.3/H-2Kb /pBM1, the 2C/ H-2Kbm3 /dEV8 structure (19), which carries an alloreactive Asp77Ser mutation buried in the H-2Kbm3 molecule, can be compared with the syngeneic 2C/H-2Kb /dEV8 complex (12). This analysis revealed a shift of the TCR/pMHC contacts from predominantly Vα contributions in 2C/H-2Kb /dEV8 to a preponderance of Vβ interactions in 440

Rudolph

·

Stanfield

·

Wilson

2C/H-2Kbm3 /dEV8 (Table 2). In the 2C/H2Kb /dEV8 complex, the Vα domain of the 2C TCR mediates 69 interactions (van der Waals and polar) with pMHC versus only 17 contacts by the Vβ domain. Strikingly, the relative contribution of the variable domains in the 2C/H-2Kbm3 /dEV8 interface was reversed to 50 Vα versus 63 Vβ interactions. In the BM3.3/H-2Kb /pBM1 complex, the Vβ interactions also dominate the TCR/pMHC interface (28 Vα versus 63 Vβ interactions). However, this propensity for increased Vβ interactions in alloreactive complexes was contradicted by another TCR/pMHC complex, KB5-C20/H-2Kb /pKB1 (31). The alloreactive murine TCR KB5-C20 arises from an H-2k background and recognizes three different pKB peptides (pKB1–3) in complex with H-2Kb(87) . In the KB5-C20/H2Kb /pKB1 complex, the Vα domain contributes 52 contacts to the pMHC, whereas only 40 contacts are mediated by the Vβ domain (Table 2). However, we do not have the corresponding syngeneic complex for comparison. In addition, a recent structure of a xenoreactive TCR/pMHC complex (20) showed a preponderance of Vα interactions. In xenoreactive complexes, a TCR selected in one species now exercises cross-species reactivity. Murine AHIII 12.2 TCR cross-reacts with human class I MHC HLA-A2 bound to peptide p1049 and acts in a CD8-independent manner, as mouse CD8 does not bind to human MHC. The crystal structure of this complex (20) provided some insight not only into the CD8-independence of T cell signaling (see below), but also on the structural basis of xenoreactivity. The TCR, again, is poised diagonally (67◦ ) across the pMHC interface, with no obvious reorientation or other characteristics that would easily rationalize this biological distinctiveness. Thus, in this case, xenoreactivity is indistinguishable from self- and allorecognition at the overall TCR/pMHC structural level. Vα contributes twice as many interactions to the pMHC as Vβ (103 Vα and 49 Vβ interactions), again

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

suggesting that Vβ interactions do not necessarily dominate in alloreactive or xenobiotic complexes. Taken together, neither alloreactivity nor xenoreactivity seems distinguishable from syngeneic TCR recognition by analysis of the TCR/pMHC interfaces of their respective crystal structures. Apparently, alloreactivity is a natural consequence of the essential requirement for TCRs to be able to rapidly screen the repertoire of MHC/antigen complexes on the cell surface. A certain number of contacts must be made and/or landmarks recognized on the pMHC surface for the TCR to successfully interact and remain docked with its antigen. However, the TCR/pMHC associations selected in the thymus are of relatively low afﬁnity (1–100 μM), creating opportunities for cross-reactivity in the periphery. Thymic selection cannot select for substantially higher or lower afﬁnity or else too few (i.e., too highly restricted or too promiscuous, respectively) TCRs would emerge to combat the ever-changing antigenic repertoire of microorganisms.

TCR SELECTION, SELECTIVITY, CHAIN BIAS, AND CROSS-REACTIVITY Certain antigens select a very restricted TCR repertoire, such as the immunodominant antigen derived from EBV. This virus causes persistent infections in up to 90% of adults (93), and an antigen derived from it is presented by HLA-B8. The conformational changes associated with binding of the LC13 TCR to the HLA-B8/EBV pMHC antigen have been described above and are on a scale similar to other changes seen between free TCRs and their complexes with pMHC. Also, the afﬁnity of the LC13/HLA-B8/EBV complex (Kd ∼10 μM) is within the range of most other TCR/pMHC systems (24). Why then, in this particular case, is chain bias so extreme that most CTLs use the LC13 TCR for combating EBV infections? The structural explanation for this immunodominance is proposed

to be the induced ﬁt of the CDRs, which included changes in the canonical structures of germ line–encoded CDRs 1α and 2α, that enhance complementarity with the pMHC. The TCR/pMHC interaction was then proposed to induce further conformational changes in the TCR Cα domain to enhance its interaction with CDR3ε (24). This speciﬁcity advantage of LC13 was then proposed to increase avidity of the expanded T cell lineages, or lead to superior signaling or more efﬁcient formation of the immunological synapse. However, LC13 does not exhibit the highest complementarity seen so far as measured by the Sc index. LC13 has a Sc coefﬁcient of 0.61, whereas in other TCR/pMHC complexes this value varies from 0.41 to 0.75, with several around 0.70 (Table 1). However, the buried surface area for the LC13 complex is among the largest at 2020 A˚ 2 compared with an average of 1791 A˚ 2 . Furthermore, other TCRs undergo conformational changes, and only this TCR has the Cα-induced change. Another example of immunodominance in TCR chain bias is the Vβ17-Vα10.2 TCR ( JM22) in complex with HLA-A2 that presents an inﬂuenza matrix protein antigen (21). In general, the anti-inﬂuenza response in HLA-A2-positive adults relies predominantly on TCRs containing Vβ17 and is directed against the inﬂuenza matrix protein ˚ (M1) residues 56–66. The 1.4 A-resolution structure of the complex is the highestresolution TCR/pMHC crystal structure determined to date and notably deﬁnes four water molecules that are completely buried in the TCR/pMHC interface and strongly contribute to the overall shape complementarity (Sc = 0.73 in the presence and 0.63 in the absence of these water molecules), underscoring the essential role of solvent in immune recognition, as in antibody-antigen interfaces (94). The JM22/HLA-A2/MP58-66 structure indicates a more perpendicular orientation (62◦ ) of the TCR with respect to the MHC that differs strongly from TCRs B7 (48◦ ) or A6 (34◦ ), but not from 1G4 (69◦ ), that all www.annualreviews.org • MHC/ TCR Interactions

441

ARI

15 February 2006

0:51

bind to the same MHC (Table 2, Figure 5). This TCR footprint is more focused on the C-terminal end of the pMHC groove, with substantially more Vβ (71) than Vα (29) interactions. Thus, as in the examples of H-2Kb recognition by different TCRs (2C, BM3.3, KB5-C20), these HLA-A2 TCRs ﬁnd different solutions to binding the same MHC, albeit with different peptides, such as has been found for different antibodies that bind to the same protein antigen (95). Does the binding mode of JM22 to HLAA2/MP58-66 then reveal the molecular reason for its immunodominance? Normally, the centrally located P5 peptide side chain in HLA-A2 complexes wedges into a notch between the CDR3α and CDR3β loops. In the JM22/HLA-A2/MP58-66 complex, this situation is reversed in that a large side chain from the TCR, CDR3β Arg89, now binds into a notch between peptide residue Phe-P5 and the MHC α2 -helix and establishes ﬁve hydrogen bonds. As CDR3β residues Arg89 and Ser99 are conserved in the majority of TCRs active against the HLA-A2/MB58-66 epitope (96, 97), this region is likely responsible for the immunodominance. Markedly more contacts to the pMHC are mediated by the TCR β-chain than by the α-chain (71 versus 29 contacts; Table 2), and all speciﬁc contacts to the peptide are made by the βchain (21). In addition, CDR1β Asp32 and CDR2β Gln52 bind to the MP58-66 peptide, suggesting that these four residues apparently are sufﬁcient for Vβ17 chain bias. Selection of this particular TCR is probably reinforced by repeated inﬂuenza infections, as the Vβ17 chain becomes dominant during the ﬁrst years of life (98). The N-terminal domain of inﬂuenza matrix protein M1 (99), from which the antigen is derived, is composed of a dimer of two four-helix bundles, where the TCR epitope residues 56–66 form one of the central helices. Thus, this sequence may critically contribute to stability of M1 so that it is not easily mutable during inﬂuenza evolution, and may explain why this epitope would

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

442

Rudolph

·

Stanfield

·

Wilson

give such a conserved and durable cytotoxic response.

AUTOIMMUNE TCR/pMHC CLASS II COMPLEXES In a recent class II TCR/pMHC complex, a novel docking mode was identiﬁed where the Ob.1A12 TCR slid along the binding groove toward the N-terminal region of the bound peptide (Figure 5). The crystal structure of the human autoimmune TCR Ob.1A12 in complex with HLA-DR2 and a self-peptide from myelin basic protein (MBP), which has been linked to multiple sclerosis, represented the second example of an autoimmune TCR/pMHC complex (27). It was suggested that this docking mode pertains to autoimmune complexes in general. The translation of the Ob.1A12 TCR along the groove indeed represents another facet of MHC restriction (100) in which the TCR has moved its center of mass to focus only on half of the available peptide epitope and binds in an orthogonal orientation (84◦ ) (Table 3; Figure 5). As a consequence, only the N-terminal residues of the autoimmune MBP peptide (P-4, P-2, P-1, P2, P3, and P5) are contacted by the TCR, leaving the C-terminal half of the peptide unsurveyed, as far as the informational content is concerned (Table 4). Why are such autoimmune TCRs not deleted via positive and negative selection? A possible explanation would be that in the thymus, the canonical diagonal docking geometry is used during the selection process for low-afﬁnity interactions (MHC) to self pMHC. TCRs would then bind to pMHCs containing self-peptides in the periphery, but in a noncanonical, yet immunologically productive, manner, which would effectively undermine the selection process. In addition, coreceptor binding (CD4 or CD8) could aid in rescue of lowafﬁnity complexes, such as the Ob.1A12 complex (27). Although the Ob.1A12/HLADR2/MBP complex certainly expands the

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

range of TCR/pMHC orientations and translations, its docking geometry is not the answer to this question of autoimmune TCR/pMHC recognition as a related complex, 172.10/ I-Au /MBP (28), docks canonically in a diagonal mode (60◦ ) and in the center of the binding groove. Unlike Ob.1A12, which focuses on the N-terminal residues of MBP, the 172– 10 TCR binds only to MBP residues at the C-terminal end of the groove. As for allogeneic complexes, Vβ dominates the interaction. In this case, the preponderance of Vβ interactions is due not to TCR translation along the groove but to a two-residue register shift of the bound peptide within the I-Au groove (101). This shift places the ﬁrst peptide residue in the P3 pocket of the MHC, and the peptide N terminus is now buried under the CDR3α loop, out of reach of CDRs 1α and 2α, and, as a consequence, the majority of the MBP peptide is accessible only to Vβ (Table 1).

γδ TCR/NONCLASSICAL MHC COMPLEX Although a database of αβ TCRs and αβ TCR/pMHC complexes has now been accumulated that is large enough to permit some general conclusions on TCR binding modes, the γδ lineage of TCRs was severely underrepresented until the ﬁrst crystal structure of a human γδ TCR G155 (76). A recent crystal structure of the γδ TCR G8 bound to its ligand, the nonclassical MHC T22 (102), provided some indication of a ligand recognition strategy by γδ TCRs. Because γδ TCRs are normally stimulated by small molecule antigens, such as phosphoantigens derived from pathogens like Mycobacteria (72, 73), or by whole protein molecules, such as herpes simplex virus glycoprotein gI (74) and CD1 (75), the γδ TCR/T22 complex is an exception to the normal binding repertoire of γδ T cells, as this TCR is in complex with an MHC molecule. The most striking observation of ˚ this 3.4 A-resolution complex structure is the

severely tilted orientation of the γδ TCR with respect to its MHC-like ligand that by far exceeds the range of tilting angles seen in αβ TCRs (Figure 4) and is also slightly different within the two molecules present in the crystallographic asymmetric unit. The γδ TCR binding mode may then be more like those of antibodies rather than of αβ TCRs, which are restricted to MHC molecules. Indeed, structural comparison of the γδ TCR chains to antibodies and αβ TCR chains showed more antibody-like characteristics of the γδ TCR, which is also reﬂected in their diverse ligand speciﬁcities (103). The CDR loops of the two γδ TCR molecules in the crystallographic asymmetric unit are coaligned such that they could recognize two T22 ligands on a target cell simultaneously. Multimerization for some αβ TCRs has been observed by quasi-elastic light scattering in solution upon ligand binding (104), but no crystallographic evidence yet supports this notion. Thus, other crystal structures of γδ TCR/ligand complexes are needed to verify any possible increased multimerization propensity for γδ TCRs compared to αβ TCRs. The tilted orientation of the γδ TCR with respect to its ligand (Figure 4) almost entirely abolishes any contacts of the Vγ domain with T22; only two (complex 1) or four (complex 2) interactions between CDR3γ and T22 are present in the complex, whereas, surprisingly, the CDR1γ and CDR2γ loops are not utilized at all. The paucity of Vγ interactions (3 van der Waals contacts) is in stark contrast to the 116 Vδ interactions, in which the CDR3δ loop predominates (71 van der Waals contacts). CDR3δ loops in γδ TCRs are generally longer than in αβ TCRs and, owing to the sideways binding mode of G8, can fully access the exposed region of the T22 groove. The disordered α2 -helix region in the unliganded T22, which is also unstructured in T10 (56), does not become ordered upon TCR binding as might have been expected from comparison of an unrelated complex of the NK cell

www.annualreviews.org • MHC/ TCR Interactions

443

ARI

15 February 2006

0:51

receptor NKG2D with the nonclassical MHC MICA (105). In unliganded MICA, the partially unwound α2 -helix (66) refolds upon receptor binding, but this loop ordering is not observed with T22 as the γδ TCR does not engage this part of the MHC-like antigen. Some docking ﬂexibility of G8 with respect to ligand binding is apparent where the TCR pivots around CDR3δ, which leads to a rotation of 5◦ (Vδ) and 13◦ (Vγ) of the TCR domains when the two complexes in the asymmetric unit are compared with each other. Such TCR ﬂexibility was also seen in a recent αβ TCR complex, where the SB27 TCR binds to a bulged 13-mer peptide derived from EBV that is presented by the class I MHC HLA-B3508 (23). As the peptide termini are ﬁxed in the MHC class I groove, the additional central residues bulge outward and away from the MHC, limiting access of the TCR to the MHC αhelices; only two direct hydrogen bonds and 40 van der Waals contacts are present between the SB27 TCR and the HLA-B3508 MHC, whereas the peptide contributes 9 hydrogen bonds and 66 van der Waals contacts to the TCR/pMHC interface (Tables 2, 4). With the bulged peptide dominating the interface, the TCR “swivels” on top of the peptide, and the two copies of the TCR in this crystal form differ by a 12◦ rotation when compared with each other, as no other interactions with the MHC α-helices are observed that would restrict its orientation and docking angle (23). Thus, the γδ TCR docking ﬂexibility is not unique to this class of TCRs and does not seem to be a consequence of smaller buried surface area (1750 A˚ 2 ) or lower shape complementarity (Sc of 0.66), as both of these parameters are in the same range as αβ complexes (Tables 2, 3). In addition, as the precision of buried surface area and Sc calculations is also dependent on the resolution of the structure determination, more crystal structures of higher resolution and with different γδ TCRs are needed for statistically meaningful analyses on ligand recognition by γδ TCRs.

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

444

Rudolph

·

Stanfield

·

Wilson

TCR ASSEMBLY AND SIGNALING TCR/pMHC engagement is only the ﬁrst step in the assembly of what is now referred to as the immunological synapse, wherein not only TCRs but also coreceptors (CD4 and CD8) and additional signaling modules (CD3) interact to form the signaling-competent, supramolecular complex. No structural information is available on the entire T cell receptor assembly, but steric restrictions imposed by the shape and properties of the individual domains and subcomplexes whose structures are known provide some essential limitations on the architecture of the complex.

CD4 AND CD8 CORECEPTORS AND THEIR MHC COMPLEXES In addition to their cognate TCRs, class I and class II MHCs are recognized by their respective coreceptors CD8 and CD4. The current database for CD8 coreceptor consists of human CD8αα/HLA-A2 (106), murine CD8αα/H-2Kb (107) (Figure 8a), and murine CD8αα/TL (68) structures. In all complexes, the CD8αα homodimer binds primarily to the α3 domain of the MHC molecule in an antibody-like fashion, with the MHC α3 CD loop wedged between two corresponding CDR-like loops from the CD8αα dimers (Figure 8a). The structure and relative conformation of the C-terminal stalk region of CD8, which connects the coreceptor to the T cell surface, are still unclear, so the disposition of the TCR relative to the CD8αα/MHC scaffold is unresolved. The determination of a crystal structure of a CD8αβ heterodimer has also been elusive. CD8αβ modeling based on CD8αα, mutagenesis, and the different stalk lengths of the α and β subunits have been used to predict the orientation of the CD8αβ heterodimer with respect to the MHC (Figure 8). The CD8αβ heterodimer is not simply a functional homolog of the CD8αα homodimer,

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

Figure 8 CD8 and CD4 coreceptor binding to class I and class II MHC. (a) The MHC is colored blue, CD8αα in yellow and orange. It is not yet known which of the domains of CD8αα homodimer correspond to the CD8αβ heterodimer (upper and lower right panel ). The MHC is colored blue, and the CD4 (two N-terminal domains) is colored in yellow (lower left panel ). (b) Human CD3εδ (top), human CD3εγ (middle), and mouse CD3εγ (bottom). In all three panels, the common ε-chain is colored in blue. (c) Hypothetical TCR/pMHC/CD3εδ/CD3εγ/CD8 complex. The TCR/pMHC-CD8αα and putative CD8αβ interaction is modeled by superimposing two structures, the HLA-A2/CD8αα complex (1akj) and the TCR A6/HLA-A2/TaxP6A complex (1qrn) on their MHC residues α1–180, with TCR (green), MHC (dark blue), peptide (red ) and CD8 ( yellow and orange). The CD3εδ (1xiw, pink and blue) and CD3εγ (1sy6, gold and blue) are shown “docked” at the top of the ﬁgure, with the common ε-chains colored in blue. This “docking” merely represents placing of the CD3 structures in the vicinity of where they are thought to bind, roughly following the cartoon diagram in Reference 125. Lines are drawn in to depict tethers connecting the different subunits to the TCR cell membrane (top, green) or the antigen-presenting cell membrane (brown, bottom).

as CD8αβ is the true αβ TCR coreceptor, whereas the main function of CD8αα on intraepithelial lymphocytes is to aid in adaptation and survival in the gut (108). Clearly, a CD8αβ structure in complex with pMHC is needed to derive the structural basis of why CD8αα cannot functionally replace or complement CD8αβ. ˚ crystal strucThe low-resolution (4.3 A) ture of CD4 (domains 1 and 2) bound to I-Ak shows how the analogous coreceptor interaction is achieved in the MHC class II system

(109). Whereas both domains of CD8 cooperate to bind class I MHCs, only one domain (the N-terminal variable-like region) of CD4 makes contact with the MHC with the second tandem CD4 domain being distal to the interface. Comparison of the CD4/pMHC and CD8/pMHC structures exposes a surprising structural dichotomy of the MHC class I and class II architectures, implying profoundly different modes of organization in their respective immunological synapses (Figure 8). As the complete, four-domain crystal www.annualreviews.org • MHC/ TCR Interactions

445

ANRV270-IY24-14

ARI

15 February 2006

0:51

structure of human CD4 is also known (110), superposition of the MHC-proximal CD4 domains 1 and 2 permits assembly of a complete class II TCR/pMHC/CD4 model that suggests a V-shape with the T cell membraneproximal ends of the TCR and CD4 ˚ This separaseparated by around 100 A. tion would exclude direct TCR/CD4 interactions, but leaves ample room for binding of CD3.

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CD3 SIGNALING MODULES CD3 consists of subunits δ, ε, γ, and ζ that noncovalently associate to form CD3ζζ homodimers and CD3εδ and CD3εγ heterodimers (111, 112). Whereas sustained T cell responses rely on coreceptor binding and TCR aggregation (113–116), early TCR signaling is independent of these events but may instead rely on conformational changes in the CD3ε subunit (84). In addition, stable cell surface expression and normal development of αβ TCRs rely on the presence of the CD3 components (117–119). Sequence comparisons predicted that the extracellular domains of the CD3ε- and CD3γ-chains would adopt an immunoglobulin fold (120). A cavity formed by the FG loop in the Cβ domain of αβ TCRs was suggested to host a binding site for such an Ig-domain (9), but other studies reported the dispensability of any Igdomain residues in CDεδ for TCR α-chain binding, limiting the key residues to the conserved charged transmembrane residues on CD3 and the TCR (121). The ﬁrst insights into these signaling modules came from an NMR structure in which the extracellular domains of the mouse CD3ε and -γ subunits that lacked the conserved RxCxxCxE stalk region motif, considered important for dimerization, were converted to a single-chain format by a 26-residue linker that ensured close proximity during folding from inclusion bodies (120) (Figure 8b). The solution structure of this construct indeed revealed an Igfold of canonical type C2 (A Greek key motif) for the CD3ε and CD3γ subunits (122). The 446

Rudolph

·

Stanfield

·

Wilson

two subunits are connected via an intermolecular β-sheet, which would put the conserved RxCxxCxE stalk region motifs into close proximity to each other (Figure 8b). The crystal structure of the human CD3εγ heterodimer in complex with the therapeutic antibody Fab OKT3 (89) revealed a topology for CD3ε slightly different from that suggested by the solution structure (Figure 8b). In the human CD3ε, an eight-residue sequence insertion between β-strands C and E adds an additional β-strand D on the surface of CD3ε. This β-strand is distal to the subunit interface and signiﬁcantly alters the surface shape and electrostatic properties of CD3ε (see below). As a result of the additional β-strand, human CD3ε adopts a C1 Ig-fold rather than the C2 Ig-fold present in mouse CD3ε(122). The higher precision of this crystal structure allowed reliable calculation of Sc values and buried surface areas (Sc of 0.76 and 1840 A˚ 2 , respectively), which explains the high afﬁnity of the subunits for each other and the fact that the cysteine-rich stalk region is not necessary for CD3εγ complex formation (123, 124). Comparison of the NMR and crystal structures indicated a 23◦ rotation of the domains about the pseudo-twofold axis relating the ε and γ subunits. Thus, variability in domain associations seems to arise among these accessory modules, as they do in other Ig proteins, such as TCRs and antibodies. Also, compared to mouse CD3ε, signiﬁcant differences are present in the surface potential of human CD3ε, mainly due to the presence of the sequence Asp-Glu-Asp, which connects β-strands D and E and considerably increases the size of an electronegative patch that is also present in mouse CD3ε. Both CD3εγ structures, therefore, support a TCR binding model that is based mainly on electrostatic interactions, although their detailed interactions may vary. In the human OKT3/CD3εγ complex, the antibody Fab binds at a site remote from the proposed TCR interacting surface of CD3εγ. Both antibody and TCR appear able to bind

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

ARI

15 February 2006

0:51

the CD3εγ module if it is kinked. This kinking, as the authors speculate, could transduce a conformational change across the plasma membrane that could represent the molecular basis for the induction of early T cell signaling by OKT3. The elusive human CD3εδ structure was recently determined in complex with the scFv fragment of the UCHT1 antibody (125) (Figure 8b). In contrast to the CD3εγ structures, no linker was used in the production of a CD3εδ complex, and each domain included the conserved cysteine-rich stalk region. The ε and δ ectodomains were produced separately and refolded with the antibody scFv fragment. Similar to CD3γ, CD3δ adopts a C2 Ig-fold and pairs with CD3ε via an intersubunit β-sheet that buries a similar surface area (1740 A˚ 2 ) between the ectodomains as CD3εγ. Although the complete ectodomains were crystallized, no interpretable electron density was observed for the stalk regions containing the conserved CxxCxExD motif. The presence of a disulﬁde bond in the stalk region was detected in most of the material by nonreducing SDS-PAGE, which should nevertheless have reduced the ﬂexibility of the stalk region. The earlier proposal (122) that pairing of the CD3 subunits via the G-strand should lead to ordering of this region is apparently not the case, at least for these crystal forms of CD3. Although the interfaces of CD3εγ and CD3εδ are conserved, their molecular surfaces are quite different, with CD3δ being more electronegative than CD3γ (calculated pIs of 5 and 9, respectively). Of the 13 conserved surface residues in CD3δ, 11 are absent in CD3γ. Some of the conserved CD3δ residues (Glu9, Asp10, Arg11, and Lys41) form a charged patch on CD3δ that may constitute a TCR or coreceptor binding site (125). CD3δ and CD3γ are both N-glycosylated, with two sites each at residues 38, 74 and 52, 92, respectively, whereas CD3ε is not glycosylated. Both the OKT3 and UCHT1 antibodies bind at sites distal from the proposed

TCR-interaction sites, and their binding sites overlap. Thus, this binding site may constitute an immunodominant epitope (125). The larger buried surface area by UCHT1 (1790 A˚ 2 ) compared to OKT3 (1140 A˚ 2 ) may explain the higher afﬁnity of UCHT1 for CD3 (Kd = 0.5 nM versus 2.6 μM). However, the mechanism of action of these antibodies seems to be the same, regardless of their afﬁnities.

TCR/CD3 ASSEMBLY The stoichiometry of the signaling-competent αβ TCR complex, as well as the sequence of its assembly and the chemical nature of the interactions between the subunits, has long been enigmatic. The early signaling TCR complex seems to consist of heterodimers of αβ TCR, CD3εγ, and CD3εδ, and a homodimer of CD3ζ ζ (121,126). Nine conserved charged residues in the transmembrane segments of the αβ TCR (three basic residues, an arginine and a lysine in the α-chain, and a lysine in the β-chain), CD3ε (aspartate), CD3γ (aspartate), CD3δ (glutamate), and CD3ζ (aspartate), could electrostatically steer docking of the subunits (127, 128). This electrostatic interaction would be stronger in the membrane than in water owing to the smaller dielectric constant in a membrane environment (129). How then are these charged residues shielded from the energetically unfavorable membrane environment in the absence of complex formation? Perturbation of the pKa’s of the side chains could eliminate formal charges in the noncomplexed dimers but would still leave unsatisﬁed and, therefore, destabilizing membrane-inserted hydrogen bond donors and acceptors. The TCR α chain binds CD3εδ (130), which would allow CD3εγ to interact with the TCR β-chain (131), but no structural information is available for the transmembrane regions of these TCR components. If an α-helical conformation is assumed, which is likely given their

www.annualreviews.org • MHC/ TCR Interactions

447

ARI

15 February 2006

0:51

hydrophobic sequence propensities, the two basic residues in the transmembrane region of the TCR α-chain would lie on opposite faces, which could enable binding of the α-chain not only to CD3εδ, but also to CD3ζ ζ . The sequence of binding events to the TCR has been suggested to occur in the order CD3εδ, CD3εγ, and then CD3ζ ζ (127). With almost all the extracellular domains of the αβ TCR signaling complex known, a tentative structural model can be put forward that contains all the αβ TCR, MHC, CD3εγ, CD3εδ, and CD8 components, lacking only the CD3ζ ζ -chains (Figure 8c). In addition to stereochemical requirements, construction of this supramolecular assembly relies on the electrostatic interactions of the conserved residues in the transmembrane regions of the TCR and CD3s, and on the fact that carbohydrates generally do not participate in protein-protein interactions and, therefore, shield surfaces that do not participate in speciﬁc complex formation (132). A model proposed for a TCR/CD3εγ/CD3εδ complex (125) can be extended to include the CD8 coreceptor. The main features of this model (125) include a compact TCR/CD3 complex with trimeric transmembrane contacts among all components (α-ε-δ, β-ε-γ, and α-ζ -ζ ). To facilitate interactions between the transmembrane regions of the components, the bulky CD3 dimer ectodomains probably lie at an angle with respect to the membrane and the TCR globular domains. This notion is also supported by the shape of the membraneproximal surface of the TCR, which would nicely ﬁt the CD3 dimers. The CD3εγ and CD3εδ dimers would interact with conserved, nonglycosylated regions of the TCR surface. Although not present in any of the current TCR structures, the length of the peptide sequences connecting the TCR α- and β-chains to the membrane suggests that the TCR would be located further from the membrane than the CD3 dimers and, hence, sit “above” them (Figure 8c).

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14

448

Rudolph

·

Stanfield

·

Wilson

MECHANISMS OF pMHC/TCR DOCKING Not every engagement of a pMHC with its cognate TCR results in T cell activation. Indeed, antagonist peptides can inhibit T cell activation, and, similarly, engagement with a partial agonist elicits some, but not all, of the responses that characterize T cell activation by fully agonistic ligands. It was expected that, by examination of crystal structures of the respective altered peptide ligands (APL) TCR/pMHC complexes, this range of responses could be correlated with structural features, such as domain or CDR loop rearrangements. Thus far, the crystal structures do not explain the large biological differences or outcomes that can arise in T cell signaling from binding of APLs (14, 17) other than some minor changes in complementarity between the surfaces that can affect the halflife of the complexes. The crystal structures of TCR/pMHC complexes deﬁne the endpoints of the docking process. Most likely, this endpoint structure initiates TCR signaling. Interesting new results conﬁrm that the overall dimensions of the TCR/pMHC complex dictate TCR triggering (133), where the relatively small TCR/pMHC complex brings the T cell and antigen-presenting cell membranes close enough together so that larger molecules such as the CD45 phosphatase are occluded, allowing the TCR-CD3 ITAMs to remain phosphorylated and thus to initiate downstream signaling events. But understanding of the early steps of TCR docking is also important because they deﬁne the antigen and TCR selection processes. The precise steps of this binding and signaling mechanism are largely unknown, despite extensive data on TCR-MHC complexation, including data on association and dissociation kinetics, half-life determination, and relative afﬁnities (134– 140). The sheer number of antigen complexes that have to be scanned by the TCR necessitates a rather cursory screen, i.e., a rapid mechanism to discriminate self from nonself in the periphery. Still, the scan needs to be

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-II

ARI

15 February 2006

0:56

comprehensive enough to select reasonable afﬁnity TCRs that allow for T cell activation. To this end, several hypotheses have been put forward to account for these requirements, including a two-step mechanism and electrostatic steering for TCR docking. The two-step mechanism is based on forming an encounter complex of the TCR with the MHC α1 - and α2 -helices, followed by a more intensive sampling of the content of the MHC peptide-binding groove by the TCR CDR3 loops (141). Such a mechanism would generally be consistent with TCR/pMHC crystal structures, which show that the CDR1 and CDR2 loops primarily contact the MHC, whereas the highly diverse CDR3 loops mainly interact with the peptide (Figure 9, 10; Tables 5–8). For recognition of class II pMHC complexes, the two-step mechanism may be more appropriate, as the peptide lies deeper in the binding groove, so the ﬁrst contact of the pMHC with the TCR would be dominated by encounter with the MHC α-helices. Indeed, using surface plasmon resonance, this mechanism has been supported by investigating the 2B4 TCR and its interactions with MHC class II I-Ek (141) containing a moth cytochrome C peptide. A two-step process in the class I system of B6 TCR Tax-HLA-A2 pMHC was detectable but less pronounced (142, 143). This distinction of consecutive scanning and reading steps may be arbitrary where peptide bound to a class I MHC (45) bulges extensively out of the groove so that the TCR encounters the peptide and the MHC simultaneously (22, 23). In such a case, only longrange electrostatic steering could preorient the TCR relative to the MHC without direct antigen contact. However, these surfaces should also not be too highly charged or they would bind other counter-ions that would need to be removed and hence would compete with the TCR for interaction. Along these lines, some short-(salt bridges) to longrange (>4 A˚ distance) electrostatic interactions have been found in TCR/pMHC crystal structures, for example between TCR residue

Figure 9 Conserved contacts formed between MHC class I and class II residues and αβ TCR. The MHC Cα backbone is shown for class I (top), class II (middle), and class I and II combined (bottom) in gray. On each backbone, spheres are placed at Cα positions of residues that contact TCR. The spheres are drawn so that their diameters are in proportion to their numbers of contacts to TCR (so that the large spheres represent residues with the most contacts). The numbers of contacts are listed in Tables 5 and 6 for MHC residues and in Tables 7 and 8 for TCR residues. www.annualreviews.org • MHC/TCR Interactions

449

ARI

15 February 2006

0:56

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-II

Figure 10 Conserved contacts formed between TCR residues and MHC. The TCR Cα backbone is shown for three different class I TCRs (left column) and three different class II TCRs (right column), with one TCR repeated on the bottom of each column. On each of the top three TCRs in each column, spheres are placed at Cα positions of residues that contact MHC. The spheres are drawn so that their diameters are in proportion to their actual numbers of contacts to TCR (thus, the large spheres represent residues with the most contacts). The numbers of contacts are listed in Tables 7 and 8 for TCR residues. The “conserved” contacts for TCRs of each class are shown as spheres representing the average number of contacts for each residue (bottom row). 450

Rudolph

·

Stanfield

·

Wilson

10

2

1

11

6

A150

2 1

1

5

www.annualreviews.org • MHC/TCR Interactions 3

4

A170

1

2

5

2

2

A167

A166

2

3

6

2

1

2

15

6

2

14

A163

1

3

3

1

1

1

6

7

6

2

21

8

A162

1

5

1

15

1

1

4

11

3

1

3

1

9

3

1

A159

2

4

A158

9

9

16

A155

7

1

7

4

6

12

A154 24

4

7

1

1

1

6

6

A152

A151

3

A149 2

1

2

A146

A147

4

A79

5

1

3 2

6

15

5

5

12

15

3

3

4

1

4

A76

3

1

3

4

1

1

2

2

2

5

4

11

1oga

A75

A73

A72

3

1

14

1lp9

PDB code

2

16

2

1

7

3

2

10

8

13

1

1

17

10

6

6

3

1mi5

3

7

7

9

1

1

3

2

2

4

1

4

9

2

4

17

1qrn

4

5

4

4

1

4

10

1

11

2

4

7

9

2

4

20

1qse

5

5

6

4

1

3

9

11

5

6

2

5

20

1

1qsf

9

1

4

6

3

6

16

4

12

4

16

2bnr

13

2

9

7

2

7

15

4

11

3

9

2bnq

4

3

15

8

5

4

1

1

3

2ak4

15

29

26

50

8

7

63

171

35

9

71

91

14

3

42

20

42

2

29

86

5

79

35

41

168

85

4

20

Sum of contacts

0.9

1.7

1.5

2.9

0.5

0.4

3.7

10.1

2.1

0.5

4.2

5.4

0.8

0.2

2.5

1.2

2.5

0.1

1.7

5.1

0.3

4.7

2.1

2.4

9.9

5.0

0.2

1.2

Average

15 February 2006

A70

8

1

A69

7

1

4

12

A68

2

11

7

6

A66

9

1

2

13

1bd2

ARI

6

8

9

1ao7

2

8

1nam

28

2

1mwa

A65

2

1g6r

A62

14

1fo0

4

6

2ckb

A59

A58

1kj2

Numbers of contacts made by class I MHC on TCR sorted by MHC residue

Residue #

Table 5

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-II 0:56

451

ANRV270-IY24-14-II

ARI

15 February 2006

Table 6

0:56

Numbers of contacts made by class II MHC on αβTCR sorted by MHC residue PDB code

Residue #

1u3h

1j8h

1fyt

A39

6

5

10

A54

1

A55

3

A57

13

6

A58

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

1zgl

16

9

12

6

4

1

3

9

A60

2

2

1

A61

31

9

7

A62

1ymm

2

Sum of contacts

Average

5

26

4.3

1

0.2

28

4.7

57

9.5

13

2.2

16

2

3

10

1.7

8

19

76

12.7

1

1

2

0.3

A64

1

7

9

3

20

3.3

A65

1

2

5

4

12

2.0

A67

1

4

4

9

1.5

A68

3

2

1

6

1.0

4

0.7

1

0.2

14

2.3

9

1.5

12

28

4.7

8

31

5.2

16

76

12.7

4

4

0.7

B60

3

B61

1

B64

4

2

B66

1

2

6 1

B67

5

6

4

B69

3

4

1

15

B70

15

13

11

9

B73

3

12

3

B76 B77

5

5

3

8

B80 B81

5

10

9

1 8

B72

12

6

12

2.0

10

1

11

1.8

4

4

29

4.8

8

8

1.3

13

49

8.2

B84

3

3

0.5

B85

3

3

0.5

Lys68 in the HV4α loop and Asp76β in MHC class II or Glu166α in MHC class I (144). More examples include the electrostatic interaction in the MHC class I complex LC13/HLA-B8/EBV (24) between CDR2β residue Glu52 and HLA-B8 residue Arg79, and the interaction seen in two MHC class II complexes (HLA-DR1 and HLA-DR4) (18, 26), between Lys39α in a loop that projects up and away from the ﬂoor of the β-sheet and Glu56β of the HA1.7 TCR CDR2β. In a recent study, a single point mutation in the CDR3β loop of the 2C TCR (Gly95Arg) increased its afﬁnity by a factor of

452

1d9k

Rudolph

·

Stanfield

·

Wilson

1000 to the QL9/Ld pMHC, most likely due to direct electrostatic interaction of the TCR arginine side chain with an aspartate residue at P8 (145). Thus, although such salt bridges and hydrogen bonds have not been conserved in all TCR/pMHC class I complexes, electrostatic effects, especially for orienting purposes, can work at a distance (146), so their inﬂuence on orienting the TCR relative to the pMHC at an early stage during antigen recognition must be considered. The glycan shield around these molecules may also inﬂuence the docking and help orient and exclude certain modes of binding (147).

5

8

2

α100

β27

α103

α102

3

2

9

15

α101

9

6

22

α99

6

1

5

10

3

3

22

α97

α98

4

α96

α95

α94

2

4

α93

2

2

6

1

8

6

1

2

2

10

1nam

10

5

1

9

4

1

2

15

1

2

2

4

2

1

4

7

4

7

7

4

1mwa

α68

α55

α53

α52

α51

1

1

1g6r

7 8

1fo0

1

11

8

17

2

1

4

1

1

2

4

5

1

7

7

2

1ao7

4

1

5

6

4

11

5

2

5

6

10

4

1bd2

7

15

2

6

9

11

5

3

20

10

1

16

1lp9

2

6

3

1

10

9

1oga

10

2

10

8

2

3

3

6

3

14

5

1

1

1mi5

3

17

10

25

3

1

11

1

3

4

8

1

9

1qrn

5

19

11

17

1

6

2

6

5

6

2

8

1qse

3

14

7

20

4

7

2

2

5

4

6

5

8

1

1qsf

21

2

9

1

4

9

5

4

3

1

4

41

2bnr

22

10

1

4

7

5

5

8

6

3

6

40

2bnq

3

12

24

12

3

4

3

4

4

2ak4

3

4

69

99

100

125

63

69

45

18

3

35

31

3

19

45

18

93

4

70

126

10

47

72

25

2

Sum of contacts

www.annualreviews.org • MHC/TCR Interactions

(Continued )

0.2

0.2

4.1

5.8

5.9

7.4

3.7

4.1

2.7

1.1

0.2

2.1

1.8

0.2

1.1

2.7

1.1

5.5

0.2

4.1

7.4

0.6

2.8

4.2

1.5

0.1

Average

15 February 2006

α50

α48

α31

α30

α29

α28

5

2ckb

ARI

α27

α26

α1

1kj2

PDB code

Numbers of contacts by TCR on class I MHC sorted by αβTCR residue

Residue #

Table 7

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-II 0:56

453

454

Rudolph

·

Stanfield

·

Wilson

2

β31

1

β51

3

3

β104

β103

β102

2

15

β100

β101

4

β99

β98

1

1

β97

3

2

4

3

β96

β95

β72

β56

β55

β54

β53

β52

5

β50

β48

β33

4

β30

6

1

2ckb

2

14

18

25

1

6

4

1

1fo0

4

12

1

2

1

14

2

1

1g6r

3

11

2

1

1

1

7

21

7

11

1mwa

1

6

6

9

1

6

3

1nam

5

15

7

3

1

16

1

3

1ao7

9

1

1

18

2

1

2

1bd2

5

1

2

23

6

4

5

2

2

1

1lp9

PDB code

13

5

1

20

1

4

3

5

2

6

8

5

1oga

17

11

15

3

3

2

2

6

8

1

1

1mi5

6

3

4

3

2

15

2

4

1qrn

13

12

4

1

6

19

1

1

4

1qse

10

9

3

3

1qsf

9

6

16

1

10

8

16

5

3

4

5

1

11

3

5

1

2

14

3

2bnq

5

12

3

3

1

6

15

4

2bnr

1

6

1

5

1

4

4

10

37

51

39

61

51

119

127

43

8

7

19

2

31

5

12

32

50

13

15

98

29

20

Sum of contacts

6

8

11

2ak4

0.6

2.2

3.0

2.3

3.6

3.0

7.0

7.5

2.5

0.5

0.4

1.1

0.1

1.8

0.3

0.7

1.9

2.9

0.8

0.2

0.9

5.8

1.7

1.2

Average

15 February 2006

β29

β28

1kj2

(Continued )

ARI

Residue #

Table 7

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-II 0:56

ANRV270-IY24-14-III

ARI

Table 8

15 February 2006

1:1

Numbers of contacts made by αβTCR on class II MHC sorted by TCR residue PDB code

Residue # α26

1u3h

3

1d9k

Sum of contacts

Average

1

2

0.3

10

1

12

2.0

6

7

24

4.0

25

1

50

8.3

14

16

27.0

5

0.8

2

29

4.8

1zgl

5

3

7

8

9

2 3

α31

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

1ymm

1

α29 α30

1fyt

1

α27 α28

1j8h

2

α50

4

1

22

α51

1

1

1

1

4

0.7

α52

2

3

1

13

19

3.2

α68

1

1

0.2

α93

1

1

0.2

15

2.5

α94

4

6

5

α96

3

5

8

1.3

α97

13

13

26

4.3

1

0.2

5

50

8.3

4

14

2.3

16

27

4.5

40

6.7

4

0.7

1

α98 α99

16

16

α100

3

7

α101

5

1

5

α102

4

18

4

β28 β30

3

β31

3

β48

2

β50

27

β51 β52

8

6

2

2

10

12

5

13

12

42

7.0

1

6

1.0

7

12

2.0

25

75

12.5

20

3.3

1

0.2

10

10

1.7

5

0.8

2 3 7

10

6

7

9

4

1

β53 β54

2

2

1

β55

1

3

10

2

16

2.7

9

10

3

5

36

6.0

1

1

7

1.2

1

0.2

1

2

0.3

7

39

6.5

β56

9

5

β57 1

β72 β96

1

β97

5

6

11

10

β98

7

10

8

8

1

9

43

7.2

5

17

9

31

5.2

12

4

34

5.7

6

6

1.0

3

3

0.5

5

25

4.2

β99 β100

5

9

4

β101 β103 β104

20

www.annualreviews.org • MHC/TCR Interactions

455

ARI

15 February 2006

1:1

Ionic interactions may also be implicated in the interaction of pMHC with the plasma membrane where the pMHC has been observed as being severely tilted with its long axis parallel to the membrane in a supine orientation (148). In addition, a kinked orientation of the CD3 modules relative to the T cell membrane was assumed to be an essential prerequisite for TCR binding (125). This “fourth dimension” in T cell signaling, which becomes available to any of the modules participating in the assembly of the TCR, may prove more important than suggested by the isolated crystal structures, particularly in light of the lipid rafts that have been implicated in membraneregulated signaling events (149, 150). So we must return again to the issue of MHC restriction. Detailed analyses of these 24 TCR/pMHC complexes do not readily identify a conserved set of interactions that would dictate a common binding orientation of the TCR on the pMHC. A variety of docking orientations from diagonal to near orthogonal (range 22◦ –87◦ ), and some additional lateral mobility along the groove can displace some TCRs from their roughly central location over the middle of the peptide to either end of the binding groove (Figure 5, 6; Tables 2–4). If we list the contacts that each pMHC residue makes with TCR, no absolutely conserved interactions are made (Table 7, 8). However, trends develop when these complexes are considered as a whole. Most TCRs that recognize class I or class II MHCs clearly focus their binding interactions on the central regions of the α1 - and α2 /β1 helices. Several MHC residues, such as α65 and α155 of class I and the corresponding α57 and β70 of class II, have the highest average number of contacts (Figure 9). These conserved contact residues also stand out when both classes are grouped together and correspond to α65/α57, α69/α61, and α155/β70 for class I/class II MHC. No such compelling picture arises from similar analyses of TCR contact residues (Figure 10). High variability in the location and number of contacts

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

456

Rudolph

·

Stanfield

·

Wilson

is found in each individual TCR that does not correlate well with average values. This ﬁnding is perhaps not unexpected because of the enormous repertoire of TCRs that can be produced against the very limited arsenal of MHCs. However, it might have been predicted that the germ line–encoded CDRs 1 and 2 would have the most conserved contacts. This prediction is true to some extent, but the data are not really convincing. Residues α30 (CDR1) and α50 (CDR2) make the most frequent contacts for TCRs to class I MHC, and, similarly, α29 and α50 for TCRs to class II MHC. Residue β30 (CDR1) is the only relatively conserved contact residue in the TCR β-chain, and some variation in the use of β50 (CDR2) as a contact residue is noted between TCRs interacting with class I and class II MHC. Several residues in CDRs 3α (α99) and 3β (β97, β98) have the most conserved contacts in both classes. What dictates this variable but still relatively conserved docking orientation? At present, we must fall back on overall shape complementarity, restricted orientation through interaction with coreceptors, and electrostatic steering.

CONCLUSIONS AND FUTURE PERSPECTIVES More than 400 antibody structures have now been determined to delineate the full extent of antibody-antigen interactions and the general principles that governed antibody-antigen recognition. But even now, completely novel modes of binding and unexpected features continue to emerge from human and other antibody complexes. From the comparatively limited number of TCR/pMHC structures, we conclude that TCRs bind MHC class I and class II in a somewhat conserved way, but with some considerable structural variation in the details of the interaction. A common docking mode would enable the αβ TCR to quickly survey the contents of the MHC binding groove. However, the 13 independent complex structures determined so far have not yet

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

ARI

15 February 2006

1:1

revealed the basis for this conserved orientation and hence the basis for MHC restriction. The variability in the pitch, twist, and roll of the TCR indicates that individual solutions to the docking problem are found that differ substantially in their details. In many cases, the TCR Vα interactions with the pMHC seem to predominate and thus provide some basis for a conserved orientation. But in several alloreactive TCR/pMHC complexes, the β-chain seems to provide most of the interactions with pMHC. Also unresolved is how the exceedingly small changes in the TCR/pMHC interface in response to different APLs can lead to such drastically different biological outcomes. The TCR itself seems to adapt to small changes in the pMHC ligand by small conformational changes or rearrangements of its central CDR loops. The complementarity, buried surface area, or number of contacts in agonist versus antagonist complexes are very similar and are difﬁcult to reconcile with the substantial differences in T cell responses. Proposals that changes in the CDR conformations themselves through induced ﬁt provide some discrimination (138) seem hard to reconcile. Therefore, differentiation of strong from weak agonists, or agonists from antagonists, by visual inspection of the crystal structures is not yet possible. The future direction still demands more TCR/pMHC complex structures to address these key issues and to garner the general principles that govern TCR/pMHC recognition.

So far, these soluble TCR/pMHC complexes are not in their native context on the membrane surface, nor are they surrounded by the other signaling components of the TCR, such as CD3, nor in the vicinity of their coreceptors or costimulatory receptors. Therefore, the most important breakthrough would be the determination of a complete αβ TCR signaling complex, including CD4/CD8 and the CD3γ-, δ-, ε-, and ζ -chains. This assembly would deﬁne the global changes that inﬂuence TCR signaling events. However, the lack of the membraneanchoring domains in the constructs used for the current structure determinations will remain a problem until intact membrane proteins can be routinely crystallized. Meanwhile, models of the TCR/pMHC in complex with coreceptors (CD4/CD8) and signaling modules (CD3εγ and CD3εδ) can be assembled from the component pieces (Figure 8c), but have to be interpreted with caution. Notwithstanding, substantial advances have certainly been made in the past two years in our understanding of the recognition of MHC class I and class II by αβ TCRs, and now of antigen recognition by γδ TCRs, as well as obtaining structural insights into alloreactivity and graft rejection, response to APLs and bulged ligands, autoimmunity, and TCR selection and bias. Future studies should also deal with the extent to which other bulky ligands, especially glycolipids or lipopeptides in the case of CD1 (68), can be accommodated within the TCR/pMHC interface.

ACKNOWLEDGMENTS We thank Christopher Garcia, Roy Mariuzza, and Jamie Rossjohn for providing coordinates prior to publication. The authors are supported by DFG SFB 523 (M.G.R.) and NIH grants CA-58896 and AI-42266 (R.L.S. and I.A.W.). This is manuscript #17589-MB from The Scripps Research Institute.

LITERATURE CITED 1. Fields BA, Ober B, Malchiodi EL, Lebedeva MI, Braden BC, et al. 1995. Crystal structure of the Vα domain of a T cell antigen receptor. Science 270:1821–24 www.annualreviews.org • MHC/TCR Interactions

457

ARI

15 February 2006

1:1

2. Fields BA, Malchiodi EL, Li H, Ysern X, Stauffacher CV, et al. 1996. Crystal structure of a T-cell receptor β-chain complexed with a superantigen. Nature 384:188–92 3. Li H, Lebedeva MI, Ward ES, Mariuzza RA. 1997. Dual conformations of a T cell receptor Vα homodimer: implications for variability in Vα Vβ domain association. J. Mol. Biol. 269:385–94 4. Plaksin D, Chacko S, Navaza J, Margulies DH, Padlan EA. 1999. The X-ray crystal structure of a Vα2.6Jα38 mouse T cell receptor domain at 2.5 A˚ resolution: alternate modes of dimerization and crystal packing. J. Mol. Biol. 289:1153–61 5. Machius M, Cianga P, Deisenhofer J, Ward ES. 2001. Crystal structure of a T cell receptor Vα11 (AV11S5) domain: new canonical forms for the ﬁrst and second complementarity determining regions. J. Mol. Biol. 310:689–98 6. Rudolph MG, Huang M, Teyton L, Wilson IA. 2001. Crystal structure of an isolated Vα domain of the 2C T-cell receptor. J. Mol. Biol. 314:1–8 7. Garcia KC, Degano M, Stanﬁeld RL, Brunmark A, Jackson MR, et al. 1996. An αβ T cell receptor structure at 2.5 A˚ and its orientation in the TCR-MHC complex. Science 274:209–19 8. Housset D, Mazza G, Gregoire C, Piras C, Malissen B, Fontecilla-Camps JC. 1997. The three-dimensional structure of a T-cell antigen receptor VαVβ heterodimer reveals a novel arrangement of the Vβ domain. EMBO J. 16:4205–16 9. Wang J, Lim K, Smolyar A, Teng M, Liu J, et al. 1998. Atomic structure of an αβ T cell receptor (TCR) heterodimer in complex with an anti-TCR fab fragment derived from a mitogenic antibody. EMBO J. 17:10–26 10. Kjer-Nielsen L, Clements CS, Brooks AG, Purcell AW, McCluskey J, Rossjohn J. 2002. The 1.5 A˚ crystal structure of a highly selected antiviral T cell receptor provides evidence for a structural basis of immunodominance. Structure 10:1521–32 11. Garboczi DN, Ghosh P, Utz U, Fan QR, Biddison WE, Wiley DC. 1996. Structure of the complex between human T-cell receptor, viral peptide and HLA-A2. Nature 384:134–41 12. Garcia KC, Degano M, Pease LR, Huang M, Peterson PA, et al. 1998. Structural basis of plasticity in T cell receptor recognition of a self peptide-MHC antigen. Science 279:1166– 72 13. Ding YH, Smith KJ, Garboczi DN, Utz U, Biddison WE, Wiley DC. 1998. Two human T cell receptors bind in a similar diagonal mode to the HLA-A2/Tax peptide complex using different TCR amino acids. Immunity 8:403–11 14. Ding YH, Baker BM, Garboczi DN, Biddison WE, Wiley DC. 1999. Four A6TCR/peptide/HLA-A2 structures that generate very different T cell signals are nearly identical. Immunity 11:45–56 15. Reinherz EL, Tan K, Tang L, Kern P, Liu J, et al. 1999. The crystal structure of a T cell receptor in complex with peptide and MHC class II. Science 286:1913–21 16. Reiser JB, Darnault C, Guimezanes A, Gregoire C, Mosser T, et al. 2000. Crystal structure of a T cell receptor bound to an allogeneic MHC molecule. Nat. Immun. 1:291–97 17. Degano M, Garcia KC, Apostolopoulos V, Rudolph MG, Teyton L, Wilson IA. 2000. A functional hot spot for antigen recognition in a superagonist TCR/MHC complex. Immunity 12:251–61 18. Hennecke J, Carﬁ A, Wiley DC. 2000. Structure of a covalently stabilized complex of a human αβ T-cell receptor, inﬂuenza HA peptide and MHC class II molecule, HLA-DR1. EMBO J. 19:5611–24 19. Luz JG, Huang M, Garcia KC, Rudolph MG, Apostolopoulos V, et al. 2002. Structural comparison of allogeneic and syngeneic T cell receptor-peptide-major histocompatibility

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

458

Rudolph

·

Stanfield

·

Wilson

ANRV270-IY24-14-III

20.

21. 22.

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

23.

24.

25.

26.

27.

28.

29.

30. 31.

32. 33. 34. 35. 36.

ARI

15 February 2006

1:1

complex complexes: a buried alloreactive mutation subtly alters peptide presentation substantially increasing Vβ Interactions. J. Exp. Med. 195:1175–86 Buslepp J, Wang H, Biddison WE, Appella E, Collins EJ. 2003. A correlation between TCR Vα docking on MHC and CD8 dependence: implications for T cell selection. Immunity 19:595–606 Stewart-Jones GB, McMichael AJ, Bell JI, Stuart DI, Jones EY. 2003. A structural basis for immunodominant human T cell receptor recognition. Nat. Immun. 4:657–63 Tynan FE, Borg NA, Miles JJ, Beddoe T, El-Hassen D, et al. 2005. High resolution structures of highly bulged viral epitopes bound to major histocompatibility complex class I. Implications for T-cell receptor engagement and T-cell immunodominance. J. Biol. Chem. 280:23900–9 Tynan FE, Burrows SR, Buckle AM, Clements CS, Borg NA, et al. 2005. T cell receptor recognition of a ‘super-bulged’ major histocompatibility complex class I-bound peptide. Nat. Immunol. 6:1114–22 Kjer-Nielsen L, Clements CS, Purcell AW, Brooks AG, Whisstock JC, et al. 2003. A structural basis for the selection of dominant αβ T cell receptors in antiviral immunity. Immunity 18:53–64 Chen JL, Stewart-Jones G, Bossi G, Lissin NM, Wooldridge L, et al. 2005. Structural and kinetic basis for heightened immunogenicity of T cell vaccines. J. Exp. Med. 201:1243– 55 Hennecke J, Wiley DC. 2001. Structure of a complex of the human αβ-T cell receptor HA1.7, Inﬂuenza HA peptide, and MHC class II molecule, HLA-DR4 (DRA∗ 0101, DRB∗ 0401)—insight into TCR cross-restriction and alloreactivity. J. Exp. Med. 195:571– 81 Hahn M, Nicholson MJ, Pyrdol J, Wucherpfennig KW. 2005. Unconventional topology of self peptide-major histocompatibility complex binding by a human autoimmune T cell receptor. Nat. Immun. 6:490–96 Maynard J, Petersson K, Wilson DH, Adams EJ, Blondelle SE, et al. 2005. Structure of an autoimmune T cell receptor complexed with class II peptide-MHC: insights into MHC bias and antigen speciﬁcity. Immunity 22:81–92 Li Y, Huang Y, Lue J, Quandt JA, Martin R, Mariuzza RA. 2005. Structure of a human autoimmune TCR bound to a myelin basic protein self-peptide and a multiple sclerosisassociated MHC class II molecule. EMBO J. 24:2968–79 Reiser JB, Darnault C, Gregoire C, Mosser T, Mazza G, et al. 2003. CDR3 loop ﬂexibility contributes to the degeneracy of TCR recognition. Nat. Immun. 4:241–47 Reiser JB, Gregoire C, Darnault C, Mosser T, Guimezanes A, et al. 2002. A T cell receptor CDR3β loop undergoes conformational changes of unprecedented magnitude upon binding to a peptide/MHC class I complex. Immunity 16:345–54 Garcia KC, Teyton L, Wilson IA. 1999. Structural basis of T cell recognition. Annu. Rev. Immunol. 17:369–97 Garcia KC. 1999. Molecular interactions between extracellular components of the T-cell receptor signaling complex. Immunol. Rev. 172:73–85 Garcia KC, Degano M, Speir JA, Wilson IA. 1999. Emerging principles for T cell receptor recognition of antigen in cellular immunity. Rev. Immunogenet. 1:75–90 Wang J, Reinherz EL. 2000. Structural basis of cell-cell interactions in the immune system. Curr. Opin. Struct. Biol. 10:656–61 Rudolph MG, Wilson IA. 2002. The speciﬁcity of TCR/pMHC interaction. Curr. Opin. Immunol. 14:52–65 www.annualreviews.org • MHC/TCR Interactions

459

ARI

15 February 2006

1:1

37. Rudolph MG, Luz JG, Wilson IA. 2002. Structural and thermodynamic correlates of T-cell signaling. Annu. Rev. Biophys. Biomol. Struct. 31:121–49 38. Allison TJ, Garboczi DN. 2002. Structure of γδ T cell receptors and their recognition of non-peptide antigens. Mol. Immunol. 38:1051–61 39. Falk K, Rotzschke O, Stevanovic S, Jung G, Rammensee HG. 1991. Allele-speciﬁc motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature 351:290–96 40. Rudensky AY, Mazel SM, Yurin VL. 1990. Presentation of endogenous immunoglobulin determinant to immunoglobulin-recognizing T cell clones by the thymic cells. Eur. J. Immunol. 20:2235–39 41. van Bleek GM, Nathenson SG. 1991. The structure of the antigen-binding groove of major histocompatibility complex class I molecules determines speciﬁc selection of selfpeptides. Proc. Natl. Acad. Sci. USA 88:11032–36 42. Fremont DH, Matsumura M, Stura EA, Peterson PA, Wilson IA. 1992. Crystal structures of two viral peptides in complex with murine MHC class I H-2Kb . Science 257:919–27 43. Madden DR, Garboczi DN, Wiley DC. 1993. The antigenic identity of peptide-MHC complexes: a comparison of the conformations of ﬁve viral peptides presented by HLAA2. Cell 75:693–708 44. Stern LJ, Wiley DC. 1994. Antigenic peptide binding by class I and class II histocompatibility proteins. Structure 2:245–51 45. Speir JA, Stevens J, Joly E, Butcher GW, Wilson IA. 2001. Two different, highly exposed, bulged structures for an unusually long peptide bound to rat MHC class I RT1-Aa . Immunity 14:81–92 ˜ AR, Segelke BW, Stura EA, Peterson PA, Wilson IA. 1997. Crystal 46. Zeng Z, Castano structure of mouse CD1: An MHC-like fold with a large hydrophobic binding groove. Science 277:339–45 47. Zajonc DM, Elsliger MA, Teyton L, Wilson IA. 2003. Crystal structure of CD1a in ˚ Nat. Immun. 4:808–15 complex with a sulfatide self antigen at a resolution of 2.15 A. 48. Zajonc DM, Crispin MD, Bowden TA, Young DC, Cheng TY, et al. 2005. Molecular mechanism of lipopeptide presentation by CD1a. Immunity 22:209–19 49. Koch M, Stronge VS, Shepherd D, Gadola SD, Mathew B, et al. 2005. The crystal structure of human CD1d with and without α-galactosylceramide. Nat. Immun. 6:819– 26 50. Batuwangala T, Shepherd D, Gadola SD, Gibson KJ, Zaccai NR, et al. 2004. The crystal structure of human CD1b with a bound bacterial glycolipid. J. Immunol. 172:2382–88 51. Gadola SD, Zaccai NR, Harlos K, Shepherd D, Castro-Palomino JC, et al. 2002. Struc˚ a maze for alkyl chains. Nat. Immun. ture of human CD1b with bound ligands at 2.3 A, 3:721–26 52. De Libero G, Mori L. 2005. Recognition of lipid antigens by T cells. Nat. Rev. Immunol. 5:485–96 53. Van Rhijn I, Zajonc DM, Wilson IA, Moody DB. 2005. T-cell activation by lipopeptide antigens. Curr. Opin. Immunol. 17:222–29 54. Sullivan BA, Nagarajan NA, Kronenberg M. 2005. CD1 and MHC II ﬁnd different means to the same end. Trends Immunol. 26:282–88 55. Moody DB, Zajonc DM, Wilson IA. 2005. Anatomy of CD1-lipid antigen complexes. Nat. Rev. Immunol. 5:387–99 56. Rudolph MG, Wingren C, Crowley MP, Chien YH, Wilson IA. 2004. Combined pseudomerohedral twinning, non-crystallographic symmetry and pseudo-translation in a monoclinic crystal form of the γδ T-cell ligand T10. Acta Cryst. D60:656–64

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

460

Rudolph

·

Stanfield

·

Wilson

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

ARI

15 February 2006

1:1

57. Wingren C, Crowley MP, Degano M, Chien Y, Wilson IA. 2000. Crystal structure of a γδ T cell receptor ligand T22: a truncated MHC-like fold. Science 287:310–14 58. Wolan DW, Teyton L, Rudolph MG, Villmow B, Bauer S, et al. 2001. Crystal structure ˚ Nat. Immun. 2:248–54 of the murine NK cell-activating receptor NKG2D at 1.95 A. 59. Vance RE, Tanamachi DM, Hanke T, Raulet DH. 1997. Cloning of a mouse homolog of CD94 extends the family of C-type lectins on murine natural killer cells. Eur. J. Immunol. 27:3236–41 60. Lanier LL. 2005. NKG2D in innate and adaptive immunity. Adv. Exp. Med. Biol. 560:51– 56 61. Hamerman JA, Ogasawara K, Lanier LL. 2005. NK cells in innate immunity. Curr. Opin. Immunol. 17:29–35 62. Vivier E, Tomasello E, Paul P. 2002. Lymphocyte activation via NKG2D: towards a new paradigm in immune recognition? Curr. Opin. Immunol. 14:306–11 63. Stephens HA. 2001. MICA and MICB genes: Can the enigma of their polymorphism be resolved? Trends Immunol. 22:378–85 64. Yamamoto K, Fujiyama Y, Andoh A, Bamba T, Okabe H. 2001. Oxidative stress increases MICA and MICB gene expression in the human colon carcinoma cell line (CaCo-2). Biochim. Biophys. Acta 1526:10–12 65. Tieng V, Le Bouguenec C, du Merle L, Bertheau P, Desreumaux P, et al. 2002. Binding of Escherichia coli adhesin AfaE to CD55 triggers cell-surface expression of the MHC class I-related molecule MICA. Proc. Natl. Acad. Sci. USA 99:2977–82 66. Li P, Willie ST, Bauer S, Morris DL, Spies T, Strong RK. 1999. Crystal structure of the MHC class I homolog MIC-A, a γδ T cell ligand. Immunity 10:577–84 67. Li P, McDermott G, Strong RK. 2002. Crystal structures of RAE-1β and its complex with the activating immunoreceptor NKG2D. Immunity 16:77–86 68. Liu Y, Xiong Y, Naidenko OV, Liu JH, Zhang R, et al. 2003. The crystal structure of a TL/CD8αα complex at 2.1 A˚ resolution: implications for modulation of T cell activation and memory. Immunity 18:205–15 69. Claverie JM, Prochnicka-Chalufour A, Bougueleret L. 1989. Implications of a Fab-like structure for the T-cell receptor. Immunol. Today 10:10–14 70. Davis MM, Bjorkman PJ. 1988. T-cell antigen receptor genes and T-cell recognition. Nature 334:395–402 71. Li H, Lebedeva MI, Llera AS, Fields BA, Brenner MB, Mariuzza RA. 1998. Structure of the Vδ domain of a human γδ T-cell antigen receptor. Nature 391:502–6 72. Morita CT, Beckman EM, Bukowski JF, Tanaka Y, Band H, et al. 1995. Direct presentation of nonpeptide prenyl pyrophosphate antigens to human γδ T cells. Immunity 3:495–507 73. Belmant C, Espinosa E, Poupot R, Peyrat MA, Guiraud M, et al. 1999. 3-Formyl-1butyl pyrophosphate, a novel mycobacterial metabolite activating human γδ T cells. J. Biol. Chem. 274:32079–84 74. Chien YH, Jores R, Crowley MP. 1996. Recognition by γδ T cells. Annu. Rev. Immunol. 14:511–32 75. Moody DB, Besra GS, Wilson IA, Porcelli SA. 1999. The molecular basis of CD1mediated presentation of lipid antigens. Immunol. Rev. 172:285–96 76. Allison TJ, Winter CC, Fournie J, Bonneville M, Garboczi DN. 2001. Structure of a human γδ T-cell antigen receptor. Nature 411:820–24 77. Yewdell JW, Haeryfar SM. 2005. Understanding presentation of viral antigens to CD8+ T cells in vivo: the key to rational vaccine design. Annu. Rev. Immunol. 23:651–82 www.annualreviews.org • MHC/TCR Interactions

461

ARI

15 February 2006

1:1

78. Melief CJ. 2005. Escort service for cross-priming. Nat. Immun. 6:543–44 79. Norbury CC, Basta S, Donohue KB, Tscharke DC, Princiotta MF, et al. 2004. CD8+ T cell cross-priming via transfer of proteasome substrates. Science 304:1318–21 80. Connolly ML. 1993. The molecular surface package. J. Mol. Graph. 11:139–41 81. Wang JH, Reinherz EL. 2002. Structural basis of T cell recognition of peptides bound to MHC molecules. Mol. Immunol. 38:1039–49 82. Hulsmeyer M, Chames P, Hillig RC, Stanﬁeld RL, Held G, et al. 2005. A major histocompatibility complex-peptide-restricted antibody and T cell receptor molecules recognize their target by distinct binding modes: crystal structure of human leukocyte antigen (HLA)-A1-MAGE-A1 in complex with FAB-HYB3. J. Biol. Chem. 280:2972–80 83. Sim BC, Travers PJ, Gascoigne NR. 1997. Vα3.2 selection in MHC class I mutant mice: evidence for an alternate orientation of TCR-MHC class I interaction. J. Immunol. 159:3322–29 84. Gil D, Schamel WW, Montoya M, Sanchez-Madrid F, Alarcon B. 2002. Recruitment of Nck by CD3ε reveals a ligand-induced conformational change essential for T cell receptor signaling and synapse formation. Cell 109:901–12 85. Hennecke J, Wiley DC. 2001. T cell receptor-MHC interactions up close. Cell 104: 1–4 86. Speir JA, Garcia KC, Brunmark A, Degano M, Peterson PA, et al. 1998. Structural basis of 2C TCR allorecognition of H-2Ld peptide complexes. Immunity 8:553–62 87. Guimezanes A, Barrett-Wilt GA, Gulden-Thompson P, Shabanowitz J, Engelhard VH, et al. 2001. Identiﬁcation of endogenous peptides recognized by in vivo or in vitro generated alloreactive cytotoxic T lymphocytes: distinct characteristics correlated with CD8 dependence. Eur. J. Immunol. 31:421–32 88. Rudolph MG, Speir JA, Brunmark A, Mattsson N, Jackson MR, et al. 2001. The crystal structures of Kbm1 and Kbm8 reveal that subtle changes in the peptide environment impact thermostability and alloreactivity. Immunity 14:231–42 89. Kjer-Nielsen L, Dunstone MA, Kostenko L, Ely LK, Beddoe T, et al. 2004. Crystal structure of the human T cell receptor CD3εγ heterodimer complexed to the therapeutic mAb OKT3. Proc. Natl. Acad. Sci. USA 101:7675–80 90. Argaet VP, Schmidt CW, Burrows SR, Silins SL, Kurilla MG, et al. 1994. Dominant selection of an invariant T cell antigen receptor in response to persistent infection by Epstein-Barr virus. J. Exp. Med. 180:2335–40 91. Al-Lazikani B, Lesk AM, Chothia C. 2000. Canonical structures for the hypervariable regions of T cell αβ receptors. J. Mol. Biol. 295:979–95 92. Sherman LA, Chattopadhyay S. 1993. The molecular basis of allorecognition. Annu. Rev. Immunol. 11:385–402 93. Moss DJ, Burrows SR, Silins SL, Misko I, Khanna R. 2001. The immunology of EpsteinBarr virus infection. Philos. Trans. R. Soc. London Ser. B. 356:475–88 94. Bhat TN, Bentley GA, Boulot G, Greene MI, Tello D, et al. 1994. Bound water molecules and conformational stabilization help mediate an antigen-antibody association. Proc. Natl. Acad. Sci. USA 91:1089–93 95. Malby RL, Tulip WR, Harley VR, McKimm-Breschkin JL, Laver WG, et al. 1994. The structure of a complex between the NC10 antibody and inﬂuenza virus neuraminidase and comparison with the overlapping binding site of the NC41 antibody. Structure 2:733– 46 96. Moss DJ, Burrows SR, Baxter GD, Lavin MF. 1991. T cell-T cell killing is induced by speciﬁc epitopes: evidence for an apoptotic mechanism. J. Exp. Med. 173:681–86

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

462

Rudolph

·

Stanfield

·

Wilson

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

ARI

15 February 2006

1:1

97. Lehner PJ, Wang EC, Moss PA, Williams S, Platt K, et al. 1995. Human HLA-A0201restricted cytotoxic T lymphocyte recognition of inﬂuenza A is dominated by T cells bearing the Vβ17 gene segment. J. Exp. Med. 181:79–91 98. Lawson TM, Man S, Williams S, Boon AC, Zambon M, Borysiewicz LK. 2001. Inﬂuenza A antigen exposure selects dominant Vβ17+ TCR in human CD8+ cytotoxic T cell responses. Int. Immunol. 13:1373–81 99. Arzt S, Petit I, Burmeister WP, Ruigrok RW, Baudin F. 2004. Structure of a knockout mutant of inﬂuenza virus M1 protein that has altered activities in membrane binding, oligomerisation and binding to NEP (NS2). Virus Res. 99:115–19 100. Wilson IA, Stanﬁeld RL. 2005. MHC restriction: slip-sliding away. Nat. Immun. 6:434– 35 101. He XL, Radu C, Sidney J, Sette A, Ward ES, Garcia KC. 2002. Structural snapshot of aberrant antigen presentation linked to autoimmunity: the immunodominant epitope of MBP complexed with I-Au . Immunity 17:83–94 102. Adams EJ, Chien YH, Garcia KC. 2005. Structure of a γδ T cell receptor in complex with the nonclassical MHC T22. Science 308:227–31 103. Wilson IA, Stanﬁeld RL. 2001. Unraveling the mysteries of γδ T cell recognition. Nat. Immun. 2:579–81 104. Reich Z, Boniface JJ, Lyons DS, Borochov N, Wachtel EJ, Davis MM. 1997. Ligandspeciﬁc oligomerization of T-cell receptor molecules. Nature 387:617–20 105. Li P, Morris DL, Willcox BE, Steinle A, Spies T, Strong RK. 2001. Complex structure of the activating immunoreceptor NKG2D and its MHC class I-like ligand MICA. Nat. Immun. 2:443–51 106. Gao GF, Tormo J, Gerth UC, Wyer JR, McMichael AJ, Stuart DI, et al. 1997. Crystal structure of the complex between human CD8αα and HLA-A2. Nature 387:630–34 107. Kern PS, Teng MK, Smolyar A, Liu JH, Liu J, et al. 1998. Structural basis of CD8 coreceptor function revealed by crystallographic analysis of a murine CD8αα ectodomain fragment in complex with H-2Kb . Immunity 9:519–30 108. Gangadharan D, Cheroutre H. 2004. The CD8 isoform CD8αα is not a functional homologue of the TCR co-receptor CD8αβ. Curr. Opin. Immunol. 16:264–70 109. Wang J, Meijers R, Xiong Y, Liu J, Sakihama T, et al. 2001. Crystal structure of the human CD4 N-terminal two domain fragment complexed to a class II MHC molecule. Proc. Natl. Acad. Sci. USA 98:10799–804 110. Wu H, Kwong PD, Hendrickson WA. 1997. Dimeric association and segmental variability in the structure of human CD4. Nature 387:527–30 111. Manolios N, Letourneur F, Bonifacino JS, Klausner RD. 1991. Pairwise, cooperative and inhibitory interactions describe the assembly and probable structure of the T-cell antigen receptor. EMBO J. 10:1643–51 112. Koning F, Maloy WL, Coligan JE. 1990. The implications of subunit interactions for the structure of the T cell receptor-CD3 complex. Eur. J. Immunol. 20:299–305 113. Grakoui A, Bromley SK, Sumen C, Davis MM, Shaw AS, et al. 1999. Immunological synapse: A molecular machine controlling T cell activation. Science 285:221–27 114. Davis SJ, van der Merwe PA. 2001. The immunological synapse: required for T cell receptor signalling or directing T cell effector function? Curr. Biol. 11:R289–91 115. Davis SJ, van der Merwe PA. 2003. TCR triggering: co-receptor-dependent or -independent? Trends Immunol. 24:624–26 116. Bromley SK, Burack WR, Johnson KG, Somersalo K, Sims TN, et al. 2001. The immunological synapse. Annu. Rev. Immunol. 19:375–96 www.annualreviews.org • MHC/TCR Interactions

463

ARI

15 February 2006

1:1

117. Berkhout B, Alarcon B, Terhorst C. 1988. Transfection of genes encoding the T cell receptor-associated CD3 complex into COS cells results in assembly of the macromolecular structure. J. Biol. Chem. 263:8528–36 118. Wang N, Wang B, Salio M, Allen D, She J, Terhorst C. 1998. Expression of a CD3ε transgene in CD3ε(null) mice does not restore CD3γ and δ expression but efﬁciently rescues T cell development from a subpopulation of prothymocytes. Int. Immunol. 10:1777–88 119. Kappes DJ, Alarcon B, Regueiro JR. 1995. T lymphocyte receptor deﬁciencies. Curr. Opin. Immunol. 7:441–47 120. Kim KS, Sun ZY, Wagner G, Reinherz EL. 2000. Heterodimeric CD3εγ extracellular domain fragments: production, puriﬁcation and structural analysis. J. Mol. Biol. 302:899– 916 121. Call ME, Pyrdol J, Wiedmann M, Wucherpfennig KW. 2002. The organizing principle in the formation of the T cell receptor-CD3 complex. Cell 111:967–79 122. Sun ZJ, Kim KS, Wagner G, Reinherz EL. 2001. Mechanisms contributing to T cell receptor signaling and assembly revealed by the solution structure of an ectodomain fragment of the CD3εγ heterodimer. Cell 105:913–23 123. Manolios N, Kemp O, Li ZG. 1994. The T cell antigen receptor α and β chains interact via distinct regions with CD3 chains. Eur. J. Immunol. 24:84–92 124. Manolios N, Li ZG. 1995. The T cell antigen receptor β chain interacts with the extracellular domain of CD3-γ. Immunol. Cell Biol. 73:532–36 125. Arnett KL, Harrison SC, Wiley DC. 2004. Crystal structure of a human CD3-ε/δ dimer in complex with a UCHT1 single-chain antibody fragment. Proc. Natl. Acad. Sci. USA 101:16268–73 126. Call ME, Wucherpfennig KW. 2005. The T cell receptor: critical role of the membrane environment in receptor assembly and function. Annu. Rev. Immunol. 23:101–25 127. Call ME, Wucherpfennig KW. 2004. Molecular mechanisms for the assembly of the T cell receptor-CD3 complex. Mol. Immunol. 40:1295–305 128. Call ME, Pyrdol J, Wucherpfennig KW. 2004. Stoichiometry of the T-cell receptorCD3 complex and key intermediates assembled in the endoplasmic reticulum. EMBO J. 23:2348–57 129. Engelman DM. 2003. Electrostatic fasteners hold the T cell receptor-CD3 complex together. Mol. Cell 11:5–6 130. Backstrom BT, Muller U, Hausmann B, Palmer E. 1998. Positive selection through a motif in the αβ T cell receptor. Science 281:835–38 131. Ghendler Y, Smolyar A, Chang HC, Reinherz EL. 1998. One of the CD3epsilon subunits within a T cell receptor complex lies in close proximity to the Cβ FG loop. J. Exp. Med. 187:1529–36 132. Rudd PM, Wormald MR, Stanﬁeld RL, Huang M, Mattsson N, et al. 1999. Roles for glycosylation of cell surface receptors involved in cellular immune recognition. J. Mol. Biol. 293:351–66 133. Choudhuri K, Wiseman D, Brown MH, Gould K, van der Merwe PA. 2005. T-cell receptor triggering is critically dependent on the dimensions of its peptide-MHC ligand. Nature 436:578–82 134. Margulies DH. 1997. Interactions of TCRs with MHC-peptide complexes: a quantitative basis for mechanistic models. Curr. Opin. Immunol. 9:390–95 135. Alam SM, Davies GM, Lin CM, Zal T, Nasholds W, et al. 1999. Qualitative and quantitative differences in T cell receptor binding of agonist and antagonist ligands. Immunity 10:227–37

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

464

Rudolph

·

Stanfield

·

Wilson

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-14-III

ARI

15 February 2006

1:1

136. Alam SM, Travers PJ, Wung JL, Nasholds W, Redpath S, et al. 1996. T-cell-receptor afﬁnity and thymocyte positive selection. Nature 381:616–20 137. Lyons DS, Lieberman SA, Hampl J, Boniface JJ, Chien Y, et al. 1996. A TCR binds to antagonist ligands with lower afﬁnities and faster dissociation rates than to agonists. Immunity 5:53–61 138. Boniface JJ, Reich Z, Lyons DS, Davis MM. 1999. Thermodynamics of T cell receptor binding to peptide-MHC: evidence for a general mechanism of molecular scanning. Proc. Natl. Acad. Sci. USA 96:11446–51 139. Garcia KC, Tallquist MD, Pease LR, Brunmark A, Scott CA, et al. 1997. αβ T-cell receptor interactions with syngeneic and allogeneic ligands: afﬁnity measurements and crystallization. Proc. Natl. Acad. Sci. USA 94:13838–43 140. Willcox BE, Gao GF, Wyer JR, Ladbury JE, Bell JI, et al. 1999. TCR binding to peptideMHC stabilizes a ﬂexible recognition interface. Immunity 10:357–65 141. Wu LC, Tuot DS, Lyons DS, Garcia KC, Davis MM. 2002. Two-step binding mechanism for T-cell receptor recognition of peptide MHC. Nature 418:552–56 142. Baker BM, Turner RV, Gagnon SJ, Wiley DC, Biddison WE. 2001. Identiﬁcation of a crucial energetic footprint on the α1 helix of human histocompatibility leukocyte antigen (HLA)-A2 that provides functional interactions for recognition by tax peptide/HLA-A2speciﬁc T cell receptors. J. Exp. Med. 193:551–62 143. Baker BM, Ding YH, Garboczi DN, Biddison WE, Wiley DC. 1999. Structural, biochemical, and biophysical studies of HLA-A2/altered peptide ligands binding to viralpeptide-speciﬁc human T-cell receptors. Cold Spring Harbor Symp. Quant. Biol. 64:235– 41 144. Wilson IA. 1999. Class-conscious TCR? Science 286:1867–68 145. Chlewicki LK, Holler PD, Monti BC, Clutter MR, Kranz DM. 2005. High-afﬁnity, peptide-speciﬁc T cell receptors can be generated by mutations in CDR1, CDR2 or CDR3. J. Mol. Biol. 346:223–39 146. McCoy AJ, Chandana-Epa V, Colman PM. 1997. Electrostatic complementarity at protein/protein interfaces. J. Mol. Biol. 268:570–84 147. Rudd PM, Elliott T, Cresswell P, Wilson IA, Dwek RA. 2001. Glycosylation and the immune system. Science 291:2370–76 148. Mitra AK, Celia H, Ren G, Luz JG, Wilson IA, Teyton L. 2004. Supine orientation of a murine MHC class I molecule on the membrane bilayer. Curr. Biol. 14:718– 24 149. Viola A. 2001. The ampliﬁcation of TCR signaling by dynamic membrane microdomains. Trends Immunol. 22:322–27 150. Vogt AB, Spindeldreher S, Kropshofer H. 2002. Clustering of MHC-peptide complexes prior to their engagement in the immunological synapse: lipid raft and tetraspan microdomains. Immunol. Rev. 189:136–51 151. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, et al. 2000. The Protein Data Bank. Nucleic Acids Res. 28:235–42 152. Bjorkman PJ, Saper MA, Samraoui B, Bennett WS, Strominger JL, Wiley DC. 1987. The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature 329:512–18 153. Hennecke J, Wiley DC. 2002. Structure of a complex of the human α/β T cell receptor (TCR) HA1.7, inﬂuenza hemagglutinin peptide, and major histocompatibility complex class II molecule, HLA-DR4 (DRA∗ 0101 and DRB1∗ 0401): insight into TCR crossrestriction and alloreactivity. J. Exp. Med. 195:571–81 www.annualreviews.org • MHC/TCR Interactions

465

ANRV270-IY24-14-III

ARI

15 February 2006

1:1

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

154. CCP4. 1994. The Collaborative Computational Project Number 4, suite programs for protein crystallography. Acta Cryst. D50:760–63 155. McDonald IK, Thornton JM. 1994. Satisfying hydrogen bonding potential in proteins. J. Mol. Biol. 238:777–93 156. Sheriff S, Hendrickson WA, Smith JL. 1987. Structure of myohemerythrin in the azidomet state at 1.7/1.3 A˚ resolution. J. Mol. Biol. 197:273–96

466

Rudolph

·

Stanfield

·

Wilson

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:419-466. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin1 and Andrew C. Chan1,2 1

Department of Immunology, 2 Department of Antibody Engineering, Genentech, Inc., South San Francisco, California 94080; email: ﬂ[email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:467–96 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090517 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0467$20.00

Key Words CD20, Rituxan, BAFF/BLyS, BR3

Abstract The pathogenic roles of B cells in autoimmune diseases occur through several mechanistic pathways that include autoantibodies, immune complexes, dendritic and T cell activation, cytokine synthesis, chemokine-mediated functions, and ectopic neolymphogenesis. Each of these pathways participate to different degrees in autoimmune diseases. The use of B cell–targeted and B cell subset–targeted therapies in humans is illuminating the mechanisms at work in a variety of human autoimmune diseases. In this review, we highlight some of these recent ﬁndings that provide insights into both murine models of autoimmunity and human autoimmune diseases.

467

ANRV270-IY24-15

ARI

15 February 2006

1:8

ROLES OF B CELLS IN AUTOIMMUNITY

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Catalyzed by the recent clinical success of B cell modulatory therapies in the treatment of B cell cancers and autoimmune diseases, we are witnessing a resurgence in exploring the mechanisms by which these B lineage cells participate in the induction and maintenance of human autoimmune diseases. These include the ability of B cells to produce self-reactive antibodies, secrete inﬂammatory cytokines, participate in antigen presentation, augment T cell activation, and generate ectopic lymphogenesis (1, 2). In this review, we examine the emerging immunologic concepts from a number of clinical trials that used B cell–targeted therapies. These data represent merely a small percentage of the wealth of biology yet to be learned regarding human immunology.

Autoantibodies and Immune Complexes Historically the role of B lymphocytes in autoimmunity has been associated largely with the capacity of plasmablasts and plasma cells, the end products of the B cell differentiation pathway, to produce self-damaging antibodies. Fulﬁllment of “Koch’s postulates” by recapitulation of pathology through transfer of pathogenic antibodies has been documented in multiple human autoimmune diseases. These autoreactive antibodies bind self-antigens and can interfere with normal cellular functions as well as harness immune effector mechanisms to generate autoimmune pathology (Figure 1). As an example of the former, antibodies to the thyroid-stimulating hormone (TSH) receptor mimic TSH activation but are not subject to the negative feedback loop of endogenous TSH. Because these antibodies are not subject to the negative feedback loop, the result is prolonged thyroid hormone (thyroxine and triiodothyronine) production and hyperthyroidism. 468

Martin

·

Chan

Autoantibodies also induce disease through Fc-mediated activation of the complement system. Although autoantibodies to acetylcholine receptors may, in part, account for the loss of acetylcholine receptors from the neuromuscular junctions in myasthenia gravis patients, complement activation and deposition of the membrane attack complex likely account for the major pathogenic mechanisms involved in disruption and loss of post-synaptic folds manifested by altered neuromuscular transmission in myasthenia gravis patients (3). Similarly, in anti–glomerular basement membrane (GBM) syndrome (also known as Goodpasture’s syndrome), deposition of autoantibodies against basement membrane collagen type IV in the kidney and lung alveoli induces glomerulonephritis and necrotizing hemorrhagic interstitial pneumonitis, respectively, through complement activation and further recruitment of inﬂammatory cells. Finally, pathogenic antibodies can also mediate end-organ damage through the formation of immune complexes. As a sequelae of hepatitis C infection, complement-activating antibodies directed against hepatitis C viral proteins can deposit in the skin, kidney, and peripheral nerves to induce cutaneous purpuric lesions, glomerulonephritis, and mononeuritis multiplex, respectively (Figure 1). Similarly, autoantibodies in the form of anti-dsDNA immune complexes can activate complement to mediate autoimmune nephritis, a major component in lupus nephritis disease pathogenesis. In addition to complement activation, autoantibodies and immune complexes also activate Fc receptors (FcRs) expressed on both myeloid and lymphoid cells. Monocytes, macrophages, dendritic cells (DCs), and neutrophils express CD64, CD32A, and CD16, activating FcRs. Autoimmune-prone mice lacking these activating FcγRs have attenuated forms of autoimmune disorders and demonstrate that FcγRs’ roles in disease are clearly not redundant. In a murine model of inﬂammatory arthritis, CD16 plays a key

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

Figure 1 Antibody-driven pathogenic mechanisms of human B cell autoimmunity. Classic antibody-dependent mechanisms are responsible either directly (top) or through immune complexes (bottom) for the pathogenesis of speciﬁc autoimmune diseases. (Abbreviations: SLE, systemic lupus erythematosus; ANCA, antineutrophil cytoplasmic antibody–associated vasculitis; AIHA, autoimmune hemolytic anemia; ITP, immune-mediated thrombocytopenia.)

role in the initiation phase of disease, whereas CD64 appears more important in the later phases of cartilage destruction (4). In contrast, in murine models of experimental allergic encephalitis (EAE) and autoimmune glomerulonephritis, CD64 appears dispensable for disease, whereas CD16 is paramount to dis-

ease initiation and maintenance (5, 6). The signals generated by these activating receptors are counterbalanced by the FcγRIIb inhibitory receptor. Mice deﬁcient in FcγRIIb develop spontaneous lupus-like autoimmunity, autoimmune models of lupus express lower levels of FcγRIIb on germinal center www.annualreviews.org • B Cell Depletion

469

ANRV270-IY24-15

ARI

15 February 2006

GC: germinal center RF: rheumatoid factor RA: rheumatoid arthritis

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ASC: antibody-secreting cell

470

1:8

(GC) B cells, and forced expression of FcγRIIb in these lupus-prone mice reverses autoimmunity (7–12). Hence, the balance of Fc-derived signals activated through the innate immune efforts is important in deﬁning the immunopathology of disease. Immune complexes can also modulate DC function through interaction and activation of Fcγ, Fcα, and Fcε receptors. Immune complex uptake and activation of DCs through FcRs have been demonstrated to augment antigen presentation in both human and mouse DC populations. In turn, crosspresentation of antigen can promote T cell activation at >1000-fold lower concentration of antigen. In a variety of autoimmune disorders, both antigen and antibody appear to synergize to activate and mediate pathogenic processes. For example, chromatin-containing immune complexes, present in autoimmune disorders such as systemic lupus erythematosus (SLE), drive FcγRIII and Toll-like receptor (TLR)9-dependent and -independent pathways on DCs (13). Interestingly, these costimuli selectively induce BAFF expression, a tumor necrosis factor (TNF) superfamily survival factor implicated in the pathogenesis of SLE (discussed in more detail below). Additionally, DNA- and RNA-containing immune complexes, also present in SLE, activate human plasmacytoid DCs through FcRs to stimulate production of IFN-α, a key cytokine implicated in the pathogenesis of human SLE (14, 15). Hence, both antigen and antibody within immune complexes can activate cellular mechanisms to mediate autoimmune disorders. Although the above examples provide clear evidence of pathogenicity of antibodies, most autoantibodies described to date have not fulﬁlled Koch’s postulates and are less likely to play a direct pathogenic role in disease (16). However, detection of autoantibodies can aid in disease diagnosis [e.g., rheumatoid factor (RF) for rheumatoid arthritis (RA) and antinuclear antibodies (ANA) for SLE] and can potentially act as surrogate markers of disease activity (e.g., anti-dsDNA antibodMartin

·

Chan

ies in SLE). Nonetheless, evolving technologies to measure autoantibodies against the proteome and greater understanding of the underlying biology of B cell autoimmunity will likely lend greater signiﬁcance to these autoantibody markers in the understanding of B cell functions in disease initiation and progression.

Origins of Autoantibodies Investigators have long appreciated that normal antibody immune responses can have one of two outcomes. In the ﬁrst scenario, antigen-speciﬁc antibodies are gradually lost after immune challenge, and upon antigen reencounter there is little memory of the ﬁrst exposure, suggesting that the plasma cells secreting these antibodies are short lived and vanish soon after the initiation of the immune response. Most of these antigens belong to the T-independent class, and reexposure is necessary for continuous immunity. In the second scenario, T-dependent generation of antibodies persists long after the antigen is cleared, implicating the existence of a long-lived source of antibody-secreting cells (ASCs). Although the distinctive rules and developmental signals for each of these normal immune response scenarios are not entirely clear, investigators have even less understanding of cells involved in producing autoantibodies in human autoimmune diseases. Recent insights into the development of ASCs— plasmablasts and plasma cells in both mice and humans—indicate that they are heterogeneous in their phenotype, location, and life span (17–20). Targeting of a green ﬂuorescent protein cassette into the mouse blimp-1 locus, a transcription factor essential for plasma cell differentiation, enabled real-time tracking of cycling plasmablasts at sites of immune response, as well as of their short- and longlived cellular progeny (17). During primary and secondary immune responses in mice and humans, plasmablasts are generated in secondary immune organs in both extrafollicular and follicular sites. Investigators believe

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

that most long-lived plasma cells are derived from post-GC plasmablasts, whereas shortlived plasma cells can emerge from both extrafollicular and GC sites. In extrafollicular sites from the spleen and lymph nodes and potentially in lymphogenic structures in various autoimmune diseases, such as RA synovium, short-lived plasmablasts differentiate quickly into short-lived plasma cells with a limited life span of a few days. In contrast, post-GC plasmablasts from secondary lymphoid organs appear transiently in blood and trafﬁc toward the bone marrow, where, if they ﬁnd favorable (but yet uncharacterized) microenvironments, they can then develop into long-lived plasma cells. Although the majority of autoimmune ASCs secrete high-afﬁnity somatically mutated antibodies, it is unclear whether they are derived solely from the classical passage through GCs or through the recently described extrafollicular process (21). Studies in autoimmune-prone lpr/lpr mice indicate that the somatically mutated sequences represented in the ASC compartment, appear to originate from the GC compartment (22, 23). Recently, however, this concept has been challenged for autoimmune cells by the description of somatically mutated RF-generating cells emerging from sites outside the GC (21, 24). The origins of autoantibodies are critical for the interpretation of existing and emerging clinical data as well as for the development of future therapeutic approaches. For example, autoantibodies may be products of a GC-like reaction in the RA synovium. These synovial GC-like structures will generate plasmablasts that ultimately can differentiate locally into short-lived plasma cells as well as long-lived plasma cells that might be sustained if they reach an appropriate bone marrow environment. Alternatively, through an extrafollicular process in the synovium, but outside of the organized follicular structures, RF might be generated through a GCand T cell–independent fashion using a completely different set of molecular cues. This extrafollicular process, in turn, would avoid

the censorship and checkpoints speciﬁc for preventing autoimmunity in GCs. Because only a minority of RA and Sjogren’s syndrome specimens from the synovium (23%) or salivary glands (17%) contain GC-like structures, these structures may not be the only sites that could enable local propagation of autoimmunity (25). In a transgenic, autoimmune-prone MRL/lpr mouse model that expresses a RF heavy chain, generation of high-afﬁnity RF occurs independently of GCs (26). Rather, formation of high-afﬁnity, somatically hypermutated RF occurs predominantly through the generation of short-lived plasmablasts and, in this model, is limited to the spleen and absent in the lymph nodes and bone marrow. Although contributions of extra-GC generation of autoantibodies in nontransgenic systems needs to be further assessed, these observations are in concordance with the current clinical experiences in RA, factor VIII (FVIII) deﬁciency, and subsets of immune-mediated thrombocytopenia (ITP) patients where RF, anti-FVIII, and antithrombocyte antibodies decrease post-B cell–depleting therapy (see below), suggesting that they originate from short-lived plasmablast or plasma cell compartments and, once exhausted, will not be replenished due to the lack of immediate B cell progenitors.

B Cell–T Cell Interactions Although causality related to autoantibodies and pathology has been demonstrated only in a minority of autoimmune diseases, most complex autoimmunity involves additional B cell functions. These include the ability of B cells and their secreted products (in addition to antibodies and immune complexes) to modulate T cells and DCs through antigen presentation and costimulation (Figure 2). The ﬁrst direct experimental demonstration of such function was provided by the generation of autoimmune-prone MRL/lpr mice with compromised B cell functions. Consistent with a requisite role for B cells in the development of autoimmune and lymphoproliferative www.annualreviews.org • B Cell Depletion

471

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

1:8

Figure 2 Antibody-independent mechanisms of B cell autoimmunity. Several other mechanisms, including antigen presentation (APC, left), costimulatory functions (surface molecules and secreted cytokines, left), and the ability to support de novo lymphoid tissue organization (neolymphogenesis, right) may explain other parts of B cell involvement in complex autoimmune conditions.

disease, MLR/lpr mice rendered completely B cell deﬁcient (μMT MRL/lpr) exhibited minimal production of autoantibodies, lymphoproliferation, or glomerulonephritis and exhibited substantial improvement in mortality (27, 28). Surprisingly, mIgM MRL/lpr mice that express only the membrane form of IgM and are incapable of antibody or autoantibody secretion developed substantially less glomerulonephritis or mortality compared with wild-type MLR/lpr littermates (29). Similar results were seen in the mIgM nonobese diabetic (NOD) mouse model of diabetes, for which insulitis and overall incidence of diabetes was lower than in NOD nontransgenic littermates but was higher than in NOD mice completely lacking B cells (μMT NOD) (30). Conversely, expression of a germ line immunoglobulin (IgH) transgene in the NZB/W SLE-prone mouse strain reduced antibody and B cell antigen recep472

Martin

·

Chan

tor (BCR) repertoire diversity, while still allowing clonal anti-dsDNA antibody generation reduced the incidence and severity of glomerulonephritis and mortality compared with NZB/W littermates (31). These data indicate that antibody-independent mechanisms play important pathogenic roles in autoimmune disorders and may involve important regulatory roles of B cells in cytokine synthesis and secretion as well as in directing T cell and DC functions. In contrast to MLR/lpr mice, the μMT MLR/lpr mouse did not develop autoantibodies or immune complex–mediated diseases; it also exhibited a massive decrease in CD4 and CD8 T cell activation and lymphocytic inﬁltration in end organs (28, 32). In addition, spontaneous accumulation of memory CD8+ T cells was inhibited by ∼10-fold in the absence of B cells. In principle, B cells could be activating CD8+ T cells either

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

directly via crosspresentation or indirectly via CD4+ T cell activation. By mixing bone marrow derived from JH−/− MLR/lpr and β2m−/− MRL/lpr mice, investigators tested chimeric mice in which only B cells lack β2m and hence lack MHC class I expression, yet are still able to activate CD4 and CD8 T cells (33). Hence, B cell–directed T cell activation can occur independently of antigen crosspresentation (self-antigen presented through MHC class I on B cells to CD8 T cells). Finally, in a system of redirected antigen presentation, using the proteoglycan-induced arthritis model in mIgM MRL/lpr mice antigen presentation by B cells is required for T cell activation (34). Development of severe inﬂammatory disease required the presence of both pathogenic antibodies and activated T cells. Together, these data indicate a signiﬁcant role for B cell–directed activation of T cells in a variety of autoimmunity models. Additional evidence indicating an important role for B cells in directing T cell function is provided by experiments involving CD40/CD40L blockade. Blockade of this TNF superfamily member through antiCD40L monoclonal antibodies (mAbs) has been demonstrated to be beneﬁcial in several mouse models of SLE, EAE, inﬂammatory arthritis, and colitis (35, 36). As predicted, anti-CD40L blocks Ig class switching and somatic mutation, reduces anti-dsDNA IgG antibodies and B cell inﬁltration in target organs, and induces a general B cell unresponsive state. Consistent with the ability of B cells to modulate T cell functions, anti-CD40L treatment also induces a period of T cell unresponsiveness. In addition to the role of immune complexes in DC maturation and APC functions of DCs (above), B cells also secrete cytokines and chemokines, including IL-16, MIP1α, and MIP1β, that can modulate DC migration and function (37, 38). Moreover, primed DCs from B cell–deﬁcient mice produce higher levels of IL-12 and impaired differentiation of IL-4-secreting T cells (39). Hence, B cells can also modulate DC mat-

uration, migration, and function. Together, these data demonstrate the existence of multiple mechanisms by which B lymphocytes can modulate both T and DC functions.

MS: multiple sclerosis

B Cells and Neolymphogenesis Dissection of attraction, repulsion, and retention cues have demonstrated that lymphoid organs and structures are formed based on signaling networks of TNF/TNFR family molecules, integrins, and chemokines (40– 43). These studies have also revealed that an important function for B cells in both the formation and maintenance of new lymphoid foci since blockade of select TNF/TNFR family molecules, integrins, and chemokine signaling pathways results in dissolution of established lymphoid structures (Figure 2). This requisite for B cells in lymphogenesis has clear bearing for human pathology, as lymphoid-like follicles have been described in the synovium in RA patients, inﬂamed salivary glands in Sjogren’s syndrome patients, the ventricular-meningeal compartment in multiple sclerosis (MS) patients, thyroid lobes in autoimmune thyroiditis patients, and kidneys in lupus nephritis patients (44 –53). In RA, three distinct histological presentations have been described. The ﬁrst type has all the attributes of a lymphoid structure indistinguishable from lymph nodes, containing follicular DCs in the GC and B cell zones together with interdigitating DCs in T cell zones. The second has only T cells, B cells, and DCs in a semi-organized fashion. The third type of histology is characterized by isolated T and B cells without typical DCs in their proximity. As the ﬁrst two types of neolymphogenic structures do not appear in all patients, these structures are likely not requisite for all disease pathogenesis. However, given the heterogeneity of autoimmune disorders and the ectopic presence of neolymphogenic structures in the diseased target organs, these structures likely play an important role in disease pathogenesis in subsets of patients. Within these structures, B cells may provide a local www.annualreviews.org • B Cell Depletion

473

ANRV270-IY24-15

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

BCR: B cell antigen receptor

1:8

catalyzing context for the generation of inﬂammatory signals and for activating T cells and DCs, contributing to inﬂammatory disease. The requirement for such structures has been suggested by studies in which human RA synovium is implanted subcutaneously into scid mice. Within this model, interruption of T-B costimulatory signals as well as depletion of B cells abolished ectopic follicle maintenance as well as T and myeloid cell activation (54). These evolving data suggest an important role for B cells and ectopic neolymphogenesis in autoimmunity and could serve as a therapeutic mechanism by which B cell depletion or modulation could improve disease.

B CELL DEPLETION AND INHIBITION Following the discovery of the B cell/T cell dichotomy in birds and rodents, most efforts to manipulate B lineage cells have been based on the ability to deplete or inhibit these cells or subsets to assess the function of different compartments during immunity and autoimmunity. Three main strategies have provided the mainstay approaches: (a) the use of depleting antibodies against B cell surface proteins, (b) generation of fusion proteins or antibodies that block B cell survival signals, and (c) generation of antibodies that activate proapoptotic signals (e.g., the surface Ig BCR in immature B cell stages). The ﬁrst tools to be introduced for B cell depletion were polyclonal antibodies (typically produced in goat, sheep, or rabbit) directed against the IgM surface immunoglobulin receptor. With the recognition that naked mAbs against cell surface glycoproteins can be used successfully to deplete speciﬁc cell lineages and subsets, several molecules have been used in both preclinical models and clinical trials to attempt pathogenic B cell elimination. Below, we review a variety of efforts but focus mainly on reagents against the CD20 antigen that has accumulated the greatest clinical experience.

474

Martin

·

Chan

B Cell–Depleting Therapies—B Cell Antigen Receptor (BCR) Initial work with anti-Ig reagents demonstrated that pre- and perinatal treatment of mice with anti-Ig polyclonal antibodies successfully depleted surface-Ig positive B lineage cells (55). The success of this strategy can be explained by the exquisite sensitivity to cell death induced by anti-Ig reagents in newly formed immature B cells that are just being released from the primary lymphoid generation site (neonatal liver, spleen, and bone marrow). Although this strategy is less applicable to adult animals with high levels of circulating Ig, which sequesters the depleting antiIg reagents, a number of protocols have been tried with varying degrees of success. An insightful protocol used mAbs directed to surface IgD, thus bypassing the high levels of serum IgM; this technique was successful in inducing decreases in circulating B cells and supports the notion that sIg crosslinking in the absence of coreceptor engagement can induce B cell apoptosis in mature B cells (56– 58). Although successful, this strategy is likely to have limited therapeutic use because most pathogenic cells have switched isotypes and do not express surface IgD. Regardless of these limitations, the anti-Ig therapies can deplete B cells and provide evidence of an important role for B cells in the development of autoimmunity in murine models of SLE (59–61). To avoid the sequestration of the therapeutic antibodies, alternative approaches in targeting the surface BCR have been directed toward the BCR-associated transmembrane signaling components, CD79a and b. Initial experiments in mice demonstrate that these antibodies are indeed agonistic and, similar to anti-CD3 for T cell development in the thymus, promote development of proB to pre-B cells in the bone marrow (62, 63). However, development of antihuman reagents have lagged, in part owing to the lower degree of BCR crosslinking compared with polyclonal anti-Ig reagents, which results in a much less efﬁcient ability to induce B cell

ANRV270-IY24-15

ARI

15 February 2006

1:8

apoptosis and in only minimal B cell depletion (64).

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell–Depleting Therapies—CD20 Antibodies against the human CD20 molecule [Rituxan (Rituximab), Genentech/BiogenIDEC; MabThera , Roche] represent the most advanced among the B cell–depleting armamentarium in autoimmune disorders. CD20, a four-transmembrane protein, is expressed at low levels on late pre-B cells, upregulated to ∼90,000 copies/cell on most normal and malignant B cells, with its expression extinguished on plasma cells (65, 66). Anti-CD20 mAbs successfully deplete peripheral human B lymphocytes for periods ranging from three months to more than one year through mechanisms involving Fc- and complement-dependent killing as well as possible proapoptotic and other signals. Recent in vivo experiments using tumor xenografts as well as a panel of antimouse and antihuman CD20 antibodies [the latter in human CD20 BAC (bacterial artiﬁcial chromosome) transgenic mice] demonstrate a dominant role for FcRs in anti-CD20 mAb immunotherapy (67–70). The kinetics of B cell depletion vary between different B cell compartments. As shown in studies of a hCD20 BAC transgenic mouse, depletion occurs rapidly for circulating blood cells (reaching >90% depletion in minutes), slower for lymph node and splenic B cells (∼60% to 70% at day 1), and slowest for B cells residing within the peritoneal cavity (signiﬁcant depletion only after day 7). Analysis of splenic B cell subsets reveals that distinct microenvironments exhibit different sensitivities to B cell depletion. Whereas >90% follicular (FO) B cells in the spleen are depleted within two days of anti-CD20 mAb administration, marginal zone (MZ) B cells are depleted to only ∼25%–50% of basal levels. In addition to the relative resistance of the MZ B cell compartment, the peritoneal cavity (particularly B1) and the GC B cell compartments demonstrate the greatest relative

resistance to anti-CD20 mAb treatment. Resistance of these compartments is not due to lack of CD20 expression, as MZ, GC, and peritoneal B1 B cells express comparable or higher levels of CD20 than do the relatively sensitive FO B cells. Moreover, these differential sensitivities are not due to differences in drug bioavailability, as all CD20 molecules on these more resistant populations are saturated with the therapeutic anti-CD20 mAb. Hence, a hierarchy of kinetics and sensitivity of depletion exists that can be categorized based on location and microenvironment (68–70). The relative resistance of MZ B cells to depletion is not B cell intrinsic because their mobilization from the spleen into the vasculature with anti-integrin (anti-α4 and anti-αL ) antibodies renders them susceptible to anti-CD20 mAb depletion. Conversely, inhibition of B cell egress from lymph nodes protects them from being depleted by anti-CD20 mAb. Together, these data underscore the importance of intravascular access for anti-CD20 mAb– mediated depletion of B cells (69). In addition to vascular access, the B cell microenvironment also contributes to factors deﬁning sensitivities of anti-CD20 mAb killing. In mixed bone marrow chimera experiments, MZ B cells are depleted entirely when competitor human CD20-negative B cells are present, suggesting that microenvironments participate in protecting cells in the absence of competition for local rescue factors. Of these, blockade of the BAFF/BLyS B cell survival factor with a BR3-Fc fusion protein synergizes with anti-CD20 mAb to enhance B cell depletion (see section “BAFF/BLyS and Its Receptors BAFFR/BR3, TACI, and BCMA,” below). The differential B cell sensitivities are also explained by the different in vivo mechanisms of anti-CD20 mAb–mediated depletion. The requirement for intravascular access is due to the requirement for FcR-mediated depletion of circulatory, lymph node, peritoneal, and splenic FO B cells by liver Kupffer cells. In contrast, the noncirculatory MZ B cell compartment within the hCD20 BAC transgenic www.annualreviews.org • B Cell Depletion

475

ANRV270-IY24-15

ARI

15 February 2006

1:8

model is not dependent on FcR but rather on complement function. Hence, the differential B cell sensitivities reﬂect the different mechanisms required for distinct compartments and microenvironments.

NHL: non-Hodgkin’s lymphoma

B Cell–Depleting Therapies—CD52, CD40, and CD22

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Aside from inducing B cell apoptosis through BCR signaling, investigators have directed signiﬁcant efforts toward B cell immunodepletion by targeting surface antigens preferentially expressed on B cells. Although not speciﬁc for B cells, Alemtuzumab (CamPath , Genzyme, Berlex/Schering, Millennium, and ILEX) is a humanized mAb directed against CD52, an antigen expressed on normal as well as malignant B and T cells, a subset of CD34+ bone marrow cells, NK cells, monocytes, macrophages, and tissues within the male reproductive system (71). Alemtuzumab depletes both T and B cells and, in several open-label, noncomparative studies, has demonstrated efﬁcacy in the treatment of ﬂudarabine-resistant B cell chronic lymphocyte leukemias (CLLs) (72). CD40 is highly expressed on B cells and also transduces a positive B cell–activating signal. Treatment of mice with agonistic antiCD40 mAbs have resulted in enhanced in vivo T cell–independent responses rather than inhibition of immunity (73). In contrast, administration of a partially agonistic chimeric mouse/human anti-CD40 mAb demonstrated efﬁcacy in renal transplantation and inhibition of anticytomegalovirus immune responses in nonhuman primates (NHPs) (74). A BCR signaling threshold is balanced by both positive (e.g., CD19 and CD21) and negative (e.g., CD22) coreceptors. CD22 is a B cell–restricted transmembrane protein, expressed on both normal and nonHodgkin’s B cell lymphoma (NHL), and controls BCR signaling thresholds through its three cytoplasmic inhibitory motifs that recruit SHIP and SHP1 inhibitory signaling proteins. Epratuzumab (Immunomedics, 476

Martin

·

Chan

Morris Plains, NJ), a humanized anti-CD22 mAb, has demonstrated clinical activity in NHL. Based on CD22’s ability to internalize following crosslinking, Epratuzumab conjugated to chemotherapeutic agents (e.g., calicheamicin) or radioactive agents (e.g., 90 YDOTA) has also demonstrated activity in NHL. Although the safety hurdles of using such toxin-conjugated agents are far greater in autoimmune disorders, a calicheamicinconjugated antimouse CD22 mAb not only resulted in B cell depletion, but also demonstrated efﬁcacy in a murine model of collageninduced arthritis (75). Interestingly, anticollagen antibodies were not dramatically altered following B cell depletion, which again supports nonautoantibody-related roles for B lineage cells in the pathogenesis of inﬂammatory arthritis. Our current understanding of the factors governing the success or failure of B cell immunotherapy has limitations. Current animal model experience with antibodies against murine MHC class II (Ia) showed B cell depletion and efﬁcacy in mouse models of MS (EAE in SJL mice) and thyroiditis (76, 77). Ongoing development of therapies directed against additional B cell targets, including CD23, CD80, and HLA-DR, will provide greater insights into the principles of cellular immunotherapy.

Blocking B Cell Survival and Activation A second major class of B cell modulators is secreted and membrane-form of growth factors that drive B cell development and survival. Several cytokines play critical roles in B cell development and differentiation. The classic lineage growth factor in mice, IL-7, acts on early pro- and pre-B cells, expanding them in primary lymphoid organs before they differentiate into immature and mature B cells (78). Blockade of murine IL-7 function with anti-IL-7 or by blocking anti-IL-7R mAbs inhibits early B cell development and generation of new immature B cells into the periphery.

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

Although IL-7 plays a critical role in mouse B cell differentiation, IL-7 in humans plays an important role in early T but not B cell development (79). However, more importantly for therapeutic approaches, a second phase of IL-7 dependence in the B cell lineage exists for both human and murine memory cells (80, 81). The differences between human and mouse biology increase the hurdles in evaluating therapeutic intervention, and further experimentation is required to assess potential differences in IL-7 biology in human memory B and T cell maintenance and function. In addition to IL-7, other common gamma (γc) chain signaling cytokines (IL-2, IL-4, and IL-21) participate in B cell activation and function (82, 83). Owing to their effects on many other cell types and their overlapping functions, therapeutic inhibition or activation has been challenging and is primarily pursued within oncologic indications. The more recently described IL-21 is produced speciﬁcally by an activated T cell subset that enters B cell follicles after antigen encounter and provides early T cell help for antigenspeciﬁc B cell activation (83, 84). These FO T cells are interesting therapeutic targets and have been recently shown to be dysregulated in an autoimmune-prone mouse with a point mutation in “roquin,” a ubiquitin-ligase involved in regulating messenger RNA translation and stability (85). This roquin mutation induces activation of FO T helper cells, enhanced expression of the inducible costimulator (ICOS), and hypersecretion of IL-21, resulting ultimately in lupus-like pathology characterized by focal proliferative glomerulonephritis with deposition of IgG-containing immune complexes, necrotizing hepatitis, anemia, and autoimmune thrombocytopenia. IL-21 has pleiotropic effects as IL-21R is expressed on B, T, and NK cells and DCs. Additional studies are required to deﬁne the potential therapeutic value and to reveal the toxicities for IL-21 neutralization in autoimmune diseases (83, 86, 87). In addition to cytokines that use the γc chain, IL-10 has also been implicated in hu-

man B cell growth (82). IL-10 synergizes in vitro with IL-2, IL-4, and IL-5 to augment B cell survival and activation (88). IL-10’s synergistic effects with these growth factors on B cell function, along with its pleiotropic effects on antigen presentation by DCs and macrophages, its inhibitory effects on T cell function, and its functions in regulatory T cell differentiation make IL-10 an interesting, but also challenging, target (89, 90). CD40 also serves as a major B cell coreceptor in B cell proliferation, survival, and Ig class switching. As mentioned above, depleting or blocking antibodies against CD40 on B cells as well as CD40L on T cells have been widely used to inhibit B and T cell activation and have demonstrated efﬁcacy in a diverse set of mouse disease models (91–94). Because CD40 is expressed on a variety of antigenpresenting cells besides B cells, and because CD40L can be upregulated on platelets, vascular endothelia, and smooth muscle cells, the biological effects of targeting this pathway extend beyond the T cell–B cell interaction (95). In turn, clinical development of blockers for the CD40/CD40L pathway, although very attractive from a scientiﬁc point of view, has been hampered by safety signals in NHPs and humans (96, 97).

BAFF: B cell–activating factor; also known as BLyS (B lymphocyte stimulator)

BAFF/BLyS and Its Receptors BAFFR/BR3, TACI, and BCMA Although CD40 serves as a major regulator of acute B cell activation and Ig class switching, peripheral B cell maturation and maintenance as well as plasma cell generation and survival are not dependent on CD40/CD40L. Mice and humans with defects in the CD40/CD40L pathway have normal or only slightly reduced B cell numbers and maturation and can mount immune responses restricted to the IgM isotype (93, 98, 99). Over the past few years, the TNF family member BAFF/BLyS has been shown to be critical for the survival of B lineage cells beyond the transitional T1 stage (100, 101). BAFF/BLyS is produced by www.annualreviews.org • B Cell Depletion

477

ANRV270-IY24-15

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

NHP: nonhuman primate

478

1:8

radioresistant stromal cells and activated myeloid cells and has receptors on B lineage cells—BAFFR/BR3 and, to a lesser extent, TACI and BCMA. BAFF, binding through BAFFR/BR3, triggers activation of the NFκB2 transcriptional factor and the serinethreonine kinase PIM2, both of which exert antiapoptotic effects through Bcl-2 activation and Bad inhibition, respectively (102– 105). Mice deﬁcient in BAFF or BAFFR, as well as wild-type mice that have a blockade of BAFF/BAFFR interactions through BAFFR/BR3-Fc, TACI-Fc, and BCMA-Fc, demonstrate marked reduction of cells within the B lineage compartment (100). Compared with anti-CD20 treatment, the kinetics of mouse B cell modulation using BAFF inhibitors are slightly slower, starting at day 3 and reaching 80%–95% depletion in 7 to 14 days. In contrast to anti-CD20 mAb– mediated B cell depletion, both lymph node and splenic B cells (FO and MZ) are equivalently sensitive to BAFF blockade. Moreover, both acute (sheep red blood cell responses) and chronically (Payer’s patches) generated GC B cells are partially dependent on BAFF for their maintenance, although to a lesser degree compared with the rest of the splenic and lymph node compartments (106). In contrast, development or maintenance of peritoneal cavity B1 B cells is BAFF independent (107). BAFF has also been implicated in setting the survival threshold of B cells, as autoreactive B cells compete with normal B cells for this critical survival factor (102, 108). Autoreactive B cells appear to exhibit greater dependence on BAFF for their survival. In turn, selfreactive B cells are competitively deleted in environments where BAFF is limiting. However, at this time it is unclear if the increased sensitivity of autoreactive B cells for survival signals derived through the BAFF axis extends to downstream plasmablasts or plasma cells. Transgenic mice overexpressing BAFF/BLyS induce expansion of all splenic B cell subsets, with a greater effect on transitional and MZ B cells, which results in the development Martin

·

Chan

of SLE- and Sjogren’s syndrome–like manifestations (49, 109–111). Conversely, blockade of BAFF/BLyS (BAFFR/BR3-Fc, TACIFc, and BCMA-Fc) in murine models of lupus inhibits both B and secondary T cell activation, and reduces autoantibody titers, thus improving survival (104, 110, 112, 113). Combined with the observation that serum BAFF levels are elevated in a variety of human autoimmune disorders, BAFF blockade offers an intriguing opportunity by which to preferentially affect autoimmune B cells (see the section “Anti-BAFF/BlyS mAbs,” below) (100, 114).

TRANSLATION FROM MICE TO HUMANS IN NONHUMAN PRIMATES Although much of our understanding of B cell biology has been based on mice studies, increasing insights into B cell biology are emerging from B cell–targeted therapies in humans and NHPs. In addition to the differences observed in IL-7 in human and mouse B cell biology, some interesting parallels and differences have also been observed with studies of anti-CD20 mAbs and BAFF/BLyS blockade in NHPs and humans. In NHPs, a range in kinetics and sensitivity of B cell subset depletion is also observed with anti-CD20 mAb treatment. As in rodents, treatment of NHPs with anti-CD20 mAbs does not completely deplete all B cells, even at doses several-fold greater than the clinically approved dose (115, 116). Furthermore, circulating B cells and other recirculating B cell subsets (follicular mantle and splenic memory B cells) are exquisitely sensitive to depletion, whereas other B cell subsets (activated and lymph node GC B cells) appear more resistant (116, 117). By comparison, BAFF/BLyS blockade with an anti-BLyS antagonistic antibody or BR3-Fc fusion protein has a quantitatively distinct outcome in NHPs when compared with mice. First, the kinetics of peripheral B cell depletion are signiﬁcantly slower in NHPs than in mice, with

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

∼50% depletion seen only after 8–12 weeks of treatment, compared with 3–5 days in mice (118). Second, although there is an increased relative sensitivity of NHP MZ B cells to BAFF/BLyS blockade, when compared with other mature B cell subsets, the decrease in NHP MZ B cells does not reach the complete depletion observed in mice even after 12–26 weeks. These comparative studies indicate that anti-CD20 mAb–mediated B cell depletion translates well at ﬁrst approximation and that, although there are quantitative differences when blocking the BAFF/BLyS pathway, the hierarchy of B cell reduction is comparable between mice and NHPs. Some of the more detailed differences in B cell depletion proﬁles between the primate and rodent systems are due to particularities in individual life span and lymphoid microarchitecture between these two species. Although GC B cells are lacking (outside of the Peyers’ patches) in pathogen-free mice, GC B cells are abundant in NHPs (and humans) in spleen and lymph nodes. In addition, in normal NHPs (but not in normal humans) a relatively large population of GC-like cells (as phenotyped by FACS) is observed in blood that is exquisitely sensitive to anti-CD20 mAb– mediated depletion but not to BAFF/BLyS blockade (117). Likewise, the MZ areas in spleen (and lymph nodes in NHPs) have different microanatomical organization and cellular composition compared with mice. As a consequence, MZ B cells in NHPs contain largely circulating memory B cells and are more sensitive to anti-CD20 mAb–mediated depletion than the mouse splenic MZ B cells, which are noncirculating naive B cells (69, 116, 119). Finally, anti-CD20 mAb treatment in NHPs results in near complete depletion of splenic B cells (including GC and MZ memory B cells), whereas the same B cell subsets within the lymph node are more resistant. Hence, although the general rules regarding the contributions of circulatory dynamics and microenvironment to immunotherapy are similar across species, differences in topography and composition of speciﬁc B cell sub-

sets result in somewhat different B cell depletion proﬁles that have to be considered carefully when translating research from mouse to human.

THERAPEUTIC HUMAN B CELL DEPLETION The success of targeted B cell therapy in patients with NHL and its known paraneoplastic associations opened the therapeutic application of B cell–directed therapies to the ﬁeld of autoimmunity. Two approaches have been taken to manipulate the B cell compartment: (a) depleting antibodies against antigens preferentially or speciﬁcally expressed on B cells (i.e., CD20, CD22, and CD52), and (b) blockade of cell survival factors and signals (i.e., CD40/CD40L and BLyS/BR3 pathways) essential for B cell activation and survival.

Clinical Experience with Anti-CD20 mAb Therapies Recent reviews have covered the extensive clinical experience in oncology with Rituximab (Rituxan , Genentech, Inc., South San Francisco, CA, and Biogen-IDEC, Cambridge, MA; Mabthera , F. HoffmanLaRoche, Ltd., Basel, Switzerland), a chimeric mAb directed against the B cell– speciﬁc CD20 molecule (120, 121). In this section, we review the emerging clinical experience with Rituxan in a variety of autoimmune diseases and how the clinical and immunologic consequences of this therapeutic intervention provide insights into human immunology and disease. The application of Rituxan in primary autoimmune disorders was catalyzed by observations that treatment of B cell lymphomas results in improvement of lymphoma-associated autoimmune phenomenon. The malignant B cell clones that are causative in NHL can also produce low-afﬁnity antibodies (paraproteins) that bind a variety of self-antigens. Examples include antibodies against antigens expressed on red blood cells and the complement www.annualreviews.org • B Cell Depletion

479

ANRV270-IY24-15

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ITP: immuno-mediated thrombocytopenia

480

1:8

inhibitor (C1 esterase inhibitor) that, in turn, manifest in autoimmune hemolytic anemia (AIHA) (e.g., IgM cold agglutinin disease) and angioedema, respectively (122–128). These initial anecdotes prompted greater investigation of Rituxan in the treatment of primary cytopenias. To date, several clinical studies, totaling more than 25 children and 15 adults (some with associated malignancies), have demonstrated clinical utility of Rituxan in both primary and secondary AIHA (129–132). Investigation of Rituxan in ITP was a natural extension of the work in AIHA. In four case series reports, 33%–54% of ITP patients, who had failed corticosteroids and splenectomy, experienced partial response (PR) or complete response (CR) to Rituxan, in most cases in the absence of concomitant corticosteroids (133– 136). Among the CRs (those who achieve a normal platelet count of >150 × 109 /l), three distinct patterns of CRs have been observed. Early responders experienced immediate (1–3 weeks) increases in platelet counts after the ﬁrst or second antibody infusions and normalization of platelet counts within the ﬁrst month. A second group had little change in platelet counts within the ﬁrst month, but then achieved normal platelet counts by the second or third month following therapy. Finally, a third CR phenotype had slow, sustained increases, reaching >50 × 109 /l by 2 months, but ultimately achieved normal platelet counts between months 3 and 8. These distinguishing patterns suggest that Rituxan likely operates through multiple mechanisms in ITP. In the ﬁrst group of CRs, the effects of Rituxan are too rapid to be accounted for by reduction of circulating antiplatelet antibodies, but more likely are mediated through FcR-mediated functions, as described for intravenous immunoglobulin therapies (137). In the latter CR groups, the sustained effects may function through alterations in antiplatelet antibodies, interruption of T cell–B cell cooperation, and/or other B cell–dependent pathogenic mechanisms. However, there appears to be no diMartin

·

Chan

rect correlation of antiplatelet antibody levels with clinical responses in the more than 100 patients treated to date, which emphasizes the likelihood of non-autoantibody-mediated mechanisms by which B lineage cell depletion modulates ITP. In contrast to ITP, where there does not appear to be a correlation between autoantibody titers and clinical response, several clinical syndromes have demonstrated a reasonable association between clinical response and decreases in autoantibody titers. One of the ﬁrst demonstrations that eradication of B cells may provide clinical beneﬁt in antibodymediated disorders was in the treatment of IgM-mediated neuropathies (138–141). IgM neuropathy is characterized by expansion of an IgM-producing B cell clone with speciﬁcity for neural antigens, the most abundant (50%– 60%) being myelin-associated glycoprotein, which induces axonal destruction resulting in neuropathy and loss of motor strength. Treatment with Rituxan results in decreases in pathogenic IgM titers, decreases that are associated with clinical improvements in muscle strength (142). Patients with hemophilia, as a result of FVIII deﬁciency, are treated with chronic FVIII replacement but develop FVIII inhibitors (i.e., anti-FVIII autoantibodies) that limit therapeutic efﬁcacy of future FVIII replacement therapy. A recent study of four such patients treated with Rituxan resulted in rapid decreases in their acquired FVIII inhibitor and return of their endogenous FVIII levels (143). Interestingly, one of these four patients had mild hemophilia A and received perioperative recombinant human FVIII (rFVIII) that appeared to trigger an increase in FVIII inhibitor activity to further compromise his already low endogenous FVIII level. Administration of Rituxan in this patient resulted in a distinct response of the autoantibody or alloantibody against FVIII. Although the autoantibody against endogenous FVIII rapidly resolved, the alloantibody against rFVIII decreased, but was still measurable for more than two months following Rituxan

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

administration. These anecdotes indicate that the sensitivity of self-reactive antibodyproducing cells to antigen are likely distinct from nonself-antibody-producing cells. Because Rituxan does not directly target plasma cells (that inherently do not express CD20), this differential reactivity likely reﬂects distinct longevities for autoreactive and nonselfreactive plasma cells, with autoreactive plasma cells potentially having a shorter-lived CD20bearing progenitor than do nonself-reactive plasma cells. Two open-label studies totaling 20 patients with antineutrophil cytoplasmic antibody (ANCA)-positive vasculitis (that include Wegener’s Granulomatosis and microscopic polyarteritis) have reported signiﬁcant improvements in clinical outcome following treatment with Rituxan in combination with a corticosteroid, mycophenolate, short courses of cyclophosphamide, or azathioprine (144, 145). These patients develop a spectrum of disease ranging from limited necrotizing granulomas to generalized vascular inﬂammation involving the upper and lower respiratory airways, kidneys, and skin that typically require treatment with high doses of corticosteroids and cyclophosphamide. In one study, all 11 patients achieved clinical remission (as measured by the Birmingham Vasculitis Activity Score) within six months of therapy. In this study, all patients experienced signiﬁcant decreases in their serum antiproteinase 3 (PR3) antibody titers. Two patients experienced a relapse in disease concomitant with peripheral B cell return and elevations in ANCA titers that again responded to Rituxan and corticosteroids. In contrast, a second series of nine patients with ANCAassociated vasculitis also reported similar clinical improvement following treatment with a combination of Rituxan and other immunosuppressive therapies. However, in this series, no signiﬁcant decreases were observed in ANCA or anti-PR3 antibodies measured by immunoﬂuorescence or ELISA, respectively. Although these immunologic differences need to be resolved in future clinical

trials, it is of interest that passive transfer of polyclonal antisera against mouse PR3 in conjunction with intradermal TNF signiﬁcantly enhances inﬂammation of the dermis and subcutaneous tissues, although antimouse PR3 antisera alone does not induce disease. This corequirement for TNF may relate to the contributions of concomitant inﬂammatory signals that are known to serve as potential triggers of disease. Patients who are chronic nasal carriers of Staphylococcus aureus are more prone to disease relapses (146). Because a variety of S. aureus constituents are known to serve as immunomodulators for a variety of immune, endothelial, and epithelial cells, some of these immunomodulators may directly or indirectly promote autoantibody production in ANCA-associated vasculitis and thus contribute to disease. Finally, clinical improvement following treatment with Rituxan in pemphigus vulgaris, pemphigus foliaceus, and myasthenia gravis is also associated with decreases in pathogenic anti-desmoglein 3, anti-desmoglein 1, and anti-acetylcholine receptor antibody titers. Together, B cell–depleting therapies through anti-CD20 antibodies likely provide clinical beneﬁt in many autoimmune disorders by targeting CD20-bearing, short-lived, autoreactive plasma cell progenitors. The greatest experience among autoimmune disorders with Rituxan, to date, is in the treatment of RA. This disease is characterized by a chronic inﬂammatory process in the joints that advances to erosions of bone architecture and decrement in joint function. Multiple disease pathogenic mechanisms are involved, including the inﬁltration of activated lymphocytes and granulocytes, proliferation of synovial ﬁbroblast and macrophages, and neovascularization of the lining surrounding joints. Therapeutics targeted at distinct cellular compartments/pathways (e.g., anti-TNF and IL-1 cytokines, CTLA4-Fc, anti-IL-6R, and Rituxan) provide clinical beneﬁt, which further supports a complex pathogenesis for RA. In a multicenter, randomized, doubleblind controlled study of 161 patients, patients www.annualreviews.org • B Cell Depletion

481

ANRV270-IY24-15

ARI

15 February 2006

1:8

treated with Rituxan and a short course of corticosteroids in conjunction with methotrexate or cyclophosphamide achieved greater clinical improvement than patients treated with corticosteroids and methotrexate alone (147, 148). The diagnosis of RA is aided typically by the presence of several autoantibodies. These include low-afﬁnity IgM, IgG, and IgA RF autoantibodies directed against the Fc portion of IgG as well as anti-CCP (cyclic citrulline peptide) antibodies. The latter are autoantibodies to antigens containing one or more citrulline residues and have been shown to have a high diagnosis/prognosis value in patients with RA and a speciﬁcity of more than 90%. Analysis of serologic changes in a smaller series of 22 RA patients following Rituxan treatment demonstrated a 60%, 80%, and 75% decrease in IgM RF, IgG RF, and anti-CCP antibodies, respectively (149). In contrast, three months following treatment, no change was detected in antipneumococcal capsular polysaccharide IgG antibodies, and only a minor (23%) decrease was shown in antitetanus toxoid IgG antibodies. Researchers have observed similar serologic effects with regard to total Ig and RF levels in two additional studies with Rituxan treatment (150, 151). This preferential decrease in autoantibodies suggests that autoreactive B cells and their corresponding plasma cells likely have short life spans, whereas nonself-reactive plasma cells (responsible for post-vaccination nonself responses) are relatively more resistant to Rituxan and have longer life spans. Unlike the KRN serum transfer model of inﬂammatory arthritis in mice, serum transfer of RF in humans does not confer arthritis and implicates non-RF-mediated mechanisms for RA (16). Undoubtedly, these mechanisms involve the nonautoantibody-mediated mechanisms of antigen presentation, cytokine secretion, costimulatory and survival signals, and neolymphogenic-mediated mechanisms in which B cells also likely participate in human RA (1, 54, 152). One of the most fertile diseases for understanding human B cell immunology is

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CCP: cyclic citrulline peptide

482

Martin

·

Chan

SLE. This autoimmune disease is characterized by a diverse repertoire of autoantibodies in which speciﬁc autoantibodies are associated with clinical subsets of patients as well as with clinical severity of disease. Three small open-label studies and several case reports are encouraging, but the overall efﬁcacy of Rituxan treatment is unknown at this time (153–155). Similar to the experience in other diseases described above, immunologic parameters in these studies demonstrate clinical responses disconnected from autoantibody reduction but correlated with B cell depletion (156). Although decreases in speciﬁc autoantibodies against nuclear constituents (antinuclear and antiextractable nuclear antibodies), dsDNA, Smith antigen, and ribonuclear proteins have been described in some patients, these decreases are not uniform (157). Additionally, as a large number of circulating plasmablasts, memory B cells, and plasma cells are detected in the blood of SLE patients, a signiﬁcant amount of B cell biology is to be learned from the present and future trials in SLE. In sum, the lessons taken from anti-CD20 mAb–mediated B cell depletion suggest that for most autoimmune diseases the mechanistic relationship between pathogenic B cells, their autoantibody products, and clinical manifestations will be addressed within the next decade as the clinical experience with anti-CD20 B cell–depleting therapies evolves.

Anti-CD52 In contrast to Rituxan treatment that targets only B cells, Campath-1H (Berlex/Schering, Millennium, and ILEX), an anti-CD52 humanized mAb, binds to CD52, which has a broader expression pattern that includes B cells, T cells, monocytes, and eosinophils. Currently approved for patients with chemotherapy-resistant CLL, Campath1H is thought to act by combining immunosuppressive as well as immunoablative properties. For many years, Campath-1H’s clinical application has been attempted for

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

various autoimmune diseases and an inductive regimen for allogeneic bone marrow transplantation to prevent graft-versus-host disease. More than 300 patients have been treated with Campath-1H for various autoimmune diseases that include RA, MS, vasculitis, scleroderma, and various autoimmune cytopenias (158). Results in both RA and MS have been disappointing due to a low therapeutic index, infectious complications associated with prolonged T cell lymphopenias, and, in a third of MS-treated patients, development of autoimmune thyroiditis (159– 161). Treatment in autoimmune cytopenias (ITP, AIHA, and autoimmune neutropenia) has proven more successful (162, 163). Owing to the relatively high incidence of side effects, this therapy is currently investigated for severe, refractory cytopenias who fail to respond to conventional immunosuppressive therapy.

Anti-CD22 CD22 (epratuzumab, Immunomedics, Morris Plains, NJ), as described above, is a B cell–speciﬁc inhibitory coreceptor that sets the threshold for B cell activation (164, 165). Again, success in Phase I/II clinical trials in NHL catalyzed investigation of anti-CD22 antibodies in SLE and Sjogren’s syndrome. In these early investigations, treatment of patients typically resulted in ∼50%–60% decrease in peripheral B cells and was associated with improvement in BILAG scores (a composite score of clinical parameters) in patients with SLE. In studies of Sjogren’s syndrome, all of the 15 patients with primary Sjogren’s syndrome enrolled in an open-label study showed improvement in tender joints, and 33% of patients had increased salivary ﬂow. As CD22 expression is highly modulated during B cell development and activation, analysis of immunologic consequences and deﬁnition of the targeted B cell subsets will further illuminate our understanding of human B cell immunobiology.

Anti-CD40L A complementary strategy to depleting antibodies is the inhibition of B cell survival and activation. The best-studied pathway involved in B cell survival is the signal provided by CD40L on T cells to its CD40 receptor on B cells. Engagement of CD40 results in NF-κB activation and promotes B cell proliferation and survival. With active SLE, in contrast to most other autoimmune diseases and despite their generalized lymphopenia, patients have signiﬁcantly elevated numbers of circulating activated B lineage cells: memory, GC, and plasma cells. Some of these cells have variable levels of inappropriate CD40L expression and in vitro can promote spontaneous proliferation and Ig secretion in a CD40/CD40Ldependent fashion (96, 166, 167). Treatment of SLE patients with an anti-CD40L antibody (BG9588, 5c8, Biogen-IDEC) results in substantial decreases in anti-dsDNA antibody titers, proteinuria, and an improved clinical SLEDAI score (a composite measure of SLE disease severity). Correspondingly, decreases in GC cells and plasmablasts, subsets that normally do not circulate through blood, were observed in treated patients (96). Consistent with this ﬁnding, patients’ peripheral B cells cultured in vitro spontaneously proliferated and secreted Ig in a manner that was inhibited by anti-CD40L blocking antibody. Together, these results indicate that patients with active lupus exhibit abnormalities in the peripheral B cell compartment that are consistent with intensive GC activity driven via CD40LCD40 interactions, which contributes to the propensity of these patients to produce autoantibodies. Hence, CD40/CD40L blockade operates at multiple levels, including inhibiting (a) overall B cell survival; (b) GC formation, maintenance, and differentiation into plasmacytes; and (c) autoantibody generation. Although side effects due to CD40L expression on human platelets and CD40 expression on activated endothelial cells have slowed down clinical development of CD40/CD40L inhibitors, this pathway remains a very

www.annualreviews.org • B Cell Depletion

483

ANRV270-IY24-15

ARI

15 February 2006

ADCC: antibody-dependent cell-mediated cytotoxicity

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CDC: complementdependent cytotoxicity

484

1:8

attractive therapeutic target for diseases involving uncontrolled, pathogenic B cell activation.

Anti-BAFF/BLyS mAb A second major B cell survival signal is delivered by the soluble TNF family member BAFF/BLyS through its receptors, BR3/BAFFR, TACI, and BCMA. As described above, BAFF/BLyS–BR3/BAFFR interaction provides a survival signal for B cell development in mice and primates (118). Although experimental evidence in rodents and NHPs suggests that B cell dependence on BAFF signaling is somehow less in humans than in mice, recent preliminary results from a neutralizing anti-BLyS antibody (Lymphostat B, Human Genome Sciences, Rockville, MD) demonstrated clinical beneﬁt in RA. In this Phase II trial of 283 patients with moderate to severe RA, which included patients that have failed anti-TNF therapies, 31% of patients achieved >20% improvement in their composite clinical scores (ACR20). These data provide the ﬁrst demonstration that sequestration of the BAFF/BLyS survival factor is important in the immunopathogenesis of human RA. As mice that overexpress BAFF/BLyS develop a lupus-like syndrome, and, conversely, as the blockade of BAFF/BLyS within a variety of murine models of lupus improves immunologic parameters and mortality, an ongoing Phase II study in SLE with Lymphostat will illuminate the BAFF/BLyS contribution in SLE. Finally, these mice overexpressing BAFF/ BLyS also develop secondary pathology reminiscent of Sjogren’s syndrome, manifested by severe sialadenitis, decreased saliva production, and destruction of submaxillary glands (49). A likely explanation for disease in BAFF/ BLyS transgenic mice is excessive and sustained survival signals to autoreactive B cells, possibly as they pass through selection checkpoints during their maturation. The MZ B cell compartment, one of the enlarged B cell subsets in the spleen of BAFF/BLyS transgenic Martin

·

Chan

mice, is a potential reservoir of autoreactive B cells. Interestingly, B cells with an MZlike phenotype inﬁltrate the salivary glands of these transgenic mice, suggesting that cells of this phenotype potentially participate in tissue damage in Sjogren’s syndrome. Because in humans with Sjogren’s syndrome elevated levels of circulating and salivary gland–expressed BAFF have been well documented, additional investigation of BLyS/BR3 blockade appears warranted in this disease (49, 168, 169).

MECHANISMS OF ANTIBODY-MEDIATED B CELL DEPLETION AND RESISTANCE TO DEPLETION In this ﬁnal section, we brieﬂy overview the mechanisms by which naked antibodies induce in vivo B cell depletion, the potential mechanisms of B cell resistance to depletion, and opportunities for enhancing therapeutic efﬁcacy. Owing to the substantially larger experience in the ﬁeld with anti-CD20, we use it as a case study, although many of these observations on Fc-mediated effector functions may be translatable to a number of other target molecules and cell types. The combined experience from hCD20 transgenic mice and a panel of antimouse CD20 antibodies suggests that ADCC and to a lesser extent CDC are the main mechanisms by which various subsets of normal mouse B cells are depleted (68, 69, 170). These observations likely apply to human autoimmunity because self-reactive cells recirculate and are probably activated using patterns and pathways that follow a comparable course to mouse and NHP normal B cells. In contrast, the mechanisms of human tumor B cell depletion are less clear because of the marked differences between normal and malignant B cells as well as the existence of diverse tumorspeciﬁc microenvironments that control the availability of effector mechanisms for tumor killing. This ﬁeld has been elegantly reviewed in several recent publications, and ADCC, CDC, apoptosis, and alterations in cellular

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

signaling have all been demonstrated to contribute, in variable degree, to the killing of tumor B cells by anti-CD20 antibodies (67, 171–176). The large heterogeneity of tumor B cells across and within each disease, together with epitope-speciﬁc properties for various agents directed against the same target, complicate this issue further; however, multiple mechanisms likely contribute to tumor B cell depletion. A summary of the mechanisms and factors involved in depletion suggests that they can be grouped into three categories, depending on (a) the molecular target (presence on target cells, expression levels, modulation); (b) effector functions [speciﬁc either for the antibody used for depletion (epitope, afﬁnity, ADCC, CDC) or for the host (availability and activation of cellular effectors such as macrophages, NK cells, and complement)]; and (c) target cell susceptibility (kinetics of recirculation, important mostly for autoimmune and tumor-circulating cells, microenvironment rescue, and cellular competition for available survival factors). As a consequence, the incomplete B cell depletion described for normal as well as self-reactive and malignant B cells can be due to any one, or to a combination, of these factors. Any of the following mechanisms may alter the level of B cell depletion: (a) suboptimal expression or downregulation of target antigen, (b) inappropriate effector mechanisms of the antibody or a host compromised in cell killing pathways, (c) poor circulatory dynamics, or (d) the presence of microenvironmental rescue factors. Constant efforts to maximize all these factors have succeeded in various in vitro and in vivo animal models, and their comprehensive evaluation will be paramount in order to select the most promising for further clinical development. For the CD20 molecule in particular, efforts to identify antibodies against different target epitopes resulted in a series of molecules with slightly different properties, and various methods to upregulate the CD20 molecule either in vitro or in vivo have resulted in only modest incremental results

(172, 174). By far, the most studied pathway to enhance B cell killing is to improve the effector function of the therapeutic antibody. This is based, in part, on the observations that Rituxan-treated patients with NHL have a better prognosis as measured by longer times to disease progression if they express the highafﬁnity alleles of CD16 (V158) and CD32 (H131) (177, 178). Preliminary data in SLE also suggest that patients with CD16 (V158) have a greater ability to deplete peripheral B cells following Rituxan therapy (154). Modulation of FcγR function can be accomplished through amino acid substitutions within the Fc portion as well as by reducing the fucose content of the therapeutic antibody (179). In addition to modiﬁcation of the effector functions, another approach is to fuse activators of immune effectors (e.g., IL-2 and GM-CSF) to therapeutic antibodies to enhance efﬁcacy (65, 161). Combinations of GM-CSF or IL-2 with Rituxan are efﬁcacious in the treatment of NHL (180–183). With the identiﬁcation of target cell–speciﬁc factors, such as the circulatory dynamics and the cellular microenvironment, additional possibilities of optimizing B cell depletion are emerging (1, 184). For example, the use of integrin blockade holds some promise in preclinical models of multiple myeloma (184 –186). Finally, blockade of survival factors (BAFF/BLyS, TLR, CD40L, and other cytokines promoting growth) appears to be within reach and will be tested in conjunction with anti-CD20 B cell depletion in both autoimmunity and speciﬁc malignancy. In summary, B lineage depletion and the inhibition of B cell activation and survival are beneﬁcial in multiple autoimmune diseases and provide a basis to further explore the role of B cell subsets and their function in normal and diseased humans. Over the next few years, results from many ongoing clinical trials targeting several B cell molecules and pathways will allow a more complete understanding and dissection of the mechanisms by which B cells are deregulated, which will contribute to understanding human pathogenesis. www.annualreviews.org • B Cell Depletion

485

ANRV270-IY24-15

ARI

15 February 2006

1:8

ACKNOWLEDGMENTS The authors thank Mercedesz Balazs for help with the graphics. Flavius Martin and Andrew Chan are employees of Genentech, Inc.

LITERATURE CITED

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

1. Martin F, Chan AC. 2004. Pathogenic roles of B cells in human autoimmunity; insights from the clinic. Immunity 20:517–27 2. Silverman GJ, Carson DA. 2003. Roles of B cells in rheumatoid arthritis. Arthritis Res. Ther. 5(Suppl. 4):S1–6 3. Ruff RL, Lennon VA. 1998. End-plate voltage-gated sodium channels are lost in clinical and experimental myasthenia gravis. Ann. Neurol. 43:370–79 4. van Lent PL, Nabbe K, Blom AB, Holthuysen AE, Sloetjes A, et al. 2001. Role of activatory FcγRI and FcγRIII and inhibitory FcγRII in inﬂammation and cartilage destruction during experimental antigen-induced arthritis. Am. J. Pathol. 159:2309–20 5. Abdul-Majid KB, Stefferl A, Bourquin C, Lassmann H, Linington C, et al. 2002. Fc receptors are critical for autoimmune inﬂammatory damage to the central nervous system in experimental autoimmune encephalomyelitis. Scand. J. Immunol. 55:70–81 6. Fujii T, Hamano Y, Ueda S, Akikusa B, Yamasaki S, et al. 2003. Predominant role of FcγRIII in the induction of accelerated nephrotoxic glomerulonephritis. Kidney Int. 64:1406–16 7. Pritchard NR, Cutler AJ, Uribe S, Chadban SJ, Morley BJ, Smith KG. 2000. Autoimmune-prone mice share a promoter haplotype associated with reduced expression and function of the Fc receptor FcγRII. Curr. Biol. 10:227–30 8. Bolland S, Ravetch JV. 2000. Spontaneous autoimmune disease in FcγRIIB-deﬁcient mice results from strain-speciﬁc epistasis. Immunity 13:277–85 9. McGaha TL, Sorrentino B, Ravetch JV. 2005. Restoration of tolerance in lupus by targeted inhibitory receptor expression. Science 307:590–93 10. Jiang Y, Hirose S, Abe M, Sanokawa-Akakura R, Ohtsuji M, et al. 2000. Polymorphisms in IgG Fc receptor IIB regulatory regions associated with autoimmune susceptibility. Immunogenetics 51:429–35 11. Jiang Y, Hirose S, Sanokawa-Akakura R, Abe M, Mi X, et al. 1999. Genetically determined aberrant down-regulation of FcγRIIB1 in germinal center B cells associated with hyper-IgG and IgG autoantibodies in murine systemic lupus erythematosus. Int. Immunol. 11:1685–91 12. Xiu Y, Nakamura K, Abe M, Li N, Wen XS, et al. 2002. Transcriptional regulation of FcγR2b gene by polymorphic promoter region and its contribution to humoral immune responses. J. Immunol. 169:4340–46 13. Boule MW, Broughton C, MacKay F, Akira S, Marshak-Rothstein A, Rifkin IR. 2004. Toll-like receptor 9-dependent and -independent dendritic cell activation by chromatinimmunoglobulin G complexes. J. Exp. Med. 199:1631–40 14. Vallin H, Perers A, Alm GV, Ronnblom L. 1999. Anti-double-stranded DNA antibodies and immunostimulatory plasmid DNA in combination mimic the endogenous IFN-α inducer in systemic lupus erythematosus. J. Immunol. 163:6306–13 15. Bave U, Magnusson M, Eloranta ML, Perers A, Alm GV, Ronnblom L. 2003. FcγRIIa is expressed on natural IFN-α-producing cells (plasmacytoid dendritic cells) and is required for the IFN-α production induced by apoptotic cells combined with lupus IgG. J. Immunol. 171:3296–302 486

Martin

·

Chan

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

16. Harris J, Vaughan JH. 1961. Transfusion studies in rheumatoid arthritis. Arthritis Rheum. 4:47–55 17. Kallies A, Hasbold J, Tarlinton DM, Dietrich W, Corcoran LM, et al. 2004. Plasma cell ontogeny deﬁned by quantitative changes in Blimp-1 expression. J. Exp. Med. 200:967–77 18. Dorner T, Radbruch A. 2005. Selecting B cells and plasma cells to memory. J. Exp. Med. 201:497–99 19. Odendahl M, Mei H, Hoyer BF, Jacobi AM, Hansen A, et al. 2005. Generation of migratory antigen-speciﬁc plasma blasts and mobilization of resident plasma cells in a secondary immune response. Blood 105:1614–21 20. Blink EJ, Light A, Kallies A, Nutt SL, Hodgkin PD, Tarlinton DM. 2005. Early appearance of germinal center-derived memory B cells and plasma cells in blood after primary immunization. J. Exp. Med. 201:545–54 21. William J, Euler C, Christensen S, Shlomchik MJ. 2002. Evolution of autoantibody responses via somatic hypermutation outside of germinal centers. Science 297:2066–70 22. Shlomchik MJ, Marshak-Rothstein A, Wolfowicz CB, Rothstein TL, Weigert MG. 1987. The role of clonal selection and somatic mutation in autoimmunity. Nature 328:805–11 23. Shlomchik M, Mascelli M, Shan H, Radic MZ, Pisetsky D, et al. 1990. Anti-DNA antibodies from autoimmune mice arise by clonal expansion and somatic mutation. J. Exp. Med. 171:265–92 24. Shlomchik MJ, Euler CW, Christensen SC, William J. 2003. Activation of rheumatoid factor (RF) B cells and somatic hypermutation outside of germinal centers in autoimmuneprone MRL/lpr mice. Ann. NY Acad. Sci. 987:38–50 25. Dorner T, Egerer K, Feist E, Burmester GR. 2004. Rheumatoid factor revisited. Curr. Opin. Rheumatol. 16:246–53 26. William J, Euler C, Shlomchik MJ. 2005. Short-lived plasmablasts dominate the early spontaneous rheumatoid factor response: differentiation pathways, hypermutating cell types, and afﬁnity maturation outside the germinal center. J. Immunol. 174:6879–87 27. Shlomchik MJ, Madaio MP, Ni D, Trounstein M, Huszar D. 1994. The role of B cells in lpr/lpr-induced autoimmunity. J. Exp. Med. 180:1295–306 28. Chan O, Shlomchik MJ. 1998. A new role for B cells in systemic autoimmunity: B cells promote spontaneous T cell activation in MRL-lpr/lpr mice. J. Immunol. 160:51–59 29. Chan OT, Hannum LG, Haberman AM, Madaio MP, Shlomchik MJ. 1999. A novel mouse with B cells but lacking serum antibody reveals an antibody-independent role for B cells in murine lupus. J. Exp. Med. 189:1639–48 30. Wong FS, Wen L, Tang M, Ramanathan M, Visintin I, et al. 2004. Investigation of the role of B-cells in type 1 diabetes in the NOD mouse. Diabetes 53:2581–87 31. Wellmann U, Letz M, Schneider A, Amann K, Winkler TH. 2001. An Ig μ-heavy chain transgene inhibits systemic lupus erythematosus immunopathology in autoimmune (NZB × NZW)F1 mice. Int. Immunol. 13:1461–69 32. Chan OT, Madaio MP, Shlomchik MJ. 1999. B cells are required for lupus nephritis in the polygenic, Fas-intact MRL model of systemic autoimmunity. J. Immunol. 163:3592–96 33. Chan OT, Shlomchik MJ. 2000. Cutting edge: B cells promote CD8+ T cell activation in MRL-Fas(lpr) mice independently of MHC class I antigen presentation. J. Immunol. 164:1658–62 34. O’Neill SK, Shlomchik MJ, Glant TT, Cao Y, Doodes PD, Finnegan A. 2005. Antigenspeciﬁc B cells are required as APCs and autoantibody-producing cells for induction of severe autoimmune arthritis. J. Immunol. 174:3781–88 www.annualreviews.org • B Cell Depletion

487

ARI

15 February 2006

1:8

35. Kalled SL, Cutler AH, Datta SK, Thomas DW. 1998. Anti-CD40 ligand antibody treatment of SNF1 mice with established nephritis: preservation of kidney function. J. Immunol. 160:2158–65 36. De Jong YP, Comiskey M, Kalled SL, Mizoguchi E, Flavell RA, et al. 2000. Chronic murine colitis is dependent on the CD154/CD40 pathway and can be attenuated by anti-CD154 administration. Gastroenterology 119:715–23 37. Kaser A, Dunzendorfer S, Offner FA, Ludwiczek O, Enrich B, et al. 2000. B lymphocytederived IL-16 attracts dendritic cells and Th cells. J. Immunol. 165:2474 –80 38. Krzysiek R, Lefevre EA, Zou W, Foussat A, Bernard J, et al. 1999. Antigen receptor engagement selectively induces macrophage inﬂammatory protein-1α (MIP-1α) and MIP1β chemokine production in human B cells. J. Immunol. 162:4455–63 39. Moulin V, Andris F, Thielemans K, Maliszewski C, Urbain J, Moser M. 2000. B lymphocytes regulate dendritic cell (DC) function in vivo: increased interleukin 12 production by DCs from B cell-deﬁcient mice results in T helper cell type 1 deviation. J. Exp. Med. 192:475–82 40. Fu YX, Huang G, Matsumoto M, Molina H, Chaplin DD. 1997. Independent signals regulate development of primary and secondary follicle structure in spleen and mesenteric lymph node. Proc. Natl. Acad. Sci. USA 94:5739–43 41. Ansel KM, Cyster JG. 2001. Chemokines in lymphopoiesis and lymphoid organ development. Curr. Opin. Immunol. 13:172–79 42. Ansel KM, Ngo VN, Hyman PL, Luther SA, Forster R, et al. 2000. A chemokine-driven positive feedback loop organizes lymphoid follicles Nature 406:309–14 43. Cyster JG, Ngo VN, Ekland EH, Gunn MD, Sedgwick JD, Ansel KM. 1999. Chemokines and B-cell homing to follicles. Curr. Top. Microbiol. Immunol. 246:87–92 44. Weyand CM, Kang YM, Kurtin PJ, Goronzy JJ. 2003. The power of the third dimension: tissue architecture and autoimmunity in rheumatoid arthritis. Curr. Opin. Rheumatol. 15:259–66 45. Corcione A, Casazza S, Ferretti E, Giunti D, Zappia E, et al. 2004. Recapitulation of B cell differentiation in the central nervous system of patients with multiple sclerosis. Proc. Natl. Acad. Sci. USA 101:11064 –69 46. Seraﬁni B, Rosicarelli B, Magliozzi R, Stigliano E, Aloisi F. 2004. Detection of ectopic Bcell follicles with germinal centers in the meninges of patients with secondary progressive multiple sclerosis. Brain Pathol. 14:164 –74 47. Uccelli A, Aloisi F, Pistoia V. 2005. Unveiling the enigma of the CNS as a B-cell fostering environment. Trends Immunol. 26:254 –59 48. Barone F, Bombardieri M, Manzo A, Blades MC, Morgan PR, et al. 2005. Association of CXCL13 and CCL21 expression with the progressive organization of lymphoid-like structures in Sjogren’s syndrome. Arthritis Rheum. 52:1773–84 49. Groom J, Kalled SL, Cutler AH, Olson C, Woodcock SA, et al. 2002. Association of BAFF/BLyS overexpression and altered B cell differentiation with Sjogren’s syndrome. J. Clin. Invest. 109:59–68 50. Armengol MP, Cardoso-Schmidt CB, Fernandez M, Ferrer X, Pujol-Borrell R, Juan M. 2003. Chemokines determine local lymphoneogenesis and a reduction of circulating CXCR4+ T and CCR7 B and T lymphocytes in thyroid autoimmune diseases. J. Immunol. 170:6320–28 51. Weyand CM, Kurtin PJ, Goronzy JJ. 2001. Ectopic lymphoid organogenesis: a fast track for autoimmunity. Am. J. Pathol. 159:787–93

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

488

Martin

·

Chan

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

52. Aust G, Sittig D, Becherer L, Anderegg U, Schutz A, et al. 2004. The role of CXCR5 and its ligand CXCL13 in the compartmentalization of lymphocytes in thyroids affected by autoimmune thyroid diseases. Eur. J. Endocrinol. 150:225–34 53. Peterson KS, Huang JF, Zhu J, D’Agati V, Liu X, et al. 2004. Characterization of heterogeneity in the molecular pathogenesis of lupus nephritis from transcriptional proﬁles of laser-captured glomeruli. J. Clin. Invest.113:1722–33 54. Takemura S, Klimiuk PA, Braun A, Goronzy JJ, Weyand CM. 2001. T cell activation in rheumatoid synovium is B cell dependent. J. Immunol. 167:4710–18 55. Cooper MD, Kearney JF, Gathings WE, Lawton AR. 1980. Effects of anti-Ig antibodies on the development and differentiation of B cells. Immunol. Rev. 52:29–53 56. Chentouﬁ AA, Nizet Y, Havaux X, De La Parra B, Cormont F, et al. 1999. Differential effects of injections of anti-μ and anti-δ monoclonal antibodies on B-cell populations in adult mice: regulation of xenoreactive natural antibody-producing cells. Transplantation 68:1728–36 57. Chentouﬁ AA, Nizet Y, Havaux X, Nisol F, Bazin H, Latinne D. 2000. B-cell suppression in adult mice injected with anti-δ followed by anti-μ mAb. Cell. Immunol. 205:40–51 58. Finkelman FD, Holmes JM, Dukhanina OI, Morris SC. 1995. Cross-linking of membrane immunoglobulin D, in the absence of T cell help, kills mature B cells in vivo. J. Exp. Med. 181:515–25 59. Cerny A, Starobinski M, Hugin AW, Sutter S, Zinkernagel RM, Izui S. 1987. Treatment with high doses of anti-IgM prevents, but with lower doses accelerates autoimmune disease in (NZW x BXSB)F1 hybrid mice. J. Immunol. 138:4222–28 60. Cerny A, Sutter S, Bazin H, Hengartner H, Zinkernagel RM. 1988. Clearance of lymphocytic choriomeningitis virus in antibody- and B-cell-deprived mice. J. Virol. 62:1803–7 61. Cheever AW, Byram JE, Hieny S, von LF, Lunde MN, Sher A. 1985. Immunopathology of Schistosoma japonicum and S. mansoni infection in B cell depleted mice. Parasite Immunol. 7:399–413 62. Koyama M, Ishihara K, Karasuyama H, Cordell JL, Iwamoto A, Nakamura T. 1997. CD79α/CD79β heterodimers are expressed on pro-B cell surfaces without associated μ heavy chain. Int. Immunol. 9:1767–72 63. Nagata K, Nakamura T, Kitamura F, Kuramochi S, Taki S, et al. 1997. The Igα/Igβ heterodimer on μ-negative proB cells is competent for transducing signals to induce early B cell differentiation. Immunity 7:559–70 64. Zhang L, French RR, Chan HT, O’Keefe TL, Cragg MS, et al. 1995. The development of anti-CD79 monoclonal antibodies for treatment of B-cell neoplastic disease. Ther. Immunol. 2:191–202 65. Riley JK, Sliwkowski MX. 2000. CD20: a gene in search of a function. Semin. Oncol. 27:17–24 66. Uchida J, Lee Y, Hasegawa M, Liang Y, Bradney A, et al. 2004. Mouse CD20 expression and function. Int. Immunol. 16:119–29 67. Clynes RA, Towers TL, Presta LG, Ravetch JV. 2000. Inhibitory Fc receptors modulate in vivo cytoxicity against tumor targets. Nat. Med. 6:443–46 68. Uchida J, Hamaguchi Y, Oliver JA, Ravetch JV, Poe JC, et al. 2004. The innate mononuclear phagocyte network depletes B lymphocytes through Fc receptor-dependent mechanisms during anti-CD20 antibody immunotherapy. J. Exp. Med. 199:1659–69 69. Gong Q, Ou Q, Ye S, Lee WP, Cornelius J, et al. 2005. Importance of cellular microenvironment and circulatory dynamics in B cell immunotherapy. J. Immunol. 174:817–26 www.annualreviews.org • B Cell Depletion

489

ARI

15 February 2006

1:8

70. Hamaguchi Y, Uchida J, Cain DW, Venturi GM, Poe JC, et al. 2005. The peritoneal cavity provides a protective niche for B1 and conventional B lymphocytes during antiCD20 immunotherapy in mice. J. Immunol. 174:4389–99 71. Flynn JM, Byrd JC. 2000. Campath-1H monoclonal antibody therapy. Curr. Opin. Oncol. 12:574–81 72. Robak T. 2005. Alemtuzumab in the treatment of chronic lymphocytic leukemia. BioDrugs 19:9–22 73. Dullforce P, Sutton DC, Heath AW. 1998. Enhancement of T cell-independent immune responses in vivo by CD40 antibodies. Nat. Med. 4:88–91 74. Pearson TC, Trambley J, Odom K, Anderson DC, Cowan S, et al. 2002. Anti-CD40 therapy extends renal allograft survival in rhesus macaques. Transplantation 74:933–40 75. Dunussi-Joannopoulos K, Hancock GE, Kunz A, Hegen M, Zhou XX, et al. 2005. B-cell depletion inhibits arthritis in a collagen-induced arthritis (CIA) model but does not adversely affect humoral responses in a respiratory syncytial virus (RSV) vaccination model. Blood 106:2235–43 76. Waldor MK, Hardy RR, Hayakawa K, Steinman L, Herzenberg LA, Herzenberg LA. 1984. Disappearance and reappearance of B cells after in vivo treatment with monoclonal anti-I-A antibodies. Proc. Natl. Acad. Sci. USA 81:2855–58 77. Vladutiu AO, Steinman L. 1987. Inhibition of experimental autoimmune thyroiditis in mice by anti-I-A antibodies. Cell. Immunol. 109:169–80 78. Grabstein KH, Waldschmidt TJ, Finkelman FD, Hess BW, Alpert AR, et al. 1993. Inhibition of murine B and T lymphopoiesis in vivo by an anti-interleukin 7 monoclonal antibody. J. Exp. Med. 178:257–64 79. Giliani S, Mori L, de Saint BG, Le DF, Rodriguez-Perez C, et al. 2005. Interleukin-7 receptor α (IL-7Rα) deﬁciency: cellular and molecular bases. Analysis of clinical, immunological, and molecular features in 16 novel patients. Immunol. Rev. 203:110–26 80. Dahlenborg K, Pound JD, Gordon J, Borrebaeck CA, Carlsson R. 2000. Signals sustaining human immunoglobulin V gene hypermutation in isolated germinal centre B cells. Immunology 101:210–17 81. Pound JD, Gordon J. 1997. Maintenance of human germinal center B cells in vitro. Blood 89:919–28 82. Liu YJ, Banchereau J. 1997. Regulation of B-cell commitment to plasma cells or to memory B cells. Semin. Immunol. 9:235–40 83. Mehta DS, Wurster AL, Grusby MJ. 2004. Biology of IL-21 and the IL-21 receptor. Immunol. Rev. 202:84–95 84. Chtanova T, Tangye SG, Newton R, Frank N, Hodge MR, et al. 2004. T follicular helper cells express a distinctive transcriptional proﬁle, reﬂecting their role as non-Th1/Th2 effector cells that provide help for B cells. J. Immunol. 173:68–78 85. Vinuesa CG, Cook MC, Angelucci C, Athanasopoulos V, Rui L, et al. 2005. A RINGtype ubiquitin ligase family member required to repress follicular helper T cells and autoimmunity. Nature 435:452–58 86. Ozaki K, Spolski R, Ettinger R, Kim HP, Wang G, et al. 2004. Regulation of B cell differentiation and plasma cell generation by IL-21, a novel inducer of Blimp-1 and Bcl-6. J. Immunol. 173:5361–71 87. Jin H, Carrio R, Yu A, Malek TR. 2004. Distinct activation signals determine whether IL21 induces B cell costimulation, growth arrest, or Bim-dependent apoptosis. J. Immunol. 173:657–65 88. Moore KW, de Waal Malefyt R, Coffman RL, O’Garra A. 2001. Interleukin-10 and the interleukin-10 receptor. Annu. Rev. Immunol. 19:683–765

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

490

Martin

·

Chan

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

89. Asadullah K, Sterry W, Volk HD. 2003. Interleukin-10 therapy—review of a new approach. Pharmacol. Rev. 55:241–69 90. Asadullah K, Sabat R, Friedrich M, Volk HD, Sterry W. 2004. Interleukin-10: an important immunoregulatory cytokine with major impact on psoriasis. Curr. Drug Targets Inﬂamm. Allergy 3:185–92 91. Campbell KA, Ovendale PJ, Kennedy MK, Fanslow WC, Reed SG, Maliszewski CR. 1996. CD40 ligand is required for protective cell-mediated immunity to Leishmania major. Immunity 4:283–89 92. De Jong YP, Comiskey M, Kalled SL, Mizoguchi E, Flavell RA, et al. 2000. Chronic murine colitis is dependent on the CD154/CD40 pathway and can be attenuated by anti-CD154 administration. Gastroenterology 119:715–23 93. Grewal IS, Foellmer HG, Grewal KD, Xu J, Hardardottir F, et al. 1996. Requirement for CD40 ligand in costimulation induction, T cell activation, and experimental allergic encephalomyelitis. Science 273:1864 –67 94. Kalled SL, Cutler AH, Ferrant JL. 2001. Long-term anti-CD154 dosing in nephritic mice is required to maintain survival and inhibit mediators of renal ﬁbrosis. Lupus 10:9–22 95. Sidiropoulos PI, Boumpas DT. 2004. Lessons learned from anti-CD40L treatment in systemic lupus erythematosus patients. Lupus 13:391–97 96. Grammer AC, Slota R, Fischer R, Gur H, Girschick H, et al. 2003. Abnormal germinal center reactions in systemic lupus erythematosus demonstrated by blockade of CD154CD40 interactions. J. Clin. Invest. 112:1506–20 97. Ferrant JL, Benjamin CD, Cutler AH, Kalled SL, Hsu YM, et al. 2004. The contribution of Fc effector mechanisms in the efﬁcacy of anti-CD154 immunotherapy depends on the nature of the immune challenge. Int. Immunol. 16:1583–94 98. Oxenius A, Campbell KA, Maliszewski CR, Kishimoto T, Kikutani H, et al. 1996. CD40CD40 ligand interactions are critical in T-B cooperation but not for other anti-viral CD4+ T cell functions. J. Exp. Med. 183:2209–18 99. Aruffo A, Hollenbaugh D, Wu LH, Ochs HD. 1994. The molecular basis of X-linked agammaglobulinemia, hyper-IgM syndrome, and severe combined immunodeﬁciency in humans. Curr. Opin. Hematol. 1:12–18 100. Kalled SL. 2005. The role of BAFF in immune function and implications for autoimmunity. Immunol. Rev. 204:43–54 101. MacKay F, Schneider P, Rennert P, Browning J. 2003. BAFF and APRIL: a tutorial on B cell survival. Annu. Rev. Immunol. 21:231–64 102. Lesley R, Xu Y, Kalled SL, Hess DM, Schwab SR, et al. 2004. Reduced competitiveness of autoantigen-engaged B cells due to increased dependence on BAFF. Immunity 20:441–53 103. Xu LG, Wu M, Hu J, Zhai Z, Shu HB. 2002. Identiﬁcation of downstream genes upregulated by the tumor necrosis factor family member TALL-1. J. Leukoc. Biol. 72:410–16 104. Kayagaki N, Yan M, Seshasayee D, Wang H, Lee W, et al. 2002. BAFF/BLyS receptor 3 binds the B cell survival factor BAFF ligand through a discrete surface loop and promotes processing of NF-κB2. Immunity 17:515–24 105. Claudio E, Brown K, Park S, Wang H, Siebenlist U. 2002. BAFF-induced NEMOindependent processing of NF-κB2 in maturing B cells. Nat. Immunol. 3:958–65 106. Vora KA, Wang LC, Rao SP, Liu ZY, Majeau GR, et al. 2003. Cutting edge: germinal centers formed in the absence of B cell-activating factor belonging to the TNF family exhibit impaired maturation and function. J. Immunol. 171:547–51 107. Schiemann B, Gommerman JL, Vora K, Cachero TG, Shulga-Morskaya S, et al. 2001. An essential role for BAFF in the normal development of B cells through a BCMAindependent pathway. Science 293:2111–14 www.annualreviews.org • B Cell Depletion

491

ARI

15 February 2006

1:8

108. Thien M, Phan TG, Gardam S, Amesbury M, Basten A, et al. 2004. Excess BAFF rescues self-reactive B cells from peripheral deletion and allows them to enter forbidden follicular and marginal zone niches. Immunity 20:785–98 109. MacKay F, Woodcock SA, Lawton P, Ambrose C, Baetscher M, et al. 1999. Mice transgenic for BAFF develop lymphocytic disorders along with autoimmune manifestations. J. Exp. Med. 190:1697–710 110. Gross JA, Johnston J, Mudri S, Enselman R, Dillon SR, et al. 2000. TACI and BCMA are receptors for a TNF homologue implicated in B-cell autoimmune disease. Nature 404:995–99 111. Khare SD, Sarosi I, Xia XZ, McCabe S, Miner K, et al. 2000. Severe B cell hyperplasia and autoimmune disease in TALL-1 transgenic mice. Proc. Natl. Acad. Sci. USA 97:3370–75 112. Ramanujam M, Wang X, Huang W, Schiffer L, Grimaldi C, et al. 2004. Mechanism of action of transmembrane activator and calcium modulator ligand interactor-Ig in murine systemic lupus erythematosus. J. Immunol. 173:3524 –34 113. Liu W, Szalai A, Zhao L, Liu D, Martin F, et al. 2004. Control of spontaneous B lymphocyte autoimmunity with adenovirus-encoded soluble TACI. Arthritis Rheum. 50:1884 –96 114. Stohl W, Metyas S, Tan SM, Cheema GS, Oamar B, et al. 2003. B lymphocyte stimulator overexpression in patients with systemic lupus erythematosus: longitudinal observations. Arthritis Rheum. 48:3475–86 115. Reff ME, Carner K, Chambers KS, Chinn PC, Leonard JE, et al. 1994. Depletion of B cells in vivo by a chimeric mouse human monoclonal antibody to CD20. Blood 83:435–45 116. Vugmeyster Y, Beyer J, Howell K, Combs D, Fielder P, et al. 2005. Depletion of B cells by a humanized anti-CD20 antibody PRO70769 in Macaca fascicularis. J. Immunother. 28:212–19 117. Vugmeyster Y, Howell K, McKeever K, Combs D, Canova-Davis E. 2003. Differential in vivo effects of Rituximab on two B-cell subsets in cynomolgus monkeys. Int. Immunopharmacol. 3:1477–81 118. Baker KP, Edwards BM, Main SH, Choi GH, Wager RE, et al. 2003. Generation and characterization of LymphoStat-B, a human monoclonal antibody that antagonizes the bioactivities of B lymphocyte stimulator. Arthritis Rheum. 48:3253–65 119. Martin F, Kearney JF. 2002. Marginal-zone B cells. Nat. Rev. Immunol. 2:323–35 120. Rastetter W, Molina A, White CA. 2004. Rituximab: expanding role in therapy for lymphomas and autoimmune diseases. Annu. Rev. Med. 55:477–503 121. Avivi I, Robinson S, Goldstone A. 2003. Clinical use of Rituximab in haematological malignancies. Br. J. Cancer 89:1389–94 122. Silberstein LE. 1994. B-cell origin of cold agglutinins. Adv. Exp. Med. Biol. 347:193–205 123. Pascual M, Mach-Pascual S, Schifferli JA. 1997. Paraproteins and complement depletion: pathogenesis and clinical syndromes. Semin. Hematol. 34:40–48 124. Ward JH. 2001. Autoimmunity in chronic lymphocytic leukemia. Curr. Treat. Options Oncol. 2:253–57 125. Lee EJ, Kueck B. 1998. Rituxan in the treatment of cold agglutinin disease. Blood 92:3490– 91 126. Schiller G, Tillisch J, Rosen P. 1999. Use of anti-CD20 monoclonal antibody for the treatment of Evans syndrome and monoclonal gammopathy/neuropathy syndrome. Blood 94:84B (Abstr.) 127. Zaja F, Russo D, Fuga G, Patriarca F, Ermacora A, Baccarani M. 1999. Rituximab for the treatment of type II mixed cryoglobulinemia. Haematologica 84:1157–58 128. Treon SP, Anderson KC. 2000. The use of Rituximab in the treatment of malignant and nonmalignant plasma cell disorders. Semin. Oncol. 27:79–85

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

492

Martin

·

Chan

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

129. Robak T. 2004. Monoclonal antibodies in the treatment of autoimmune cytopenias. Eur. J. Haematol. 72:79–88 130. Raj A, Bertolone S, Cheerva A. 2004. Successful treatment of refractory autoimmune hemolytic anemia with monthly Rituximab following nonmyeloablative stem cell transplantation for sickle cell disease. J. Pediatr. Hematol. Oncol. 26:312–14 131. Zecca M, Nobili B, Ramenghi U, Perrotta S, Amendola G, et al. 2003. Rituximab for the treatment of refractory autoimmune hemolytic anemia in children. Blood 101:3857–61 132. Quartier P, Brethon B, Philippet P, Landman-Parker J, Le Deist F, et al. 2001. Treatment of childhood autoimmune haemolytic anaemia with Rituximab. Lancet 358:1511–13 133. Saleh MN, Gutheil J, Moore M, Bunch PW, Butler J, et al. 2000. A pilot study of the anti-CD20 monoclonal antibody Rituximab in patients with refractory immune thrombocytopenia. Semin. Oncol. 27:99–103 134. Stasi R, Pagano A, Stipa E, Amadori S. 2001. Rituximab chimeric anti-CD20 monoclonal antibody treatment for adults with chronic idiopathic thrombocytopenic purpura. Blood 98:952–57 135. Zaja F, Vianelli N, Sperotto A, De Vita S, Iacona I, et al. 2003. B-cell compartment as the selective target for the treatment of immune thrombocytopenias. Haematologica 88:538–46 136. Cooper N, Stasi R, Cunningham-Rundles S, Feuerstein MA, Leonard JP, et al. 2004. The efﬁcacy and safety of B-cell depletion with anti-CD20 monoclonal antibody in adults with chronic immune thrombocytopenic purpura. Br. J. Haematol. 125:232–39 137. Samuelsson A, Towers TL, Ravetch JV. 2001. Anti-inﬂammatory activity of IVIG mediated through the inhibitory Fc receptor. Science 291:484–86 138. Levine TD, Pestronk A. 1999. IgM antibody-related polyneuropathies: B-cell depletion chemotherapy using Rituximab. Neurology 52:1701–4 139. Pestronk A, Florence J, Miller T, Choksi R, Al Lozi MT, Levine TD. 2003. Treatment of IgM antibody associated polyneuropathies using Rituximab. J. Neurol. Neurosurg. Psychiatry 74:485–89 140. Rojas-Garcia R, Gallardo E, de Andres I, de Luna N, Juarez C, et al. 2003. Chronic neuropathy with IgM anti-ganglioside antibodies: lack of long term response to Rituximab. Neurology 61:1814–16 141. Renaud S, Gregor M, Fuhr P, Lorenz D, Deuschl G, et al. 2003. Rituximab in the treatment of polyneuropathy associated with anti-MAG antibodies. Muscle Nerve 27:611– 15 142. Nobile-Orazio E. 2004. IgM paraproteinaemic neuropathies. Curr. Opin. Neurol. 17:599– 605 143. Wiestner A, Cho HJ, Asch AS, Michelis MA, Zeller JA, et al. 2002. Rituximab in the treatment of acquired factor VIII inhibitors. Blood 100:3426–28 144. Keogh KA, Wylam ME, Stone JH, Specks U. 2005. Induction of remission by B lymphocyte depletion in eleven patients with refractory antineutrophil cytoplasmic antibodyassociated vasculitis. Arthritis Rheum. 52:262–68 145. Eriksson P. 2005. Nine patients with anti-neutrophil cytoplasmic antibody-positive vasculitis successfully treated with Rituximab. J. Intern. Med. 257:540–48 146. Stegeman CA, Tervaert JW, Sluiter WJ, Manson WL, de Jong PE, Kallenberg CG. 1994. Association of chronic nasal carriage of Staphylococcus aureus and higher relapse rates in Wegener granulomatosis. Ann. Intern. Med. 120:12–17 147. Shaw T, Quan J, Totoritis MC. 2003. B cell therapy for rheumatoid arthritis: the Rituximab (anti-CD20) experience. Ann. Rheum. Dis. 62(Suppl. 2):ii55–59 www.annualreviews.org • B Cell Depletion

493

ARI

15 February 2006

1:8

148. Edwards JC, Szczepanski L, Szechinski J, Filipowicz-Sosnowska A, Emory P, et al. 2004. Efﬁcacy of B-cell-targeted therapy with rituximab in patients with rheumatoid arthritis. N. Engl. J. Med. 350:2572–81 149. Cambridge G, Leandro MJ, Edwards JC, Ehrenstein MR, Salden M, et al. 2003. Serologic changes following B lymphocyte depletion therapy for rheumatoid arthritis. Arthritis Rheum. 48:2146–54 150. De Vita S, Zaja F, Sacco S, De Candia A, Fanin R, Ferraccioli G. 2002. Efﬁcacy of selective B cell blockade in the treatment of rheumatoid arthritis: evidence for a pathogenetic role of B cells. Arthritis Rheum. 46:2029–33 151. Tuscano JM. 2002. Successful treatment of inﬂiximab-refractory rheumatoid arthritis with Rituximab. Arthritis Rheum. 46:3420 152. Weyand CM, Goronzy JJ. 2003. Ectopic germinal center formation in rheumatoid synovitis. Ann. NY Acad. Sci. 987:140–49 153. Leandro MJ, Edwards JC, Cambridge G, Ehrenstein MR, Isenberg DA. 2002. An open study of B lymphocyte depletion in systemic lupus erythematosus. Arthritis Rheum. 46:2673–77 154. Anolik JH, Campbell D, Felgar RE, Young F, Sanz I, et al. 2003. The relationship of FcγRIIIa genotype to degree of B cell depletion by Rituximab in the treatment of systemic lupus erythematosus. Arthritis Rheum. 48:455–59 155. Eisenberg R. 2003. Mechanisms of autoimmunity. Immunol. Res. 27:203–18 156. Looney RJ, Anolik JH, Campbell D, Felgar RE, Young F, et al. 2004. B cell depletion as a novel treatment for systemic lupus erythematosus: a phase I/II dose-escalation trial of Rituximab. Arthritis Rheum. 50:2580–89 157. Tomietto P, Gremese E, Tolusso B, Venturini P, De VS, Ferraccioli G. 2004. B cell depletion may lead to normalization of anti-platelet, anti-erythrocyte and antiphospholipid antibodies in systemic lupus erythematosus. Thromb. Haemost. 92:1150–53 158. Reiff A. 2005. A review of Campath in autoimmune disease: biologic therapy in the gray zone between immunosuppression and immunoablation. Hematology 10:79–93 159. Isaacs JD, Manna VK, Rapson N, Bulpitt KJ, Hazleman BL, et al. 1996. CAMPATH-1H in rheumatoid arthritis—an intravenous dose-ranging study. Br. J. Rheumatol. 35:231–40 160. Weinblatt ME, Maddison PJ, Bulpitt KJ, Hazleman BL, Urowitz MB, et al. 1995. CAMPATH-1H, a humanized monoclonal antibody, in refractory rheumatoid arthritis. An intravenous dose-escalation study. Arthritis Rheum. 38:1589–94 161. Coles AJ, Wing M, Smith S, Coraddu F, Greer S, et al. 1999. Pulsed monoclonal antibody treatment and autoimmune thyroid disease in multiple sclerosis. Lancet 354:1691–95 162. Willis F, Marsh JC, Bevan DH, Killick SB, Lucas G, et al. 2001. The effect of treatment with Campath-1H in patients with autoimmune cytopenias. Br. J. Haematol. 114:891–98 163. Marsh JC, Gordon-Smith EC. 2001. CAMPATH-1H in the treatment of autoimmune cytopenias. Cytotherapy 3:189–95 164. Leonard JP, Coleman M, Ketas JC, Chadburn A, Ely S, et al. 2003. Phase I/II trial of epratuzumab (humanized anti-CD22 antibody) in indolent non-Hodgkin’s lymphoma. J. Clin. Oncol. 21:3051–59 165. Juweid M. 2003. Technology evaluation: epratuzumab, Immunomedics/Amgen. Curr. Opin. Mol. Ther. 5:192–98 166. Huang W, Sinha J, Newman J, Reddy B, Budhai L, et al. 2002. The effect of anti-CD40 ligand antibody on B cells in human systemic lupus erythematosus. Arthritis Rheum. 46:1554–62

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

494

Martin

·

Chan

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-15

ARI

15 February 2006

1:8

167. Grammer AC, Lipsky PE. 2002. CD154-CD40 interactions mediate differentiation to plasma cells in healthy individuals and persons with systemic lupus erythematosus. Arthritis Rheum. 46:1417–29 168. Mariette X, Roux S, Zhang J, Bengoufa D, Lavie F, et al. 2003. The level of BLyS (BAFF) correlates with the titre of autoantibodies in human Sjogren’s syndrome. Ann. Rheum. Dis. 62:168–71 169. Lavie F, Miceli-Richard C, Quillard J, Roux S, Leclerc P, Mariette X. 2004. Expression of BAFF (BLyS) in T cells inﬁltrating labial salivary glands from patients with Sjogren’s syndrome. J. Pathol. 202:496–502 170. Hamaguchi Y, Uchida J, Cain DW, Venturi GM, Poe JC, et al. 2005. The peritoneal cavity provides a protective niche for B1 and conventional B lymphocytes during antiCD20 immunotherapy in mice. J. Immunol. 174:4389–99 171. Cragg MS, Walshe CA, Ivanov AO, Glennie MJ. 2005. The biology of CD20 and its potential as a target for mAb therapy. Curr. Dir. Autoimmun. 8:140–74 172. Teeling JL, French RR, Cragg MS, van den Brakel J, Pluyter M, et al. 2004. Characterization of new human CD20 monoclonal antibodies with potent cytolytic activity against non-Hodgkin lymphomas. Blood 104:1793–800 173. Stanglmaier M, Reis S, Hallek M. 2004. Rituximab and alemtuzumab induce a nonclassic, caspase-independent apoptotic pathway in B-lymphoid cell lines and in chronic lymphocytic leukemia cells. Ann. Hematol. 83:634 –45 174. Jazirehi AR, Bonavida B. 2005. Cellular and molecular signal transduction pathways modulated by Rituximab (rituxan, anti-CD20 mAb) in non-Hodgkin’s lymphoma: implications in chemosensitization and therapeutic intervention. Oncogene 24:2121–43 175. Harjunpaa A, Junnikkala S, Meri S. 2000. Rituximab (anti-CD20) therapy of B-cell lymphomas: direct complement killing is superior to cellular effector mechanisms. Scand. J. Immunol. 51:634–41 176. Cardarelli PM, Quinn M, Buckman D, Fang Y, Colcher D, et al. 2002. Binding to CD20 by anti-B1 antibody or F(ab )(2) is sufﬁcient for induction of apoptosis in B-cell lines. Cancer Immunol. Immunother. 51:15–24 177. Cartron G, Dacheux L, Salles G, Solal-Celigny P, Bardos P, et al. 2002. Therapeutic activity of humanized anti-CD20 monoclonal antibody and polymorphism in IgG Fc receptor FcγRIIIa gene. Blood 99:754–58 178. Weng WK, Levy R. 2003. Two immunoglobulin G fragment C receptor polymorphisms independently predict response to Rituximab in patients with follicular lymphoma. J. Clin. Oncol. 21:3940–47 179. Shields RL, Lai J, Keck R, O’Connell LY, Hong K, et al. 2002. Lack of fucose on human IgG1 N-linked oligosaccharide improves binding to human FcγRIII and antibodydependent cellular toxicity. J. Biol. Chem. 277:26733–40 180. Friedberg JW, Neuberg D, Gribben JG, Fisher DC, Canning C, et al. 2002. Combination immunotherapy with Rituximab and interleukin 2 in patients with relapsed or refractory follicular non-Hodgkin’s lymphoma. Br. J. Haematol. 117:828–34 181. Gluck WL, Hurst D, Yuen A, Levine AM, Dayton MA, et al. 2004. Phase I studies of interleukin (IL)-2 and Rituximab in B-cell non-Hodgkin’s lymphoma: IL-2 mediated natural killer cell expansion correlations with clinical response. Clin. Cancer Res. 10:2253– 64 182. Olivieri A, Lucesole M, Capelli D, Gini G, Montanari M, et al. 2005. A new schedule of CHOP/Rituximab plus granulocyte-macrophage colony-stimulating factor is an effective rescue for patients with aggressive lymphoma failing autologous stem cell transplantation. Biol. Blood Marrow Transplant. 11:627–36 www.annualreviews.org • B Cell Depletion

495

ANRV270-IY24-15

ARI

15 February 2006

1:8

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

183. Niitsu N, Hayama M, Okamoto M, Khori M, Higashihara M, et al. 2004. Phase I study of Rituximab-CHOP regimen in combination with granulocyte colony-stimulating factor in patients with follicular lymphoma. Clin. Cancer Res. 10:4077–82 184. Mori Y, Shimizu N, Dallas M, Niewolna M, Story B, et al. 2004. Anti-α4 integrin antibody suppresses the development of multiple myeloma and associated osteoclastic osteolysis. Blood 104:2149–54 185. Huang YW, Richardson JA, Vitetta ES. 1995. Anti-CD54 (ICAM-1) has antitumor activity in SCID mice with human myeloma cells. Cancer Res. 55:610–16 186. Olson DL, Burkly LC, Leone DR, Dolinski BM, Lobb RR. 2005. Anti-α4 integrin monoclonal antibody inhibits multiple myeloma growth in a murine model. Mol. Cancer Ther. 4:91–99

496

Martin

·

Chan

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:467-496. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-16

ARI

15 February 2006

1:25

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

The Evolution of Adaptive Immunity Zeev Pancer1 and Max D. Cooper2 1

Center of Marine Biotechnology, University of Maryland Biotechnology Institute, Baltimore, Maryland 21202; email: [email protected]

2

Howard Hughes Medical Institute, University of Alabama, Birmingham, Alabama 35294; email: [email protected]

Annu. Rev. Immunol. 2006. 24:497–518

Key Words

First published online as a Review in Advance on January 16, 2006

invertebrate, vertebrate, agnatha, gnathostome, innate immunity, variable lymphocyte receptors (VLRs), leucine-rich repeat (LRR)–containing proteins

The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090542 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0497$20.00

Abstract Approximately 500 mya two types of recombinatorial adaptive immune systems appeared in vertebrates. Jawed vertebrates generate a diverse repertoire of B and T cell antigen receptors through the rearrangement of immunoglobulin V, D, and J gene fragments, whereas jawless ﬁsh assemble their variable lymphocyte receptors through recombinatorial usage of leucine-rich repeat (LRR) modular units. Invariant germ line–encoded, LRR-containing proteins are pivotal mediators of microbial recognition throughout the plant and animal kingdoms. Whereas the genomes of plants and deuterostome and chordate invertebrates harbor large arsenals of recognition receptors primarily encoding LRR-containing proteins, relatively few innate pattern recognition receptors sufﬁce for survival of pathogeninfected nematodes, insects, and vertebrates. The appearance of a lymphocyte-based recombinatorial system of anticipatory immunity in the vertebrates may have been driven by a need to facilitate developmental and morphological plasticity in addition to the advantage conferred by the ability to recognize a larger portion of the antigenic world.

497

ANRV270-IY24-16

ARI

15 February 2006

1:25

INTRODUCTION PAMPs: pathogen-associated molecular patterns PRR: pattern recognition receptor LRR: leucine-rich repeat

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

VLR: variable lymphocyte receptor

498

The earth biomass consists primarily of microorganisms, many of which are pathogens capable of killing and converting other organisms into copies of themselves. In response to this threat, eukaryotes have constantly evolved antipathogen devices. In turn, microorganisms continually evolve new ways to evade host defense tactics in what has been called the host-versus-pathogen arms race. The ﬁrst line of host responses to pathogen invasion is the innate immune defenses. Innate immunity depends on germ line–encoded receptors that have evolved to recognize highly conserved pathogenassociated molecular patterns (PAMPs). These receptors have therefore been termed pattern recognition receptors (PRRs). In addition to the innate defense mechanisms, jawed vertebrates (gnathostomes) have evolved an adaptive immune system mediated primarily by lymphocytes. By virtue of rearrangeable immunoglobulin (Ig) V, D, and J gene segments, the jawed vertebrates generate a lymphocyte receptor repertoire of sufﬁcient diversity to recognize the antigenic component of any potential pathogen or toxin. All jawed vertebrates, beginning with cartilaginous ﬁsh, rearrange their V(D)J gene segments to assemble complete genes for the antigen receptors expressed by T and B lymphocytes. Antigen-mediated triggering of T and B cells initiates speciﬁc cell-mediated and humoral immune responses (1, 2). The Ig domains are an ancient protein superfamily, and in the adaptive immune system of jawed vertebrates, the IgV (variable) domains are the cardinal molecular elements of antigen receptors. In invertebrate animals, however, evidence of a role for Ig domains in pathogen recognition or self/nonself discrimination was ﬁrst reported for hemolin, a unique hemolymph protein of lepidopteran insects comprised of four Ig-like domains. Microbial challenge induces secretion of the hemolin protein, which can bind to bacteria and yeast (3). Remarkably, a diverse

Pancer

·

Cooper

repertoire of Ig domain–containing receptors is generated in insects through alternative splicing of the Downs syndrome adhesion molecule gene transcripts (3a). Other immunerelated, Ig-containing molecules have been reported in invertebrates, but their direct role in immune recognition has not been demonstrated. For instance, the freshwater snail, Biomphalaria glabrata, has a diverse family of ﬁbrinogen-related hemolymph proteins (FREPs) with one or two N-terminal Iglike domains. FREPs are expressed in increased abundance by circulating phagocytic cells, called hemocytes, following infection with trematode parasites, and they can bind to soluble trematode antigens. FREP genes in the hemocytes, central nervous system tissue, and stomach wall muscle may undergo some type of somatic diversiﬁcation (4). Another family of genes that encode IgV region– containing chitin-binding proteins (VCBPs) was identiﬁed ﬁrst in the cephalochordate amphioxus, Branchiostoma ﬂoridae (5), and later in the genome of the tunicate Ciona intestinalis (6). These amphioxus VCBP molecules are encoded by ﬁve or more multigene families that are polymorphic within the population. VCBP gene products are secreted into the intestine, where they may play a role in preventing microbial invasion. Evidence has recently been obtained that two very different recombinatorial systems for lymphocyte antigen receptor diversiﬁcation appeared at the dawn of vertebrate evolution ∼500 mya (Figure 1). Lamprey and hagﬁsh, which are the only surviving jawless ﬁsh (agnathans) belonging to the oldest vertebrate taxon (7, 8), have been found to assemble diverse lymphocyte antigen receptor genes through the genomic rearrangement of leucine-rich repeat (LRR)–encoding modules (9, 10). These cell surface receptors are designated variable lymphocyte receptors, or VLRs. Recombinatorial mechanisms for the generation of anticipatory receptors thus evolved in both the jawless and jawed vertebrates, but each vertebrate group employs a different kind of modular protein domain.

ANRV270-IY24-16

ARI

15 February 2006

1:25

Agnathan VLR gene LRR NT

LRR 1

LRR LRR 1 V

LRR LRR V V

SP 5`LRR 5`LRR 3`LRR Stalk NT CT CT

Rearrangement with flanking

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LRR SP NT

Gnathostome antibody genes 5`LRR CT

Lymphocyte incomplete V germ line genes

V

J J J

LRR CT

C

V

V

V

Stalk

Lymphocyte recombinatorial genes

V J C

J J J

V D J C

Cell-bound anticipatory receptors

Humoral effector molecules Figure 1 Rearranging antigen receptors of jawless and jawed vertebrates. Agnathan variable lymphocyte receptors (VLRs) are assembled by insertion of diverse LRR modules from ﬂanking genomic cassettes into the germ line incomplete VLR gene. The mature VLR gene consists of a signal peptide (SP), an N-terminal LRR (LRRNT), ﬁrst 18-residue LRRs (LRR1), a variable number of 24-residue LRRs (LRRV), a connecting peptide (CP), a C-terminal LRR (LRRCT), and a threonine/proline-rich stalk. Portions of LRRNT and LRRCT that are not encoded in the germ line VLRs are hatched. Jawed vertebrates antibody genes are assembled via random joining of Ig gene fragments consisting of variable (V), diversity (D), and joining ( J) elements as well as Ig constant (C) domains. Following somatic DNA rearrangement these antigen receptors are expressed on the surface of lymphocytes via GPI anchorage in VLRs or via a transmembrane domain in the antibody IgM cell surface form. Upon activation VLRs can be released to the plasma via GPI-speciﬁc phospholipase cleavage, while secreted antibodies result from isotype switching to the secretory forms.

The appearance of two types of recombinatorial immune systems within a relatively short evolutionary period of ∼40 million years during the Cambrian raises intriguing questions. What was the selective pressure to evolve acquired immunity? Why were LRRcontaining modules selected as the recombinatorial units of antigen receptors in agnathans, and why were Ig domains selected by the gnathostomes? The question of gnathostome Igs is not addressed here. On the other hand, the issue of LRR-based antigen receptors of jawless ﬁsh may be more easily addressed. Because LRR-containing proteins are ancient mediators of antimicrobial responses in both kingdoms of multicellular

D D

RAG-mediated rearrangment

LRR casettes

LRR LRR LRR LRR LRR LRR LRR CP V V V V V 1 V

Heavy chain

Light chain V

organisms, it is reasonable to suggest that the last common ancestor of plants and animals used some version(s) of LRR-containing proteins for microbial detection (11). LRRcontaining proteins therefore would have provided natural molecular candidates for early agnathan experimentation with somatic DNA rearrangement to achieve receptor gene diversiﬁcation. We begin this review with a consideration of the emergence of lymphocytes as a novel circulatory cell type in vertebrates and then consider phylogenetic aspects of the superfamily of LRR-containing proteins and their role in immunity. We conclude with an evolutionary scenario that may explain the sudden www.annualreviews.org • Evolution of Adaptive Immunity

499

C

ANRV270-IY24-16

ARI

15 February 2006

1:25

appearance of a lymphocyte-based recombinatorial system of anticipatory immunity in vertebrates.

MIGRATORY PHAGOCYTIC CELLS APPEARED BEFORE IMMUNOCOMPETENT LYMPHOCYTES

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Phagocytic cells form the cellular arm of innate immune defenses in almost all animals (metazoans) that have been studied, except for the nematode Caenorhabditis elegans, which may lack cellular immune defenses. Following pathogen invasion in C. elegans, there is activation of an inducible defense system marked by an increased expression of genes encoding lectins, antimicrobial peptides, and lysozymes, but exactly how the host perceives infection is not yet understood (12). In Drosophila melanogaster, another protostome invertebrate whose genome has been sequenced, plasmatocytes are the predominant phagocytic blood cells involved in clearance of invading microorganisms. The Drosophila plasmatocytes are considered the functional equivalents of monocytes/macrophages in the vertebrates (13, 14). Monocyte/macrophage-type cells have a relatively short life span in both invertebrates and vertebrates. Proliferation of these innate immunocytes appears to be conﬁned to the generative hematopoietic tissues. As mature circulatory cells, these nondividing phagocytes can be activated to become effector cells. These innate immunocytes may express a surprisingly large repertoire of surface receptors. As one example, a multigene family in the sea urchin Strongylocentrotus purpuratus (an echinoderm) encodes proteins featuring scavenger receptor cysteine-rich repeats. The circulatory coelomocytes of individual sea urchins express unique and temporally varying scavenger receptor cysteine-rich repertoires that are selected from an arsenal of hundreds of genes (15). It is therefore conceivable that diversiﬁcation mechanisms for immune receptors evolved before the ap500

Pancer

·

Cooper

pearance of long-lived circulatory immune cells that can undergo clonal expansion following ligand engagement of their unique receptors (16). Evidence suggesting somatic diversiﬁcation of the snail FREPs may support this view, although this diversiﬁcation process may not be limited to circulating hemocytes (4). Inevitably, self-reactivity among receptors generated by random somatic diversiﬁcation mechanisms would present a problem. In principle, though, autoimmunity could be avoided by a developmental program including a transitory interval of immunocyte hypersensitivity to receptor-mediated triggering, leading to apoptosis, or another type of inactivation mechanism for cells with autoreactive receptors. In this way, cells bearing nonselfreactive receptors could selectively continue their maturation to become migratory immunocytes.

THE APPEARANCE OF LYMPHOCYTES IN VERTEBRATES A new type of circulatory cell with the potential for self-renewal and clonal expansion appeared near the beginning of vertebrate radiation in the form of the long-lived lymphocyte. In the jawed vertebrates, T and B lymphocytes are the acknowledged cellular pillars of adaptive immunity. T lymphocytes are primarily responsible for cell-mediated immunity, and B lymphocytes are responsible for humoral immunity, but they work together and with other types of cells to mediate effective adaptive immunity. Along with the natural killer cells, these specialized lymphoid cells are derived from committed progenitors in hematopoietic tissues, which then undergo unique V(D)J rearrangements of their antigen receptors to become clonally diverse lymphocytes. Newly formed T and B lymphocytes bearing autoreactive receptors can be eliminated by self-antigen contact in the thymus and bone marrow, respectively. The surviving T and B cells then migrate via the bloodstream to peripheral lymphoid tissues, where,

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

following antigen recognition, they may undergo clonal expansion and differentiation into effector T lymphocytes or antibodyproducing plasma cells or otherwise become memory cells that await reexposure to their speciﬁc antigens. Exactly when during evolution the lymphocytes appeared as a specialized type of immunocompetent cells is unknown, but cells comparable to the lymphocytes in jawed vertebrates have never been characterized in invertebrates. On the other hand, increasing evidence for bona ﬁde lymphocytes in lamprey and hagﬁsh suggests that lymphocytes must have evolved in the common ancestor of the vertebrates. Most of the lymphoid cells in lamprey and hagﬁsh are small round cells composed mainly of a nucleus with condensed chromatin and a small rim of surrounding cytoplasm that contains relatively few organelles (17). Following antigen and/or mitogen stimulation, agnathan lymphocytes can transform into large lymphoblast-like cells (9). Morphological studies have led to the view that agnathan lymphocytes are generated in the hematopoietic tissues, primarily the intestineassociated hematopoietic organ in lamprey larvae, called the typhlosole, and the protovertebral arch of adult lamprey (17–19). Evidence for a thymus-like organ in agnathans is equivocal at best, however. Collections of lymphoid cells have been found in pharyngeal gutters of the lamprey gill region, but there is no recognizable capsular, stromal, or lymphoepithelial organization of the type that characterizes the lymphopoietic thymus in jawed vertebrates (17, 20–22). The lamprey lymphocytes express homologs of many genes expressed during jawed vertebrate lymphocyte differentiation, proliferation, migration, and intracellular signaling and perhaps also express the relatives of genes that gnathostomes use in antigen processing and intracellular transport of antigenic peptides (23–27). It is important to note that proteins with sequence similarity to the jawed vertebrate rearrangeable Ig genes or MHC genes have not been found in extensive surveys of

lamprey and hagﬁsh leukocyte transcripts (28, 29). Until very recently, there was no credible evidence for lymphocyte receptor diversity in lamprey or hagﬁsh. This led to considerable skepticism about the earlier reports of agnathan adaptive immunity (17, 20, 30–32). This picture changed dramatically with the identiﬁcation of VLR genes in the lamprey and hagﬁsh (9, 10). These genes are assembled by a special recombinatorial mechanism used to generate a diverse repertoire of anticipatory receptors. The lymphocytes of lamprey and hagﬁsh rearrange modular LRR cassettes to create functional mature VLR genes. A VLR of unique sequence is expressed by each lymphocyte in a monoallelic fashion. As in the case for the gnathostome lymphocytes, the agnathan lymphocytes may undergo lymphoblastoid transformation following antigen and/or mitogen stimulation. Clonal ampliﬁcation appears to occur during antigen-induced proliferative responses in lamprey and hagﬁsh, although more experimental evidence will be required to conﬁrm this conclusion. Lamprey lymphocytes can respond to immunization by the release of their antigen-speciﬁc VLRs into the plasma, thus providing the potential basis for humoral immunity (M.N. Alder, M.D. Cooper & Z. Pancer, unpublished data). Although many questions about the development and function of agnathan lymphocytes are still unanswered, it is clear that the jawless vertebrates have a lymphocyte-based recombinatorial immune system that differs radically from the Ig-based recombinatorial immune system in the jawed vertebrates.

IMMUNE-RELATED LRR PROTEINS OF INVERTEBRATES AND PLANTS Drosophila has two distinct families of PRRs that are used selectively to activate one or the other NFκB-like signaling pathways, the Toll or the Immune Deﬁciency. Cellular activation via these pathways results in induction www.annualreviews.org • Evolution of Adaptive Immunity

501

ANRV270-IY24-16

ARI

15 February 2006

1:25

Gnathostomes (jawed vertebrates)

Cephalochordates (Amphioxus) Urochordates (ascidians)

Hemichordates (acorn worms)

Figure 2 Metazoan tree of life. A simpliﬁed representation of the two major groups of multicellular animals, protostomes and deuterostomes, that include the chordate invertebrates and vertebrates.

Annelids Molluscs

Metazoans

Nematodes Arthropods

PROTOSTOMES

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Echinoderms

DEUTEROSTOMES

Agnathans (jawless vertebrates)

Cnidarians

of antimicrobial peptide genes in the fat body and the secretion of antimicrobial peptides into the hemolymph (33). It was ﬁrst realized in 1996 that the dorsoventral patterning receptor Toll is also involved in protection of ﬂies against fungi, by way of inducing expression of the antifungal peptide gene Drosomycin (34). The Toll pathway has recently been shown to be essential for protection against the Drosophila X virus as well (35). The Toll receptor has an extracellular LRR-containing domain, a transmembrane region, and a cytoplasmic Toll/interleukin-1 receptor homology domain (TIR). In addition to Toll, the Drosophila genome contains eight Toll homologs, and the mosquito Anopheles gambiae genome contains ten Toll homologs. All but one of these Tolls in both insects appear to be linked to developmental functions rather than to immunity (36). Likewise, immune function could not be attributed to the single C. elegans and Caenorhabditis briggsae

TLR: Toll-like receptor

502

Pancer

·

Cooper

Toll homolog, nor to the Toll-like receptor (TLR) reported in the horseshoe crab Tachypleus tridentatus (37). In striking contrast to the TLR saga in these protostome invertebrates, the vertebrate homologs of Drosophila Toll appear to be dedicated solely to host defense (38, 39). This leads us to consider what is known about the LRR-containing proteins in deuterostome invertebrates, which are in the ancestral evolutionary lineage of the vertebrates (Figure 2). It is known that TLR polypeptides are frequently encoded by a continuous open reading frame that is uninterrupted by introns. Standard examples of this type of gene structure include Tollo, Toll 6, and Toll 7 from Drosophila and human TLRs 1, 2, 4D, 5, 6, and 10 (intronless genes database) (40). The genome of the sea urchin S. purpuratus abounds with intronless LRR-containing genes. A sample of 52 intronless TLRs, derived from those identiﬁed

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

in the draft genome sequence (Sequencing Project Version 0.3, Baylor College of Medicine; http://www.hgsc.bcm.tmc.edu), is illustrated in Figure 3. These TLRs cluster in a very unusual branching pattern consisting of sequences with nearly identical values of genetic distance among branch members (Figure 4), thereby indicating multiple events of expansion and diversiﬁcation among branch members. The total number of sea urchin TLRs is estimated to be ∼340 members, or half this number if all these genes prove to be polymorphic (41). Cloning and expression analysis indicate that multiple TLRs may function in sea urchin immunity (Z. Pancer & E.H. Davidson, unpublished data). In addition to the TLR genes, the sea urchin genome harbors other intronless LRR-encoding genes. One group with at least 47 members consists solely of LRR motifs (not shown); some of these proteins have transmembrane domains (N = 13/47). Another group also has C-terminal Ig domains (N = 10), and some of these are also predicted to have membrane anchorage domains (N = 6/10) (Z. Pancer & M.D. Cooper, unpublished data). In the chordate phylum, the genome of the solitary tunicate Ciona savignyi contains between 7 and 19 TLR genes (41), only 2 of which are intronless genes, rendering gene prediction uncertain. At least 22 other intronless LRR-containing genes were identiﬁed in the C. savignyi genome (Sequencing Project data, April 25, 2003, version, Whitehead Institute and MIT Center for Genome Research; http://www-genome. wi.mit.edu), and 8 of these appear to be cell surface receptors (Z. Pancer & M.D. Cooper, unpublished data). Only three TLRs were reported in the genome of another solitary tunicate, C. intestinalis (6), and all of these are interrupted by introns. Notably, no other intronless LRR genes could be identiﬁed in this species. In the amphioxus B. ﬂoridae, we identiﬁed at least 42 intronless TLR genes (trace archive, WGS Sequencing Project, DOE

Joint Genome Institute; ftp://ftp.ensembl. org/pub/traces/branchiostoma floridae/). These TLR genes cluster in equidistant branches of the genetic distance tree, like TLRs from the sea urchin. In addition, the B. ﬂoridae genome harbors at least 211 intronless LRR-containing genes, or half this number if all alleles are polymorphic. Seventy-one of these LRR-containing genes are illustrated in Figure 5. Fifty-one consist only of LRR motifs, and 12 of these include transmembrane domains; the other 20 have both LRR and C-terminal Ig-like domains, and 12 of these are predicted cell surface proteins (Z. Pancer & M.D. Cooper, unpublished data). Computational analysis indicates that several more of the sea urchin and amphioxus intronless LRR-containing genes may encode proteins that are tethered to the cell surface via glycosyl-phosphatidyl-inositol (GPI) anchors. Intriguing unanswered questions abound: Why is the solitary tunicate C. intestinalis different from other deuterostome invertebrates that utilize multiple LRR-containing proteins? What may be the strategy employed by colonial tunicates, such as Botryllus schlosseri, that are known for their highly elaborate and polymorphic self/nonself-recognition systems (42)? Plant genomes harbor very large families of LRR-containing genes, and many of these mediate disease resistance. The most important plant disease resistance genes encode the STAND ATPase domain (43) or nucleotide-binding site (NBS)–LRR proteins, some of which include N-terminal TIR domains. There are also the LRR receptor– like kinases and the membrane-bound LRR receptor–like proteins. Several of these proteins control resistance to a wide variety of plant pathogens and pests, including viruses, bacteria, fungi, nematodes, and insects (44, 45). In response to pathogen challenge, diverse resistance responses in plants are activated by disease resistance proteins. These responses include the production of antimicrobial peptides and a form of programmed www.annualreviews.org • Evolution of Adaptive Immunity

503

ARI

15 February 2006

1:25

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

Figure 3 A large superfamily of sea urchin TLRs. A sample of 52 intronless TLR genes identiﬁed in the genome of Strongylocentrotus purpuratus. Prediction of domain architecture was via the SMART server (http://smart.embl-heidelberg.de). (N-terminal LRR: light blue rectangle; LRR: green rectangle; C-terminal LRR: light blue oval; transmembrane domain: dark blue rectangles; C-terminal TIR domains: green diamond.) 504

Pancer

·

Cooper

TL R.3 72 Sp TL R. 19 Sp 7 TL R. 27 7

R.7

Sp

R.260

SpT L

258

3 R.1 TL

L R. SpT

Sp

SpT

R.

6

Sp TL

R.

TL R. 39

0.1

5 44

TL

1 .2 LR

Sp

R. TL Sp

T Sp

Sp

1:25

SpTL

15 February 2006

SpTLR.44 8

ARI

SpTLR.106

ANRV270-IY24-16

T Sp

28

35

8

28 .1 R L

Sp

7

0 .24

LR SpT

.94

74

Sp T

LR.1

22

SpTLR .182

.262 SpTLR

SpTLR.308

SpTLR.410

SpTLR.214

SpTLR.198 SpTLR .37

.11 SpTLR

LR.3 SpT

4

SpT LR.3

20

SpT

4 .40

LR

Sp

6

7 .6 3

SpTLR.251

SpTLR.28

R1 .2 SpT LR. 201 SpTL R.151

1

136

Sp

33

TL

11 0

R.

.1 9

.1

TL

R.

R1

Sp

TL R

TL

Sp

R.

Sp

TL

.10

R TL

R.21

SpTL

LR. SpT

Sp

TL

Sp

S

LR pT

Sp

14

TL

Sp

R. 23

Sp

12 R.

R.

R

TL Sp

LR

.35

TL

.27

TL

SpT

4 SpTLR.16

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LR.2

1

29 R.

TL Sp

LR

SpT

8

.14

R TL

Figure 4 Genetic distance among 52 sea urchin intronless TLRs. UPGMA tree constructed from the predicted amino acid sequence using the pairwise deletion option (MEGA 3 Molecular Evolutionary Genetics Analysis) (99).

cell death called the hypersensitive response (46). In rice, Oryza sativa, 585 predicted NBSLRR genes account for approximately 1% of the genes identiﬁed in the genome; a similar fraction of the Arabidopsis thaliana genome is dedicated to disease resistance genes (44, 45).

Even with knowledge of this large arsenal of disease resistance genes, we still do not understand how plants can detect the multitude of infectious pathogens with a limited number of PRR genes. The “guard” hypothesis postulates that resistance proteins constitute www.annualreviews.org • Evolution of Adaptive Immunity

505

29

5

3

41

9

ARI

15 February 2006

1:25

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

Figure 5 A large superfamily of amphioxus intronless LRR-containing genes. A sample of 71 genes identiﬁed in the genome of the Florida lancelet Branchiostoma ﬂoridae. Prediction of domain architecture was via the SMART server (http://smart.embl-heidelberg.de). N-terminal LRR: light blue rectangle; LRR: green rectangle; C-terminal LRR: light blue oval; Ig superfamily domain: green oval; transmembrane domain: dark blue rectangle. 506

Pancer

·

Cooper

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16 ARI 15 February 2006 1:25

Figure 5

(Continued )

www.annualreviews.org • Evolution of Adaptive Immunity

507

ARI

15 February 2006

1:25

components of larger signal perception complexes that are activated in response to pathogen-induced perturbance of the normal function of host proteins. Hence, instead of recognition via a speciﬁc receptor for each pathogen, detection may occur indirectly via the damage inﬂicted by pathogen-derived virulence proteins (45–47). High levels of polymorphism have been noted among members of plant resistance genes, and the putative ligand-binding surfaces in many families of LRR-containing resistance genes appear to be undergoing rapid diversifying selection (48). Interestingly, there is also evidence of pronounced selection for somatic variants of disease resistance genes that may affect the ligand speciﬁcities of particular resistance proteins (49). This variation may be generated via an unknown diversiﬁcation mechanism of plant PRR genes, or it may reﬂect the increased frequency in homologous recombination that has been observed for virus-infected plants (50). Intronless LRR-containing genes are relatively rare in animal genomes other than deuterostome and chordate invertebrates. Apart from the few intronless TLRs in mammals and insects, there are only 10 other intronless LRR-containing genes in the human and mouse genomes, none in Drosophila, and only 25 of the NBS-LRR genes in Arabidopsis (intronless genes database) (40). Intronless genes most likely result from retroposition and subsequent genomic integration thought to occur via the reverse transcription activity of endogenous retrotransposons, such as the human LINE elements (51). Intronless genes may be excellent gene templates to generate rapidly evolving arsenals of diverse germ line– encoded receptors.

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

IMMUNE-RELATED LRR PROTEINS OF JAWED VERTEBRATES The typical TLR complement for vertebrates is approximately one dozen genes. The only exception that has been noted thus far is ﬁsh 508

Pancer

·

Cooper

that have retained both copies of duplicated TLRs resulting from a whole genome duplication that seems to have occurred after the divergence of bony ﬁsh and tetrapods. For example, at least 17 predicted TLR genes were expressed in the zebraﬁsh, Danio rerio (52). Nearly all vertebrate TLRs belong to one of six major families (TLR1–6), and each TLR family is capable of recognizing a general class of PAMPs. TLR2 family members bind lipopeptide; TLR3 family members bind double-stranded RNA; TLR4 family members bind LPS; TLR5 family members bind ﬂagellin; members of the TLR7, TLR8, and TLR9 subfamilies bind nucleic acid and heme motifs; and TLR1 family members associate with TLR2 members as heterodimeric receptors (41). Mouse TLR11 has recently been implicated in the response to a proﬁlin-like protein of the protozoan parasite Toxoplasma gondii (53). Soluble TLR forms, consisting of the extracellular portions only, may also participate in immunity. Amphibians and ﬁsh have a soluble form of the TLR5 gene that arose by duplication of the region encoding the extracellular domain. In rainbow trout, Onchorhynchus mikiss, bacterial ﬂagellin interacts with membrane-bound TLR5 to induce the expression of the soluble TLR5 gene in the liver, resulting in efﬁcient clearance of ﬂagellin from the circulation (54). In chicken, alternatively spliced forms of TLR3 and TLR5 yield soluble products (55), and human plasma and breast milk also contain a functional soluble form of TLR2 that is generated by a posttranslational modiﬁcation (56). The recently identiﬁed CATERPILLER family of LRR-containing immune-regulatory genes encodes cytoplasmic proteins that are structurally similar to some of the plant disease resistance genes (43). The N terminus of these proteins may function as an effector domain, mediating homotypic or heterotypic interactions; a central NBS domain has regulatory function, whereas the C terminus is composed of variable sets of LRR motifs that may function in ligand binding.

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

Heterodimers formed between different family members may in some cases increase the combinatorial binding potential (57). Nucleotide oligomerization domain (NOD) 1 and NOD2 members of this family are implicated as sensors of intracellular bacterial products and activators of host responses against invading pathogens (57, 58). In gut regions that are rich in commensal bacteria, NOD1 functions as the detector of enteroinvasive bacteria that have evolved the means to prevent intestinal epithelial cell signaling through the TLRs. NOD2 is critical for regulation of bacterial immunity within the intestine by controlling the expression of cryptdins, which are intestinal antimicrobial peptides (59). More than 20 members of this family have been identiﬁed in mammalian genomes, whereas no CATERPILLER orthologs have been identiﬁed in Drosophila or C. elegans (57). One strategy that is employed to maintain balance in the arms race between insects and pathogens involves natural selection of random mutations in insect PRR genes. In wild Drosophila populations, for example, ∼10% of the polymorphic sites in genes encoding antibacterial peptides, Toll receptors, signal transduction molecules, and other pathogen recognition molecules are associated with disease resistance phenotypes to a single pathogen (60). However, the vertebrate TLRs are not rapidly evolving genes. They appear to be under strong purifying selection to maintain their PAMP recognition speciﬁcity in order to discriminate between pathogens and the host. This may be because selfreactive TLRs would be detrimental to the host, as none of the self-tolerance mechanisms that can purge self-reactive T and B lymphocytes or prevent development of potentially harmful natural killer cells have been identiﬁed for cells bearing TLRs (41). The large multigene families of LRR-containing proteins in deuterostome invertebrates are especially remarkable given that microbial recognition is served by only a handful of PRRs in nematodes, insects, and vertebrates (61).

STRUCTURE AND FUNCTION OF LRR-CONTAINING PROTEINS LRRs of 20–29 amino acids per repeat are present in more than 2000 proteins from viruses, bacteria, archaea, and eukaryotes. Family members of the LRR-containing proteins participate in nearly all known biological functions, including plant and animal immunity, apoptosis, cell adhesion, signal transduction, DNA repair, DNA recombination and transcription, RNA processing, and ice nucleation. Nonetheless, the existence of many types of LRR-containing immune gene families in the genomes of both plants and animals argues for the special role of these proteins in host defense. Sixteen crystallized LRR-containing proteins all adopt an arc or horseshoe-like shape, with the individual LRR motifs forming parallel loops that are stacked into a coil. Most of the LRR-containing proteins have characteristic N-terminal and C-terminal LRR domains capping the ends of the hollow tube. The concave face of the coil consists of a parallel β-sheet, whereas there may be α-, 310 -, or pII helices in the convex face (62). Ligand-binding sites have been determined for several LRR-containing receptors. The mammalian ribonuclease inhibitor, the ﬁrst LRR structure solved, interacts with the ribonuclease via multiple contact points located on the concave LRR surface of the inhibitor (63). Glycoprotein Ib, a platelet LRRcontaining receptor for the von Willebrand factor, binds its ligand via exposed residues on the concave LRR face and via a ﬁngerlike insertion in the C-terminal LRR (64). CD14, one of the major LPS receptor units, is an LRR-containing protein expressed on myelomonocytic cells as a GPI-linked glycoprotein or released into the plasma as a soluble form. The crystal structure of CD14 reveals a dimer in which each horseshoeshaped monomer consists of 13 β-strands. The large hydrophobic pocket is located on the side of the horseshoe near the N terminus

www.annualreviews.org • Evolution of Adaptive Immunity

509

ARI

15 February 2006

1:25

(65). The polygalacturonase-inhibiting protein of Phaseolus vulgaris is the product of a plant disease resistance gene. It acts as an inhibitor of cell wall–degrading enzymes produced by pathogenic fungi. A negatively charged surface on the concave LRR face is probably involved in binding fungal polygalacturonases (66). The tomato, Lycopersicon pimpinellifolium, Cf-9 protein confers fungal resistance via residues in its N-terminal LRR; putative glycosylation sites in the outer α-helices are also essential for binding (67). Mammalian receptors for the glycoprotein hormones (thyrotropin, lutropin, chorionic gonadotropin, and follitropin) are G protein– coupled proteins that bind their ligands via the LRR motifs in their extracellular domains (68, 69). Internalins A and B of Listeria monocytogenes are LRR-containing surface proteins that mediate speciﬁc host-cell invasion by the bacteria. Internalin A mediates bacterial adhesion and initiates invasion of human intestinal epithelia through speciﬁc interaction with the E-cadherin receptor. The crystal structure of Internalin A complexed with the N-terminal domain of E-cadherin reveals tight interaction sites on the concave surface of the LRR coil (70). The N terminus of Internalin B consists of an LRR domain that is C-terminally capped by an Ig-like domain, and this portion is sufﬁcient to induce bacterial internalization into host cells (71). Promiscuity of ligand binding has been suggested in the case of Nogo and its LRRcontaining GPI-anchored receptor, which plays a key role in inhibition of mammalian axon regeneration. The Nogo receptor has a putative ligand-binding site within the concave LRR face in which multiple solventexposed hydrophobic and aromatic residues create high potential for binding crossreactivity (72). Decorin and Opticin are small LRR-containing extracellular matrix proteoglycans that form antiparallel homodimers via highly speciﬁc interactions at their concave LRR surfaces. It therefore seems likely that their binding of different ligands occurs

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

510

Pancer

·

Cooper

through protein surface sites other than those of the concave sheet (73, 74). The extracellular domains of TLRs consist of 19–25 tandem LRR motifs, most of which conform to a 24-residue consensus motif. Peptide insertions are present within some of the LRRs, and these have been predicted to mediate recognition of PAMPs (75). The only TLR crystal structure reported to date, the human TLR3 ectodomain, provides experimental support for this prediction (76). The TLR3 solenoid consists of 25 LRR motifs that are stacked and stabilized through hydrogen bonds formed by conserved asparagine residues. Glycosylation-free faces of the solenoid provide interfaces for a homodimeric conﬁguration maintained via conserved surface residues and a loop formed by a peptide insertion in LRR20, whereas a second peptide insertion in LRR12 and two clusters of positively charged residues form the putative binding site for double-stranded RNA. Some TLRs may recognize only a limited number of PAMPs. For instance, TLR9 directly interacts with particular sequences of unmethylated CpG-DNA found in bacterial DNA (77, 78), and TLR7 on the surface of plasmacytoid dendritic cells and B cells mediates the recognition of singlestranded RNA from vesicular stomatitis virus and inﬂuenza virus. Thus, TLR7 recognizes single-stranded RNA viruses, whereas either TLR3 or TLR9 detects double-stranded RNA viruses (79, 80). Other TLRs have the remarkable potential to interact with structurally unrelated ligands. TLR2 mediates host responses to peptidoglycan and lipoteichoic acid from Gram-positive bacteria, lipoarabinomannan from mycobacteria, neisserial porins, bacterial tripalmitoylated and mycoplasmal diacylated lipoproteins, and yeast products and GPI-anchored proteins of the protozoan Trypanosoma cruzi. Results from mutagenesis of the extracellular LRRs in TLR2 imply the existence of different binding sites for different ligands (81). TLR4 can bind LPS from Gram-negative bacteria, viral proteins, bacterial and host heat shock

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

proteins, host oligosaccharides derived from heperan sulfate and hyaluronic acid, host taxol, and ﬁbrinogen (75). In spite of their sequence and functional diversity, the horseshoe-like structure with its concave binding surface is remarkably conserved in all the LRR-containing proteins. This may reﬂect the modular structure formed by the tandem array of stacked LRR motifs. Each such structure consists of a mosaic of conserved scaffold residues interspersed with highly variable residues to account for the enormous diversity in ligandbinding sites.

HYPOTHESIS: EVOLUTION OF ADAPTIVE IMMUNITY IN VERTEBRATES Are there any fundamental differences between invertebrates and vertebrates in terms of their potential pathogens? We cannot go back to the time when the earliest jawless ﬁsh diverged from a common cephalochordate ancestor (82), but we can speculate that the answer to this question is no, based on the fossil record and studies of contemporary species. There is no indication for massive eradication of species in the Cambrian that could imply new types of potentially devastating pathogens, and it seems unlikely that the newly evolved vertebrates became the favorite hosts for new kinds of pathogens soon after their emergence. Despite 500 million years of vertebrate existence, there is a dearth of evidence for signiﬁcant numbers of vertebrate-speciﬁc pathogens. Conversely, many pathogens are known to infect both invertebrate and vertebrate hosts, including, for example, more than 500 varieties of the arboviruses (83). Germ line–encoded innate immune barriers protect both invertebrates and vertebrates from potential pathogens, although vertebrates may be better protected against some of the frequently recurring pathogens. Even in vertebrates, however, innate immunity provides the ﬁrst line of defense against

pathogens because a protective level adaptive immune response takes at least several days to mount. Innate immune mechanisms of invertebrates must therefore be as efﬁcient as those in vertebrates for combating the rapidly evolving pathogens that these animals inevitably encounter (16). Why then do deuterostome invertebrates need a vastly expanded arsenal of germ line– encoded receptors when only a handful of PRRs sufﬁces for immunity in nematodes, insects, and vertebrates? The LRR-containing proteins and other multigene families of immune receptors of deuterostome invertebrates may play a pivotal role in the maintenance and surveillance of the endosymbiotic microbial communities that these animals harbor. For example, it has been estimated that more than 60% of echinoderm species associate with bacterial symbionts (84). The intestinal ﬂoral symbionts in sea urchins may be needed to ferment and detoxify the poorly nutrient kelp and algae on which these animals graze. Such a complex mode of long-term coexistence between animals and microorganisms may have favored the evolution of large arsenals of speciﬁc microbial recognition molecules, whereas the strategy of PAMP recognition indiscriminately targets practically all microorganisms as nonself. This complex mode of coexistence with endosymbiotic microbes most likely was transmitted from invertebrate ancestors to their vertebrate descendents because complex microbial communities exist in the intestines of all vertebrates; 400–1000 species are estimated to live in the human gastrointestinal tract (85, 86), and the commensal microbiota have been shown to shape the Ig repertoire of peripheral B lymphocytes (87). Furthermore, maintenance of the mammalian gut ﬂora appears to require highly elaborate immune mechanisms and an active cross talk between the microﬂora and the host mucosal immune system (88). It may be interesting to explore the mechanisms that invertebrates employ to distinguish between symbionts and potential pathogens in animals belonging to the deuterostome www.annualreviews.org • Evolution of Adaptive Immunity

511

ARI

15 February 2006

1:25

lineage as well as in protostome invertebrates that harbor endosymbiotic microorganisms, for example, the wood-feeding termites (89) and molluscan cephalopods (octopus, squid, and cuttleﬁsh) that harbor symbiotic bacteria in their light organs (90). Thus far we have emphasized the similarities between deuterostome invertebrates and their vertebrate evolutionary descendents. It may be more relevant to consider differences between these animal groups that may have favored the development of a completely new mode of antigen recognition in vertebrates on the basis of receptor gene rearrangement. To address this enigma, we need to look back at the Cambrian explosion ∼500 mya, when a unique and stunning burst of evolutionary diversiﬁcation of new vertebrate species began. In a relatively brief evolutionary period, a variety of free swimming jawless ﬁsh appeared in the oceans. These ﬁsh descended from small amphioxus-like ancestors that lived as suspension feeders buried in the sand in shallow coastal waters. Early skulled vertebrates (Craniates) had a unique feature that separated them from their cephalochordate ancestor, namely a whole genome duplication that most likely occurred at the beginning of vertebrate divergence (91, 92). This genome duplication may have fueled the dramatic “big leap” in vertebrate developmental, morphological, and functional innovation during the Cambrian period (93). We can speculate that a large arsenal of diverse LRR-containing proteins was also part of the ancestral cephalochordate heritage. Members of these abundant cell surface and soluble receptors may have engaged in serendipitous interactions with newly evolving molecular determinants of early agnathans. If so, this interference may have been a rate-limiting factor in the process of rapid vertebrate evolution. In consideration of the enormous binding versatility of LRRcontaining proteins, it is conceivable that their self-reactivity presented serious autoimmunity problems at a time of rapid developmental and morphologic innovation. Further-

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

512

Pancer

·

Cooper

more, the transition from invertebrates to vertebrates may have been associated with changes in the endosymbiotic communities, thereby rendering many of these microbial surveillance proteins obsolete. In any case, the rate of readjustment required to maintain large multigene families of germ line receptors as strictly nonself-reactive may have become overly burdensome. Consequently, early vertebrates may have been forced to abandon the invertebrate deuterostome strategy of large arsenals of germ line–encoded immune receptors. This line of reasoning leads us to speculate that an adjustable immune system based on randomly generated receptor diversity evolved in part to enable the burst of vertebrate speciation in the Cambrian. Lymphocytes bearing uniquely rearranged surface antigen receptors, which could undergo negative selection to purge self-reactive lymphocytes while sparing clones expressing potentially beneﬁcial antigen receptors of sufﬁcient diversity, could have replaced the function of ancestral germ line arsenals. It is also likely that the newly evolved vertebrate lymphocytes performed innate immune functions concomitantly with the stepwise acquisition of acquired immune functions. There is ample evidence that lymphocytes have retained innate immune functionality. For example, B lymphocytes express TLRs and respond to their ligands by proliferation, expression of costimulatory molecules, and plasma cell differentiation (39). B cells of the peritoneal cavity and spleen marginal zones mediate microbial destruction via secretion of polyreactive antibodies that are essentially germ line encoded (94). Plasmacytoid dendritic cells derived from a common lymphoid progenitor express TLR7 and -9. These professional producers of IFN-α/β are important in protection against a wide range of viruses, bacteria, and parasites (95). The T lymphocytes are professional producers of IFN-γ. The concerted integration between innate and adaptive immune functions of lymphocytes may explain why the outcome of genetic defects that prevent the development

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

of T and B lymphocytes in infants with severe combined immunodeﬁciency diseases precludes survival from viral, bacterial, and fungal infections (96). Only hagﬁsh, lamprey, and the jawed vertebrates survived from the early vertebrate radiation, and only a sparse fossil record remains from the short period that separates the emergence of jawless ﬁsh and the appearance of jawed vertebrates (97, 98). It therefore will be difﬁcult to determine whether the agnathan LRR-containing VLRs were forerunners of vertebrate immune receptors or if

the rearranging VLRs and Igs evolved as independent solutions to similar necessities. The development of two very different modes of lymphocyte-based receptor diversiﬁcation at the dawn of vertebrate evolution nevertheless strongly attests to the enormous ﬁtness value of anticipatory immunity. The beneﬁts from this innovative strategy may not be limited to the ability to recognize a nearly inﬁnite antigenic world. Rather, the immediate selective pressure may instead have been facilitation of the developmental and morphological plasticity of the vertebrates.

ACKNOWLEDGMENTS We thank Hui-Hsien Chou of Iowa State University at Ames for providing Perl scripts to analyze the trace archive sequences. We also thank Matthew Alder of the University of Alabama at Birmingham; L. Aravind, Lakshminarayan Iyer, and Igor Rogozin of the National Center for Biotechnology Information, National Library of Medicine at the National Institutes of Health; and Gerardo Vasta of the Center of Marine Biotechnology, University of Maryland Biotechnology Institute at Baltimore, for helpful discussion. We additionally thank Ann Brookshire for her role in manuscript preparation. Z.P. was funded by National Science Foundation grants MCB-0317460 and IBN-0321461. M.D.C. is an investigator at the Howard Hughes Medical Institute. This paper is contribution #05-120 from the Center of Marine Biotechnology.

LITERATURE CITED 1. Cannon JP, Haire RN, Rast JP, Litman GW. 2004. The phylogenetic origins of the antigenbinding receptors and somatic diversiﬁcation mechanisms. Immunol. Rev. 200:12–22 2. Flajnik MF, Du Pasquier L. 2004. Evolution of innate and adaptive immunity: Can we draw a line? Trends Immunol. 25:640–44 3. Yu XQ, Kanost MR. 2002. Binding of hemolin to bacterial lipopolysaccharide and lipoteichoic acid. An immunoglobulin superfamily member from insects as a pattern-recognition receptor. Eur. J. Biochem. 269:1827–34 3a. Watson FL, Puttmann-Holgado R, Thomas F, Lamar DL, Hughes M, et al. 2005. Extensive diversiﬁcation of Ig-superfamily proteins in the immune system of insects. Science 309:1874– 78 4. Zhang SM, Adema CM, Kepler TB, Loker ES. 2004. Diversiﬁcation of Ig superfamily genes in an invertebrate. Science 305:251–54 5. Cannon JP, Haire RN, Schnitker N, Mueller MG, Litman GW. 2004. Individual protochordates have unique immune-type receptor repertoires. Curr. Biol. 14:R465–66 6. Azumi K, De Santis R, De Tomaso A, Rigoutsos I, Yoshizaki F, et al. 2003. Genomic analysis of immunity in a Urochordate and the emergence of the vertebrate immune system: “waiting for Godot.” Immunogenetics 55:570–81 7. Meyer A, Zardoya R. 2003. Recent advances in the (molecular) phylogeny of vertebrates. Annu. Rev. Ecol. Evol. Syst. 34:311–38 www.annualreviews.org • Evolution of Adaptive Immunity

513

ARI

15 February 2006

1:25

8. Takezaki N, Figueroa F, Zaleska-Rutczynska Z, Klein J. 2003. Molecular phylogeny of early vertebrates: monophyly of the agnathans as revealed by sequences of 35 genes. Mol. Biol. Evol. 20:287–92 9. Pancer Z, Amemiya CT, Ehrhardt GRA, Ceitlin J, Gartland GL, Cooper MD. 2004. Somatic diversiﬁcation of variable lymphocyte receptors in the agnathan sea lamprey. Nature 430:174–80 10. Pancer Z, Saha NR, Kasamatsu J, Suzuki T, Amemiya CT, et al. 2005. Variable lymphocyte receptors in hagﬁsh. Proc. Natl. Acad. Sci. USA 102:9224–29 11. Meyerowitz EM. 2002. Plants compared to animals: the broadest comparative study of development. Science 295:1482–85 12. Millet ACM, Ewbank JJ. 2004. Immunity in Caenorhabditis elegans. Curr. Opin. Immunol. 16:4–9 13. Meister M. 2004. Blood cells of Drosophila: cell lineages and role in host defense. Curr. Opin. Immunol. 16:10–15 14. Irving P, Ubeda JM, Doucet D, Troxler L, Lagueux M, Zachary D. 2005. New insights into Drosophila larval haemocyte functions through genome-wide analysis. Cell Microbiol. 7:335–50 15. Pancer Z. 2000. Dynamic expression of multiple scavenger receptor cysteine-rich genes in coelomocytes of the purple sea urchin. Proc. Natl. Acad. Sci. USA 97:13156–61 16. Loker ES, Adema CM, Zhang SM, Kepler TB. 2004. Invertebrate immune systems—not homogeneous, not simple, not well understood. Immunol. Rev. 198:10–24 17. Good RA, Finstad J, Litman GW. 1972. Immunology. In The Biology of Lampreys, ed. MV Hardisty, IC Potter, 2:405–32. London: Academic 18. Ardavin CF, Zapata A. 1987. Ultrastructure and changes during metamorphosis of the lympho-hemopoietic tissue of the larval anadromous sea lamprey Petromyzon marinus. Dev. Comp. Immunol. 11:79–93 19. Finstad J, Papermaster BW, Good RA. 1964. Evolution of the immune response. II. Morphologic studies on the origin of the thymus and organized lymphoid tissue. Lab. Invest. 13:490–512 20. Ardavin CF, Zapata A. 1988. The pharyngeal lymphoid tissue of lampreys. a morphofunctional equivalent of the vertebrate thymus? Thymus 11:59–65 21. Wallin IE. 2005. The relationships and histogenesis of thymus-like structures in ammocoetes. Am. J. Anat. 22:127–67 22. Shintani S, Terzic J, Sato A, Saraga-Babic M, O’hUigin C, et al. 2000. Do lampreys have lymphocytes? The Spi evidence. Proc. Natl. Acad. Sci. USA 97:7417–22 23. Mayer WE, Uinuk-ool T, Tichy H, Gartland LA, Klein J, Cooper MD. 2002. Isolation and characterization of lymphocyte-like cells from a lamprey. Proc. Natl. Acad. Sci. USA 99:14350–55 24. Uinuk-ool T, Mayer WE, Sato A, Dongak R, Cooper MD, Klein J. 2002. Lamprey lymphocyte-like cells express homologs of genes involved in immunologically relevant activities of mammalian lymphocytes. Proc. Natl. Acad. Sci. USA 99:14356–61 25. Uinuk-ool TS, Mayer WE, Sato A, Takezaki N, Benyon L, et al. 2003. Identiﬁcation and characterization of a TAP-family gene in the lamprey. Immunogenetics 55:38–48 26. Rothenberg EV, Pant R. 2004. Origins of lymphocyte developmental programs: transcription factor evidence. Semin. Immunol. 16:227–38 27. Pancer Z, Mayer WE, Klein J, Cooper MD. 2004. Prototypic T-cell receptor and CD4like coreceptor expressed in lymphocytes of the agnathan sea lamprey. Proc. Natl. Acad. Sci. USA 101:13273–78

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

514

Pancer

·

Cooper

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

28. Suzuki T, Shin-I T, Kohara Y, Kasahara M. 2004. Transcriptome analysis of hagﬁsh leukocytes: a framework for understanding the immune system of jawless ﬁshes. Dev. Comp. Immunol. 28:993–1003 29. Acton RT, Weinheimer PF, Hildemann WH, Evans EE. 1969. Induced bactericidal response in the hagﬁsh. J. Bacteriol. 99:626–28 30. Hildemann WH, Thoenes GH. 1969. Immunological responses of Paciﬁc hagﬁsh. Transplantation 7:506–21 31. Fujii T, Nakagawa H, Murakawa S. 1979. Immunity in lamprey. I. Production of haemolytic and haemagglutinating antibody to sheep red blood cells in Japanese lampreys. Dev. Comp. Immunol. 3:441–51 32. Ferrandon D, Imler JL, Hoffmann JA. 2004. Sensing infection in Drosophila: Toll and beyond. Semin. Immunol. 16:43–53 33. Hoffmann JA. 2003. The immune response of Drosophila. Nature 426:33–38 34. Zambon RA, Nandakumar M, Vakharia VN, Wu LP. 2005. The Toll pathway is important for an antiviral response in Drosophila. Proc. Natl. Acad. Sci. USA 102:7257–62 35. Imler JL, Zheng LB. 2004. Biology of Toll receptors: lessons from insects and mammals. J. Leukoc. Biol. 75:18–26 36. Inamori K, Ariki S, Kawabata S. 2004. A Toll-like receptor in horseshoe crabs. Immunol. Rev. 198:106–15 37. Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388:394–97 38. Pasare C, Medzhitov R. 2004. Toll-like receptors: linking innate and adaptive immunity. Microbes Infect. 6:1382–87 39. Sakharkar MK, Kangueane P. 2004. Genome SEGE: a database for “intronless” genes in eukaryotic genomes. BMC Bioinformatics 5:67 40. Roach JC, Glusman G, Rowen L, Kaur A, Purcell MK, et al. 2005. The evolution of vertebrate Toll-like receptors. Proc. Nat. Acad. Sci. USA 102:9577–82 41. Pancer Z, Gershon H, Rinkevich B. 1995. Coexistence and possible parasitism of somatic and germ cell lines in chimeras of the colonial urochordate Botryllus schlosseri. Biol. Bull. 9:106–12 42. Leipe DD, Koonin EV, Aravind L. 2004. STAND, a class of P-loop NTPases including animal and plant regulators of programmed cell death: multiple, complex domain architectures, unusual phyletic patterns, and evolution by horizontal gene transfer. J. Mol. Biol. 343:1–28 43. Monosi B, Wisser RJ, Pennill L, Hulbert SH. 2004. Full-genome analysis of resistance gene homologues in rice. Theor. Appl. Genet. 109:1434–47 44. Nurnberger T, Brunner F, Kemmerling B, Piater L. 2004. Innate immunity in plants and animals: striking similarities and obvious differences. Immunol. Rev. 198:249–66 45. Jones DA, Takemoto D. 2004. Plant innate immunity—direct and indirect recognition of general and speciﬁc pathogen-associated molecules. Curr. Opin. Immunol. 16:48–62 46. Innes RW. 2004. Guarding the goods. New insights into the central alarm system of plants. Plant Physiology 135:695–701 47. Rose LE, Bittner-Eddy PD, Langley CH, Holub EB, Michelmore RW, Beynon JL. 2004. The maintenance of extreme amino acid diversity at the disease resistance gene, RPP13, in Arabidopsis thaliana. Genetics 166:1517–27 48. Xu M, Korban SS. 2004. Somatic variation plays a key role in the evolution of the Vf gene family residing in the Vf locus that confers resistance to apple scab disease. Mol. Phylogenet. Evol. 32:57–65 www.annualreviews.org • Evolution of Adaptive Immunity

515

ARI

15 February 2006

1:25

49. Kovalchuk I, Kovalchuk O, Kalck V, Boyko V, Filkowski J, et al. 2003. Pathogen-induced systemic plant signal triggers DNA rearrangements. Nature 423:760–62 50. Esnault C, Maestre J, Heidmann T. 2000. Human LINE retrotransposons generate processed pseudogenes. Nat. Genet. 24:363–67 51. Meijer AH, Krens SFG, Rodriguez IAM, He SN, Bitter W, et al. 2004. Expression analysis of the Toll-like receptor and TIR domain adaptor families of zebraﬁsh. Mol. Immunol. 40:773–83 52. Yarovinsky F, Zhang DK, Andersen JF, Bannenberg GL, Serhan CN, et al. 2005. TLR11 activation of dendritic cells by a protozoan proﬁlin-like protein. Science 308:1626– 29 53. Tsujita T, Tsukada H, Nakao M, Oshiumi H, Matsumoto M, Seya T. 2004. Sensing bacterial ﬂagellin by membrane and soluble orthologs of Toll-like receptor 5 in rainbow trout (Onchorhynchus mikiss). J. Biol. Chem. 279:48588–97 54. Yilmaz A, Shen S, Adelson DL, Xavier S, Zhu JJ. 2005. Identiﬁcation and sequence analysis of chicken Toll-like receptors. Immunogenetics 56:743–53 55. LeBouder E, Rey-Nores JE, Rushmere NK, Grigorov M, Lawn SD, et al. 2003. Soluble forms of Toll-like receptor (TLR)2 capable of modulating TLR2 signaling are present in human plasma and breast milk. J. Immunol. 171:6680–89 56. Ting JP-Y, Davis BK. 2005. CATERPILLER: a novel gene family important in immunity, cell death, and diseases. Annu. Rev. Immunol. 23:387–414 57. Chamaillard M, Inohara N, Nunez G. 2004. Battling enteroinvasive bacteria: Nod1 comes to the rescue. Trends Microbiol. 12:529–32 58. Kobayashi KS, Chamaillard M, Ogura Y, Henegariu O, Inohara N, et al. 2005. Nod2dependent regulation of innate and adaptive immunity in the intestinal tract. Science 307:731–34 59. Kwiatkowski DP. 2005. The complexity of genetic variation in a simple immune system. Trends Genet. 21:197–99 60. Hoebe K, Janssen E, Beutler B. 2004. The interface between innate and adaptive immunity. Nat. Immun. 5:971–74 61. Enkhbayar P, Kamiya M, Osaki M, Matsumoto T, Matsushima N. 2004. Structural principles of leucine-rich repeat (LRR) proteins. Proteins 54:394–403 62. Kobe B, Deisenhofer J. 1996. Mechanism of ribonuclease inhibition by ribonuclease inhibitor protein based on the crystal structure of its complex with ribonuclease A. J. Mol. Biol. 264:1028–43 63. Uff S, Clemetson JM, Harrison T, Clemetson KJ, Emsley J. 2002. Crystal structure of the platelet glycoprotein Ibα N-terminal domain reveals an unmasking mechanism for receptor activation. J. Biol. Chem. 277:35657–63 64. Kim JI, Lee CJ, Jin MS, Lee CH, Paik SG, et al. 2005. Crystal structure of CD14 and its implications for lipopolysaccharide signaling. J. Biol. Chem. 280:11347–51 65. Di Matteo A, Federici L, Mattei B, Salvi G, et al. 2003. The crystal structure of polygalacturonase-inhibiting protein (PGIP), a leucine-rich repeat protein involved in plant defense. Proc. Natl. Acad. Sci. USA 100:10124–28 66. van der Hoorn RAL, Wulff BBH, Rivas S, Durrant MC, van der Ploeg A, et al. 2005. Structure-function analysis of cf-9, a receptor-like protein with extracytoplasmic leucinerich repeats. Plant Cell 17:1000–15 67. Smits G, Campillo M, Govaerts C, Janssens V, Richter C, et al. 2003. Glycoprotein hormone receptors: determinants in leucine-rich repeats responsible for ligand speciﬁcity. EMBO J. 22:2692–703

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

516

Pancer

·

Cooper

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

ARI

15 February 2006

1:25

68. Vischer HF, Granneman JC, Noordam MJ, Mosselman S, Bogerd J. 2003. Ligand selectivity of gonadotropin receptors. Role of the β-strands of extracellular leucine-rich repeats 3 and 6 of the human luteinizing hormone receptor. J. Biol. Chem. 278:15505–13 69. Schubert WD, Urbanke C, Ziehm T, Beier V, Machner MP, et al. 2002. Structure of internalin, a major invasion protein of Listeria monocytogenes, in complex with its human receptor E-cadherin. Cell 111:825–36 70. Freiberg A, Machner MP, Pfeil W, Schubert WD, Heinz DW, Seckler R. 2004. Folding and stability of the leucine-rich repeat domain of internalin B from Listeria monocytogenes. J. Mol. Biol. 337:453–61 71. He XL, Bazan JF, McDermott G, Park JB, Wang K, et al. 2003. Structure of the Nogo receptor ectodomain: a recognition module implicated in myelin inhibition. Neuron 38:177– 85 72. Le Goff MM, Hindson VJ, Jowitt TA, Scott PG, Bishop PN. 2003. Characterization of opticin and evidence of stable dimerization in solution. J. Biol. Chem. 278:45280–87 73. Scott PG, McEwan PA, Dodd CM, Bergmann EM, Bishop PN, et al. 2004. Crystal structure of the dimeric protein core of decorin, the archetypal small leucine-rich repeat proteoglycan. Proc. Natl. Acad. Sci. USA 101:15633–38 74. Bell JK, Mullen GE, Leifer CA, Mazzoni A, Davies DR, Segal DM. 2003. Leucine-rich repeats and pathogen recognition in Toll-like receptors. Trends Immunol. 24:528–33 75. Choe J, Kelker MS, Wilson IA. 2005. Crystal structure of human Toll-like receptor 3 (TLR3) ectodomain. Science 309:581–85 76. Cornelie S, Hoebeke J, Schacht AM, Bertin B, Vicogne J, et al. 2004. Direct evidence that Toll-like receptor 9 (TLR9) functionally binds plasmid DNA by speciﬁc cytosinephosphate-guanine motif recognition. J. Biol. Chem. 279:15124–29 77. Rutz M, Metzger J, Gellert T, Luppa P, Lipford GB, et al. 2004. Toll-like receptor 9 binds single-stranded CpG-DNA in a sequence- and pH-dependent manner. Eur. J. Immunol. 34:2541–50 78. Diebold SS, Kaisho T, Hemmi H, Akira S, Sousa CRE. 2004. Innate antiviral responses by means of TLR7-mediated recognition of single-stranded RNA. Science 303:1529–31 79. Lund JM, Alexopoulou L, Sato A, Karow M, Adams NC, et al. 2004. Recognition of singlestranded RNA viruses by Toll-like receptor 7. Proc. Natl. Acad. Sci. USA 101:5598–603 80. Meng G, Grabiec A, Vallon M, Ebe B, Hampel S, et al. 2003. Cellular recognition of tri-/ di-palmitoylated peptides is independent from a domain encompassing the N-terminal seven leucine-rich repeat (LRR)/LRR-like motifs of TLR2. J. Biol. Chem. 278:39822– 29 81. Forey PL, Janvier P. 1993 Agnathans and the origin of jawed vertebrates. Nature 361:129– 34 82. Borucki MK, Kempf BJ, Blitvich BJ, Blair CD, Beaty BJ. 2002. La Crosse virus: replication in vertebrate and invertebrate hosts. Microbes Infect. 4:341–50 83. Burnett WJ, McKenzie JD. 1997. Subcuticular bacteria from the brittle star Ophiactis balli (Echinodermata: Ophiuroidea) represent a new lineage of extracellular marine symbionts in the alpha subdivision of the class Proteobacteria. Appl. Environ. Microbiol. 63:1721–24 84. Noverr MC, Huffnagle GB. 2004. Does the microbiota regulate immune responses outside the gut? Trends Microbiol. 12:562–68 85. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, et al. 2005. Diversity of the human intestinal microbial ﬂora. Science 308:1635–38 86. Lanning DK, Rhee KJ, Knight KL. 2005. Intestinal bacteria and development of the Blymphocyte repertoire. Trends Immunol. 26:419–25 www.annualreviews.org • Evolution of Adaptive Immunity

517

ARI

15 February 2006

1:25

87. Kelly D, Conway S, Aminov R. 2005. Commensal gut bacteria: mechanisms of immune modulation. Trends Immunol. 26:326–33 88. Stingl U, Radek R, Yang H, Brune A. 2005. “Endomicrobia”: cytoplasmic symbionts of termite gut protozoa form a separate phylum of prokaryotes. Appl. Environ. Microbiol. 71:1473– 79 89. Nishiguchi MK, Lopez JE, Boletzky S. 2004. Enlightenment of old ideas from new investigations: more questions regarding the evolution of bacteriogenic light organs in squids. Evol. Dev. 6:41–49 90. Abi-Rached L, Gilles A, Shiina T, Pontarotti P, Inoko H. 2002. Evidence of en bloc duplication in vertebrate genomes. Nat. Genet. 31:100–5 91. Durand D. 2003. Vertebrate evolution: doubling and shufﬂing with a full deck. Trends Genet. 19:2–5 92. Ohno S. 1970. Evolution by Gene Duplication. New York: Springer-Verlag 93. Viau M, Zouali M. 2005. B-lymphocytes, innate immunity, and autoimmunity. Clin. Immunol. 114:17–26 94. McKenna K, Beignon AS, Bhardwaj N. 2005. Plasmacytoid dendritic cells: linking innate and adaptive immunity. J. Virol. 79:17–27 95. Rosen FS, Cooper MD, Wedgwood RJP. 1984. The primary immunodeﬁciencies. N. Engl. J. Med. 311(Pt. I):235–42, (Pt. II):300–10 96. Mallatt J, Chen JY. 2003. Fossil sister group of craniates: predicted and found. J. Morphol. 258:1–31 97. Kumar S, Hedges SB. 1998. A molecular timescale for vertebrate evolution. Nature 392:917–20 98. Kumar S, Tamura K, Nei M. 2004. MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief. Bioinformatics 5:150–63

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-16

518

Pancer

·

Cooper

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:497-518. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-17

ARI

15 February 2006

1:34

Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How∗ Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Flora Castellino and Ronald N. Germain Lymphocyte Biology Section, Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892; email: [email protected]

Annu. Rev. Immunol. 2006. 24:519–40 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.23.021704.115825 c 2006 by Copyright Annual Reviews. All rights reserved ∗ The U.S. Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

0732-0582/06/0423-0519$20.00

Key Words dendritic cells, T cell help, cell-mediated immunity, chemokines

Abstract Concepts of cell-cell interactions in adaptive immunity have alternated between the simple and the complex. The notion that one population of small, circulating lymphocytes is responsible for adaptive immunity was sequentially supplanted by the concept of separate T and B lymphocyte populations that cooperate to produce IgG antibody responses, by a three-cell model in which a myeloid APC initiates these cooperative lymphoid responses, by the recognition of T cell subsets, and by the idea that CD8+ T cell subset responses to graft antigens depend on CD4+ T cell subset activity. Simplicity was reintroduced with the revelation that CD8+ T cells can act independently of CD4+ T cells against acute viral infections. The pendulum has swung again toward complexity with recognition of the distinct and conjoint contributions of innate stimuli, APCs, NK and NKT cells, Tregs, and CD4+ helper T cells to CD8+ T cell behavior during acute and chronic infections or as memory cells. The renewed appreciation that multiple, sometimes rare cell types must communicate during cell-mediated immune responses has led to questions about how such interactions are orchestrated within organized lymphoid tissues. We review recent advances in deciphering the speciﬁc contribution of CD4+ T cells to physiologically useful CD8+ T cell responses, the signals involved in producing acute effectors versus long-lived memory cells, and the mechanisms underlying the cell-cell associations involved in delivery of such signals. We propose a model based on these new ﬁndings that may serve as a general paradigm for cellular interactions that occur in an inﬂamed lymph node during the initiation of immune responses.

519

ANRV270-IY24-17

ARI

15 February 2006

1:34

INTRODUCTION

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CD8+ T cells contribute to host defense during acute and chronic infection with viruses, intracellular bacteria, or single-cell as well as multi-cellular parasites, and they also participate in the elimination of transformed cells (1–8). There is thus a substantial interest at the basic, translational, and clinical levels in understanding the antigens recognized by this T cell subset, the processing pathways involved in the generation of these receptor ligands, the cellular interactions that participate in facilitating and regulating responses by CD8+ T lymphocytes, and the soluble or cellassociated molecular signals that inﬂuence the differentiation and survival of this cell type during the acute, chronic, and memory phases of a response. Over the past several decades, a substantial literature has dealt with these various aspects of cell-mediated immunity. The key role of major histocompatibility complex (MHC) class I molecules in presenting short peptide antigens to the clonal receptors of CD8+ T cells (TCRs) has been amply documented and the cell biology of antigen processing that leads to formation of these peptide-MHC ligands well delineated (9–11). In more recent years, there has been a renewed interest in better understanding the contributions of other cell types to the proliferation, effector function, and memory cell behavior of the CD8+ T cell subset. A growing appreciation of the impact of innate signals arising in response to both pathogen-derived and host stress response stimuli (12–14) has led to a reexamination of the early events in the activation of CD8+ T cells and in their development of effector function. Intimately connected with this issue is the role of dendritic cells (DCs), not merely in presenting antigen, but in supplying critical costimulatory and differentiation-inducing mediators that guide the developing lymphocyte response (15, 16). In addition, there has been increasing attention to the role of CD4+ T cells in CD8+ T cell responses (5, 17), driven in part by a de-

520

Castellino

·

Germain

sire to understand better the basis for immune dysfunction in individuals with defects in their CD4+ T cell population. After many years of back and forth debate over whether interactions between these two αβ T cell subsets are required in vivo for effective cell-mediated immunity, a consensus has begun to emerge that distinguishes the requirements for acute effector responses of CD8+ T cells from the development of robust memory and in many circumstances assigns to CD4+ T cells a primary role in the latter rather than the former. Here, we review the evidence that CD4+ T cell “help” (TH) has a key impact on CD8+ memory cell formation and function, and describe the ongoing controversy about whether such help is antigen speciﬁc. We also consider whether help is delivered early or late in the response and examine the events occurring within lymphoid tissues that allow effective communication between rare antigen-speciﬁc cells in the two T cell lineages. We end with some speculations about the roles of the many distinct mediators that have been reported to determine the nature and magnitude of CD8+ T cell responses and the possibility that how cellular cooperation is orchestrated for these responses might represent a general strategy for cell-cell interactions in the immune system.

WHY CD4+ T CELL HELP? CELL-CELL INTERACTIONS AS CHECKPOINTS IN LIMITING AUTOIMMUNITY The goal of the immune system is to protect the organism from foreign infectious agents while avoiding pathology during such useful responses or owing to undesirable reactivity to self (18). This goal is achieved via multiple checkpoints operating on both precursor lymphocytes during their development in primary lymphoid tissues (central tolerance) and on mature cells in secondary lymphoid tissues as well as parenchymal sites (peripheral tolerance) (19, 20).

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

In the early 1970s, investigators ﬁrst proposed that B cell activation requires two signals: one provided by the recognition of the antigen and the other by the interaction with helper cells recognizing the same or a physically linked antigen. Failure to receive the second signal was postulated to induce B cell tolerance and therefore to limit reactivity to self in light of somatic hypermutation. This two-signal model was based on the presumption that B cells whose immunoglobulin receptors gained self-reactivity by the mutation process would require, for their full activation, a second signal from activated self-reactive CD4+ T cells that should have been deleted or inactivated as a result of thymic and peripheral tolerance mechanisms (21). Since then, several observations have validated the core conceptual (although not the speciﬁc mechanistic) features of this model both for B cells and CD8+ T cells (22–26). With respect to T cells, many of the immature cells that exhibit high afﬁnity for selfpeptides presented in the thymus are deleted there, but such negative selection is not a foolproof mechanism, and some T cells capable of damaging autoimmune reactivity escape to the periphery (27). To further limit effector responses to both self as well as harmless environmental antigens, several mechanisms regulate lymphocyte activation in the periphery. The dose and afﬁnity of the antigen required to activate mature T lymphocytes is higher than for immature thymocytes (28– 30), thus facilitating activation-related ignorance of many self-peptides presented in the periphery. In addition, T cell responses are regulated by requirements for concurrent or sequentially ordered signaling. Prior to their interaction with naive T cells, antigenpresenting cells (APCs), in particular DCs, require stimulation through innate receptors, such as Toll-like receptors (TLRs), to acquire the capacity for highly effective triggering of a T cell and for the production of mediators that augment T cell clonal expansion, viability, and development of effector capacity (12, 31). A requirement for multiple antigen-speciﬁc

lymphocytes to interact, either directly with each other or through the intermediation of DCs, further reduces the risk of inappropriate responses to nonthreatening antigens, because inadequate activation/maturation of any of the cooperating cell types limits the clonal expansion and/or effector capacity of potentially auto-aggressive cells in the other subsets. A lack of CD4+ T cell help in particular clearly does not prevent initiation of potentially pathologic responses by other cell types; rather, the absence of such help controls the duration of such responses, limiting the damage to the host. As detailed below, a lack of CD4+ T cell help will allow an acute but not a sustained or memory CD8+ T cell response. Similarly, in the absence of CD4+ T cell help, B cell responses can be initiated, but the somatic hypermutation, isotype switching, and clonal selection necessary for production of high-afﬁnity immunoglobulins is restricted (32, 33). Such observations provide support for the thesis that a major reason for evolving an immune system with critical requirements for cell-cell cooperation is to impose controls on the development of inappropriate responses, as originally postulated in the two-signal model (21).

ROLE OF CD4+ T CELL HELP IN THE GENERATION OF ACUTE EFFECTOR VERSUS MEMORY CD8+ T CELL RESPONSES If the involvement of CD4+ T cells in CD8+ T cell responses is to provide a check on unwanted, possibly tissue-damaging reactions, at what point in the development of such responses does this regulation take place? In addressing this question, it is useful to summarize the prototypic course of CD8+ lymphocyte responses as they are currently understood from studies using various acute infection models in mice (7) and, to a more limited extent, from the analysis of infected humans (34). Naive CD8+ T cells are long-lived resting cells that migrate continuously from the blood to the www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

521

ARI

15 February 2006

1:34

secondary lymphoid tissues and survive with only limited division via the recognition of MHC–self-peptide ligands and exposure to what are called homeostatic cytokines, of which IL-7 is the most important (35–38). Exposure of a naive host to an infectious agent, especially a rapidly replicating virus, evokes a cell-mediated response with three distinct phases: activation/expansion, contraction, and maintenance/memory (4). The activation/expansion phase typically peaks one to two weeks after infection begins and is characterized by the exponential proliferation of antigen-speciﬁc T cells. Once the T cells are fully activated, they leave the lymph nodes and accumulate in the peripheral tissues, where they come in direct contact with infected parenchymal cells and exert their effector functions. Pathogen numbers increase dramatically during the early part of this ﬁrst phase, then decline prior to the peak of CD8+ T cell expansion, as a cumulative result of innate defenses and the antigen-speciﬁc activity of the expanding CD8+ effector population. A week to ten days after the initiation of the immune response, a contraction period begins. This is characterized by the apoptosis of more than 90% of the activated T cells, massive cell death that is required to maintain quasi-constant numbers of T cells throughout the life of the individual. If the response was effective in clearing the infection, and in an intact host with all other aspects of the immune system acting in a physiological fashion, this death phase will spare a small number of T cells that are maintained over time through limited divisions in response to cytokines, particularly IL-15 (39, 40). This small pool of memory cells is believed by most, but not all, investigators to be crucial in protecting the individual upon reinfection with the same or a related organism because its members are capable of more rapid and potent effector responses and because these cells possess the ability to migrate and patrol peripheral tissues (4, 41, 42). The proportion of CD8+ T cells that survive as memory cells is remarkably constant (5%–10%

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

522

Castellino

·

Germain

of the initial burst size) (43, 44), suggesting that loss during the contraction phase is not a passive default pathway owing to the intrinsic short life of activated T cells, but rather a tightly regulated process involving several death-inducing mechanisms and the selective sparing of certain activated lymphocytes. The ﬁeld’s view of where CD4+ T cells participate in this sequence of events has changed over time. Early in vivo experiments involving allograft rejection and in vitro studies of allogeneic mixed lymphocyte reactions concluded that MHC class II–speciﬁc CD4+ T cells were necessary for the generation of cytotoxic CD8+ T cell responses and led to the original concept of CD4+ T cell “help” (TH) as essential for the clonal expansion of naive CD8+ T cells (1, 26, 45–50). On the basis of studies ﬁrst conducted more than two decades ago (51–53) and conﬁrmed by more recent experiments (54), investigators believed that the processed antigens seen by CD4+ T cells and CD8+ T cells involved in such cooperative responses must reside on the membrane of a single APC for effective delivery of TH. Evidence that neutralization of IL2 or blockade of the IL-2 receptor in culture severely limited cytotoxic CD8+ T cell responses (55–59), and data showing that CD4+ T cells were a major source of IL-2 (60), all were consistent with TH function being dependent on the local paracrine delivery of this cytokine. This concept of TH function led to a simple hypothesis: CD4+ and CD8+ T cells have to be stimulated by antigen on the same APC so that IL-2 secreted by the CD4+ T cell can act on a neighboring CD8+ T cell expressing high-afﬁnity IL-2 receptors. This cytokine secretion-signaling link was presumed to play a key role in vivo in controlling the early phase of expansion of the small number of naive CD8+ precursors reactive with the antigens of an infectious organism or minor H antigens. Since then, several ﬁndings have challenged this original model of CD4+ -CD8+ T cell cooperation. The observation that both

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

DCs and CD8+ T cells can transiently secrete IL-2 upon activation (61–65) raised questions about the role of activated CD4+ T cells as unique producers of this cytokine and of IL2 as the key mediator of TH. The picture became more complex when several studies showed that CD4+ T cells were dispensable in the clearance of acute viral infections, arguing against an absolute dependence of CD8+ T cells on TH for development of a clonally expanded effector population (66). This ﬁnding led to the hypothesis that TH was essential for immunity to antigens such as tissue grafts that lacked molecular stimuli for cells of the innate immune system [adjuvants, or PAMPs (pathogen-associated molecular patterns) in Janeway’s terminology (67)], but that help was not required for responses to antigens associated with potent sources of these stimuli, such as live viruses and bacteria. In this model, TH was required to activate quiescent APCs (DCs), a functionality that was provided in the case of infectious agents by ligands for TLRs, such as dsRNA (TLR3), lipopolysaccharide (TLR4), and CpG-containing oligonucleotides (TLR9) along with their induced products, such as type 1 interferons (68, 69). In the absence of such signals, CD40L (CD154) expression by antigen-activated CD4+ T cells appeared to be responsible for promoting a similar activation of APCs through interaction with CD40 expressed on these latter cells (54, 70–72). Doubts about this new formulation arose with the publication of new studies of antiviral responses after the acute phase. In both chronic infections and recall (memory) responses, a dependence on CD4+ T cells was noted that was at odds with the lack of such a requirement for acute CD8+ effector responses to the same agents (73–83). The apparent correspondence between the absence of strong inﬂammation and a need for CD4+ T cell help was lost. Such studies showed that depletion of CD4+ T cells did not affect pathogen clearance mediated by CD8+ T cells, if mice were acutely infected with low numbers of organisms, but led to persis-

tent infection when higher doses of the same agent were used. In the latter animals, responses were impaired owing both to the deletion of some virus-speciﬁc CD8+ clones and to the persistence of CD8+ T cells without effector functions, and the exhaustion of the CD8+ T cell response was more pronounced in the absence of CD4+ T cells (73–75, 78). Although these studies ﬁrst documented a critical role for CD4+ T cells in sustaining rather than initiating CD8+ effector responses, they did not reveal whether the CD4+ subset contributed directly to the survival of activated CD8+ T cells or indirectly, for example, because of the inability of CD4+ T cell–deﬁcient animals to mount a strong B cell response that in turn could lead to persistence of antigen and inﬂammation, both of which have detrimental effects on the survival of functional cytotoxic CD8+ T cells (84, 85). Newer methods, such as tetramer tracking of speciﬁc T cells and/or the use of TCR transgenic clonal T cell populations, have allowed resolution of this issue. A series of recent studies have all reached a similar conclusion, namely that in both classical TH-dependent and TH-independent immunization models, CD4+ T cells are very often dispensable for early clonal expansion and the generation of primary CD8+ cytotoxic effectors (although they may augment both aspects of the CD8+ T cell response to a greater or lesser degree), but are required for the generation of an optimal pool of functional memory CD8+ T cells (80–83).

WHEN IS TH REQUIRED FOR MEMORY CELL DEVELOPMENT? If the preponderance of evidence is now that TH has its primary role in the formation of a useful memory cell pool [some evidence to the contrary notwithstanding (50)], when does this contribution occur during the several phases of the CD8+ T cell response? Proliferation of antigen-speciﬁc CD8+ T cells www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

523

ARI

15 February 2006

1:34

begins within 24–30 h of antigen introduction in vivo (86). The activated CD8+ T cells then typically undergo more than 7–8 cell divisions in a process that appears to be independent of further foreign antigen recognition by the TCR of these CD8+ T cells (87, 88), although a role for self-MHC recognition in driving this expansion has not been ruled out. If the CD8+ T cells operate on autopilot during this proliferative phase, could TH acting at this very early point possibly have an effect on memory cell survival weeks or months later? Some might consider it more sensible that TH acts when the expanded CD8+ T cells start to undergo rapid attrition, or even very late, in what would be called the maintenance phase. As is typical in immunology, different studies have claimed to provide evidence for all three possibilities. Support for an early effect of TH on long-term memory CD8+ T cell survival and function came from experiments in which CD4+ T cells were removed from the priming environment, or antigen-activated CD8+ T cells were removed from the CD4+ T cell– containing host within a few days of initial antigen exposure. CD4+ T cell depletion later than three days after antigen injection in any of several model systems produced a memory CD8+ T cell pool whose size and functionality was the same as in animals with an intact CD4+ T cell cohort (81–83). Furthermore, such helped CD8+ T cells, if transferred into a CD4+ T cell–deﬁcient recipient, survived, developed into functional memory cells, and protected the host from a subsequent challenge with the same infectious agent used for the priming. The conclusion drawn from these studies was that once CD8+ T cells have been primed in the presence of CD4+ T cells, the latter are no longer required for the eventual emergence of useful memory cells many weeks later. Because these studies analyzed only the acute secondary response, it remains unclear whether CD4+ T cells are also required at the time of boosting for the second wave of effectors to survive and develop into a memory population as well.

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

524

Castellino

·

Germain

In at least one model involving bland priming with tumor cells, the helpless CD8+ T cells showed a defect in clonal expansion upon antigen reexposure in vitro (81). A follow-up study has provided evidence that this proliferative defect is the result of TRAIL production by the helpless CD8+ T cells, which leads to their death by apoptosis shortly after antigen stimulation (89). Although these recent data can explain the lack of vigorous secondary responses or sustained memory cell activity upon tumor or infectious challenge, they do not account for the smaller number of antigen-speciﬁc memory CD8+ T cells in the maintenance phase, when priming occurs in the absence of CD4+ T cells. Nor do these data account for the more limited ability of those cells that remain several weeks after priming in the absence of CD4+ T cells to show acute cytokine responses within hours after restimulation (80–83), before TRAIL-induced death could affect the outcome. These considerations suggest that the TRAIL limitation of secondary expansion is just one of a number of defects imposed on the CD8+ T cells by an early lack of TH. The opposite conclusion about the time when TH affects the memory pool was reached by different investigators. In a study using viral or bacterial antigen delivery, the absence of a polyclonal CD4+ T cell population in the host led to a late, gradual decline in the number of memory CD8+ cells even when priming occurred in the presence of antigen-speciﬁc TH (90). These investigators concluded that antigen-speciﬁc CD4+ T cell activation was not required at priming to program a small population of CD8+ T cells to develop into memory cells, but rather that an antigen-unspeciﬁc bystander effect, probably mediated by cytokines, operated late in the response to maintain the viability of the CD8+ memory pool (90). A key point is that in this study the conclusion that antigen-speciﬁc TH was not necessary at priming for the generation of a maximal pool of functional memory CD8+ T cells relied exclusively on the

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

detection of similar numbers of activated IL7Rαhi CD8+ T cells in wild-type and CD4+ T cell–deﬁcient animals at the peak of the primary response (90). Kaech et al. had previously proposed that high surface expression of IL-7Rα on CD8+ T cells during the early primary response could be used to identify the prememory CD8+ T cell pool, a subpopulation of activated T cells that preferentially gave rise to long-lived memory cells (91, 92). The principle evidence that cells of this phenotype were programmed for survival as memory cells involved their twofold decline in number during the contraction phase of the response, compared with a more than 20-fold decline for IL-7Rαlo cells, as well their persistence after adoptive transfer into naive animals, compared with the nearly complete loss of IL-7Rαlo cells under the same conditions (91). However, although the correlation between IL-7Rα expression and survival was numerically robust in this one study, in other experimental situations this was not the case. Under a variety of priming conditions involving either live agents or nonreplicating vaccine formulations, from 30% to 70% of the activated, dividing CD8+ T cells at the peak of the response have the IL-7Rαhi phenotype, yet only ∼5% of the cells remain several weeks later (90, 92; F. Castellino, manuscript submitted). Thus, although a high level of this cytokine receptor is a necessary condition for long-term survival of an activated CD8+ T cell, it is not a sufﬁcient condition. Furthermore, we have conﬁrmed that the frequency of CD8+ T cells marked by high IL-7Rα levels does not differ early after priming in the presence or absence of TH (F. Castellino, manuscript submitted). However, we can detect an enhanced frequency of a phenotypically distinct subpopulation of IL7Rαhi -activated CD8+ T cells arising early after priming in the presence of TH, as compared with the same priming conditions in the absence of TH. The number of these cells corresponds precisely to the increment in long-term memory cells provided by antigen-

activation of CD4+ T cells, that is, by TH (F. Castellino, manuscript submitted). Together, these various observations indicate that it may have been misleading to assume, on the basis of a failure to observe an increment in IL7Rαhi cells, that no programming for memory cell development occurred in the presence of TH. We wish to stress, however, that beyond the contribution of early TH-dependent programming of CD8+ T cells to long-term memory, an additional role for bystander polyclonal CD4+ T cells in the maintenance of committed memory CD8+ T cells is not ruled out by available data.

WHERE IS TH DELIVERED EARLY AFTER PRIMING? PART 1 The “when” of TH delivery now seems reasonably well answered, with the bulk of evidence in favor of an early programming effect of TH on CD8+ T cells that enhances the ability of at least a fraction of the responding population to survive as long-term memory cells and to show heightened effector function upon recall stimulation (93). Thus, CD4+ T cells must deliver one or more signals to the CD8+ T cells directly or indirectly at the time of, or shortly after, initial contact with antigen-bearing APCs. This leaves the “where” and “how” to be addressed. Given that in a normal naive repertoire, the frequency of CD4+ and CD8+ T cells speciﬁc for antigens from a single infectious organism is quite low, a mechanism for bringing these rare cells together in time and/or space is required. The obvious platform for such signal transfer is the APC, and indeed, a substantial literature indicates that both the MHC class II–bound ligand that stimulates the CD4+ T cells responsible for TH and the MHC class I–bound ligand involved in activating the CD8+ T cells must be on the same APC membrane for cooperation to be effective (51, 54). The question that has preoccupied the ﬁeld for some time involves temporal details of such cooperation: whether the CD4+ and CD8+ T cells must simultaneously engage the same APC www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

525

ARI

15 February 2006

1:34

(the three-cell cluster model) or if the antigenactivated CD4+ T cell can leave its mark by licensing the DC for delivery of help to a CD8+ T cell after the CD4+ T cell has left the scene (94). The three-cell cluster model is consistent with the early hypothesis that CD4+ and CD8+ T cells need to recognize their speciﬁc antigens simultaneously on the same APC, so that local paracrine action of cytokines such as IL-2 secreted by an APC-bound CD4+ T cell can support clonal expansion by the colocalized CD8+ T cell (46). A recent variation of this paracrine three-cell cluster model suggests that antigen-primed CD4+ T cells can directly stimulate CD8+ T cells via CD40L when the two lymphocytes are in close proximity on an APC (80). Here, again, there is controversy, because other investigators ﬁnd no critical role for CD40 expression by CD8+ T cells in primary or memory responses (95, 96). In the minds of many immunologists, the strongest argument against the three-cluster model is the low probability that three rare cells (an antigen-bearing DC coming from an infected site, an antigen-speciﬁc CD4+ T cell, and an antigen-speciﬁc CD8+ T cell) will ﬁnd each other in the same place at the same time (26, 54). The intuitive feeling that such clusters could not be efﬁciently generated led to a new model in which antigen-stimulated CD4+ cells activate DCs via CD40L-CD40 interaction, and the resultant “licensed DCs” become fully competent to activate naive CD8+ T cells, even in the absence of an associated CD4+ T cell (54, 70–72). In one version of this model, the CD4+ T cell then moves from DC to DC, anointing each with the capacity to provide help to a CD8+ T cell that subsequently binds to one of these APCs, which also bears the appropriate peptideMHC class I ligand. This ﬂitting around was proposed to amplify TH function by creating many licensed DCs for each antigen-activated CD4+ T cell. There is clear evidence for an enhanced capacity of DCs preexposed to activated CD4+ T cells to support CD8+ T cell responses

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

526

Castellino

·

Germain

in vitro (54). The in vivo data are less clear, although some experiments appear consistent with the licensing concept (54, 70–72, 97). However, this model has several conceptual limits, and a number of assumptions were made when it was proposed: that CD4+ T cell binding to antigen-bearing DCs is shortlived; that naive T cells cannot sense each other’s presence; and that in lymph nodes, naive T cells randomly scan APCs. Regarding this last point, one key argument made for CD4+ T cells moving around rapidly and licensing many DCs was to prevent tolerization of CD8+ T cells upon engagement of cognate antigen on a nonlicensed DC (68). But if there is no speciﬁc mechanism that attracts CD8+ T cells to the licensed DC, it is still highly likely that a CD8+ T cell would ﬁnd an antigen-bearing, unlicensed DC before that CD8+ T cell has a chance to engage a DC that had already engaged a CD4+ T cell. Given the few speciﬁc CD8+ T cells in a naive repertoire (98), this tolerance induction would be a potentially devastating blow to the development of a useful immune response. One way around this limitation is to postulate that there is a period of time after antigen recognition involving an unlicensed DC during which delivery of the proper signal can rescue a CD8+ T cell from inactivation. This period would allow the CD8+ T cell much more time to ﬁnd a DC that had at some point interacted with a CD4+ T cell, but it would also remove one of the main arguments in support of the licensing model. Second, the assumption that the rare nature of speciﬁc lymphocytes would limit CD4+ and CD8+ T cell coengagement of a single DC is based on the idea that the length of time a T cell spends on an APC is short relative to the time it takes a T cell to ﬁnd that APC. But studies 30 years ago with guinea pig cells showed that antigenspeciﬁc CD4+ T cells could stay associated with APCs bearing speciﬁc antigen for at least several hours (99). This may be more than enough time for a speciﬁc CD8+ T cell to arrive, given that such cells complete a passage through a lymph node in about the same time

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

span (100). Third, and most important, the model assumes that cell-cell interactions are completely random; if the antigen-dependent interaction of a CD4+ T cell and a DC generates a signal that attracts naive CD8+ T cells, then the likelihood of a useful three-cell cluster arising increases greatly. New data relevant to these issues has recently become available from imaging studies of lymphocyte-DC interactions within intact lymph nodes. In terms of cell movement, several groups have conducted detailed analyses of the rate and directionality of migration for both CD4+ and CD8+ T cells in lymph nodes in the absence and presence of speciﬁc antigen and/or inﬂammatory stimuli (86, 101– 104). Each of these studies concluded that naive T cells move at an average speed of ∼8–12 μm/min in a manner that is consistent with a random walk, that is, with unguided movements not directed toward a particular location. The principle analytic method used to draw this conclusion is a plot of displacement versus time1/2 . A straight-line relationship in such a plot is well recognized in chemical physics to be consistent with a random walk, like gas molecules diffusing by Brownian motion. The studies of Cahalan and colleagues (101, 102) are the most extensive in this regard and involve both steady state and immunogenic conditions. They found no evidence that CD4+ T cells showed any deviation in their paths of travel near activated DCs, from which they concluded that contact between these cell types was dictated by two unguided search strategies, one the rapid movement of the dendrites of the DC and the other the random migration of the CD4+ T cells in the T cell zone of the lymph node (101, 102). In the case of CD8+ T cells studied without CD4+ T cell activation, two studies saw no inﬂuence of activated but nonantigen-bearing DCs on the behavior of the naive lymphocytes (103, 104), whereas a third saw some deviation from the expected straight-line random walk plot with activated DCs present, but only for a subset of the T cells and with no clear mechanism (86). However, none of these studies

involved a combination of antigen-activated CD4+ T cells and naive CD8+ T cells, which is the situation relevant to the delivery of TH. Returning to the issue of how long a CD4+ T cell remains on an antigen-bearing DC, there is a reasonable consensus that following an early phase in the ﬁrst few hours after antigen introduction during which transient CD4+ -DC interactions occur (101), a period of long-lived association of the two cell types ensues that can last for many hours (101, 105). Given the lack of evidence for cytokine gene expression by CD4+ T cells during the early transient interaction stage, it is unlikely that extensive DC licensing occurs during this period. During the second phase (between 6 h and 30 h following antigen introduction), most speciﬁc CD4+ T cells do not make contacts with multiple DCs but rather show monogamous interactions (101, 105, 106). During this time span, initiation of the CD8+ proliferative phase begins, so it does not seem likely that the ﬂitting hypothesis in which a single CD4+ T cells licenses many DCs early on ﬁts the data. Except for a model in which rescue of function among CD8+ T cells initiating their response on unlicensed DCs is the predominant mode for delivery of TH, one is left with two possibilities: (a) that random interaction among the relevant cells is the underlying mechanism, with all the problems this hypothesis poses for avoiding tolerance and ensuring an efﬁcient response at low precursor frequency; or (b) that guided interactions occur under conditions not analyzed in these existing studies of CD4+ or CD8+ T cells alone. Our very recent experiments suggest that the latter is the correct answer.

WHERE IS TH DELIVERED EARLY AFTER PRIMING? PART 2 To understand how the immune system could organize itself to promote interactions effectively among rare cells, a summary of the dynamics of antigen, DC, and inﬂammatory www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

527

ARI

15 February 2006

1:34

mediator trafﬁcking to draining lymph nodes and of the process of lymphocyte recirculation is helpful. Antigens injected subcutaneously reach a draining lymph node in two temporally distinct waves. Soluble molecules directly drain via the afferent lymph together with tissue-produced inﬂammatory cytokines such as TNF. This trafﬁcking via lymphatics allows antigens and cytokines to reach the lymph nodes within minutes, where they enter the conduit network that is the distribution system for lymph within the lymph node proper (107, 108) and are taken up by subcapsular macrophages and by conduitassociated, resident DCs, to be processed and presented by these cells within hours (109– 111). A second wave of presentation relies on tissue-resident DCs that acquire the antigen in the periphery (109, 112). Local inﬂammation induces rapid modiﬁcations in the pattern of expression of chemokine receptors that regulate the migration of DCs toward and their positioning within lymph nodes (113–117). Immature DCs located in peripheral tissues express receptors (CCR1, CCR2, CCR5, CCR6, CXCR1) for inﬂammatory chemokines (CCL3, CCL4, CCL5, CCL20). This pattern of expression allows DCs or their blood precursors to be attracted to inﬂammatory sites, where they internalize antigens and receive maturation stimuli. These stimuli (e.g., TNF-α) induce the DCs to decrease expression of the receptors for inﬂammatory chemokines and to upregulate the chemokine receptor CCR7, making them responsive to CCL21 (SLC) present on the lymphatic endothelium (100, 115, 118, 119). This in turn promotes DC migration via the lymphatics into the subcapsular space of the lymph node and from there into the lymph node parenchyma along pathways between the B cell follicles, from which they eventually enter the paracortical T cell zone with their antigen cargo. Imaging studies suggest that these newly arriving DCs integrate themselves into a large complex network of resident DCs, where they then are available to interact with T cells (120).

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

528

Castellino

·

Germain

During the period in which this delivery of free antigen, DC-associated antigen, and inﬂammatory signals is occurring, T cells in the naive and central memory recirculating pool are entering and trafﬁcking through the lymph node. These T cells express CCR7 and CD62L; the interactions of CCR7 with CCL21, which has been produced by stromal elements and decorates high endothelial venules (HEV), and of CD62L with its carbohydrate ligand peripheral lymph node addressin, which is highly expressed on the luminal surface of these endothelial cells, lead to rolling, integrin activation, and transendothelial migration of the T cells into the lymph node (42, 100, 121–123). HEV are concentrated in the interfollicular region (IFR) of noninﬂamed lymph nodes, between primary B cell follicles and at what many investigators consider the outer rim of the T cell zone (124). These specialized vessels also occur in certain areas of the paracortex (125). HEV are closely associated with the conduit network through which antigens and cytokines reach the lymph node interior, and the IFR is the zone through which tissue-resident DCs trafﬁc on their way to the deeper paracortical region (107, 108, 124). Thus, recirculating naive and central memory T cells ﬁrst have access to antigen-bearing DCs immediately upon exit from the HEV, where they encounter a dense meshwork of resident DCs that have sampled antigen from the conduit ﬂow and have potentially been activated by PAMPs and inﬂammatory cytokines in this same ﬂuid. Indeed, static sequential imaging studies have shown that T cells ﬁrst bind stably to antigen-bearing DCs in this IFR, rather than more generally throughout the paracortical T zone (110, 124). How do naive CD8+ T cells ﬁnd the right APC? Encounters between naive T cells and DCs have been postulated to depend on their common attraction to CCL19 and CCL21 (126). Such an attraction would bring both DCs and T cells to the same area of the lymph node, but it would not necessarily favor the attraction of naive CD8+ T cells to licensed

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

DCs or DC-CD4+ T cell clusters. Such considerations bring the discussion back to the critical issue of discriminating between random and focused interactions within lymph nodes. One limitation of the existing analyses of lymphocyte movement is that observing a linear relationship between cell displacement and time1/2 is a necessary but not sufﬁcient condition for establishing random walk behavior. Especially if short-range chemoattraction is present, this analysis can be misleading (127, 128). Mempel et al. (86) dealt with this limitation by examining the local turning angle of CD8+ T cells as they encountered DCs, ﬁnding little evidence for a signiﬁcant deviation of naive cells toward the preactivated, antigen-bearing DCs used in this study. However, these experiments did not include antigen-activated CD4+ T cells capable of delivering TH. This left open the possibility that in the presence of a suitable antigen-driven interaction between CD4+ T cells and DCs there would be a directed attraction of CD8+ T cells to the site of this preexisting productive cell-cell contact. Given the discussion above, just such a mechanism in which CD8+ T cells are attracted selectively to DCs that have been or are being licensed by CD4+ T cells would make the most physiological sense. Based on these considerations, we have recently analyzed whether antigen-dependent interactions of CD4+ T cells with DCs affects the migratory properties of naive CD8+ T cells within lymph nodes. Our studies showed that CD8+ T cells preferentially accumulated in an antigen-independent manner in lymph nodes in which CD4+ T cells were undergoing antigen-speciﬁc activation. Intravital 2-photon imaging documented that naive T cells interacted three- to fourfold more often with DCs that bore antigenic complexes recognized by available CD4+ T cells than with neighboring DCs that lacked such antigen but were equivalent in activation state (F. Castellino, manuscript submitted). Because chemokines are the key molecules guiding cell movement in lymphoid com-

partments, some combination of chemokinesis and chemoattraction was the most logical explanation for the focusing of naive CD8+ T cells on DCs engaged in productive interactions with CD4+ T cells. Using a blocking antibody strategy, we looked for a chemokine whose neutralization would eliminate the CD4+ T cell–dependent accumulation and DC interaction behavior we had observed. The results of this screen were unexpected. CCL3 and CCL4 (MIP1β/α), two well-known inﬂammatory chemokines, were responsible for focusing naive CD8+ T cells on licensed DCs and/or antigen-dependent CD4+ T cell–DC clusters. Neither of these chemokines had been reported as playing a role in the initiation of T cell responses. The cognate receptor, CCR5, was believed to be upregulated only after TCR engagement with antigen as part of a differentiation process that produced effector cells with a capacity to home to peripheral sites of inﬂammation (129). This seemingly paradoxical result was explained by our observation that (a) a substantial subpopulation of naive CD8+ T cells in the lymph node draining site of inﬂammation expresses CCR5 prior to cognate antigen recognition; and (b) these cells are sensitive to attraction by CCL3 and CCL4 (F. Castellino, manuscript submitted). The anatomy of the lymph node is quite consistent with these ﬁndings. The conduits that bring inﬂammatory signals from a peripheral infected site open onto the cells that invest HEV and through which naive CD8+ T cells enter the lymph node parenchyma (100, 107, 108). Thus, as they make their way into the lymph node and engage the dense array of antigen-bearing DCs that surround the HEV, naive CD8+ T cells would be exposed to a high concentration of the mediators that we have found can induce CCR5 upregulation, thus preparing these cells for enhanced rates of interaction with the subset of DCs bearing foreign antigen and interacting with the CD4+ T cells responsible for TH function. Blocking CCL3/4mediated focusing of CD8+ T cells on DCs www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

529

ARI

15 February 2006

1:34

that interact with antigen-activated CD4+ T cells leads to the elimination of measurable effects of TH on the long-term development of memory in a nonreplicating vaccine model (F. Castellino, manuscript submitted), illustrating the crucial role of directed migration in this aspect of CD8+ T cell immunity under at least certain immunization conditions. The absence of any known gross defect in antiviral immunity in humans with CCR5 deﬁciency (130, 131) raises the interesting question of whether these ﬁndings on cell recruitment are restricted to the mouse or to certain conditions of immune activation. Our experiments did not resolve the issue of whether the generation of a three-cell cluster, the licensing of the DCs, or both are required for functional delivery of help. We did ﬁnd that CCL3 and CCL4 are secreted by DCs in response to TLR signaling and that this secretion is synergistically increased by CD40 ligation. We also found abundant secretion of the above chemokines by TCRstimulated naive CD4+ and CD8+ T cells (F. Castellino, manuscript submitted). Such a pattern suggests that the DCs that receive a TLR signal and then engage in a productive interaction with CD4+ T cells would be most attractive to a naive CD8+ T cell that has been exposed to inﬂammatory signals and that has expressed CCR5. Interestingly, CCL3 and CCL4 are produced in large amounts by TLR-activated DCs, but their production is not sustained (123). Such transient production by the DCs might have the purpose of requiring the presence of antigen-activated CD4+ T cells to maintain the chemokine gradient, thus rendering activated DC-CD4+ T cell clusters rather than licensed DCs alone as the most likely to recruit naive CD8+ T cells. The transient secretion of CCL3/4 by DCs may also have the role of preferentially attracting naive CD8+ T cells to freshly activated DCs instead of DCs that have been similarly activated earlier in the response and that are already exhausted (132, 133). The key message from these studies is that in lymph nodes,

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

530

Castellino

·

Germain

under immunogenic conditions, interactions between the few relevant cells are optimized by chemokine guidance mechanisms. In the case of CD8+ T cells, both inﬂammation and antigen-speciﬁc activation of CD4+ T cell are required for such a focusing pathway to operate. This allows the immune system to efﬁciently activate CD8+ T cells even though their frequency is low in the naive repertoire, while maintaining the cell-cell cooperation checkpoints required the keep these potentially damaging effector responses under strict control.

THE HOW: SIGNALS INVOLVED IN DELIVERY OF TH The two-signal hypothesis discussed at the beginning of this review was originally based on the activity of two cell types, but it has morphed over the years to “two signaling events by distinct receptor types” involving a single cell (134). More recently, this simple view has grown much more complex, with a large array of simultaneously and sequentially delivered stimuli recognized to have key roles in guiding T cell clonal expansion, effector differentiation, survival, homing propensity, and memory cell performance. Antigen together with a plethora of cell membrane–associated costimulatory molecules (135) as well as secreted and cell-associated cytokines (136, 137) contribute to driving a resting T cell into cell division, modulating the balance of pro- and anti-apoptotic proteins that control cell survival, and altering chromatin structure to permit new gene expression (138, 139). Ligand/ receptor pairs in the immunoglobulin superfamily [particularly CD28-CD80, CD86 and ICOS-ICOS-L (135, 140, 147)], as well as in the tumor necrosis factor superfamily [including CD40-CD40L, 4–1BB-4–1BBL, CD27CD70, CD30-CD30L, and HVEM-LIGHT (141)] have all been reported to play critical and sometimes apparently redundant roles in CD8+ T cell responses. In addition, there is evidence for critical contributions from numerous cytokines during acute phases of

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

CD8+ T cell activation and differentiation (61, 142, 143) and during later phases of the response, when they regulate the balance between proliferation and death of activated and memory T cells (40, 144–148). Studies in animals with selective defects in expression of individual members of the above ligand/receptor pairs have shown that none of them, perhaps with the exception of CD28, has an absolutely essential function in the development of CD8+ T cell responses, but rather each adjusts the balance between proliferation and death, thus affecting the number of T cells that accumulates at the peak of the response and/or that survives to generate the T cell memory pool. This growing body of molecular mediators makes the identiﬁcation of the speciﬁc mechanism(s) by which CD4+ T cells affect CD8+ T cell responses very difﬁcult to unravel. Among all the molecules examined for a role in TH, CD40 has most consistently been reported as critical in this process (72, 135). Maximal expression of CD40 on DCs requires speciﬁc activating signals. PAMP stimulation of TLRs is the best-characterized pathway leading to such upregulation (149, 150), although various forms of mechanical or chemical stress (151) or signaling by other receptors such as C-type lectins (152) may do the same. This increased expression of CD40 in turn appears to be important for effective stimulation of the partially activated DCs by CD40L expressed on antigen-activated CD4+ T cells (149), thus linking innate stimulation with antigen-driven responses to produce optimally matured and functional DCs. CD40 signaling increases MHC display, costimulatory molecule expression, cytokine secretion, and chemokine production, a list that encompasses nearly all the players that amplify or sustain CD8+ T cell responses during either the acute or memory phases. Indeed, in vitro studies are consistent with the notion that CD40-stimulated DCs gain the capacity to promote CD8+ T cell responses under conditions in which CD4+ TH is otherwise required (54). This raises the question of

whether the be-all and end-all of TH activity is promotion of this last step in DC activation, generating APCs that have an optimal display of all the elements that enhance CD8+ activation, effector differentiation, and memory cell survival. A conservative interpretation of these data that we favor is that several different receptorligand combinations with partially overlapping functions and sequential actions all contribute to maintaining CD8+ T cell viability both by synergizing to reach a critical threshold of necessary anti-apoptotic molecules and/or by sustaining the level of such proteins throughout the different phases of the immune response. This concept implies that TH does not inﬂuence CD8+ T cell memory formation through a single master molecule; rather, it produces a population of optimized DCs expressing a host of relevant molecules and at the same time helps insure that naive CD8+ T cells interact preferentially with these special APCs. This concept of TH reﬂecting CD8+ T cell contact with properly signaled DCs raises an intriguing question. If the initial interaction of a naive T cell with a licensed versus an unlicensed DC were to dictate irrevocably the ultimate fate of the activated T cell, then it would be difﬁcult to explain data showing that for each antigen speciﬁcity in an antiviral response a similar fraction of the activated cells present at the peak of the response survives in the memory pool (153). The difﬁculty in reconciling these data arises from the fact that there should be substantial statistical ﬂuctuations in the fraction of naive precursors of each speciﬁcity that ﬁnd one or the other type of APC at the beginning of a response because of the stochastic nature of events at very low precursor numbers. We envision two scenarios able to reconcile these data with our model for delivery of TH. The less likely one is that a statistically similar fraction of the precursors of each speciﬁcity ﬁnds a licensed DC. The alternative is that there is a window of time for antigen-stimulated CD8+ T cell and/or for their early progeny to be rescued by www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

531

ARI

15 February 2006

1:34

receiving help. Such a proposal can be reconciled with our observations if we postulate that the offspring of individual responding CD8+ T cells are attracted to licensed DCs in the same way as naive precursors. This scenario is not unreasonable considering the strong evidence in support of CCR5 upregulation in response to TCR engagement (129, 131). If a subset of each clonally expanded pool made effective contact with licensed DCs, thus acquiring the capacity to survive during the contraction phase of the response and enter the memory pool, this process would greatly reduce the impact of stochastic behavior at a low initial precursor number and could explain why the residual memory population reﬂects the frequency of the expanded pool at the peak of the response. Imaging studies suggest that after initial activation on one antigen-bearing DC, CD8+ T cells contact many other DCs, providing at least some evidence consistent with this proposal (86, 104), but further experiments are needed to test its validity.

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

CONCLUSION The analysis of CD4+ -CD8+ T cell interactions has followed a long and winding road. Although there is still a minority report pending, the present day majority view suggests that TH is a manifestation of CD4+ T cell– induced differentiation of DCs that in turn optimally supports antigen-activated CD8+ T cell survival during the rigors of extensive clonal expansion and the genetic reprogramming required for effector functionality. Our present understanding of the events in this pathway can be summarized as follows: A combination of microbial products and tissue-derived cytokines act on DCs both within lymphoid tissues and in the periphery to induce differentiation events that include the enhanced surface display of MHC and costimulatory molecules, the internalization and processing of protein antigens, and a switch in the pattern of expression of chemokine receptors among peripheral DCs that enables them to migrate to the T cell 532

Castellino

·

Germain

area of draining lymph node. In response to inﬂammation, the DCs increase CD40 expression, becoming more responsive to further stimulation via CD4+ T cells that display CD40L on their membranes because of TCR signaling induced by processed antigen on these innately activated DCs. TLR stimulation and CD40 ligation also synergistically induce the DCs to secrete inﬂammatory cytokines and chemokines, including CCL3 and CCL4. Upon entry into a lymph node, antigen-naive CD8+ T cells, which through exposure to inﬂammatory stimuli have increased their expression of CCR5, are attracted to licensed DCs and/or antigenspeciﬁc clusters of DCs and CD4+ T cells. If the recruited naive CD8+ T cells do not ﬁnd cognate antigen on the DC, their association will last only a few minutes, after which they will leave and engage other DCs in a search for a suitable TCR ligand. If instead the DC does present the relevant antigen, the CD8+ T cell will stop and adhere for a prolonged time, estimated to be up to several hours (103). The effect of chemokine (CCL3/4) production by the CD4+ T cell– DC pair will thus be to enhance the likelihood that rare antigen-speciﬁc CD8+ T cells will ﬁnd this optimal stimulatory environment rather than being incompletely stimulated by less matured DCs that also bear the cognate ligand. In addition, once the recruited CD8+ T cell ﬁnds its cognate antigen on such an optimally activated DC, it can also begin to secrete CCL3 and CCL4. The combined secretion from the CD8+ T cells and the DC will sustain the chemokine gradient toward these licensed DCs, even if the antigen-stimulated CD4+ T cell has already dissociated. This may recruit more naive CD8+ T cells, but perhaps plays a predominant role in attracting the initial progeny of antigen-activated CD8+ T cells, even those ﬁrst stimulated by nonlicensed DCs, thus preparing a representative fraction of all CD8+ speciﬁcities for entry into the memory pool. This concatenation of events ﬁts well with our opening theme of checkpoints for

ANRV270-IY24-17

ARI

15 February 2006

1:34

protection of the host against sustained immune responses that can cause pathology. As just outlined, a coordinated series of events must transpire in the proper order to promote effective memory formation and/or sustain chronic T cells responses, beginning with innate stimuli, progressing to an initial pairwise interaction of lymphocytes and myeloid cells that must be productive to drive recruit-

ment of a second lymphocyte subset and to provide the full range of signals necessary for long-lived responses. This scheme of initial innate signaling followed by chemokinedriven coclustering of cooperating lymphocytes may be a general solution to the immune system’s needle in a haystack problem, without the risk of unwanted autopathological responses.

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ACKNOWLEDGMENTS The authors thank all the members of the Lymphocyte Biology Section, LI, NIAID for helpful discussion and technical advice during the course of our studies on TH. We are especially grateful to Alex Huang for his collaboration in the imaging studies in lymph nodes that helped give rise to some of the concepts developed in this review. We also thank many other colleagues for discussions about the cellular and molecular mechanisms involved in CD8+ T cell responses and memory cell generation.

LITERATURE CITED 1. Bevan MJ. 1989. Immunology. Stimulating killer cells. Nature 342:478–79 2. Doherty PC. 1998. The numbers game for virus-speciﬁc CD8+ T cells. Science 280:227 3. Butz EA, Bevan MJ. 1998. Massive expansion of antigen-speciﬁc CD8+ T cells during an acute virus infection. Immunity 8:167–75 4. Kaech SM, Wherry EJ, Ahmed R. 2002. Effector and memory T-cell differentiation: implications for vaccine development. Nat. Rev. Immunol. 2:251–62 5. Seder RA, Ahmed R. 2003. Similarities and differences in CD4+ and CD8+ effector and memory T cell generation. Nat. Immunol. 4:835–42 6. Rocha B, Tanchot C. 2004. CD8 T cell memory. Semin. Immunol. 16:305–14 7. Miller JD, Masopust D, Wherry EJ, Kaech S, Silvestri G, Ahmed R. 2005. Differentiation of CD8 T cells in response to acute and chronic viral infections: implications for HIV vaccine development. Curr. Drug Targets Infect. Disord. 5:121–29 8. Klenerman P, Hill A. 2005. T cells and viral persistence: lessons from diverse infections. Nat. Immunol. 6:873–79 9. Germain RN. 1994. MHC-dependent antigen processing and peptide presentation: providing ligands for T lymphocyte activation. Cell 76:287–99 10. Lehner PJ, Cresswell P. 2004. Recent developments in MHC-class-I-mediated antigen presentation. Curr. Opin. Immunol. 16:82–89 11. Watts C. 2004. The exogenous pathway for antigen presentation on major histocompatibility complex class II and CD1 molecules. Nat. Immunol. 5:685–92 12. Iwasaki A, Medzhitov R. 2004. Toll-like receptor control of the adaptive immune responses. Nat. Immunol. 5:987–95 13. Lanier LL. 2005. NK cell recognition. Annu. Rev. Immunol. 23:225–74 14. Raulet DH. 2003. Roles of the NKG2D immunoreceptor and its ligands. Nat. Rev. Immunol. 3:781–90 15. Banchereau J, Steinman RM. 1998. Dendritic cells and the control of immunity. Nature 392:245–52 www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

533

ARI

15 February 2006

1:34

16. Lane PJ, McConnell FM. 2001. Dendritic cell subsets and costimulation for effector T-cell responses. Curr. Opin. Pharmacol. 1:409–16 17. Bevan MJ. 2004. Helping the CD8 T cell response. Nat. Rev. Immunol. 4:595–602 18. Segel LA, Bar-Or RL. 1999. On the role of feedback in promoting conﬂicting goals of the adaptive immune system. J. Immunol. 163:1342–49 19. Lohr J, Knoechel B, Nagabhushanam V, Abbas AK. 2005. T-cell tolerance and autoimmunity to systemic and tissue-restricted self-antigens. Immunol. Rev. 204:116–27 20. Redmond WL, Sherman LA. 2005. Peripheral tolerance of CD8 T lymphocytes. Immunity 22:275–84 21. Bretscher P, Cohn M. 1970. A theory of self-nonself discrimination. Science 169:1042–49 22. Metcalf ES, Schrater AF, Klinman NR. 1979. Murine models of tolerance induction in developing and mature B cells. Immunol. Rev. 43:142–83 23. Goodnow CC, Adelstein S, Basten A. 1990. The need for central and peripheral tolerance in the B cell repertoire. Science 248:1373–79 24. Goodnow CC, Brink R, Adams E. 1991. Breakdown of self-tolerance in anergic B lymphocytes. Nature 352:532–36 25. Hartley SB, Crosbie J, Brink R, Kantor AB, Basten A, Goodnow CC. 1991. Elimination from peripheral lymphoid tissues of self-reactive B lymphocytes recognizing membranebound antigens. Nature 353:765–69 26. Guerder S, Matzinger P. 1989. Activation versus tolerance: a decision made by T helper cells. Cold Spring Harb. Symp. Quant. Biol. 54(Pt. 2):799–805 27. Anderson AC, Nicholson LB, Kuchroo VK. 2001. The origin and regulation of autopathogenic T cells. J. Clin. Immunol. 21:74–80 28. Pircher H, Rohrer UH, Moskophidis D, Zinkernagel RM, Hengartner H. 1991. Lower receptor avidity required for thymic clonal deletion than for effector T-cell function. Nature 351:482–85 29. Davey GM, Schober SL, Endrizzi BT, Dutcher AK, Jameson SC, Hogquist KA. 1998. Preselection thymocytes are more sensitive to T cell receptor stimulation than mature T cells. J. Exp. Med. 188:1867–74 30. Lucas B, Stefanova I, Yasutomo K, Dautigny N, Germain RN. 1999. Divergent changes in the sensitivity of maturing T cells to structurally related ligands underlies formation of a useful T cell repertoire. Immunity 10:367–76 31. Reis e Sousa C. 2001. Dendritic cells as sensors of infection. Immunity 14:495–98 32. Bishop GA, Hostager BS. 2001. B lymphocyte activation by contact-mediated interactions with T lymphocytes. Curr. Opin. Immunol. 13:278–85 33. Mills DM, Cambier JC. 2003. B lymphocyte activation during cognate interactions with CD4+ T lymphocytes: molecular dynamics and immunologic consequences. Semin. Immunol. 15:325–29 34. van Lier RA, ten Berge IJ, Gamadia LE. 2003. Human CD8+ T-cell differentiation in response to viruses. Nat. Rev. Immunol. 3:931–39 35. Goldrath AW, Sivakumar PV, Glaccum M, Kennedy MK, Bevan MJ, et al. 2002. Cytokine requirements for acute and basal homeostatic proliferation of naive and memory CD8+ T cells. J. Exp. Med. 195:1515–22 36. Kovanen PE, Leonard WJ. 2004. Cytokines and immunodeﬁciency diseases: critical roles of the γc -dependent cytokines interleukins 2, 4, 7, 9, 15, and 21, and their signaling pathways. Immunol. Rev. 202:67–83 37. Surh CD, Sprent J. 2005. Regulation of mature T cell homeostasis. Semin. Immunol. 17:183–91

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

534

Castellino

·

Germain

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

38. Bradley LM, Haynes L, Swain SL. 2005. IL-7: maintaining T-cell memory and achieving homeostasis. Trends Immunol. 26:172–76 39. Lodolce JP, Boone DL, Chai S, Swain RE, Dassopoulos T, et al. 1998. IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation. Immunity 9:669–76 40. Schluns KS, Lefrancois L. 2003. Cytokine control of memory T-cell development and survival. Nat. Rev. Immunol. 3:269–79 41. Zinkernagel RM. 2000. What is missing in immunology to understand immunity? Nat. Immunol. 1:181–85 42. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401:708– 12 43. Hou S, Hyland L, Ryan KW, Portner A, Doherty PC. 1994. Virus-speciﬁc CD8+ T-cell memory determined by clonal burst size. Nature 369:652–54 44. Gourley TS, Wherry EJ, Masopust D, Ahmed R. 2004. Generation and maintenance of immunological memory. Semin. Immunol. 16:323–33 45. Bach FH, Grillot-Courvalin C, Kuperman OJ, Sollinger HW, Hayes C, et al. 1977. Antigenic requirements for triggering of cytotoxic T lymphocytes. Immunol. Rev. 35:76– 96 46. Keene JA, Forman J. 1982. Helper activity is required for the in vivo generation of cytotoxic T lymphocytes. J. Exp. Med. 155:768–82 47. Rosenberg AS, Mizuochi T, Sharrow SO, Singer A. 1987. Phenotype, speciﬁcity, and function of T cell subsets and T cell interactions involved in skin allograft rejection. J. Exp. Med. 165:1296–315 48. Bennett SR, Carbone FR, Karamalis F, Miller JF, Heath WR. 1997. Induction of a CD8+ cytotoxic T lymphocyte response by cross-priming requires cognate CD4+ T cell help. J. Exp. Med. 186:65–70 49. Ossendorp F, Mengede E, Camps M, Filius R, Melief CJ. 1998. Speciﬁc T helper cell requirement for optimal induction of cytotoxic T lymphocytes against major histocompatibility complex class II negative tumors. J. Exp. Med. 187:693–702 50. Wang JC, Livingstone AM. 2003. Cutting edge: CD4+ T cell help can be essential for primary CD8+ T cell responses in vivo. J. Immunol. 171:6339–43 51. Mitchison NA. 1983. Linked help in the cytotoxic T-cell response revealed by adoptive transfer. Transplant. Proc. 15:2121–24 52. Mitchison NA, O’Malley C. 1987. Three-cell-type clusters of T cells with antigenpresenting cells best explain the epitope linkage and noncognate requirements of the in vivo cytolytic response. Eur. J. Immunol. 17:1579–83 53. Cassell D, Forman J. 1988. Linked recognition of helper and cytotoxic antigenic determinants for the generation of cytotoxic T lymphocytes. Ann. NY Acad. Sci. 532:51–60 54. Ridge JP, Di Rosa F, Matzinger P. 1998. A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393:474–78 55. Gillis S, Smith KA. 1977. Long term culture of tumour-speciﬁc cytotoxic T cells. Nature 268:154–56 56. Smith KA. 1988. Interleukin-2: inception, impact, and implications. Science 240:1169– 76 57. Cousens LP, Orange JS, Biron CA. 1995. Endogenous IL-2 contributes to T cell expansion and IFN-γ production during lymphocytic choriomeningitis virus infection. J. Immunol. 155:5690–99 www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

535

ARI

15 February 2006

1:34

58. Suzuki H, Kundig TM, Furlonger C, Wakeham A, Timms E, et al. 1995. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor beta. Science 268:1472–76 59. Willerford DM, Chen J, Ferry JA, Davidson L, Ma A, Alt FW. 1995. Interleukin-2 receptor alpha chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3:521–30 60. Wagner H, Kronke M, Solbach W, Scheurich P, Rollinghoff M, Pﬁzenmaier K. 1982. Murine T cell subsets and interleukins: relationships between cytotoxic T cells, helper T cells and accessory cells. Clin. Haematol. 11:607–30 61. Deeths MJ, Kedl RM, Mescher MF. 1999. CD8+ T cells become nonresponsive (anergic) following activation in the presence of costimulation. J. Immunol. 163:102–10 62. Granucci F, Vizzardelli C, Pavelka N, Feau S, Persico M, et al. 2001. Inducible IL-2 production by dendritic cells revealed by global gene expression analysis. Nat. Immunol. 2:882–88 63. Granucci F, Feau S, Angeli V, Trottein F, Ricciardi-Castagnoli P. 2003. Early IL-2 production by mouse dendritic cells is the result of microbial-induced priming. J. Immunol. 170:5075–81 64. Feau S, Facchinetti V, Granucci F, Citterio S, Jarrossay D, et al. 2005. Dendritic cellderived IL-2 production is regulated by IL-15 in humans and in mice. Blood 105:697–702 65. Rogers NC, Slack EC, Edwards AD, Nolte MA, Schulz O, et al. 2005. Syk-dependent cytokine induction by Dectin-1 reveals a novel pattern recognition pathway for C type lectins. Immunity 22:507–17 66. Buller RM, Holmes KL, Hugin A, Frederickson TN, Morse HC 3rd. 1987. Induction of cytotoxic T-cell responses in vivo in the absence of CD4 helper cells. Nature 328:77–79 67. Janeway CJ. 1989. Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harb. Symp. Quant. Biol. 1:1–13 68. Matzinger P. 1994. Tolerance, danger, and the extended family. Annu. Rev. Immunol. 12:991–1045 69. Kaisho T, Akira S. 2003. Regulation of dendritic cell function through Toll-like receptors. Curr. Mol. Med. 3:759–71 70. Bennett SR, Carbone FR, Karamalis F, Flavell RA, Miller JF, Heath WR. 1998. Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 393:478–80 71. Schoenberger SP, Toes RE, van der Voort EI, Offringa R, Melief CJ. 1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393:480– 83 72. Grewal IS, Flavell RA. 1998. CD40 and CD154 in cell-mediated immunity. Annu. Rev. Immunol. 16:111–35 73. Moskophidis D, Lechner F, Pircher H, Zinkernagel RM. 1993. Virus persistence in acutely infected immunocompetent mice by exhaustion of antiviral cytotoxic effector T cells. Nature 362:758–61 74. Matloubian M, Concepcion RJ, Ahmed R. 1994. CD4+ T cells are required to sustain CD8+ cytotoxic T-cell responses during chronic viral infection. J. Virol. 68:8056–63 75. Battegay M, Moskophidis D, Rahemtulla A, Hengartner H, Mak TW, Zinkernagel RM. 1994. Enhanced establishment of a virus carrier state in adult CD4+ T-cell-deﬁcient mice. J. Virol. 68:4700–4 76. Cardin RD, Brooks JW, Sarawar SR, Doherty PC. 1996. Progressive loss of CD8+ T cell-mediated control of a gamma-herpesvirus in the absence of CD4+ T cells. J. Exp. Med. 184:863–71

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

536

Castellino

·

Germain

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

77. von Herrath MG, Yokoyama M, Dockter J, Oldstone MB, Whitton JL. 1996. CD4deﬁcient mice have reduced levels of memory cytotoxic T lymphocytes after immunization and show diminished resistance to subsequent virus challenge. J. Virol. 70:1072–79 78. Zajac AJ, Blattman JN, Murali-Krishna K, Sourdive DJ, Suresh M, et al. 1998. Viral immune evasion due to persistence of activated T cells without effector function. J. Exp. Med. 188:2205–13 79. Belz GT, Wodarz D, Diaz G, Nowak MA, Doherty PC. 2002. Compromised inﬂuenza virus-speciﬁc CD8+ -T-cell memory in CD4+ -T-cell-deﬁcient mice. J. Virol. 76:12388– 93 80. Bourgeois C, Rocha B, Tanchot C. 2002. A role for CD40 expression on CD8+ T cells in the generation of CD8+ T cell memory. Science 297:2060–63 81. Janssen EM, Lemmens EE, Wolfe T, Christen U, von Herrath MG, Schoenberger SP. 2003. CD4+ T cells are required for secondary expansion and memory in CD8+ T lymphocytes. Nature 421:852–56 82. Shedlock DJ, Shen H. 2003. Requirement for CD4 T cell help in generating functional CD8 T cell memory. Science 300:337–39 83. Sun JC, Bevan MJ. 2003. Defective CD8 T cell memory following acute infection without CD4 T cell help. Science 300:339–42 84. Gallimore A, Glithero A, Godkin A, Tissot AC, Pluckthun A, et al. 1998. Induction and exhaustion of lymphocytic choriomeningitis virus-speciﬁc cytotoxic T lymphocytes visualized using soluble tetrameric major histocompatibility complex class I-peptide complexes. J. Exp. Med. 187:1383–93 85. Appay V, Dunbar PR, Callan M, Klenerman P, Gillespie GM, et al. 2002. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections. Nat. Med. 8:379–85 86. Mempel TR, Henrickson SE, von Andrian UH. 2004. T-cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature 427:154–59 87. Kaech SM, Ahmed R. 2001. Memory CD8+ T cell differentiation: initial antigen encounter triggers a developmental program in naive cells. Nat. Immunol. 2:415–22 88. van Stipdonk MJ, Lemmens EE, Schoenberger SP. 2001. Naive CTLs require a single brief period of antigenic stimulation for clonal expansion and differentiation. Nat. Immunol. 2:423–29 89. Janssen EM, Droin NM, Lemmens EE, Pinkoski MJ, Bensinger SJ, et al. 2005. CD4+ T-cell help controls CD8+ T-cell memory via TRAIL-mediated activation-induced cell death. Nature 434:88–93 90. Sun JC, Williams MA, Bevan MJ. 2004. CD4+ T cells are required for the maintenance, not programming, of memory CD8+ T cells after acute infection. Nat. Immunol. 5:927–33 91. Kaech SM, Tan JT, Wherry EJ, Konieczny BT, Surh CD, Ahmed R. 2003. Selective expression of the interleukin 7 receptor identiﬁes effector CD8 T cells that give rise to long-lived memory cells. Nat. Immunol. 4:1191–98 92. Huster KM, Busch V, Schiemann M, Linkemann K, Kerksiek KM, et al. 2004. Selective expression of IL-7 receptor on memory T cells identiﬁes early CD40L-dependent generation of distinct CD8+ memory T cell subsets. Proc. Natl. Acad. Sci. USA 101:5610–15 93. Masopust D, Kaech SM, Wherry EJ, Ahmed R. 2004. The role of programming in memory T-cell development. Curr. Opin. Immunol. 16:217–25 94. Lanzavecchia A. 1998. Immunology. Licence to kill. Nature 393:413–14 95. Sun JC, Bevan MJ. 2004. Cutting edge: long-lived CD8 memory and protective immunity in the absence of CD40 expression on CD8 T cells. J. Immunol. 172:3385–89 www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

537

ARI

15 February 2006

1:34

96. Lee BO, Hartson L, Randall TD. 2003. CD40-deﬁcient, inﬂuenza-speciﬁc CD8 memory T cells develop and function normally in a CD40-sufﬁcient environment. J. Exp. Med. 198:1759–64 97. Borrow P, Tishon A, Lee S, Xu J, Grewal IS, et al. 1996. CD40L-deﬁcient mice show deﬁcits in antiviral immunity and have an impaired memory CD8+ CTL response. J. Exp. Med. 183:2129–42 98. Blattman JN, Antia R, Sourdive DJ, Wang X, Kaech SM, et al. 2002. Estimating the precursor frequency of naive antigen-speciﬁc CD8 T cells. J. Exp. Med. 195:657– 64 99. Lipsky PE, Rosenthal AS. 1975. Macrophage-lymphocyte interaction. II. Antigenmediated physical interactions between immune guinea pig lymph node lymphocytes and syngeneic macrophages. J. Exp. Med. 141:138–54 100. von Andrian UH, Mackay CR. 2000. T-cell function and migration. Two sides of the same coin. N. Engl. J. Med. 343:1020–34 101. Miller MJ, Safrina O, Parker I, Cahalan MD. 2004. Imaging the single cell dynamics of CD4+ T cell activation by dendritic cells in lymph nodes. J. Exp. Med. 200:847–56 102. Miller MJ, Hejazi AS, Wei SH, Cahalan MD, Parker I. 2004. T cell repertoire scanning is promoted by dynamic dendritic cell behavior and random T cell motility in the lymph node. Proc. Natl. Acad. Sci. USA 101:998–1003 103. Bousso P, Robey E. 2003. Dynamics of CD8+ T cell priming by dendritic cells in intact lymph nodes. Nat. Immunol. 4:579–85 104. Hugues S, Fetler L, Bonifaz L, Helft J, Amblard F, Amigorena S. 2004. Distinct T cell dynamics in lymph nodes during the induction of tolerance and immunity. Nat. Immunol. 5:1235–42 105. Stoll S, Delon J, Brotz TM, Germain RN. 2002. Dynamic imaging of T cell-dendritic cell interactions in lymph nodes. Science 296:1873–76 106. Shakhar G, Lindquist RL, Skokos D, Dudziak D, Huang JH, et al. 2005. Stable T celldendritic cell interactions precede the development of both tolerance and immunity in vivo. Nat. Immunol. 6:707–14 107. Gretz JE, Anderson AO, Shaw S. 1997. Cords, channels, corridors and conduits: critical architectural elements facilitating cell interactions in the lymph node cortex. Immunol. Rev. 156:11–24 108. Gretz JE, Kaldjian EP, Anderson AO, Shaw S. 1996. Sophisticated strategies for information encounter in the lymph node: the reticular network as a conduit of soluble information and a highway for cell trafﬁc. J. Immunol. 157:495–99 109. Itano AA, McSorley SJ, Reinhardt RL, Ehst BD, Ingulli E, et al. 2003. Distinct dendritic cell populations sequentially present antigen to CD4 T cells and stimulate different aspects of cell-mediated immunity. Immunity 19:47–57 110. Bajenoff M, Granjeaud S, Guerder S. 2003. The strategy of T cell antigen-presenting cell encounter in antigen-draining lymph nodes revealed by imaging of initial T cell activation. J. Exp. Med. 198:715–24 111. Sixt M, Kanazawa N, Selg M, Samson T, Roos G, et al. 2005. The conduit system transports soluble antigens from the afferent lymph to resident dendritic cells in the T cell area of the lymph node. Immunity 22:19–29 112. Itano AA, Jenkins MK. 2003. Antigen presentation to naive CD4 T cells in the lymph node. Nat. Immunol. 4:733–39 113. Cyster JG. 1999. Chemokines and cell migration in secondary lymphoid organs. Science 286:2098–102

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

538

Castellino

·

Germain

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

ARI

15 February 2006

1:34

114. Cyster JG. 1999. Chemokines and the homing of dendritic cells to the T cell areas of lymphoid organs. J. Exp. Med. 189:447–50 115. Sallusto F, Lanzavecchia A. 2000. Understanding dendritic cell and T-lymphocyte trafﬁc through the analysis of chemokine receptor expression. Immunol. Rev. 177:134–40 116. Cyster JG. 2003. Lymphoid organ development and cell migration. Immunol. Rev. 195:5– 14 117. Cyster JG. 2005. Chemokines, sphingosine-1-phosphate, and cell migration in secondary lymphoid organs. Annu. Rev. Immunol. 23:127–59 118. Sozzani S, Allavena P, D’Amico G, Luini W, Bianchi G, et al. 1998. Differential regulation of chemokine receptors during dendritic cell maturation: a model for their trafﬁcking properties. J. Immunol. 161:1083–86 119. Yanagihara S, Komura E, Nagafune J, Watarai H, Yamaguchi Y. 1998. EBI1/CCR7 is a new member of dendritic cell chemokine receptor that is up-regulated upon maturation. J. Immunol. 161:3096–102 120. Lindquist RL, Shakhar G, Dudziak D, Wardemann H, Eisenreich T, et al. 2004. Visualizing dendritic cell networks in vivo. Nat. Immunol. 5:1243–50 121. Gunn MD, Tangemann K, Tam C, Cyster JG, Rosen SD, Williams LT. 1998. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc. Natl. Acad. Sci. USA 95:258–63 122. Campbell JJ, Bowman EP, Murphy K, Youngman KR, Siani MA, et al. 1998. 6-C-kine (SLC), a lymphocyte adhesion-triggering chemokine expressed by high endothelium, is an agonist for the MIP-3β receptor CCR7. J. Cell Biol. 141:1053–59 123. Sallusto F, Kremmer E, Palermo B, Hoy A, Ponath P, et al. 1999. Switch in chemokine receptor expression upon TCR stimulation reveals novel homing potential for recently activated T cells. Eur. J. Immunol. 29:2037–45 124. Katakai T, Hara T, Lee JH, Gonda H, Sugai M, Shimizu A. 2004. A novel reticular stromal structure in lymph node cortex: an immuno-platform for interactions among dendritic cells, T cells and B cells. Int. Immunol. 16:1133–42 125. Anderson AO, Anderson ND, White JD. 1982. Lymphocyte locomotion, lymphatic tissues and lymphocyte circulation in the rat. In Animal Models of Immunological Processes, ed. JB Hay, pp. 26–95. London: Academic 126. Luther SA, Tang HL, Hyman PL, Farr AG, Cyster JG. 2000. Coexpression of the chemokines ELC and SLC by T zone stromal cells and deletion of the ELC gene in the plt/plt mouse. Proc. Natl. Acad. Sci. USA 97:12694–99 127. Huang AY, Qi H, Germain RN. 2004. Illuminating the landscape of in vivo immunity: insights from dynamic in situ imaging of secondary lymphoid tissues. Immunity 21:331– 39 128. Germain RN, Castellino F, Chieppa M, Egen J, Huang AYC, et al. 2005. An extended vision for dynamic high-resolution intravital immune imaging. Sem. Immunol. 17:431–41 129. Oppermann M. 2004. Chemokine receptor CCR5: insights into structure, function, and regulation. Cell. Signal. 16:1201–10 130. Berger EA, Murphy PM, Farber JM. 1999. Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu. Rev. Immunol. 17:657–700 131. Galvani AP, Novembre J. 2005. The evolutionary history of the CCR5-Delta32 HIVresistance mutation. Microbes Infect. 7:302–9 132. Reis e Sousa C, Yap G, Schulz O, Rogers N, Schito M, et al. 1999. Paralysis of dendritic cell IL-12 production by microbial products prevents infection-induced immunopathology. Immunity 11:637–47 www.annualreviews.org • CD 4+ and CD8+ T Cell Cooperation

539

ARI

15 February 2006

1:34

133. Langenkamp A, Messi M, Lanzavecchia A, Sallusto F. 2000. Kinetics of dendritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat. Immunol. 1:311–16 134. Schwartz RH. 2003. T cell anergy. Annu. Rev. Immunol. 21:305–34 135. Greenwald RJ, Freeman GJ, Sharpe AH. 2005. The B7 family revisited. Annu. Rev. Immunol. 23:515–48 136. Trinchieri G. 2003. Interleukin-12 and the regulation of innate resistance and adaptive immunity. Nat. Rev. Immunol. 3:133–46 137. Watts TH. 2005. TNF/TNFR family members in costimulation of T cell responses. Annu. Rev. Immunol. 23:23–68 138. Ansel KM, Lee DU, Rao A. 2003. An epigenetic view of helper T cell differentiation. Nat. Immunol. 4:616–23 139. Reiner SL. 2005. Epigenetic control in the immune response. Hum. Mol. Genet. 14(Suppl. 1):R41–46 140. Acuto O, Michel F. 2003. CD28-mediated co-stimulation: a quantitative support for TCR signalling. Nat. Rev. Immunol. 3:939–51 141. Croft M. 2003. Co-stimulatory members of the TNFR family: keys to effective T-cell immunity? Nat. Rev. Immunol. 3:609–20 142. Trinchieri G. 1995. Interleukin-12: a proinﬂammatory cytokine with immunoregulatory functions that bridge innate resistance and antigen-speciﬁc adaptive immunity. Annu. Rev. Immunol. 13:251–76 143. Schmidt CS, Mescher MF. 1999. Adjuvant effect of IL-12: conversion of peptide antigen administration from tolerizing to immunizing for CD8+ T cells in vivo. J. Immunol. 163:2561–67 144. Marrack P, Bender J, Hildeman D, Jordan M, Mitchell T, et al. 2000. Homeostasis of αβ TCR+ T cells. Nat. Immunol. 1:107–11 145. Nakajima H, Noguchi M, Leonard WJ. 2000. Role of the common cytokine receptor gamma chain (γc) in thymocyte selection. Immunol. Today 21:88–94 146. Plas DR, Rathmell JC, Thompson CB. 2002. Homeostatic control of lymphocyte survival: potential origins and implications. Nat. Immunol. 3:515–21 147. Croft M. 2003. Co-stimulatory members of the TNFR family: keys to effective T-cell immunity? Nat. Rev. Immunol. 3:609–20 148. Zeng R, Spolski R, Finkelstein SE, Oh S, Kovanen PE, et al. 2005. Synergy of IL-21 and IL-15 in regulating CD8+ T cell expansion and function. J. Exp. Med. 201:139–48 149. Schulz O, Edwards AD, Schito M, Aliberti J, Manickasingham S, et al. 2000. CD40 triggering of heterodimeric IL-12 p70 production by dendritic cells in vivo requires a microbial priming signal. Immunity 13:453–62 150. O’Sullivan B, Thomas R. 2003. CD40 and dendritic cell function. Crit. Rev. Immunol. 23:83–107 151. Gallucci S, Lolkema M, Matzinger P. 1999. Natural adjuvants: endogenous activators of dendritic cells. Nat. Med. 5:1249–55 152. Geijtenbeek TB, van Vliet SJ, Engering A, t’Hart BA, van Kooyk Y. 2004. Self- and nonself-recognition by C-type lectins on dendritic cells. Annu. Rev. Immunol. 22:33–54 153. Blattman JN, Sourdive DJ, Murali-Krishna K, Ahmed R, Altman JD. 2000. Evolution of the T cell repertoire during primary, memory, and recall responses to viral infection. J. Immunol. 165:6081–90

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-17

540

Castellino

·

Germain

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:519-540. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung,1 Cosmas Giallourakis,2 Raul Mostoslavsky,1 and Frederick W. Alt1 1

Howard Hughes Medical Institute, Children’s Hospital, CBR Institute for Biomedical Research, and Department of Genetics, Harvard Medical School, Boston, Massachusetts 02115; email: david [email protected]; [email protected]; [email protected]

2

Gastrointestinal Unit, Massachusetts General Hospital, CBR Institute for Biomedical Research, Boston, Massachusetts 02115; email: [email protected]

Annu. Rev. Immunol. 2006. 24:541–70 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.23.021704.115830 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0541$20.00

Key Words IgH, lymphocyte development, transcriptional control elements, allelic exclusion, gene expression

Abstract V(D)J recombination assembles antigen receptor variable region genes from component germline variable (V), diversity (D), and joining ( J) gene segments. For B cells, such rearrangements lead to the production of immunoglobulin (Ig) proteins composed of heavy and light chains. V(D)J is tightly controlled at the Ig heavy chain locus (IgH) at several different levels, including cell-type speciﬁcity, intra- and interlocus ordering, and allelic exclusion. Such controls are mediated at the level of gene segment accessibility to V(D)J recombinase activity. Although much has been learned, many longstanding questions regarding the regulation of IgH locus rearrangements remain to be elucidated. In this review, we summarize advances that have been made in understanding how V(D)J recombination at the IgH locus is controlled and discuss important areas for future investigation.

541

ANRV270-IY24-18

ARI

15 February 2006

1:38

INTRODUCTION V(D)J: variable (V), diversity (D), and joining ( J) gene segments IgH: immunoglobulin heavy chain

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

IgL: immunoglobulin light chain TCR: T cell receptor RAG: recombinationactivating gene DSB: double-strand break RS: recombination signal sequence

Lymphocytes can initiate speciﬁc immune responses against antigens by generating a nearly inﬁnite diversity of antigen receptors within the constraints of a ﬁnite genome (1). This remarkable feat is achieved in large part by a somatic recombination process known as V(D)J recombination. Through V(D)J recombination, the variable region of antigen receptor genes is assembled from component V, D, and J gene segments. There are seven different loci that are rearranged to generate the antigen receptors of T and B lymphocytes. These include the immunoglobulin (Ig) heavy chain locus (IgH) and the Ig light chain (IgL) loci (Igκ and Igλ), which encode the antigen receptor and secreted antibodies of B cells, and the T cell receptor (TCR) β,δ and α,γ loci. The rearrangement of these loci is tightly controlled in a lineage-, stage-, and allelespeciﬁc manner. Regulation of V(D)J recombination in these contexts requires a complex orchestration of DNA, chromatin structure, and trans-acting factors in collaboration with transcriptional activation and DNA breakage and repair mechanisms. As the immune system is not required for survival in a pathogenfree environment, the antigen receptor loci provide a system uniquely suited to advance our basic understanding of these fundamental cellular processes, as well as their roles in both normal and dysfunctional states. In this review, we examine advances in understanding regulation of V(D)J recombination at the IgH locus.

OVERVIEW OF THE BASIC V(D)J RECOMBINATION REACTION AND DIVERSITY GENERATION The basic subunit of an Ig or antibody molecule is a pair of identical Ig heavy chains and a pair of identical Ig light chains. The N-terminal portion of the heavy chain and light chain has a variable and unique amino acid sequence (variable region), and is involved in speciﬁc antigen binding. By con542

Jung et al.

trast, the C-terminal portion of these chains comes in only a few different forms and is termed the constant region, which prescribes the class and effector function of an antibody molecule. Variable regions of antigen receptors, including IgH and IgL chains and TCR chains, are assembled from germline V, D, and J gene segments through a site-speciﬁc recombination reaction known as V(D)J recombination (2, 3). V(D)J recombination only functions in developing T and B lymphocytes. The recombination-activating genes 1 and 2 (RAG1 and RAG2) together form the RAG endonuclease (RAG), which is sufﬁcient in vitro (4, 5) and necessary in vivo (6, 7) to initiate the cleavage phase of V(D)J recombination. RAG introduces DNA double-strand breaks (DSBs) speciﬁcally at the borders between two coding segments and their ﬂanking recombination signal sequences (RSs). RSs are comprised of a highly conserved heptamer and nonamer, separated by a relatively nonconserved spacer of either 12 or 23 base pairs (bp) (8). RAG function requires that one RS have a 12-bp spacer and the other a 23-bp spacer for efﬁcient recognition and DSB formation (9, 10), a restriction referred to as the 12/23 rule (11). Following further processing of the ends, the resolved DSBs are repaired by ubiquitously expressed nonhomologous end joining proteins to generate coding and RS joints (12).

OVERVIEW OF THE IgH VARIABLE REGION LOCUS: GENOMIC STRUCTURE AND REARRANGEMENT The murine IgH locus spans approximately 3 Mb near the telomeric end of chromosome 12 (Figure 1) (13). Depending on the strain, 150 or more functional VH gene segments comprise 15 VH segment families, arrayed over 2.7 Mb, starting approximately 100 kb upstream of 12–13 DH gene segments. VH genes that lie closer to the DH gene segments have been termed “proximal,” whereas VH genes that lie further upstream have been termed distal.

ANRV270-IY24-18

ARI

15 February 2006

1:38

2.7Mb

97kb

52kb

0.8kb

200kb

iEμ Cμ 15 V gene families J558 (~76)

D

Cα 3’RR

J

7183(~26) 81x

DFL16.1

DQ52

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Figure 1 Genomic organization of the IgH locus (not drawn to scale). The VH 81X, DFL16.1, and DQ52 gene segments are highlighted in yellow. Distances shown are from the murine 129 strain and vary between strains.

The DH segments are located in a roughly 50 kb region upstream of the four JH gene segments, which are, in turn, located within a 2 kb region starting approximately 700 bp downstream of DQ52, the most 3 DH gene segment. Starting with the Cμ exons that lie about 7 kb downstream of the JH gene segments, the various constant region exons are spread across 200 kb (13). Although this review focuses primarily on the mouse IgH variable region locus, the human locus has a similar organization (Figure 1) (13). Based on past usage, we refer to a DH to JH rearrangement as DJH and a VH to DJH rearrangement as VH DJH . Germline VH and DH gene segments are ﬂanked by upstream transcriptional promoters, which initiate transcription in a developmentally regulated fashion and may be important for activation and regulation of IgH rearrangements (14–16). Assembly of the complete VH DJH variable region at the JH locus places the germline VH promoter in close proximity to a strong enhancer element (referred to as iEμ) that lies in the intron between the JH and Cμ exons. Transcripts initiated from the promoter of the rearranged VH segment run through the Cμ constant region exons, and RNA splicing assembles the mature heavy chain mRNA with the variable region exon spliced to the Cμ constant region exons via splice donor and acceptor signals encoded 3 to the JH segment and 5 to the ﬁrst Cμ exon (Figure 2b) (17)

Several different mechanisms generate IgH variable region diversity with respect to V(D)J recombination (Figure 2). In assembled IgH variable exons, the bulk of the exon is encoded by the germline VH segment, which also contains two of the three complementarity determining (antigen contact) regions (CDRs). These two CDRs are different in different germline VH segments, which provides germline-encoded diversity. In addition, the VH , DH , JH junctional region encodes the third CDR (Figure 2b). Random assortment of the different germline VH , DH , and JH segments is therefore a source of somatic diversity. Finally, the combinatorial diversity generated by the different combinations of germline V, D, and J segments is further augmented by diversiﬁcation of D to J and V to D junctions during V(D)J recombination through deletion of nucleotides and nontemplated nucleotide additions (N regions) by terminal transferase to generate a vast diversity of different IgH variable region gene exons (18). Although junctional diversiﬁcation mechanisms typically only lead to deletion or addition of a limited number of nucleotides, the actual number is usually random. Therefore, given that the ATG translation initiation codon at the 5 end of the VH exon is ﬁxed within a triplet reading frame, and given that the genetically encoded in-frame splice junction between the JH and the Cμ exon is also ﬁxed, only a fraction of www.annualreviews.org • V(D)J Recombination

CDR: complementarity determining region

543

ANRV270-IY24-18

ARI

15 February 2006

1:38

a VH

23

12

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

VH

DH

12

23

12

23

JH

DJH

VDJH

b L

V 1

DJH

2 3 CDR

iEμ

Cμ s m

Figure 2 Assembly and expression of IgH genes. (a) Variable (VH ), diversity (DH ), and joining ( JH ) gene segments are shown, along with their ﬂanking recombination signal sequences (RSs). RS heptamers are depicted as yellow triangles, whereas RS nonamers are depicted as white triangles. Spacer lengths are indicated above the various RSs. (b) Location of the three complementarity determining regions (CDRs) on the assembled VH DJH exon is shown. Transcription initiates upstream of the assembled VH DJH exon and proceeds through the four Cμ exons and the membrane (m) and secreted (s) exons. Possible splicing events are indicated.

VH DJH rearrangements place the VH ATG in-frame with the coding sequence of the JH gene segment to generate a μ protein (termed productive; Figure 2). Out-of-frame rearrangements encode full-length and processed transcripts but cannot encode a fulllength heavy chain protein and, therefore, are termed nonproductive. Nonproductive rearrangements, which ﬁgure signiﬁcantly into mechanisms that regulate IgH locus V(D)J recombination, can therefore be considered a necessary by-product of the imprecise mechanisms that create junctional diversity. VH and JH gene segments are ﬂanked by 23 RSs, whereas DH gene segments are ﬂanked 544

Jung et al.

on both sides by 12 RSs. The 12/23 rule therefore dictates that the vast majority of IgH locus rearrangements involving VH gene segments will contain a DH gene segment sandwiched between a VH and a JH gene segment (Figure 2a) (11, 19). Inversional DH to JH rearrangements, which are technically permitted by the 12/23 rule, have been described (20–22), but the vast majority of DH to JH joints occur by deletion (21, 22). Investigators have suggested that RS preferences may underlie this bias, as JH 23 RSs displayed a roughly 20-fold preference for 3 DH 12 RSs versus 5 DH 12 RSs in transfected recombination substrates (23, 24). Another

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

well-described example of rearrangement patterns driven by RS preference has been elucidated for the TCRβ locus, where Vβ segments are ﬂanked by 23 RSs and Jβ segments are ﬂanked by 12 RSs, whereas Dβ segments are ﬂanked by 5 12 RSs and 3 23 RSs to mediate joining to the Vβ and Jβ gene segments, respectively. With this organization, Vβ segments could theoretically join directly to Jβ segments under the 12/23 rule, but they are prevented from doing so by a “beyond 12/23” restriction under which 3 Dβ 23 RSs can mediate joining efﬁciently with the Jβ 12 RSs but the Vβ 23 RSs cannot (25, 26). In this regard, replacement of a Vβ 23 RS with a 3 Dβ1 23 RS can greatly increase the frequency of Vβ utilization and can promote direct Vβ to Jβ rearrangement (27). Similar replacement experiments to exchange a 5 DH 12 RS with a 3 DH 12 RS could directly test whether beyond 12/23 restrictions also prevent inversional DH to JH rearrangements. Progressive stages of early B cell development have been well deﬁned on the basis of expression of various cell surface markers and ordered patterns of IgH and IgL chain gene rearrangements (28). Assembly of VH , DH , and JH segments during the pro-B cell differentiation stage into a complete VH DJH variable region gene generally occurs before that of IgL chain genes (29). Productive assembly of a VH DJH exon leads to the generation of a μ heavy chain protein that signals differentiation to the pre-B cell stage by signaling through the pre-B cell receptor (pre-BCR; see below). RAG mRNA and protein expression are also shut off by various mechanisms once a productive IgH rearrangement occurs but are reexpressed during IgL rearrangement in subsequent developmental stages (30–32). Igκ and Igλ light chain genes are assembled in the pre-B cell stage, with Igκ rearrangement generally preceding that of Igλ (29). Productive IgL gene assembly leads to production of a light chain protein that can associate with the preexisting heavy chain and together be deposited on the cell surface in the form of a membrane-bound IgM surface antigen recep-

tor known as the B cell receptor (BCR). The resulting immature B cells can migrate to the periphery where, upon appropriate stimulation with a cognate antigen, they can be activated to secrete their previously membranebound BCR as IgM antibodies. The shift from production of a membrane-bound IgM BCR to a secreted IgM antibody is mediated at the level of differential RNA processing of a set of membrane versus secreted exons at the 3 end of the Cμ coding exons (Figure 2a) (33, 34).

BCR: B cell receptor

ORDERED REARRANGEMENT OF THE IgH LOCUS Early studies demonstrated that μ heavy chains are expressed before IgL chains during B cell development (35). Subsequently, investigators showed that this ordered expression of IgH versus IgL chains occurs because IgH genes rearrange before IgL genes (36–38). One set of studies showed that Abelson murine leukemia virus (AMuLV)-transformed pro-B cell lines contain rearrangements of both JH gene segments in the absence of IgL rearrangements (39). Further studies showed that, in normal ex vivo sorted cell populations, early B cells similarly rearrange both of their IgH alleles before proceeding to rearrange their IgL genes (40). Such studies therefore demonstrated that rearrangement of IgH versus IgL loci is ordered during B cell development, a process that is thought to, at least in part, have evolved to allow proper regulation for variable region assembly in the context of feedback regulation (41). Ordered Ig gene rearrangement also extends to the order in which the gene segments of the IgH variable region are assembled. Thus, IgH DH to JH joints can occur in some developing T cells, whereas VH to DJH joints do not (42). Subsequent studies of a large number of immature and mature B cell lines, with ﬁndings conﬁrmed for sorted normal B cell populations, clearly demonstrated that DH to JH joining occurs on both IgH www.annualreviews.org • V(D)J Recombination

545

ANRV270-IY24-18

ARI

15 February 2006

1:38

Pre–Pro–B cell

Pro–B cell

Pre–B cells Feedback inhibition

μ+ ~60% VDJH+ DJH

DJH VH

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

DH – JH DH – JH

D

JH

DJH DJH

~1/3 productive μ chain expression

VH

Feedback inhibition

VDJH– DJH

VDJH– VDJH+

DJH

VH

DJH

~2/3 non– productive

induced by μ protein

~1/3 productive μ chain expression VH

DJH ~2/3 non– productive

μ+ ~40%

DEATH

sIg μ+ L+

VL

JL

Joining (κ before λ)

Productive Nonproductive μ+ L– μ+

VDJH– VDJH–

Germline VH transcripts Antisense genic/intergenic VH transcripts Figure 3 The regulated model of allelic exclusion. Partially adapted from Reference 16. See text for details.

alleles before VH to DJH joining on one allele, as nearly all the cell lines analyzed harbored DJH /DJH , VH DJH /DJH , or VH DJH /VH DJH rearrangements at their two IgH loci (43). Furthermore, only one VH DJH rearrangement in each cell line produced a functional μ chain (43). Moreover, although VH to DH rearrangements (in the absence of DH to JH joining) are technically permitted by the 12/23 rule, they were not observed (43). Taken together, these analyses showed that IgH locus rearrangements are ordered, with DH to JH joining occurring on both alleles before the initiation of VH to DJH joining (Figure 3). These observations, coupled with the ﬁndings of DH to JH but not VH to DJH rearrangement in T cells (42), led to the suggestion that the VH to DJH step is the one that is regulated in the context of lineage speciﬁcity and, as discussed below, in the context of feedback regulation and allelic exclusion (38, 43). Later studies showed that assembly of Vβ, Dβ, and Jβ gene segments at the TCRβ locus is sim546

Jung et al.

ilarly ordered, with DJβ joining taking place before Vβ to DJβ joining (44), despite the fact that Vβ to Dβ joining would be permitted by the 12/23 rule, and that Vβ to DJβ rearrangement also appears to be the regulated step (45).

ALLELIC EXCLUSION OF IgH LOCUS REARRANGEMENTS The clonal selection theory of acquired immunity is based on the assumption that, despite the fact that two alleles of every antigen receptor locus are theoretically available for rearrangements in every developing lymphocyte, B and T cells of the adaptive immune system each express a unique antigen receptor. This unique expression results in a functional allelic exclusion of the nonexpressed alleles. The clonal selection theory further postulates that speciﬁc binding between an antigen receptor and its cognate antigen stimulates cellular proliferation and, ultimately,

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

increased production of the receptor molecule (46). Conversely, the clonal selection theory also suggests a solution to the related problem of autoimmunity, whereby antigen receptor recognition of self-antigens results in elimination of self-reactive lymphocytes. The observation of allelic exclusion supports this theory, as over 99% of peripheral B cells express a single antigen receptor composed of a unique heavy chain generated from only one of two IgH alleles, and a unique light chain generated from only one of several IgL (two for Igκ and four for Igλ) alleles (47, 48). At the IgH locus, cells expressing two different μ chains occur at an estimated frequency of less than 1 in 104 (49). Other antigen receptor loci that are similarly allelically excluded include the Igκ and Igλ light chain loci, as well as the TCRβ locus in T cells.

Models for Allelic Exclusion Despite intense investigation, a precise mechanistic understanding of the enforcement of allelic exclusion remains elusive (50). Although there have been many models for allelic exclusion, two general models have prevailed, the regulated models (43, 51) and the unregulated, stochastic models (52–54). All these theories must accommodate the fact that V(D)J rearrangements are inherently imprecise owing to mechanisms that maximize junctional diversity. Junctional diversiﬁcation mechanisms dictate that only about one in three V(D)J joints will align the joined gene segments in the correct translational reading frame, although this fraction is merely an estimate for the IgH locus, as some VH segments have crippling germline mutations (13), and many DJH joints cannot form functional μ chains owing to in-frame stop codons within the DJH joint or counterselection of certain DJH joints (55, 56). At the Igκ locus in immature B cells that are undergoing receptor editing, nonfunctional Vκ to Jκ joints can be replaced by “leapfrogging” of an upstream Vκ to a downstream Jκ, a process in accord with the 12/23 rule (57). As VH to DJH joining

deletes all intervening DH segments, however, VH DJH rearrangements are usually ﬁxed because of the 12/23 incompatibility of direct VH to JH joining. A minority of nonfunctional VH DJH joints may be rescued by VH editing mechanisms that use an internal VH heptamer to join an upstream germline VH to an assembled VH DJH (58, 59), thereby potentially rescuing a nonproductive rearrangement (60). Of the prevailing theories, much evidence supports a regulated model for the establishment of IgH allelic exclusion (Figure 3) (50). A regulated model was ﬁrst invoked to explain control of IgL rearrangements, and proposed that IgL chain V gene assembly must proceed on one chromosome at a time and that protein products generated from a functional IgL (i.e., that could associate with the preexisting heavy chain) rearrangement mediate allelic exclusion via feedback inhibition of further IgL assembly (51). This model was based, in part, on the observation that Igκ-expressing B cells could be divided into a subset of about 40% that had rearranged both Igκ alleles (one nonfunctionally) and another subset of about 60% that had rearranged only one allele and left the other in germline conﬁguration (52, 53). The observation that IgH locus rearrangements are ordered revealed a potential parallel between feedback regulation of both IgL and IgH rearrangements. Therefore, because both IgH alleles are always rearranged in B lineage cells in the form of DJH rearrangements, if the DJH rearrangement is considered equivalent to an Igκ germline allele, then VH to DJH joining occurs in the same patterns and proportions of productive, nonproductive, and nonrearranged alleles as described for Igκ alleles (Figure 3) (43). These ﬁndings led to the notion that for both IgL and IgH chain rearrangements, control of V segment rearrangement is the allelically excluded step. Under a regulated model for IgH allelic exclusion, analogous to that originally proposed for Igκ rearrangement, pro-B cells that make a productive V(D)J rearrangement on the ﬁrst attempt do not initiate VH to DJH joining on the second allele, which freezes www.annualreviews.org • V(D)J Recombination

547

ARI

15 February 2006

1:38

a DJH on the second allele (Figure 3). Further, this model predicts that cells that make a nonfunctional VH DJH joint on the ﬁrst allele will subsequently rearrange VH to DJH on the second allele, leading to the population of mature lymphocytes carrying VH DJH rearrangements on both alleles (one productive and one nonproductive). B cells that are nonproductively rearranged VH DJH on both alleles cannot express Ig on their surface and therefore generally do not survive (61), although some may be rescued by VH replacement mechanisms (see below). Approximately 50%–60% of mature B cells have one IgH allele rearranged as a VH DJH and one allele frozen at DJH , whereas about 40%–50% have two VH DJH rearrangements in productive/nonproductive conﬁguration (Figure 3) (43, 50). These percentages are consistent with a productive rearrangement level of 33% or less and with feedback regulation by a μ chain to inhibit further rearrangement speciﬁcally at the VH to DJH joining step on the second allele (Figure 3) (43, 53). This proportion of productive versus nonproductive rearrangements is referred to as the 60/40 ratio for historical reasons, even though this ratio is an estimate owing to DH reading frame incompatibility and VH pseudogenes (56). As noted above, further evidence that VH to DJH joining is the allelically excluded and regulated step comes from studies showing that DJH joints occur in thymocytes, whereas VH DJH joints do not (42, 62, 63); in addition, IgH alleles that are DJH rearranged can continue to undergo DJH joining during IgL chain rearrangement, whereas VH DJH joining does not occur at this stage (64). The fact that rearrangement patterns of Igκ alleles also ﬁt the 60/40 ratio supported the notion that Vκ to Jκ rearrangements are feedback regulated in a similar fashion (16, 53). Finally, the TCRβ locus is also allelically excluded and undergoes ordered rearrangement in a pattern similar to the IgH locus, consistent with feedback regulation by productive TCRβ chains (65). In summary, control of V gene segment rearrangement is the

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

548

Jung et al.

regulated step under a feedback model for allelic exclusion. An alternative stochastic model to explain IgL allelic exclusion has proposed that the ﬁrst allele to rearrange is selected by chance, and that the V(D)J rearrangement process is so inefﬁcient that the likelihood of generating two productive rearrangements in the same cell is vanishingly low (52, 53). One recent extension of this proposal to explain the basis for Igκ allelic exclusion is that variegated, lowprobability activation of a locus for V(D)J recombination, in conjunction with allelic competition for a limited number of activating proteins (such as transcription factors), can result in allelic exclusion (54). This model is based on data generated from a Igκ reporter allele that suggested only a small percentage of cells transcriptionally activate the Igκ locus and that those cells go on to rearrange their Igκ loci (54). However, one prediction of this model is that, among normal pre-B cells, a very large population should be unrearranged on both alleles, which thus far has not been found (66). In any case, if the inefﬁcient rearrangement model were to be applied to the IgH locus, it would have to account for the fact that all IgH alleles undergo DH to JH rearrangements, indicating that DH to JH joining is quite efﬁcient. The inefﬁciency would therefore have to be speculated to occur at the VH to DJH rearrangement stage. However, as for Vκ to Jκ joining at the Igκ locus, the observation that nearly half of all mature B cells contain nonproductive VH to DJH rearrangements (43) suggests that, in its simplest form, an inefﬁcient rearrangement model cannot fully explain IgH locus allelic exclusion (50). Yet, there are still scenarios whereby these rearrangement patterns might be rationalized with inefﬁciency (50), and additional experiments are needed to fully address this model. Therefore, although aspects of both stochastic and regulated models of allelic exclusion with regard to the rearrangement patterns that are observed at the IgH locus could apply, most current evidence strongly supports a feedback regulated model.

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

The precise mechanistic relationship between ordered rearrangement and allelic exclusion remains to be fully elucidated. However, it is worth noting that allelically excluded antigen receptor loci containing D gene segments, including the IgH and TCRβ loci, rearrange in an ordered, regulated manner, whereas the TCRδ locus, which also contains V, D, and J segments, does not undergo ordered rearrangement and is not allelically excluded (67). These observations are consistent with the idea that ordered rearrangement is important to allow time for a V to DJ rearrangement to be functionally evaluated by a feedback mechanism (43). This long-standing hypothesis has not been directly tested, although in principle it could be evaluated by generating gene-targeted mutations that break the order of rearrangement and allow VH segments to join directly to unrearranged DH segments or to rearrange directly to JH segments. If the timing of VH gene segment rearrangements is directly linked with the order in which DH to JH and VH to DJH rearrangements take place, such a mutation would be predicted to disrupt allelic exclusion.

Feedback Regulation of IgH Locus Rearrangements: Evidence and Open Questions Studies on the regulation of IgL recombination revealed that functional light chain protein production—as opposed to light chain gene expression—is required to block further light chain rearrangements (51). This line of reasoning was extended to test the functional requirements for feedback regulation of IgH locus rearrangements and, in particular, to see if a membrane-bound μ chain is required to signal allelic exclusion. Transgenic studies ﬁrst provided direct experimental evidence for regulated allelic exclusion, where expression of a functional antigen receptor protein partially blocked the rearrangement of endogenous loci (68–70). In most of these early studies, however, both the secreted and membrane-bound forms of the μ chain were

expressed (69), necessitating further work to address whether cell surface expression of the μ chain is necessary for feedback regulation. Direct evidence for this proposal has subsequently come both from additional transgenic studies (71), which showed that transmembrane but not secreted μ chains blocked endogenous IgH rearrangements, and from endogenous deletion of the heavy chain transmembrane exon (72), which resulted in allelic inclusion owing to inability to shut off rearrangements of the other allele. Thus, the requirement for cell surface expression of a μ chain for allelic exclusion directly supports a feedback regulated model for IgH rearrangements. Similar experiments also showed that it was the membrane-bound form of the μ chain that directs progression from the pro-B to the pre-B stage, where IgL rearrangement is initiated (64). In this regard, the observation that, in accordance with a feedback regulated model for rearrangements, roughly 40% of mature B cells that have completed all V(D)J rearrangements are VH DJH rearranged on one allele and remain DJH rearranged on the second allele provides strong evidence that continued VH to DJH rearrangements do not occur at the pre-B cell stage (43). DSBs at 5 DH 12 RSs, which would correspond to VH to DJH joining, are only present in pro-B cells and are not present in pre-B cells, when IgL loci rearrange (73). This retargeting of RAG appears to be speciﬁc for the VH to DJH joining step, however, as studies in early B cell lines have demonstrated that a DJH joint can be replaced by successive DJH joining during Vκ to Jκ joining (64), further supporting the notion that VH to DJH joining is the allelically excluded step in IgH locus recombination. Thus, the available evidence strongly supports the idea that VH to DJH joining is limited to the pro-B cell stage and does not occur during the pre-B cell stage even though RAG is reexpressed to rearrange the IgL loci, implying that VH to DJH joining has been rendered inaccessible to RAG relative to IgL joining and DH to JH joining (see below). www.annualreviews.org • V(D)J Recombination

549

ARI

15 February 2006

1:38

The fact that IgH rearrangements usually occur before IgL rearrangements and that surface expression of μ chains in mature B cells depends on IgL expression (74) suggested that an alternative complex is involved in pairing with newly generated heavy chains. Subsequent studies demonstrated that production of a functional μ chain at the pro- to pre-B cell transition leads to its cell surface expression in association with protein products of the V preB and λ5 genes (75–77), which together form the surrogate light chain (78). Ability to pair with the surrogate light chain is a quality control parameter by which μ chains are assessed. For example, a study of rare B cells with two productive VH DJH rearrangements demonstrated that in every case, only one of the two expressed μ chains was capable of pairing with the surrogate light chain (79). However, allelic exclusion is normal in animals with deletion of the known surrogate light chain elements (78), suggesting that other functionally redundant elements can take the place of surrogate light chain in its absence. Although normal light chains do rearrange during IgH rearrangement in a small fraction of developing B cells (80), it is unlikely that BCRs frequently take the place of pre-BCRs because insertion of a productive Igκ rearrangement into the Igκ locus cannot fully rescue B cells in the absence of surrogate light chain (81). Therefore, although the proteins that can substitute for surrogate light chain elements in their absence remain unknown, existing data strongly support the model that cell surface expression of the pre-BCR produces a feedback signal that shuts off further IgH rearrangements and promotes development to the pre-B cell stage. Although production of a functional, fulllength μ chain results in positive B cell selection, generation of a shortened μ chain can subvert this process. The production of truncated μ chains was ﬁrst demonstrated to occur in certain early B cell lines undergoing IgH rearrangements (82), and these shortened chains were later shown to be membrane-bound and encoded by a DJH joint and the μ constant region; these “Dμ” chains are encoded in one

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

550

Jung et al.

of the three possible DH reading frames (14). Further work showed that the production of Dμ chains is counterselected during B cell development owing to arrest of VH to DJH joining (55). In addition, generation of mice transgenic for a Dμ chain has shown that Dμ cannot allow developing B cells to transition from the pro-B to the pre-B cell stage (83). However, the reasons Dμ protein production should be evolutionarily conserved remain unclear, as do the qualitative differences between μ chain and Dμ chain signaling. Additional details regarding downstream signaling effectors of feedback regulation remain to be characterized, but some of the proximal events have been identiﬁed. The pre-BCR pairs with the signal-transducing transmembrane proteins Igα and Igβ (84, 85) to generate signals that ultimately shutdown further IgH rearrangements and drive progression to the pre-B cell stage, where IgL is rearranged (86). Association of the pre-BCR with Igα and Igβ is required for pre-BCR signaling because mutations blocking μ chain association with Igα and Igβ, as well as mutations in the cytoplasmic domains of either of these molecules, disrupt allelic exclusion and pre-B cell development (87). Stimulation through Igα and Igβ leads to activation of various members of the Src family kinases (86). Thus, concomitant deletion of Syk and ZAP-70 disrupts allelic exclusion at the IgH locus (88), and deletion of the analogous kinase SLP-76 disrupts allelic exclusion at the TCRβ locus (89), suggesting that at least the proximal elements of pathways signaling allelic exclusion are conserved between both B and T cell lineages. Ultimately, activation of these kinases probably triggers a phosphorylation cascade to signal developmental progression to the pre-B cell stage in cells that express a functional pre-BCR. In T cells, there appear to be separate signals that cause feedback regulation of TCRβ rearrangements versus proliferation and onset of TCRα rearrangements (90, 91). It remains to be shown whether promotion of developing B cell proliferation and IgL rearrangements can be similarly

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

dissociated from signals for IgH feedback regulation. Such approaches may help to elucidate the ultimate signals that model accessibility of the IgH and IgL loci (see below). IgL rearrangements occur in the pre-B cell stage, after expression of a pre-BCR. Several lines of experimental evidence suggest that heavy chain expression activates IgL rearrangements, as ﬁrst inferred from the observation that IgH loci rearrange before IgL loci (39). For example, B cell lines that do not make endogenous heavy chain because they have rearranged VH DJH nonproductively on both IgH alleles can undergo Vκ to Jκ rearrangements when transfected with an expression vector encoding a membrane-bound μ chain, but not when transfected with a vector encoding a secreted μ chain (64). Moreover, initiation of IgL rearrangements in these contexts appears to depend on signaling through the pre-BCR (92). Membrane-bound heavy chain expression is not absolutely required for IgL rearrangement, however, as some transformed human pre-B cell lines can express IgL in the absence of VH to DJH joining (93); moreover, animals that express only secreted μ chains (80) or harbor deletion of the JH gene segments (94) both initiate a low level of Vκ to Jκ joining. The overall impact of this “disordered” rearrangement is not yet understood, as the developmental arrest incurred by such mutations may enhance this alternative pathway for IgL rearrangements (29). The weight of evidence therefore supports the idea that a membrane-bound μ chain signals both the cessation of further VH to DJH joining and progression to the pre-B cell stage and the initiation of IgL rearrangement. In this regard, there have been several transformed cell lines (95) as well as in vitro culture models derived from normal early B cells (96) that reproduce this ordered program of IgH and IgL rearrangement. Feedback models of allelic exclusion at the Igκ locus were based on the premise that one allele is rearranged and tested for productivity before the other is rearranged (51). The asynchrony required for this process has been sub-

divided into two phases: initiation and maintenance (50). With respect to allelic exclusion of the IgH locus, feedback regulation can explain maintenance of a frozen DJH upon a productive VH DJH rearrangement on the other allele, but it cannot explain how cells initiate VH to DJH rearrangements on only one allele ﬁrst (51, 97) over a period in which the feedback signal can take place. Many potential general mechanisms have been proposed as to how this might occur, including differential nuclear localization and locus contraction (29, 98, 99) and allelic marking associated with early replication and demethylation (97, 100), although the potential mechanistic roles of these phenomena remain to be directly tested. Regardless, these ﬁndings still would not explain how the second allele becomes activated for VH to DJH rearrangement in those cells that have made a nonproductive VH DJH rearrangement on the ﬁrst allele. Two mechanisms in this regard—permissive or instructive—are theoretically possible, or a combination of these two mechanisms may be at work. First, a permissive model would postulate that both IgH alleles in a pro-B cell operate on a recombination clock, where one allele has a head start and has initiated rearrangement ﬁrst. In this scenario, the second allele eventually undergoes VH to DJH rearrangement, unless a signal from the pre-BCR prevents it from doing so (54, 97). Conversely, in an instructive model, a nonproductive rearrangement somehow informs the cell that it is necessary to initiate VH to DJH rearrangement on the second allele. VH to DJH rearrangement on the second allele would, thus, only occur upon production of a nonproductive VH DJH joint on the ﬁrst allele. Investigators have argued that such a mechanism also operates in activating Igλ for V(D)J recombination upon deletion or nonproductive rearrangement of both Igκ alleles (29). Within the context of permissive and instructive models for activating VH to DJH rearrangement on the second allele, we should also consider the implications that both models hold for the maintenance of IgH allelic www.annualreviews.org • V(D)J Recombination

551

ARI

15 February 2006

1:38

exclusion when RAG is reactivated for IgL rearrangement, which takes place after progression to the pre-B cell stage. On one hand, a permissive model predicts that, while initiating VH to DJH asynchronously, both alleles are destined to rearrange VH to DJH in the absence of a pre-BCR signal. This model therefore would postulate that the pre-BCR sends a signal that actively shuts down a DJH allele with respect to VH to DJH rearrangements at the transition between the pro- to pre-B cell stage. On the other hand, an instructive model predicts that a DJH allele will only rearrange VH to DJH upon a signal that the ﬁrst allele has generated a nonproductive VH DJH joint. If the ﬁrst allele generates a productive VH DJH joint, therefore, such a signal would never be sent, and the DJH allele would never be available for VH to DJH rearrangement. In either scenario, both permissive and instructive models for activating VH to DJH rearrangement on the second allele imply that DJH rearranged alleles are unavailable for VH to DJH joining events at the pre-B cell stage. By contrast, the second population of B cells that transit to the pre-B cell stage have terminally rearranged both IgH alleles VH DJH /VH DJH (in productive/nonproductive conﬁguration) and would likely not require such mechanisms to block further rearrangements (Figure 3).

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

Accessibility Control of IgH Locus Rearrangements As outlined above, V(D)J recombination at the IgH locus is strictly controlled with respect to cell-type speciﬁcity (complete rearrangements in B cells but not in T cells), intralocus order (DH to JH on both alleles before VH to DJH ), interlocus order (no continued VH to DJH rearrangement during light chain rearrangement), and feedback regulation by productive VH DJH joints in the context of allelic exclusion. Although the mechanisms of IgH locus ordered rearrangement and feedback regulation remain to be fully elucidated, both processes are likely based on differential accessibility of VH , DH , and JH gene seg552

Jung et al.

ments to the RAG recombinase (16, 101). The accessibility model for control of V(D)J recombination in all the above contexts initially was hypothesized based on the discovery that murine VH gene segments are transcribed and produce germline (i.e., noncoding and unrearranged) transcripts in cell lines and normal early B lineage cells that are actively undergoing VH to DJH joining (15). However, such transcripts are not found in mature cell lines or adult spleen or in newborn or adult thymus, all of which are lymphoid tissues where VH to DJH joining does not occur. These germline transcripts initiate from VH promoters, are spliced, and are observed in the cytoplasm, where they could theoretically be translated into VH chains (15), although the endogenous production of such chains has never been reported. Proof for the accessibility model for regulation of V(D)J recombination, along with evidence for its association with higher-order controls that act to selectively open or close antigen receptor loci, came from the demonstration that transfected TCRβ gene segments and endogenous IgH gene segments rearranged in a pro-B cell line, whereas endogenous TCRβ gene segments did not (102). Furthermore, transfected TCRβ gene segments and endogenous IgH gene segments in this cell line were sensitive to DNase I, whereas endogenous TCRβ gene segments were not, which directly correlated recombinational potential with a physical correlate of open chromatin with respect to the various gene segments (102). The ability of antigen receptor gene segments to undergo V(D)J recombination has therefore been broadly deﬁned as physical accessibility of these segments to the recombinase machinery. In a further demonstration of this general principle, RAG-mediated cleavage events occur in a cell- and stage-speciﬁc manner both in ex vivo lymphocytes (73) and in nuclei isolated from lymphocytes in vitro (103). In addition, a large number of studies have demonstrated that known transcriptional control elements, most extensively studied for

ANRV270-IY24-18

ARI

15 February 2006

1:38

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

transcriptional enhancers, function to target V, D, and J gene segments for recombinational accessibility, both in transgenic constructs and in endogenous loci (104). Although a considerable amount of recent effort has been invested in deﬁning the nature of accessibility and the role of transcriptional regulatory elements in controlling V(D)J recombination of different Ig and TCR loci (104), we focus primarily on progress related to the IgH locus.

CONTROL OF DH TO JH REARRANGEMENT Various cis-acting elements associated with transcription have been studied for their potential roles in activating DH to JH rearrangement. The intronic μ enhancer (iEμ) is located within the intron between the last JH gene segment and the Cμ exons about 0.8 kb downstream of JH 4 (105, 106). Ectopic expression of E2A and EBF, factors that bind to iEμ, activates DH to JH rearrangement in nonlymphoid cells (107). In addition, it has been shown that iEμ can activate D to J rearrangement in a transgenic recombination substrate (108). Prior to DH to JH rearrangement, μ0 sterile transcripts are initiated upstream of DQ52, the most JH -proximal DH gene segment (109, 110). Transcription also initiates upstream of other germline DH gene segments and generates higher levels of steadystate transcript upon DH to JH joining, a property that is thought to depend on proximity to iEμ (14, 109, 110). At the endogenous locus, replacement of iEμ with a highly transcribed pgk-Neor element blocks all detectable V(D)J recombination of the JH gene segments (111, 112). However, clean deletion of iEμ results in a partial defect in DH to JH rearrangement, in which roughly 30% of peripheral B cells harbor a germline IgH locus (112–114). Levels of μ0 transcription reﬂect these rearrangement defects, as replacement of iEμ with a pgk-Neor cassette removes all detectable levels of μ0 transcripts, whereas a clean deletion of iEμ retains roughly 20% of wild-type μ0

transcripts (112). In addition to the described defect in DH to JH joining, deletion of iEμ results in a stronger inhibition of VH to DJH joining (112–114). This partial inhibition of DH to JH joining on alleles harboring the iEμ deletion, along with complete inhibition of DH to JH rearrangements on alleles harboring replacement of iEμ with a pgk-Neor cassette, together suggest that additional elements are required to activate DH to JH rearrangement in vivo. As a potential example of such an element, a promoter/enhancer element has been identiﬁed upstream of DQ52 (115), although deletion of this sequence together with the DQ52 gene segment does not completely abolish μ0 transcription and also does not appear to dramatically inhibit overall levels of DH to JH joining (116). In this regard, the pgk-Neor cassette is known to inhibit the activities of the long-range 3 IgH regulatory region (RR) (117) that lies 200 Kb downstream of the JH segments, just beyond the last set of constant region exons, and contains enhancers (hs3b and hs4) that are important for activating transcription from I promoters that ﬂank switch regions upstream of the various constant regions. The 3 IgH RR functions to enhance germline constant region transcription, to regulate germline constant region genes, and to inﬂuence expression of rearranged VH DJH exons assembled upstream at the JH region (118). The pgk-Neor cassette appears to affect the 3 IgH RR through promoter competition (117). Thus, the 3 IgH RR may well be another candidate for an element that works in addition to the iEμ element to inﬂuence V(D)J recombination at the JH locus (112). Although deletion of the 3 IgH RR in isolation does not result in an obvious defect in DH to JH (or VH to DJH ) rearrangement (118), iEμ may contribute overlapping activities. Therefore, inactivating both the 3 IgH RR and the DQ52 promoter/enhancer element, or some other elements, including unknown elements within the VH locus or upstream, in conjunction with deleting iEμ, is relevant to see if these various elements www.annualreviews.org • V(D)J Recombination

iEμ: intronic μ enhancer

553

ANRV270-IY24-18

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CSR: class switch recombination

554

1:38

cooperate with iEμ to enhance DH to JH rearrangements. In addition to the study of endogenous transcriptional control elements in the initiation of DH to JH joining, a number of chromatin modiﬁcations, many ﬁrst identiﬁed in transcriptional studies as markers of “open” chromatin, correlate with V(D)J recombinational accessibility at this locus. Thus, the genomic region comprising the DH and JH gene segments is associated with hyperacetylated chromatin in RAG-deﬁcient pro-B cells, which should be poised to rearrange but are unable to initiate V(D)J rearrangements (119–121). Additional studies of transformed RAG-deﬁcient pro-B cell lines have further shown that the DH and JH gene segments are also associated with BRG1, the catalytic subunit of the chromatin remodeling complex SWI/SNF, suggesting that chromatin remodeling may be important for accessibility of genomic DNA to the RAG recombinase machinery (121). Perhaps the most direct genetic evidence that such chromatin modiﬁcations control V(D)J recombinational accessibility comes from studies of TCRβ locus gene segments, which also undergo ordered rearrangement and feedback regulation in vivo (65). Thus, a TCRβ rearrangement substrate in which a H3-K9 methyltransferase is tethered in close proximity to a Dβ promoter has increased local methylation of H3K9 and decreased transcription through and V(D)J recombination of the Dβ gene segment (122). Further work using such creative approaches at endogenous loci may shed light on whether transcription and its associated chromatin modiﬁcations are a cause or effect of an antigen locus primed for rearrangement. In this regard, it is instructive to note basic differences between the requirements for transcription during V(D)J recombination and during class switch recombination (CSR). CSR changes the receptor effector function of a VH DJH exon through site-speciﬁc recombination between large, repetitive switch regions that lie upstream of the various constant region exons. The AID (activation-induced Jung et al.

deaminase) enzyme is required for this process (123), and deaminates cytidine residues in single-stranded DNA generated within S regions undergoing transcription (124). The deaminated cytidines eventually yield strand lesions that may ultimately lead to DSBs by mechanisms that are still under intense investigation (125). By contrast, it appears that, during V(D)J recombination, transcription may only be required to render gene segments accessible to RAG, rather than playing a direct mechanistic role in the recombination reaction (101). In support of this idea, recent experiments have shown that, in transgenic V(D)J recombination substrates, the presence of a transcriptional promoter upstream of a recombining gene segment can activate its rearrangement independent of its orientation (126). Thus, although the precise role of transcription in regulating V(D)J recombination remains to be fully understood, it appears that, in contrast to CSR, transcription through recombining V, D, and J gene segments may be dispensable for this process, and the role of transcriptional control elements may instead be to provide more general accessibility via modiﬁcations in chromatin structure.

CONTROL OF VH TO DJH REARRANGEMENT Activation of VH Gene Segment Rearrangement Because DH to JH joining occurs on both IgH alleles in an allelically included manner, the VH to DJH joining step is the controlled step for feedback-regulated, ordered rearrangement of the IgH locus. Several targeted mutations that impair VH to DJH joining have been generated. First, VH to DH rearrangement does not occur on an allele harboring a deletion of the JH gene segments, suggesting that DH to JH rearrangement may be a prerequisite for VH gene segment rearrangement (94). However, this mutation retains the pgk-Neor cassette, which may contribute confounding effects, possibly through

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

promoter competition for necessary activating factors (117). In addition, deletion of iEμ severely disrupts VH to DJH joining and generates a distribution skewed away from cells that are rearranged VH DJH /VH DJH on both IgH alleles and toward VH DJH /DJH rearranged cells (112–114). Finally, various endogenous mutations of RAG2 speciﬁcally decrease VH to DJH joining. In this regard, the core RAG mutations have been identiﬁed as the minimal regions necessary to support RAG-mediated cleavage in vitro, although these truncated proteins clearly lack functions that may be important during V(D)J recombination in vivo (32). For example, on the one hand, targeted replacement of the RAG2 gene with the core RAG2 mutation in vivo resulted in a dramatic defect in VH to DJH rearrangement, along with a lesser but significant disruption of DH to JH rearrangement (127, 128). On the other hand, replacement of the endogenous RAG1 gene with the core RAG1 mutation caused an overall decrease in efﬁciency of rearrangement, without any obvious differential effect on DH to JH versus VH to DJH joining (129). With respect to requirements for VH to DJH joining, it is worth noting that clones harboring unrearranged JH alleles were identiﬁed at increased frequency among mature B cell hybridomas generated from animals harboring the two core RAG mutations (128, 129), as well as from those with the iEμ deletion (112, 114). This ﬁnding demonstrates that DH to JH joining is not required on both IgH alleles to activate VH to DJH joining. However, whether or not DH to JH joining is required in cis to activate VH to DJH joining on the same allele remains an open question. The different phenotypes of the two RAG core mutations may perhaps be explained by different functions encoded in the noncore regions of the RAGs; for example, the noncore region of RAG2 contains a PHD (plant homeodomain) ﬁnger, a domain found in many chromatin-associated proteins, raising the tantalizing possibility that RAG2 may be involved in directly modulating chromatin

structure during V(D)J recombination (130). Consistent with this potential general role for RAG2 is the ﬁnding that the noncore portion of RAG2 binds various histones through residues that also appear to be critical for VH to DJH rearrangement, although these residues lie outside the PHD ﬁnger (131). However, further work is needed to show that RAG2 physically targets chromatin modiﬁcation activities in vivo, either by itself or in conjunction with additional factors.

Molecular Correlates of VH Gene Segment Rearrangement As the VH gene segment cluster is spread across three megabases of DNA, elucidation of how these segments are brought into DJH joints should illuminate mechanisms of largescale chromosomal interactions in other gene regulatory processes. Moreover, understanding how one allele is chosen to rearrange VH to DJH ﬁrst is critical for a complete understanding of the processes underlying allelic exclusion, although no mechanism has yet been shown to be directly involved in controlling this process. Asynchronous, early replication of one IgH allele is closely correlated with rearrangement of that allele about 90% of the time (100), although the precise link between the two processes remains unclear. In addition, although allele-speciﬁc demethylation serves as a mark for preferential rearrangement at the Igκ locus (97), it has not to date been reported for the IgH locus and, moreover, has not been clearly demonstrated to be mechanistically related to allele-speciﬁc choice. By analogy with the Igκ locus (97), one might expect that allele-speciﬁc demethylation or some other allele-speciﬁc mark, if one exists, will localize to the region between the VH and DH gene segments, and further work should address this possibility. In this regard, a recent study of the Igκ locus followed various stages of Igκ activation from pro-B cells through pre-B cells to outline steps of nuclear relocalization, histone modiﬁcation, monoallelic heterochromatinization, and monoallelic www.annualreviews.org • V(D)J Recombination

555

ARI

15 February 2006

1:38

demethylation (132). Similar comprehensive studies of the IgH locus may help elucidate how one allele is ﬁrst chosen to initiate VH to DJH joining. The observations of chromosomal relocalization of the IgH locus and IgH locus contraction and looping in pro-B cells have been recent visual correlates of VH to DJH joining (62, 99, 133–135). Transport to a localized recombination center in the nucleus, followed by export out of such a center upon the cessation of rearrangement, is an attractive means by which cell stage-speciﬁc control of V(D)J recombination might be established (98). In this regard, recent studies have shown that in pro-B cells, which are actively undergoing IgH locus rearrangement, both IgH alleles are centrally located (133). In non-B cells, however, both IgH loci are located at the nuclear periphery (134). In addition, the phenomena of IgH locus contraction and looping, as visualized by 3D-FISH, are thought to represent the linking of distal VH gene segments with DJH complexes (62, 99, 135). Importantly, IgH locus looping occurs in the absence of RAG2 (62), suggesting that the relocalization of VH gene segments in threedimensional space occurs prior to the assembly of a RAG-RS complex. Thus, further understanding the control of IgH locus localization, contraction, and looping may provide important insights into the determination of stage-speciﬁc accessibility in the context of ordered rearrangement and allelic exclusion. In another correlate of the accessibility mechanisms that drive VH gene segment recombination, VH segments are transcribed prior to their rearrangement (15, 136). More recently, researchers have also shown that antisense transcripts are generated in the vicinity of unrearranged VH gene segments (137). Like sense germline VH transcripts, antisense transcripts are found in pro-B cell populations that are actively undergoing VH to DJH joining, and are switched off in cells that have successfully completed VH to DJH joining (137). Moreover, antisense VH transcripts appear to be generated in wide regions of the

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

556

Jung et al.

VH gene segment cluster, including in intergenic regions. In the context of IgH rearrangements, antisense transcription may be a property speciﬁc to the VH cluster, as it was not observed near JH gene segments (137). Although the functional signiﬁcance of these transcripts is not yet clear, several possibilities were proposed, including remodeling of IgH locus chromatin and generation of doublestranded RNA (by the coordinated generation of sense and antisense transcripts) to target the RNA interference machinery and thereby recruit other proteins, such as histone methyltransferases, that may be important for VH gene segment rearrangement (137). Strikingly, about 80% of the clones expressed antisense germline transcripts monoallelically, as would be expected for a mechanism involved in choosing one allele to rearrange ﬁrst (137), although we do not know whether the allele expressing antisense transcripts actually rearranged ﬁrst. Further work to deﬁne antisense promoters and/or enhancers may elucidate whether antisense VH transcripts play a role in rendering VH gene segment chromatin accessible to RAG or, alternatively, are a nonspeciﬁc by-product of opening the locus by some other means. Regardless, it is striking that a sense/antisense transcription proﬁle appears to operate at many known mammalian genes, perhaps to mediate gene silencing through degradation of sense transcripts by targeting RNA interference proteins (138), and understanding the control of VH gene segment rearrangement may shed light on the general functional importance of this phenomenon.

Preferential Rearrangement of Proximal VH Gene Segments A striking feature of murine VH gene segment recombination is that VH gene segments proximal to the DH gene segments are preferentially rearranged (139, 140). Although the functional signiﬁcance of this phenomenon remains unresolved, it has been suggested that rearrangements using proximal VH gene segments generate self-reactive antibodies and

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

that such antibodies may play some role in shaping the adult antibody repertoire (141). In the human IgH locus, by contrast, although the two most proximal VH gene segments are highly rearranged (142, 143), other frequently rearranged VH gene segments are scattered throughout the human VH cluster (144), as is also observed for the Igκ and TCRβ loci. The mechanistic basis for this difference between murine and human VH gene segment utilization remains unclear and may depend on various factors, including RS strength and differential accessibility of individual VH gene segments versus VH gene segment families. Preferential rearrangement of proximal VH gene segments at the murine IgH locus was ﬁrst demonstrated in fetal liver hybridomas and in A-MuLV-transformed pre-B cell lines (139, 140). Although this process was (and by some still is) considered representative of developmental programming of VH gene rearrangement with proximal VH segments rearranging earlier in ontogeny than distal VH segments, this possibility was ruled out by studies showing that the newly generated repertoire was similarly biased in both fetal liver and adult bone marrow B lineage cells (145). Thus, 3 VH genes are preferentially rearranged at all stages of development. However, before reaching the periphery, the VH repertoire appears to be normalized by selection mechanisms such that the population of rearranged VH genes expressed by peripheral B cells roughly correlates with the number of members in each VH gene segment family (146), although the degree to which this is achieved may depend on other factors, such as the number of nonfunctional gene segments within a particular family (147, 148). An example of one process that may be involved in normalizing the VH repertoire comes from studies that show VH 81X, the most proximal and most frequently rearranged VH gene segment in the murine 129 strain, is often counterselected, most likely because of its relative inability to pair with the surrogate light chain (149, 150). The VH repertoire might therefore be normalized

through a VH replacement mechanism (139). In this regard, rearranged VH /DJH joints can undergo VH replacement through recombination between a germline VH gene segment and a cryptic heptamer within the rearranged VH coding sequence (58, 59). However, how often VH replacements occur in vivo remains unclear, and they likely occur infrequently (60).

Possible Mechanisms Involved in Preferential VH Gene Segment Rearrangement The mechanistic basis for the greater recombinational potential of proximal VH gene segments has remained an outstanding question in IgH locus V(D)J recombination. Several factors, either singly or in combination, may be responsible for the preferential rearrangement of proximal VH gene segments. Such factors may include more potent RSs ﬂanking proximal versus distal VH gene segments and/or increased accessibility of proximal VH gene segments to the RAG proteins. Regarding RS strength, recombination substrate transfection assays have demonstrated that a consensus VH 7183 RS, derived from a proximal VH family, can mediate higher levels of rearrangement than a consensus VH J558 RS, derived from a distal VH family (151). However, other studies have shown that among proximal VH gene segments, in vivo rearrangement frequencies are correlated with a more proximal position and not with RS potency (152). A deﬁnitive answer in this regard thus likely awaits genetic swaps between proximal and distal VH gene segments and/or RSs. However, the likelihood that RS differences can play a role is reinforced by the results of such an experiment done in the context of the TCRβ locus, which showed a major increase in the rearrangement frequency and change in preference for a Vβ gene segment by replacing its 23 RS with a 3 Dβ1 23 RS (27). Various mutations in factors and signaling pathways that may act in trans with respect to the IgH locus appear to demonstrate a www.annualreviews.org • V(D)J Recombination

557

ARI

15 February 2006

1:38

speciﬁc defect in distal VH rearrangement. Several of these involve elements of the IL-7 signaling cascade. Deletion of the IL-7 receptor causes a developmental block very early in B cell development in vivo (153). In addition, pro-B cells sorted from IL-7R−/− mice are deﬁcient in distal VH J558 germline transcription and rearrangements, as detected by PCR (154). Consistent with a role for IL-7 signaling in the activation of distal VH J558 gene segments, VH J558 gene segments have increased levels of histone acetylation and nuclease sensitivity upon IL-7 treatment (119, 155). In further support of a role for IL-7 signaling in recombinational activation of distal VH gene segments, STAT5, an IL-7-responsive transcription factor, appears to be important for distal VH J558 germline transcription and rearrangement (156, 157). In addition to elements of the IL-7 signaling cascade, B-cellspeciﬁc deletion of Ezh2, a histone methyltransferase, results in a decrease in VH J558 rearrangement (again as measured by PCR on pro-B cell DNA), but in the presence of normal germline transcripts and histone acetylation; instead, methylation of H3-K27 appears to be impaired in Ezh2−/− pro-B cells (158). Pax5 is a transcription factor that has been extensively analyzed for its role in controlling the rearrangement of VH gene segments and for its more general role in B cell commitment and differentiation (159). In this regard, Pax5-deﬁcient pro-B cells appear to have a speciﬁc defect in distal VH gene rearrangement, as detected by PCR on pro-B cell DNA (160). This defect is not correlated with differences in certain measures of accessibility, as germline transcription and histone acetylation are normal at both proximal and distal VH gene segments from Pax5-deﬁcient pro-B cells (160); conversely, demethylation of H3K9 appears to be globally disrupted across the VH gene segment cluster in pro-B cells lacking Pax5 (161). However, contraction of the IgH locus, which has been correlated with rearrangement of distal VH gene segments, is only observed in pro-B cells that express Pax5 (62).

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

558

Jung et al.

Recent studies have reported the striking ﬁnding that ectopic expression of Pax5 in developing T cells results in (a) the relocalization of IgH alleles to the center of the nucleus (62), (b) induction of proximal and distal VH germline transcription (62), and (c) rearrangement of proximal versus distal VH gene segments (62, 63). Thus, with respect to V(D)J recombination of the IgH locus, early T cells expressing Pax5 appear to be phenotypically nearly identical to early B cells lacking Pax5 by all measures so far examined. One interpretation of the available data on Pax5, therefore, is that Pax5 is sufﬁcient in early T cells for the rearrangement of proximal VH gene segments and necessary in early B cells for the rearrangement of distal VH gene segments, which may suggest the existence of another factor in B cells that is required for IgH locus contraction and distal VH gene segment rearrangement (62). Taken together, experiments on roles of IL-7R signaling and Pax5 in driving distal versus proximal VH gene segment rearrangement are compelling studies that warrant further investigation. However, there may be additional explanations than simply activating distal versus proximal VH segments that could also, at least in part, account for the data. For example, mutations related to IL-7 signaling, Pax5 function, or other genetic modiﬁcations, including transgenes, which appear to differentially affect distal versus proximal VH genes, may also reﬂect changes in B cell development that result in an unnormalized repertoire. Notably, proximal VH gene segments are preferentially rearranged in all developing B cells (145); thus, preferential rearrangement of the 3 VH gene segments in different transgenic and knock-out models could, theoretically, reﬂect the normally biased patterns found in the newly generated repertoire. Further work, perhaps aimed at directly measuring recombinational accessibility of proximal versus distal VH gene segments in different contexts, may help to distinguish among these possibilities.

ANRV270-IY24-18

ARI

15 February 2006

1:38

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Transgenic Studies of VH Feedback Regulation and Proximal versus Distal VH Rearrangement In accordance with a feedback-regulated model for IgH rearrangement control, expression of transgenic constructs that contain membrane-bound prerearranged VH DJH μ chains results in a substantial decrease of endogenous VH to DJH rearrangements (69, 71). Such studies may be confounded, however, by evidence that suggests μ transgenes drive artiﬁcially accelerated B cell development. Because the VH to DJH joining step is the allelically excluded step, one would predict that, if μ transgenes signal normal allelic exclusion, virtually all B cells from μ-transgenic mice would contain DJH /DJH rearrangements on both alleles. In fact, a signiﬁcant fraction of hybridomas generated from μ-transgenic B cells retain at least one of their endogenous JH clusters in unrearranged conﬁguration, which is not observed in normal B cells (69, 162). Further, in μ-transgenic B cells, DH to JH rearrangements do not begin until the preB cell stage, during normal light chain rearrangement (162). Studies on other membrane μ transgenes, however, have reported that levels of DH to JH rearrangements are normal, based on PCR assays (163, 164). Such differences in interpretation may be due to some aspect of transgene expression levels or the methods used to detect rearrangements. Although the majority of VH to DJH rearrangements are abrogated in μ-transgenic mice, some VH to DJH rearrangements do occur (163, 165, 166). As in normal mice, the VH to DJH rearrangements in μ-transgenic mice are skewed toward proximal VH gene segments (163, 166), a ﬁnding that has been interpreted to suggest that proximally located VH gene segments may be less regulated with respect to allelic exclusion (99, 163). Given that proximal VH gene segments are preferentially rearranged in both fetal liver and adult bone marrow (145), however, a possible alternative explanation of these ﬁndings might be that, in the context of accelerated B cell development,

proximal VH gene segment rearrangements are observed preferentially because they are normally more likely to occur. Thus, shortening the time available for IgH rearrangements in general to occur might nonspeciﬁcally select for proximal VH to DJH rearrangements by decreasing the overall level of VH to DJH rearrangements. Taken together, the data suggest that results generated using μ transgenes should be interpreted with some caution, as should similar experiments generated with TCRβ transgenes (45, 90).

Receptor Editing of VH DJH Joints Receptor editing refers to the process by which secondary V(D)J rearrangements can replace a nonfunctional or self-reactive variable region exon with another one (57). In B cells, this general phenomenon can occur frequently at the Igκ locus, as Vκ gene segments are allowed to rearrange, by the 12/23 rule, to available downstream Jκ gene segments. However, for the IgH locus, VH replacements appear to take place much less frequently because a VH to DJH rearrangement deletes all remaining DH segments, which are required for VH gene segment recombination that obeys the 12/23 rule. However, VH replacement events did occur in murine cell lines through a VH to VH DJH recombination reaction that was hypothesized to employ a conserved heptamer sequence in the downstream body of the rearranged VH segments (58, 59). More recently, VH replacement was also shown to occur in human B cell lines (167). Furthermore, a database generated from normal human mature B cells was screened for a VH replacement footprint, revealing that 5%–10% of VH DJH joints may have resulted from VH replacement events (167). However, mature B cells represent a selected population, and there still remains no evidence that murine VH replacement has any signiﬁcant role in gross normalization of primary VH repertoires (168). In models in which self-reactive VH DJH joints have been targeted to replace the JH gene segments, initiation of www.annualreviews.org • V(D)J Recombination

559

ANRV270-IY24-18

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

cRS: cryptic recombination signal sequence

560

1:38

VH editing events appears to take place during the immature B cell stage, after the completion of IgL rearrangements (169). Recombination substrate studies conﬁrmed that VH replacements involve the usage of a cryptic RS (cRS) that contains only an isolated heptamer and is located near the 3 end of most VH gene segments, which mediates V(D)J recombination with the 23 RS of an upstream VH segment (170). The cRS element is highly conserved from mammals to cartilaginous ﬁsh, suggesting that it has been selected during the course of evolution (168). Moreover, the cRS can mediate RAG-induced DSBs in cell lines and in cleavage reactions in vitro, although at a lower frequency relative to RS-RS cleavage (167). Thus, the fact that V(D)J recombination efﬁciently occurs only in the context of two complete RSs suggests that immature B cells that undergo editing may have some way of relaxing the speciﬁcity of RAG, perhaps through the expression of as yet undeﬁned factors, to accommodate rearrangements between a complete RS and a cRS. Notably, the Igκ light chain locus sometimes may be inactivated during editing by joining a Vκ to an isolated heptamer in the Jκ-Cκ intron (57), again raising the possibility of relaxed RAG speciﬁcity during editing. The regulation of VH replacements may have implications for certain aspects of the allelic exclusion process. In this regard, it seems likely that in VH DJH /DJH rearranged B cells expressing a self-reactive μ chain, VH replacement may be speciﬁcally targeted to occur on the VH DJH allele, whereas VH genes on the DJH rearranged allele are kept in an inaccessible state to maintain allelic exclusion. Thus, receptor editing occurs preferentially on the rearranged allele as opposed to the germline allele in the Igκ locus (73, 171). Furthermore, in VH DJH /VH DJH rearranged B cells that express a self-reactive μ chain from one IgH allele and are nonproductively rearranged on the other IgH allele, VH replacement is likely similarly targeted speciﬁcally to the self-reactive VH DJH joint to maintain allelic exclusion. By contrast, such considerJung et al.

ations do not seem relevant for B cells that harbor two nonfunctional VH DJH rearrangements and attempt to salvage a productive μ chain through VH replacement. Understanding the mechanisms by which various alleles choose to pursue VH replacement events may also be valuable in understanding allelic exclusion and feedback regulation.

PERSPECTIVE Research into the mechanisms underlying V(D)J recombination at the IgH locus appears poised to generate many exciting insights over the next several years, as emerging technologies should facilitate the study of IgH rearrangements. Longstanding issues include how one allele is chosen to ﬁrst rearrange VH to DJH , how the second allele is activated to rearrange VH to DJH if the ﬁrst is nonproductive, and what the signals are that redirect the RAG proteins to the κ locus at the pre-B cell stage. To address these questions, new approaches are likely required. Improvements in genomic engineering technology should allow for simpliﬁcation of the VH gene segment cluster (i.e., deleting it down to a few VH gene segments), which would allow for detailed analyses of the requirements for accessibility at individual VH gene segments, provided such large-scale deletions are normally regulated. Such an approach would also make it easier to characterize chromatin modiﬁcations and factors such as antisense transcripts and their potential function. Another long-standing issue is the function of ordered D to JH and VH to DJH rearrangement. To address the role of ordered rearrangement, it is important to disrupt the normal timing and/or order of events to determine effects on feedback regulation. Investigators must also have more direct measurements of accessibility throughout the locus. One such approach could be to use reporters to measure transcription [as has been done for the κ locus, (54)], an approach that could be particularly useful for analyzing antisense transcription. Another approach might be the insertion of V(D)J

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

recombination reporter cassettes into speciﬁc locations within endogenous antigen receptor loci to assess whether these locations are available to RAG in various stages and contexts. Further elucidation of the implications of IgH locus contraction and looping may provide additional mechanistic insights. Recent reports have described the role of nuclear actin in activating transcription by helping to form a preinitiation complex (172), and additional reports have shown that actin is important for chromosome association during meiosis (173), suggesting that the potential role of actin and other such proteins in relocalization of IgH alleles should be examined. In addition, IgH locus contraction may

involve the spatial reorganization of chromatin to facilitate physical interaction between the different VH genes and the DJH region. Thus, some of the known regulatory elements, as well as some that perhaps await discovery, might function as a locus control region (LCR) to coordinate such looping events in bringing together DJH rearrangements and germline VH gene segments for rearrangement. Such a putative element might function analogously to the way the β-globin LCR (174) and the TH 2 cytokine LCR (175) regulate stage-speciﬁc gene expression. Many or all of the future advances in our understanding of V(D)J recombination at the IgH locus likely hold broad implications for the control of gene expression in different systems.

LCR: locus control region

ACKNOWLEDGMENTS We apologize to those whose work we could not directly cite owing to space limitations, and we urge interested readers to access the more specialized reviews that we have referenced throughout the text. D. Jung received support from the Medical Scientist Training Program, Harvard Medical School. C. Giallourakis is supported by the Crohn’s and Colitis Foundation of America. R. Mostoslavsky is a senior postdoctoral fellow of the Leukemia and Lymphoma Society. F.W. Alt is an investigator of the Howard Hughes Medical Institute and is supported by NIH grant AI-20047.

LITERATURE CITED 1. Davis MM, Bjorkman PJ. 1988. T-cell antigen receptor genes and T-cell recognition. Nature 334:395–402 2. Jung D, Alt FW. 2004. Unraveling V(D)J recombination; insights into gene regulation. Cell 116:299–311 3. Bassing CH, Swat W, Alt FW. 2002. The mechanism and regulation of chromosomal V(D)J recombination. Cell 109(Suppl.):S45–55 4. McBlane JF, van Gent DC, Ramsden DA, Romeo C, Cuomo CA, et al. 1995. Cleavage at a V(D)J recombination signal requires only RAG1 and RAG2 proteins and occurs in two steps. Cell 83:387–95 5. van Gent DC, McBlane JF, Ramsden DA, Sadofsky MJ, Hesse JE, Gellert M. 1995. Initiation of V(D)J recombination in a cell-free system. Cell 81:925–34 6. Shinkai Y, Rathbun G, Lam KP, Oltz EM, Stewart V, et al. 1992. RAG-2-deﬁcient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68:855–67 7. Mombaerts P, Iacomini J, Johnson RS, Herrup K, Tonegawa S, Papaioannou VE. 1992. RAG-1-deﬁcient mice have no mature B and T lymphocytes. Cell 68:869–77 8. Fugmann SD, Lee AI, Shockett PE, Villey IJ, Schatz DG. 2000. The RAG proteins and www.annualreviews.org • V(D)J Recombination

561

ANRV270-IY24-18

ARI

15 February 2006

9. 10.

11.

12. Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

13. 14. 15. 16. 17. 18. 19.

20.

21. 22. 23. 24.

25.

26.

27.

28.

562

1:38

V(D)J recombination: complexes, ends, and transposition. Annu. Rev. Immunol. 18:495– 527 Eastman QM, Leu TMJ, Schatz DG. 1996. Initiation of V(D)J recombination in vitro obeying the 12/23 rule. Nature 380:85–88 Sawchuk DJ, Weis-Garcia F, Malik S, Besmer E, Bustin M, et al. 1997. V(D)J recombination: modulation of RAG1 and RAG2 cleavage activity on 12/23 substrates by whole cell extract and DNA-bending proteins. J. Exp. Med. 185:2025–32 Early P, Huang H, Davis M, Calame K, Hood L. 1980. An immunoglobulin heavy chain variable region gene is generated from three segments of DNA: VH, D and JH. Cell 19:981–92 Rooney S, Chaudhuri J, Alt FW. 2004. The role of the non-homologous end-joining pathway in lymphocyte development. Immunol. Rev. 200:115–31 Riblet R. 2004. Immunoglobulin heavy chain genes of mouse. In Molecular Biology of B Cells, ed. T Honjo, FW Alt, MS Neuberger, pp. 19–26. London: Elsevier Acad. Reth MG, Alt FW. 1984. Novel immunoglobulin heavy chains are produced from DJH gene segment rearrangements in lymphoid cells. Nature 312:418–23 Yancopoulos GD, Alt FW. 1985. Developmentally controlled and tissue-speciﬁc expression of unrearranged VH gene segments. Cell 40:271–81 Yancopoulos GD, Alt FW. 1986. Regulation of the assembly and expression of variableregion genes. Annu. Rev. Immunol. 4:339–68 Tonegawa S. 1983. Somatic generation of antibody diversity. Nature 302:575–81 Benedict CL, Gilﬁllan S, Thai TH, Kearney JF. 2000. Terminal deoxynucleotidyl transferase and repertoire development. Immunol. Rev. 175:150–57 Schilling J, Clevinger B, Davie JM, Hood L. 1980. Amino acid sequence of homogeneous antibodies to dextran and DNA rearrangements in heavy chain V-region gene segments. Nature 283:35–40 Alt FW, Baltimore D. 1982. Joining of immunoglobulin heavy chain gene segments: implications from a chromosome with evidence of three D-JH fusions. Proc. Natl. Acad. Sci. USA 79:4118–22 Meek KD, Hasemann CA, Capra JD. 1989. Novel rearrangements at the immunoglobulin D locus. Inversions and fusions add to IgH somatic diversity. J. Exp. Med. 170:39–57 Sollbach AE, Wu GE. 1995. Inversions produced during V(D)J rearrangement at IgH, the immunoglobulin heavy-chain locus. Mol. Cell. Biol. 15:671–81 Gauss GH, Lieber MR. 1992. The basis for the mechanistic bias for deletional over inversional V(D)J recombination. Genes Dev. 6:1553–61 Pan PY, Lieber MR, Teale JM. 1997. The role of recombination signal sequences in the preferential joining by deletion in DH-JH recombination and in the ordered rearrangement of the IgH locus. Int. Immunol. 9:515–22 Bassing CH, Alt FW, Hughes MM, D’Auteuil M, Wehrly TD, et al. 2000. Recombination signal sequences restrict chromosomal V(D)J recombination beyond the 12/23 rule. Nature 405:583–86 Jung D, Bassing CH, Fugmann SD, Cheng HL, Schatz DG, Alt FW. 2003. Extrachromosomal recombination substrates recapitulate beyond 12/23 restricted VDJ recombination in nonlymphoid cells. Immunity 18:65–74 Wu C, Bassing CH, Jung D, Woodman BB, Foy D, Alt FW. 2003. Dramatically increased rearrangement and peripheral representation of Vβ14 driven by the 3 Dβ1 recombination signal sequence. Immunity 18:75–85 Hardy RR, Hayakawa K. 2001. B cell development pathways. Annu. Rev. Immunol. 19:595– 621

Jung et al.

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

29. Gorman JR, Alt FW. 1998. Regulation of immunoglobulin light chain isotype expression. Adv. Immunol. 69:113–81 30. Desiderio S, Lin WC, Li Z. 1996. The cell cycle and V(D)J recombination. Curr. Top. Microbiol. Immunol. 217:45–59 31. Grawunder U, Winkler TH, Melchers F. 1996. Regulation of recombination activating gene expression during lymphocyte development. Curr. Top. Microbiol. Immunol. 217:31– 43 32. Sadofsky MJ. 2004. Recombination-activating gene proteins: more regulation, please. Immunol. Rev. 200:83–89 33. Alt FW, Bothwell AL, Knapp M, Siden E, Mather E, et al. 1980. Synthesis of secreted and membrane-bound immunoglobulin mu heavy chains is directed by mRNAs that differ at their 3 ends. Cell 20:293–301 34. Early P, Rogers J, Davis M, Calame K, Bond M, et al. 1980. Two mRNAs can be produced from a single immunoglobulin mu gene by alternative RNA processing pathways. Cell 20:313–19 35. Levitt D, Cooper MD. 1980. Mouse pre-B cells synthesize and secrete mu heavy chains but not light chains. Cell 19:617–25 36. Maki R, Kearney J, Paige C, Tonegawa S. 1980. Immunoglobulin gene rearrangement in immature B cells. Science 209:1366–69 37. Perry RP, Kelley DE, Coleclough C, Kearney JF. 1981. Organization and expression of immunoglobulin genes in fetal liver hybridomas. Proc. Natl. Acad. Sci. USA 78:247–51 38. Alt FW. 1983. The ordered sequence of Ig gene rearrangements. J. Mol. Cell. Immunol. 1:39–41 39. Alt F, Rosenberg N, Lewis S, Thomas E, Baltimore D. 1981. Organization and reorganization of immunoglobulin genes in A-MULV-transformed cells: rearrangement of heavy but not light chain genes. Cell 27:381–90 40. Coffman RL, Weissman IL. 1983. Immunoglobulin gene rearrangement during pre-B cell differentiation. J. Mol. Cell. Immunol. 1:31–41 41. Alt FW, Blackwell TK, DePinho RA, Reth MG, Yancopoulos GD. 1986. Regulation of genome rearrangement events during lymphocyte differentiation. Immunol. Rev. 89:5–30 42. Kurosawa Y, von Boehmer H, Haas W, Sakano H, Trauneker A, Tonegawa S. 1981. Identiﬁcation of D segments of immunoglobulin heavy-chain genes and their rearrangement in T lymphocytes. Nature 290:565–70 43. Alt FW, Yancopoulos GD, Blackwell TK, Wood C, Thomas E, et al. 1984. Ordered rearrangement of immunoglobulin heavy chain variable region segments. EMBO J. 3:1209–19 44. Born W, Yague J, Palmer E, Kappler J, Marrack P. 1985. Rearrangement of T-cell receptor β-chain genes during T-cell development. Proc. Natl. Acad. Sci. USA 82:2925–29 45. Uematsu Y, Ryser S, Dembic Z, Borgulya P, Krimpenfort P, et al. 1988. In transgenic mice the introduced functional T cell receptor beta gene prevents expression of endogenous beta genes. Cell 52:831–41 46. Burnet FM. 1957. A modiﬁcation of Jerne’s theory of antibody production using the concept of clonal selection. Austr. J. Sci. 20:67–69 47. Pernis B, Chiappino G, Kelus AS, Gell PGH. 1965. Cellular localization of immunoglobulins with different allotypic speciﬁcities in rabbit lymphoid tissues. J. Exp. Med. 122:853– 75 48. Cebra JJ, Colberg JE, Dray S. 1966. Rabbit lymphoid cells differentiated with respect to α-, γ-, and μ-heavy polypeptide chains and to allotypic markers Aa1 and Aa2. J. Exp. Med. 123:547–58 www.annualreviews.org • V(D)J Recombination

563

ARI

15 February 2006

1:38

49. Barreto V, Cumano A. 2000. Frequency and characterization of phenotypic Ig heavy chain allelically included IgM-expressing B cells in mice. J. Immunol. 164:893–99 50. Mostoslavsky R, Alt FW, Rajewsky K. 2004. The lingering enigma of the allelic exclusion mechanism. Cell 118:539–44 51. Alt FW, Enea V, Bothwell AL, Baltimore D. 1980. Activity of multiple light chain genes in murine myeloma cells producing a single, functional light chain. Cell 21:1–12 52. Coleclough C, Perry RP, Karjalainen K, Weigert M. 1981. Aberrant rearrangements contribute signiﬁcantly to the allelic exclusion of immunoglobulin gene expression. Nature 290:372–78 53. Coleclough C. 1983. Chance, necessity and antibody gene dynamics. Nature 303:23–26 54. Liang HE, Hsu LY, Cado D, Schlissel MS. 2004. Variegated transcriptional activation of the immunoglobulin kappa locus in pre-B cells contributes to the allelic exclusion of light-chain expression. Cell 118:19–29 55. Gu H, Kitamura D, Rajewsky K. 1991. B cell development regulated by gene rearrangement: arrest of maturation by membrane-bound D mu protein and selection of DH element reading frames. Cell 65:47–54 56. Rajewsky K. 1996. Clonal selection and learning in the antibody system. Nature 381:751– 58 57. Nemazee D. 2000. Receptor selection in B and T lymphocytes. Annu. Rev. Immunol. 18:19–51 58. Reth M, Gehrmann P, Petrac E, Wiese P. 1986. A novel VH to VHDJH joining mechanism in heavy-chain-negative (null) pre-B cells results in heavy-chain production. Nature 322:840–42 59. Kleinﬁeld R, Hardy RR, Tarlinton D, Dangl J, Herzenberg LA, Weigert M. 1986. Recombination between an expressed immunoglobulin heavy-chain gene and a germline variable gene segment in a Ly 1+ B-cell lymphoma. Nature 322:843–46 60. Nemazee D, Weigert M. 2000. Revising B cell receptors. J. Exp. Med. 191:1813–17 61. Lam KP, Kuhn R, Rajewsky K. 1997. In vivo ablation of surface immunoglobulin on mature B cells by inducible gene targeting results in rapid cell death. Cell 90:1073–83 62. Fuxa M, Skok J, Souabni A, Salvagiotto G, Roldan E, Busslinger M. 2004. Pax5 induces V-to-DJ rearrangements and locus contraction of the immunoglobulin heavy-chain gene. Genes Dev. 18:411–22 63. Hsu LY, Liang HE, Johnson K, Kang C, Schlissel MS. 2004. Pax5 activates immunoglobulin heavy chain V to DJ rearrangement in transgenic thymocytes. J. Exp. Med. 199:825–30 64. Reth M, Petrac E, Wiese P, Lobel L, Alt FW. 1987. Activation of Vκ gene rearrangement in pre-B cells follows the expression of membrane-bound immunoglobulin heavy chains. EMBO J. 6:3299–305 65. Khor B, Sleckman BP. 2002. Allelic exclusion at the TCRβ locus. Curr. Opin. Immunol. 14:230–34 66. Yamagami T, ten Boekel E, Andersson J, Rolink A, Melchers F. 1999. Frequencies of multiple IgL chain gene rearrangements in single normal or κL chain-deﬁcient B lineage cells. Immunity 11:317–27 67. Sleckman BP, Khor B, Monroe R, Alt FW. 1998. Assembly of productive T cell receptor delta variable region genes exhibits allelic inclusion. J. Exp. Med. 188:1465–71 68. Ritchie KA, Brinster RL, Storb U. 1984. Allelic exclusion and control of endogenous immunoglobulin gene rearrangement in kappa transgenic mice. Nature 312:517–20 69. Weaver D, Costantini F, Imanishi-Kari T, Baltimore D. 1985. A transgenic immunoglobulin mu gene prevents rearrangement of endogenous genes. Cell 42:117–27

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

564

Jung et al.

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

¨ 70. Rusconi S, Kohler G. 1985. Transmission and expression of a speciﬁc pair of rearranged immunoglobulin mu and kappa genes in a transgenic mouse line. Nature 314:330–34 71. Nussenzweig MC, Shaw AC, Sinn E, Danner DB, Holmes KL, et al. 1987. Allelic exclusion in transgenic mice that express the membrane form of immunoglobulin mu. Science 236:816–19 72. Kitamura D, Rajewsky K. 1992. Targeted disruption of mu chain membrane exon causes loss of allelic exclusion. Nature 69:823–31 73. Constantinescu A, Schlissel MS. 1997. Changes in locus-speciﬁc V(D)J recombinase activity induced by immunoglobulin gene products during B cell development. J. Exp. Med. 185:609–20 74. Reth M. 1992. Antigen receptors on B lymphocytes. Annu. Rev. Immunol. 10:97–121 75. Sakaguchi N, Melchers F. 1986. Lambda 5, a new light-chain-related locus selectively expressed in pre-B lymphocytes. Nature 324:579–82 76. Kudo A, Melchers F. 1987. A second gene, VpreB in the lambda 5 locus of the mouse, which appears to be selectively expressed in pre-B lymphocytes. EMBO J. 6:2267–72 77. Pillai S, Baltimore D. 1987. Formation of disulphide-linked mu 2 omega 2 tetramers in pre-B cells by the 18K omega-immunoglobulin light chain. Nature 329:172–74 78. Martensson IL, Rolink A, Melchers F, Mundt C, Licence S, Shimizu T. 2002. The preB cell receptor and its role in proliferation and Ig heavy chain allelic exclusion. Semin. Immunol. 14:335–42 79. ten Boekel E, Melchers F, Rolink AG. 1998. Precursor B cells showing H chain allelic inclusion display allelic exclusion at the level of pre-B cell receptor surface expression. Immunity 8:199–207 80. Ehlich A, Schaal S, Gu H, Kitamura D, Muller W, Rajewsky K. 1993. Immunoglobulin heavy and light chain genes rearrange independently at early stages of B cell development. Cell 72:695–704 81. Pelanda R, Schaal S, Torres RM, Rajewsky K. 1996. A prematurely expressed Igκ transgene, but not VκJκ gene segment targeted into the Igκ locus, can rescue B cell development in λ5-deﬁcient mice. Immunity 5:229–39 82. Alt FW, Rosenberg N, Enea V, Siden E, Baltimore D. 1982. Multiple immunoglobulin heavy-chain gene transcripts in Abelson murine leukemia virus-transformed lymphoid cell lines. Mol. Cell. Biol. 2:386–400 83. Malynn BA, Shaw AC, Young F, Stewart V, Alt FW. 2002. Truncated immunoglobulin Dmu causes incomplete developmental progression of RAG-deﬁcient pro-B cells. Mol. Immunol. 38:547–56 84. Brouns GS, de Vries E, van Noesel CJ, Mason DY, van Lier RA, Borst J. 1993. The structure of the mu/pseudo light chain complex on human pre-B cells is consistent with a function in signal transduction. Eur. J. Immunol. 23:1088–97 85. Matsuo T, Nomura J, Kuwahara K, Igarashi H, Inui S, et al. 1993. Cross-linking of B cell receptor-related MB-1 molecule induces protein tyrosine phosphorylation in early B lineage cells. J. Immunol. 150:3766–75 86. Jumaa H, Hendriks RW, Reth M. 2005. B cell signaling and tumorigenesis. Annu. Rev. Immunol. 23:415–45 87. Papavasiliou F, Jankovic M, Gong S, Nussenzweig MC. 1997. Control of immunoglobulin gene rearrangements in developing B cells. Curr. Opin. Immunol. 9:233–38 88. Schweighoffer E, Vanes L, Mathiot A, Nakamura T, Tybulewicz VL. 2003. Unexpected requirement for ZAP-70 in pre-B cell development and allelic exclusion. Immunity 18:523–33 www.annualreviews.org • V(D)J Recombination

565

ARI

15 February 2006

1:38

89. Aifantis I, Pivniouk VI, Gartner F, Feinberg J, Swat W, et al. 1999. Allelic exclusion of the T cell receptor beta locus requires the SH2 domain-containing leukocyte protein (SLP)-76 adaptor protein. J. Exp. Med. 190:1093–102 90. Gartner F, Alt FW, Monroe R, Chu M, Sleckman BP, et al. 1999. Immature thymocytes employ distinct signaling pathways for allelic exclusion versus differentiation and expansion. Immunity 10:537–46 91. Anderson SJ, Levin SD, Perlmutter RM. 1993. Protein tyrosine kinase p56lck controls allelic exclusion of T-cell receptor beta-chain genes. Nature 365:552–54 92. Tsubata T, Tsubata R, Reth M. 1992. Crosslinking of the cell surface immunoglobulin (mu-surrogate light chains complex) on pre-B cells induces activation of V gene rearrangements at the immunoglobulin kappa locus. Int. Immunol. 4:637–41 93. Kubagawa H, Cooper MD, Carroll AJ, Burrows PD. 1989. Light-chain gene expression before heavy-chain gene rearrangement in pre-B cells transformed by Epstein-Barr virus. Proc. Natl. Acad. Sci. USA 86:2356–60 94. Chen J, Trounstine M, Alt FW, Young F, Kurahara C, et al. 1993. Immunoglobulin gene rearrangement in B cell deﬁcient mice generated by targeted deletion of the JH locus. Int. Immunol. 5:647–56 95. Reth MG, Ammirati P, Jackson S, Alt FW. 1985. Regulated progression of a cultured pre-B-cell line to the B-cell stage. Nature 317:353–55 96. Rolink A, Ghia P, Grawunder U, Haasner D, Karasuyama H, et al. 1995. In-vitro analyses of mechanisms of B-cell development. Semin. Immunol. 7:155–67 97. Mostoslavsky R, Singh N, Kirillov A, Pelanda R, Cedar H, et al. 1998. Kappa chain monoallelic demethylation and the establishment of allelic exclusion. Genes Dev. 12:1801– 11 98. Alt FW, Oltz EM, Young F, Gorman J, Taccioli G, Chen J. 1992. VDJ recombination. Immunol. Today 13:306–14 99. Roldan E, Fuxa M, Chong W, Martinez D, Novatchkova M, et al. 2005. Locus ‘decontraction’ and centromeric recruitment contribute to allelic exclusion of the immunoglobulin heavy-chain gene. Nat. Immunol. 6:31–41 100. Mostoslavsky R, Singh N, Tenzen T, Goldmit M, Gabay C, et al. 2001. Asynchronous replication and allelic exclusion in the immune system. Nature 414:221–25 101. Mostoslavsky R, Alt FW, Bassing CH. 2003. Chromatin dynamics and locus accessibility in the immune system. Nat. Immunol. 4:603–6 102. Yancopoulos GD, Blackwell TK, Suh H, Hood L, Alt FW. 1986. Introduced T cell receptor variable region gene segments recombine in pre-B cells: evidence that B and T cells use a common recombinase. Cell 44:251–59 103. Stanhope-Baker P, Hudson KM, Shaffer AL, Constantinescu A, Schlissel MS. 1996. Cell type-speciﬁc chromatin structure determines the targeting of V(D)J recombinase activity in vitro. Cell 85:887–97 104. Krangel MS. 2003. Gene segment selection in V(D)J recombination: accessibility and beyond. Nat. Immunol. 4:624–30 105. Banerji J, Olson L, Schaffner W. 1983. A lymphocyte-speciﬁc cellular enhancer is located downstream of the joining region in immunoglobulin heavy chain genes. Cell 33:729–40 106. Gillies SD, Morrison SL, Oi VT, Tonegawa S. 1983. A tissue-speciﬁc transcription enhancer element is located in the major intron of a rearranged immunoglobulin heavy chain gene. Cell 33:717–28 107. Romanow WJ, Langerak AW, Goebel P, Wolvers-Tettero IL, van Dongen JJ, et al. 2000. E2A and EBF act in synergy with the V(D)J recombinase to generate a diverse immunoglobulin repertoire in nonlymphoid cells. Mol. Cell 5:343–53

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

566

Jung et al.

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

108. Ferrier P, Krippl B, Blackwell TK, Furley AJ, Suh H, et al. 1990. Separate elements control DJ and VDJ rearrangement in a transgenic recombination substrate. EMBO J. 9:117–25 109. Alessandrini A, Desiderio SV. 1991. Coordination of immunoglobulin DJH transcription and D-to-JH rearrangement by promoter-enhancer approximation. Mol. Cell. Biol. 11:2096–107 110. Schlissel MS, Corcoran LM, Baltimore D. 1991. Virus-transformed pre-B cells show ordered activation but not inactivation of immunoglobulin gene rearrangement and transcription. J. Exp. Med. 173:711–20 111. Chen J, Young F, Bottaro A, Stewart V, Smith RK, Alt FW. 1993. Mutations of the intronic IgH enhancer and its ﬂanking sequences differentially affect accessibility of the JH locus. EMBO J. 12:4635–45 112. Perlot T, Alt FW, Bassing C, Suh H, Pinaud E. 2005. Elucidation of IgH intronic enhancer functions via germ-line deletion. Proc. Natl. Acad. Sci. USA 102:14362–67 113. Serwe M, Sablitzky F. 1993. V(D)J recombination in B cells is impaired but not blocked by targeted deletion of the immunoglobulin heavy chain intron enhancer. EMBO J. 12:2321–27 114. Sakai E, Bottaro A, Davidson L, Sleckman BP, Alt FW. 1999. Recombination and transcription of the endogenous Ig heavy chain locus is effected by the Ig heavy chain intronic enhancer core region in the absence of the matrix attachment regions. Proc. Natl. Acad. Sci. USA 96:1526–31 115. Kottmann AH, Zevnik B, Welte M, Nielsen PJ, Kohler G. 1994. A second promoter and enhancer element within the immunoglobulin heavy chain locus. Eur. J. Immunol. 24:817–21 116. Nitschke L, Kestler J, Tallone T, Pelkonen S, Pelkonen J. 2001. Deletion of the DQ52 element within the Ig heavy chain locus leads to a selective reduction in VDJ recombination and altered D gene usage. J. Immunol. 166:2540–52 117. Seidl KJ, Manis JP, Bottaro A, Zhang J, Davidson L, et al. 1999. Position-dependent inhibition of class-switch recombination by PGK-neor cassettes inserted into the immunoglobulin heavy chain constant region locus. Proc. Natl. Acad. Sci. USA 96:3000–5 118. Cogne M, Lansford R, Bottaro A, Zhang J, Gorman J, et al. 1994. A class switch control region at the 3 end of the immunoglobulin heavy chain locus. Cell 77:737–47 119. Chowdhury D, Sen R. 2001. Stepwise activation of the immunoglobulin mu heavy chain gene locus. EMBO J. 20:6394–403 120. Chowdhury D, Sen R. 2004. Regulation of immunoglobulin heavy-chain gene rearrangements. Immunol. Rev. 200:182–96 121. Morshead KB, Ciccone DN, Taverna SD, Allis CD, Oettinger MA. 2003. Antigen receptor loci poised for V(D)J rearrangement are broadly associated with BRG1 and ﬂanked by peaks of histone H3 dimethylated at lysine 4. Proc. Natl. Acad. Sci. USA 100:11577–82 122. Osipovich O, Milley R, Meade A, Tachibana M, Shinkai Y, et al. 2004. Targeted inhibition of V(D)J recombination by a histone methyltransferase. Nat. Immunol. 5:309–16 123. Muramatsu M, Kinoshita K, Fagarasan S, Yamada S, Shinkai Y, Honjo T. 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102:553–63 124. Chaudhuri J, Tian M, Khuong C, Chua K, Pinaud E, Alt FW. 2003. Transcriptiontargeted DNA deamination by the AID antibody diversiﬁcation enzyme. Nature 422:726– 30 125. Chaudhuri J, Alt FW. 2004. Class-switch recombination: interplay of transcription, DNA deamination and DNA repair. Nat. Rev. Immunol. 4:541–52 www.annualreviews.org • V(D)J Recombination

567

ARI

15 February 2006

1:38

126. Sikes ML, Meade A, Tripathi R, Krangel MS, Oltz EM. 2002. Regulation of V(D)J recombination: a dominant role for promoter positioning in gene segment assembly. Proc. Natl. Acad. Sci. USA 99:12309–14 127. Liang HE, Hsu LY, Cado D, Cowell LG, Kelsoe G, Schlissel MS. 2002. The “dispensable” portion of RAG2 is necessary for efﬁcient V-to-DJ rearrangement during B and T cell development. Immunity 17:639–51 128. Akamatsu Y, Monroe R, Dudley DD, Elkin SK, Gartner F, et al. 2003. Deletion of the RAG2 C terminus leads to impaired lymphoid development in mice. Proc. Natl. Acad. Sci. USA 100:1209–14 129. Dudley DD, Sekiguchi J, Zhu C, Sadofsky MJ, Whitlow S, et al. 2003. Impaired V(D)J recombination and lymphocyte development in core RAG1-expressing mice. J. Exp. Med. 198:1439–50 130. Elkin SK, Ivanov D, Ewalt M, Ferguson CG, Hyberts SG, et al. 2005. A PHD ﬁnger motif in the C terminus of RAG2 modulates recombination activity. J. Biol. Chem. 280:28701– 10 131. West KL, Singha NC, De Ioannes P, Lacomis L, Erdjument-Bromage H, et al. 2005. A direct interaction between the RAG2 C terminus and the core histones is required for efﬁcient V(D)J recombination. Immunity 23:203–12 132. Goldmit M, Ji Y, Skok J, Roldan E, Jung S, et al. 2005. Epigenetic ontogeny of the Igk locus during B cell development. Nat. Immunol. 6:198–203 133. Skok JA, Brown KE, Azuara V, Caparros ML, Baxter J, et al. 2001. Nonequivalent nuclear location of immunoglobulin alleles in B lymphocytes. Nat. Immunol. 2:848–54 134. Kosak ST, Skok JA, Medina KL, Riblet R, Le Beau MM, et al. 2002. Subnuclear compartmentalization of immunoglobulin loci during lymphocyte development. Science 296:158– 62 135. Sayegh C, Jhunjhunwala S, Riblet R, Murre C. 2005. Visualization of looping involving the immunoglobulin heavy-chain locus in developing B cells. Genes Dev. 19:322–27 136. Haines BB, Brodeur PH. 1998. Accessibility changes across the mouse Igh-V locus during B cell development. Eur. J. Immunol. 28:4228–35 137. Bolland DJ, Wood AL, Johnston CM, Bunting SF, Morgan G, et al. 2004. Antisense intergenic transcription in V(D)J recombination. Nat. Immunol. 5:630–37 138. Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, et al. 2005. Antisense transcription in the mammalian transcriptome. Science 309:1564–66 139. Yancopoulos GD, Desiderio SV, Paskind M, Kearney JF, Baltimore D, Alt FW. 1984. Preferential utilization of the most JH-proximal VH gene segments in pre-B-cell lines. Nature 311:727–33 140. Perlmutter RM, Kearney JF, Chang SP, Hood LE. 1985. Developmentally controlled expression of immunoglobulin VH genes. Science 227:1597–601 141. Kearney JF, Bartels J, Hamilton AM, Lehuen A, Solvason N, Vakil M. 1992. Development and function of the early B cell repertoire. Int. Rev. Immunol 8:247–57 142. Schroeder HW Jr, Walter MA, Hofker MH, Ebens A, Willems van Dijk K, et al. 1988. Physical linkage of a human immunoglobulin heavy chain variable region gene segment to diversity and joining region elements. Proc. Natl. Acad. Sci. USA 85:8196–200 143. Berman JE, Nickerson KG, Pollock RR, Barth JE, Schuurman RK, et al. 1991. VH gene usage in humans: biased usage of the VH6 gene in immature B lymphoid cells. Eur. J. Immunol. 21:1311–14 144. Willems van Dijk K, Milner LA, Sasso EH, Milner EC. 1992. Chromosomal organization of the heavy chain variable region gene segments comprising the human fetal antibody repertoire. Proc. Natl. Acad. Sci. USA 89:10430–34

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

568

Jung et al.

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

ARI

15 February 2006

1:38

145. Malynn BA, Yancopoulos GD, Barth JE, Bona CA, Alt FW. 1990. Biased expression of JH-proximal VH genes occurs in the newly generated repertoire of neonatal and adult mice. J. Exp. Med. 171:843–59 146. Yancopoulos GD, Malynn BA, Alt FW. 1988. Developmentally regulated and strainspeciﬁc expression of murine VH gene families. J. Exp. Med. 168:417–35 147. Sheehan KM, Brodeur PH. 1989. Molecular cloning of the primary IgH repertoire: a quantitative analysis of VH gene usage in adult mice. EMBO J. 8:2313–20 148. Sheehan KM, Mainville CA, Willert S, Brodeur PH. 1993. The utilization of individual VH exons in the primary repertoire of adult BALB/c mice. J. Immunol. 151:5364–75 149. Keyna U, Beck-Engeser GB, Jongstra J, Applequist SE, Jack HM. 1995. Surrogate light chain-dependent selection of Ig heavy chain V regions. J. Immunol. 155:5536–42 150. ten Boekel E, Melchers F, Rolink AG. 1997. Changes in the V(H) gene repertoire of developing precursor B lymphocytes in mouse bone marrow mediated by the pre-B cell receptor. Immunity 7:357–68 151. Connor AM, Fanning LJ, Celler JW, Hicks LK, Ramsden DA, Wu GE. 1995. Mouse VH7183 recombination signal sequences mediate recombination more frequently than those of VHJ558. J. Immunol. 155:5268–72 152. Williams GS, Martinez A, Montalbano A, Tang A, Mauhar A, et al. 2001. Unequal VH gene rearrangement frequency within the large VH7183 gene family is not due to recombination signal sequence variation, and mapping of the genes shows a bias of rearrangement based on chromosomal location. J. Immunol. 167:257–63 153. Peschon JJ, Morrissey PJ, Grabstein KH, Ramsdell FJ, Maraskovsky E, et al. 1994. Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deﬁcient mice. J. Exp. Med. 180:1955–60 154. Corcoran AE, Riddell A, Krooshoop D, Venkitaraman AR. 1998. Impaired immunoglobulin gene rearrangement in mice lacking the IL-7 receptor. Nature 391:904–7 155. Chowdhury D, Sen R. 2003. Transient IL-7/IL-7R signaling provides a mechanism for feedback inhibition of immunoglobulin heavy chain gene rearrangements. Immunity 18:229–41 156. Bertolino E, Reddy K, Medina KL, Parganas E, Ihle J, Singh H. 2005. Regulation of interleukin 7-dependent immunoglobulin heavy-chain variable gene rearrangements by transcription factor STAT5. Nat. Immunol. 6:836–43 157. Stanton ML, Brodeur PH. 2005. Stat5 mediates the IL-7-induced accessibility of a representative D-distal VH gene. J. Immunol. 174:3164–68 158. Su IH, Basavaraj A, Krutchinsky AN, Hobert O, Ullrich A, et al. 2003. Ezh2 controls B cell development through histone H3 methylation and Igh rearrangement. Nat. Immunol. 4:124–31 159. Busslinger M. 2004. Transcriptional control of early B cell development. Annu. Rev. Immunol. 22:55–79 160. Hesslein DG, Pﬂugh DL, Chowdhury D, Bothwell AL, Sen R, Schatz DG. 2003. Pax5 is required for recombination of transcribed, acetylated, 5 IgH V gene segments. Genes Dev. 17:37–42 161. Johnson K, Pﬂugh DL, Yu D, Hesslein DG, Lin KI, et al. 2004. B cell-speciﬁc loss of histone 3 lysine 9 methylation in the V(H) locus depends on Pax5. Nat. Immunol. 5:853–61 162. Chang Y, Bosma MJ, Bosma GC. 1999. Extended duration of DH-JH rearrangement in immunoglobulin heavy chain transgenic mice: implications for regulation of allelic exclusion. J. Exp. Med. 189:1295–305 www.annualreviews.org • V(D)J Recombination

569

ARI

15 February 2006

1:38

163. Costa TE, Suh H, Nussenzweig MC. 1992. Chromosomal position of rearranging gene segments inﬂuences allelic exclusion in transgenic mice. Proc. Natl. Acad. Sci. USA 89:2205–8 164. Schlissel MS, Morrow T. 1994. Ig heavy chain protein controls B cell development by regulating germ-line transcription and retargeting V(D)J recombination. J. Immunol. 153:1645–57 165. Iacomini J, Yannoutsos N, Bandyopadhay S, Imanishi-Kari T. 1991. Endogenous immunoglobulin expression in mu transgenic mice. Int. Immunol. 3:185–96 166. Lam KP, Herzenberg LA, Stall AM. 1993. A high frequency of hybridomas from M54 mu heavy chain transgenic mice initially co-express transgenic and rearranged endogenous mu genes. Int. Immunol. 5:1011–22 167. Zhang Z, Zemlin M, Wang YH, Munfus D, Huye LE, et al. 2003. Contribution of VH gene replacement to the primary B cell repertoire. Immunity 19:21–31 168. Zhang Z, Burrows PD, Cooper MD. 2004. The molecular basis and biological signiﬁcance of VH replacement. Immunol. Rev. 197:231–42 169. Chen C, Nagy Z, Prak EL, Weigert M. 1995. Immunoglobulin heavy chain gene replacement: a mechanism of receptor editing. Immunity 3:747–55 170. Covey LR, Ferrier P, Alt FW. 1990. VH to VHDJH rearrangement is mediated by the internal VH heptamer. Int. Immunol. 2:579–83 171. Chen C, Prak EL, Weigert M. 1997. Editing disease-associated autoantibodies. Immunity 6:97–105 172. de Lanerolle P, Johnson T, Hofmann WA. 2005. Actin and myosin I in the nucleus: what next? Nat. Struct. Mol. Biol. 12:742–46 173. Lenart P, Bacher CP, Daigle N, Hand AR, Eils R, et al. 2005. A contractile nuclear actin network drives chromosome congression in oocytes. Nature 436:812–18 174. Trouche D, Khochbin S, Dimitrov S. 2003. Chromatin and epigenetics: dynamic organization meets regulated function. Mol. Cell 12:281–86 175. Lee GR, Spilianakis CG, Flavell RA. 2005. Hypersensitive site 7 of the TH2 locus control region is essential for expressing TH2 cytokine genes and for long-range intrachromosomal interactions. Nat. Immunol. 6:42–48

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-18

570

Jung et al.

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:541-570. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-19

ARI

23 February 2006

15:27

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

A Central Role for Central Tolerance Bruno Kyewski1 and Ludger Klein2 1

Division of Developmental Immunology, Tumor Immunology Program, German Cancer Research Center, 69120 Heidelberg, Germany; email: [email protected]

2

Research Institute of Molecular Pathology, 1030 Vienna, Austria; email: [email protected]

Annu. Rev. Immunol. 2006. 24:571–606 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.23.021704.115601 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0571$20.00

Key Words thymic epithelium, promiscuous gene expression, autoimmune regulator, regulatory T cells, autoimmunity

Abstract Recent elucidation of the role of central tolerance in preventing organ-speciﬁc autoimmunity has changed our concepts of self/nonself discrimination. This paradigmatic shift is largely attributable to the discovery of promiscuous expression of tissuerestricted self-antigens (TRAs) by medullary thymic epithelial cells (mTECs). TRA expression in mTECs mirrors virtually all tissues of the body, irrespective of developmental or spatio-temporal expression patterns. This review summarizes current knowledge on the cellular and molecular regulation of TRA expression in mTECs, outlines relevant mechanisms of antigen presentation and modes of tolerance induction, and discusses implications for the pathogenesis of autoimmune diseases and other biological processes such as fertility, pregnancy, puberty, and tumor defense.

571

ANRV270-IY24-19

ARI

23 February 2006

15:27

INTRODUCTION TRA: tissue-restricted antigen T1DM: type 1 diabetes mellitus MS: multiple sclerosis

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MG: myasthenia gravis cTEC(s): cortical thymic epithelial cell(s) mTEC(s): medullary thymic epithelial cell(s) Promiscuous gene expression (pGE): the expression of a highly diverse set of genes in mTECs, which otherwise are expressed in a strictly tissue-restricted fashion

572

The immune system is delicately balanced between self-antigen-driven tolerance and pathogen-driven immunity. In the healthy individual these two states represent a sliding scale of responsiveness. A shift toward the extreme ends of this scale, i.e., lack of a response (immunodeﬁciency) or an inappropriate, excessive response (autoimmunity or allergy), results in pathophysiological conditions and often overt disease. This complexity of the immune response is also reﬂected in the control of self-tolerance. Notwithstanding all the intricacies of how the immune system evades harmful self-reactivity, there is consensus that T cells, especially the CD4 subset, play a central role in this process. Imposition and regulation of self-tolerance within the T cell repertoire is exerted at two levels. First, the development and selection of T cells in the thymus strongly biases the naive T cell repertoire against self-reactivity (central tolerance) (1, 2). Second, mature T cells are subject to secondary selection (deletion, anergy) in lymphoid and nonlymphoid organs (peripheral tolerance) (3, 4). In addition, regulatory T cells (Tregs) may suppress the activation of those self-reactive T cells that escape selection (5, 6). The relative importance attributed to either central or peripheral mechanisms of tolerance has varied periodically according to prevailing schools of thought and experimental progress. Of recurrent interest is the extent to which central mechanisms contribute to T cell tolerance toward the plethora of tissue-restricted self-antigens (TRAs). [To accord with the data published on gene expression, genes that are expressed in fewer than 5 of 45 tissues tested are operationally deﬁned as tissue-restricted (7)]. Elucidation of the physiological mechanisms underlying tolerance toward dominant auto-antigens implicated in autoimmune diseases might also stimulate new therapeutic approaches. Examples would be type 1 diabetes mellitus (T1DM), multiple sclerosis (MS), or myasthenia gravis

Kyewski

·

Klein

(MG), all of which are restricted to one or a few target organs. The extent of central self-tolerance is speciﬁed by the diversity of self-antigens that are accessible to the nascent repertoire within the thymus. The pool of self-epitopes available for repertoire selection comprises intrathymically expressed ubiquitous antigens and antigens speciﬁc to various types of thymic antigen-presenting cells (APCs), i.e., cortical and medullary thymic epithelial cells (cTECs and mTECs), thymic dendritic cells (DCs), macrophages, and thymic B cells. Furthermore, self-antigens can gain access to the thymus either via the circulation or by association with immigrating cells (8) (Figure 1). Because the latter routes are unlikely to afford a comprehensive representation of extrathymic tissue antigens, peripheral tolerance mechanisms have been invoked to explain the absence of auto-reactivity toward tissue-restricted self-constituents. This view has undergone a gradual but profound change; the essential contribution of the thymus to tissue-speciﬁc tolerance is now recognized (9, 10). The unorthodoxy of the underlying phenomena might explain why this insight has taken so long to gain wider acceptance. Thus, in apparent contravention of the rules established for cell type–speciﬁc regulation of gene expression, TECs express a wide host of TRAs. This phenomenon, possibly unique to TECs, has been termed promiscuous gene expression (pGE) (11). Except for the involvement of the autoimmune regulator Aire, the molecular and cellular regulation of this unorthodox gene expression pattern is poorly understood. Self-antigens expressed by mTECs represent virtually all parenchymal organs, thereby mirroring the peripheral self and pre-empting the peripheral encounter of potentially dangerous T cells with TRAs. Here, we place this surprising twist to original concepts of self/nonself discrimination into the wider context of thymic T cell development, review recent progress in this area, and discuss the biological and medical implications.

ANRV270-IY24-19

ARI

23 February 2006

15:27

Cortex

Capsule Subcapsular epithelial cell Thymocyte

Medulla

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Trabecule

Cortical epithelial cell Macrophage Medullary epithelial cell B cell

Blood vessel

Dendritic cell

Figure 1 Cellular composition of the thymus. The major cell types and the sequential cell-cell interactions along the migratory route of developing thymocytes are depicted. The different APCs are color-coded. mTECs, highlighted in red, play an essential role in self-tolerance induction toward tissue-restricted self-antigens. Shaded areas depict functionally distinct stratiﬁed microenvironments as recently proposed (193).

THYMIC SELECTION BY NUMBERS: MATCHING NEGATIVE SELECTION WITH SELF-ANTIGEN DIVERSITY During postnatal life, ∼10–100 hematopoietic precursors on average enter the thymus from the bloodstream per day (12) in what is believed to be a discontinuous, “gated” process (13). Upon commitment to the T lineage, these cells undergo approximately 20 divisions, mostly within the double-negative (DN) stage of T cell development that extends over about two weeks (12, 14). This massive expansion of the precursor pool results in the generation of about 5 × 107 T cells daily, a ﬁgure in remarkable contrast to the estimated 1–2 × 106 mature T cells that are actually released daily into the circulation. The loss of over 95% of thymocytes reﬂects the stringent selection processes that shape the developing T cell repertoire. The ﬁrst checkpoint, so-called beta-selection, is con-

tingent upon pre-TCR signaling and ensures that only those DN thymocytes that have successfully rearranged their TCRβ locus progress to the CD4+ CD8+ (DP) compartment (reviewed in 15). After progression to the DP stage and rearrangement of the TCRα locus, all subsequent developmental decisions of thymocytes are dictated by interactions with peptide/MHC (pMHC) ligands on stromal cells within the thymic microenvironment. This fascinating paradigm of one receptor driving diverse and even diametrically opposite cellular responses—differentiation, proliferation, or apoptosis—has captured the imagination of immunologists ever since the basic principles of positive and negative selection in the thymus were ﬁrst discovered (reviewed in 1). Failure of a TCR to interact with pMHC ligands within a certain window of afﬁnity/avidity is interpreted as reﬂecting a useless speciﬁcity, i.e., lack of self-MHC restriction, resulting in so-called death by www.annualreviews.org • Central T Cell Tolerance

TCR: T cell receptor

573

ARI

23 February 2006

15:27

neglect of the respective DP thymocytes. Most apoptotic cell death in the thymus has been attributed to failure of positive selection, indicating that the quantitative effect of selection for self-MHC restriction (positive selection) far exceeds that of selection against potentially dangerous auto-reactivity (negative selection) (16). The loss of cells during positive selection is mitigated by two mechanisms. First, TCR structures, although generated by random rearrangement, appear to have an intrinsic afﬁnity for their polymorphic MHC ligands, presumably a result of their coevolution (17–21). Second, sequential rounds of recombination at the TCRα locus within the 3- to 4-day life span of DP T cells allow several distinct TCR speciﬁcities to be generated, thereby increasing the chance of an appropriate TCR/MHC match and thus the survival of the individual thymocyte (22). Nevertheless, an estimated 90%–95% of thymocytes are lost due to death by neglect (23, 24). How does the quantitative impact of intrathymic tolerance induction compare with that of positive selection? Approximately 50%–70% of positively selected, or rather selectable, T cells are thought to be subject to negative selection (25–28), a conservative estimate based mainly on experimental approaches that allowed quantiﬁcation of the fraction of T cells that are deleted by hematopoietic cells, i.e., thymic DCs, whereas it is more difﬁcult to address experimentally the contribution and extent of negative selection by radio-resistant thymic stromal cells. Using reaggregation thymic organ culture (RTOC), Anderson et al. (29) concluded that TECs are far less efﬁcient mediators of negative selection than thymic DCs. Although this observation supports the concept of stromal cell specialization for negative selection, other studies have shown, at least in vitro, that deletion does not require antigen recognition on a dedicated stromal interaction partner. Thus, almost all thymic stromal cells or even unrelated cells expressing the appropriate MHC were able to deliver death signals to immature thymocytes in suspension cul-

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

574

Kyewski

·

Klein

ture, fetal thymic organ culture, or RTOC (30–32), compatible with the notion that negative selection can, in principle, occur in both the cortex and the medulla (33). Nonetheless, there is strong argument that the medulla is the specialized site to rid the thymus of autoreactive T cells (33a). First, the expression of the TCR is not fully upregulated on DP T cells in the cortex (34). Consequently, this “down-tuned” TCR sensitivity entails a bias toward low-avidity interactions with pMHC ligands on cortical epithelial cells and may therefore tip the balance in favor of survival (positive selection) instead of death (negative selection). Second, upon transition into the medulla, both T cell intrinsic, i.e., full upregulation of the TCR, and T cell extrinsic factors, i.e., costimulatory molecules expressed by DCs and mTECs, provide a bias toward high-avidity signals. Third, and most pertinent here, the medullary microenvironment provides a unique representation of selfantigens.

PROMISCUOUS EXPRESSION OF PERIPHERAL ANTIGENS BY THYMIC STROMAL CELLS The ﬁnding that over half of the positively selected T cell repertoire does not pass thymic censorship challenged the long-standing view that the majority of the genome or, more precisely, proteome might not be visible to T cells during intrathymic maturation, such that thymic mechanisms of tolerance should not apply to a signiﬁcant fraction of potentially auto-reactive T cells. If a limited representation of self within the thymus eliminates almost two thirds of T cell speciﬁcities, why then is the immune system not functionally crippled through the subsequent action of peripheral tolerance mechanisms? To resolve this paradox, it has been argued that the immune system may indeed not be tolerized at all toward many peripheral antigens and that the existence of auto-reactive T cells per se may not pose an autoimmune hazard in the healthy individual (35, 36). Auto-reactive

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

lymphocytes are indeed part of the peripheral T cell repertoire, as best illustrated by the fact that deliberate immunization with particular self-antigens can elicit experimental autoimmune diseases. This is especially well documented for auto-antigens of the central nervous system (CNS) (37, 38). Alternatively, the assumption that only a limited spectrum of “self ” is represented in the thymus for tolerance induction may be wrong, and central tolerance may cover a much wider range of self-antigens than previously assumed. As early as 1989 it was hypothesized that the thymus represents a patchwork quilt of diverse tissues at the RNA expression level (39). Indeed, numerous individual reports on intrathymic expression of TRAs entertained the possibility that the scope of intrathymically expressed self-antigens might extend beyond the limits imposed by general rules of cell type–speciﬁc gene expression. Two areas of research independently furnished support for a more extended role of the thymus in conferring tolerance to peripheral tissues. First, Le Douarin and associates demonstrated in the chicken/quail model and subsequently in mice that transplantation of the thymus anlage, i.e., pure thymic epithelium, confers tolerance to transplanted tissues such as limb buds or skin (40–43). In a separate series of studies, neonatal thymectomy up to day 3 was shown to lead to a multiorgan autoimmune syndrome including gastritis, sialadenitis, hepatitis, and diabetes (44 and references therein). In both experimental systems, dominant rather than recessive tolerance mechanisms such as deletion seemed to play the more important role. Second, mice transgenic for neo-self-antigens, which had originally been designed to study mechanisms of peripheral tolerance, frequently demonstrated a role of “ectopic” expression of the respective antigen in the thymus (45–48). However, many of these observations initially received little attention and were interpreted as artifacts of transgenesis. Hanahan and colleagues (46), in a study of tolerance toward antigens implicated in di-

abetes, were the ﬁrst to reinterpret the intrathymic expression of transgenes under the control of the insulin promoter as a physiological property of the endogenous insulin gene locus. This seminal study was supported by several reports challenging the simplistic concept of a clear demarcation between central and peripheral self-antigens (49–52a), and the existence of specialized peripheral antigen expressing (PAE) cells in the thymus was being postulated concurrently on the basis of histological evidence (53). Early attempts to unambiguously characterize these PAE cells were unsuccessful (54) until a rigorous puriﬁcation protocol was established that allowed for analysis of extremely pure thymic stromal cell preparations. This revealed the expression of over 30 prototypic peripheral antigens, largely restricted to mTECs (11). The scope of pGE in mTECs has now been characterized in much greater detail (Figures 2 and 5) (7, 51, 55–57). The surprising ﬁnding was the functional and structural diversity of antigens expressed in TECs representing essentially all organs, including not only genes expressed in a spatially restricted fashion but also developmentally and temporally regulated genes. An interspecies comparison showed the size and complexity of this gene pool to be highly conserved between mouse and human. Moreover, pGE was fully maintained in the involuting thymus of aging mice, suggesting that a diverse array of self-ligands is provided as long as the thymus continues to replenish the peripheral T cell pool (11). There is now consensus that epithelial cells of the thymic cortex and medulla originate from a common endoderm-derived progenitor, as originally proposed by Le Douarin (58– 62). This precursor differentiates into cTECs and mTECs, where considerable phenotypic and functional heterogeneity is found among the latter in particular (63, 64). Whether these subsets of mTECs develop in parallel or represent sequential stages of progressive differentiation is not clear. www.annualreviews.org • Central T Cell Tolerance

575

ANRV270-IY24-19

ARI

23 February 2006

15:27

exp1 Macrophages

exp2 exp1

Dendritic cells

exp2 1.5 exp1

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

cTECs

Liver

exp2

1.0

exp2

0.5

exp3

0.0

exp1

0.5

exp1 mTECs exp2

1.0 1.5

exp1 Hippocampus

exp2 exp3

Figure 2 mTECs are specialized in promiscuous gene expression. Analysis of variance (ANOVA) of global gene expression of four thymic stromal cell subsets and two tissues. Note that mTECs display the highest proportion of differentially expressed genes among thymic stromal cells and even whole liver. This transcriptional diversity is surpassed only by the hippocampus ( yellow, upregulated genes; blue, downregulated genes; black, same expression as the mean for that gene across all samples). For details, see Reference 7.

Aire: autoimmune regulator

576

An initial analysis at the level of a selected cohort of individual genes suggested that pGE varies among thymic stromal cell types (11). While the expression of the majority of genes was conﬁned to mTECs, certain genes were equally expressed in cTECs and mTECs or at low levels in DCs. An extended analysis of gene expression in thymic stromal cells using DNA microarrays disclosed at least three distinct gene pools (7). Pool 1 includes genes expressed at comparable levels in cTECs and mTECs; pool 2 includes genes conﬁned to the mTEC lineage irrespective of phenotypic heterogeneity; pool 3 includes genes expressed in mTECs with a “mature” (MHC class IIhi CD80hi ) phenotype (Figure 3). Genes in pool Kyewski

·

Klein

3 can be further subdivided by whether their expression is dependent or not on the putative transcription factor Aire (discussed below). While these pools were deﬁned by assessment of gene expression in cDNA samples prepared from bulk-sorted thymic stromal cells, a second distinctive feature of pGE regards expression at the single cell level. When probed with antibodies or by in situ hybridization (ISH), a number of individual TRAs were found to be expressed by a minor subset of mTECs (1%–3%) (11, 65). These ﬁgures are in line with earlier observations regarding the frequency of insulin- or somatostatinexpressing cells of unknown identity in cryosections of the mouse thymus (54).

ANRV270-IY24-19

ARI

23 February 2006

15:27

Endoderm

TEC precursor

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

mTEC

cTEC

CD80 lo “Closed” chromatin

PLP

1

CRP 1 + 2

CD80 hi Boundary “Open” chromatin

“Closed” chromatin

Ins2

Boundary

Aire

X

Aire X

Aire

X

Aire X

“Closed” chromatin

1+2+3

Csnb Degree of promiscuous gene expression Figure 3

Terminal differentiation model of promiscuous gene expression. Progressive differentiation of mTECs correlates with increased pGE as reﬂected by sequential addition of distinct gene pools (1 to 3). This is thought to be accomplished by the cooperation of genetic and epigenetic mechanisms. Promiscuous expression of TRAs may be either strictly or partially Aire-dependent or Aire-independent. This model is based on the assumption that upregulation of CD80 and MHC class II expression denotes progressive differentiation of mTECs (prototypical antigens for each pool: CRP, C-reactive protein; Csnb, casein beta; Ins2, insulin 2; PLP, proteolipid protein). www.annualreviews.org • Central T Cell Tolerance

577

ANRV270-IY24-19

ARI

23 February 2006

15:27

CELLULAR REGULATION OF PROMISCUOUS GENE EXPRESSION: COMPETING MODELS

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Two competing models, the “terminal differentiation model” and the “developmental” or “progressive restriction” model, have been proposed to integrate the intricate intrathymic expression patterns of TRAs into a coherent model. In this section, we outline the basic features of the two models. The ﬁnding that the complexity of pGE increases from cTECs to MHC class IIlo CD80lo mTECs to MHC class IIhi CD80hi mTECs led to the proposition of the terminal differentiation model. This model holds that two differentiation steps are required to unfold the full extent of pGE in mTECs: commitment to the mTEC lineage and development of a “mature” phenotype, as deﬁned by high expression of surface markers such as MHC class II and CD80/86 (Figure 3). Concomitant with this developmental sequence, expression of TRAs becomes increasingly diverse. At the same time, mTECs acquire full APC-competence by upregulating components of the antigenprocessing and -presentation pathway, including certain cathepsins, the immunoproteasome, and nonclassical MHC class molecules such as H-2 DM (7). As a consequence, the mTEC subset that appears functionally most competent to present antigen also shows the highest degree of TSA expression and therefore is the candidate cell mediating self-tolerance via direct interactions with T cells. In the context of the terminal differentiation model, note that CD80hi mTECs are short-lived, with a half-life of the order of three weeks (66; B. Kyewski, unpublished information). Thus, within individual medullary areas, maturation of mTECs would proceed in an outward direction with the cortico-medullary junction, which harbors the most “mature” mTECs, representing a corridor of TRA display through which thymocytes have to pass upon entry into the

578

Kyewski

·

Klein

medulla (Figure 4). This proposition concurs with the conspicuous location of small clusters of TSA-expressing cells in the outer medulla (54, 65, 67, 68). Furthermore, Aireexpressing cells that represent a subset of mature mTECs tend to congregate in the outer medulla (69). There is still no formal proof of a direct precursor product relationship between MHC class IIlo CD80lo and MHC class IIhi CD80hi mTECs during embryogenesis or in the adult thymus. However, the terminal differentiation model is supported by the observation that MHC class IIhi CD80hi mTECs resemble other terminally differentiated epithelial cells such as keratinocytes as far as upregulated expression of the keratinocyte differentiation complex, tight junction components (i.e., claudin 4 and 7), or the differentiation markers involucrin and loricrin are concerned (70, 71). These ﬁndings indicate that mTECs, albeit exceptional among epithelial cells in forming a three-dimensional network, undergo a differentiation sequence akin to stratiﬁed epithelia of skin and gut, and pGE in mTECs clearly segregates with markers of terminally differentiated epithelium when assessed in situ or in isolated cells. The competing model termed developmental or progressive restriction model [an extension of the mosaic model (72)] differs from the terminal differentiation model in assuming that pGE is a property of an immature, possibly still multipotent, progenitor stage (73). Differentiation of mTECs would progressively restrict transcriptional promiscuity. Concomitant with this restriction, tissuespeciﬁc gene expression programs would be enacted with individual mTECs emulating gene expression of a particular native tissue. This model is adopted from a concept emerging from studies of early hematopoiesis (74, 75). Here, it was demonstrated that multipotent hematopoetic stem cells coexpress multiple genes afﬁliated to various more mature hematopoetic lineages at low levels. Upon commitment into a particular lineage, this broad pGE becomes progressively restricted. Thus, in hematopoesis pGE is a hallmark of a

ANRV270-IY24-19

ARI

23 February 2006

15:27

mTEC subclone 1 - Aire high Chromosome 3 Chromosome 1

Aire Aire

mTEC subclone 2 - Aire negative and/or low Chromosome 12 Chromosome 1

Outer medulla

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Inner medulla

Figure 4 The multi-clonal model of pGE. Small clones of terminally differentiated mTECs are continuously generated de novo from committed mTEC precursors. Each clone expresses a distinct set of Aire-dependent or Aire-independent TRAs clustered on various chromosomes. Gene expression within the same cluster may be complementary in Aire+ and Aire− mTEC clones. This model concurs with the multi-clonal origin of mTECs (58) and the observation that TRAs are frequently expressed in small clusters of two to four cells in the outer medulla. The small cluster size might result from a limited number of cell divisions after a certain stage of differentiation is reached and expression of TRAs is switched on. The aggregation of these TRA-expressing clones in the outer medulla creates a corridor of dense self-antigen display that thymocytes have to pass during their transit from cortex to the medulla. This spatial arrangement is reminiscent of the model of stratiﬁed cortical microenvironments (193).

multipotent, uncommitted stage, rather than a terminal differentiation stage. By assuming a similar mechanism to be at work in mTECs, the developmental model depicts intrathymic pGE as a reﬂection of a more general principle of gene regulation during cellular differentiation. In contrast, the terminal differentiation model posits that pGE is a mechanism unique to mTECs superimposed on their developmental program. The developmental model implies that mTECs mimic lineage-afﬁliated gene expression programs. This prediction can be tested only by gene expression analysis at the single cell level, which has not been reported to date. Moreover, the regulation of TRA expression in mature mTECs should be similar to that operating in the respective peripheral tissue, again a prediction that has not yet been rigorously tested. However, certain observations cannot be easily reconciled with this prediction. For example, the level of insulin pro-

duction in beta cells of the mouse pancreas is independent of the copy number of the insulin1 and insulin2 genes in mice; this feedback mechanism, however, does not operate in the thymus, where there is a direct correlation between gene copy number and RNA level (76). Taken together, proving or disproving either model rests on the clariﬁcation of two unresolved issues: ﬁrst, the establishment of a developmental relationship among mTEC subsets expressing different degrees of TRAs and second, a deﬁnition of the coexpression pattern of TRAs at the single cell level.

MOLECULAR REGULATION OF PROMISCUOUS GENE EXPRESSION: AIRE AND BEYOND Any model of the molecular regulation of pGE has to accommodate the following salient features: (a) Different thymic stromal www.annualreviews.org • Central T Cell Tolerance

579

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

Autoimmune polyglandular syndrome type 1 (APS-1): a rare monogenic, autosomal, recessive immune disorder affecting multiple organs, with a predilection for endocrine organs; it results in hypoparathyroidism and adrenocortical failure

580

15:27

cell subsets express different pools of TRAs, (b) promiscuously expressed genes have no obvious structural and functional commonalities, (c) gene regulation in mTEC is uncoupled from developmental regulation, (d) promiscuously expressed genes are highly clustered in the genome, (e) the expression of a particular TRA appears to be restricted to a minor subset of 1%–3% of mTECs, and (f) allelic polymorphisms in the regulatory region of TRAs inﬂuence the level of promiscuous transcription. Only one molecular determinant has been identiﬁed to date. The autoimmune regulator (Aire), the gene mutated in the rare autoimmune disorder autoimmune polyglandular syndrome type 1 (APS-1), controls the expression of a large portion of promiscuously expressed genes in mTECs (77). Aire is a 54.5kDa molecule with a nuclear localization signal and several potential DNA-binding and protein interaction domains (for review, see 78). It is highly expressed in human and mouse mTECs in which it localizes to nuclear bodies. In a minor subset of mTECs, the general transcriptional regulator CBP colocalizes to these nuclear bodies. Aire, in cooperation with CBP, trans-activates the transcription of the interferon-β gene upon transfection in HEK293 cells (79). Furthermore, Aire was reported to bind distinct DNA sequence motifs in vitro (80, 81). No experimental evidence exists to date for direct binding of Aire to cisacting sequences of target genes in a physiological context. Likewise, common sequence motifs in regulatory regions of Aire target genes have not been reported. Although we currently have no molecular explanation of how this single molecule controls the transcription of such an array of genes, certain observations offer clues to its mode of action. Thus, intrathymic ectopic expression of transgenic neo-self-antigens under the direction of tissue-restricted promoters often retained its dependency on Aire, as observed for the respective genes in the native genetic context (82; J. Buer, personal communication). This dependency is robust even across species barriers and pertains to promoters of genes whose Kyewski

·

Klein

expression is strictly or only partially dependent on Aire (77, 82a). As this ﬁnding appears to be independent of the integration site, it is most easily reconciled with the existence of cis-acting sequences at the level of individual genes specifying the action of Aire directly or indirectly. Aire might form part of higher-order transcriptional complexes (e.g., in conjunction with CBP) and thus modulate the transcriptional activity without necessarily binding directly to DNA. It could thereby inﬂuence transcription, depending on the particular composition of the transcriptional complex and the particular promotor, which may explain the varying degrees to which promiscuously expressed genes depend on the presence of Aire. The recent observation that intrathymic expression of adjacent genes in the mouse casein locus is regulated by Aire in an alternating fashion also supports the notion that Aire acts in a gene-speciﬁc manner rather than by altering higher-order chromatin conﬁgurations (7, 83). This is not to say that differential expression of directly neighboring genes excludes epigenetic regulation, e.g., regulation of the imprinting status of the adjacent imprinted genes Igf2 and H19 is indeed controlled by higher-order chromatin conﬁgurations (84). A contribution of epigenetic mechanisms to the control of pGE is strongly implied by the following ﬁndings: (a) promiscuously expressed genes tend to localize in clusters; (b) the imprinting status of Igf2, which is 17fold upregulated in mTECs, is speciﬁcally lost in these cells (7); (c) derepression of certain cancer germ cell antigens like MAGE-A1 and -A3 in mTECs correlates with promotor hypomethylation (C. deSmedt, unpublished information). Signiﬁcant clustering of TRAs expressed in mTECs is highly conserved between mouse and human. Clusters identiﬁed so far contain up to 16 genes, including homologous members of gene families and nonhomologous genes known to be restricted to different tissues. When the mouse casein gene region on chromosome 5 was analyzed in bulk

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

mTEC preparations, expression of 11 genes within 1 Mb was contiguous (7). This pattern was observed only in mTECs but not in the respective tissues to which the genes encoded in this cluster were restricted. Transcription of genes within this cluster seems to be regulated at two levels. First, six genes forming the core region of this cluster are coordinately upregulated in CD80hi mTECs, implying that this domain becomes accessible to transcription only upon mTEC differentiation. Second, individual genes of the casein family are, however, differentially regulated by Aire, with directly neighboring genes being induced, repressed, or unaffected by Aire (7, 83). In the context of the terminal differentiation model, one interpretation of these ﬁndings is that upon differentiation of mTECs, widespread alterations in the accessibility of deﬁned local regions in the genome would allow DNA-binding complexes to differentially control transcription of individual genes within such delineated domains. Aire would be one component of such complexes (Figure 3). The ability of Aire to control such an array of different genes would thus be contingent on epigenetic alterations speciﬁc to mTECs, a presumption that accords with the observation that Aire, when overexpressed in human monocytic cells, does not mimic typical pGE (85, 86). Gene clustering across the genome is found in many species and is a feature of both organ-speciﬁc and housekeeping genes (87). It has been interpreted as the result of shufﬂing of genetic material to facilitate coregulation. Alternatively, it may be due to gene duplication events, which occurred at a time during evolution too recent to have allowed for their spreading in the genome. Thus, the selection of promiscuously expressed genes would not be arbitrary, as it seems at ﬁrst sight, but would rather be inﬂuenced by the evolution of gene order. In this context, note that, when comparing different pools of promiscuously expressed genes, the degree of clustering correlates with the content of bona ﬁde TRAs and this in turn correlates

with the content of genes regulated by Aire (Figure 5). The molecular mechanisms underlying pGE might thus have exploited a propensity of TRAs to cluster in the genome in the ﬁrst place (87). Genes for which epigenetic regulation has been extensively documented are imprinted genes. At least four imprinted genes, Igf2, Cdkn1c, Plagl1, and H19, are overexpressed in mTECs. Incidentally, coexpression of insulin2, Igf2, and H19 represents another example of a cluster with contiguous gene expression. Strikingly, the imprinting of Igf2, but not of the H19 and Cdkn1c genes, is selectively lost in mTECs. The methylation status of so-called differentially methylated regions (DMR) within the Igf2/H19 locus is known to control imprinting, and changes in DNA methylation might contribute to promiscuous expression of imprinted genes and selective loss of imprinting (88). This view is supported by the observation that pGE resembles the gene expression pattern in hypomethylated ﬁbroblasts (89). Given the intricate regulation of the Igf2/H19 locus, it is, however, too simplistic to assume that global demethylation during mTEC differentiation explains the observed changes in the expression of imprinted genes and other TSAs regulated by promotor methylation (90). To complicate matters further, Igf2 is not only biallelically expressed but is also dependent on Aire, as is the adjacent insulin2 gene. A detailed analysis of alterations of the methylation status and the chromatin conﬁguration of this well-studied locus in mTECs will, it is hoped, disclose basal mechanisms directing pGE. Taken together, the regulation of pGE clearly entails a degree of speciﬁcity inconsistent with the proposition of “random gene derepression” (56).

MECHANISM(S) OF TOLERANCE INDUCED BY PROMISCUOUS GENE EXPRESSION IN mTECS The contribution of thymic epithelium to intrathymic induction of tolerance has long www.annualreviews.org • Central T Cell Tolerance

581

ANRV270-IY24-19

ARI

15:27

mTEC versus cTEC

cTEC versus mTEC

mTEC CD80 hi versus mTEC CD80 lo

33 tissues

11 tissues

32 tissues

28 (Relative % of tissue-restrictive genes)

28 (Relative % of tissue-restrictive genes)

33 (Relative % of tissue-restrictive genes)

leukocytes epidermal CNS

liver

intestine

lymphocytes mammary gland thyroid bone heart oocyte stomach tongue umbilical cord kidney pancreas testis endothelium lung placenta blastocyst eye

epidermal

thymus

CNS

CNS hematopoietic cells lymphoid

intestine

leukocytes epidermal oocyte adipose tissue liver bone kidney

liver

others

others

150

Frequency

30 20 10 0

60

P<0.001

0

50

100

150

P=0.018 100

50

Number of neighbors (Window size 200 kb)

50

P<0.001

40 30 20 10

0

200

Frequency

50 40

leukocytes bone thyroid oocyte lymphocytes tongue mammary gland stomach testis heart salivary gland umbilical cord kidney prostate blastocyst adrenal gland endothelium

medial uterus lung olfactor y eye epithelium fetal erythrocytes trachea adipose tissue placenta pancreas ameloblast vomeronasal organ

spleen PNS trachea adrenal gland ameloblast salivary gland fetal erythrocytes parathyroid prostate uterus vomeronasal organ

Frequency

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

23 February 2006

0 0

5

10

15

Number of neighbors (Window size 200 kb)

0

20 40 60 80 100 120

Number of neighbors (Window size 200 kb)

Figure 5 Tissue representation in different gene pools. Genes identiﬁed as overexpressed in comparisons of different thymic stromal cell populations were assigned to tissues according to their predominant expression using information that was combined from various public databases. Genes with expression restricted to less than 5 of 45 tissues were designated “tissue restricted.” Note that a decrease in tissue diversity is accompanied by a decrease in genomic clustering, as deduced from neighborhood analysis. Shown here is the distribution of the number of neighbors located on the same chromosome within a distance of 200 kb in 1000 random gene lists. The actually observed number of neighbors in the respective gene pools is indicated by a red line (for more details, see Reference 7). This correlation might be due to pGE exploiting a pre-existing clustering of tissue-restricted genes in the genome. This analysis was done in C57BL/6 mice except for Aire knockout mice, which were of mixed genetic background.

582

Kyewski

·

Klein

ANRV270-IY24-19

ARI

23 February 2006

Aire–/– mTEC CD80hi versus Aire–/– mTEC CD80lo

WT mTEC CD80hi versus Aire–/– mTEC CD80hi

21 tissues

23 tissues

21 (Relative % of tissue-restrictive genes)

45 (Relative % of tissue-restrictive genes)

leukocytes liver

CNS lymphocytes

epidermis/epithelium leukocytes

intestine

mammary gland

epidermis intestine

testis mammary gland placenta bone heart oocyte

CNS

liver

others

tongue

trachea

hematopoietic blastocyst eye prostate fertilized egg umbilical cord salivary gland

uterus

stomach

40 20 0

gallbladder

ameloblast

Frequency

60

salivary gland lung

heart

P=0.013

80

bone umbilical cord oocyte, fertilized egg tongue testis stomach pancreas kidney

thyroid

others

Frequency

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

15:27

fetal erythrocytes

400

P<0.001

300 200 100

0

10

20

30

40

Number of neighbors (Window size 200 kb)

0

0

10

20

30

40

Number of neighbors (Window size 200 kb)

Figure 5 (Continued)

been controversial. Conﬂicting results ranging from no role in self-tolerance to clear demonstrations of deletional or nondeletional mechanisms might, in hindsight, be attributable to differences in the read-out systems used (43, 91–99). One common denom-

inator of many of these studies was the observation that radio-resistant thymic stromal cells, i.e., TECs, were far less efﬁcient than bone marrow–derived cells. How can this notion be reconciled with the recurrent observation that promiscuously expressed genes are www.annualreviews.org • Central T Cell Tolerance

583

ARI

23 February 2006

15:27

so efﬁcient in deleting auto-reactive T cells, a fact that is all the more astounding given that these genes were often reported to be expressed by only a minor subset of mTECs (11, 54, 65, 67)? Two mutually nonexclusive explanations have been suggested. First, TRAs expressed in mTECs may be transferred to and cross-presented by thymic DCs. This “antigen spreading” would increase the availability of tolerizing antigen to the vicinity of expressing mTECs. Second, deletion may be mediated by rare mTECs themselves, a scenario that may involve extensive scanning of stromal cells in the medulla by thymocytes (100, 101). Based on recent in vivo imaging of the dynamics of T cell–DC interactions in lymph nodes, a single DC may scan an estimated 5000 T cells per hour (102). Assuming similar dynamics for mTEC-thymocyte encounters and the presence of about 1000 mTECs per mouse thymus expressing a given TRA, about 5 × 106 thymocytes (or approximately one third of all medullary thymocytes) can be scanned per hour. CD80hi mTECs expressing the bulk of TRAs represent about one third of all mTECs, i.e., ∼30,000 cells per thymus. Accordingly, scanning of this subset by 15 × 106 thymocytes would take 90 h or 4 days. This time frame is well in accord within the residence time of thymocytes in the medulla (103). The requirement for cross-presentation (used here for both MHC class I– and II– restricted antigen presentation) of mTECderived TRAs by DCs has been analyzed in two models. In the case of a soluble neo-selfantigen, antigen presentation by hematopoietic cells was dispensable for efﬁcient deletion (65). Because deletion in this model occurred at the DP stage and the antigen was secreted, it could not be unequivocally ascribed to presentation by mTECs and a role for antigen transfer and cross-presentation by cTECs could not be formally excluded. Using a membrane-bound form of ovalbumin expressed under the control of the rat insulin promoter, Gallegos & Bevan (104) showed that Ova-expressing mTECs autonomously

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

584

Kyewski

·

Klein

deleted speciﬁc CD8 T cells, but that crosspresentation by DCs was required for deletion of speciﬁc CD4 T cells. These studies were complemented by another approach in which deletion of HEL-speciﬁc CD4 T cells was compared when the neo-antigen was expressed in membrane-bound or soluble form. Despite lower protein expression levels, the soluble form mediated more efﬁcient deletion (105). Taken together, these studies make two points. First, rare mTECs by themselves can, in principle, delete T cells, which implies a highly efﬁcient scanning mechanism. Second, cross-presentation contributes to deletion and appears essential for particular TCR speciﬁcities (Figure 6). Which parameters dictate the requirement for cross-presentation of mTEC-derived TRAs is unclear; the avidity of the TCR for its cognate peptide-MHC ligand and/or the density of self-epitopes displayed by mTECs are likely candidates. Selfepitope density may vary considerably among different TRAs, depending on the stability of the respective protein and the efﬁciency of antigen processing by the MHC class I or class II pathway. Cross-presentation of mTEC-derived antigens appears to occur constitutively and very efﬁciently in vivo, as assessed by direct ex vivo presentation of TRAs by puriﬁed thymic DC (B. Kyewski, unpublished information). Compatible with this ﬁnding, most thymic DC belongs to the CD8 positive lineage of DCs, which is the most efﬁcacious DC subset in mediating cross-presentation (106, 107). By inference from studies on peripheral DCs and in vitro models, several routes by which thymic DCs may acquire protein material from mTECs are plausible. First, DCs efﬁciently cross-present antigens derived from apoptotic cells (108–110). Given that CD80hi mTEC turn over within three weeks under steady-state conditions (66; B. Kyewski, unpublished information), thymic DCs may be supplied continuously in situ with TRAs via fragments of apoptotic mTECs. Second, DCs might acquire cellular material from viable

ANRV270-IY24-19

ARI

23 February 2006

15:27

Deletion MHC class I or II

mTEC

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MHC class II

Transfer of tissuespecific antigens

DC

TCR Developing T cell

Treg

Selection of Tregs

Potential mechanisms of antigen spreading: • Exosome transfer

• Uptake of apoptotic material

• “Nibbling”

• Gap junctions

Figure 6 Recessive and dominant mechanisms cooperatively mediate tolerance toward TRAs. Various experimental models suggest a division of labor during central tolerance induction, in that Treg induction is efﬁciently mediated by TECs, whereas DCs are specialized for T cell deletion (reviewed in 176). Constitutive cross-presentation by thymic DCs ensures that both tolerance modes operate concomitantly for TRAs, most of which are restricted in their expression to mTECs. Note that the apparent redundancy of antigen presentation is limited by the unidirectional ﬂow of antigenic material from mTECs to DCs. Four possible routes of intercellular transfer of antigens are indicated. Uptake of apoptotic fragments from neighbor cells has been demonstrated for lymph node DCs. Antigen transfer via exosomes, capture of small portions of cytoplasm/membrane (“nibbling”), or gap junctions has been reported in particular systems in vitro, but their relevance in vivo remains to be established.

mTECs via either secreted exosomes (111) or a less well-deﬁned process referred to as nibbling (112). Indeed, intercellular transfer of MHC-bound peptides from TECs to DCs has been reported (113–115; B. Kyewski, unpublished information). Third, antigen has recently been found to be transferred between living cells through gap junctions, whereby transferred antigenic determinants appear to be restricted to the size range of processed peptides (116). Such transfer of antigenic de-

terminants via gap junctions would also open the possibility of antigen spreading among mTECs, which are notoriously inefﬁcient in capturing extracellular antigen (65). Taken together, the deletional mode of tolerance-induction through expression of TRAs in mTECs has been put on a ﬁrm experimental basis in several model systems, and cross-presentation of TRAs by DCs contributes to this process. Further support for a role for deletion in conferring tolerance www.annualreviews.org • Central T Cell Tolerance

585

ANRV270-IY24-19

ARI

23 February 2006

15:27

to TRAs expressed by mTECs comes from studies with AIRE-deﬁcient mice. Thus, Aire deﬁciency resulted in impaired deletion of speciﬁc CD4 T cells in the RIP-hen-egg lysozyme (HEL)/3A9 TCR double transgenic model of negative selection (82, 117), and similar observations were also reported for the RIP-Ova/OTII system (82a, 118). Unexpectedly, in the latter model, expression of ovalbumin under control of the insulin promoter in mTECs was not affected by the absence of AIRE (in contrast to expression of the endogenous insulin gene), suggesting a role for AIRE that extends beyond that of a regulator of pGE. Indeed, puriﬁed AIREdeﬁcient mTECs were found to be less efﬁcient in presenting endogenous proteins or exogenous peptides to T cells in vitro. This observation is currently not understood and may point to additional functions of Aire. Thus, it has been reported that Aire has E3 ligase activity (118a); this contention, however, has subsequently been questioned (118b).

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Treg(s): T regulatory cell(s)

PROMISCUOUS GENE EXPRESSION AND INDUCTION OF REGULATORY T CELLS Soon after the initial description of the immune regulatory function of CD4+ CD25+ suppressor or regulatory T cells (Tregs), an essential role of thymic selection processes in their generation was proposed (119). A consensus now holds that the majority of these cells represent a separate, thymus-derived lineage of CD4 T cells (120). A number of TCR transgenic systems in which CD25+ Treg cells with known speciﬁcity for neo-selfantigens are generated in the thymus strongly supports an essential role of thymic epithelium for Treg development. Jordan and colleagues found that inﬂuenza hemagglutinin (HA)-speciﬁc CD4 T cells were selected into the CD25+ lineage when the agonist ligand was expressed under the control of a ubiquitous promoter (121). Similar observations were subsequently reported in other mod586

Kyewski

·

Klein

els using either the same transgenic TCR (122, 123) or the ovalbumin-speciﬁc DO11.10 TCR (124, 125). In the two HA-speciﬁc systems, the role of thymic epithelium was rigorously tested and conﬁrmed, because transplantation of thymi from antigen-transgenic mice into otherwise antigen-deﬁcient hosts was sufﬁcient to direct the selection of a large fraction of speciﬁc CD4 T cells into the Treg lineage (Figure 6). In apparent contrast to the conclusion drawn from murine models, a role for human thymic DCs in inducing Tregs has been proposed based on in vitro studies (125a). Does promiscuous expression of TRAs in mTECs contribute to the generation of a TRA-speciﬁc Treg repertoire? This tempting assumption is intimately related to the spatiotemporal regulation of the induction and selection of the Treg lineage repertoire, as it would imply a medullary origin of Tregs. By contrast, evidence for a cortical origin of the Treg lineage has been obtained in transgenic mice that purportedly express MHC class II molecules only on cTECs under control of the Keratin 14 promoter (K14-IAb ) (126). In these mice, the fraction of CD4 SP thymocytes that express the CD25 marker was similar to that observed in wild-type controls (127). More important, the peripheral repertoire of CD4 T cells in these mice contained about 10% of CD25+ T cells that displayed normal suppressive function in a standard in vitro assay, and these were able to prevent wasting disease in the adoptive transfer model of colitis in vivo. At face value, this model system indicates that MHC class II expression by cTECs is sufﬁcient to generate a fully functional suppressor T cell repertoire, whereas self-antigen encounter in the medulla appears dispensable. However, an important caveat of the K14 model is that the exclusive restriction of MHC class II expression to cTECs has not been rigorously veriﬁed by functional assays, e.g., testing the capacity of isolated mTECs from K14IAb mice to present MHC class II–restricted antigens in vitro. In light of the propensity of mTECs to express otherwise tissue-speciﬁc

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

molecules in a promiscuous fashion and the fact that K14 is a marker for mTECs, it remains questionable whether mTECs are indeed devoid of any MHC class II molecules in these mice. The timing and location of suppressor lineage commitment within the thymic microenvironment has recently been analyzed using a reporter mouse that expresses a fusion protein of FoxP3 and green ﬂuorescent protein. FoxP3-reporter expression could be detected in thymocytes within medullary areas but not in the cortex (120, 127a). If Treg lineage commitment is promptly followed by FoxP3 induction, lineage induction would thus appear to occur after transition into the medulla. However, depending on the temporal gap between delivery of the lineage induction signal and the appearance of FoxP3, these data are also compatible with a cortical origin of the Treg lineage. In support of a role of mTECs in the selection of Tregs, a number of targeted mutations that selectively disturb the medullary architecture and mTEC differentiation, while apparently preserving an intact cortical microenvironment, lead to greatly reduced numbers of CD25+ Tregs (128–130). A putative contribution of mTECs to Treg induction does not necessarily imply a particular qualitative property of this cell type. Extrathymic antigen encounter can induce the de novo generation of Treg from mature, naive CD4 T cells under certain circumstances (131–133a). Thus, unlike other branch points of hematopoietic lineage commitment, there appear to be a least two windows of opportunity (intra- and extrathymic) within which Treg differentiation can occur. This plasticity of lineage commitment may well already apply to selection processes in the thymus, in that antigen encounter on stromal cells in the medulla may induce a “second wave” of intrathymic suppressor T cells from (semi)mature thymocytes under conditions that may be akin to peripheral induction of Tregs.

AUTOIMMUNITY AS A CONSEQUENCE OF SUBVERTED CENTRAL TOLERANCE: LESSONS FROM ANIMAL MODELS Different genetic models in mice clearly show that pGE is indeed essential for self-tolerance, whereby compromised pGE is invariably associated with elevated susceptibility to autoimmunity. Tolerance induction by expression of TRAs in mTECs can be affected at several levels, such as defective development or spatial organization of mTECs or through quantitative or qualitative alterations in pGE in an otherwise unperturbed medullary microenvironment. In a number of gene-targeted mice or natural mutant strains, autoimmunity correlates with perturbed integrity of the medullary microenvironment. For instance, lymphotoxin β receptor (LTβR)–deﬁcient mice exhibit autoimmune symptoms that have been attributed to defects in mTECs (66, 134). However, the two published reports reach conﬂicting conclusions as to the nature of the underlying defect. Boehm and colleagues concluded that reduced numbers of otherwise normal mTECs (with respect to expression of TRAs) and a perturbed medullary architecture are responsible for inefﬁcient tolerance induction (66). By contrast, Chin et al. claimed that mTEC development and medullary architecture were unaffected by lack of lymphotoxin signaling. Rather, these authors suggested that expression of Aire, and consequently a number of TRAs, was downregulated in the absence of lymphotoxinmediated cross-talk between thymocytes and mTECs (134). A related study reported that aly/aly mice, carrying a natural mutation of the NF-κBinducing kinase (NIK) gene that acts downstream of the LTβR, develop a phenotype very similar to that of RelB-deﬁcient mice, i.e., multi-inﬂammatory lesions, reduced Aire expression, a disorganized cortico-medullary junction, and a strong reduction of UEA-1+

www.annualreviews.org • Central T Cell Tolerance

FoxP3: forkhead box family transcription factor 3 LTβR: lymphotoxin beta receptor UEA-1: Ulex europaeus 1 lectin

587

ARI

23 February 2006

15:27

mTECs (69, 130, 135–137). Similar observations were recently reported for mice with a targeted disruption of the tumor necrosis factor receptor-associated factor (TRAF) 6 gene, which also acts upstream of NF-κB, without directly intersecting with lymphotoxin signaling (129). Although the exact molecular and cellular consequences of these genetic lesions, e.g., impaired differentiation and/or proliferation of mTECs versus reduction of AIRE and TRA expression in otherwise normally differentiated mTECs, remain to be clariﬁed, NF-κBmediated regulation of the thymic microenvironment clearly is critical for efﬁcient induction of central tolerance. It will be informative to assess in more detail which mode of tolerance induction is affected in the respective mutants, i.e., negative selection and/or induction of Tregs. pGE can also modulate the susceptibility to autoimmunity through quantitative variations in TRA expression. Thus, natural allelic or strain-dependent variations in the intrathymic expression levels of insulin (76, 138–140), certain ocular antigens (141, 142), or myelin basic protein (MBP) (143) were found to correlate with the susceptibility to diabetes, experimental autoimmune uveitis (EAU), or experimental autoimmune encephalomyelitis (EAE), respectively. Finally, qualitative features of intrathymic TRA expression can result in a difference in self-epitope display between thymic and peripheral APCs and thus be a critical determinant of tolerance induction. One such example is the differential splicing of proteolipid protein (PLP) by mTECs and oligodendrocytes that explained the exquisite susceptibility of SJL mice to the induction of EAE by immunization with PLP (144, 145). While PLP is expressed predominantly in its full-length form in the CNS, expression in thymic epithelium is largely restricted to a shorter variant (DM20) lacking a portion of 35 amino acids, the so-called PLP loop. As a consequence, this segment of the molecule can be classiﬁed as a truly peripheral antigen and is therefore ex-

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

588

Kyewski

·

Klein

cluded from central tolerance induction. As this pattern of expression is conserved among SJL mice and several EAE-resistant mouse strains, it cannot by itself account for the discrepancy in EAE susceptibility. Rather, the MHC haplotype-speciﬁc choice of an I-As restricted epitope within the PLP loop renders SJL mice susceptible, whereas the immune system of EAE-resistant BL/6 mice is intrinsically unresponsive to the PLP loop, as no I-Ab -restricted epitopes are generated from this amino acid stretch. Together, the EAE model exempliﬁes some of the complexities and potential pitfalls inherent in central tolerance toward peripheral antigens. Thus, several qualitative features of gene expression in the thymus beyond mere variations in the level of auto-antigen expression, such as alternative splicing or isoform expression (146– 149), incomplete assembly of multimeric receptors (149a), RNA-editing, glycosylation (150), and other posttranslational modiﬁcations, need to be considered as factors potentially predisposing to autoimmunity.

LINKING PROMISCUOUS GENE EXPRESSION TO HUMAN AUTOIMMUNITY As discussed in the previous section, there are several possibilities by which inherent pitfalls, genetic polymorphisms, and spontaneous mutations can compromise tolerance induction through intrathymic expression of TRAs. Depending on its severity, such a defect may either lead to spontaneous autoimmunity or only predispose to autoimmunity, which can be detected under certain experimental conditions. How relevant are these insights from animal models for our understanding of the etiology and pathogenesis of human autoimmune diseases? The resemblance of the autoimmune syndrome observed in Aire knockout mice with the rare human autoimmune disease APS-1 provides an instructive example of the validity of animal models for human diseases and the opportunity to link a host of experimental ﬁndings to the human

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

situation. APS-1 is a rare monogenic, autosomal recessive autoimmune syndrome with a predilection for endocrine organs (10, 77, 118, 151). It is invariably linked to mutations in the Aire gene (152) and is prevalent among certain ethnic groups such as regional populations in Finland and Sardinia as well as in Iranian Jews. The severity of the pathology and the range of affected organs vary at the population and at the individual levels, which indicates a modifying inﬂuence of background genes and/or environmental factors (152a). There is a strong conservation in pGE between mouse and human, including a collection of TRAs expressed in organs that are preferentially afﬂicted in APS-1 patients and recognized by prevalent auto-antibodies (153). In addition, the expression pattern of Aire is very similar in the human and mouse thymus, with strong expression in mTECs. Hence, in analogy to mice, the pathophysiology of APS-1 is currently best explained by the reduced expression of the Aire-dependent set of TRAs in the thymus that results in a corresponding lower threshold or lack of central tolerance. However, direct proof of this contention is still lacking. Also lacking is an explanation for the predilection of certain endocrine organs to autoimmune attack in humans, which is not observed to the same extent in mice. The relative dependence on Aire for intrathymic transcription of essential auto-antigens is unlikely to be the only determinant for targeting a particular organ. Thus, insulin is assumed to be the inciting auto-antigen in T1DM (154–156). Its expression is strictly Aire dependent in mice and its levels in humans also closely correlate with those of Aire (158; B. Kyewski, unpublished information), yet T1DM does not feature prominently among the immune pathology in APS-1 patients (151) or in Aire knockout mice (77, 82, 152a). Aire mutations exemplify an informative but rare simple genetic trait with high penetrance. Unlike APS-1, most common organrestricted autoimmune diseases are caused by

polygenic traits involving multiple susceptibility alleles of low penetrance, each contributing a minor risk to disease development. The identiﬁcation of such low-penetrance alleles and proof of their causal contribution to the disease process has been demanding. Moreover, certain immunological pathways are shared by multiple autoimmune diseases, whereas others are speciﬁc to a particular disease (157). The regulation of pGE seems to inﬂuence multiple autoimmune diseases, as exempliﬁed by contributing low-risk alleles in the case of two common organ-restricted autoimmune diseases, namely T1DM and MG. In T1DM, the IDDM2 (insulin-dependent diabetes mellitus 2) locus shows the highest correlation with disease susceptibility of all identiﬁed disease-associated non-MHC loci. This trait reﬂects a polymorphism in Variable Number of Tandem Repeats (VNTRs) upstream of the promotor region of the insulin gene. Whereas the class III alleles with high copy numbers of repeats (>120) are protective, the low copy number class I alleles (<60) are nonprotective (138, 139, 158). These VNTR alleles correlate with the level of insulin transcription in the thymus. The protective class III allele confers three- to fourfold higher expression levels than the class I allele. This important ﬁnding was originally deduced from the analysis of allele-speciﬁc transcripts in class I/III heterozygotes and has recently been conﬁrmed by quantitative PCR analysis of insulin expression in puriﬁed human mTECs (B. Kyewski, unpublished information). In the same vein, the relative expression levels of the α-chain of the acetylcholine receptor in mTECs correlates with a single nucleotide polymorphism (SNP) in its promotor region and this in turn correlates with the age of onset of MG (H.J. Garchon & B. Kyewski, unpublished information). The relative expression levels of both self-antigens—insulin and α-chain of the acetylcholine receptor—in different individuals correlate with the respective Aire levels. This correlation is strong in the case of the protective and weak in the case of the nonprotective allele. www.annualreviews.org • Central T Cell Tolerance

589

ANRV270-IY24-19

ARI

23 February 2006

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MBP: myelin basic protein

590

15:27

How genetic polymorphisms in the regulatory regions of TRA genes might modulate the intrathymic expression levels of the respective gene remains elusive. They might affect either the access or the DNA binding of transcription (co)factors such as Aire. This in turn shifts the threshold of selftolerance to an extent that translates into signiﬁcant alterations of disease susceptibility on the appropriate genetic background. The surprising ﬁnding in this context is the exquisite sensitivity of the central tolerance process to moderate quantitative variations in TRA expression. Thus, subtle differences in the range of two- to fourfold in intrathymic expression of auto-antigens as obtained in mice with deﬁned copy numbers of pancreatic or nervous system–speciﬁc antigens can modulate the susceptibility to autoimmunity (76, 140, 159). The threshold range for self-antigens in humans is less well deﬁned. Carriers of a single allele of an Aire mutation do not develop APS-1 (151; O. Winqvist, personal communication). Assuming a direct gene-dosage effect between Aire and target gene expression levels, as suggested by ﬁndings in the Aire knockout mouse model (82), minor variations obviously do not predispose to disease in a polygenic setting in humans. Nevertheless, the examples cited above—insulin and the acetylcholine receptor—indicate that there is only limited buffering of the tolerance threshold in humans on a susceptible background. Remarkably, interindividual variations in gene expression levels of several human auto-antigens range between 2to more than 30-fold (158, 160; B. Kyewski, unpublished information). The signiﬁcance of these variations for setting the threshold of self-tolerance in each individual on a highly heterogeneous genetic background remains to be determined. While the examples discussed above illustrate how genetic polymorphisms in the target gene locus inﬂuence expression levels of auto-antigens, pGE also affords new insights into how MHC polymorphisms inﬂuence disease susceptibility. Qualitative differences in Kyewski

·

Klein

transcription and posttranslational modiﬁcations between mTECs and the respective tissue cells and peripheral APCs could result in a discrepancy in self-epitope display between the thymus and the periphery. Central tolerance might thus miss certain self-epitopes. These epitope-speciﬁc holes in self-tolerance would obviously pose a danger only in conjunction with MHC haplotypes to which these particular epitopes bind. By inference from the mouse models discussed in the previous section, this may be the case for certain MHC class II alleles associated with susceptibility to MS and epitopes within the loop region of PLP (144). Pitfalls inherent to tolerance induction through pGE may also undermine selftolerance to another presumptive target antigen in MS, namely MBP. During ontogeny, oligodendrocytes ﬁrst express the fetal form termed golli-MBP, which postnatally is replaced by the adult form, termed classic MBP. Both isoforms differ in their transcriptional start sites, whereby sequences of classic MBP are appended to the golli-speciﬁc transcript (161). Due to the uncoupling of pGE from developmental regulation of gene expression in speciﬁc tissues, the switch from golli-MBP to classic MBP does not occur in mTECs. Instead, golli-MBP remains the predominant transcript throughout adulthood, possibly favoring self-tolerance to epitopes present in this isoform (B. Kyewski, unpublished information). Noncensored T cells speciﬁc for epitopes encoded only by the full-length form of PLP or classic MBP, respectively, may thus contribute to the initiation and maintenance of the early phase of MS. As evident from the LTβR- and TRAF6deﬁcient mouse strains, a disorganized microenvironment and perturbed homeostasis of mTEC development also indirectly favor autoimmunity (66, 129). This might also be a factor contributing to the intriguing association of diverse autoimmune symptoms such as MG with human thymomas. These malignant transformations of TECs lead to a profound

ANRV270-IY24-19

ARI

23 February 2006

15:27

disruption of the thymic microenvironment (162, 163).

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

TOLERANCE TO “LATE” SELF-ANTIGENS The bulk of self-antigens become available to the developing mammalian immune system during the late fetal and early postnatal period when organogenesis is being completed. There are, however, notable exceptions. During puberty self-antigens are newly expressed or upregulated in the male and female reproductive tract and the mammary gland (for examples, see 7, 55). It has always been a conundrum how the fully developed repertoire of effector T cells copes with this “new” set of self-constituents. In male germ cells, sequestration from circulating lymphocytes and the lack of MHC expression afford a degree of protection from auto-aggression (“immune privilege”). Tolerance is further backed up by the local expression of the apoptosisinducing CD95 ligand (“T cell counterattack”). Owing to the abolition of developmental gene regulation, mTECs also express a range of male germ cell–associated self-antigens, thus presumably contributing to self-tolerance. Breaking self-tolerance to male germ cells is one cause of male infertility (164, 165). The observation that 14% of male APS-1 patients show gonadal failure clearly supports this supposition (151). As a corollary, tolerance of females toward nonY chromosome–encoded male antigens may also play a role in curbing an immune response against sperm cells, which may interfere with successful conception (166). pGE also contributes to tolerance to selfantigens expressed by the female reproductive tract and thus to fertility. Oophoritis can be observed after d3 thymectomy of Balb/c mice (167) and in mutant strains in which mTEC development is impeded. Likewise, over 50% of female APS-1 patients are infertile and display auto-antibodies against granulose/thecal cells of the Grafﬁan follicles and the corpus luteum (168, 169).

Pregnancy represents yet another situation during which the established T cell repertoire is confronted with a qualitative and quantitative change in the display of self-antigens. These include antigens expressed by the placenta (i.e., HLA-G) or the lactating mammary gland (e.g., milk proteins such as casein). Moreover, cells of fetal origin leak into the maternal circulation (re-)exposing the maternal immune system to fetal antigens such as the γ-chain of the acetylcholine receptor or certain male antigens. Indeed, APS-1 patients harbor antibodies against, for instance, the syncytio-trophoblast of the placenta (168). Promiscuous expression of members of these different categories of pregnancy-associated antigens (55) should curb the danger of a pathogenic immune response as observed in spontaneous abortions (170). The identiﬁcation of several sex-dependent self-antigens, which are equally expressed in mTECs of males and females and thus transgress sexual identity, should allow their tolerogenic potential to be assessed.

TUMOR IMMUNITY AND CENTRAL TOLERANCE Depending on the nature of the target antigens, the immune response to tumors can be categorized either as a response directed against self or foreign (171). The foreign category will include tumor antigens arising from genetic alterations unique to the tumor, i.e., point mutations or chromosomal rearrangements generating new T and B cell epitopes. The immune response against such foreign epitopes will likely follow rules established for pathogen-driven immunity and thus is not further considered here. In contrast, the immune response directed against socalled tumor-associated self-antigens, which are induced or upregulated in tumor cells, should be quenched by central or peripheral self-tolerance imposition. Numerous studies in animal models and more recent results from clinical immunotherapy trials conﬁrm this prediction (172, 173, 173a). pGE adds an www.annualreviews.org • Central T Cell Tolerance

591

ARI

23 February 2006

15:27

additional twist to this issue by demonstrating that several categories of tumor-associated self-antigens are reproducibly detectable in mouse or human mTECs (55, 160): differentiation antigens (e.g., MART1 or tyrosinase), cancer germ cell antigens (e.g., MAGE family proteins), and onco-fetal antigens (e.g., CEA family members). Because of the spatially and temporally restricted expression patterns of these antigens, they were considered preferred targets for active vaccination against various nonhematopoietic tumors, foremost melanomas, since it was assumed that speciﬁc T cells are not subject to tolerance induction (174, 175). This assumption, however, can no longer be upheld (for discussion, see 176). For example, expression of the human CEA antigen in mTECs of transgenic mice has been shown to induce central tolerance (177). As a result, only T cells of low avidity escape tolerance induction, and this blunted repertoire is incapable of eradicating CEA-expressing mouse tumor cells even after repeated immunization. This is not to say, however, that such a compromised T cell repertoire, when appropriately recruited, might not be effective as an adjuvant tumor therapy. As noted above, central tolerance toward promiscuously expressed genes may also include the induction of antigen-speciﬁc Tregs, and this should also apply to tumor-associated self-antigens. Hence, the speciﬁc elimination of Tregs, preferably antigen-speciﬁc ones, holds great promise to optimally recruit the available T cell repertoire. The feasibility of this strategy has been directly demonstrated in animal models (178–180) and can be inferred from recent clinical trials in humans (181). An additional point to consider pertains to the emerging perception that intrathymic levels of self-antigens, determining the threshold of self-tolerance, may vary considerably between individuals (160). A systematic analysis of the expression patterns and levels of nonmutated tumor-associated antigens in puriﬁed human mTECs could thus guide the choice of suitable candidates for future clinical trials (182).

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

592

Kyewski

·

Klein

ON THE PHYLOGENY OF PROMISCUOUS GENE EXPRESSION The evolution of the immune system can be divided into two periods either predating or following the advent of lymphocyte receptor generation through somatic gene rearrangement. This transition from invertebrate immune systems that solely rely on innate mechanisms to vertebrate immune systems that combine innate and adaptive arms of immunity occurred about 450–500 million years ago. Self/nonself discrimination is brought about in fundamentally different ways by both arms of the immune system. In innate immunity, a limited set of germ line–encoded receptors speciﬁcally recognizes pathogen-derived components. Matching of appropriate receptors with ligands occurred over many generations by classical Darwinian selection. In contrast, the adaptive immune system is characterized by the generation of a highly diverse set of antigen-speciﬁc receptors that are somatically generated in each individual by gene rearrangement. Potential autoreactivity is an inherent consequence of such random generation of millions of antigen-speciﬁc receptors. The ability to recognize a vast array of pathogen-derived structures with a high degree of speciﬁcity is inevitably coupled to potential recognition of a multitude of selfantigens with equal precision. Hence, with the emergence of a diverse T cell repertoire there must have been pressure to select for a mechanism allowing for the display of self-ligands serving as tolerogenic substrate for the selection of antigen receptor-expressing cells. Provision of a structure allowing somatic generation and selection of a diverse T cell repertoire might have been the driving force for the development of the thymus as a distinct organ. Indeed, this structure-function relationship is inseparable as far as evolutionary records are concerned. If this scenario is correct, pGE should date back to early vertebrates, a prediction still to be tested. An ortholog of Aire, serving here

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

as a molecular tracer of pGE, has been identiﬁed in zebraﬁsh (N. Trede, personal communication). Furthermore, the identiﬁcation of a zebraﬁsh ortholog of FoxP3 is compatible with the notion that the control of selftolerance through Tregs dates back at least 400 million years (183). The paramount importance of intrathymically imprinted dominant and recessive tolerance mechanisms is further illustrated by the fact that two of the ﬁve known monogenic autoimmune diseases, namely APS-1 and IPEX, affect either pGE or Treg development, respectively (157). Interestingly, the complete lack of the Treg lineage in IPEX leads to a more severe phenotype than the partial defect in pGE resulting from AIRE-deﬁciency in humans (184) and mice (185). An intriguing question in the context of reconstructing the evolution of the adaptive immune system is the identiﬁcation of molecules, signaling pathways, and regulatory mechanisms that already existed in invertebrates and that may have been adopted by vertebrates. Although gene rearrangement was clearly novel, whether this is also true for pGE remains open. The domain composition of the Aire molecule including putative DNAbinding domains and nuclear localization signals is typical for transcriptional regulators, and the evolutionary forerunners of Aire may have served and still may serve another role in gene regulation or other functions (82a, 118a). Aire is expressed in many different cell types in mice and humans in addition to mTECs, albeit at much lower levels (186, 187). Yet, none of these cells, with the possible exception of male germ cells, exhibits typical pGE. By targeting gene clusters, pGE may have exploited preexisting and/or coevolving colocalization of genes. It is commonly assumed that colocalization is linked to coregulation, e.g., to serve a common biological process such as signal transduction or lineage speciﬁcation. Functionally unrelated tissuerestricted or housekeeping genes may also colocalize to chromosomal domains with a

high density of actively transcribed genes (188). Analysis of a possible link between the evolution of gene order and the selection of promiscuously expressed genes should be possible in the near future as more detailed physical gene maps and cell type–speciﬁc geneexpression proﬁles in different species become available (87). Whatever the answers to these intriguing questions, pGE has clearly been indispensable to the survival of mice and humans and possibly all vertebrates by preventing debilitating or even lethal autoimmunity. The selection pressure to develop and maintain this central tolerance mechanism is best illustrated by the high degree of infertility among female APS1 patients and Aire knockout mice as a result of autoimmune ovarian failure.

IPEX: immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome

EPILOGUE While this review focuses mainly on the unique role of mTECs, there are other more conventional routes by which TRAs may become available for intrathymic selection of the T cell repertoire and thus blur the distinction between central and peripheral tolerance. Thymic and splenic DCs express Aire and TRAs such as insulin at 10- and 100-fold lower levels than mTECs, respectively (11, 68, 189). Whatever the contribution of this hematopoietic pGE to self-tolerance (190), it is insufﬁcient to substitute for pGE by mTECs in animal models (77, 82). It has also been proposed that DCs might pick up TRAs in draining lymph nodes, which are continuously supplied by a ﬂux of tissue-derived material due to physiological cell turnover (191, 192). These antigen-laden DCs may then trafﬁc to the thymic medulla and display TRAs for repertoire selection. The tolerogenic potential of this route could conceivably vary, depending on the ontogeny, turnover rate, and size of the respective tissue and the sampling rate of various peripheral lymph nodes. A convincing demonstration of continuous DC trafﬁcking from lymph nodes to the thymus is still lacking, however. www.annualreviews.org • Central T Cell Tolerance

593

ANRV270-IY24-19

ARI

23 February 2006

15:27

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

While keeping an open mind on a role of these alternative routes, we feel justiﬁed in stating that the expression of TRAs representing virtually all the body’s tissues by mTECs enlarges the scope of central tolerance in a manner not anticipated by original concepts of self/nonself discrimination. The implications go beyond our understanding of what we consider typical autoimmune diseases, impinging on biological processes as diverse as fertility, pregnancy, puberty, and tumor de-

fense. A major challenge is identifying the cellular and molecular strategies that allow for mimicking the transcriptome and possibly epigenome of most cell lineages. Progress in this area is expected to contribute new insights into the complex genetic regulation of most common autoimmune diseases. Ultimately, it would be most gratifying if elucidation of this seemingly arcane phenomenon opened up new and successful strategies to prevent or even cure autoimmune diseases.

SUMMARY POINTS 1. Promiscuous expression of otherwise strictly TRAs in mTECs extends the scope of central tolerance to essentially all tissues. 2. MTECs are heterogenous; a subset with high expression of MHC class II and costimulatory molecules displays the strongest and most diverse expression of TRAs. 3. pGE is regulated at multiple levels; Aire is the only molecular determinant identiﬁed to date. 4. Expression of TRAs in mTECs contributes to negative selection of self-reactive T cells, through direct antigen presentation by mTECs and cross-presentation by thymic DCs. 5. Perturbation of pGE through genetic lesions that either directly affect pGE or indirectly affect the integrity of the mTEC microenvironment can lead to autoimmunity, thereby documenting the essential, nonredundant role of mTECs in self-tolerance.

FUTURE ISSUES TO BE RESOLVED 1. A clearer understanding of the developmental biology of TECs and the precursor/product relationships among TEC subsets is an essential prerequisite for a molecular understanding of pGE. 2. Additional transcriptional regulators and epigenetic mechanisms involved in pGE await identiﬁcation. 3. Pathways of antigen presentation by mTECs are only poorly understood, e.g., deletion of CD4 SP thymocytes by mTECs implies the existence of an efﬁcient endogenous MHC class II loading pathway. 4. A possible role of pGE in the generation of a TRA-speciﬁc repertoire of Tregs needs to be clariﬁed. 5. The role of pGE in common and rare human autoimmune diseases needs to be identiﬁed and speciﬁed; genetic polymorphisms versus inherent pitfalls must be delineated.

594

Kyewski

·

Klein

ANRV270-IY24-19

ARI

23 February 2006

15:27

ACKNOWLEDGMENTS

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B.K. is supported by the German Cancer Research Center and the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 405). L.K.’s research at the Research Institute of Molecular Pathology (funded by Boehringer Ingelheim) is supported by the Austrian National Science Fund (grant Z58-B01 and Sonderforschungsbereich F023). Both authors receive funds from the European Union (FP6 Integrated Project Eurothymaide). We thank all members of our labs for many critical and constructive discussions and suggestions. Special thanks to S. ¨ for providing gene expression Jonnakuty and B. Brors for bioinformatic data analysis, G. Schutz data, J. Derbinski for critical reading, and J. Arnold, J. G¨abler, and C. Koble for help with the preparation of ﬁgures.

LITERATURE CITED 1. Starr TK, Jameson SC, Hogquist KA. 2003. Positive and negative selection of T cells. Annu. Rev. Immunol. 21:139–76 2. Palmer E. 2003. Negative selection—clearing out the bad apples from the T-cell repertoire. Nat. Rev. Immunol. 3:383–91 3. Alferink J, Aigner S, Reibke R, Hammerling GJ, Arnold B. 1999. Peripheral T-cell tolerance: the contribution of permissive T-cell migration into parenchymal tissues of the neonate. Immunol. Rev. 169:255–61 4. Walker LS, Abbas AK. 2002. The enemy within: keeping self-reactive T cells at bay in the periphery. Nat. Rev. Immunol. 2:11–19 5. Shevach EM. 2002. CD4+ CD25+ suppressor T cells: more questions than answers. Nat. Rev. Immunol. 2:389–400 6. Sakaguchi S. 2004. Naturally arising CD4+ regulatory t cells for immunologic selftolerance and negative control of immune responses. Annu. Rev. Immunol. 22:531–62 7. Derbinski J, Gabler J, Brors B, Tierling S, Jonnakuty S, et al. 2005. Promiscuous gene expression in thymic epithelial cells is regulated at multiple levels. J. Exp. Med. 202:33–45 8. Klein L, Kyewski B. 2000. Self-antigen presentation by thymic stromal cells: a subtle division of labor. Curr. Opin. Immunol. 12:179–86 9. Mathis D, Benoist C. 2004. Back to central tolerance. Immunity 20:509–16 10. Villasenor J, Benoist C, Mathis D. 2005. AIRE and APECED: molecular insights into an autoimmune disease. Immunol. Rev. 204:156–64 11. Derbinski J, Schulte A, Kyewski B, Klein L. 2001. Promiscuous gene expression in medullary thymic epithelial cells mirrors the peripheral self. Nat. Immunol. 2:1032–39 12. Shortman K, Egerton M, Spangrude GJ, Scollay R. 1990. The generation and fate of thymocytes. Semin. Immunol. 2:3–12 13. Foss DL, Donskoy E, Goldschneider I. 2001. The importation of hematogenous precursors by the thymus is a gated phenomenon in normal adult mice. J. Exp. Med. 193:365–74 14. Porritt HE, Gordon K, Petrie HT. 2003. Kinetics of steady-state differentiation and mapping of intrathymic-signaling environments by stem cell transplantation in nonirradiated mice. J. Exp. Med. 198:957–62 15. von Boehmer H, Aifantis I, Gounari F, Azogui O, Haughn L, et al. 2003. Thymic selection revisited: how essential is it?Immunol. Rev. 191:62–78 16. Surh CD, Sprent J. 1994. T-cell apoptosis detected in situ during positive and negative selection in the thymus. Nature 372:100–3 17. Zerrahn J, Held W, Raulet DH. 1997. The MHC reactivity of the T cell repertoire prior to positive and negative selection. Cell 88:627–36 www.annualreviews.org • Central T Cell Tolerance

595

ARI

23 February 2006

15:27

18. Merkenschlager M, Graf D, Lovatt M, Bommhardt U, Zamoyska R, et al. 1997. How many thymocytes audition for selection? J. Exp. Med. 186:1149–58 19. Blackman M, Yague J, Kubo R, Gay D, Coleclough C, et al. 1986. The T cell repertoire may be biased in favor of MHC recognition. Cell 47:349–57 20. Jerne NK. 1971. The somatic generation of immune recognition. Eur. J. Immunol. 1:1–9 21. Huseby ES, White J, Crawford F, Vass T, Becker D, et al. How the T cell repertoire becomes peptide and MHC speciﬁc. Cell 122:247–60 22. Petrie HT, Livak F, Schatz DG, Strasser A, Crispe IN, et al. 1993. Multiple rearrangements in T cell receptor alpha chain genes maximize the production of useful thymocytes. J. Exp. Med. 178:615–22 23. Shortman K, Vremec D, Egerton M. 1991. The kinetics of T cell antigen receptor expression by subgroups of CD4+ 8+ thymocytes: delineation of CD4+ 8+ 3(2+ ) thymocytes as post-selection intermediates leading to mature T cells. J. Exp. Med. 173:323–32 24. Huesmann M, Scott B, Kisielow P, von Boehmer H. 1991. Kinetics and efﬁcacy of positive selection in the thymus of normal and T cell receptor transgenic mice. Cell 66:533–40 25. Ignatowicz L, Kappler J, Marrack P. 1996. The repertoire of T cells shaped by a single MHC/peptide ligand. Cell 84:521–29 26. Fukui Y, Ishimoto T, Utsuyama M, Gyotoku T, Koga T, et al. 1997. Positive and negative CD4+ thymocyte selection by a single MHC class II/peptide ligand affected by its expression level in the thymus. Immunity 6:401–10 27. van Meerwijk JP, Marguerat S, Lees RK, Germain RN, Fowlkes BJ, et al. 1997. Quantitative impact of thymic clonal deletion on the T cell repertoire. J. Exp. Med. 185:377–83 28. Tourne S, Miyazaki T, Oxenius A, Klein L, Fehr T, et al. 1997. Selection of a broad repertoire of CD4+ T cells in H-2Ma0/0 mice. Immunity 7:187–95 29. Anderson G, Partington KM, Jenkinson EJ. 1998. Differential effects of peptide diversity and stromal cell type in positive and negative selection in the thymus. J. Immunol. 161:6599–603 30. Pircher H, Brduscha K, Steinhoff U, Kasai M, Mizuochi T, et al. 1993. Tolerance induction by clonal deletion of CD4+ 8+ thymocytes in vitro does not require dedicated antigen-presenting cells. Eur. J. Immunol. 23:669–74 31. Spain LM, Berg LJ. 1994. Quantitative analysis of the efﬁciency of clonal deletion in the thymus. Dev. Immunol. 4:43–53 32. Volkmann A, Zal T, Stockinger B. 1997. Antigen-presenting cells in the thymus that can negatively select MHC class II-restricted T cells recognizing a circulating self antigen. J. Immunol. 158:693–706 33. Baldwin KK, Trenchak BP, Altman JD, Davis MM. 1999. Negative selection of T cells occurs throughout thymic development. J. Immunol. 163:689–98 33a. Hogquist KA, Baldwin TA, Jameson SC. 2005. Central tolerance: learning self-control in the thymus. Nat. Rev. Immunol. 5:772–82 34. Davey GM, Schober SL, Endrizzi BT, Dutcher AK, Jameson SC, et al. 1998. Preselection thymocytes are more sensitive to T cell receptor stimulation than mature T cells. J. Exp. Med. 188:1867–74 35. Matzinger P. 1994. Tolerance, danger, and the extended family. Annu. Rev. Immunol. 12:991–1045 36. Matzinger P. 2002. The danger model: a renewed sense of self. Science 296:301–5 37. Wekerle H, Bradl M, Linington C, Kaab G, Kojima K. 1996. The shaping of the brainspeciﬁc T lymphocyte repertoire in the thymus. Immunol. Rev. 149:231–43

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

596

Kyewski

·

Klein

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

38. Kuchroo VK, Anderson AC, Waldner H, Munder M, Bettelli E, et al. 2002. T cell response in experimental autoimmune encephalomyelitis (EAE): role of self and crossreactive antigens in shaping, tuning, and regulating the autopathogenic T cell repertoire. Annu. Rev. Immunol. 20:101–23 39. Linsk R, Gottesman M, Pernis B. 1989. Are tissues a patch quilt of ectopic gene expression?Science 246:261 40. Salaun J, Bandeira A, Khazaal I, Calman F, Coltey M, et al. 1990. Thymic epithelium tolerizes for histocompatibility antigens. Science 247:1471–74 41. Salaun J, Corbel C, Le Douarin NM. 2005. Regulatory T cells in the establishment and maintenance of self-tolerance: role of the thymic epithelium. Int. J. Dev. Biol. 49:137–42 42. Le Douarin N, Corbel C, Bandeira A, Thomas-Vaslin V, Modigliani Y, et al. 1996. Evidence for a thymus-dependent form of tolerance that is not based on elimination or anergy of reactive T cells. Immunol. Rev. 149:35–53 43. Ohki H, Martin C, Corbel C, Coltey M, Le Douarin NM. 1987. Tolerance induced by thymic epithelial grafts in birds. Science 237:1032–35 44. Asano M, Toda M, Sakaguchi N, Sakaguchi S. 1996. Autoimmune disease as a consequence of developmental abnormality of a T cell subpopulation. J. Exp. Med. 184:387–96 45. Antonia SJ, Geiger T, Miller J, Flavell RA. 1995. Mechanisms of immune tolerance induction through the thymic expression of a peripheral tissue-speciﬁc protein. Int. Immunol. 7:715–25 46. Jolicoeur C, Hanahan D, Smith KM. 1994. T-cell tolerance toward a transgenic betacell antigen and transcription of endogenous pancreatic genes in thymus. Proc. Natl. Acad. Sci. USA 91:6707–11 47. Heath WR, Allison J, Hoffmann MW, Schonrich G, Hammerling G, et al. 1992. Autoimmune diabetes as a consequence of locally produced interleukin-2. Nature 359:547– 49 48. von Herrath MG, Dockter J, Oldstone MB. 1994. How virus induces a rapid or slow onset insulin-dependent diabetes mellitus in a transgenic model. Immunity 1:231–42 49. Heath VL, Moore NC, Parnell SM, Mason DW. 1998. Intrathymic expression of genes involved in organ speciﬁc autoimmune disease. J. Autoimmun. 11:309–18 50. Klein L, Klein T, Ruther U, Kyewski B. 1998. CD4 T cell tolerance to human C-reactive protein, an inducible serum protein, is mediated by medullary thymic epithelium. J. Exp. Med. 188:5–16 51. Sospedra M, Ferrer-Francesch X, Dominguez O, Juan M, Foz-Sala M, et al. 1998. Transcription of a broad range of self-antigens in human thymus suggests a role for central mechanisms in tolerance toward peripheral antigens. J. Immunol. 161:5918–29 52. Werdelin O, Cordes U, Jensen T. 1998. Aberrant expression of tissue-speciﬁc proteins in the thymus: a hypothesis for the development of central tolerance. Scand. J. Immunol. 47:95–100 52a. Geenen V, Legros JJ, Franchimont P, Baudrihaye M, Defresne MP, Boniver J. 1986. The neuroendocrine thymus: coexistence of oxytocin and neurophysin in the human thymus. Science 232:508–11 53. Hanahan D. 1998. Peripheral-antigen-expressing cells in thymic medulla: factors in self-tolerance and autoimmunity. Curr. Opin. Immunol. 10:656–62 54. Smith KM, Olson DC, Hirose R, Hanahan D. 1997. Pancreatic gene expression in rare cells of thymic medulla: evidence for functional contribution to T cell tolerance. Int. Immunol. 9:1355–65 www.annualreviews.org • Central T Cell Tolerance

597

ARI

23 February 2006

15:27

55. Gotter J, Brors B, Hergenhahn M, Kyewski B. 2004. Medullary epithelial cells of the human thymus express a highly diverse selection of tissue-speciﬁc genes colocalized in chromosomal clusters. J. Exp. Med. 199:155–66 56. Kyewski B, Derbinski J, Gotter J, Klein L. 2002. Promiscuous gene expression and central T-cell tolerance: more than meets the eye. Trends Immunol. 23:364–71 57. Bruno R, Sabater L, Sospedra M, Ferrer-Francesch X, Escudero D, et al. 2002. Multiple sclerosis candidate autoantigens except myelin oligodendrocyte glycoprotein are transcribed in human thymus. Eur. J. Immunol. 32:2737–47 58. Rodewald HR, Paul S, Haller C, Bluethmann H, Blum C. 2001. Thymus medulla consisting of epithelial islets each derived from a single progenitor. Nature 414:763– 68 59. Bennett AR, Farley A, Blair NF, Gordon J, Sharp L, et al. 2002. Identiﬁcation and characterization of thymic epithelial progenitor cells. Immunity 16:803–14 60. Gill J, Malin M, Hollander GA, Boyd R. 2002. Generation of a complete thymic microenvironment by MTS24+ thymic epithelial cells. Nat. Immunol. 3:635–42 61. Gordon J, Wilson VA, Blair NF, Sheridan J, Farley A, et al. 2004. Functional evidence for a single endodermal origin for the thymic epithelium. Nat. Immunol. 5:546–53 62. Blackburn CC, Manley NR. 2004. Developing a new paradigm for thymus organogenesis. Nat. Rev. Immunol. 4:278–89 63. Surh CD, Gao EK, Kosaka H, Lo D, Ahn C, et al. 1992. Two subsets of epithelial cells in the thymic medulla. J. Exp. Med. 176:495–505 64. Nelson AJ, Hosier S, Brady W, Linsley PS, Farr AG. 1993. Medullary thymic epithelium expresses a ligand for CTLA4 in situ and in vitro. J. Immunol. 151:2453–61 65. Klein L, Roettinger B, Kyewski B. 2001. Sampling of complementing self-antigen pools by thymic stromal cells maximizes the scope of central T cell tolerance. Eur. J. Immunol. 31:2476–86 66. Boehm T, Scheu S, Pfeffer K, Bleul CC. 2003. Thymic medullary epithelial cell differentiation, thymocyte emigration, and the control of autoimmunity require lymphoepithelial cross talk via LTβR. J. Exp. Med. 198:757–69 67. Avichezer D, Grajewski RS, Chan CC, Mattapallil MJ, Silver PB, et al. 2003. An immunologically privileged retinal antigen elicits tolerance: major role for central selection mechanisms. J. Exp. Med. 198:1665–76 68. Pugliese A, Brown D, Garza D, Murchison D, Zeller M, et al. 2001. Self-antigenpresenting cells expressing diabetes-associated autoantigens exist in both thymus and peripheral lymphoid organs. J. Clin. Invest. 107:555–64 69. Zuklys S, Balciunaite G, Agarwal A, Fasler-Kan E, Palmer E, et al. 2000. Normal thymic architecture and negative selection are associated with Aire expression, the gene defective in the autoimmune-polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED). J. Immunol. 165:1976–83 70. Candi E, Schmidt R, Melino G. 2005. The corniﬁed envelope: a model of cell death in the skin. Nat. Rev. Mol. Cell Biol. 6:328–40 71. Langbein L, Pape UF, Grund C, Kuhn C, Praetzel S, et al. 2003. Tight junction-related structures in the absence of a lumen: occludin, claudins and tight junction plaque proteins in densely packed cell formations of stratiﬁed epithelia and squamous cell carcinomas. Eur. J. Cell Biol. 82:385–400 72. Farr AG, Dooley JL, Erickson M. 2002. Organization of thymic medullary epithelial heterogeneity: implications for mechanisms of epithelial differentiation. Immunol. Rev. 189:20–27

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

598

Kyewski

·

Klein

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

73. Gillard GO, Farr AG. 2005. Contrasting models of promiscuous gene expression by thymic epithelium. J. Exp. Med. 202:15–19 74. Akashi K, He X, Chen J, Iwasaki H, Niu C, et al. 2003. Transcriptional accessibility for genes of multiple tissues and hematopoietic lineages is hierarchically controlled during early hematopoiesis. Blood 101:383–89 75. Miyamoto T, Iwasaki H, Reizis B, Ye M, Graf T, et al. 2002. Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment. Dev. Cell 3:137–47 76. Thebault-Baumont K, Dubois-Laforgue D, Krief P, Briand JP, Halbout P, et al. 2003. Acceleration of type 1 diabetes mellitus in proinsulin 2-deﬁcient NOD mice. J. Clin. Invest. 111:851–57 77. Anderson MS, Venanzi ES, Klein L, Chen Z, Berzins SP, et al. 2002. Projection of an immunological self shadow within the thymus by the aire protein. Science 298:1395–401 78. Pitkanen J, Peterson P. 2003. Autoimmune regulator: from loss of function to autoimmunity. Genes Immun. 4:12–21 79. Pitkanen J, Rebane A, Rowell J, Murumagi A, Strobel P, et al. 2005. Cooperative activation of transcription by autoimmune regulator AIRE and CBP. Biochem. Biophys. Res. Commun. 333:944–53 80. Purohit S, Kumar PG, Laloraya M, She JX. 2005. Mapping DNA-binding domains of the autoimmune regulator protein. Biochem. Biophys. Res. Commun. 327:939–44 81. Kumar PG, Laloraya M, Wang CY, Ruan QG, Davoodi-Semiromi A, et al. 2001. The autoimmune regulator (AIRE) is a DNA-binding protein. J. Biol. Chem. 276:41357–64 82. Liston A, Gray DH, Lesage S, Fletcher AL, Wilson J, et al. 2004. Gene dosage–limiting role of Aire in thymic expression, clonal deletion, and organ-speciﬁc autoimmunity. J. Exp. Med. 200:1015–26 82a. Anderson MS, Venanzi ES, Chen Z, Berzins SP, Benoist C, Mathis D. 2005. The cellular mechanism of Aire control of T cell tolerance. Immunity 23:227–39 83. Johnnidis JB, Venanzi ES, Taxman DJ, Ting JP, Benoist CO, et al. 2005. Chromosomal clustering of genes controlled by the aire transcription factor. Proc. Natl. Acad. Sci. USA 102:7233–38 84. Murrell A, Heeson S, Reik W. 2004. Interaction between differentially methylated regions partitions the imprinted genes Igf2 and H19 into parent-speciﬁc chromatin loops. Nat. Genet. 36:889–93 85. Sillanpaa N, Magureanu CG, Murumagi A, Reinikainen A, West A, et al. 2004. Autoimmune regulator induced changes in the gene expression proﬁle of human monocytedendritic cell-lineage. Mol. Immunol. 41:1185–98 86. Kogawa K, Nagafuchi S, Katsuta H, Kudoh J, Tamiya S, et al. 2002. Expression of AIRE gene in peripheral monocyte/dendritic cell lineage. Immunol. Lett. 80:195–98 87. Hurst LD, Pal C, Lercher MJ. 2004. The evolutionary dynamics of eukaryotic gene order. Nat. Rev. Genet. 5:299–310 88. Reik W, Walter J. 2001. Genomic imprinting: parental inﬂuence on the genome. Nat. Rev. Genet. 2:21–32 89. Jackson-Grusby L, Beard C, Possemato R, Tudor M, Fambrough D, et al. 2001. Loss of genomic methylation causes p53-dependent apoptosis and epigenetic deregulation. Nat. Genet. 27:31–39 90. Gotter J, Kyewski B. 2004. Regulating self-tolerance by deregulating gene expression. Curr. Opin. Immunol. 16:741–45 91. Schonrich G, Momburg F, Hammerling GJ, Arnold B. 1992. Anergy induced by thymic medullary epithelium. Eur. J. Immunol. 22:1687–91 www.annualreviews.org • Central T Cell Tolerance

599

ARI

23 February 2006

15:27

92. Jenkinson EJ, Jhittay P, Kingston R, Owen JJ. 1985. Studies of the role of the thymic environment in the induction of tolerance to MHC antigens. Transplantation 39:331–33 93. von Boehmer H, Hafen K. 1986. Minor but not major histocompatibility antigens of thymus epithelium tolerize precursors of cytolytic T cells. Nature 320:626–28 94. Good MF, Pyke KW, Nossal GJ. 1983. Functional clonal deletion of cytotoxic Tlymphocyte precursors in chimeric thymus produced in vitro from embryonic Anlagen. Proc. Natl. Acad. Sci. USA 80:3045–49 95. Jordan RK, Robinson JH, Hopkinson NA, House KC, Bentley AL. 1985. Thymic epithelium and the induction of transplantation tolerance in nude mice. Nature 314:454–56 96. Gao EK, Lo D, Sprent J. 1990. Strong T cell tolerance in parent—F1 bone marrow chimeras prepared with supralethal irradiation. Evidence for clonal deletion and anergy. J. Exp. Med. 171:1101–21 97. Ramsdell F, Lantz T, Fowlkes BJ. 1989. A nondeletional mechanism of thymic self tolerance. Science 246:1038–41 98. Hoffmann MW, Allison J, Miller JF. 1992. Tolerance induction by thymic medullary epithelium. Proc. Natl. Acad. Sci. USA 89:2526–30 99. Bonomo A, Matzinger P. 1993. Thymus epithelium induces tissue-speciﬁc tolerance. J. Exp. Med. 177:1153–64 100. Witt CM, Raychaudhuri S, Schaefer B, Chakraborty AK, Robey EA. 2005. Directed migration of positively selected thymocytes visualized in real time. PLoS Biol. 3:e160 101. Witt CM, Robey EA. 2005. Thymopoiesis in 4 dimensions. Semin. Immunol. 17:95–102 102. Miller MJ, Hejazi AS, Wei SH, Cahalan MD, Parker I. 2004. T cell repertoire scanning is promoted by dynamic dendritic cell behavior and random T cell motility in the lymph node. Proc. Natl. Acad. Sci. USA 101:998–1003 103. Scollay R, Godfrey DI. 1995. Thymic emigration: conveyor belts or lucky dips? Immunol. Today 16:268–73; discussion 73–74 104. Gallegos AM, Bevan MJ. 2004. Central tolerance to tissue-speciﬁc antigens mediated by direct and indirect antigen presentation. J. Exp. Med. 200:1039–49 105. Zhang M, Vacchio MS, Vistica BP, Lesage S, Egwuagu CE, et al. 2003. T cell tolerance to a neo-self antigen expressed by thymic epithelial cells: the soluble form is more effective than the membrane-bound form. J. Immunol. 170:3954–62 106. den Haan JM, Lehar SM, Bevan MJ. 2000. CD8+ but not CD8− dendritic cells crossprime cytotoxic T cells in vivo. J. Exp. Med. 192:1685–96 107. Belz GT, Behrens GM, Smith CM, Miller JF, Jones C, et al. 2002. The CD8α+ dendritic cell is responsible for inducing peripheral self-tolerance to tissue-associated antigens. J. Exp. Med. 196:1099–104 108. Albert ML, Sauter B, Bhardwaj N. 1998. Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs. Nature 392:86–89 109. Albert ML, Pearce SF, Francisco LM, Sauter B, Roy P, et al. 1998. Immature dendritic cells phagocytose apoptotic cells via αvβ5 and CD36, and cross-present antigens to cytotoxic T lymphocytes. J. Exp. Med. 188:1359–68 110. Bellone M, Iezzi G, Rovere P, Galati G, Ronchetti A, et al. 1997. Processing of engulfed apoptotic bodies yields T cell epitopes. J. Immunol. 159:5391–99 111. Thery C, Zitvogel L, Amigorena S. 2002. Exosomes: composition, biogenesis and function. Nat. Rev. Immunol. 2:569–79 112. Harshyne LA, Watkins SC, Gambotto A, Barratt-Boyes SM. 2001. Dendritic cells acquire antigens from live cells for cross-presentation to CTL. J. Immunol. 166:3717–23

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

600

Kyewski

·

Klein

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

113. Humblet C, Rudensky A, Kyewski B. 1994. Presentation and intercellular transfer of self antigen within the thymic microenvironment: expression of the E alpha peptide-I-Ab complex by isolated thymic stromal cells. Int. Immunol. 6:1949–58 114. Barlow AK, He X, Janeway C Jr. 1998. Exogenously provided peptides of a self-antigen can be processed into forms that are recognized by self-T cells. J. Exp. Med. 187:1403–15 115. Viret C, Barlow AK, Janeway CA Jr. 1999. On the intrathymic intercellular transfer of self-determinants. Immunol. Today 20:8–10 116. Neijssen J, Herberts C, Drijfhout JW, Reits E, Janssen L, et al. 2005. Cross-presentation by intercellular peptide transfer through gap junctions. Nature 434:83–88 117. Liston A, Lesage S, Wilson J, Peltonen L, Goodnow CC. 2003. Aire regulates negative selection of organ-speciﬁc T cells. Nat. Immunol. 4:350–54 118. Su MA, Anderson MS. 2004. Aire: an update. Curr. Opin. Immunol. 16:746–52 118a. Uchida D, Hatakeyama S, Matsushima A, Han H, Ishido S, et al. 2004. AIRE functions as an E3 ubiquitin ligase. J. Exp. Med. 199:167–72 118b. Bottomley MJ, Stier G, Pennacchini D, Legube G, Simon B, et al. 2005. NMR structure of the ﬁrst PHD ﬁnger of autoimmune regulator protein (AIRE1). Insights into autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) disease. J. Biol. Chem. 280:11505–12 119. Itoh M, Takahashi T, Sakaguchi N, Kuniyasu Y, Shimizu J, et al. 1999. Thymus and autoimmunity: production of CD25+ CD4+ naturally anergic and suppressive T cells as a key function of the thymus in maintaining immunologic self-tolerance. J. Immunol. 162:5317–26 120. Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, et al. 2005. Regulatory T cell lineage speciﬁcation by the forkhead transcription factor Foxp3. Immunity 22:329– 41 121. Jordan MS, Boesteanu A, Reed AJ, Petrone AL, Holenbeck AE, et al. 2001. Thymic selection of CD4+ CD25+ regulatory T cells induced by an agonist self-peptide. Nat. Immunol. 2:301–6 122. Apostolou I, Sarukhan A, Klein L, von Boehmer H. 2002. Origin of regulatory T cells with known speciﬁcity for antigen. Nat. Immunol. 3:756–63 123. Klein L, Khazaie K, von Boehmer H. 2003. In vivo dynamics of antigen-speciﬁc regulatory T cells not predicted from behavior in vitro. Proc. Natl. Acad. Sci. USA 100:8886–91 124. Kawahata K, Misaki Y, Yamauchi M, Tsunekawa S, Setoguchi K, et al. 2002. Generation of CD4+ CD25+ regulatory T cells from autoreactive T cells simultaneously with their negative selection in the thymus and from nonautoreactive T cells by endogenous TCR expression. J. Immunol. 168:4399–405 125. Walker LS, Chodos A, Eggena M, Dooms H, Abbas AK. 2003. Antigen-dependent proliferation of CD4+ CD25+ regulatory T cells in vivo. J. Exp. Med. 198:249–58 125a. Watanabe N, Wang YH, Lee HK, Ito T, Cao W, Liu YJ. 2005. Hassall’s corpuscles instruct dendritic cells to induce CD4+ CD25+ regulatory T cells in human thymus. Nature 436:1181–85 126. Laufer TM, DeKoning J, Markowitz JS, Lo D, Glimcher LH. 1996. Unopposed positive selection and autoreactivity in mice expressing class II MHC only on thymic cortex. Nature 383:81–85 127. Bensinger SJ, Bandeira A, Jordan MS, Caton AJ, Laufer TM. 2001. Major histocompatibility complex class II-positive cortical epithelium mediates the selection of CD4+ 25+ immunoregulatory T cells. J. Exp. Med. 194:427–38 127a. Fontenot JD, Dooley JL, Farr AG, Rudensky AY. 2005. Developmental regulation of Foxp3 expression during ontogeny. J. Exp. Med. 202:901–6 www.annualreviews.org • Central T Cell Tolerance

601

ARI

23 February 2006

15:27

128. Derbinski J, Kyewski B. 2005. Linking signalling pathways, thymic stroma integrity and autoimmunity. Trends Immunol. 26:503–6 129. Akiyama T, Maeda S, Yamane S, Ogino K, Kasai M, et al. 2005. Dependence of self-tolerance on TRAF6-directed development of thymic stroma. Science 308:248– 51 130. Kajiura F, Sun S, Nomura T, Izumi K, Ueno T, et al. 2004. NF-κB-inducing kinase establishes self-tolerance in a thymic stroma-dependent manner. J. Immunol. 172:2067– 75 131. Thorstenson KM, Khoruts A. 2001. Generation of anergic and potentially immunoregulatory CD25+ CD4 T cells in vivo after induction of peripheral tolerance with intravenous or oral antigen. J. Immunol. 167:188–95 132. Chen W, Jin W, Hardegen N, Lei KJ, Li L, et al. 2003. Conversion of peripheral CD4+ CD25− naive T cells to CD4+ CD25+ regulatory T cells by TGF-β induction of transcription factor Foxp3. J. Exp. Med. 198:1875–86 133. Apostolou I, von Boehmer H. 2004. In vivo instruction of suppressor commitment in naive T cells. J. Exp. Med. 199:1401–8 133a. Kretschmer K, Apostolou I, Hawiger D, Khazaie K, Nussenzweig MC, von Boehmer H. 2005. Inducing and expanding regulatory T cell populations by foreign antigen. Nat. Immunol. 6:1219–27 134. Chin RK, Lo JC, Kim O, Blink SE, Christiansen PA, et al. 2003. Lymphotoxin pathway directs thymic Aire expression. Nat. Immunol. 4:1121–27 135. Burkly L, Hession C, Ogata L, Reilly C, Marconi LA, et al. 1995. Expression of relB is required for the development of thymic medulla and dendritic cells. Nature 373:531– 36 136. Weih F, Carrasco D, Durham SK, Barton DS, Rizzo CA, et al. 1995. Multiorgan inﬂammation and hematopoietic abnormalities in mice with a targeted disruption of RelB, a member of the NF-κB/Rel family. Cell 80:331–40 137. Heino M, Peterson P, Sillanpaa N, Guerin S, Wu L, et al. 2000. RNA and protein expression of the murine autoimmune regulator gene (Aire) in normal, RelB-deﬁcient and in NOD mouse. Eur. J. Immunol. 30:1884–93 138. Pugliese A, Zeller M, Fernandez A Jr., Zalcberg LJ, Bartlett RJ, et al. 1997. The insulin gene is transcribed in the human thymus and transcription levels correlated with allelic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes. Nat. Genet. 15:293–97 139. Vaﬁadis P, Bennett ST, Todd JA, Nadeau J, Grabs R, et al. 1997. Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus. Nat. Genet. 15:289–92 140. Chentouﬁ AA, Polychronakos C. 2002. Insulin expression levels in the thymus modulate insulin-speciﬁc autoreactive T-cell tolerance: the mechanism by which the IDDM2 locus may predispose to diabetes. Diabetes 51:1383–90 141. Egwuagu CE, Charukamnoetkanok P, Gery I. 1997. Thymic expression of autoantigens correlates with resistance to autoimmune disease. J. Immunol. 159:3109–12 142. Charukamnoetkanok P, Fukushima A, Whitcup SM, Gery I, Egwuagu CE. 1998. Expression of ocular autoantigens in the mouse thymus. Curr. Eye Res. 17:788–92 143. Liu H, MacKenzie-Graham AJ, Kim S, Voskuhl RR. 2001. Mice resistant to experimental autoimmune encephalomyelitis have increased thymic expression of myelin basic protein and increased MBP speciﬁc T cell tolerance. J. Neuroimmunol. 115:118– 26

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

602

Kyewski

·

Klein

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

144. Klein L, Klugmann M, Nave KA, Tuohy VK, Kyewski B. 2000. Shaping of the autoreactive T-cell repertoire by a splice variant of self protein expressed in thymic epithelial cells. Nat. Med. 6:56–61 145. Anderson AC, Nicholson LB, Legge KL, Turchin V, Zaghouani H, et al. 2000. High frequency of autoreactive myelin proteolipid protein-speciﬁc T cells in the periphery of naive mice: mechanisms of selection of the self-reactive repertoire. J. Exp. Med. 191:761–70 146. Li HS, Carayanniotis G. 2005. Detection of thyroglobulin mRNA as truncated isoform(s) in mouse thymus. Immunology 115:85–89 147. Bruno R, Sabater L, Tolosa E, Sospedra M, Ferrer-Francesch X, et al. 2004. Different patterns of nicotinic acetylcholine receptor subunit transcription in human thymus. J. Neuroimmunol. 149:147–59 148. Diez J, Park Y, Zeller M, Brown D, Garza D, et al. 2001. Differential splicing of the IA-2 mRNA in pancreas and lymphoid organs as a permissive genetic mechanism for autoimmunity against the IA-2 type 1 diabetes autoantigen. Diabetes 50:895–900 149. Talib S, Okarma TB, Lebkowski JS. 1993. Differential expression of human nicotinic acetylcholine receptor alpha subunit variants in muscle and non-muscle tissues. Nucleic Acids Res. 21:233–37 149a. Allen S, Read S, DiPaolo R, McHugh RS, Shevach EM, et al. 2005. Promiscuous thymic expression of an autoantigen gene does not result in negative selection of pathogenic T cells. J. Immunol. 175:5759–64 150. Backlund J, Treschow A, Bockermann R, Holm B, Holm L, et al. 2002. Glycosylation of type II collagen is of major importance for T cell tolerance and pathology in collageninduced arthritis. Eur. J. Immunol. 32:3776–84 151. Vogel A, Strassburg CP, Obermayer-Straub P, Brabant G, Manns MP. 2002. The genetic background of autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy and its autoimmune disease components. J. Mol. Med. 80:201–11 152. Heino M, Peterson P, Kudoh J, Shimizu N, Antonarakis SE, et al. 2001. APECED mutations in the autoimmune regulator (AIRE) gene. Hum. Mutat. 18:205–11 152a. Jiang W, Anderson MS, Bronson R, Mathis D, Benoist C. 2005. Modiﬁer loci condition autoimmunity provoked by Aire deﬁciency. J. Exp. Med. 202:805–15 153. Soderbergh A, Myhre AG, Ekwall O, Gebre-Medhin G, Hedstrand H, et al. 2004. Prevalence and clinical associations of 10 deﬁned autoantibodies in autoimmune polyendocrine syndrome type I. J. Clin. Endocrinol. Metab. 89:557–62 154. Jaeckel E, Lipes MA, von Boehmer H. 2004. Recessive tolerance to preproinsulin 2 reduces but does not abolish type 1 diabetes. Nat. Immunol. 5:1028–35 155. Nakayama M, Abiru N, Moriyama H, Babaya N, Liu E, et al. 2005. Prime role for an insulin epitope in the development of type 1 diabetes in NOD mice. Nature 435:220–23 156. Kent SC, Chen Y, Bregoli L, Clemmings SM, Kenyon NS, et al. 2005. Expanded T cells from pancreatic lymph nodes of type 1 diabetic subjects recognize an insulin epitope. Nature 435:224–28 157. Rioux JD, Abbas AK. 2005. Paths to understanding the genetic basis of autoimmune disease. Nature 435:584–89 158. Sabater L, Ferrer-Francesch X, Sospedra M, Caro P, Juan M, Pujol-Borrell R. 2005. Insulin alleles and autoimmune regulator (AIRE) gene expression both inﬂuence insulin expression in the thymus. J. Autoimmun. 25:312–18 159. Miyamoto K, Miyake S, Schachner M, Yamamura T. 2003. Heterozygous null mutation of myelin P0 protein enhances susceptibility to autoimmune neuritis targeting P0 peptide. Eur. J. Immunol. 33:656–65 www.annualreviews.org • Central T Cell Tolerance

603

ARI

23 February 2006

15:27

160. Takase H, Yu CR, Mahdi RM, Douek DC, Dirusso GB, et al. 2005. Thymic expression of peripheral tissue antigens in humans: a remarkable variability among individuals. Int. Immunol. 17:1131–40 161. Voskuhl RR. 1998. Myelin protein expression in lymphoid tissues: implications for peripheral tolerance. Immunol. Rev. 164:81–92 162. Muller-Hermelink HK, Marx A. 2000. Thymoma. Curr. Opin. Oncol. 12:426–33 163. Hoffacker V, Schultz A, Tiesinga JJ, Gold R, Schalke B, et al. 2000. Thymomas alter the T-cell subset composition in the blood: a potential mechanism for thymoma-associated autoimmune disease. Blood 96:3872–79 164. Shibahara H, Shiraishi Y, Hirano Y, Suzuki T, Takamizawa S, et al. 2003. Diversity of the inhibitory effects on fertilization by anti-sperm antibodies bound to the surface of ejaculated human sperm. Hum. Reprod. 18:1469–73 165. Diekman AB, Norton EJ, Westbrook VA, Klotz KL, Naaby-Hansen S, et al. 2000. Antisperm antibodies from infertile patients and their cognate sperm antigens: a review. Identity between SAGA-1, the H6–3C4 antigen, and CD52. Am. J. Reprod. Immunol. 43:134–43 166. Taneichi A, Shibahara H, Takahashi K, Sasaki S, Kikuchi K, et al. 2003. Effects of sera from infertile women with sperm immobilizing antibodies on fertilization and embryo development in vitro in mice. Am. J. Reprod. Immunol. 50:146–51 167. Tung KS, Setiady YY, Samy ET, Lewis J, Teuscher C. 2005. Autoimmune ovarian disease in day 3-thymectomized mice: the neonatal time window, antigen speciﬁcity of disease suppression, and genetic control. Curr. Top. Microbiol. Immunol. 293:209–47 168. Maclaren N, Chen QY, Kukreja A, Marker J, Zhang CH, et al. 2001. Autoimmune hypogonadism as part of an autoimmune polyglandular syndrome. J. Soc. Gynecol. Investig. 8:S52–54 169. Laml T, Preyer O, Umek W, Hengstschlager M, Hanzal H. 2002. Genetic disorders in premature ovarian failure. Hum. Reprod. Update 8:483–91 170. Zenclussen AC, Gerlof K, Zenclussen ML, Sollwedel A, Bertoja AZ, et al. 2005. Abnormal T-cell reactivity against paternal antigens in spontaneous abortion: adoptive transfer of pregnancy-induced CD4+ CD25+ T regulatory cells prevents fetal rejection in a murine abortion model. Am. J. Pathol. 166:811–22 171. Pardoll D. 2003. Does the immune system see tumors as foreign or self? Annu. Rev. Immunol. 21:807–39 172. Boon T, Old LJ. 1997. Cancer tumor antigens. Curr. Opin. Immunol. 9:681–83 173. Rosenberg SA, Yang JC, Restifo NP. 2004. Cancer immunotherapy: moving beyond current vaccines. Nat. Med. 10:909–15 173a. Boon T, Coulie PG, van den Eynde BJ, van der Bruggen P. 2005. Human T cell responses against melanoma. Annu. Rev. Immunol. 24:175–208 174. Wang RF, Rosenberg SA. 1999. Human tumor antigens for cancer vaccine development. Immunol. Rev. 170:85–100 175. Van Pel A, van der Bruggen P, Coulie PG, Brichard VG, Lethe B, et al. 1995. Genes coding for tumor antigens recognized by cytolytic T lymphocytes. Immunol. Rev. 145:229–50 176. Kyewski B, Derbinski J. 2004. Self-representation in the thymus: an extended view. Nat. Rev. Immunol. 4:688–98 177. Bos R, van Duikeren S, van Hall T, Kaaijk P, Taubert R, et al. 2005. Expression of a natural tumor antigen by thymic epithelial cells impairs the tumor-protective CD4+ T-cell repertoire. Cancer Res. 65:6443–49

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

604

Kyewski

·

Klein

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-19

ARI

23 February 2006

15:27

178. Antony PA, Piccirillo CA, Akpinarli A, Finkelstein SE, Speiss PJ, et al. 2005. CD8+ T cell immunity against a tumor/self-antigen is augmented by CD4+ T helper cells and hindered by naturally occurring T regulatory cells. J. Immunol. 174:2591–601 179. Sutmuller RP, van Duivenvoorde LM, van Elsas A, Schumacher TN, Wildenberg ME, et al. 2001. Synergism of cytotoxic T lymphocyte-associated antigen 4 blockade and depletion of CD25+ regulatory T cells in antitumor therapy reveals alternative pathways for suppression of autoreactive cytotoxic T lymphocyte responses. J. Exp. Med. 194:823– 32 180. Overwijk WW, Theoret MR, Finkelstein SE, Surman DR, de Jong LA, et al. 2003. Tumor regression and autoimmunity after reversal of a functionally tolerant state of self-reactive CD8+ T cells. J. Exp. Med. 198:569–80 181. Dudley ME, Wunderlich JR, Robbins PF, Yang JC, Hwu P, et al. 2002. Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298:850–54 182. Renkvist N, Castelli C, Robbins PF, Parmiani G. 2001. A listing of human tumor antigens recognized by T cells. Cancer Immunol. Immunother. 50:3–15 183. Trede NS, Langenau DM, Traver D, Look AT, Zon LI. 2004. The use of zebraﬁsh to understand immunity. Immunity 20:367–79 184. Ramsdell F, Ziegler SF. 2003. Transcription factors in autoimmunity. Curr. Opin. Immunol. 15:718–24 185. Chen Z, Benoist C, Mathis D. 2005. How defects in central tolerance impinge on a deﬁciency in regulatory T cells. Proc. Natl. Acad. Sci. USA 102:14735–40 186. Heino M, Peterson P, Kudoh J, Nagamine K, Lagerstedt A, et al. 1999. Autoimmune regulator is expressed in the cells regulating immune tolerance in thymus medulla. Biochem. Biophys. Res. Commun. 257:821–25 187. Bjorses P, Pelto-Huikko M, Kaukonen J, Aaltonen J, Peltonen L, et al. 1999. Localization of the APECED protein in distinct nuclear structures. Hum. Mol. Genet. 8:259–66 188. Lercher MJ, Urrutia AO, Pavlicek A, Hurst LD. 2003. A uniﬁcation of mosaic structures in the human genome. Hum. Mol. Genet. 12:2411–15 189. Garcia CA, Prabakar KR, Diez J, Cao ZA, Allende G, et al. 2005. Dendritic cells in human thymus and periphery display a proinsulin epitope in a transcription-dependent, capture-independent fashion. J. Immunol. 175:2111–22 190. Zheng X, Yin L, Liu Y, Zheng P. 2004. Expression of tissue-speciﬁc autoantigens in the hematopoietic cells leads to activation-induced cell death of autoreactive T cells in the secondary lymphoid organs. Eur. J. Immunol. 34:3126–34 191. Goldschneider I, Cone RE. 2003. A central role for peripheral dendritic cells in the induction of acquired thymic tolerance. Trends Immunol. 24:77–81 192. Huseby ES, Sather B, Huseby PG, Goverman J. 2001. Age-dependent T cell tolerance and autoimmunity to myelin basic protein. Immunity 14:471–81 193. Petrie HT. 2003. Cell migration and the control of post-natal T-cell lymphopoiesis in the thymus. Nat. Rev. Immunol. 3:859–66

RELATED RESOURCES Shevach EM. 2000. Regulatory T cells in autoimmunity. Annu. Rev. Immunol. 18:423–50 Heath WR, Carbone FR. 2001. Cross-presentation, dendritic cells, tolerance, and immunity. Annu. Rev. Immunol. 19:47–64 www.annualreviews.org • Central T Cell Tolerance

605

ANRV270-IY24-19

ARI

23 February 2006

15:27

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Steinman RM, Hawiger D, Nussenzweig MC. 2003. Tolerogenic dendritic cells. Annu. Rev. Immunol. 21:423–50 Anderson MS, Bluestone JA. 2005. The NOD mouse: a model of immune dysregulation. Annu. Rev. Immunol. 23:447–86 Sospedra M, Martin R. 2005. Immunology of multiple sclerosis. Annu. Rev. Immunol. 23:663– 748

606

Kyewski

·

Klein

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:571-606. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao Harvard Medical School, CBR Institute for Biomedical Research, Boston, Massachusetts 02115; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:607–56 First published online as a Review in Advance on January 24, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.23.021704.115821 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0607$20.00

Key Words cytokine, transcription factors, epigenetic regulation, chromatin, RNA interference/RNAi

Abstract Helper T cells coordinate immune responses through the production of cytokines. Th2 cells express the closely linked Il4, Il13, and Il5 cytokine genes, whereas these same genes are silenced in the Th1 lineage. The Th1/Th2 lineage choice has become a textbook example for the regulation of cell differentiation, and recent discoveries have further reﬁned and expanded our understanding of how Th2 differentiation is initiated and reinforced by signals from antigen-presenting cells and cytokine-driven feedback loops. Epigenetic changes that stabilize the active or silent state of the Il4 locus in differentiating helper T cells have been a major focus of recent research. Overall, the ﬁeld is progressing toward an integrated model of the signaling and transcription factor networks, cis-regulatory elements, epigenetic modiﬁcations, and RNA interference mechanisms that converge to determine the lineage fate and gene expression patterns of differentiating helper T cells.

607

ANRV270-IY24-20

ARI

23 February 2006

15:32

INTRODUCTION

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

An established property of effector Th1 and Th2 cells is the ability to produce high levels of the effector cytokines IFN-γ and IL-4, respectively (1). These two cell types differentiate from a common precursor, the naive T cell that has not yet encountered antigen. The role of chromatin-based structural changes and epigenetic mechanisms in Th1/Th2 differentiation was ﬁrst proposed in 1998 (2). Since then, many laboratories have worked on the problem and many excellent reviews have appeared that cover epigenetic and other aspects of helper T cell differentiation and type II cytokine production (3–10 and references therein). In this review, we focus primarily on recent advances in our understanding of the transcriptional and epigenetic basis of Th2 differentiation, while providing enough background to put these recent discoveries into a historical perspective. In the ﬁrst section, we describe the roles of transcription factors and epigenetic processes in the three stages of Th2 differentiation: initiation, reinforcement, and maintenance. In the second section, we provide a brief general introduction to epigenetic regulation of gene expression. In the third section, we focus on epigenetic regulation of the Il4 locus (by which we mean the extended genomic region containing Il5, Rad50, Il13, and Il4). We discuss the properties of cisregulatory sequences in the Il4 locus and their functions as determined by gene targeting in mice. Finally, in the fourth section, we discuss the very recent emerging evidence that the RNAi machinery plays a role in Th1/Th2 lineage speciﬁcation. We hope that the review will be useful both to newcomers and to established scientists in the ﬁeld, and apologize to colleagues whose work we have been unable to cover fully owing to space limitations.

THREE STAGES OF Th2 DIFFERENTIATION Helper T cell (Th) differentiation is deﬁned by acquisition of the ability to produce, se608

Ansel et al.

lectively, large amounts of speciﬁc effector cytokines upon antigen exposure. The process of differentiation can be conceptually divided into three stages: initiation, reinforcement, and maintenance. IL-4 itself potently initiates Th2 differentiation, but the early source of IL-4 and even its importance as a Th2-initiating factor in physiological conditions remain unclear. Once the differentiation process is set in motion, powerful feedback loops and cytokine signals reinforce the lineage decisions of individual Th2 cells as well as the overall polarization of the T cell response at the population level. These mechanisms are sufﬁcient to establish the functional characteristics of Th2 cells, including resistance to Th1-inducing signals. However, epigenetic mechanisms eventually liberate Th2 cells from the requirement for these feedback loops to maintain their gene expression patterns.

Initiation of Th2 Differentiation In in vitro studies of helper T cell differentiation, Th2 cells are typically generated by activating naive T cells through T cell antigen receptor (TCR) crosslinking in the presence of exogenous IL-4 (Figure 1). The IL-4 receptor (IL-4R), which consists of the common cytokine-receptor gamma subunit and the IL4-binding IL-4Rα chain, is expressed on naive T cells. IL-4R signals are transduced by the transcription factor STAT6 (signal transduction and activator of transcription 6), which, together with NFAT (nuclear factor of activated T cells), AP-1, NF-κB, and other TCRinduced signals, activates transcription of Il4 as well as the gene encoding the transcription factor, GATA3, a major regulator of Th2 lineage commitment. The resulting autocrine feedback loop acts positively to favor Th2 differentiation. During this process, the Il4, Il5, and Il13 genes, which reside together in a region we designate the Il4 locus, are potentiated for extremely strong transcriptional activation upon restimulation. This secondary burst of cytokine gene transcription occurs

ANRV270-IY24-20

ARI

23 February 2006

15:32

Naive T cell

antigen

antigen

NFAT STAT5 STAT6

NFAT

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Differentiated Th2 cell

IL-2R

IFN-γ

IL-4

IL-2

IL-2

Resting

Restimulated

GATA3

NFAT GATA3

IL-4R

IL-4

IL-4

HOURS DAYS (chromatin remodeling) Figure 1 Initiation and reinforcement of helper T cell differentiation. Cell-autonomous model of Th2 differentiation. Antigen stimulation of naive CD4+ T cells induces NFAT-dependent production of IL-2, IL-4, and IFN-γ. IL-4 and IL-2 reinforce Il4 locus activation and Th2 differentiation through activation of STAT6 and STAT5, respectively. Naive T cells can produce sufﬁcient IL-2 and IL-4 to induce their own differentiation into Th2 cells, and other cells that produce these cytokines may inﬂuence Th2 differentiation as well (not shown). IL-4R/STAT6 signaling is a potent inducer of Th2 differentiation, as it also induces expression of the transcription factor GATA3. GATA3 further induces its own expression and, over a period of days, directs chromatin remodeling of the Il4 locus. These changes allow fully differentiated Th2 cells to produce a large quantity of IL-4 rapidly upon antigen restimulation, without need for further cytokine signals. Th2 differentiation also involves silencing of the Ifng locus, but this process occurs more slowly and/or less efﬁciently than Il4 locus activation.

independently of instructive cytokine signals. As several recent review articles have detailed the roles of signaling pathways and transcription factors in Th2 differentiation (1, 3, 10, 11), we limit our discussion here to a brief review of only three key factors: NFAT, STAT6, and GATA3. Among the antigen-induced transcription factors, NFAT proteins have an essential role: They bind to the Il4, Il5, and Il13 promoters and several enhancer elements in the locus (12, 13; S. Kim & A. Rao, unpublished data). Treatment with the calcineurin inhibitor cyclosporin A (CsA), which inhibits the dephosphorylation and nuclear import

of NFAT proteins, blocks acute transcriptional activation of cytokine genes regardless of the differentiation status of T cells and also blocks the chromatin structural changes that occur in the Il4 locus during Th2 differentiation (14). T cells express three NFAT proteins: NFAT1/NFATc2, NFAT2/NFATc1, and NFAT4/NFATc3. Experiments with T cells carrying various combinations of targeted mutations inactivating NFAT genes conﬁrmed that NFAT is required for cytokine gene transcription and further revealed the existence of an NFAT1-mediated negative feedback loop that acts to limit the duration of Il4 gene transcription during T cell activation. www.annualreviews.org • Regulation of Th2 Differentiation

609

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Thus, lack of NFAT1 is associated with a pronounced Th2 bias that is further exacerbated in the combined absence of NFAT1 and NFAT4 (15–18; reviewed in 3, 19, and 20). IL-4 potently synergizes with TCR signals to induce Th2 differentiation, and overexpression of constitutively active STAT6 in activated T cells also efﬁciently induces Il4 locus potentiation (21). Although STAT6 binds directly to the Il4 promoter, the physiological importance of STAT6 for the Th2 program can be largely attributed to its ability to synergize with TCR signals to upregulate Gata3. STAT6 is not absolutely necessary for Th2 differentiation, but the small minority of STAT6−/− cells that are capable of producing IL-4 show concomitant upregulation of Gata3 (22). The increased GATA3 in these cells could be stochastic, or it could be induced by other Th2 initiating pathways, as we discuss below. The importance of GATA3 in Th2 differentiation has been thoroughly established by studies from many different laboratories (23–28). Naive T cells express a basal level of Gata3, which is upregulated in the course of Th2 differentiation and extinguished in the course of Th1 differentiation. Gata3 is a downstream target of STAT6 signaling, and upregulation of Gata3 in primary T cells is dependent on TCR signals and blocked by the NF-κB inhibitor, SN50, implicating NF-κB in the regulation of GATA3 (29). Overexpression of GATA3, either in transgenic mice or retrovirally in primary T cells, leads to increased production of IL-4 and other Th2 cytokines (23–25); conversely, retroviral overexpression of a dominant-negative GATA3 in primary differentiating Th2 cells diminished Th2 cytokine production (30). In addition to directly binding to the Il5 and Il13 promoters (31, 32), GATA3 appears to play an important role in establishing many of the long-range chromatin changes that occur within the Il4 locus during Th2 differentiation (22, 23, 28, 33).

Initiating Th2 responses in vivo. Our ability to use exogenous IL-4 to drive Th2 differentiation in vitro has made this process one of the most tractable experimental systems for the investigation of cell-lineage speciﬁcation. Yet the way in which Th2 differentiation is initiated in vivo remains far less well understood. In physiological settings, naive T cells ﬁrst encounter antigens presented by dendritic cells (DCs) in the T cell zones of secondary lymphoid organs, where IL-4 is scarce (Figure 2). Inﬂammatory stimuli induce DC activation, antigen presentation, costimulatory molecule expression, and migration to sites of T-DC interaction (34). When DCs recognize bacterial and viral products via Toll-like receptors (TLRs), the cells produce IL-12 and IL-23; these cytokines, in turn, induce Th1 differentiation (35). Th2 differentiation also begins with an interaction between naive T cells and antigenpresenting DCs, but Th2-inducing stimuli activate DCs without inducing production of either Th1- or Th2-promoting cytokines (IL-12/IL-23 and IL-2/IL-4, respectively). Other cellular sources of IL-4 include mast cells, basophils, eosinophils, NKT cells, and previously differentiated Th2 cells (4, 36, 37). Many of these cell types display high basal *Erratum levels of Il4 transcripts and can be triggered to rapidly release Il4 upon stimulation: mast cells and basophils via IgE receptor crosslinking, and NKT cells by lipoprotein ligands presented by nonclassical MHC molecules of the CD1 family. Although any of these cells may help initiate Th2 differentiation if they are activated in the right place at the right time to coincide with naive T cell antigen encounter, robust Th2 responses can be elicited (a) in mice in which only CD4+ T cells can produce IL-4 (38) and (b) in mice that lack invariant NKT cells or CD1 (39–42). Early Il4 expression and Th2 differentiation. In vitro Th2 differentiation occurs in cultures of highly puriﬁed naive T cells even in the absence of exogenous IL-4 or

*Erratum (21 Mar. 2006): See online log at http://arjournals.annualreviews.org/errata/immunol 610

Ansel et al.

ANRV270-IY24-20

ARI

23 February 2006

15:32

MyD88 TLR

IL-12 IL-23

MyD88 TLR

?

PGE2, CTX, allergens?

Dendritic cell (DC) IL-12 IL-23

MHC peptide TCR

DL4 Notch1, 2 ICD

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CsA

MHC peptide TCR

Jagged

γ-secretase inhibitors

Jagged Notch1, 2 ICD

CsA

T cell NFAT?

ICD MAML RBPJκ

Tbet

NFAT AP-1 Tbet

NFAT?

IFN-γ

ICD MAML RBPJκ?

IL-4 NFAT AP-1

NFAT GATA3

GATA3

ICD MAML RBPJκ

IFN-γ

Th1 differentiation

IL-4

Th2 differentiation

Figure 2 Notch and dendritic cell (DC) inﬂuences on Th1/Th2 differentiation. Appropriately stimulated dendritic cells can induce naive T cell differentiation to Th1 or Th2 cell fates. Left panel: TLR signals enhance antigen presentation and induce strong expression of IL-12, IL-23, and the Notch ligands Jagged and DL4, generating DCs that efﬁciently induce Th1 differentiation. Notch ligation induces proteolytic cleavage of the Notch intracellular domain (ICD), which subsequently translocates to the nucleus and interacts with the transcriptional repressor RBPJκ. In cooperation with coactivators of the Mastermind-like (MAML) family, Notch ICD converts RBPJκ to a transcriptional activator. The ICD/MAML/RBPJκ complex binds to the promoter of the gene encoding the transcription factor T-bet, a key regulator of Th1 differentiation. T-bet, NFAT, and AP-1 cooperate to activate IFN-γ expression in antigen-stimulated Th1 cells. Right panel: In the absence of the signaling adaptor MyD88, TLR ligation induces Jagged but not DL4 upregulation. Other DC stimulants also induce Jagged upregulation and induce GATA3 in an IL-4R/STAT6-independent manner. Notch may also directly regulate Il4 expression via RBPJκ binding sites in a conserved downstream enhancer (CNS2). NFAT/AP-1 and NFAT/GATA3 complexes induce Il4 transcription in activated Th2 cells through the Il4 promoter and 3 enhancer, respectively. Other Notch-independent pathways of Th2 initiation by DCs may also exist.

IL-4-producing cells, implicating naive T cells themselves as the cellular source of Th2initiating IL-4 (Figure 1). Indeed, TCR signals alone moderately upregulate naive T cell transcription of Il4 (but not Gata3) with rapid

kinetics that reach a peak within three hours of stimulation (43, 44). The genes encoding the Th2 cytokines IL-13 and IL-5, as well as IFNγ, which contributes to Th1 differentiation, are similarly induced by TCR signals. Early www.annualreviews.org • Regulation of Th2 Differentiation

611

ARI

23 February 2006

15:32

Il4 and Ifng expression require TCR-induced calcineurin activity and involve NFAT1 binding to the Il4 and Ifng promoters; however, instructive cytokine signals transmitted through STAT6 and STAT4 are not required at this early stage. Although naive T cells produce very small amounts of IL-4 in comparison to Th2 cells (0.1%–1.0% at the mRNA level), they can produce enough IL-4 to initiate their own Th2 differentiation, especially under conditions of IFN-γ neutralization (45–47). The importance of cytokines produced by T cells activated in nonpolarizing culture conditions, coupled with powerful feedback mechanisms that reinforce these signals on the cell and population level, leave the differentiation fate of these cells in a delicate balance. Small perturbations in early cytokine production, cytokine, antigen, and costimulatory receptor signaling, or transcription factor induction can dramatically alter the differentiation of helper T cells in vitro. Conceivably, in vivo T cell responses to relatively innocuous antigens that do not produce a robust inﬂammatory response may also be subject to these precarious regulatory circumstances. This situation, perhaps combined with a genetic predisposition for Th2 responses, could result in allergy.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

Genetic contributions to Th2 differentiation and disease. Allergic diseases, particularly asthma, pose a serious and growing medical challenge. Numerous large population genetic studies in human patients have established the complex multigenic inheritance of allergic diseases (48). The Il4 locus itself, located on human chromosome 5q23.3/5q31 (mouse 11qB1.3), has been identiﬁed both as a susceptibility locus for a number of allergic diseases and as a genetic modiﬁer of various disease states. For example, polymorphisms in the Il13 promoter and coding regions are strongly associated with asthma susceptibility and various aspects of allergic disease (49). In laboratory investigations, Th2-prone Balb strains of mice and their Th1-prone C57BL/6 and B10.D2 counterparts provide 612

Ansel et al.

tractable experimental systems for analyzing genetic contributions to Th1/Th2 differentiation. At least one aspect of the differentiation trait is T cell autonomous and can be quantiﬁed by measuring T cell IL-4 production in nonpolarizing culture conditions (50). Strainspeciﬁc differences in Il4 transcript levels can be measured as early as 16 h after activation, and early IL-4 production accounts for at least a portion of the differences in Th1/Th2 predisposition in vitro (51). Importantly, this in vitro trait correlates with well-established in vivo models of helper T cell–dependent immune responses: Balb/c mice are highly susceptible to experimental asthma and mount ineffective Th2 responses against the intracellular parasite Leishmania major, whereas Th1prone mouse strains resist asthma induction and mount curative Th1 responses against L. major (49, 52). Inbred mouse strains have been used for linkage analyses of asthma susceptibility in vivo, and these analyses have uncovered roles for complement protein 5a (53) and the Tim genes, which reside on the same chromosome as Il4 in the mouse and human genomes (54). Similarly, linkage analyses of Th2 bias in vitro identiﬁed two quantitative trait loci on mouse chromosome 16; these loci have been dubbed Dice1.1 and Dice1.2 and account for 27% of the difference between C57BL/6 and Balb/c mice for this multigenic trait (50, 55). Despite the similar effects of the Dice1.1 and Dice1.2 loci on Th2 bias in vitro, only the region that contains Dice1.2 inﬂuenced L. major susceptibility, strongly suggesting that early IL-4 from naive T cells is not the critical factor governing the Th1/Th2 polarization of L. major responses in vivo. Indeed, Leishmania antigen (LACK)-speciﬁc T cells in the draining lymph nodes of Balb/c and B10.D2 mice produce equally high levels of IL-4 in the ﬁrst days of L. major infection (56, 57). Clearly, IL-4 is a highly efﬁcacious initiator of Th2 differentiation, and the development of effective Th2 responses in vitro and in vivo depends strongly on IL-4. However, the difﬁculty in identifying a cellular source of

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

initiating IL-4 in physiological settings, along with an accumulating log of experiments in which at least the early stages of Th2 responses can occur in the absence of IL-4R signals, fueled a search for alternative Th2initiating signals (22, 58–60). As we discuss in the next section, one of these signals is mediated by the Notch pathway, which is capable of providing an IL-4/STAT6-independent initiating signal for Il4 and Gata3 expression. Notch: a new Th2-initiating pathway. Recent reports have characterized a complex regulatory role for Notch signaling in Th1/Th2 differentiation (61–64) (Figure 2). Notch receptors direct differentiation decisions in a great variety of developmental processes, including thymopoiesis (65). Ligand binding releases the intracellular domain (ICD) of Notch by proteolytic cleavage; this allows the ICD to enter the nucleus and transactivate genes through its association with the transcription factor RBPJκ (recombination signal binding protein Jκ, also known as CSL, CBF-1, suppressor of hairless, and Lag-1) and coactivators of the Mastermind-like (MAML) family. In the absence of Notch/MAML, RBPJκ acts as a transcriptional repressor. There are ﬁve mammalian ligands [Delta-like (DL)1, DL3, DL4, Jagged1, and Jagged2], all of which can activate any of the four Notch receptors. The receptors and ligands are favorably positioned to inﬂuence helper T cell fate decisions: Naive CD4 T cells express Notch1 and Notch2, and DCs express the Notch ligands Jagged1 and Jagged2 (66). DC activation through TLR upregulates Jagged1 and additionally induces a high level of DL4 expression (61, 67). Notch signals delivered by activated DCs can contribute to their ability to promote Th1 differentiation (61, 67) (Figure 2, left). Complete abrogation of Notch signaling with γ-secretase inhibitors blocks Th1 differentiation; conversely, forced expression of the constitutively active Notch ICD in T cells potentiates Th1 differentiation (63). RBPJκ and the Notch ICD bind to the T-bet promoter,

indicating that Notch signaling may directly upregulate T-bet, which encodes a major determinant of Th1 differentiation (63, 64). However, RBPJκ knockout T cells are not deﬁcient in Th1 differentiation (61, 62). One explanation is that the requirement for Notch signaling arises from the constitutive binding of RBPJκ to the T-bet promoter; this would result in basal repression of T-bet expression that could only be countered by Notch ICD. Thus, in the absence of RBPJκ, Notch signaling may no longer be necessary for T-bet upregulation. A second explanation is that DCs are capable of inducing Th1 differentiation via both DL4 and separate, Notch-independent Th1initiating signals, such as IL-12/-23. The two explanations are not mutually exclusive, and the latter is supported by the ﬁnding that a dominant-negative MAML impaired Th2 but not Th1 differentiation both in vitro and in vivo (67a). Notch appears to play a particularly prominent role in Th2 differentiation (Figure 2, right). RBPJκ knockout T cells show a striking defect in Th2 differentiation (61, 62), as do cells that express dominant-negative MAML (67a). Expression of Notch ICD in CD4 T cells increases IL-4 production, in part by increasing expression of GATA3 (61). In addition, Notch activation may also act directly on the Il4 locus to counter RBPJκmediated repression of a conserved enhancer, conserved noncoding sequence (CNS) 2. Expression of Notch ICD–stimulated transcription of Il4 minilocus reporter transgenes that contained CNS2 and mutation of the three RBPJκ consensus binding sequences in CNS2 greatly diminished this activity (61). Strikingly, the effect of Notch ICD was independent of STAT6. Thus, Notch is able to initiate Th2 differentiation independently of IL-4. However, optimal IL-4 production and spreading of the IL-4-producing phenotype to co-cultured T cells that were not expressing Notch ICD were STAT6 dependent, indicating that Notch-initiated Th2 differentiation is subject to IL-4-directed reinforcement. www.annualreviews.org • Regulation of Th2 Differentiation

613

ARI

23 February 2006

15:32

The connection between the TLR and Notch pathways has helped to clarify the complex role of antigen-presenting cells in the regulation of Th1/Th2 differentiation. When stimulated with lipopolysaccharide, DCs that lack the TLR adapter and signal-transducing protein Myd88 upregulate Jagged1 but not DL4 (unlike wild-type DCs, which upregulate both proteins) and preferentially induce Th2 differentiation (61, 68, 69). These mutant DCs resemble wild-type DCs that have been exposed to Prostaglandin E2 or cholera toxin; following exposure, the wild-type DCs upregulate Jagged2 (but not DL4) and also preferentially induce Th2 differentiation (61). Consistent with these ﬁndings, Jaggedexpressing antigen-presenting cell lines were shown to induce Th2 cytokine production preferentially, whereas Delta-expressing lines induced IFN-γ production. The mechanism controlling this ligand-speciﬁc effect remains unclear.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

Reinforcement of Th1/Th2 Phenotype Cytokine-driven feedback loops. Regardless of the initiating event, the onset of Th1 or Th2 cytokine gene expression engages powerful positive and negative feedback loops that reinforce lineage speciﬁcation. IL-4 reinforces its own expression in an autocrine and paracrine fashion: It strengthens Th2 commitment and recruits additional T cells into the Th2 response through STAT6-mediated Gata3 upregulation (Figure 1). Similarly, IFN-γ reinforces Th1 commitment through the induction of T-bet expression via STAT1 signaling downstream of the IFN-γ receptor (IFNGR1) (70). Upon TCR stimulation, IFNGR1 is rapidly mobilized to developing immunological synapses that contain the TCR (71). IL-4 treatment blocks this process. However, the block is not mediated through acute IL-4 signaling but rather requires STAT6 and therefore, presumably, prior transcription. Although the mechanism and biological signiﬁcance of this 614

Ansel et al.

phenomenon are not yet understood, it may represent a means for rapid IL-4-mediated negative regulation of Th1 differentiation. Developing Th1/Th2 cells also modulate their responsiveness to IL-12. T-bet transcriptionally upregulates expression of the IL-12 receptor β chain, and thereby sensitizes developing Th1 cells to IL-12 signals (70). Conversely, GATA3 represses STAT4 expression and thus decreases IL-12 responsiveness in developing Th2 cells (72). Cell-intrinsic feedback mechanisms also reinforce Th1/Th2 lineage commitment. GATA3 directly autoactivates its own expression (22), whereas T-bet forms an autoregulatory loop with the homeobox transcription factor Hlx, which is both a direct target and a transcriptional activator of T-bet (73, 74). The Th2-speciﬁc transcription factor c-Maf is critical for maximal Il4 transcription and is also responsible for the relatively high expression of the IL-2 receptor α chain (CD25) in developing Th2 cells in comparison with Th1 cells (75, 76). IL-2/STAT5 signals stabilize Il4 expression. IL-2 has long been implicated in the regulation of Il4 expression and Th2 differentiation (77). Recent reports substantially clarify its role in Th2 differentiation (33, 76, 78–81). IL-2 signals transmitted through STAT5 stabilize Il4 expression in developing Th2 cells independently of the effects of IL-2 on cell growth and survival (78, 81). In contrast to IL-4, which can deliver a potent Th2-initiating signal if delivered simultaneously with antigen receptor signals, IL-2 is required in the period that follows T cell activation (76, 78) (Figure 1). Even in optimal Th2-inducing conditions, IL-2-blocking antibodies or STAT5a deﬁciency impaired Il4 expression (33, 80). In the absence of STAT6, STAT5a appears indispensable for Th2 differentiation in vitro and in vivo (80). A constitutively active mutant of STAT5a (STAT5A1∗ 6) bypassed the requirement for IL-2 in IL-4-induced Th2 differentiation and was also able to support limited Th2 differentiation in the absence of STAT6 or IL-4Rα

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

(33). STAT5A1∗ 6 did not induce GATA3; rather, GATA3 was required for STAT5A1∗ 6induced IL-4 production (27, 33). Forced expression of both STAT5A1∗ 6 and GATA3 in Th1 cells additively activated Il4 expression through parallel cis-regulatory pathways (33, 78). IL-2/STAT5 signaling also favors Th2 over Th1 differentiation by increasing expression of SOCS3, an inhibitor of IL-12-induced STAT4 activation (79). Plasticity. In spite of these reinforcing mechanisms, differentiating T cells maintain a moderate degree of plasticity. Although commitment to expression of either Th1 or Th2 cytokines occurs rapidly, there seems to be some ﬂexibility and/or a delay in shutting down the opposite pathway (43, 82). Put another way, the process that leads to high-level expression of cytokine genes seems to be faster and more efﬁcient than the converse process of cytokine gene silencing. This characteristic is especially apparent in human Th1 and Th2 cells, which appear to be particularly receptive to changes in cytokine-mediated instructive signals in comparison with their counterparts in the mouse (82, 83). Culture of established IFN-γ-producing human Th1 clones under strong Th2-skewing conditions yielded cells that expressed high levels of both IFNγ and IL-4 (82). Thus, the original Th1inducing signal had not irreversibly silenced the Il4 gene, which remained capable of being activated by the second Th2-inducing signal; furthermore, the second Th2-inducing signal, which would have sufﬁced to depress IFNG expression if applied to naive T cells, was insufﬁcient to silence the previously activated IFNG gene in Th1 cells. Culture of Th2 clones under Th1-skewing conditions also produced cells that expressed both IFNγ and IL-4, although a subset of primary Th2 cells that expressed CRTH2 (chemoattractant receptor-homologous molecule expressed on Th2 cells) was resistant to induction of IFN-γ production. Similar effects have been observed even when cytokine signaling pathways are by-

passed by ectopic expression of T-bet and GATA3. Introduction of T-bet into differentiated Th2 cells suppresses IL-4 production to some degree, but the induction of IFN-γ is far more efﬁcient and results in the generation of many cells that produce both cytokines (83–85). GATA3 overexpression in Th1 cells produces the same effect (22, 23). Thus, immunomodulatory therapies that aim to reprogram established or ongoing helper T cell responses may be successful in inducing Th2 cytokines in Th1-mediated diseases (and vice versa), and this effect will likely be ameliorative in some cases. However, in cases in which repression of established cytokine expression is required, strategies for more generalized suppression of T cell responses or direct engagement of cytokine gene silencing mechanisms will likely be more successful than attempts to deviate the Th1/Th2 balance.

Maintenance of Th2 Differentiation and Cytokine Expression Massive reprogramming of gene expression occurs during helper T cell differentiation, with hundreds of genes differentially expressed in Th1 and Th2 cells (86–89). Many of these genes encode transcription factors, and some of these transcription factors directly regulate the Il4 and Ifng cytokine genes. For example, although the Th2-speciﬁc and Il4 promoter-binding transcription factor c-Maf is not required for Th2 differentiation, Th2 cells that lack c-Maf produce more than 100fold less IL-4 (75). It seems obvious that differential expression of transcription factors, cell surface receptors, and signaling molecules that transduce cytokine signals restrict the ability of developing helper T cells to change course and switch from a Th1 to a Th2 fate. Indeed, helper T cell differentiation can be modeled as a bistable system that simply requires a small push in one direction or another to set off feedback mechanisms that ultimately establish a new equilibrium in the form of resting Th1 or Th2 cells (90). www.annualreviews.org • Regulation of Th2 Differentiation

615

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Layered feedback mechanisms certainly contribute strongly to the establishment of polarized helper T cell responses, and Tbet and GATA3 play central roles in these feedback loops. It has been more difﬁcult, however, to determine to what extent these transcription factors are required for maintenance of Th1 and Th2 differentiation, or whether fully differentiated cells can “remember” their identities and function in the absence of continuous expression of the factors that directed their differentiation. GATA3 is not required to maintain Il4 expression. GATA3 is strictly required for both embryonic development and thymopoiesis. Therefore, conventional Gata3deﬁcient mice cannot be used to analyze the obligatory role of Gata3 in initiation and maintenance of Th2 differentiation. These obstacles have now been overcome through the use of conditional Gata3 deletion (26– 28). Cre-mediated deletion of ﬂoxed Gata3 alleles at the earliest stages of Th2 differentiation led to a very substantial (∼85%) reduction in the number of cells capable of producing IL-4; this number may be an underestimate, since loss of Gata3 also selectively impaired proliferation of Th2 cells, such that prolonged culture led to accumulation of cells bearing undeleted alleles. In contrast, deletion of Gata3 in fully differentiated Th2 cells led to a ∼60% decrease in the average amount of IL-4 produced per cell without affecting the total number of cells capable of producing IL-4. Thus, Gata3 is not absolutely essential for IL-4 production by differentiated Th2 cells, but it is important to ensure optimal levels of IL-4 production per cell, potentially because NFAT-GATA3 cooperation is essential for activity of the 3 Il4 enhancer [DNase I hypersensitivity site (HS) VA ] (91). Despite the minor effect on Il4 expression, acute deletion of Gata3 in differentiated Th2 cells led to pronounced reduction of Il5 and Il13 expression, consistent with the fact that both the Il5 and Il13 promoters contain functional GATA3 binding sites (31, 32). 616

Ansel et al.

Human patients with only one functional copy of GATA3 express reduced levels of GATA3 protein and also display both deﬁciencies in Th2 responses in vitro and reduced Th2 frequency and serum IgE levels in vivo (92). Overexpression of a dominant-negative T-bet protein, constructed by fusing the DNA-binding domain of T-bet with the transcriptional repression domain from Drosophila Engrailed, suggested that T-bet may also become dispensable for Ifng expression after Th1 differentiation is established (73). This ﬁnding is somewhat surprising, as T-bet directly binds the Ifng promoter and an upstream enhancer and acts synergistically with NFAT to activate Ifng transcription (93). Conditional deletion of T-bet would facilitate the precise delineation of T-bet requirements for maintenance of Th1 differentiation, Ifng expression, and Th2 cytokine repression. Epigenetic memory. If the differentially expressed transcription factors that establish Th1 and Th2 differentiation are not necessary to maintain Il4 activation or silencing, how do these cells “remember” whether or not they should transcribe Th2 cytokine genes? As discussed in the following sections, the answer lies in the way those genes are packaged in chromatin. The regulation of chromatin structure, or epigenetic regulation, confers heritable stabilization of gene expression patterns and thereby plays a major role in determining and maintaining cell differentiation. Differential accessibility at cis-regulatory elements appears to underlie stochastic as well as heritable variations in the ability of restimulated Th2 cells to produce IL-4 (94, 95) and has also been invoked to address the often contentious issue of mono-allelic versus bi-allelic expression of Th2 cytokine genes (96–98). As these complex phenomena can be investigated in primary cell populations stimulated under physiological conditions, the Il4 locus has become a popular model system for exploring the mechanisms of epigenetic regulation and long-range cis-regulatory control of gene transcription and silencing (6–9).

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

EPIGENETIC REGULATION: GENERAL ASPECTS Gene expression in eukaryotic cells is determined not only by interactions between transcription factors and DNA sequences in promoters and other cis-regulatory regions, but also by epigenetic changes in chromatin structure. These changes include DNA methylation, covalent posttranslational modiﬁcation of nucleosomal histones, incorporation of linker histones and core histone variants, and remodeling of nucleosome structure and position on the chromosome. Epigenetic changes control the accessibility of cis-regulatory elements—promoters, enhancers, insulators, and silencers—to transcriptional regulators and the RNA polymerase machinery. Conversely, transcription factors orchestrate epigenetic changes by binding cis-regulatory elements and recruiting chromatin-remodeling complexes and histone-modifying enzymes. Epigenetic changes can occur over remarkably long distances, and often spread or reach over chromosomal domains spanning hundreds of kilobases both 5 and 3 of genes. The fundamental unit of chromatin is the nucleosome, an octamer of histones that is subject to acetylation, phosphorylation, methylation, and ubiquitylation catalyzed by several families of histone-modifying enzymes (see Table 1 or follow the Supplemental Material link from the Annual Reviews home page at http://www.annualreviews.org for online resources). These modiﬁcations form a histone code that constitutes an important epigenetic determinant of the transcriptional state (99). Individual histone modiﬁcations can be characterized as “permissive” or “repressive” for locus accessibility and transcriptional activity. However, in employing these terms one should remain mindful that modiﬁcations can have distinct effects when they are presented in combination, and that other exceptions and new discoveries complicate any simple classiﬁcation system for histone modiﬁcations. For example, methylation of His-

tone (H) 3 Lysine (K) 9, a well-established characteristic of the silenced heterochromatin in which centromeres and telomeres reside, was recently also shown to be associated with transcriptional elongation within active genes (100). The histone code is translated into transcriptional states by chromatin-binding proteins that are able to modulate gene expression in both a short-term and a long-term manner (101). Proteins containing bromodomains and chromodomains recognize acetylated and methylated histones, respectively, whereas SANT domain-containing proteins have the ability to bind unmodiﬁed histone tails (102). Many chromatin-binding proteins are themselves chromatin-modifying enzymes or participate in large multiprotein complexes that include chromatin-modifying enzymes. Thus, the establishment of a covalent mark on a nucleosome can engage a cascade of epigenetic regulation that can spread to adjacent nucleosomes and form active or silenced chromatin domains. These cascades can be initiated (or reversed) by sequence-speciﬁc transcription factors that recruit chromatin-modifying complexes to discrete loci.

The Establishment and Maintenance of Permissive Histone Modifications Histone acetylation is catalyzed by several families of histone acetyltransferases (HATs), some of which have long been studied for their role as transcriptional coactivators (103, 104) (Table 1). Among the bromodomaincontaining factors that bind to acetylated histones are ATP-dependent chromatin remodeling complexes that can increase the accessibility of the nucleosomal DNA for further binding of transcription factors, including the RNA polymerase machinery (105). Histones can also be modiﬁed on lysine residues with up to three methyl groups (106, 107). Lysine methylation is mediated by the SET domain, which is named after the three main classes of histone www.annualreviews.org • Regulation of Th2 Differentiation

617

ANRV270-IY24-20

Table 1

ARI

23 February 2006

15:32

Histone modifying enzymesa

Enzyme activity/family

HUGO gene symbol

Histone Acetyltransferasesb Gnat family

GCN5L2

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

PCAF

Myst (SAS/MOZ) family

Ansel et al.

GCN5 general control of amino-acid synthesis 5-like 2

CAF GCN5 GCN5L P/CAF

p300/CBP-associated factor

histone acetyltransferase 1

ELP3

elongation protein 3 homolog

HTATIP

TIP60 cPLA2 HTATIP1

HIV-1 Tat interacting protein (60kDa)

MYST1 MYST2

MOF HBO1 HBOA MOZ ZNF220 RUNXBP2 MORF MOZ2

MYST histone acetyltransferase 1 MYST histone acetyltransferase 2

CREBBP

CBP

CREB binding protein [Rubinstein-Taybi syndrome]

EP300

P300

E1A binding protein p300

MYST4

618

Gene description

HAT1

MYST3

p300/CBP family

Gene aliases (alternate gene symbols)

Specificity H3:K9 H3:K14 H3:K18 H3:K23 H3:K27 H4:K8 H4:K16 H3:K14 H4:K8

H4:K5 H4:K12 H3:K14 H4:K8 H2A:K5 H3:K14 H4:K5 H4:K8 H4:K12 H4:K16 H4:K16

MYST histone acetyltransferase (monocytic leukemia) 3 MYST histone acetyltransferase (monocytic leukemia) 4 H2A:K5 H2B:K12 H2B:K15 H2B:K20 H3:K14 H3:K18 H3:K23 H4:K5 H4: K8 H2A:K5 H2B:K12 H2B:K15 H2B:K20 H3:K14 H3:K18 H3:K23 H4:K5 H4:K8 (Continued )

ANRV270-IY24-20

ARI

Table 1

23 February 2006

(Continued )

Enzyme activity/family Basal transcription factors

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

15:32

Nuclear receptor cofactors (SRC/p160 nuclear receptor coactivator family)

HUGO gene symbol

Gene aliases (alternate gene symbols)

TAF1

CCG1 TAFII250

GTF3C4

TFIIIC90 TFIIICdelta

NCOA1

SRC1 RIP160 ACTR AIB1 RAC3 SRC3 pCIP

NCOA3

Gene description

Specificity

TAF1 RNA polymerase II, TATA box binding protein (TBP)-associated factor (250kDa) general transcription factor IIIC, polypeptide 4 (90kDa)

H3:K14

nuclear receptor coactivator 1

H3:K9 H3:K14

H3:K14

nuclear receptor coactivator 3

Histone deacetylases Type 1 HDAC subfamily

HDAC11 HDAC1 HDAC2 HDAC3 HDAC8

Type 2 HDAC subfamily

HDAC4

HDAC5 HDAC6 HDAC7A HDAC9

HD1 RPD3L1 YAF1 HD3 HDACL1 HD4 HDACA HA6116 NY-CO-9 HD5 HD6 JM21 HDAC7 HD7 MITR HDAC7B

HDAC10 Type 3 HDACs (sirtuin family)

Histone methyltransferases N-Lysine histone methyltransferases

histone deacetylase 11 histone deacetylase 1 histone deacetylase 2 histone deacetylase 3 histone deacetylase 8 histone deacetylase 4

histone deacetylase 5 histone deacetylase 6 histone deacetylase 7A histone deacetylase 9

histone deacetylase 10

SIRT1

SIR2L1

SIRT2

SIR2L SIR2L2

SIRT3

SIR2L3

ASH1L

ASH1

NSD1

STO SOTOS ARA267

sirtuin (silent mating type information regulation 2 homolog) 1 sirtuin (silent mating type information regulation 2 homolog) 2 sirtuin (silent mating type information regulation 2 homolog) 3 ash1 (absent, small, or homeotic)-like [Sotos syndrome and Weaver syndrome] nuclear receptor binding SET domain protein 1

H3:K4 H3:K9 H4:K20 H3:K36 H4:K20 (Continued )

www.annualreviews.org • Regulation of Th2 Differentiation

619

ANRV270-IY24-20

Table 1

ARI

23 February 2006

15:32

(Continued )

Enzyme activity/family

HUGO gene symbol

Gene aliases (alternate gene symbols)

EHMT1

GLP

EHMT2

SET7

G9A BAT8 NG36 ESET KG1T CLLD8 CLLL8 SET9

DOT1L

DOT1

SUV420H1

CGI-85

SETDB1

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

SETDB2

SUV420H2

Gene description

Specificity

euchromatic histone-lysine N-methyltransferase 1 euchromatic histone-lysine N-methyltransferase 2

H3:K9

SET domain, bifurcated 1

H3:K9

H3:K9 H3:K27

SET domain, bifurcated 2 SET domain-containing protein 7 DOT1-like, histone H3 methyltransferase suppressor of variegation 4-20 homolog 1 suppressor of variegation 4-20 homolog 2 SET and MYND domain containing 3

H3:K4 H4:K20 H3:K79 H4:K20 H4:K20 H3:K4

SMYD3

ZMYND1 ZNFN3A1

PR/SET subfamily

SET8

SET07

PR/SET domain containing protein 8

H4:K20

SUVAR3-9 subfamily

SUV39H1

SUV39H

suppressor of variegation 3-9 homolog 1 suppressor of variegation 3-9 homolog 2

H3:K9

SUV39H2

H3:K9

TRX/MLL family

MLL

HRX TRX1 CXXC7 HTRX1 MLL1A.

myeloid/lymphoid or mixed-lineage leukemia

H3:K4

EZ family

EZH1 EZH2

ENX2 ENX1

enhancer of zeste homolog 1 enhancer of zeste homolog 2

H3:K27 H3:K9 H3:K27 H1:K26

N-Arginine histone methyltransferases

HRMT1L2

ANM1 HCP1 IR1B4 PRMT1 PRMT4

HMT1 hnRNP methyltransferase-like 2

H4:R3

coactivator-associated arginine methyltransferase 1

SKB1

JBP1 IBP72 PRMT5

SKB1 homolog

H3:R2 H3:R17 H3:R26 H3:R8

AOF2

LSD1

amine oxidase (ﬂavin containing) domain 2

CARM1

Histone demethylases Flavin monoamine oxidase family

H3:K4 (Continued )

620

Ansel et al.

ANRV270-IY24-20

ARI

Table 1

23 February 2006

15:32

(Continued )

Enzyme activity/family Protein arginine deiminase family Histone ubiquitylation Ubiquitin-conjugating enzyme family

HUGO gene symbol

Gene aliases (alternate gene symbols)

PADI4

PAD PADI5

peptidyl arginine deiminase, type IV

UBE2A

UBC2 HHR6A RAD6A HR6B UBC2 RAD6B

ubiquitin-conjugating enzyme E2A (RAD6 homolog)

H2A H2B

ubiquitin-conjugating enzyme E2B (RAD6 homolog)

H2A H2B

P18 UBC9

ubiquitin-conjugating enzyme E2I (UBC9 homolog, yeast)

H4

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

UBE2B

Histone sumoylation Ubiquitin-conjugating enzyme family

UBE2I

Gene description

Specificity

a For an online version of this Table, with links to UniProt, NCBI Entrez Gene database, and the UCSC genome browser for each human and murine histone modifying enzyme, follow the Supplemental Material link from the Annual Reviews home page at http://www.annualreviews.org. b Enzyme activity and protein family classiﬁcations were made according to the BRENDA comprehensive enzyme information system database (www.brenda.uni-koeln.de), and UniProtKB/SwissProt database (www.ebi.ac.uk/swissprot/), respectively.

methyltransferases (HMTs) found in Drosophila: Su(var)3-9, Ez, and Trithorax, which methylate H3K9, H3K27, and H3K4, respectively. H3K27 trimethylation and H3K9 di- and trimethylation correlate with gene silencing, as we discuss below. H3K4 dimethylation is broadly distributed throughout the genome at transcriptionally active genes, inactive genes that are kept in a transcriptionally permissive state, and intergenic sites, including regulatory elements such as distal enhancers and insulators (108– 110). H3K4 dimethylation is therefore generally considered a permissive modiﬁcation, although it is also observed in combination with H3K9 and H4K20 methylation in the silenced chromatin of centromeres. Trimethylation of H3K4 correlates closely with transcriptional activity and is found primarily at the 5 ends of active genes together with RNA polymerase II and H3K4 methyl transferases such as the yeast Set1 and MLL1 (110, 111). In addition to lysines, histone arginines can be methylated and dimethylated by a separate class of HMTs (112) (Table 1). Arginine methylation in H3 by CARM1 and H4

by PRMT1 occurs at promoters during transcriptional activation but is reversed upon disengagement of RNA Pol II from the promoters. Arginine demethylation is catalyzed by peptidyl arginine deiminase PADI4, which deiminates arginine to citrulline (113). In addition to histones, arginine methyl transferases can regulate transcription by methylating other proteins such as the transcriptional coactivator CBP/p300 as well as some transcriptional elongation factors (112). In helper T cells, arginine methylation of the cofactor NIP45 facilitates its interaction with NFAT and thereby regulates the expression of NFAT-dependent cytokine genes, including Ifng and Il4 (114). HATs and HDACs can also affect transcription by modifying other proteins, including the transcription factors BCL6, Runx, and p53, and thereby altering their stability and/or transcriptional potential (104, 115–118). Other histone modiﬁcations include phosphorylation and ubiquitylation, which are implicated in modulating methylation and acetylation. Ubiquitylation of H2BK123 by the ubiquitin-conjugating enzyme Rad6 is required for telomeric gene silencing by www.annualreviews.org • Regulation of Th2 Differentiation

621

ARI

23 February 2006

15:32

priming H3K4 and K79 methylation, two modiﬁcations that regulate silencing in yeast (119–121). Interestingly, ubiquitylation was not required for monomethylation by Set1 and Dot1 but rather for the processivity of methyltransferases and formation of diand trimethyl states (122). Phosphorylation of H3S10 interferes with the repressive effect of methylation on the adjacent H3K9 residue and inhibits binding of the chromodomain protein HP1 (123). Nucleosomes are also modiﬁed by the incorporation of histone variants and linker histones (124). It is evident that additional studies will be needed to ﬁnd and interpret other combinations of histone modiﬁcations to decipher the histone code. Some HATs and H3K4 HMTs contain bromodomains, which bind acetylated histones and thus provide a means for perpetuating the activating mark and a mechanism for stable locus activation. However, histone acetylation is readily reversed by histone deacetylases (HDACs), which are common components of corepressor complexes. In yeast, dimethylated H3K4 can recruit the HAT SAGA through its interaction with the chromodomain protein CHD1 (125). Methylation is a less labile modiﬁcation than histone acetylation, but the recent discovery of LSD1, an enzyme capable of oxidatively demethylating dimethyl- and monomethylH3K4, has introduced further complexity into the dynamics of histone lysine modiﬁcation (126). LSD1 and continuous removal of H3K4 dimethylation are required to maintain repression of neuronal-speciﬁc genes in HeLa cells, underscoring the vital importance of H3K4 dimethylation in controlling gene activity.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ATP-Dependent Chromatin Remodeling Complexes Accessibility of DNA is also regulated by chromatin-remodeling complexes that use the energy of ATP hydrolysis to remove and assemble nucleosomes, to slide them along the DNA strand, or to create loops of DNA on the 622

Ansel et al.

nucleosome surface (105, 127). Like histonemodifying enzymes, chromatin-remodeling complexes are recruited to DNA via their chromatin-binding domains or by sequencespeciﬁc transcription factors. There are several classes of chromatin remodelers, including the SWI/SNF (BAF), ISWI, INO80, SWR, and Mi-2/CHD, each specialized for a different function that can positively or negatively regulate transcription (127). The active ATPase motors of these large complexes associate with a variety of other proteins that regulate their activity and help target them to chromatin. Chromatin remodeling plays critical and speciﬁc roles in gene regulation at various stages of T cell development, as evidenced by both thymic and peripheral abnormalities of T cells deﬁcient in Brg1 (the ATPase subunit of the BAF complex) and by the disregulation of CD4 gene silencing in T cells carrying a targeted mutation of the high mobility domain of the DNA binding subunit BAF57 (128, 129). The role of ATPdependent chromatin remodelers in Il4 locus regulation during Th cell differentiation has yet to be investigated.

Repressive Histone Modifications and the Formation of Silent Chromatin Complex multicellular organisms such as ourselves require complex yet precise spatial and temporal regulation of gene expression to create and maintain the diversity of cell types that constitute the organs and tissues of our bodies. Gene repression is as important as gene activation in this process, and more than half of the genome is kept in a silent state in any given cell type (other than embryonic stem cells). Chromatin structure and histone modiﬁcations play a vital role in mediating this repression. Although several ways of establishing and maintaining silent chromatin (generally termed heterochromatin) have been described, each requires histone methylation at H3K9 and/or H3K27, as well as histone deacetylation (106, 107).

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

Analogous to the way in which HATs create binding sites for bromodomain-containing proteins, HMTs for H3K9 and H3K27 create binding sites for chromodomaincontaining proteins such as HP-1 (heterochromatin protein-1) and chromodomaincontaining protein complexes such as NURD (nucleosome-remodeling deacetylase) and Polycomb, all of which are implicated in transcriptional repression and the formation of heterochromatin (106, 107, 130, 131). Heterochromatin can be constitutive, such as that found in pericentromeric regions and telomeres where it remains condensed throughout the organism’s lifespan, or facultative, such as that formed during X inactivation and on genes silenced during development. In the ﬁssion yeast S. pombe, silencing of the mating-type region is initiated at a discrete site that undergoes deacetylation of histones H3 and H4, methylation of H3 at K9, and recruitment of HP-1 to methylated histones via its chromodomain (132). Repressive modiﬁcations then spread by sequential methylation and binding of Clr4, the yeast homolog of Su(var)3–9 (a chromodomain-containing H3K9 methyltransferase), and eventually lead to chromatin compaction. This process generates a structure that is relatively inaccessible to transcriptional activators, remodelers, and the RNA polymerase machinery. The assembly of heterochromatin at Dntt, a mammalian gene that encodes TdT, in double-positive thymocytes is remarkably similar. Loss of H3K4 methylation and acetylation precedes the acquisition of H3K9 methylation at the Dntt promoter; these modiﬁcations spread outward from the promoter and are accompanied by intranuclear repositioning of the gene to areas of centromeric heterochromatin (133, 134). The key difference between mating-type locus silencing in yeast and Dntt silencing in thymocytes is in how the silencing is initiated. In the mating-type region, repetitive DNA sequences homologous to centromeric repeats serve as originating points for bidirectional noncoding transcripts that are pro-

cessed into short interfering RNAs (siRNAs) that target the RNA-induced transcriptional silencing (RITS) complex and its associated histone deacetylases and methyltransferases (132, 135–137). RNAi is not required for Dntt silencing, as this process is unaffected in thymocytes lacking Dicer, the enzyme that processes double-stranded RNA precursors into siRNAs (138). The mechanism by which Dntt silencing is initiated may instead involve the binding of the heterochromatinassociated transcription factor Ikaros to the Dntt promoter (139). However, a redundant role for RNAi is not ruled out; heterochromatin formation at the mating-type locus is partially preserved in the absence of RNAi through a parallel pathway that involves direct locus binding by ATF/CREB family transcription factors (140). For developmentally regulated genes, the mechanism and durability of gene silencing can be determined by sequence-speciﬁc transcription factors that recruit different repressive chromatin-modifying activities to regulatory elements. Another instructive example is provided by the Runx proteins, which bind to an intronic CD4 silencer element to mediate two different forms of CD4 gene silencing at different stages of thymocyte development (141). In doublenegative thymocytes, Runx1 binding to the CD4 silencer is important for an (as yet) undetermined mechanism of active repression that depends on the continued presence of the silencer element. Later in thymocyte development, Runx3 binds to the CD4 silencer in double-positive thymocytes and mediates stable epigenetic silencing of the gene; this silencing persists even if the silencer element is subsequently removed. Runx proteins associate with histone deacetylases and the H3K9 methyltransferase Suv39h1, providing a possible link between sequencespeciﬁc transcription factors, repressive histone modiﬁcations, and developmentally regulated gene-speciﬁc heterochromatinization. Consistent with this possibility, repositioning of the CD4 gene to centromeric www.annualreviews.org • Regulation of Th2 Differentiation

623

ARI

23 February 2006

15:32

heterochromatin takes place early in the transition of double-positive to CD8 singlepositive thymocytes, though the requirements for Runx proteins and the CD4 silencer element for repositioning have not been tested (142, 143). Polycomb group (PcG) proteins were ﬁrst identiﬁed in Drosophila as negative regulators of homeobox genes during embryonic development when they work in tandem with positive regulators, Trithorax proteins, to ensure proper expression patterns of homeobox genes (144). Like HP-1, the founding member, Polycomb, has a chromodomain that binds methylated histones and is also a structural component of heterochromatin (145). PcG proteins assemble into at least two complexes, PRC1 (which contains HPC2, HPH, Bmi-1, and RING1) and PRC2 (which contains Eed, Suz12, and EZH2). Neither complex has intrinsic sequence speciﬁcity; however, sequence-speciﬁc factors recruit PcG complexes to regions called Polycomb responsive elements (PREs) in the regulatory regions of speciﬁc target genes (144, 146). EZH2 is a methyltransferase that methylates H3K27 (147–149), and this modiﬁcation recruits the PRC1 complex to PREs where it inhibits transcription and ultimately causes formation of a highly repressive chromatin structure by compaction of the nucleosomal array (150). EZH2 and H3K27 methylation are also able to cause gene repression in a PRC1-independent manner (151). No mammalian PREs have been identiﬁed so far, but it is clear from the phenotypes of mice deﬁcient in various PcG proteins that PcG complexes play an important role in mammalian development. PcG proteins are highly expressed in the lymphoid lineages, and various defects in B and T cell development result from disruption of PcG genes (152, 153).

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

DNA Methylation Status DNA methylation is an ancient adaptation used to distinguish an organism’s own DNA from that of invaders, such as viruses. In 624

Ansel et al.

eukaryotes, DNA methylation has further evolved into an important mechanism for restricting endogenous gene activity (8, 154). DNA methyltransferases (Dnmts) methylate most of the genome’s cytosines that are found in CpG dinucleotides, but clusters of CpGs known as CpG islands found at gene promoters are able to escape the methylation mark. DNA methylation patterns are faithfully propagated through cell division by the maintenance methyltransferase Dnmt1, which copies existing methylation patterns during DNA replication using the parental strand as a template. Other Dnmts are capable of de novo methylation of symmetrically unmethylated CpGs. Although no methylcytosine demethylases have been identiﬁed, their existence is predicted by rapid demethylation events that occur early in embryogenesis and in activated T cells (154, 155). Like methylated histones, methylated CpGs recruit proteins involved in gene repression and heterochromatin formation. Notable methyl CpG-binding proteins include MeCP2 and MBD2, which are found in corepressor complexes with HDACs, HMTs, and ATP-dependent chromatin remodelers (8). Although DNA methylation is often a late event in gene silencing processes, the synergy between DNA and histone methylation appears to be required to maintain heritable silent states.

EPIGENETIC REGULATION OF THE Il4 LOCUS General Considerations The complex processes of Th1 and Th2 differentiation are associated with elaborate epigenetic changes occurring across ∼200 kb of the extended Il4 locus (6–8). Functionally, these changes are responsible for regulated differences in accessibility of the Il4 and Ifng genes in naive versus differentiated T cells. For instance, both the Il4 and Ifng promoters are poised for rapid transcription in naive CD4+ T cells, and both are accessible to

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

binding of NFAT, an antigen-induced transcription factor (44). In contrast, NFAT binds only to the Il4 promoter in activated Th2 cells and only to the Ifng promoter in activated Th1 cells, in close correlation with the reciprocal activation and silencing of these genes (13, 44). Thus, as a result of the epigenetic changes occurring during helper T cell differentiation, the common transcription factor NFAT acquires a selective and cell-type-speciﬁc ability to activate the Ifng and Il4 genes in Th1 and Th2 cells, respectively, although it is expressed at similar levels in both cell types. The regulated accessibility of cytokine gene promoters is just the tip of the epigenetic iceberg. Although naive T cells and Th2 cells activate Il4, Il13, and Il5 transcription with similar kinetics, the peak steadystate level of transcripts and the resulting amounts of cytokine protein are dramatically greater in Th2 cells (43–45). This difference is partly explained by increased expression of lineage-speciﬁc transcription factors such as c-Maf and GATA3, which bind directly to cytokine gene promoters (32, 75). However, numerous additional cis-regulatory sequences contribute to the regulation of the Il4 locus (Table 2), including a recently discovered group of elements within the introns of Rad50, which encodes a ubiquitously expressed protein involved in DNA repair. Genetic analysis of these cis-regulatory elements is providing clues to their functional roles. In parallel, several aspects of chromatin structure, including DNA accessibility, histone modiﬁcation, and DNA methylation, are being monitored across the Il4 locus of naive, Th1, and Th2 cells. Together, these studies are beginning to provide mechanistic insights into the processes of regulated activation and silencing of the Il4 locus. Cis-regulatory elements. Using a variety of experimental and bioinformatic methods, a number of distal regulatory elements have been identiﬁed in the Il4 and Il13 genes (13, 14, 156–164) (see Figure 3 and Table 2). The “territory” of these two genes

encompasses a ∼85 kb region, which extends from the 3 introns of the Rad50 gene to the 3 end of the KIF3A gene. The chromatin structure of this entire region undergoes signiﬁcant changes during helper T cell differentiation (Figure 3). This region contains several candidates for discrete cis-regulatory regions, identiﬁed either bioinformatically as CNSs, or experimentally as regions showing increased accessibility (hypersensitivity) to DNase I in naive, Th1, and/or Th2 cells (see Figure 3 and Table 2). There is a strong concordance between CNS regions and the HS regions identiﬁed in the Il4 locus; this observation supports the notion that CNS/HS regions regulate gene activity in cis (14, 160, 164). Consistent with their coordinate expression in Th2 cells, mast cells, and basophils, the Il4 and Il13 genes share some cis-regulatory elements, but local control elements also make important contributions to the probability of expression as well as the expression level of each gene in any given cell type (Table 2; also see Mast Cells, below). The activities of many of the HS sites/CNS regions have been demonstrated or validated by germ line deletion or transgenesis in mice (44, 163, 165–168), and speciﬁc transcription factors that bind to the regions have been identiﬁed in many cases. HSS3 and HS IV are notable in that they are the only HSs observed in naive T cells (157), and targeted deletion of HS IV revealed that it is required for Il4 gene silencing in Th1 cells (44). Several conserved elements at the 3 end of the Rad50 gene constitute a Th2 locus control region (LCR) that, like the intergenic element CNS1, coordinately regulates expression of the Il4 and Il13 genes (156, 163, 166, 168). A number of other elements are selectively DNase I hypersensitive in Th2 cells and may function as local enhancers; of these, the best-documented are the 3 Il4 enhancer (HS VA ) and HS V/CNS2 (13, 157, 167). The properties of selected cis-regulatory regions are brieﬂy discussed in a subsequent section. www.annualreviews.org • Regulation of Th2 Differentiation

625

ANRV270-IY24-20

ARI

23 February 2006

Table 2

15:32

DNase I hypersensitivity (HS) sites and cis-regulatory activity in the Il4 locus

DNase I HS site(s)

Alternative name

HS patterna

RHS 5 RHS 6 RHS 6 RHS 7

Th2, Th1, MC Th2, Th1, MC Th2, Stimulated Th1, MC Stimulated Th2, MC

CGRE Il13 promoter

Th2, MC Th2, MC Th2, MC MC only

Effect of deletionb

3

Rad50 introns LCR O LCR A LCR B LCR C

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Il13 HS I HS II HS III MCHS Il13/Il4 intergenic HSS3 HSS2 and HSS1 HS 0 Il4 HS I — HS II

CNS1

Th2, Th1, naive T, MC Th2 Th2, MC

Il4 promoter CIRE Il4 intronic enhancer

Th2, MC None Th2, MC

Il4 silencer

Th2, Th1, naive T, MC

Il4 3 enhancer CNS2

Stimulated Th2, MC Th2, MC

↓IL-4/IL-13/IL-5 in Th2 cells ↓LCR/promoter interactions

↓IL-4/IL-13 in Th2 cells

HS III Il4/Kif3a intergenic HS IV HS VA HS V

↑IL-4 in naive T, Th1, Th2 cells, MC ↓IL-4/IL-13 in Th2, MCc

a

List of cell types that display DNase I hypersensitivity at each site. MC, mast cells. Changes in cytokine gene regulation in cells bearing deletions in the indicated regions, compared with wild-type cells. c HS VA and HS V were deleted in tandem. b

Histone modifications. Several groups have examined histone modiﬁcations in the Il4 locus of naive, Th1, and Th2 cells. The analyses hint at a multifaceted mechanism of epigenetic regulation during Th differentiation, involving many conserved cis-regulatory elements spread throughout the locus. In naive T cells, the locus is poised for early transcription of the Il4 and Il13 genes. Histone code proﬁling of the Il4 locus in these cells, with some bias toward known or suspected cis-regulatory elements (e.g., HSs and conserved regions), revealed a low level of basal H3K9/14 acetylation at the Il4 promoter, a peak of H3K9/14 acetyla626

Ansel et al.

tion and H3K4 dimethylation at HS IV, and a prominent peak of H3K4 dimethylation without H3K9/14 acetylation at HS V (91, 162, 169–171) (Figure 3). HS V is also unusual for being DNA hypomethylated in naive T cells and becoming de novo methylated during Th1 differentiation (172). Unexpectedly, naive T cells also display a silencing modiﬁcation, H3K27 trimethylation, weakly at HS IV and more prominently at HSS3 (173). The H3K27 methyltransferase EZH2 is also associated with both of these sites in naive T cells. Presumably, these elements bind trans-acting factors that maintain these epigenetic modiﬁcations and may be responsible

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

Figure 3 Chromatin structure of the Il4 locus. DNase I hypersensitivity sites (HS) and regions associated with covalently modiﬁed histone-3 are summarized for naive T, Th1, Th2, and mast cells. HS names and positions are provided at the top of the ﬁgure, and the sites that are accessible in each cell type are indicated in the HS pattern tracks below (red bars). Asterisks indicate regions that become DNase I hypersensitive only upon stimulation. Association with histone-3 acetylated on lysine-9 and -14 (AcK9/14, black bars), dimethylated on lysine-4 (Me2K4, blue bars), or trimethylated on lysine-27 (Me3K27, brown bars) was determined by chromatin immunoprecipitation (data are from References 169, 173; bars indicate regions found to be enriched above an arbitrary threshold, and darker shading indicates an increased level of association with the modiﬁed histones). Only regions speciﬁcally measured by quantitative PCR are shown. A data annotation ﬁle and instructions for viewing an interactive form of this ﬁgure in the UCSC Genome Browser are provided as supplemental online materials. Further explanation of the interspecies sequence conservation tracks are also available. Follow the Supplemental Material link from the Annual Reviews home page at http://www.annualreviews.org.

for the poised state of the locus in naive T cells. The early Il4 transcription that occurs upon activation of naive T cells is accompanied by the appearance of permissive chromatin modiﬁcations and increased accessibility throughout the Il4 locus (14, 91, 169, 171). If the cells are activated under Th2 conditions, this epigenetic state is stabilized, whereas under Th1 conditions, the permissive modiﬁcations are erased and replaced with repressive modiﬁcations (173). Only the Il4 silencer at HS IV remains associated with the permissive histone modiﬁcations, H3K9/14 acetylation, and H3K4 dimethylation, even in Th1 cells (Figure 3), suggesting that transacting factors bound to this region repress transcription from the Il4 promoter despite recruit-

ing permissive histone-modifying enzymes (44, 169). In Th2 cells, a domain of abundant H3K9/14 acetylation, H3S10 phosphorylation, and H3K4 dimethylation extends from HS V/CNS2, an enhancer element located ∼6.5 kb 3 of the Il4 gene, to HS I/CGRE (conserved GATA3 response element), a conserved GATA3-binding enhancer located 1.6 kb 5 of the Il13 promoter (162, 169) (see Figure 3 and Table 2). Especially strong signals were observed at the Il4 and Il13 promoters and intronic/coding regions, and at CNS1 and HS V. The actual territory of the Il4 and Il13 genes extends beyond the CGRE, however, to introns at the 3 end of Rad50 (14, 160). An interesting question is whether HS V, the CGRE, or elements within Rad50 www.annualreviews.org • Regulation of Th2 Differentiation

627

ARI

23 February 2006

15:32

actively maintain the boundaries of this permissive chromatin domain, as does HS4, a heavily acetylated insulator element in the chicken β-globin locus (174). Chromatin mapping in T cells has been extended beyond the Il4 locus, and efforts to describe the entire “epigenome” have started to bear fruit (172, 175, 176). For example, random cloning and sequencing of HSs in naive CD4+ T cells revealed a higher than expected frequency of putative cisregulatory sequences downstream of proteincoding genes (175). Interestingly, binding of Sp1 and Myc on chromosomes 21 and 22 in Jurkat T cells was also enriched immediately downstream of genes (177). Regions of H3K9/14 acetylation in naive CD4+ T cells have also been randomly cloned and sequenced from chromatin immunoprecipitations on a large scale (172). These early reports have already provided valuable data and a ﬁrst glimpse of the epigenome, but the power of chromatin mapping will only be fully realized as data are compiled for comparison between different types of analyses (e.g., DNase HS and histone modiﬁcations), different cell types and conditions, and a large number of genomic analytes at high enough resolution to ensure that isolated islands of modiﬁed chromatin do not go undetected (109). Because of the large number of analytes and conditions to be tested, development of reliable, affordable high throughput systems for epigenetic mapping will be key to the success of this effort (176).

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

DNA methylation status. The role of DNA methylation in Il4 gene regulation has been extensively studied (178–181; reviewed in 8). Much of the Il4 locus is heavily methylated (>90% of all CpGs) in naive T cells; exceptions are the Il4 promoter, in which ∼60% of all CpGs are methylated, and HS V/CNS2, where quantitative estimates have not yet been made. In T cells activated under nonpolarizing conditions, CpGs that become hemi-methylated during S phase are symmetrically remethylated by the mainte628

Ansel et al.

nance DNA methyltransferase Dnmt1; however, if the cells are cultured under Th2 conditions, the level of Dnmt1 recruitment to the Il4 locus is substantially decreased (180). Consistent with decreased Dnmt1 recruitment, DNA methylation at the 5 end (ﬁrst and second introns) of the Il4 gene is progressively lost in T cells differentiating toward the Th2 lineage, in a passive process that requires cell division (179) (Figure 4). This passive demethylation is not required for chromatin remodeling or initial transcription of the Il4 gene, but it is required for efﬁcient, high-level Il4 transcription by differentiated Th2 cells (161, 179). Early Il4 expression in naive T cells is strongly inﬂuenced by the level of basal methylation of the Il4 gene and the associated recruitment of methyl-CpG-binding proteins. T cells lacking the maintenance methyltransferase Dnmt1 exhibited substantially decreased DNA methylation across the Il4 locus and showed increased production of IL-4 under all culture conditions, especially in the CD8 lineage (182, 183). Th1 cells lacking the methyl-CpG-binding protein MBD2 produced IL-4 inappropriately, and whereas naive T cells normally produce very little IL4 before the ﬁrst cell division (during which DNA demethylation is initiated), naive T cells from Mbd2-deﬁcient mice produced substantial amounts (184). Increased IL-4 expression in Dnmt1−/− and Mbd2−/− T cells was not associated with increased levels of GATA3, and the cells responded appropriately to polarizing conditions by enhancing and repressing production of the relevant opposing cytokines (183, 184). There is evidence for competition between Il4 locus activation by GATA3 and the repressive effects of MBD2: In wild-type T cells overexpression of GATA3 correlated with decreased MBD2 binding at Il4 intron 2 and CNS1 and in Mbd2−/− T cells, overexpressed GATA3 promoted higher levels of IL-4 expression than in Mbd2+/+ Th1 cells (184). These data need to be related to the ﬁnding that a short conserved intronic regulatory

ANRV270-IY24-20

ARI

23 February 2006

15:32

Naive

exon 1

2

3

4

2

3

4

2

3

4

Il4

Early Th2

exon 1 GATA3

Il4

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Late Th2

exon 1

GATA3

Progressive demethylation

Il4

Figure 4 Progressive DNA demethylation of the Il4 gene during Th2 differentiation. In naive T cells, the DNA of the Il4 gene is methylated at CpG dinucleotides, and methyl-binding domain proteins recruit repressor complexes that inhibit IL-4 transcription (CpG and bound proteins are shown as purple ovals). During Th2 differentiation, DNA methylation is progressively lost from 5 to 3 along the Il4 gene in a passive process that requires cell division. Demethylation may be facilitated by GATA3, which occupies an originally heavily methylated element (CIRE) in the ﬁrst intron of the Il4 gene and may compete with MBD2 for binding or otherwise interfere with Dnmt1 function. Alternatively, demethylation of the CIRE may permit binding of GATA3. For details see text.

element (CIRE) in the ﬁrst intron of the Il4 gene binds GATA3, and that CpGs surrounding this element undergo rapid and Th2-selective demethylation during Th2 differentiation (161). One possibility is that binding of GATA3 at the CIRE excludes Dnmt1 and thereby promotes early demethylation at the 5 end of the Il4 gene; however, this simple explanation is ruled out by the ﬁnding that overexpression of GATA3 did not interfere with maintenance methylation or promote passive demethylation of the Il4 gene (184). Rather, CIRE demethylation through a different mechanism may be permissive for the early binding of GATA3 and thereby potentiate Th2 differentiation (Figure 4).

Properties of Cis-Regulatory Elements in the Il4 Locus A simple functional classiﬁcation of cisregulatory elements in the Il4 locus is precluded by their diversity and by the limited availability of genetic data regarding their functional roles. The classiﬁcation used in this

section is based on the pattern of accessibility of the elements to DNase I (13, 14, 157– 160, 185) (see Figure 3 and Table 2). Two sites, HS IV and HS S3, are accessible in naive T cells, where they display unusual patterns of histone modiﬁcation; these sites continue to be accessible in both Th1 and Th2 cells. The large majority of sites, however, are not DNase I hypersensitive in naive T cells, but become selectively accessible during Th2 differentiation. Sites contained within the Th2 LCR, located at the 3 end of the Rad50 gene, are considered separately, since they form a distinct functional unit. Although none of the sites in the Th2 LCR are DNase I hypersensitive in naive T cells, three of the major elements are accessible in both Th1 and Th2 cells, with only the 3 -most element (LCRC) being DNase I hypersensitive only in Th2 cells (14, 160). There is evidence that the different cisregulatory elements in the Il4 locus bind distinct but overlapping combinations of trans-acting factors. In chromatin immunoprecipitation assays, GATA3 is not detected at the Il4 promoter but does bind strongly www.annualreviews.org • Regulation of Th2 Differentiation

629

ARI

23 February 2006

15:32

to the Il5 and Il13 promoters and to the Il13 CGRE and Il4 CIRE (13, 31, 32, 161, 162). GATA3 is also recruited to HS VA and LCR-C in stimulated Th2 cells (13, 186). NFAT and STAT6, which translocate to the nucleus following stimulation with antigen and IL-4, respectively, bind to the Il4 promoter, LCR-B, LCR-C, and HS VA (14, 91). Consistent with the distinct actions of IL2-induced STAT5 and IL-4-induced STAT6 and GATA3 during Th2 differentiation (Figure 1), STAT5a binding is uniquely detected at the intronic enhancers at Il4 HS II and HS III (33). Two important questions for the future are (a) how the binding of these diverse transcription factors is orchestrated and (b) how they contribute to the long-range interactions among regulatory elements that have been detected in the Il4 locus (186). These questions are addressed further below.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

Sites that are DNase I hypersensitive in naive, Th1, and Th2 cells. Two regions in the Il4 locus, HSS3 and HS IV, stand out for their high degree of accessibility in naive T cells, which transcribe the Il4 and Il13 genes only at low levels, as well as in Th1 cells, which silence the Il4 and Il13 genes. These two regions are also DNase I hypersensitive in Th2 cells, but are completely inaccessible in unrelated cells such as ﬁbroblasts (157, 159). HSS3. The function of HSS3 has yet to be determined by targeted deletion. Presumably, the function will be consistent with the diverse changes in histone modiﬁcations observed in this region. HSS3 is a strong focus of H3K27 trimethylation and EZH2 binding in naive T cells, and both properties are maintained in Th1 cells (173). Although the H3K27 trimethylation is gradually lost with progressive Th2 differentiation, EZH2 binding is maintained (alternatively, the kinetics of loss of EZH2 binding may be delayed with respect to loss of H3K27 methylation). HSS3 is also a focus for activation-dependent changes: The level of H3K4 acetylation at HSS3 and an adjacent region, CNS1, increased markedly 630

Ansel et al.

after activation of human CD3 T cells with anti-CD3 and anti-CD28 (172). HS IV. Targeted deletion of HS IV in mice indicates that HS IV contains an Il4 silencer (44). Loss of HS IV derepresses Il4 expression in all cell types that display DNase I hypersensitivity at this site—naive T cells, Th1 and Th2 cells, and mast cells. HS IV−/− Th1 cells expressed high levels of both Il4 and Ifng in vitro and in vivo. Although HS IV-deﬁcient mice mounted a vigorous Th1 response with normal levels of IFN-γ production, they failed to control infection with Leishmania major, indicating that Il4 silencing is a critical component of Th1-mediated immunity. The mechanism of HS IV-mediated silencing remains unclear. One clue derives from the ﬁnding that DNase I hypersensitivity and permissive histone modiﬁcations are maintained at HS IV even in Th1 cells that silence the Il4 gene (44, 157, 169). This site also harbors a small peak of H3K27 trimethylation in naive T cells that increases with Th1 differentiation and spreads into the neighboring HS VA and HS V regions (173). Potentially, HS IV could function as an “enhancer blocker” and interfere with communication between the Il4 promoter and the 3 Il4 enhancers at HS VA and HS V. The high levels of permissive histone modiﬁcations at HS IV recall the HS4 element in the chicken βglobin locus, a potent enhancer blocker and chromatin insulator that also stands out as a peak of permissive histone modiﬁcations in erythroid cells (174, 187). The mechanisms that contribute to Il4 gene silencing are discussed in more detail in a subsequent section. Sites that are selectively DNase I hypersensitive in Th2 cells. Th2 differentiation is accompanied by increased accessibility at several cis-regulatory elements that contribute to the transcriptional activation of Th2 cytokine genes. These elements include the Il4 and Il13 promoters, the Il4 intronic enhancer, and the Il4 3 enhancer. These elements have in

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

common the fact that they display a selectively increased accessibility to DNase I and restriction enzymes in Th2 cells, relative to naive T cells and Th1 cells (13, 157–159). LCR-C also shares these properties but is discussed in a separate section below. Promoters. The Il4 and Il13 promoters both appear as DNase I hypersensitive sites in Th2 cells, but not in Th1 cells or naive T cells (157, 158). The promoters are promiscuous in that they each bind many different transcriptional regulators: The Il4 promoter binds NFAT, AP-1 (especially JunB), c-Maf, and STAT6, whereas the Il13 promoter binds NFAT, AP-1, and GATA3. These transcription factors tend to synergize strongly in reporter assays, with combinations of factors yielding greater transcriptional activity than individual factors alone. Both the Il4 and Il13 promoters (as well as many other regions of the locus) show low or undetectable levels of histone modiﬁcations in naive T cells, but rapidly acquire the permissive modiﬁcations H4 acetylation and H3K9/K14 acetylation upon activation (91, 162, 169–171). Detailed kinetic analyses suggest that H4 acetylation initially occurs equivalently under both Th1 and Th2 conditions but is subsequently lost in differentiating Th1 cells, whereas it is stabilized and maintained in differentiating Th2 cells (91, 171). Intronic elements. Both the Il4 and Il13 genes contain HS sites in their introns, although intronic enhancer function has been formally demonstrated only for Il4 (157, 158, 165, 188). The CIRE in the ﬁrst intron of the Il4 gene binds GATA3 and has been suggested to regulate passive demethylation of Il4 (161). The second intron of the Il4 gene contains HS II and HS III, which are DNase I hypersensitive and function as Il4 intronic enhancer elements in both Th2 cells and mast cells (36, 94, 185, 188). Both elements contain conserved STAT binding sites, and both bind endogenous STAT5a and possibly other STAT factors in Th2 cells and mast cells but not in Th1 cells (33, 36, 78). Although GATA1 and

GATA2 bind to HS II in mast cells, GATA3 has not been detected at this region in Th2 cells (189). The second intron of the Il4 gene rapidly acquires high levels of the H3K9/14 acetylation during Th2 differentiation (162, 169), possibly as a result of the early binding of STAT factors to HS II and HS III. Blockade of IL-2 signaling during Th2 differentiation decreased the accessibility of HS II and HS III to restriction enzymes; conversely, Th1 cells expressing constitutively active STAT5a exhibited increased accessibility at these sites (33). The introns of the Il13 gene contain two sites of increased accessibility to DNase I and restriction enzymes: Il13 HS III, which is present in both Th2 cells and mast cells, and MCHS (mast cell hypersensitivity site), which coincides with a nucleosome-free region and is only apparent in mast cells (157, 158, 185; see also Mast Cells). Neither the precise regulatory functions of these sites nor the factors binding them are known at present. Although the MCHS region does not function as a conventional enhancer element in transient reporter assays, it may contribute to the particularly robust Il13 expression of mast cells, as compared with Th2 cells (185). Intergenic and distal enhancers. Il13 HS I has the characteristics of a distal enhancer for the Il13 gene. This element, also termed CGRE, lies at the edge of a ∼30 kb region that spans the Il4 and Il13 genes and is selectively hyperacetylated in Th2 cells (162, 169). The CGRE contains several conserved binding sites for GATA3 and functions in transient reporter assays as a GATA3-dependent enhancer (32, 162). The most highly conserved noncoding sequence in the Il4 locus, CNS1, lies between Il4 and Il13 and contains two Th2-speciﬁc HS sites, HSS1 and HSS2 (156, 159). Histone acetylation at CNS1 correlates with a high probability of Il4 expression (95). In human CD3 T cells, H3 acetylation at CNS1 showed a marked increase after activation, suggesting that this element functions as an activation-dependent as well as a www.annualreviews.org • Regulation of Th2 Differentiation

631

ARI

23 February 2006

15:32

developmentally regulated enhancer (172). Consistent with these in vitro observations, deletion of CNS1 decreased the probability of Il4 and Il13 expression in Th2 differentiation, and in vivo Th2 immune responses were impaired in CNS1-deﬁcient mice (166, 170). The Il4 3 enhancer was ﬁrst discovered as an inducible hypersensitive site (HS VA ) located 3 of the Il4 gene (13). This enhancer contains functional binding sites for NFAT, STAT6, and GATA3 (13, 91). HS VA accessibility is induced by TCR triggering in Th2 cells, and among Th2 cells active IL-4 producers display a higher degree of HS VA accessibility than nonproducers (13, 94). Induction of hypersensitivity at HS VA derives from binding of NFAT together with GATA3; induction of HS VA is blocked by cyclosporin A, and GATA3 binding to HS VA is only detected upon TCR triggering (13, 91, 94). GATA3 overexpression in Th1 cells increased restriction enzyme accessibility at HS VA (94). Indeed, GATA3 overexpression in Th1 cells induces a Th2-like pattern of DNase I hypersensitivity and histone acetylation across the entire Il4 gene and 3 cis-regulatory elements (22, 23, 28), suggesting that a major role for this transcription factor in Th2 differentiation is to direct chromatin remodeling in the Il4 locus. A second region of very high evolutionary sequence conservation, CNS2, maps to HS V in the Il4/Kif3a intergenic region (156, 157). In naive T cells, this element bears the permissive modiﬁcation H3K4 dimethylation without displaying detectable levels of H3K9/14 acetylation (169). As such, HS V resembles boundary elements deﬁned in the Ig heavy chain and TCR-β loci that separate permissive from repressive chromatin (108); however, there is no evidence so far that HS V could function as a boundary element between Il4 and Kif3a in T cells. In addition to H3K4 dimethylation, HS V is also distinguished from other Th2-speciﬁc HSs in the vicinity of the Il4 gene by its relatively low level of DNA methylation in naive T cells (179). HS V remains DNA hypomethylated

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

632

Ansel et al.

and H3K4 dimethylated in Th2 cells but undergoes de novo DNA methylation and loses H3K4 dimethylation during Th1 differentiation (169, 179). Therefore, H3K4 dimethylation and DNA demethylation at HS V correlates well with T cell competence to transcribe the Il4 gene. The presence of RBPJκ binding sites in HS V has been cited as further evidence for a role for HS V in Il4 gene activation in naive T cells (61). Indeed, naive T cells from mice that lack a region that encompasses HS V and HS VA exhibit decreased Il4 transcription in response to TCR crosslinking (K.M. Ansel, unpublished data). Th2 cells from these mice exhibit decreased Il4 and Il13 expression, and particularly fail to increase cytokine production after repeated rounds of stimulation under Th2 culture conditions (167). In vivo Th2 immune responses are also impaired in HS V/VA −/− mice (K.M. Ansel, unpublished data). Further experiments are needed to dissect the individual activities of HS V and HS VA and to determine whether the mild phenotypic defects of the various mutant mice are due to functional redundancy among two or more cis-regulatory elements. The Th2 locus control region. Essentially all the cis-regulatory regions we discuss above are located in a ∼30 kb region that encompasses the Il4 and Il13 genes but excludes the neighboring genes Rad50 and Kif 3a. Minilocus reporter transgenes, in which most or all of these elements were linked to a luciferase reporter gene driven by a minimal Il4 promoter, did not recapitulate the high level of transcription of the endogenous Il4 gene (165). Moreover, luciferase expression from these minilocus transgenes varied with the integration site of the transgene and was not proportional to transgene copy number. Thus, none of these elements, individually or together, contained an LCR, deﬁned empirically as a genomic element that confers integration site– independence and copy number–dependence on linked genes in transgenic mice.

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LCRs can be located at long distances (up to hundreds of kilobases) from the genes that they regulate, and several have been shown to reside within introns of neighboring genes. For instance, the β-globin LCR spans a region between ∼6 and ∼25 kb upstream of the nearest gene (-globin) in the β-globin gene cluster, and an LCR for the Cd4 gene is located in a neighboring housekeeping gene, approximately ∼60 kb downstream of the Cd4 promoter (190, 191). Discovery and general features of the Th2 LCR. Analysis of mice bearing long bacterial artiﬁcial chromosome (BAC) transgenes showed that a 25 kb region located at the 3 end of the Rad50 gene (the Th2 LCR) was both necessary and sufﬁcient for LCR activity directed toward the neighboring Il4 and Il13 genes (163). The presence of this region in the BAC transgene led to integration site–independent, copy number–dependent expression of a linked luciferase reporter as well as the intact Il4 and Il13 genes present on the BAC. However, the region did not possess LCR activity for the Il5 gene, which is located at the 5 end of the Rad50 gene (in contrast to Il13 and Il4, which are 3 of Rad50). The same region of Rad50 was independently identiﬁed as potentially involved in helper T cell differentiation during a search for the genomic boundaries of the Il4/Il13 gene territory (14). The most distal genomic regions that differ in their DNase I hypersensitivity patterns and DNA methylation status between differentiated Th1 and Th2 cells were HS V (CNS2) at the 3 end of the Il4 gene and several conserved elements containing HSs at the 3 end of the Rad50 gene (Figure 3). In naive T cells, the Th2 LCR is hypermethylated at sites A and B but displays no DNase I hypersensitivity at any of the four sites (14). In Th1 cells, sites O and A are constitutively DNase I hypersensitive, whereas site B acquires hypersensitivity after TCR stimulation. Sites O, A, and B are all constitutively DNase I hypersensitive in Th2 cells, and site C [also known as RHS 7 (Rad50

hypersensitivity site 7)] becomes hypersensitive in an IL-4-dependent manner after TCR stimulation. The detailed pattern of hypersensitivity observed at the Th2 LCR appears to vary according to the experimental conditions employed (14, 160). The patterns we describe in the paragraph above (see Figure 3) are observed in cells cultured in the absence of IL-2 (14); in the presence of IL-2, LCR sites A and B (also collectively termed RHS 6) are constitutively hypersensitive in both Th1 and Th2 cells, while LCR-C/RHS 7 presents as a constitutive Th2-speciﬁc HS site (160; K.M. Ansel & A. Rao, unpublished data). Broad multispecies sequence comparisons reveal ancient conservation of the Rad50 exons in ﬁsh, birds, and mammals, with the LCR sites evolving in situ in the intervening Rad50 introns (Figure 3, bottom). The four intronic LCR hypersensitivity sites are not conserved in ﬁsh, which lack Il4 and Il13 genes. However, all four of these sites correspond to highly conserved noncoding sequences present in all currently available mammalian genome sequences (14, 156, 160; Figure 3). Interestingly, neither the LCR nor the cis-regulatory elements situated among the Il4 and Il13 genes score as highly conserved in chicken (Gallus gallus), although Il4 and Il13 genes are annotated between Kif3a and Rad50 in the chicken genome. LCR-C/RHS 7. What are the functions of the HS sites in the Th2 LCR? So far, only the functions of LCR-C/RHS 7 have been investigated in any detail (168). Other than HS VA at the 3 end of the Il4 gene, this is the only site in the Il4 locus that is inducibly hypersensitive in stimulated Th2 cells. Induction of hypersensitivity at LCR-C/RHS 7 has been attributed to binding of STAT6, which is induced in stimulated Th2 cells as a result of autocrine IL-4 production (14). Retroviral expression of a constitutively active STAT6 in developing Th1 cells induces LCR-C only if the cells are subsequently restimulated through the TCR, suggesting www.annualreviews.org • Regulation of Th2 Differentiation

633

ARI

23 February 2006

15:32

that TCR-induced transcription factors also contribute to induction of hypersensitivity at LCR-C. Consistent with this hypothesis, NFAT binding can be detected at LCR-C by chromatin immunoprecipitation, and induction of hypersensitivity at LCR-C is at least partially CsA sensitive (S. Kim, D.U. Lee & A. Rao, unpublished data). Retroviral transduction of GATA3 in developing Th1 cells does not induce LCR-C (14), nor does it induce reporter activity driven by the LCR (163). However, GATA3 can be detected by chromatin immunoprecipitation as binding to this region in differentiated Th2 cells (186). Thus, exactly as we suggest above for the rest of the Il4 locus, it is likely that (a) STAT6 and TCR signals initially set up the changes that lead to increased accessibility of LCR elements during Th2 differentiation and (b) GATA3 is involved in maintaining differential accessibility, at least in recently differentiated Th2 cells. In transient reporter assays and in reporter-transgenic mice, LCR-C/RHS 7 functions as a conventional enhancer (160). Consistent with this function, deletion of LCR-C/RHS 7 in the genome of mice substantially but partially diminishes Il4 and Il13 expression in naive T cells cultured under nonpolarizing and Th2 conditions, respectively (168). Furthermore, CpG dinucleotides in LCR-C/RHS 7 are rapidly demethylated during Th2 differentiation, with kinetics similar to those observed at the 5 end of the Il4 gene, including the CIRE (160, 161, 179). These observations are consistent with the physical interactions (looping and chromatin hub formation) between the LCR and the cytokine promoters in naive and differentiated T cells (see below).

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

LCR-O/RHS 5 and LCR-A, B/RHS 6. Individual deletion of these elements will be required to assess the contributions of these elements to Th2 cytokine expression at the various stages of Th1 and Th2 differentiation. On the basis of the precedent of HS4 in the chicken β-globin LCR, one might predict that the most 5 sites of the Th2 LCR, LCR-O 634

Ansel et al.

and LCR-A, could function as boundary elements that insulate the Rad50 promoter from the activation and silencing mechanisms directed at the cytokine genes in differentiating helper T cells. The functions of LCR-B could well be more complex, given that this element displays constitutive and inducible hypersensitivity in Th2 and Th1 cells, respectively. Long-range interactions among regulatory regions: role of the LCR. It has recently become possible to study long-range chromatin topology using two powerful new techniques: chromosome conformation capture (3C) and RNA-TRAP (reviewed in 192). Both methods have been applied to the β-globin locus with comparable results (193, 194); however, because the RNA-TRAP technique can only be used in cells that are actively transcribing the relevant locus, the 3C assay has been more informative in general. In the 3C assay, the native conﬁguration of chromatin is captured by formaldehyde crosslinking, and thereafter restriction enzyme digestion, religation, and quantitative PCR are used to assess which regions of a locus (potentially far apart from each other in a linear context) are associated in cells. In mouse erythroid-lineage cells that are actively transcribing β-globin genes, the 3C assay detects a core structure termed the “active chromatin hub,” which is composed of the β-globin LCR, a far upstream hypersensitive site (HS -60/-62), a downstream hypersensitive site (3 HS 1), and promoters of the relevant β-globin genes—the embryonic globin genes Hbb-y and Hbb-bh1 in primitive erythroid cells (embryonic day 10.5), but the adult genes Hbb-b1 and Hbb-b2 in cells undergoing deﬁnitive erythropoiesis (embryonic day 14.5) (194, 195). This structure is not detected in brain in which the globin genes are not transcribed. Formation of the hub does not appear to require ongoing transcription, since deletion of the human β-globin promoter did not abolish 3C interactions between the β-globin gene and the LCR (196). However, a double deletion of both the promoter and HS 3 in the LCR in cis abolished gene

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

transcription, DNAse I hypersensitivity, and formation of the chromatin hub in parallel; this effect was remarkably selective, since double deletion of the promoter and HS 2, an adjacent HS site in the LCR, had little effect (196). It is important to note, however, that the interactions measured in the 3C assays are averaged over an entire population; thus, some of the detected interactions may occur in completely different cell types within a population or—as expected for sequential interactions of LCRs with different promoters—at different times in the same cells. More recently, the 3C assay has been successfully applied to the Il4 locus (186) (Figure 5). Remarkably, the Il4, Il13, and Il5 promoters form a minimal core interacting structure even in ﬁbroblasts that do not transcribe the Th2 cytokines. In naive T cells in which the Il4 locus is “poised” for rapid expression, the Th2 LCR was additionally recruited into this minimal structure to form a chromatin hub. As in the β-globin locus, formation of the hub is required for efﬁcient transcription: Targeted deletion of the Th2-

speciﬁc LCR-C/RHS 7 HS site in the Th2 LCR was associated with decreased production of Th2 cytokines as well as diminished promoter-LCR interactions in the 3C assay (168). Surprisingly, however, a hub conﬁguration very similar to that of naive T cells was observed not only in Th2 cells that have increased their transcriptional competence at the Il4 locus, but also in Th1 cells that are in the process of silencing the cytokine genes (186). Thus, the 3C technique cannot yet detect subtle changes associated with gene activation and silencing, but, as in the β-globin locus, formation of the chromatin hub is independent of active transcription. Use of the 3C assay, especially with reﬁnements that detect activation-dependent changes or the different expected conﬁgurations in Th1 and Th2 cells, should make it possible to address several important questions in molecular detail. For instance, what proteins are involved in forming the chromatin hub in naive, Th1, and Th2 cells, and how does hub formation relate to cytokine gene transcription and silencing? As

Fibroblast

Lymphocyte

RAD50p

RAD50p LCR

IL-5p

IL-13p

IL-4p

CNS1 V/VA

Core promoter interactions

LCR IL-5p

IL-13p

IL-4p

CNS1 V/VA

"Chromatin hub"

Figure 5 Model of chromatin looping in the Il4. The cytokine gene promoters (IL-5p, IL-13p, IL-4p) interact with one another even in ﬁbroblasts that are not competent to transcribe these cytokine genes. In lymphocytes, additional cis-regulatory elements involved in type II cytokine gene regulation—the Th2 LCR and the enhancers CNS1 and V/VA —join the promoters in a “chromatin hub.” A similar hub structure has been reported for Th2, Th1, and naive T cells, as well as for B cells and NK cells. www.annualreviews.org • Regulation of Th2 Differentiation

635

ANRV270-IY24-20

ARI

23 February 2006

15:32

Silencing of the Il4 Locus in Th1 Cells

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MAST CELLS Mast cells, like Th2 cells, produce large quantities of IL-13 and IL-4 (36). Although these cells derive from the myeloid lineage and are therefore only distantly related to Th2 cells, mast cells use an overlapping set of transcription factors and cis-regulatory elements to regulate Il13 and Il4 (185, 198). Comparison of chromatin structure in the two cell types indicates that the involvement of particular cis-regulatory sequences in a given cell type can be accurately predicted by determining the accessibility of that sequence in that cell type. Most of the hypersensitivity sites detected in Th2 cells are also highly accessible in mast cells, but two clear differences have been documented: (a) Th2 cells display the HSS1 and HSS2 hypersensitivity sites at CNS1, whereas mast cells do not (185), consistent with the ﬁnding that CNS1−/− Th2 cells show impairments in Il4 and Il13 expression, whereas CNS1−/− bone marrow-derived mast cells do not (166); and (b) compared with Th2 cells, mast cells express higher levels of Il13 than of Il4. The differences between Th2 cells and mast cells are inﬂuenced by the expression of different GATA and NFAT family transcription factors, but they may also reﬂect the fact that mast cells use an additional, mast cell-speciﬁc cis-regulatory element (MCHS) within the third intron of Il13 (185). a partial answer to this question, retroviral transduction of GATA3 into NIH3T3 ﬁbroblasts followed by ionomycin stimulation was shown to lead to de novo 3C interactions between the Il4 promoter and LCR-C/RHS 7, the Th2-speciﬁc site in the Th2 LCR (186). GATA3 was shown to bind sites in the LCRC/RHS 7 element both in Th2 cells and in the transduced ﬁbroblasts, and the requirement for ionomycin stimulation was attributed to a need for cooperation between the transduced GATA3 and endogenous NFAT proteins present in the ﬁbroblasts, as previously documented for the 3 Il4 enhancer (91, 186). However, the transduced and stimulated ﬁbroblasts remained incapable of transcribing the Th2 cytokine genes, suggesting that additional factors are necessary. In this context, it is interesting that GATA1 expression in an erythroid precursor cell line led to increased interaction between the mouse β-globin LCR and the promoter of the β-major gene (197). 636

Ansel et al.

Historically, the regulation of Th2 cytokine silencing in Th1 cells has received less attention than the regulation of Th2 cytokine activation in Th2 cells. This trend has been reversed recently, driven by rapid progress in the elaboration of general mechanisms of transcriptional gene silencing (139). Repressive histone modiﬁcations have been detected within the Il4 locus, particularly in association with HS IV and HSS3, which remain highly accessible in Th1 cells (170, 173). As discussed above, deletion of HS IV in the mouse Il4 locus revealed the presence of a cis-regulatory silencer that is important for effective Th1-mediated immune responses (44). Continued investigation of the chromatin structure of the Il4 locus may provide further insights, but given the wealth of descriptive information now available, important challenges lie ahead in connecting these data to build a mechanistic model for epigenetic silencing. The transcription factors that direct Il4 locus silencing remain obscure, and even the mechanism by which transcription factors that activate transcription are excluded from the locus in Th1 cells is the subject of considerable uncertainty. How are histonemodifying enzymes recruited to the locus, and what regulates their activity? Do these modiﬁcations recruit chromatin-binding protein complexes involved in gene silencing? How do these processes reduce locus accessibility and maintain heritable Il4 and Il13 silencing in Th1 cells? Some intriguing possibilities and a collection of supporting observations are discussed below. Histone modifications associated with gene silencing. The early Il4 transcription that occurs upon activation of naive T cells is accompanied by the appearance of permissive chromatin modiﬁcations and increased accessibility throughout the Il4 locus, regardless of the priming conditions (14, 91, 169, 171). As we describe above, this epigenetic state is

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

stabilized and strengthened in cells activated under Th2 conditions. In Th1 conditions, however, most of these permissive modiﬁcations are erased and replaced with repressive modiﬁcations (91, 162, 169–171, 173). Within a few days of Th1 priming, only the Il4 silencer at HS IV remains associated with permissive histone modiﬁcations (44, 169). HDAC activity is important for repression of Il4 expression (181, 184, 199). Even the basal H3K9/14 acetylation of Il4 promoterassociated nucleosomes in naive T cells is removed within ﬁve days of Th1 priming (170). In addition, the H3K4 dimethylation observed at HS V in unstimulated naive T cells is eliminated in Th1 cells (169, 200). It is not clear whether this process involves active demethylation by LSD1 or another demethylase, or if methylated histones are simply replaced with unmethylated histones as the cells divide. The DNA of HS V, which is generally unmethylated in naive T cells, undergoes de novo methylation early in Th1 differentiation (179). Th2 differentiation and early Il4 transcription are both impaired in HS V/VA −/− mice (167; K.M. Ansel, unpublished data), and reduced or absent basal Il4 promoter H3K9/14 acetylation correlates well with the diminished early Il4 transcription and impaired Th2 differentiation of T cells that lack CD4, Itk, or CNS1 (166, 170). Taken together, these data indicate that the reversal of permissive chromatin structure, particularly at the Il4 promoter and HS V, plays an important role in reducing Il4 locus activity early in Th1 differentiation. As we discuss above (see Plasticity section, p. 615), however, stable silencing of Il4 expression occurs over a longer period of time, implying that the late acquisition of repressive epigenetic changes prevents Il4 locus reactivation when a committed Th1 cell encounters Th2-inducing signals. A recent survey of repressive histone modiﬁcations in the Il4 locus bears out this prediction (173). A modest level of H3K27 trimethylation spreads over the Il4 locus, but only after two rounds of stimulation under Th1 conditions (Figure 3).

Chromatin immunoprecipitation with an antibody raised against methylated H3K9 (but that may crossreact with methylated H3K27) also revealed histone methylation at the Il4 promoter in an established Th1 cell line, but not in a Th2 cell line or naive T cells (170). Further experiments with more speciﬁc antibodies did not detect dimethyl- or trimethylH3K9 in the Il4 locus after two rounds of Th1 priming (173), although it remains possible that H3K9 methylation may occur in the Il4 locus as a very late event in Th1 differentiation. The association of most of the Il4 locus with trimethylated H3K27 appears fairly weak, but two foci of robust H3K27 trimethylation are present in Th1 cells, and one of these is particularly prominent in naive T cells (173). These foci correspond with HSS3 and HS IV, which also share a unique pattern of accessibility in naive, Th1, and Th2 cells (157, 159) (Figure 3). H3K27 trimethylation is especially strong at HSS3 in naive T cells, and progressively diminishes at both sites during Th2 differentiation (173). In addition, the H3K27 methyltransferase EZH2 was found to be constitutively associated with both sites in naive T cells and Th1 cells. Unexpectedly, EZH2 also remains bound to HSS3 and HS IV in Th2 cells despite their loss of H3K27 trimethylation, indicating that EZH2 activity in the Il4 locus may be differentially regulated in Th1 and Th2 cells. Overall, these data suggest that HSS3 and HS IV may serve as nucleation sites for spreading of repressive chromatin modiﬁcations during Th1 differentiation. The cis-regulatory activity of HSS3 and its relationship to the HS IV silencer are under investigation. Surprisingly, the HS IV silencer is also strongly associated with permissive histone modiﬁcations in Th2, Th1, and naive T cells. In Th1 cells, HS IV appears as an isolated peak of H3K4 dimethylation and H3K9/14 acetylation in an otherwise repressive chromatin domain extending at least from HS V to the CGRE (44, 169) (Figure 3). It is not clear whether these permissive histone www.annualreviews.org • Regulation of Th2 Differentiation

637

ANRV270-IY24-20

ARI

23 February 2006

15:32

modiﬁcations coexist with H3K27 trimethylation on the same nucleosomes, but if so, the combination would be unprecedented, representing a new “character” in the histone code.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Heterochromatin and intranuclear positioning of cytokine genes. One silencing mechanism that has received considerable attention is the formation of facultative heterochromatin at the Il4 locus. Both H3K9 and H3K27 methylation can be considered markers of facultative heterochromatin, although H3K9 methylation is particularly important for binding of the canonical marker of heterochromatin, HP1 (201). In this model, repressive histone modiﬁcations lead to the recruitment of silencing complexes that draw the Il4 locus into a compact, inaccessible structure or to a nuclear microenvironment that favors inactivity. Genes thus regulated often assume intranuclear positions adjacent to constitutive heterochomatin, which is rich in repetitive DNA sequences such as those found in and near centromeres (133, 142, 143). It has been reported that up to 60% of the Il4 genes in populations of Th1 cells from Leishmania major–infected mice, as well as in Th1 cell lines established in vitro, are indeed positioned in centromeric heterochromatin (43, 170). Another group of investigators failed to reproduce these results, but instead observed selective localization of the Gata3 gene near centromeric heterochromatin and the gene that encodes c-Maf near the nuclear periphery in Th1 cells (202). If the Il4 locus becomes heterochromatinized in Th1 cells, it likely occurs via a novel pathway and certainly retains characteristics that do not conform to our current conception of heterochromatin. For example, HS IV, HSS3, and LCR-O, -A, and -B remain highly accessible even in Th1 cell lines in which Il4 expression is effectively extinguished (14, 44, 159, 160, 203), and the LCR remains joined with the Il4 and Il13 promoters in an active chromatin hub (186). Therefore, it seems unlikely that the Il4 locus undergoes widespread 638

Ansel et al.

chromatin compaction or sequestration in a highly inaccessible nuclear compartment. A comparison with the heterochromatinization of Dntt, which is rapidly silenced upon TCR engagement in developing double-positive thymocytes, is instructive. Transcriptional silencing of Dntt is dependent on Ikaros-binding sites present in its promoter (203). An ordered series of histone modiﬁcations commences at the Dntt promoter within one hour of TCR crosslinking, starting with histone deacetylation, loss of H3K4 methylation, and ﬁnally methylation of H3K9 (134). These epigenetic marks spread through the locus, which is simultaneously repositioned to centromeric heterochromatin (133). The entire process of heterochromatinization appears complete within 24 h of the initiating TCR signal (133, 134, 139). In the Il4 locus, Ikaros binds to sequences within CNS1 in electrophoretic mobility shift experiments, although this observation has not been conﬁrmed in living Th1 cells (170). Th1 priming does induce histone deacetylation, but only after an early burst of widespread acetylation. Instead of H3K9 methylation, H3K27 trimethylation is observed in two foci. These foci exist prior to Th1 priming, and spreading of the repressive modiﬁcation can only be detected after two rounds of priming (173). The difﬁculty in detecting H3K9 methylation and locus repositioning to centromeric heterochromatin underscores the inefﬁciency of Il4 heterochromatinization, and suggests that some form of active repression akin to the silencing of the CD4 gene in double-negative thymocytes must contribute to Il4 locus silencing in Th1 cells (141). Further investigation is needed to ﬂesh out our molecular understanding of Il4 locus silencing. Polycomb-mediated silencing and intergenic transcription. Several lines of evidence point to a role for Polycomb complexes in Il4 locus silencing. Although H3K27 trimethylation can participate in PRC1independent gene repression, as it does in X-inactivation, EZH2-mediated H3K27

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

trimethylation may provide binding sites for PRC1 in the Il4 locus (173). In either case, it will be of great interest to determine how EZH2 is recruited to HSS3 and HS IV, and how its activity is differentially regulated in Th1 and Th2 cells. In addition to EZH2, two other Polycomb proteins have been implicated in Il4 locus regulation. Mice lacking the PRC1 protein Mel-18 exhibit defective lymphocyte development (204), but the mature T cells that do survive thymic development are defective in Th2 differentiation (152). Although this defect was attributed to inadequate induction of GATA3 expression, additional direct effects on the Il4 locus cannot be excluded. Yin Yang-1 (YY1) is a DNA binding transcription factor that can substitute for its Drosophila homolog, Pleiohomeotic, in initiating Polycomb silencing (205). YY1 has been implicated in the activation and repression of many genes, including Il5 (repression), and both Il4 and Ifng (activation) (206–208). In Drosophila, Polycomb-responsive genes remain inactivated unless something intervenes to reverse that inactivation. Such genes are stably repressed, yet poised for developmental changes, exactly like the cytokine genes in naive T cells (209). Recently, the molecular mechanism of that activating intervention has been identiﬁed: Transcription through Drosophila PREs is sufﬁcient to reverse Polycomb-mediated silencing (210). Importantly, the presence of naturally occurring noncoding transcripts that traverse PREs correlates closely with relief from Polycombmediated silencing (210–213). Intergenic transcripts can be readily detected throughout the human IL-4 and mouse Il4 loci in Th2 cells, but not Th1 cells (162, 200, 214, 215), and a low level of RNA Polymerase II (Pol II) binding was detected throughout the domain of permissive histone modiﬁcations in Th2 cells (200). In addition, strong binding of both Pol II and TFIIB was observed at three discrete sites: the Il4 promoter, HS VA , and CNS1 (200). The latter two of these lie adjacent to HS IV and

HSS3, respectively, and may be sites of initiation of intergenic transcripts. Intergenic transcription in Th2 cells may simply be a nonfunctional side effect of maintaining the locus in an accessible, transcriptionally active state. However, it has long been suggested that these transcripts play a role in maintaining the transcriptional competence of the Il4 locus, perhaps owing to the association of HATs and H3K4 methyltransferases with active transcription complexes (reviewed in 216). In keeping with the idea that the Il4 locus may be subject to Polycomb-mediated repression, intergenic transcription in Th2 cells may proceed through a mammalian PRE (perhaps HS IV and/or HSS3) and disrupt its repressive function. Conversely, the absence of these transcripts in Th1 cells may allow stable silencing of the locus. Alternatively, rather than interfering with Polycomb-mediated repression, intergenic transcription may actively impose Il4 silencing during Th1 differentiation. Although Il4 intergenic transcripts are not observed in fully differentiated Th1 cells, transcripts that direct gene silencing often simultaneously silence their own transcription. One possibility is that antisense transcription may traverse the Il4 gene during a critical stage of Th1 differentiation. Antisense transcripts are abundant in EST databases (217), and many of them have been experimentally validated (218). Antisense transcription could impede initiation or elongation of Il4 primary transcripts, or it could generate double-stranded sense/antisense transcript pairs (219). Bidirectional transcript pairs, whether from coding or intergenic regions, can enter the RNAi pathway and initiate transcriptional gene silencing. In this context, the Il4 silencer at HS IV, which is positioned just 2 kb downstream of Il4, could function as an enhancer or promoter of silencing intergenic transcripts or as the target of an RNA-directed epigenetic silencing process. Although these models remain highly speculative, endogenous RNAi does inﬂuence cytokine gene regulation (220, 221), as discussed in the following section. www.annualreviews.org • Regulation of Th2 Differentiation

639

ANRV270-IY24-20

ARI

23 February 2006

15:32

ENDOGENOUS RNAi AND Th1/Th2 DIFFERENTIATION In the past decade, a new paradigm in gene regulation has emerged with the discovery that siRNAs direct sequence-speciﬁc repression of gene expression. Here, we provide a brief introduction to RNAi and discuss early ﬁndings on its role in helper T cell differentiation.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Introduction to RNA Interference (RNAi) Although numerous strategies are being developed to exploit RNAi-mediated repression for experimental and therapeutic purposes (222, 223), we have just begun to unravel the myriad mechanisms through which RNAi regulates gene expression (224–226). Both bidirectionally transcribed transposons and double-stranded RNA viral genomes can be processed into siRNAs, which prompts speculation that RNAi may have evolved as a cell-autonomous innate immune mechanism (227–229). However, RNAi-based mechanisms also regulate endogenous gene expression through the actions of a developmentally regulated class of siRNAs called microRNAs (miRNA) (230–232). Over 200 different human miRNA have been cataloged (233), and prevailing estimates place the likely total around 1000 (234, 235). Endogenous RNAi is required for the proper development of multicellular organisms; it regulates basic cellular processes such as proliferation, survival, and chromosome segregation and also inﬂuences cell lineage decisions (224, 230, 236). All siRNA are processed from doublestranded RNA precursors by the ribonuclease III enzyme Dicer (231, 232, 237). Doublestranded RNAs are cleaved by Dicer into double-stranded ∼21-nucleotide fragments, of which one strand becomes associated with effector ribonucleoprotein complexes as a mature siRNA. In contrast, miRNAs are processed from larger, single-stranded primary miRNA transcripts (pri-miRNA) by the sequential actions of Drosha and Dicer, 640

Ansel et al.

nuclear and cytoplasmic ribonuclease III enzymes, respectively (238, 239). Drosha recognizes and cleaves 60–70 nucleotide stem-loop structures (pre-miRNAs) embedded within the pri-miRNAs; these pre-miRNAs are then exported to the cytoplasm for further processing into mature ∼21-nucleotide miRNAs by Dicer (237). Although some of the molecular details have diverged, the basic mechanisms of RNAi are broadly conserved in diverse eukaryotic organisms: siRNA pairs with RNA targets and induces posttranscriptional silencing by target cleavage and/or translational inhibition, and nuclear siRNA can also direct transcriptional gene silencing via heterochromatin formation (225, 229, 240, 241). The choice of mRNA silencing mechanism(s) triggered by a particular siRNA is incompletely understood, but it clearly depends on the degree of complementarity between the siRNA and its target, and the composition of the siRNA-associated ribonucleoprotein complexes that mediate the biochemical functions that interfere with gene expression (226, 242). Incomplete complementarity between siRNA and mRNA targets favors translational inhibition, whereas perfect complementarity favors target cleavage through a “Slicer” endonuclease activity. Slicer activity is provided by Argonaute-2, just one of four broadly expressed mammalian Argonaute proteins, all of which associate with siRNAs in the cytoplasmic RNAinduced silencing complex (RISC) (243). A nuclear Argonaute-containing RNA-induced transcriptional silencing complex (RITS) initiates siRNA-directed heterochromatin formation in ﬁssion yeast and thereby regulates centromere integrity and switching of mating-type locus gene expression (as discussed in Repressive Histone Modiﬁcations and the Formation of Silent Chromatin, above) (244–246). An analogous vertebrate nuclear RNAi complex has yet to be described, but promoter-directed siRNAs have been reported to induce transcriptional gene silencing in human tissue culture cell lines (247, 248). Moreover, inactivation of RNAi by

ANRV270-IY24-20

ARI

23 February 2006

15:32

targeted mutation of the gene encoding Dicer disrupted epigenetic silencing of centromeres in mouse embryonic stem cells (249, 250) and additionally compromised the integrity of chromosome segregation in chicken DT40 B cells (251).

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Endogenous RNAi in T Cell Differentiation RNAi has been implicated in the regulation of immune system development and function (252–254), but direct evidence for this has only recently emerged from studies of T cells that lack Dicer function (220, 255). In some organisms, specialized Dicer proteins process different double-stranded RNA substrates, but all these functions are served by a single mammalian Dicer. Therefore, Dicer deletion disrupts all endogenous RNAi. As Dicer deﬁciency causes early embryonic lethality (249, 256), conditional gene targeting with the Cre/loxP system has been employed to analyze the role of RNAi in speciﬁc cell lineages (220, 255, 257). Dicer deletion with an Lck promoter Cre transgene that is expressed early in thymoctye development decreased thymocyte survival but did not disrupt thymocyte lineage decisions or the expression of several key developmentally regulated genes, including Cd4, Cd8, and Dntt (255). Dicer inactivation at the doublepositive stage of thymocyte development using a CD4 promoter Cre transgene had only a moderate effect on thymocyte maturation and resulted in a decreased but appreciable output of both CD8+ and CD4+ mature T cells with a normal naive T cell surface phenotype (220). Although Dicer protein was nearly undetectable and Dicer substrate pre-miRNA accumulated in these peripheral T cells, a small amount of residual mature miRNA was detected and may have been important for the successful completion of thymopoiesis. When activated in vitro, Dicer-deﬁcient helper T cells exhibited decreased proliferation and increased apoptosis and underwent

rapid, disregulated Th1 differentiation (220). Abnormally elevated IFN-γ production was detectable within 24 h of naive T cell stimulation in nonpolarizing conditions. After just two days, 70% of Dicer-deﬁcient cells had differentiated into IFN-γ-producing Th1 cells, compared with less than 20% of wild-type cells. Dicer-deﬁcient T cells were able to activate the Th2 program and produce IL-4 when activated in strong Th2-skewing culture conditions, yet a substantial fraction of these cells still differentiated into Th1 cells. Taken together, these data suggest that RNAi may normally restrain the expression of key proteins involved in Th1 differentiation, such as IFN-γ or T-bet. Alternatively, RNAi may restrain T cell differentiation per se, and the preferential Th1 differentiation observed in these experiments may be an exaggerated reﬂection of the natural Th1 bias of the C57BL/6 genetic background or may occur because IL-4 production is inefﬁcient prior to cell division (181). Thus, the rapid Th1 differentiation of Dicer-deﬁcient T cells may have obscured or bypassed RNAi-dependent pathways that control Th2 differentiation and Il4 locus regulation. Further experiments are required to determine whether RNAi regulates Th2 cytokines. How does loss of Dicer function affect the balance of Th1/Th2 differentiation? Impairment of global RNAi function could directly affect transcriptional silencing of cytokine genes, as discussed above. Alternatively, the observed defects could arise from the deﬁciency of one or more individual miRNAs. In either case, identiﬁcation of the critical missing siRNAs and derepressed targets that underlie the complex phenotype of Dicer-deﬁcient T cells presents a major challenge for future investigation. A comparison of helper T cell subsets showed few differences in miRNA expression between Th1 and Th2 cells (258). A few striking changes in miRNA expression were noted: mir-150 was abundant in naive T cells, but absent in both Th1 and Th2 effectors; and mir-146 was expressed at a low level in naive T cells and www.annualreviews.org • Regulation of Th2 Differentiation

641

ARI

23 February 2006

15:32

upregulated or downregulated in Th1 and Th2 cells, respectively. Additional cases of differential miRNA expression among T cell subsets will likely be discovered as new miRNAs are identiﬁed and other T cell subsets are analyzed. However, it will be a further challenge to identify the corresponding mRNA targets, since mammalian miRNAs generally mediate translational inhibition through imperfect base pairing with sites in the 3 untranslated region (UTR) of target mRNAs (259–261). Hobert has noted the common logic of miRNA and transcription factor regulation of gene expression (236) (Figure 6). In both cases, trans-acting factors (miRNA, transcription factors) bind to conserved sequences in cis-regulatory elements (3 UTR, promoters, enhancers, silencers) and regulate gene expression by recruiting regulatory protein complexes (RISC, RNA polymerase machinery, and/or chromatin modifying complexes). As miRNAs regulate the translation of transcription factors, and transcription factors regulate the transcription of pri-miRNAs, both pathways participate in one integrated regulatory network. Reﬁned models of helper T cell differentiation will be needed to incorporate our growing understanding of the role of endogenous RNAi.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

SUMMARY AND FUTURE DIRECTIONS Since 1998 when chromatin structural changes and epigenetic modiﬁcations were ﬁrst shown to operate in Th1/Th2 differentiation, studies from many laboratories have elucidated various aspects of the process. In this review, we have summarized our current understanding of how transcription factors, cis-regulatory elements, chromatin structural changes, epigenetic modiﬁcations, and RNAi mechanisms act in concert to determine the transcriptional competence of cytokine genes. T cell differentiation can be modeled as occurring in several stages: initiation, which depends on early inducible transcription factors; reinforcement (commitment), which depends on lineage-speciﬁc transcription factors; and maintenance of the committed transcriptional state, which may rely on more general mechanisms that involve ubiquitous transcriptional regulators and long-range interactions among cis-regulatory elements in the chromatin context. Each of these processes needs to be understood in greater detail. Important questions for future research include: How do inducible and lineage-speciﬁc transcription factors interact to cause early and late epigenetic changes at conserved cisregulatory elements in the Il4 locus? Which

Gene Enhancer

Promoter

Silencer Transcription 3'UTR

mRNA

AAAAAA

Translation RISC +miRNA

RISC +miRNA

Protein

Figure 6 The “common logic” of transcriptional and posttranscriptional gene regulation. Transcription factors bind to conserved cis-regulatory regions in chromosomal DNA and direct transcriptional activation or silencing. RISC-associated miRNAs bind to conserved sites in the 3 untranslated region (3 UTR) of mRNA and direct translational inhibition. Adapted from Reference 236. 642

Ansel et al.

ANRV270-IY24-20

ARI

23 February 2006

15:32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

transcriptional regulators and coregulatory complexes are recruited and with what kinetics, and which chromatin-remodeling complexes and histone-modifying enzymes are involved? What are the kinetics of the epigenetic changes, and how and at what point do the changes become irreversible? How is transcription through the locus inﬂuenced by long-range interactions among cis-regulatory regions including the LCR, and how are these interactions themselves regulated by tran-

scriptional regulators and epigenetic modiﬁcations? And ﬁnally, what are the roles of Dicer, noncoding intergenic transcription, the RNAi machinery and micro-RNAs? Given the ease of in vitro manipulation of Th1/Th2 differentiation and the ready availability of naive T cell precursors, the lineage decisions of helper T cells will continue to provide a valuable paradigm for investigating the biochemical mechanisms underlying cell lineage speciﬁcation.

LITERATURE CITED 1. Murphy KM, Reiner SL. 2002. The lineage decisions of helper T cells. Nat. Rev. Immunol. 2:933–44 2. Agarwal S, Rao A. 1998. Long-range transcriptional regulation of cytokine gene expression. Curr. Opin. Immunol. 10:345–52 3. Mowen KA, Glimcher LH. 2004. Signaling pathways in Th2 development. Immunol. Rev. 202:203–22 4. Stetson DB, Voehringer D, Grogan JL, Xu M, Reinhardt RL, et al. 2004. Th2 cells: orchestrating barrier immunity. Adv. Immunol. 83:163–89 5. Szabo SJ, Sullivan BM, Peng SL, Glimcher LH. 2003. Molecular mechanisms regulating Th1 immune responses. Annu. Rev. Immunol. 21:713–58 6. Ansel KM, Lee DU, Rao A. 2003. An epigenetic view of helper T cell differentiation. Nat. Immunol. 4:616–23 7. Reiner SL. 2005. Epigenetic control in the immune response. Hum. Mol. Genet. 14(Spec. Issue 1):R41–46 8. Wilson CB, Makar KW, Shnyreva M, Fitzpatrick DR. 2005. DNA methylation and the expanding epigenetics of T cell lineage commitment. Semin. Immunol. 17:105–19 9. Smale ST, Fisher AG. 2002. Chromatin structure and gene regulation in the immune system. Annu. Rev. Immunol. 20:427–62 10. Li-Weber M, Krammer PH. 2003. Regulation of IL4 gene expression by T cells and therapeutic perspectives. Nat. Rev. Immunol. 3:534–43 11. Berg LJ, Finkelstein LD, Lucas JA, Schwartzberg PL. 2005. Tec family kinases in T lymphocyte development and function. Annu. Rev. Immunol. 23:549–600 12. Rao A, Luo C, Hogan PG. 1997. Transcription factors of the NFAT family: Regulation and function. Annu. Rev. Immunol. 15:707–47 13. Agarwal S, Avni O, Rao A. 2000. Cell-type-restricted binding of the transcription factor NFAT to a distal IL-4 enhancer in vivo. Immunity 12:643–52 14. Lee DU, Rao A. 2004. Molecular analysis of a locus control region in the T helper 2 cytokine gene cluster: a target for STAT6 but not GATA3. Proc. Natl. Acad. Sci. USA 101:16010–15 15. Xanthoudakis S, Viola JP, Shaw KT, Luo C, Wallace JD, et al. 1996. An enhanced immune response in mice lacking the transcription factor NFAT1. Science 272:892–95 16. Rengarajan J, Tang B, Glimcher LH. 2002. NFATc2 and NFATc3 regulate T(H)2 differentiation and modulate TCR-responsiveness of naive T(H)cells. Nat. Immunol. 3:48– 54 www.annualreviews.org • Regulation of Th2 Differentiation

643

ARI

23 February 2006

15:32

17. Peng SL, Gerth AJ, Ranger AM, Glimcher LH. 2001. NFATc1 and NFATc2 together control both T and B cell activation and differentiation. Immunity 14:13–20 18. Ranger AM, Oukka M, Rengarajan J, Glimcher LH. 1998. Inhibitory function of two NFAT family members in lymphoid homeostasis and Th2 development. Immunity 9:627– 35 19. Hogan PG, Chen L, Nardone J, Rao A. 2003. Transcriptional regulation by calcium, calcineurin, and NFAT. Genes Dev. 17:2205–32 20. Crabtree GR, Olson EN. 2002. NFAT signaling: choreographing the social lives of cells. Cell 109(Suppl.):S67–79 21. Zhu J, Guo L, Watson CJ, Hu-Li J, Paul WE. 2001. Stat6 is necessary and sufﬁcient for IL-4’s role in Th2 differentiation and cell expansion. J. Immunol. 166:7276–81 22. Ouyang W, Lohning M, Gao Z, Assenmacher M, Ranganath S, et al. 2000. Stat6independent GATA-3 autoactivation directs IL-4-independent Th2 development and commitment. Immunity 12:27–37 23. Lee HJ, Takemoto N, Kurata H, Kamogawa Y, Miyatake S, et al. 2000. GATA-3 induces T helper cell type 2 (Th2) cytokine expression and chromatin remodeling in committed Th1 cells. J. Exp. Med. 192:105–15 24. Ouyang W, Ranganath S, Weindel K, Bhattacharya D, Murphy T, et al. 1998. Inhibition of Th1 development mediated by GATA-3 through an IL-4-independent mechanism. Immunity 9:745–55 25. Zheng W, Flavell RA. 1997. The transcription factor GATA-3 is necessary and sufﬁcient for Th2 cytokine gene expression in CD4 T cells. Cell 89:587–96 26. Pai SY, Truitt ML, Ho IC. 2004. GATA-3 deﬁciency abrogates the development and maintenance of T helper type 2 cells. Proc. Natl. Acad. Sci. USA 101:1993–98 27. Zhu J, Min B, Hu-Li J, Watson CJ, Grinberg A, et al. 2004. Conditional deletion of Gata3 shows its essential function in T(H)1-T(H)2 responses. Nat. Immunol. 5:1157– 65 28. Yamashita M, Ukai-Tadenuma M, Miyamoto T, Sugaya K, Hosokawa H, et al. 2004. Essential role of GATA3 for the maintenance of type 2 helper T (Th2) cytokine production and chromatin remodeling at the Th2 cytokine gene loci. J. Biol. Chem. 279:26983–90 29. Das J, Chen CH, Yang L, Cohn L, Ray P, Ray A. 2001. A critical role for NF-κB in GATA3 expression and TH2 differentiation in allergic airway inﬂammation. Nat. Immunol. 2:45– 50 30. Zhang DH, Yang L, Cohn L, Parkyn L, Homer R, et al. 1999. Inhibition of allergic inﬂammation in a murine model of asthma by expression of a dominant-negative mutant of GATA-3. Immunity 11:473–82 31. Zhang DH, Yang L, Ray A. 1998. Differential responsiveness of the IL-5 and IL-4 genes to transcription factor GATA-3. J. Immunol. 161:3817–21 32. Kishikawa H, Sun J, Choi A, Miaw SC, Ho IC. 2001. The cell type-speciﬁc expression of the murine IL-13 gene is regulated by GATA-3. J. Immunol. 167:4414–20 33. Zhu J, Cote-Sierra J, Guo L, Paul WE. 2003. Stat5 activation plays a critical role in Th2 differentiation. Immunity 19:739–48 34. Guermonprez P, Valladeau J, Zitvogel L, Th´ery C, Amigorena S. 2002. Antigen presentation and T cell stimulation by dendritic cells. Annu. Rev. Immunol. 20:621–67 35. Kapsenberg ML. 2003. Dendritic-cell control of pathogen-driven T-cell polarization. Nat. Rev. Immunol. 3:984–93 36. Weiss DL, Brown MA. 2001. Regulation of IL-4 production in mast cells: a paradigm for cell-type-speciﬁc gene expression. Immunol. Rev. 179:35–47

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

644

Ansel et al.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

37. Stetson DB, Mohrs M, Reinhardt RL, Baron JL, Wang ZE, et al. 2003. Constitutive cytokine mRNAs mark natural killer (NK) and NK T cells poised for rapid effector function. J. Exp. Med. 198:1069–76 38. Schmitz J, Thiel A, Kuhn R, Rajewsky K, Muller W, et al. 1994. Induction of interleukin 4 (IL-4) expression in T helper (Th) cells is not dependent on IL-4 from non-Th cells. J. Exp. Med. 179:1349–53 39. Brown DR, Fowell DJ, Corry DB, Wynn TA, Moskowitz NH, et al. 1996. Beta 2microglobulin-dependent NK1.1+ T cells are not essential for T helper cell 2 immune responses. J. Exp. Med. 184:1295–304 40. Smiley ST, Kaplan MH, Grusby MJ. 1997. Immunoglobulin E production in the absence of interleukin-4-secreting CD1-dependent cells. Science 275:977–79 41. Zhang Y, Rogers KH, Lewis DB. 1996. Beta 2-microglobulin-dependent T cells are dispensable for allergen-induced T helper 2 responses. J. Exp. Med. 184:1507–12 42. Akbari O, Stock P, Meyer E, Kronenberg M, Sidobre S, et al. 2003. Essential role of NKT cells producing IL-4 and IL-13 in the development of allergen-induced airway hyperreactivity. Nat. Med. 9:582–88 43. Grogan JL, Mohrs M, Harmon B, Lacy DA, Sedat JW, Locksley RM. 2001. Early transcription and silencing of cytokine genes underlie polarization of T helper cell subsets. Immunity 14:205–15 44. Ansel KM, Greenwald RJ, Agarwal S, Bassing CH, Monticelli S, et al. 2004. Deletion of a conserved Il4 silencer impairs T helper type 1-mediated immunity. Nat. Immunol. 5:1251–59 45. Noben-Trauth N, Hu-Li J, Paul WE. 2000. Conventional, naive CD4+ T cells provide an initial source of IL-4 during Th2 differentiation. J. Immunol. 165:3620–25 46. Croft M, Swain SL. 1995. Recently activated naive CD4 T cells can help resting B cells, and can produce sufﬁcient autocrine IL-4 to drive differentiation to secretion of T helper 2-type cytokines. J. Immunol. 154:4269–82 47. Demeure CE, Yang LP, Byun DG, Ishihara H, Vezzio N, Delespesse G. 1995. Human naive CD4 T cells produce interleukin-4 at priming and acquire a Th2 phenotype upon repetitive stimulations in neutral conditions. Eur. J. Immunol. 25:2722–25 48. Ono SJ. 2000. Molecular genetics of allergic diseases. Annu. Rev. Immunol. 18:347– 66 49. Wills-Karp M. 2000. The gene encoding interleukin-13: a susceptibility locus for asthma and related traits. Respir. Res. 1:19–23 50. Bix M, Wang ZE, Thiel B, Schork NJ, Locksley RM. 1998. Genetic regulation of commitment to interleukin 4 production by a CD4+ T cell-intrinsic mechanism. J. Exp. Med. 188:2289–99 51. Yagi R, Suzuki W, Seki N, Kohyama M, Inoue T, et al. 2002. The IL-4 production capability of different strains of naive CD4+ T cells controls the direction of the T(h) cell response. Int. Immunol. 14:1–11 52. Locksley RM, Pingel S, Lacy D, Wakil AE, Bix M, Fowell DJ. 1999. Susceptibility to infectious diseases: Leishmania as a paradigm. J. Infect. Dis. 179(Suppl. 2):S305–8 53. Karp CL, Grupe A, Schadt E, Ewart SL, Keane-Moore M, et al. 2000. Identiﬁcation of complement factor 5 as a susceptibility locus for experimental allergic asthma. Nat. Immunol. 1:221–26 54. McIntire JJ, Umetsu SE, Akbari O, Potter M, Kuchroo VK, et al. 2001. Identiﬁcation of Tapr (an airway hyperreactivity regulatory locus) and the linked Tim gene family. Nat. Immunol. 2:1109–16 www.annualreviews.org • Regulation of Th2 Differentiation

645

ARI

23 February 2006

15:32

55. Baguet A, Epler J, Wen KW, Bix M. 2004. A Leishmania major response locus identiﬁed by interval-speciﬁc congenic mapping of a T helper type 2 cell bias-controlling quantitative trait locus. J. Exp. Med. 200:1605–12 56. Scott P, Eaton A, Gause WC, di Zhou X, Hondowicz B. 1996. Early IL-4 production does not predict susceptibility to Leishmania major. Exp. Parasitol. 84:178–87 57. Stetson DB, Mohrs M, Mallet-Designe V, Teyton L, Locksley RM. 2002. Rapid expansion and IL-4 expression by Leishmania-speciﬁc naive helper T cells in vivo. Immunity 17:191– 200 58. Mohrs M, Holscher C, Brombacher F. 2000. Interleukin-4 receptor alpha-deﬁcient BALB/c mice show an unimpaired T helper 2 polarization in response to Leishmania major infection. Infect. Immun. 68:1773–80 59. Cunningham AF, Serre K, Toellner KM, Khan M, Alexander J, et al. 2004. Pinpointing IL-4-independent acquisition and IL-4-inﬂuenced maintenance of Th2 activity by CD4 T cells. Eur. J. Immunol. 34:686–94 60. Kusam S, Toney LM, Sato H, Dent AL. 2003. Inhibition of Th2 differentiation and GATA-3 expression by BCL-6. J. Immunol. 170:2435–41 61. Amsen D, Blander JM, Lee GR, Tanigaki K, Honjo T, Flavell RA. 2004. Instruction of distinct CD4 T helper cell fates by different notch ligands on antigen-presenting cells. Cell 117:515–26 62. Tanigaki K, Tsuji M, Yamamoto N, Han H, Tsukada J, et al. 2004. Regulation of αβ/γδ T cell lineage commitment and peripheral T cell responses by Notch/RBP-J signaling. Immunity 20:611–22 63. Minter LM, Turley DM, Das P, Shin HM, Joshi I, et al. 2005. Inhibitors of gammasecretase block in vivo and in vitro T helper type 1 polarization by preventing Notch upregulation of Tbx21. Nat. Immunol. 6:680–88 64. Maekawa Y, Tsukumo S, Chiba S, Hirai H, Hayashi Y, et al. 2003. Delta1-Notch3 interactions bias the functional differentiation of activated CD4+ T cells. Immunity 19:549– 59 65. Maillard I, Fang T, Pear WS. 2005. Regulation of lymphoid development, differentiation, and function by the Notch pathway. Annu. Rev. Immunol. 23:945–74 66. Yamaguchi E, Chiba S, Kumano K, Kunisato A, Takahashi T, et al. 2002. Expression of Notch ligands, Jagged1, 2 and Delta1 in antigen presenting cells in mice. Immunol. Lett. 81:59–64 67. Napolitani G, Rinaldi A, Bertoni F, Sallusto F, Lanzavecchia A. 2005. Selected Tolllike receptor agonist combinations synergistically trigger a T helper type 1-polarizing program in dendritic cells. Nat. Immunol. 6:769–76 67a. Tu L, Fang TC, Artis D, Shestova O, Pross SE, et al. 2005. Notch signaling is an important regulator of type 2 immunity. J. Exp. Med. 202:1037–42 68. Kaisho T, Hoshino K, Iwabe T, Takeuchi O, Yasui T, Akira S. 2002. Endotoxin can induce MyD88-deﬁcient dendritic cells to support T(h)2 cell differentiation. Int. Immunol. 14:695–700 69. Schnare M, Barton GM, Holt AC, Takeda K, Akira S, Medzhitov R. 2001. Tolllike receptors control activation of adaptive immune responses. Nat. Immunol. 2:947– 50 70. Mullen AC, High FA, Hutchins AS, Lee HW, Villarino AV, et al. 2001. Role of T-bet in commitment of TH1 cells before IL-12-dependent selection. Science 292:1907–10 71. Maldonado RA, Irvine DJ, Schreiber R, Glimcher LH. 2004. A role for the immunological synapse in lineage commitment of CD4 lymphocytes. Nature 431:527–32

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

646

Ansel et al.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

72. Usui T, Nishikomori R, Kitani A, Strober W. 2003. GATA-3 suppresses Th1 development by downregulation of Stat4 and not through effects on IL-12Rβ2 chain or T-bet. Immunity 18:415–28 73. Mullen AC, Hutchins AS, High FA, Lee HW, Sykes KJ, et al. 2002. Hlx is induced by and genetically interacts with T-bet to promote heritable T(H)1 gene induction. Nat. Immunol. 3:652–58 74. Zheng WP, Zhao Q, Zhao X, Li B, Hubank M, et al. 2004. Up-regulation of Hlx in immature Th cells induces IFN-γ expression. J. Immunol. 172:114–22 75. Ho I-C, Hodge MR, Rooney JW, Glimcher LH. 1996. The proto-oncogene c-maf is responsible for tissue-speciﬁc expression of interleukin-4. Cell 85:973–83 76. Hwang ES, White IA, Ho IC. 2002. An IL-4-independent and CD25-mediated function of c-maf in promoting the production of Th2 cytokines. Proc. Natl. Acad. Sci. USA 99:13026–30 77. Ben-Sasson SZ, Le Gros G, Conrad DH, Finkelman FD, Paul WE. 1990. IL-4 production by T cells from naive donors. IL-2 is required for IL-4 production. J. Immunol. 145:1127– 36 78. Cote-Sierra J, Foucras G, Guo L, Chiodetti L, Young HA, et al. 2004. Interleukin 2 plays a central role in Th2 differentiation. Proc. Natl. Acad. Sci. USA 101:3880–85 79. Takatori H, Nakajima H, Kagami S, Hirose K, Suto A, et al. 2005. Stat5a inhibits IL-12induced Th1 cell differentiation through the induction of suppressor of cytokine signaling 3 expression. J. Immunol. 174:4105–12 80. Takatori H, Nakajima H, Hirose K, Kagami S, Tamachi T, et al. 2005. Indispensable role of Stat5a in Stat6-independent Th2 cell differentiation and allergic airway inﬂammation. J. Immunol. 174:3734–40 81. Kagami S, Nakajima H, Suto A, Hirose K, Suzuki K, et al. 2001. Stat5a regulates T helper cell differentiation by several distinct mechanisms. Blood 97:2358–65 82. Messi M, Giacchetto I, Nagata K, Lanzavecchia A, Natoli G, Sallusto F. 2003. Memory and ﬂexibility of cytokine gene expression as separable properties of human T(H)1 and T(H)2 lymphocytes. Nat. Immunol. 4:78–86 83. Sundrud MS, Grill SM, Ni D, Nagata K, Alkan SS, et al. 2003. Genetic reprogramming of primary human T cells reveals functional plasticity in Th cell differentiation. J. Immunol. 171:3542–49 84. Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH. 2000. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100:655–69 85. Lametschwandtner G, Biedermann T, Schwarzler C, Gunther C, Kund J, et al. 2004. Sustained T-bet expression confers polarized human TH2 cells with TH1-like cytokine production and migratory capacities. J. Allergy Clin. Immunol. 113:987–94 86. Nagai S, Hashimoto S, Yamashita T, Toyoda N, Satoh T, et al. 2001. Comprehensive gene expression proﬁle of human activated T(h)1- and T(h)2-polarized cells. Int. Immunol. 13:367–76 87. Rogge L, Bianchi E, Bifﬁ M, Bono E, Chang SY, et al. 2000. Transcript imaging of the development of human T helper cells using oligonucleotide arrays. Nat. Genet. 25:96–101 88. Lu B, Zagouras P, Fischer JE, Lu J, Li B, Flavell RA. 2004. Kinetic analysis of genomewide gene expression reveals molecule circuitries that control T cell activation and Th1/2 differentiation. Proc. Natl. Acad. Sci. USA 101:3023–28 89. Chtanova T, Kemp RA, Sutherland AP, Ronchese F, Mackay CR. 2001. Gene microarrays reveal extensive differential gene expression in both CD4+ and CD8+ type 1 and type 2 T cells. J. Immunol. 167:3057–63 www.annualreviews.org • Regulation of Th2 Differentiation

647

ARI

23 February 2006

15:32

90. Hofer T, Nathansen H, Lohning M, Radbruch A, Heinrich R. 2002. GATA-3 transcriptional imprinting in Th2 lymphocytes: a mathematical model. Proc. Natl. Acad. Sci. USA 99:9364–68 91. Avni O, Lee D, Macian F, Szabo SJ, Glimcher LH, Rao A. 2002. T(H) cell differentiation is accompanied by dynamic changes in histone acetylation of cytokine genes. Nat. Immunol. 3:643–51 92. Skapenko A, Leipe J, Niesner U, Devriendt K, Beetz R, et al. 2004. GATA-3 in human T cell helper type 2 development. J. Exp. Med. 199:423–28 93. Lee DU, Avni O, Chen L, Rao A. 2004. A distal enhancer in the interferon-γ (IFN-γ) locus revealed by genome sequence comparison. J. Biol. Chem. 279:4802–10 94. Guo L, Hu-Li J, Zhu J, Watson CJ, Diﬁlippantonio MJ, et al. 2002. In TH2 cells the Il4 gene has a series of accessibility states associated with distinctive probabilities of IL-4 production. Proc. Natl. Acad. Sci. USA 99:10623–28 95. Guo L, Hu-Li J, Paul WE. 2004. Probabilistic regulation of IL-4 production in Th2 cells: accessibility at the Il4 locus. Immunity 20:193–203 96. Mohrs M, Shinkai K, Mohrs K, Locksley RM. 2001. Analysis of type 2 immunity in vivo with a bicistronic IL-4 reporter. Immunity 15:303–11 97. Bix M, Locksley RM. 1998. Independent and epigenetic regulation of the interleukin-4 alleles in CD4+ T cells. Science 281:1352–54 98. Guo L, Hu-Li J, Paul WE. 2005. Probabilistic regulation in TH2 cells accounts for monoallelic expression of IL-4 and IL-13. Immunity 23:89–99 99. Jenuwein T, Allis CD. 2001. Translating the histone code. Science 293:1074–80 100. Vakoc CR, Mandat SA, Olenchock BA, Blobel GA. 2005. Histone H3 lysine 9 methylation and HP1γ are associated with transcription elongation through mammalian chromatin. Mol. Cell 19:381–91 101. de la Cruz X, Lois S, Sanchez-Molina S, Martinez-Balbas MA. 2005. Do protein motifs read the histone code? BioEssays 27:164–75 102. Bottomley MJ. 2004. Structures of protein domains that create or recognize histone modiﬁcations. EMBO Rep. 5:464–69 103. Roth SY, Denu JM, Allis CD. 2001. Histone acetyltransferases. Annu. Rev. Biochem. 70:81–120 104. Kouzarides T. 2000. Acetylation: a regulatory modiﬁcation to rival phosphorylation? EMBO J. 19:1176–79 105. Narlikar GJ, Fan HY, Kingston RE. 2002. Cooperation between complexes that regulate chromatin structure and transcription. Cell 108:475–87 106. Lachner M, O’Sullivan RJ, Jenuwein T. 2003. An epigenetic road map for histone lysine methylation. J. Cell Sci. 116:2117–24 107. Kouzarides T. 2002. Histone methylation in transcriptional control. Curr. Opin. Genet. Dev. 12:198–209 108. Morshead KB, Ciccone DN, Taverna SD, Allis CD, Oettinger MA. 2003. Antigen receptor loci poised for V(D)J rearrangement are broadly associated with BRG1 and ﬂanked by peaks of histone H3 dimethylated at lysine 4. Proc. Natl. Acad. Sci. USA 100:11577–82 109. Bernstein BE, Kamal M, Lindblad-Toh K, Bekiranov S, Bailey DK, et al. 2005. Genomic maps and comparative analysis of histone modiﬁcations in human and mouse. Cell 120:169–81 110. Schneider R, Bannister AJ, Myers FA, Thorne AW, Crane-Robinson C, Kouzarides T. 2004. Histone H3 lysine 4 methylation patterns in higher eukaryotic genes. Nat. Cell Biol. 6:73–77

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

648

Ansel et al.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

111. Guenther MG, Jenner RG, Chevalier B, Nakamura T, Croce CM, et al. 2005. Global and Hox-speciﬁc roles for the MLL1 methyltransferase. Proc. Natl. Acad. Sci. USA 102:8603–8 112. Bedford MT, Richard S. 2005. Arginine methylation an emerging regulator of protein function. Mol. Cell 18:263–72 113. Cuthbert GL, Daujat S, Snowden AW, Erdjument-Bromage H, Hagiwara T, et al. 2004. Histone deimination antagonizes arginine methylation. Cell 118:545–53 114. Mowen KA, Schurter BT, Fathman JW, David M, Glimcher LH. 2004. Arginine methylation of NIP45 modulates cytokine gene expression in effector T lymphocytes. Mol. Cell 15:559–71 115. Bereshchenko OR, Gu W, Dalla-Favera R. 2002. Acetylation inactivates the transcriptional repressor BCL6. Nat. Genet. 32:606–13 116. Yamaguchi Y, Kurokawa M, Imai Y, Izutsu K, Asai T, et al. 2004. AML1 is functionally regulated through p300-mediated acetylation on speciﬁc lysine residues. J. Biol. Chem. 279:15630–38 117. Jin YH, Jeon EJ, Li QL, Lee YH, Choi JK, et al. 2004. Transforming growth factor-β stimulates p300-dependent RUNX3 acetylation, which inhibits ubiquitination-mediated degradation. J. Biol. Chem. 279:29409–17 118. Vaziri H, Dessain SK, Eagon EN, Imai SI, Frye RA, et al. 2001. hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase. Cell 107:149–59 119. Sun ZW, Allis CD. 2002. Ubiquitination of histone H2B regulates H3 methylation and gene silencing in yeast. Nature 418:104–8 120. Ng HH, Xu RM, Zhang Y, Struhl K. 2002. Ubiquitination of histone H2B by Rad6 is required for efﬁcient Dot1-mediated methylation of histone H3 lysine 79. J. Biol. Chem. 277:34655–57 121. Xiao T, Kao CF, Krogan NJ, Sun ZW, Greenblatt JF, et al. 2005. Histone H2B ubiquitylation is associated with elongating RNA polymerase II. Mol. Cell. Biol. 25:637–51 122. Shahbazian MD, Zhang K, Grunstein M. 2005. Histone H2B ubiquitylation controls processive methylation but not monomethylation by Dot1 and Set1. Mol. Cell 19:271–77 123. Fischle W, Wang Y, Allis CD. 2003. Binary switches and modiﬁcation cassettes in histone biology and beyond. Nature 425:475–79 124. Sarma K, Reinberg D. 2005. Histone variants meet their match. Nat. Rev. Mol. Cell Biol. 6:139–49 125. Pray-Grant MG, Daniel JA, Schieltz D, Yates JR 3rd, Grant PA. 2005. Chd1 chromodomain links histone H3 methylation with SAGA- and SLIK-dependent acetylation. Nature 433:434–38 126. Shi Y, Lan F, Matson C, Mulligan P, Whetstine JR, et al. 2004. Histone demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell 119:941–53 127. Cairns BR. 2005. Chromatin remodeling complexes: strength in diversity, precision through specialization. Curr. Opin. Genet. Dev. 15:185–90 128. Chi TH, Wan M, Zhao K, Taniuchi I, Chen L, et al. 2002. Reciprocal regulation of CD4/CD8 expression by SWI/SNF-like BAF complexes. Nature 418:195–99 129. Gebuhr TC, Kovalev GI, Bultman S, Godfrey V, Su L, Magnuson T. 2003. The role of Brg1, a catalytic subunit of mammalian chromatin-remodeling complexes, in T cell development. J. Exp. Med. 198:1937–49 130. Grewal SI, Moazed D. 2003. Heterochromatin and epigenetic control of gene expression. Science 301:798–802 131. Levine SS, King IF, Kingston RE. 2004. Division of labor in Polycomb group repression. Trends Biochem. Sci. 29:478–85 www.annualreviews.org • Regulation of Th2 Differentiation

649

ARI

23 February 2006

15:32

132. Hall IM, Shankaranarayana GD, Noma K, Ayoub N, Cohen A, Grewal SI. 2002. Establishment and maintenance of a heterochromatin domain. Science 297:2232–37 133. Brown KE, Guest SS, Smale ST, Hahm K, Merkenschlager M, Fisher AG. 1997. Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell 91:845–54 134. Su RC, Brown KE, Saaber S, Fisher AG, Merkenschlager M, Smale ST. 2004. Dynamic assembly of silent chromatin during thymocyte maturation. Nat. Genet. 36:502–6 135. Volpe TA, Kidner C, Hall IM, Teng G, Grewal SI, Martienssen RA. 2002. Regulation of heterochromatic silencing and histone H3 lysine-9 methylation by RNAi. Science 297:1833–37 136. Noma K, Sugiyama T, Cam H, Verdel A, Zofall M, et al. 2004. RITS acts in cis to promote RNA interference-mediated transcriptional and post-transcriptional silencing. Nat. Genet. 36:1174–80 137. Verdel A, Jia S, Gerber S, Sugiyama T, Gygi S, et al. 2004. RNAi-mediated targeting of heterochromatin by the RITS complex. Science 303:672–76 138. Cobb BS, Nesterova TB, Thompson E, Hertweck A, O’Connor E, et al. 2005. T cell lineage choice and differentiation in the absence of the RNase III enzyme Dicer. J. Exp. Med. 201:1367–73 139. Smale ST. 2003. The establishment and maintenance of lymphocyte identity through gene silencing. Nat. Immunol. 4:607–15 140. Jia S, Noma K, Grewal SI. 2004. RNAi-independent heterochromatin nucleation by the stress-activated ATF/CREB family proteins. Science 304:1971–76 141. Taniuchi I, Littman DR. 2004. Epigenetic gene silencing by Runx proteins. Oncogene 23:4341–45 142. Merkenschlager M, Amoils S, Roldan E, Rahemtulla A, O’Connor E, et al. 2004. Centromeric repositioning of coreceptor loci predicts their stable silencing and the CD4/CD8 lineage choice. J. Exp. Med. 200:1437–44 143. Delaire S, Huang YH, Chan SW, Robey EA. 2004. Dynamic repositioning of CD4 and CD8 genes during T cell development. J. Exp. Med. 200:1427–35 144. Ringrose L, Paro R. 2004. Epigenetic regulation of cellular memory by the Polycomb and Trithorax group proteins. Annu. Rev. Genet. 38:413–43 145. Fischle W, Wang Y, Jacobs SA, Kim Y, Allis CD, Khorasanizadeh S. 2003. Molecular basis for the discrimination of repressive methyl-lysine marks in histone H3 by Polycomb and HP1 chromodomains. Genes Dev. 17:1870–81 146. Kirmizis A, Bartley SM, Kuzmichev A, Margueron R, Reinberg D, et al. 2004. Silencing of human polycomb target genes is associated with methylation of histone H3 Lys 27. Genes Dev. 18:1592–605 147. Czermin B, Melﬁ R, McCabe D, Seitz V, Imhof A, Pirrotta V. 2002. Drosophila enhancer of Zeste/ESC complexes have a histone H3 methyltransferase activity that marks chromosomal Polycomb sites. Cell 111:185–96 148. Cao R, Wang L, Wang H, Xia L, Erdjument-Bromage H, et al. 2002. Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science 298:1039–43 149. Muller J, Hart CM, Francis NJ, Vargas ML, Sengupta A, et al. 2002. Histone methyltransferase activity of a Drosophila Polycomb group repressor complex. Cell 111:197– 208 150. Francis NJ, Kingston RE, Woodcock CL. 2004. Chromatin compaction by a polycomb group protein complex. Science 306:1574–77

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

650

Ansel et al.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

151. Cao R, Zhang Y. 2004. The functions of E(Z)/EZH2-mediated methylation of lysine 27 in histone H3. Curr. Opin. Genet. Dev. 14:155–64 152. Kimura M, Koseki Y, Yamashita M, Watanabe N, Shimizu C, et al. 2001. Regulation of Th2 cell differentiation by mel-18, a mammalian polycomb group gene. Immunity 15:275–87 153. Raaphorst FM, Otte AP, Meijer CJ. 2001. Polycomb-group genes as regulators of mammalian lymphopoiesis. Trends Immunol. 22:682–90 154. Reik W, Dean W, Walter J. 2001. Epigenetic reprogramming in mammalian development. Science 293:1089–93 155. Bruniquel D, Schwartz RH. 2003. Selective, stable demethylation of the interleukin-2 gene enhances transcription by an active process. Nat. Immunol. 4:235–40 156. Loots GG, Locksley RM, Blankespoor CM, Wang ZE, Miller W, et al. 2000. Identiﬁcation of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288:136–40 157. Agarwal S, Rao A. 1998. Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 9:765–75 158. Takemoto N, Koyano-Nakagawa N, Yokota T, Arai N, Miyatake S, Arai K-I. 1998. Th2speciﬁc DNAse I-hypersensitive sites in the murine IL-13 and IL-4 intergenic region. Int. Immunol. 10:1981–85 159. Takemoto N, Kamogawa Y, Jun Lee H, Kurata H, Arai KI, et al. 2000. Cutting edge: chromatin remodeling at the IL-4/IL-13 intergenic regulatory region for Th2-speciﬁc cytokine gene cluster. J. Immunol. 165:6687–91 160. Fields PE, Lee GR, Kim ST, Bartsevich VV, Flavell RA. 2004. Th2-speciﬁc chromatin remodeling and enhancer activity in the Th2 cytokine locus control region. Immunity 21:865–76 161. Tykocinski LO, Hajkova P, Chang HD, Stamm T, Sozeri O, et al. 2005. A critical control element for interleukin-4 memory expression in T helper lymphocytes. J. Biol. Chem. 280:28177–85 162. Yamashita M, Ukai-Tadenuma M, Kimura M, Omori M, Inami M, et al. 2002. Identiﬁcation of a conserved GATA3 response element upstream proximal from the interleukin-13 gene locus. J. Biol. Chem. 277:42399–408 163. Lee GR, Fields PE, Grifﬁn TJ, Flavell RA. 2003. Regulation of the Th2 cytokine locus by a locus control region. Immunity 19:145–53 164. Nardone J, Lee DU, Ansel KM, Rao A. 2004. Bioinformatics for the ‘bench biologist’: how to ﬁnd regulatory regions in genomic DNA. Nat. Immunol. 5:768–74 165. Lee GR, Fields PE, Flavell RA. 2001. Regulation of IL-4 gene expression by distal regulatory elements and GATA-3 at the chromatin level. Immunity 14:447–59 166. Mohrs M, Blankespoor CM, Wang ZE, Loots GG, Afzal V, et al. 2001. Deletion of a coordinate regulator of type 2 cytokine expression in mice. Nat. Immunol. 2:842–47 167. Solymar DC, Agarwal S, Bassing CH, Alt FW, Rao A. 2002. A 3 enhancer in the IL-4 gene regulates cytokine production by Th2 cells and mast cells. Immunity 17:41–50 168. Lee GR, Spilianakis CG, Flavell RA. 2005. Hypersensitive site 7 of the TH2 locus control region is essential for expressing TH2 cytokine genes and for long-range intrachromosomal interactions. Nat. Immunol. 6:42–48 169. Baguet A, Bix M. 2004. Chromatin landscape dynamics of the Il4-Il13 locus during T helper 1 and 2 development. Proc. Natl. Acad. Sci. USA 101:11410–15 170. Grogan JL, Wang ZE, Stanley S, Harmon B, Loots GG, et al. 2003. Basal chromatin modiﬁcation at the IL-4 gene in helper T cells. J. Immunol. 171:6672–79 www.annualreviews.org • Regulation of Th2 Differentiation

651

ARI

23 February 2006

15:32

171. Fields PE, Kim ST, Flavell RA. 2002. Cutting edge: changes in histone acetylation at the IL-4 and IFN-γ loci accompany Th1/Th2 differentiation. J. Immunol. 169:647– 50 172. Roh TY, Cuddapah S, Zhao K. 2005. Active chromatin domains are deﬁned by acetylation islands revealed by genome-wide mapping. Genes Dev. 19:542–52 173. Koyanagi M, Baguet A, Martens J, Margueron R, Jenuwein T, Bix M. 2005. EZH2 and histone 3 trimethyl lysine 27 associated with Il4 and Il13 gene silencing in TH1 cells. J. Biol. Chem. 280:31470–77 174. Burgess-Beusse B, Farrell C, Gaszner M, Litt M, Mutskov V, et al. 2002. The insulation of genes from external enhancers and silencing chromatin. Proc. Natl. Acad. Sci. USA 99(Suppl. 4):16433–37 175. Crawford GE, Holt IE, Mullikin JC, Tai D, Blakesley R, et al. 2004. Identifying gene regulatory elements by genome-wide recovery of DNase hypersensitive sites. Proc. Natl. Acad. Sci. USA 101:992–97 176. ENCODE Project Consort. 2004. The ENCODE (ENCyclopedia of DNA elements) project. Science 306:636–40 177. Cawley S, Bekiranov S, Ng HH, Kapranov P, Sekinger EA, et al. 2004. Unbiased mapping of transcription factor binding sites along human chromosomes 21 and 22 points to widespread regulation of noncoding RNAs. Cell 116:499–509 178. Santangelo S, Cousins DJ, Winkelmann NE, Staynov DZ. 2002. DNA methylation changes at human Th2 cytokine genes coincide with DNase I hypersensitive site formation during CD4+ T cell differentiation. J. Immunol. 169:1893–903 179. Lee DU, Agarwal S, Rao A. 2002. Th2 lineage commitment and efﬁcient IL-4 production involves extended demethylation of the IL-4 gene. Immunity 16:649–60 180. Makar KW, Perez-Melgosa M, Shnyreva M, Weaver WM, Fitzpatrick DR, Wilson CB. 2003. Active recruitment of DNA methyltransferases regulates interleukin 4 in thymocytes and T cells. Nat. Immunol. 4:1183–90 181. Bird JJ, Brown DR, Mullen AC, Moskowitz NH, Mahowald MA, et al. 1998. Helper T cell differentiation is controlled by the cell cycle. Immunity 9:229–37 182. Lee PP, Fitzpatrick DR, Beard C, Jessup HK, Lehar S, et al. 2001. A critical role for dnmt1 and DNA methylation in T cell development, function, and survival. Immunity 15:763–74 183. Makar KW, Wilson CB. 2004. DNA methylation is a nonredundant repressor of the Th2 effector program. J. Immunol. 173:4402–6 184. Hutchins AS, Mullen AC, Lee HW, Sykes KJ, High FA, et al. 2002. Gene silencing quantitatively controls the function of a developmental trans-activator. Mol. Cell 10:81– 91 185. Monticelli S, Lee DU, Nardone J, Bolton DL, Rao A. 2005. Chromatin-based regulation of cytokine transcription in Th2 cells and mast cells. Int. Immunol. 17:1513–24 186. Spilianakis CG, Flavell RA. 2004. Long-range intrachromosomal interactions in the T helper type 2 cytokine locus. Nat. Immunol. 5:1017–27 187. Recillas-Targa F, Pikaart MJ, Burgess-Beusse B, Bell AC, Litt MD, et al. 2002. Positioneffect protection and enhancer blocking by the chicken β-globin insulator are separable activities. Proc. Natl. Acad. Sci. USA 99:6883–88 188. Hural JA, Kwan M, Henkel G, Hock MB, Brown MA. 2000. An intron transcriptional enhancer element regulates IL-4 gene locus accessibility in mast cells. J. Immunol. 165:3239– 49

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

652

Ansel et al.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

189. Kwan M, Powell DR, Nachman TY, Brown MA. 2005. An intron GATA-binding site regulates chromatin accessibility and is essential for IL-4 gene expression in mast cells. Eur. J. Immunol. 35:1267–74 190. Engel JD, Tanimoto K. 2000. Looping, linking, and chromatin activity: new insights into β-globin locus regulation. Cell 100:499–502 191. Adlam M, Siu G. 2003. Hierarchical interactions control CD4 gene expression during thymocyte development. Immunity 18:173–84 192. Dekker J. 2003. A closer look at long-range chromosomal interactions. Trends Biochem. Sci. 28:277–80 193. Carter D, Chakalova L, Osborne CS, Dai YF, Fraser P. 2002. Long-range chromatin regulatory interactions in vivo. Nat. Genet. 32:623–26 194. Tolhuis B, Palstra RJ, Splinter E, Grosveld F, de Laat W. 2002. Looping and interaction between hypersensitive sites in the active β-globin locus. Mol. Cell 10:1453–65 195. Palstra RJ, Tolhuis B, Splinter E, Nijmeijer R, Grosveld F, de Laat W. 2003. The βglobin nuclear compartment in development and erythroid differentiation. Nat. Genet. 35:190–94 196. Patrinos GP, de Krom M, de Boer E, Langeveld A, Imam AM, et al. 2004. Multiple interactions between regulatory regions are required to stabilize an active chromatin hub. Genes Dev. 18:1495–509 197. Vakoc CR, Letting DL, Gheldof N, Sawado T, Bender MA, et al. 2005. Proximity among distant regulatory elements at the β-globin locus requires GATA-1 and FOG-1. Mol. Cell 17:453–62 198. Monticelli S, Solymar DC, Rao A. 2004. Role of NFAT proteins in IL13 gene transcription in mast cells. J. Biol. Chem. 279:36210–18 199. Valapour M, Guo J, Schroeder JT, Keen J, Cianferoni A, et al. 2002. Histone deacetylation inhibits IL4 gene expression in T cells. J. Allergy Clin. Immunol. 109:238–45 200. Yamashita M, Shinnakasu R, Nigo Y, Kimura M, Hasegawa A, et al. 2004. Interleukin (IL)-4-independent maintenance of histone modiﬁcation of the IL-4 gene loci in memory Th2 cells. J. Biol. Chem. 279:39454–64 201. Lachner M, O’Carroll D, Rea S, Mechtler K, Jenuwein T. 2001. Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature 410:116–20 202. Hewitt SL, High FA, Reiner SL, Fisher AG, Merkenschlager M. 2004. Nuclear repositioning marks the selective exclusion of lineage-inappropriate transcription factor loci during T helper cell differentiation. Eur. J. Immunol. 34:3604–13 203. Trinh LA, Ferrini R, Cobb BS, Weinmann AS, Hahm K, et al. 2001. Down-regulation of TDT transcription in CD4+ CD8+ thymocytes by Ikaros proteins in direct competition with an Ets activator. Genes Dev. 15:1817–32 204. Akasaka T, Tsuji K, Kawahira H, Kanno M, Harigaya K, et al. 1997. The role of mel-18, a mammalian Polycomb group gene, during IL-7-dependent proliferation of lymphocyte precursors. Immunity 7:135–46 205. Atchison L, Ghias A, Wilkinson F, Bonini N, Atchison ML. 2003. Transcription factor YY1 functions as a PcG protein in vivo. EMBO J. 22:1347–58 206. Sweetser MT, Hoey T, Sun YL, Weaver WM, Price GA, Wilson CB. 1998. The roles of nuclear factor of activated T cells and ying-yang 1 in activation-induced expression of the interferon-γ promoter in T cells. J. Biol. Chem. 273:34775–83 207. Guo J, Casolaro V, Seto E, Yang WM, Chang C, et al. 2001. Yin-Yang 1 activates interleukin-4 gene expression in T cells. J. Biol. Chem. 276:48871–78 208. Schwenger GT, Fournier R, Hall LM, Sanderson CJ, Mordvinov VA. 1999. Nuclear www.annualreviews.org • Regulation of Th2 Differentiation

653

ANRV270-IY24-20

ARI

23 February 2006

209. 210. 211.

212.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

213.

214. 215.

216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231.

654

15:32

factor of activated T cells and YY1 combine to repress IL-5 expression in a human T-cell line. J. Allergy Clin. Immunol. 104:820–27 Orlando V. 2003. Polycomb, epigenomes, and control of cell identity. Cell 112:599–606 Schmitt S, Prestel M, Paro R. 2005. Intergenic transcription through a polycomb group response element counteracts silencing. Genes Dev. 19:697–708 Rank G, Prestel M, Paro R. 2002. Transcription through intergenic chromosomal memory elements of the Drosophila bithorax complex correlates with an epigenetic switch. Mol. Cell. Biol. 22:8026–34 Bender W, Fitzgerald DP. 2002. Transcription activates repressed domains in the Drosophila bithorax complex. Development 129:4923–30 Hogga I, Karch F. 2002. Transcription through the iab-7 cis-regulatory domain of the bithorax complex interferes with maintenance of Polycomb-mediated silencing. Development 129:4915–22 Rogan DF, Cousins DJ, Staynov DZ. 1999. Intergenic transcription occurs throughout the human IL-4/IL-13 gene cluster. Biochem. Biophys. Res. Commun. 255:556–61 Rogan DF, Cousins DJ, Santangelo S, Ioannou PA, Antoniou M, et al. 2004. Analysis of intergenic transcription in the human IL-4/IL-13 gene cluster. Proc. Natl. Acad. Sci. USA 101:2446–51 Hampsey M, Reinberg D. 2003. Tails of intrigue: phosphorylation of RNA polymerase II mediates histone methylation. Cell 113:429–32 Herbert A. 2004. The four Rs of RNA-directed evolution. Nat. Genet. 36:19–25 Rosok O, Sioud M. 2004. Systematic identiﬁcation of sense-antisense transcripts in mammalian cells. Nat. Biotechnol. 22:104–8 Lavorgna G, Dahary D, Lehner B, Sorek R, Sanderson CM, Casari G. 2004. In search of antisense. Trends Biochem. Sci. 29:88–94 Muljo SA, Ansel KM, Kanellopoulou C, Livingston DM, Rao A, Rajewsky K. 2005. Aberrant T cell differentiation in the absence of Dicer. J. Exp. Med. 202:261–69 Jing Q, Huang S, Guth S, Zarubin T, Motoyama A, et al. 2005. Involvement of microRNA in AU-rich element-mediated mRNA instability. Cell 120:623–34 Hannon GJ, Rossi JJ. 2004. Unlocking the potential of the human genome with RNA interference. Nature 431:371–78 Dykxhoorn DM, Lieberman J. 2005. The silent revolution: RNA interference as basic biology, research tool, and therapeutic. Annu. Rev. Med. 56:401–23 Sontheimer EJ, Carthew RW. 2005. Silence from within: endogenous siRNAs and miRNAs. Cell 122:9–12 Zamore PD, Haley B. 2005. Ribo-gnome: the big world of small RNAs. Science 309:1519– 24 Meister G, Tuschl T. 2004. Mechanisms of gene silencing by double-stranded RNA. Nature 431:343–49 Baulcombe D. 2004. RNA silencing in plants. Nature 431:356–63 Voinnet O. 2005. Induction and suppression of RNA silencing: insights from viral infections. Nat. Rev. Genet. 6:206–20 Martienssen RA, Zaratiegui M, Goto DB. 2005. RNA interference and heterochromatin in the ﬁssion yeast Schizosaccharomyces pombe. Trends Genet. 21:450–56 Ambros V. 2004. The functions of animal microRNAs. Nature 431:350–55 Bartel DP. 2004. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116:281–97

Ansel et al.

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

ARI

23 February 2006

15:32

232. He L, Hannon GJ. 2004. MicroRNAs: small RNAs with a big role in gene regulation. Nat. Rev. Genet. 5:522–31 233. Grifﬁths-Jones S. 2004. The microRNA Registry. Nucleic Acids Res. 32:D109–11 234. Bentwich I, Avniel A, Karov Y, Aharonov R, Gilad S, et al. 2005. Identiﬁcation of hundreds of conserved and nonconserved human microRNAs. Nat. Genet. 37:766–70 235. Berezikov E, Guryev V, van de Belt J, Wienholds E, Plasterk RH, Cuppen E. 2005. Phylogenetic shadowing and computational identiﬁcation of human microRNA genes. Cell 120:21–24 236. Hobert O. 2004. Common logic of transcription factor and microRNA action. Trends Biochem. Sci. 29:462–68 237. Cullen BR. 2004. Transcription and processing of human microRNA precursors. Mol. Cell 16:861–65 238. Denli AM, Tops BB, Plasterk RH, Ketting RF, Hannon GJ. 2004. Processing of primary microRNAs by the Microprocessor complex. Nature 432:231–35 239. Gregory RI, Yan KP, Amuthan G, Chendrimada T, Doratotaj B, et al. 2004. The Microprocessor complex mediates the genesis of microRNAs. Nature 432:235–40 240. Bernstein E, Allis CD. 2005. RNA meets chromatin. Genes Dev. 19:1635–55 241. Wassenegger M. 2005. The role of the RNAi machinery in heterochromatin formation. Cell 122:13–16 242. Tang G. 2005. siRNA and miRNA: an insight into RISCs. Trends Biochem. Sci. 30:106– 14 243. Meister G, Landthaler M, Patkaniowska A, Dorsett Y, Teng G, Tuschl T. 2004. Human Argonaute2 mediates RNA cleavage targeted by miRNAs and siRNAs. Mol. Cell 15:185– 97 244. Verdel A, Jia S, Gerber S, Sugiyama T, Gygi S, Grewal SI, Moazed D. 2004. RNAimediated targeting of heterochromatin by the RITS complex. Science 303:672–76 245. Noma K, Sugiyama T, Cam H, Verdel A, Zofall M, et al. 2004. RITS acts in cis to promote RNA interference-mediated transcriptional and post-transcriptional silencing. Nat. Genet. 36:1174–80 246. Hall IM, Noma K, Grewal SI. 2003. RNA interference machinery regulates chromosome dynamics during mitosis and meiosis in ﬁssion yeast. Proc. Natl. Acad. Sci. USA 100:193–98 247. Morris KV, Chan SW, Jacobsen SE, Looney DJ. 2004. Small interfering RNA-induced transcriptional gene silencing in human cells. Science 305:1289–92 248. Kawasaki H, Taira K. 2004. Induction of DNA methylation and gene silencing by short interfering RNAs in human cells. Nature 431:211–17 249. Kanellopoulou C, Muljo SA, Kung AL, Ganesan S, Drapkin R, et al. 2005. Dicer-deﬁcient mouse embryonic stem cells are defective in differentiation and centromeric silencing. Genes Dev. 19:489–501 250. Murchison EP, Partridge JF, Tam OH, Chelouﬁ S, Hannon GJ. 2005. Characterization of Dicer-deﬁcient murine embryonic stem cells. Proc. Natl. Acad. Sci. USA 102:12135– 40 251. Fukagawa T, Nogami M, Yoshikawa M, Ikeno M, Okazaki T, et al. 2004. Dicer is essential for formation of the heterochromatin structure in vertebrate cells. Nat. Cell Biol. 6:784–91 252. Chen CZ, Li L, Lodish HF, Bartel DP. 2004. MicroRNAs modulate hematopoietic lineage differentiation. Science 303:83–86 253. McManus MT. 2004. Small RNAs and immunity. Immunity 21:747–56 254. Chowdhury D, Novina CD. 2005. Potential roles for short RNAs in lymphocytes. Immunol. Cell Biol. 83:201–10 www.annualreviews.org • Regulation of Th2 Differentiation

655

ARI

23 February 2006

15:32

255. Cobb BS, Nesterova TB, Thompson E, Hertweck A, O’Connor E, et al. 2005. T cell lineage choice and differentiation in the absence of the RNase III enzyme Dicer. J. Exp. Med. 201:1367–73 256. Bernstein E, Kim SY, Carmell MA, Murchison EP, Alcorn H, et al. 2003. Dicer is essential for mouse development. Nat. Genet. 35:215–17 257. Harfe BD, McManus MT, Mansﬁeld JH, Hornstein E, Tabin CJ. 2005. The RNaseIII enzyme Dicer is required for morphogenesis but not patterning of the vertebrate limb. Proc. Natl. Acad. Sci. USA 102:10898–903 258. Monticelli S, Ansel KM, Xiao C, Socci ND, Krichevsky AM, et al. 2005. MicroRNA proﬁling of the murine hematopoietic system. Genome Biol. 6:R71 259. Lewis BP, Burge CB, Bartel DP. 2005. Conserved seed pairing, often ﬂanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120:15– 20 260. Krek A, Grun D, Poy MN, Wolf R, Rosenberg L, et al. 2005. Combinatorial microRNA target predictions. Nat. Genet. 37:495–500 261. John B, Enright AJ, Aravin A, Tuschl T, Sander C, Marks DS. 2004. Human MicroRNA targets. PLoS Biol. 2:e363

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-20

656

Ansel et al.

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:607-656. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett Department of Medicine, Biomedical Sciences Program, University of California, San Francisco, California 94143–0451; email: [email protected]

Annu. Rev. Immunol. 2006. 24:657–79 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090727 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0657$20.00

Key Words cytokines, lymphoid homeostasis, trans-presentation

Abstract IL-2, IL-15, and IL-7 are cytokines that are critical for regulating lymphoid homeostasis. These cytokines stimulate similar responses from lymphocytes in vitro, but play markedly divergent roles in lymphoid biology in vivo. Their distinct physiological functions can be ascribed to distinct signaling pathways initiated by proprietary cytokine receptor chains, differential expression patterns of the cytokines or their receptor chains, and/or signals occurring in distinct physiological contexts. Recently, the discovery of a novel mechanism of cytokine signaling, trans-presentation, has provided further insights into the different ways these cytokines function. Transpresentation also raises several novel cell biological and cellular implications concerning how cytokines support lymphoid homeostasis.

657

ANRV270-IY24-21

ARI

15 February 2006

16:33

INTRODUCTION IL: interleukin NK: natural killer NKT cell: natural killer T cell

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

IEL: intraepithelial lymphocyte

IL-2, IL-15, and IL-7 bind to multimeric receptors that share the common γ chain (γc ) and can stimulate cellular activation, survival, and/or proliferation in vitro (1–4). The roles of these cytokines in regulating lymphoid homeostasis have been reviewed in recent years, but new and, in several instances, unexpected developments have highlighted the importance of these molecules as well as forced reevaluation of their functions (5, 6). IL-2 binds to heterotrimeric receptors composed of IL-2Rα, IL-2/15Rβ, and γc . IL-2Rα is a proprietary receptor chain that is speciﬁc for IL-2, binds IL-2 with low afﬁnity (Kd ∼ = 10−8 M), and possesses a short cytoplasmic domain that does not appear to recruit intracytoplasmic signaling molecules. IL-2/15Rβ is a chain shared by the IL-15 receptor that is responsible for stimulating JAK3-, STAT5-, and AKT-dependent signaling pathways that support cellular survival and proliferation. The γc chain completes the IL-2 receptor complex, raising its binding afﬁnity for IL-2 to Kd ∼ = 10−11 M. γc confers the ability of IL-2 as well as other γc cytokines to stimulate MAP kinase and PI3 kinase pathways that lead to mitogenic and antiapoptotic signals. The IL-15 receptor is thought to be a heterotrimeric receptor that closely parallels the IL-2 receptor, except that the IL-15Rα chain substitutes for IL-2Rα in complex with IL-2/15Rβ and γc . IL-15Rα differs from IL-2Rα in that it alone binds IL-15 with high afﬁnity (Kd ∼ = 10−11 M), a somewhat surprising ﬁnding given the similarities of these ligand-receptor pairs and the Table 1

Knockout phenotypes of IL-2, IL-7, IL-15 cytokines and their receptors

Cytokine

Receptor chains

Knockout phenotype

IL-2

Activated CD4+ T cell accumulation, autoimmunity in older mice

IL-7

T cell, B cell, NK cell, NKT cell, IEL deﬁciency NK cells, IEL, memory phenotype CD8+ T cells deﬁciency

IL-15

658

fact that IL-15Rα possesses only one ligandbinding (“sushi”) domain compared with IL2Rα’s two sushi domains. The IL-7 receptor is a heterodimer composed of IL-7Rα, the speciﬁc receptor chain for IL-7 that can also recruit signaling molecules involved in antiapoptotic signals, and γc . Extensive work has been done in the area of cytokine signaling, and this topic has been reviewed recently (7). Studies with gene-targeted mice demonstrate that signals from IL-2, IL-15, and IL-7 are critical for lymphoid homeostasis. The phenotypes of IL-2−/− , IL-2Rα−/− , and IL-2/ 15Rβ−/− mice all exhibit spontaneous accumulation of activated T lymphocytes and inﬂammatory disease that lead to premature death (8–10) (Table 1). This inﬂammatory disease is primarily driven by the T lymphocytes (11). Hence, despite a plethora of in vitro evidence showing that IL-2 was critical for supporting T cell activation, the dominant physiological function of IL-2 signals in vivo is to restrain T cell activation and prevent autoimmunity. The similar phenotypes of IL-2−/− and IL-2Rα−/− mice indicate that virtually all physiological IL-2 signals require IL-2Rα. Additional defects in NK cells, NKT cells, and subsets of intraepithelial lymphocytes (IELs) in IL-2/15Rβ−/− mice conﬁrm the requirement of the IL-2/15Rβ chain in supporting IL-15 as well as IL-2 signaling (12). IL-15−/− and IL-15Rα−/− mice both exhibit selective losses of memory phenotype CD8+ T cells, NK cells, NKT cells, and subsets of IELs, indicating that IL-15 signals

(IL-2)

IL-2Rα

Similar to IL-2−/− mice

(IL-7)

IL-7Rα

Similar to IL-7−/− mice but phenotype slightly more severe

(IL-15)

IL-15Rα

Similar to IL-15−/− mice

(IL-2/15)

IL-2Rβ

T cell driven autoimmunity, NK cells, NKT cells, IEL deﬁciency

(IL-2/7/15)

γc

T cell, B cell deﬁciency

· ·

Burkett

Ma

Koka

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

provide essential positive homeostatic functions for these subsets of cells (13, 14). The phenotypes of these two strains of mice are also very similar, indicating that IL-15Rα is probably required for physiologically relevant IL-15 signals. Additional IL-15Rαindependent signaling mechanisms for IL-15 have been proposed but not conﬁrmed in vivo (15). IL-7−/− and IL-7Rα−/− mice exhibit severe lymphoid hypoplasia, including deﬁciencies of both B and T lymphocytes, demonstrating that the IL-7 signals are critical for positively supporting the development and/or homeostasis of these cells (16, 17). The lymphopenia of IL-7Rα−/− mice is a bit more severe than that of IL-7−/− mice, which may be due to the ability of thymic stromal lymphopoietin (TSLP) to bind to IL-7Rα and support T cell development (18). The phenotypes of these various genetargeted mice establish that IL-2-, IL-7-, and IL-15-induced activation, survival, and/or proliferation signals regulate distinct aspects of lymphoid homeostasis in vivo. Understanding the divergent phenotypes of these mice requires an understanding of the expression patterns of these cytokines and their receptors, as well as the contexts in which their signals are delivered.

T CELL DEVELOPMENT Reconciling the cell biology of IL-2, IL-15, and IL-7 signals with their physiological functions in vivo can be approached by examining detailed studies of T lymphocytes and NK cells. The dramatic T cell lymphopenia observed in X-SCID patients (which bear mutations in γc ) and γc −/− mice indicates that one or more of the γc cytokines is important for T cell development. Expression of IL-2, IL-15, and IL-7 receptor chains have all been described in thymocytes, suggesting that they may all regulate T cell differentiation. CD3− CD4− CD8− , or triple-negative (TN), thymocytes proceed through two stages of differentiation, during which they express IL-

2Rα. IL-2/15Rβ expression is also induced on CD4+ CD8+ double-positive (DP) thymocytes during positive selection (19). These cells can produce IL-2 in response to T cell receptor (TCR) ligation. However, conventional thymic or T cell subsets are present in normal numbers in young IL-2−/− , IL2Rα−/− , and IL-2/15Rβ−/− mice, and thymic selection events appear to proceed normally in the absence of IL-2 (9, 10, 20). Hence, the physiological roles for IL-2 signaling in conventional T cell development are unclear. By contrast, IL-2 signaling may play a major role in the differentiation of regulatory T cells, or Tregs, a subset of CD4+ T cells that expresses the transcription factor FoxP3 as well as high levels of IL-2Rα and IL-2/15Rβ (21). Tregs are thought to differentiate in the thymus during the perinatal period. IL-2−/− , IL-2Rα−/− , and IL-2/15Rβ−/− mice all lack Tregs, and Tregs can be reconstituted in these mice by complementing IL-2−/− mice with IL-2, or by complementing IL-2/15Rβ−/− mice with IL-2/15Rβ+ cells or Tregs during this period (22). Hence, the essential, nonredundant functions of IL-2 signaling during T cell differentiation appear restricted to Tregs. IL-15 and IL-15Rα are expressed in the thymus, so IL-15 signals may support the intrathymic differentiation of T cells. Heterologous IL-15 can induce the expression of Bcl-x and CD44 in CD8+ thymocytes (23). Initial studies suggested that the number of singlepositive (SP) CD8+ thymocytes and naive CD8+ T cells were reduced in IL-15Rα−/− mice (13). However, follow-up studies have not revealed an obvious defect in thymic selection (P. Burkett, A. Ma, unpublished data). Thus, although it is possible that IL-15 signaling may support CD8+ T cell differentiation, compelling evidence in this regard is currently lacking. IL-7Rα is highly expressed on TN thymocytes, absent from most DP thymocytes, and induced again on thymocytes undergoing positive selection (24). IL-7 is expressed by thymic stroma, and treatment of mice with anti-IL-7 antibody causes marked depletion www.annualreviews.org • IL-2, IL-15, and IL-7

TSLP: thymic stromal lymphopoietin TCR: T cell receptor Treg: regulatory T cell

659

ANRV270-IY24-21

ARI

15 February 2006

16:33

of thymocytes (25). The numbers of thymocytes are markedly reduced in both IL7Rα−/− and IL-7−/− mice, conﬁrming that IL-7-induced signals via IL-7Rα are critical for supporting development of most T cells. In addition, IL-7Rα−/− mice contain fewer thymocytes than IL-7−/− mice, and this difference may be due to the absence of TSLP signaling, a ligand that also binds to IL-7Rα (16, 25, 26). Consistent with the expression of IL-7Rα on TN thymocytes, T cell differentiation is blocked at this stage in IL-7−/− mice (27). IL7 may promote thymocyte survival by upregulating thymocyte Bcl-2 expression because TN thymocytes from γc −/− mice have diminished Bcl-2 levels, and transgenic expression of Bcl-2 in IL-7Rα−/− thymocytes partially restores thymic cellularity (27, 28). However, as heterologous expression of Bcl-2 fails to restore thymic cellularity fully, IL-7 signals may also serve Bcl-2-independent functions (29– 31). Other relevant antiapoptotic factors include Mcl-1, a Bcl-2 family member that is induced by IL-7 and binds to the proapoptotic BH3-only Bcl-2 family member Bim, and Bcl-x, another Bcl-2 family member that is expressed in thymocytes and protects them from cell death (31, 32). IL-7 may also support TCRβ chain rearrangement during the TN stage by inducing recombination activating gene (RAG)-1 expression because RAG-1 expression is reduced in IL-7Rα−/− TN thymocytes, and a TCRαβ transgene can restore intrathymic T cell differentiation in IL-7Rα−/− mice (33). The downregulation of IL-7Rα in DP thymocytes may also be important for T cell differentiation because deregulated transgenic expression of IL-7Rα in DP cells leads to reduced Bcl-2 expression and survival of TN cells. This phenomenon could be due to competition of IL-7Rα-bearing DP cells for limiting intrathymic IL-7, suggesting that the downregulation of IL-7Rα in DP thymocytes serves to prevent such competition as the DP cells expand (24). Hence, IL-7 signaling may support intrathymic T cell differentiation in several ways.

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

RAG-1: recombinant activation gene 1

660

Ma

· · Koka

Burkett

In summary, IL-2 is clearly essential for Treg differentiation, although why this subset of T cells is uniquely sensitive to IL-2 is currently unclear. IL-15 may regulate CD8+ T cell differentiation, and IL-7 is critical for the differentiation of most T cells. Although the expression of IL-7Rα on virtually all thymocytes at certain stages of differentiation matches well with the phenotypes of IL-7−/− and IL-7Rα−/− mice, similarly broad expression of IL-2 and IL-15 receptors do not coincide well with the highly restricted roles that IL-2 and IL-15 appear to play in thymic development. T cell differentiation is also largely normal in IL-2/15Rβ−/− mice, indicating that IL-2 and IL-15, which might be predicted to deliver very similar signals, are not compensating for each other in IL-2−/− , IL-2Rα−/− , IL-15−/− , or IL-15Rα−/− mice.

NAIVE T CELL HOMEOSTASIS Once mature T cells exit the thymus, they depend on signals from MHC molecules as well as cytokines for their continued survival in the periphery (34). IL-2 would not be expected to regulate naive T cells directly because IL-2Rα and IL-2/15Rβ are not expressed at signiﬁcant levels on naive mature T cells. The major physiological role currently proven for IL-2 signals in naive T cell homeostasis involves the indirect role of IL2 in supporting Tregs, which in turn regulate naive T cell activation and memory phenotype T cells. Like IL-2Rα and IL-2/15Rβ, IL-15Rα is expressed at low levels in resting T cells, and IL-15 improves the survival of naive T cells in vitro. Naive CD8+ T cells express slightly higher levels of IL-2/15Rβ than naive CD4+ T cells, and the number of naive CD8+ T cells is slightly reduced in IL-15−/− and IL15Rα−/− mice (13, 14). This subtle phenotype could be due to modest defects in thymic production, or to modest problems in survival or proliferation of naive CD8+ T cells. IL-7Rα is expressed on resting T cells. Injection of anti-IL-7 blocking antibody into

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

thymectomized mice leads to a reduction of Bcl-2 levels in T cells and shortens the half-life of peripheral T cells to two to three weeks (35–37). IL-7 deprivation also leads to metabolic arrest, decreased cell size, reduced glycolysis, and delayed mitosis in response to stimuli in vitro. Restoring IL-7 rescues these metabolically inactive cells (36). These responses of naive mature T cells to IL7 probably reﬂect similar survival responses to those initiated by IL-7Rα- and γc -stimulated signal transduction pathways in thymocytes (38). Interestingly, CD4+ T cells, which express higher levels of IL-7Rα than do CD8+ T cells, appear to be more dependent on IL-7 signals for survival (37). In addition, IL-7 signals are also critical for supporting the homeostatic proliferation of naive T cells after adoptive transfer into lymphopenic mice. Although these cells may be partly compromised by a lack of survival signals, these studies suggest that IL-7 may also support T cell proliferation in vivo. Therefore, as in T cell differentiation, IL-7 plays the dominant role in supporting the homeostasis of mature naive T cells. The distinct functions of IL-2, IL-15, and IL-7 signals in naive T cell homeostasis correspond well with the expression patterns of the various receptor chains.

T CELL ACTIVATION The initial activation of naive T cells occurs in response to TCR engagement in the context of innate immune cell stimulation and costimulation. These events likely proceed independently of γc cytokine signals. However, the subsequent expansion of antigen-speciﬁc T cells requires cellular proliferation and survival signals that help determine the number of T cells that possess various afﬁnities for antigen and the number that differentiate along speciﬁc pathways. This early selection process is important as cell fate decisions made during this period will be ampliﬁed in the survivors and progeny generated during the early T cell response. The induction of cytokines and cytokine receptor chains for IL-2, IL-15,

and IL-7 during T cell activation suggest that these particular cytokines help regulate this process. Of these cytokines, IL-2 is induced predominantly by activated CD4+ T cells, IL15 mRNA is rapidly induced in macrophages and dendritic cells (DCs) in response to TLR stimuli and type I interferons, and the source of IL-7 early during inﬂammatory responses is less clear. Expression of IL-2, IL-2Rα, and IL-2Rβ are all induced in T cells after TCR engagement, and multiple in vitro studies demonstrate that T cell activation depends on the presence of IL-2. Recently activated T cells are the predominant source of IL-2 produced during immune responses, and these cells increase surface expression of both IL-2/15Rβ and IL-2Rα. This coincidence of expression raises the possibility that T cells may produce and respond to IL-2 in an autocrine or paracrine fashion. DCs may also express modest amounts of IL-2 in response to microbial stimuli (39), which may prime the initial T cell response. However, experiments directly interrogating the physiological requirements for DC-produced IL-2 have yet to be published. IL-2−/− , IL-2Rα−/− , and IL-2Rβ−/− mice contain supranormal rather than depleted numbers of activated T cells, demonstrating that T cell activation clearly occurs spontaneously in the complete absence of IL-2 signals (8–10, 20, 40). This homeostatic defect results predominantly from the absence of Tregs in these mice. Nevertheless, transgenic expression of IL-2 can support activated T cells in vivo (41), and activated antigenspeciﬁc IL-2Rα−/− CD8+ T cells fail to expand fully in nonlymphoid tissues of normal mice (42). These latter experiments studying adoptively transferred IL-2Rα−/− T cells in normal mice can more precisely dissect the role(s) of IL-2 signals on antigen-speciﬁc T cells because the cells are studied in mice bearing normal complements of Tregs. These studies suggest that IL-2 signals probably regulate certain aspects of conventional antigenspeciﬁc T cell activation in vivo. www.annualreviews.org • IL-2, IL-15, and IL-7

661

ARI

15 February 2006

16:33

IL-2 signals have also been proposed to restrain T cell activation by encouraging death receptor- (i.e., Fas or TNF receptor–) induced programmed cell death of activated T cells (43, 44). Proposed mechanisms for this induction include upregulating Fas ligand expression and downregulating c-FLIP expression (45). Although several studies have failed to reveal a negative role for IL-2Rα signals on activated T cells, one recent study suggests that presumptively autocrine IL-2 signals may restrict CD8+ T cell expansion (46). Among the questions that remain to be resolved are (a) how IL-2 may play both supportive and restrictive roles in the same transgenic OT-1 CD8+ T cells; (b) whether these differences are due to divergent contexts when the signals are delivered or whether some of these ﬁndings reﬂect indirect effects; and (c) whether distinct cell biological mechanisms of delivering cytokine signals distinguish physiologically positive and negative signals. Expression of both IL-15Rα and IL-2/ 15Rβ increases after TCR activation, and heterologous IL-15, like IL-2, IL-4, and IL-7, can support the survival and/or proliferation of activated T cells (41). Although IL15Rα expression is induced on both CD4+ and CD8+ T cells, the higher expression of IL-2/15Rβ on activated CD8+ T cells compared with activated CD4+ T cells renders CD8+ T cells more sensitive to IL-15 (47). Direct examination of antigen-speciﬁc CD8+ T cells’ ability to become activated in IL-15−/− mice suggested that these cells could expand normally in response to lymphocytic choriomeningitis virus (LCMV), but not to vesicular stomatitis virus (VSV) (48, 49). These cells expanded normally after either VSV or LCMV infection in IL-15Rα−/− mice. As experiments examining T cell responses to LCMV and VSV have previously revealed different conclusions about requirements for costimulatory molecules, IL-15 may regulate T cell activation during some but not all types of immune responses. Why VSV responses should be compromised in

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

662

Ma

· · Koka

Burkett

IL-15−/− but not in IL-15Rα−/− mice could be related to IL-15-dependent signals that use receptors other than IL-15Rα, such as IL-15RX, but such signals require further characterization (15). Unlike IL-2Rα and IL-15Rα, IL-7Rα expression on most naive CD8+ T cells is downregulated after TCR-induced activation, and selective retention of IL-7Rα expression correlates with the ability of effector CD8+ T cells to become memory CD8+ T cells (35, 50). Although expression of IL-7Rα can only be called a marker at this point, the tempting extrapolation from this correlation is that the differential expression of IL-7Rα on activated CD8+ T cell populations responding to TCR engagement regulates their IL-7 responsiveness and hence their ability to survive and become memory T cells (50, 51). This line of reasoning further raises the possibility that activated CD8+ T cells compete for a limited pool of IL-7 during their expansion and differentiation to effector and memory T cells. Such competition is reminiscent of studies in which aberrantly IL-7Rα-expressing thymocytes compromise the survival of cells expressing endogenous levels of IL-7Rα (24, 52). Although parallel studies with CD4+ T cells have not yet been reported, the transition from effector CD4+ T cells into memory T cells occurs poorly after adoptive transfer into IL-7−/− mice (53). Thus, IL-7 likely provides critical survival signals to both activated CD4+ and CD8+ T cells and supports their transition to memory T cells in vivo. In summary, although early T cell activation can occur in the absence of IL-2, IL15, and IL-7 signals, these cytokines probably support recently activated T cells and thereby mold the characteristics of T cell responses. The microenvironment in which paracrine or autocrine IL-2 and IL-7 signals are delivered, potential competition between activating T cells for limited cytokines, and dynamic regulation of cytokine receptor expression levels are among the factors that modulate how these cytokines speciﬁcally support distinct aspects of T cell activation.

ANRV270-IY24-21

ARI

15 February 2006

16:33

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MEMORY T CELL MAINTENANCE One of the key features of memory T cells is their ability to survive as lineages of expanded cells in the absence of antigen- or TCR-mediated signals. Memory T cells rely predominantly on IL-2, IL-15, and IL-7 cytokines for their homeostasis during these prolonged periods of quiescence. Given the roles of cytokines in regulating naive and recently activated T cells, rigorous interrogation of the functions of cytokines in maintaining memory T cells has largely relied on adoptive transfer studies where memory T cells are generated under normal conditions and then transferred into environments lacking speciﬁc cytokines or receptor chains. Memory T cells express increased levels of IL-2Rα and IL-2/15Rβ, so IL-2 might be predicted to support memory T cells. However, although IL-2 signals indirectly regulate memory T cells via their direct effect on Treg differentiation, no in vivo evidence exists yet to support a direct role for IL-2 signals on memory T cell function (54). However, as newer data regarding memory T cell subsets emerge, studies with IL-2Rα−/− memory T cells may reveal functions for IL-2 in certain aspects of memory T cell function. Compelling data exist to support the idea that memory T cells depend on IL-7 signals. As noted earlier, IL-7Rα expression is preserved on a subset of activated CD8+ T cells that progress to become memory CD8+ T cells, so IL-7 may transmit the same survival signals to memory CD8+ T cells as it does to naive CD8+ T cells (35, 55). Memory CD8+ T cells survive and proliferate poorly after adoptive transfer into IL-7−/− mice (35). IL-7Rα−/− OT-1 CD8+ T cells, which can be activated, also survive poorly as memory CD8+ T cells (35). Although it is difﬁcult to distinguish survival from proliferative signals in vivo, the large numbers of memory phenotype T cells found in IL-7 transgenic mice may be the result of improved survival rather than enhanced proliferation of these

cells (56). These survival beneﬁts may be mediated through antiapoptotic molecules such as Bcl-2 or Mcl-1 (28, 31, 36). The cytokine dependence of memory CD4+ T cells is less evident than for memory CD8+ T cells. One study showed that γc −/− memory CD4+ T cells exhibited normal homeostatic proliferation and antigenic responses, suggesting that IL-2, IL-15, and IL-7 were all dispensable for these cells (57). However, memory CD4+ T cells have been shown to speciﬁcally require IL-7 signals in several other in vivo models (57–60). Reconciling these data may require studies of mice bearing conditionally deleted alleles of γc to exclude developmental biases of these cells, or examination of compound cytokine-deﬁcient mice to unveil potential interactions between γc cytokine signals. Nevertheless, the preponderance of data suggests that IL-7 signals are probably important for the function of memory CD4+ T cells. Multiple lines of evidence suggest that IL-15 signals are important for maintaining memory CD8+ T cells. Memory phenotype CD8+ T cells are selectively expanded by heterologous IL-15, consistent with the higher expression levels of IL-2/15Rβ on these cells compared with naive CD8+ T cells or CD4+ T cells (47). Memory phenotype CD8+ T cells are depleted in IL-15−/− mice, IL-15Rα−/− mice, and in normal mice treated with blocking antibodies against IL-2Rβ (presumptively blocking both IL-2 and IL-15 signals) but not in normal mice treated with antibodies against IL-2 or IL-2Rα (blocking IL-2) (13, 14, 47, 54). Direct examination of the putative role for IL-15 in supporting antigenspeciﬁc memory CD8+ T cells revealed that these cells could be generated normally in response to nonreplicating antigens as well as live viruses in the absence of IL-15 signals, but they could not be maintained over time (48, 49, 55, 61). The gradual loss of these cells over time correlates with reduced bromodeoxyuridine (BrdU) labeling and carboxyﬂuorescein succinimidyl ester (CFSE) dilution, suggesting that IL-15 supports www.annualreviews.org • IL-2, IL-15, and IL-7

663

ARI

15 February 2006

16:33

memory CD8+ T cell proliferation (48, 49, 55, 61). The undivided subpopulation of CFSE-labeled memory CD8+ T cells does not decline in the absence of IL-15, suggesting that IL-15 may not be essential for their survival. Therefore, although IL-7 may preferentially support memory CD8+ T cell survival, IL-15 appears to play a greater role in supporting their proliferation in vivo. The roles for IL-15 in supporting memory CD4+ T cells have been less well studied. Memory phenotype CD4+ T cells express lower levels of IL-2/15Rβ than do memory phenotype CD8+ T cells, and memory phenotype CD4+ T cells are present in normal numbers in IL-15Rα−/− mice, suggesting that IL-15 signals may be less important for the homeostasis of CD4+ T cells than for CD8+ T cells (13, 47). However, IL-7Rα signals are more important for supporting antigenspeciﬁc memory cells than for memory phenotype CD4+ T cells, so IL-15 signals may also be important for antigen-speciﬁc memory CD4+ T cells. Other proposed roles for IL-15 in memory T cell function have been less well corroborated by in vivo studies. Some studies using in vitro–activated CD8+ T cells treated with IL-2 or IL-15 suggested that these cytokines might differentially regulate the homing and differentiation of central versus effector subsets of memory CD8+ T cells (62). However, no differences in these memory CD8+ T cell subsets have been observed in IL-15−/− or IL-15Rα−/− mice, so the physiological signiﬁcance of studies using heterologous IL-15 in vitro remains unclear at present. Still, as the relationships between such memory T cell subsets is just beginning to be unraveled, future studies could reveal roles for IL-15 signals in this regard. One recent study demonstrated that IL15Rα could recycle intact IL-15 through endosomes and represent IL-15 to neighboring cells (63). This phenomenon was proposed as a mechanism by which IL-15-dependent cells (e.g., CTLL-2 cells) could survive longer after withdrawal of IL-15 than after withdrawal of

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

664

Ma

· · Koka

Burkett

IL-2 (63). As the expression of IL-15Rα is increased as naive CD8+ T cells become memory CD8+ T cells, it is possible that memory CD8+ T cells could use their IL-15Rα to maintain themselves in vivo. However, the expansion and contraction of activated CD8+ T cells and the maintenance of memory CD8+ T cells are similar whether these cells express IL-15Rα or not, suggesting that IL-15Rαmediated endosomal recycling does not confer any advantage to T cells (61, 64). Indeed, this surprising result indicates that IL-15Rα appears to play no cell-autonomous role in supporting memory T cell homeostasis. Given the biochemical evidence that IL-15Rα binds IL-15 with very high afﬁnity, it is challenging to understand why IL-15Rα expression on the surface of CD8+ T cells does not confer these cells with increased binding of IL-15, IL-15-dependent signaling, and a proliferative advantage. However, in vitro studies with antigen-speciﬁc IL-15Rα+/− and IL15Rα−/− memory cells that express the same levels of IL-2/15Rβ in fact reveal no difference in responsiveness to either soluble IL-15 or to plate-bound IL-15Rα/IL-15 complexes mimicking trans-presentation (61). Thus, it may be that IL-15Rα does not confer proliferative signals in a cell-autonomous fashion to CD8+ T cells. Differences with the prior studies may be ascribed to differences between human lymphoblastoid cell lines and primary murine lymphocytes. Alternatively, cell-autonomous IL-15Rα signals may stimulate responses other than homeostatic survival or proliferation. IL-15Rα on the surface of either memory phenotype CD8+ T cells or normal memory CD8+ T cells does not support the homeostasis of these cells. Yet in IL-15Rα−/− mice, normal memory CD8+ T cells fail to proliferate or maintain themselves as a lineage (61, 64, 65). Thus, IL-15Rα supports memory CD8+ T cells via an entirely noncell-autonomous mechanism, i.e., IL-15Rα must be expressed by accessory cells and not by memory CD8+ T cells to support the proliferation of the latter. This ﬁnding conﬁrms that physiological

ANRV270-IY24-21

ARI

15 February 2006

16:33

IL-15 signals are transmitted in a distinct fashion from other γc chain cytokines.

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CYTOKINE TRANS-PRESENTATION: A NOVEL MECHANISM OF SUPPORTING LYMPHOID HOMEOSTASIS The surprising ﬁnding that IL-15Rα supports memory CD8+ T cells in a noncellautonomous manner could be explained by at least two general models (Figure 1). In the ﬁrst model, IL-15 could bind to IL15Rα receptors on (as yet, undeﬁned) accessory cells which could then produce protein(s) that support IL-15-dependent cells such as memory CD8+ T cells (Figure 1). Multiple molecules and indirect mechanisms could fall into this general model. A second model, originally proposed for IL-2 and IL-2Rα, but not substantiated by in vivo studies, would involve trans-presentation of IL-15 by IL15Rα on the surface of accessory cells to IL-2/ 15Rβ and γc receptor chains on the surface of memory CD8+ T cells (63, 66). These models could be distinguished by the prediction that accessory cells would need to express IL-15Rα, IL-2/15Rβ, and γc to respond to IL-15 and generate secondary proteins that indirectly support memory CD8+ T cells (Figure 1, Indirect Signaling Model). By contrast, accessory cells would need only to express IL-15Rα (but not IL-2/15Rβ or γc ) to bind IL-15 and present it to memory CD8+ T cells (Figure 1, Trans-Presentation Model). Hence, accessory cells lacking IL-2/ 15Rβ should be able to support CD8+ T cells if trans-presentation is the relevant model, but not if indirect signals are required. Experiments in which IL-15Rα−/− mice are reconstituted with a mixture of IL-2/15Rβ−/− and IL-15Rα−/− hematopoietic stem cells demonstrate that IL-2/15Rβ−/− hematopoietic cells indeed sustain adoptively transferred memory CD8+ T cells as well as wild-type hematopoietic cells do, indicating that transpresentation is the dominant, if not exclu-

sive mechanism by which accessory cells use IL-15Rα to support memory CD8+ T cell homeostasis (67). In addition, these mice possess memory phenotype CD44hi CD8+ T cells from IL-15Rα−/− stem cells but not from IL2/15Rβ−/− stem cells, suggesting that CD8+ T cells must express IL-2/15Rβ but not IL2/15Rα to become memory cells. Therefore, expression of IL-15Rα on accessory cells and IL-2/15Rβ on memory CD8+ T cells are essential for memory CD8+ T cell homeostasis, whereas expression of IL-2/15Rβ on accessory cells and of IL-15Rα on memory CD8+ T cells play no role in this process. This genetic dissection indicates a cell biological trans-presentation mechanism in which IL-15Rα on the surface of accessory cells presents IL-15 to IL-2/15Rβ and γc receptor chains on the surface of memory CD8+ T cells (Figure 1, Trans-Presentation Model). The observation that IL-15 signals are delivered in trans by IL-15Rα-expressing accessory cells raises the question of which speciﬁc cells serve as trans-presenting cells. Radiation-sensitive (largely hematopoietic) cells appear to support memory better than radiation-resistant (stromal) cells (61, 67). Moreover, RAG-1−/− hematopoietic cells function as well as wild-type hematopoietic cells (67). Hence, myeloid cells appear to be the most likely candidates in this regard. Among myeloid cells, DCs are the most obvious candidates, given the large body of literature indicating direct interactions of DCs with T cells. Some recent evidence suggests that IL-15 can support DC survival and that adoptively transferred DCs can induce memory phenotype CD8+ T cells (68). Future studies using lineage-speciﬁc deletions of IL-15 or IL-15Rα will more directly address this question. The requirement for memory CD8+ T cells to receive trans-presented IL-15 in vivo also raises the question of why IL15Rα-expressing cells, which respond to picomolar concentrations of soluble IL-15 cytokine in vitro, do not respond to IL-15 in IL-15Rα−/− mice. Indeed, one might expect www.annualreviews.org • IL-2, IL-15, and IL-7

665

ANRV270-IY24-21

ARI

15 February 2006

16:33

Indirect signaling model Accessory cell

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

IL-2/15Rβ β

Trans-presentation model Accessory cell

γc IL-15Rα

Unknown signal(s)

IL-15

IL-15Rα IL-15 IL-2/15Rβ β

γc

IL-15dependent cell

IL-15dependent cell

Figure 1 Trans-presentation of IL-15 by IL-15Rα. Two general mechanisms that may explain the noncell-autonomous role of IL-15Rα in supporting IL-15-dependent cells. In the indirect signaling model (left), IL-15 (red circle) binds to heterotrimeric IL-15 receptors on accessory cells and triggers signals that lead to the production of additional proteins. These proteins, or other proteins generated by additional rounds of protein synthesis and cellular responses, stimulate the survival and/or proliferation of IL-15-responsive cells. This model predicts that accessory cells must express IL-2/15Rβ ( green) and γc ( purple) as well as IL-15Rα ( yellow) receptor chains. In the trans-presentation model (right), IL-15Rα on the surface of accessory cells alone binds IL-15 and presents IL-15 to IL-2/15Rβ and γc receptor chains on the surface of IL-15-responsive cells. This interaction directly stimulates proliferative and survival signals in the responding cells. In this model, accessory cells must express IL-15Rα but not IL-2/15Rβ or γc , and IL-15-responsive cells must express IL-2/15Rβ and γc but not IL-15Rα.

soluble IL-15 to be readily available in IL15Rα−/− mice, since it is not bound by IL15Rα. This conundrum led to the hypothesis that IL-15 may not be secreted as a soluble cytokine, and that IL-15Rα may be required 666

Ma

· · Koka

Burkett

for the production of IL-15. Indeed, mixed chimera generated from IL-15−/− and IL15Rα−/− cells fail to complement each other in supporting memory CD8+ T cells in vivo (67, 69). Therefore, IL-15 and IL-15Rα

ANRV270-IY24-21

ARI

15 February 2006

16:33

Paracrine expression of IL-15 IL-15producing cell

IL-15presenting cell

Autocrine expression of IL-15 IL-15producing and -presenting cell

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

IL-15

IL-15Rα

IL-15

γc

IL-15Rα

γc IL-2/15Rβ β

IL-15dependent cell

IL-2/15Rβ IL-15dependent cell

Figure 2 Coordinate expression of IL-15 and IL-15Rα. Two mechanisms by which the production of IL-15 could lead to its trans-presentation by IL-15Rα. In a paracrine expression model (left), IL-15 (red dot) is secreted by some cells. Soluble IL-15 is then bound to IL-15Rα ( yellow) on the surface of other accessory cells, which then present IL-15 to IL-15-responsive cells. This model predicts that IL-15−/− and IL-15Rα−/− cells would be able to complement each other in vivo. It also predicts that soluble IL-15 would be produced. In an autocrine expression model (right), IL-15 and IL-15Rα are produced by the same cell and may be preassociated prior to their emergence on the cell surface. In this model, IL-15−/− and IL-15Rα−/− cells would not be expected to complement each other, and soluble IL-15 might never be secreted.

need to be coexpressed by the same accessory cells to support IL-15-dependent lymphocytes in vivo (Autocrine Expression Model, Figure 2). As IL-15−/− DCs express normal amounts of surface IL-15Rα protein, and as IL-15 mRNA is expressed at normal levels in IL-15Rα−/− DCs both before and after stimulation with lipopolysaccharide, IL15Rα might be required for the translation of IL-15 mRNA, or alternatively for the processing or secretion of IL-15 protein (65, 70). Further implications of these remarkable ﬁndings are discussed below.

In summary, IL-7 is important for the survival of both memory CD4+ and CD8+ T cells, and IL-15 is important for the proliferation of memory CD8+ T cells. Memory T cells are likely to receive physiological IL-7 and IL-15 signals from distinct cell types in different locations. In addition, the novel mechanism by which IL-15 is trans-presented to lymphocytes indicates that these cytokine signals are delivered in different cellular and molecular contexts. Despite the expression of IL-2Rα on virtually all memory T cells, no clear role for IL-2 www.annualreviews.org • IL-2, IL-15, and IL-7

667

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

668

16:33

signals on these cells has been demonstrated to date.

family receptors in controlling viral pathogens (78).

NATURAL KILLER CELL DEVELOPMENT

MATURE NATURAL KILLER CELL HOMEOSTASIS

NK cells are innate immune cells involved in tumor surveillance and immunity against viruses. Receptors for IL-2, IL-15, and IL-7 are expressed in various stages of immature NK cells, and in vitro studies indicate that IL-2, IL-15, and IL-7 can all support NK cell differentiation (71–74). However, analyses of IL-2−/− , IL-2Rα−/− , IL-7−/− , and IL7Rα−/− mice fail to exhibit signiﬁcant defects in NK cell development (16, 26, 40, 75). Thus, although IL-2 and IL-7 may contribute in a redundant fashion to NK cell differentiation, the ability of these cytokines to heterologously support NK cell differentiation in vitro more probably does not represent their physiological functions. By contrast, IL-15−/− , IL15Rα−/− , IL-2Rβ−/− , and γc −/− mice all contain dramatically reduced numbers of mature NK cells, suggesting that IL-15 signals may be important for the differentiation of NK cells in vivo (75, 76). During the differentiation of NK cells, IL-15 signals may also induce the expression of NK cell Ly49 receptors. Ly49 receptors include both inhibitory and activating receptors that regulate the cytolytic activity of mature NK cells via recognition elements that include MHC class I antigens. Ly49 receptors may be acquired during NK cell differentiation via novel stochastic mechanisms of transcriptional activation (77). IL-15Rα−/− NK cells that either differentiate in vitro or in chimeric mice reconstituted with IL-15Rα−/− hematopoietic stem cells exhibit reduced expression of virtually all Ly49 alleles (76). This observation indicates a critical role for cytokine signaling in regulating NK cell receptor acquisition that appears independent from cytokine regulation of NK cell homeostasis. This function for IL-15 is an important area of future study given the critical role of Ly49

In the absence of immunogenic stimuli (e.g., viral infection), mature NK cells persist in the periphery with a half-life of approximately eight days. Recent studies indicate that cytokines are important for maintaining this life span of mature NK cells. Normal numbers of NK cells in young IL-2−/− and IL-2Rα−/− mice (prior to the development of spontaneous autoimmune disease) and in IL-7 and IL-7Rα−/− mice indicate that these cytokines do not play critical roles in supporting mature NK cells. By contrast, whereas the paucity of mature NK cells in IL-15−/− and IL-15Rα−/− mice may be partly due to defects in NK cell differentiation, adoptive transfer of mature NK cells into either IL-15−/− or IL-15Rα−/− mice results in the abrupt loss of these cells (79, 80). The half-life of transferred cells is shortened dramatically from 7–8 days in normal mice to 5–6 h in IL-15Rα−/− mice (80). Adoptively transferred NK cells do not proliferate in these experiments, so the loss of these cells from IL-15Rα−/− mice indicates that IL15Rα provides critical survival signals (80). In addition, Bcl-2 transgenic NK cells survive for as long as wild-type cells after adoptive transfer into IL-15−/− mice (79). Thus, IL-15 signals are tonically critical for the survival of NK cells. Under basal conditions, IL-15 is critical for maintaining short-term (i.e., hours) homeostasis of NK cells by supporting their survival. By contrast, IL-15 maintains longer term (i.e., weeks) homeostasis of memory CD8+ T cells by supporting their slow proliferation. One question that then arises is whether these IL-15 signals are delivered via similar mechanisms. Adoptively transferred NK cells survive in mixed radiation chimera generated from IL-15Rα−/− and IL-2/15Rβ−/− mice, establishing the ability of IL-2/15Rβ−/− cells to use IL-15Rα to

Ma

· · Koka

Burkett

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

trans-present IL-15 to NK cells (Figure 1). NK cells do not survive in chimera generated from IL-15Rα−/− and IL-15−/− mice, establishing that these proteins must be coordinately expressed in the same trans-presenting cells to function in vivo (Figure 2). Thus, the novel trans-presentation mechanism by which IL-15Rα presents IL-15 to memory CD8+ T cells applies to NK cells as well. The rapidity with which mature NK cells disappear from IL-15−/− and IL-15Rα−/− mice suggests that NK cells need to encounter IL-15Rα-/IL-15-expressing accessory cells at least every few hours to avoid programmed cell death. Thus, the survival of resting mature NK cells is tightly regulated by transpresented IL-15 signals in vivo. This is a remarkable requirement for cells that, unlike T cells, may not regularly encounter classical accessory cells in deﬁned niches under basal conditions. This ﬁnding also emphasizes the importance of understanding NK cell trafﬁcking in vivo. In summary, whereas IL-2 and IL-7 appear to play negligible roles in the physiological support of NK cell survival, trans-presented IL-15 is critical for this function.

MATURE NATURAL KILLER CELL ACTIVATION γc chain cytokines may also promote lymphocyte activation in addition to supporting survival or proliferation. For example, although a Bcl-2 transgene reconstitutes mature NK cell numbers in IL-2Rβ−/− mice, the NK cells lack cytolytic activity (81). Both IL-2 and IL-15 synergize with IL-12 to enhance NK cell function in vitro, and stimulation of NK cells with heterologous IL-12 and IL-15 improves in vivo inﬂammatory responses, tumor surveillance, and viral clearance (82–86). However, NK cell activation is grossly intact in IL-2−/− and IL-2Rα−/− mice, so there is no compelling evidence to date to indicate that IL-2 plays a nonredundant physiological involvement in NK cell activation. Assessing the role of IL-15 signaling in NK cell acti-

vation is complicated in vivo because of the critical role of IL-15 signaling in supporting acute NK cell survival (80). Recent studies have suggested that DCs have the ability to prime NK cells in a cell contact–dependent fashion (87–91). Although DCs have been reported to produce IL-2, this function has not been linked to NK cell activation (39). By contrast, IL-15Rα expression on DCs is critical for their ability to activate NK cells (70). Enhancement of IFN-γ production requires both DC-bound IL15Rα and IL-12 whereas cytolytic activity is IL-15Rα dependent and IL-12 independent (70). Moreover, blockade of DC-NK interactions with anti-IL-15Rα or anti-IL2Rβ antibodies, but not anti-IL-2Rα antibody prevents DC priming of NK cells. Therefore, it is probable that DC-mediated trans-presentation of IL-15 activates NK cells under physiological circumstances.

CYTOKINES AND LYMPHOID HOMEOSTASIS: FUTURE CONSIDERATIONS Homeostatic cytokines such as IL-7 and IL15 must be constitutively delivered in noninﬂamed mice to support lymphoid homeostasis under basal conditions. One concept that has emerged from studies of the IL-7 and IL-15 dependence of lymphocytes is that a limiting quantity of bioavailable cytokine may deﬁne the size of lymphoid populations under basal conditions. In the case of IL-15, two experiments support this concept. First, the homeostatic expansion of IL-2/15Rβ-expressing (and thus IL-15-responsive) CD1-restricted NKT cells depends on IL-15 and on the presence of other IL-15-responsive cell types such as NK cells and memory CD8+ T cells (92). Secondly, titrating the ratio of IL-15Rαcompetent to IL-15Rα-deﬁcient hematopoietic cells in mixed radiation chimera leads to proportionate changes in the numbers of peripheral NK cells (67). Thus, regulating the number of trans-presenting cells can regulate the number of IL-15-responsive lymphocytes. www.annualreviews.org • IL-2, IL-15, and IL-7

669

ARI

15 February 2006

16:33

As maintaining appropriate cell numbers in solid tissues may be easier for multicellular organisms than counting cells in suspension (i.e., lymphocytes in serum), this mechanism of regulating lymphoid homeostasis may be a teleologically attractive beneﬁt of a trans-presentation mechanism of providing cytokines to lymphocytes. If cytokines are limiting, then lymphocytes expressing higher levels of cytokine receptors could be favored. Hence, the signaling molecules and transcription factors regulating IL-2Rα, IL-2/15Rβ, and IL-7Rα expression in lymphocytes are an important facet of lymphoid homeostasis and cytokine biology. In this regard, that IL-2 induces the expression of IL-2Rα has been known for some time, but important cell biological details such as the initiation of this apparent autocrine loop remain unclear. Multiple signals, including IL-7 itself, regulate IL-7Rα expression. With regard to IL-15 signals, preliminary studies indicate that the transcription factors T-bet and eomesodermin may be important regulators of IL-2/15Rβ expression and hence IL-15 responsiveness (93, 93a). Physiological “pools” of IL-2, IL-15, and IL-7 cytokines may expand transiently during inﬂammatory conditions, and their return to basal levels may help facilitate the return of expanded lymphocyte populations to baseline levels. These pools may partially overlap because virtually all B and T lymphocytes rely on IL-7 (16), whereas NK cells, NKT cells, CD8αα IELs, and memory CD8+ T cells depend on IL-15 (13, 14, 92). The idea that homeostatic IL-7 and IL-15 cytokines are present in limiting quantities implies that different lymphoid populations will expand at the cost of others. For example, increased numbers of memory T cells in aging animals may contribute to decreased numbers of naive T cells that share the same pool of IL-7. Similarly, infections that generate new populations of memory CD8+ T cells may compromise the numbers of memory CD8+ T cells speciﬁc for other antigens because they all share a pool of IL-7 and IL-15. Partial compensation

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

670

Ma

· · Koka

Burkett

for the absence of one cytokine signal with the other has also been suggested by studies in which the absence of both IL-7 and IL-15 compromises memory phenotype or antigenspeciﬁc memory T cells to a far greater degree than the absence of other cytokine alone (55, 94). These experiments may suggest that IL-7 and IL-15 pools can be shared under certain circumstances or may reﬂect complementary functions (e.g., survival and proliferation) for these cytokines. The magnitude of these cytokine pools may also be more complex than static quantities of soluble serum cytokine. Recent evidence indicates that these cytokines are probably delivered by different cells from different microenvironments. In the case of IL7, radiation-resistant cells such as bone marrow or thymic stromal cells are thought to constitutively produce IL-7, but many questions remain. Are these stromal cells the cells that are physiologically important for the maintenance of peripheral lymphocytes? Do other cells in the periphery produce IL-7? Do mature lymphocytes obtain soluble IL-7 from the serum or lymph or more deﬁned niches? In the case of IL-15, the physiologically important cells that trans-present IL-15 to IELs appear to be radiation resistant (95; D. Boone & A. Ma, unpublished data). These are likely to be intestinal epithelial cells because they express IL-15Rα and are in constant contact with IELs. This interaction is thus likely to be central to the recent ﬁndings that IL-15 may be pathogenic in the small intestines of celiac sprue patients (96, 97). Cells that use IL-15Rα to support NK cells appear to be evenly divided between radiation-sensitive (largely hematopoietic) and radiation-resistant (largely stromal) populations (80). These cells might exist in the blood, spleen, or liver, sites where most NK cells reside. Although activated DCs may be important for directly activating NK cells during acute immune responses, no speciﬁc cell type has been identiﬁed to support NK cell survival. Finally, the cells that support memory CD8+ T cells are largely radiation

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

sensitive and RAG-1 independent (61, 67, 95). These cells could include DCs, macrophages, or other myeloid cells. DCs clearly support T cell activation, and deliver survival signals during this process, but a role for DCs in supporting memory T cell homeostasis during basal conditions has not yet been shown. A recent clue regarding where memory CD8+ T cells may receive proliferation stimuli is that a high proportion of memory CD8+ T cells proliferate in the bone marrow of mice (98). As the bone marrow expresses IL-7, IL-15, and IL-15Rα, which support NK and B lymphocyte development, it is feasible that CD8+ T cells recirculate to the marrow to obtain these critical cytokine signals. Identifying the locations and cell populations required for these distinct homeostatic stimuli is an important challenge. The degree to which physiological cytokine pools are dispersed from their cellular sources also inﬂuences the capacity of these cytokines to support lymphocytes in vivo. In the case of IL-2, autocrine signaling may restrict the expansion of antigen-speciﬁc CD8+ T cells, whereas paracrine or systemic IL-2 signals positively support these cells (42, 46). How IL-2 signals promote two physiologically opposing functions is currently unclear, but the mechanism could involve different signals that are delivered coordinately with IL-2, different strengths of IL-2 autocrine versus paracrine signals, or other aspects of the context in which IL-2 signaling is triggered. These studies also reafﬁrm that IL-2 signals are probably not delivered by IL-2Rαmediated trans-presentation. Little is known about the extent to which IL-7 disperses, except that IL-7 is probably not delivered via autocrine signaling because lymphocytes are not known to produce IL-7. IL-7 is also probably not delivered via trans-presentation like IL-15 because T lymphocytes respond normally after adoptive transfer into IL-7Rα−/− mice (35). The genetic and physiological evidence for trans-presentation of IL-15 in memory T cell homeostasis also leads to several cell biologi-

cal predictions. First, the afﬁnity of IL-15Rα as a single receptor chain for IL-15 is quite high (Kd ∼ = 10−11 M), whereas the afﬁnity of IL-2/15Rβ and γc dimers for IL-15 is signiﬁcantly lower (Kd ∼ = 10−9 M), so IL-15 likely remains bound to IL-15Rα on the surface of accessory cells until accessory cells are in close contact with memory CD8+ T cells and NK cells, indeed perhaps until IL-15 is also bound to IL-2/15Rβ and γc chains on these IL-15-responsive cells. Memory CD8+ T cells and NK cells would thus need to make virtual cell-cell contact with IL-15Rα-/ IL-15-bearing accessory cells regularly to receive proliferation signals. This is the ﬁrst evidence suggesting that either of these cell types must make cell-cell contact for their homeostasis (67). This cellular mechanism implies that IL-15 signals may be delivered in the context of other cell surface ligands (e.g., CD40, B7) to stimulate distinct proliferation or survival signals. In addition, the accessory cell type that trans-presents IL-15/IL-15Rα may dictate the combination of signals that IL-15-responsive lymphocytes receive. These variables may help explain why NK cells receive predominantly survival signals from IL15, whereas memory CD8+ T cells receive predominantly proliferation signals. Homeostatic IL-15 proliferation signals could also be coordinated with other activation, survival, or differentiation signals that would allow ﬁnetuned regulation of the behavior of these cells. The intimate interaction between IL-15presenting and IL-15-responsive cells also raises the question of whether IL-15Rα, which possesses a 26 amino acid cytoplasmic tail, could induce signals on the presenting accessory cells. Similar to CD40-CD40L signaling between recently activated T cells and DCs, bidirectional signaling between memory CD8+ T cells or NK cells and accessory cells may allow integration of transpresented cytokine signals with other cellular cross-talk. Preliminary studies have suggested that TRAF2 or syk may be recruited to the cytoplasmic domain of IL-15Rα, but more biochemical and genetic studies are needed www.annualreviews.org • IL-2, IL-15, and IL-7

671

ARI

15 February 2006

16:33

to determine whether physiological signals emanate from this domain (99, 100). Additional studies suggest that IL-15/IL-15Rα complexes may be cleaved from cells by TNFα converting enzyme (TACE) (101, 102). Such soluble complexes resemble the IL-6/ IL-6R complexes that stimulate agonistic IL6 signals on cells bearing the gp130 receptor chain. However, preliminary evidence indicates that soluble IL-15/IL-15Rα complexes are inhibitory rather than agonistic, so IL-6 and IL-15 signaling may differ considerably (102). Understanding how these cytokines support lymphoid homeostasis also requires an understanding of how their production is regulated within cells. Following these proteins under physiological conditions has been challenging because they are produced in low quantities and typically possess short halflives. The production of IL-2 and IL-7 protein has been presumed to mirror the production of their mRNAs. Thus, if future studies can precisely localize and quantitate the production of IL-2 and IL-7 mRNAs, e.g., with BAC transgenic or gene-targeting approaches, more precise understanding of the physiological regulation of their homeostatic functions may be appreciated (103). For IL15, the inability of IL-15Rα−/− cells to complement IL-15−/− cells in supporting either NK or memory CD8+ T cells in vivo indicated that IL-15 protein is not secreted from IL-15Rα−/− cells to then bind to IL-15Rα on the surface of IL-15−/− cells (67). Hence, IL-15 may not by secreted as a freely soluble protein under basal conditions in vivo. If this is true, then the physiological relevance of any studies conducted with recombinant soluble IL-15 protein becomes less clear. Whether IL-15Rα is required for IL-15 translation, secretion, or other processes should be interesting from the standpoint of lymphoid homeostasis as well as from more a more general standpoint of ligand-receptor cellular biology. Given the high afﬁnity with which IL-15Rα alone binds to IL-15, it is possible that IL15Rα binds to IL-15 within the cell and di-

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

672

Ma

· · Koka

Burkett

rectly facilitates the secretion of IL-15. This potential preassociation of IL-15 with IL15Rα prior to emergence from the cell would be distinct from the endosomal recycling of these proteins observed after heterologous addition of soluble IL-15. Indeed, this mechanism would be entirely unique to cytokine biology. Finally, additional insights into the unique cell biology of these cytokines may be gained from a better understanding of the biochemistry of these proteins. Just as the afﬁnities of TCRs for MHC-peptide complexes are important in deﬁning the biology of the cells that express them, the different afﬁnities of IL-2, IL-15, and IL-7 for their respective receptors may play important roles in their biology. In this regard, the recent solution of IL-2 bound to IL-2Rα demonstrates how the two ligandbinding sushi domains of IL-2Rα cooperate to bind IL-2 (104). This structure should also facilitate the solution of the IL-15/IL-15Rα structure because IL-2 and IL-15 are both four-α-helix-bundle family cytokines and because IL-2Rα and IL-15Rα are highly homologous genes. Perhaps the IL-15/IL-15Rα structure will help explain why, despite its homology to IL-2Rα and its possession of only one sushi domain, IL-15Rα binds IL-15 with much higher afﬁnity than IL-2Rα binds IL-2. In conclusion, IL-2, IL-15, and IL-7 signals support multiple functions that maintain lymphoid homeostasis. The production of these cytokines, the expression of their receptor chains, the context in which the cytokines are delivered, and the cell biological responses are all highly regulated processes that contribute to appropriate immune responses. Although recent studies have identiﬁed important facets of cytokine biology, unanticipated results have forced a reevaluation of how cytokines function in vivo. These studies also highlight the need for future biochemical, cell biological, and genetic studies to complement physiological experiments in gaining a complete understanding of how cytokines function in the immune system.

ANRV270-IY24-21

ARI

15 February 2006

16:33

SUMMARY POINTS 1. IL-2 supports the differentiation of IL-2Rα-expressing Tregs, thereby maintaining T cell tolerance. IL-2 may also regulate the expansion of recently activated T lymphocytes. 2. IL-7 supports B and T lymphocyte development and homeostasis of both naive and memory T cells by binding to IL-7Rα on these cells.

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

3. IL-15 supports the homeostasis of IL-2/15Rβ-expressing NK cells and memory CD8+ T cells via trans-presentation of IL-15 by IL-15Rα-expressing cells. IL-15 and IL-15Rα must be coexpressed in accessory cells to trans-present IL-15.

FUTURE ISSUES TO BE RESOLVED 1. The physiologically relevant cell types that elaborate IL-2, IL-15, and IL-7 and their locations must be determined. 2. Investigators require an understanding of the cell biological and biochemical mechanisms by which autocrine or paracrine cis signaling of IL-2 and IL-7 and the transpresentation of IL-15 by IL-15Rα occur. 3. The regulation of both cytokine receptor chain expression and receptor initiated signals on responsive lymphocytes needs to be understood.

ACKNOWLEDGMENTS We apologize for not comprehensively referencing the relevant literature owing to space restrictions. The work from the authors’ laboratory was supported by NIH RO1 AI45860 and AI059827. We thank Barbara Malynn for critically reading the manuscript.

LITERATURE CITED 1. Smith KA, Gilbride KJ, Favata MF. 1980. Lymphocyte activating factor promotes T-cell growth factor production by cloned murine lymphoma cells. Nature 287:853–55 2. Namen AE, Lupton S, Hjerrild K, Wignall J, Mochizuki DY, et al. 1988. Stimulation of B-cell progenitors by cloned murine interleukin-7. Nature 333:571–73 3. Bamford RN, Grant AJ, Burton JD, Peters C, Kurys G, et al. 1994. The interleukin (IL) 2 receptor β chain is shared by IL-2 and a cytokine, provisionally designated IL-T, that stimulates T-cell proliferation and the induction of lymphokine-activated killer cells. Proc. Natl. Acad. Sci. USA 91:4940–44 4. Giri JG, Ahdieh M, Eisenman J, Shanebeck K, Grabstein K, et al. 1994. Utilization of the β and γ chains of the IL-2 receptor by the novel cytokine IL-15. EMBO J. 13:2822–30 5. Khaled AR, Durum SK. 2002. Lymphocide: cytokines and the control of lymphoid homeostasis. Nat. Rev. Immunol. 2:817–30 6. Schluns KS, Lefrancois L. 2003. Cytokine control of memory T-cell development and survival. Nat. Rev. Immunol. 3:269–79 www.annualreviews.org • IL-2, IL-15, and IL-7

673

ARI

15 February 2006

16:33

7. Kovanen PE, Leonard WJ. 2004. Cytokines and immunodeﬁciency diseases: critical roles of the γc -dependent cytokines interleukins 2, 4, 7, 9, 15, and 21, and their signaling pathways. Immunol. Rev. 202:67–83 8. Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, Horak I. 1993. Ulcerative colitislike disease in mice with a disrupted interleukin-2 gene. Cell 75:253–61 9. Willerford DM, Chen J, Ferry JA, Davidson L, Ma A, Alt FW. 1995. Interleukin-2 receptor α chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3:521–30 10. Suzuki H, Kundig TM, Furlonger C, Wakeham A, Timms E, et al. 1995. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor β. Science 268:1472–76 11. Ma A, Datta M, Margosian E, Chen J, Horak I. 1995. T cells, but not B cells, are required for bowel inﬂammation in interleukin 2-deﬁcient mice. J. Exp. Med. 182:1567–72 12. Suzuki H, Duncan GS, Takimoto H, Mak TW. 1997. Abnormal development of intestinal intraepithelial lymphocytes and peripheral natural killer cells in mice lacking the IL-2 receptor β chain. J. Exp. Med. 185:499–505 13. Lodolce JP, Boone DL, Chai S, Swain RE, Dassopoulos T, et al. 1998. IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation. Immunity 9:669–76 14. Kennedy MK, Glaccum M, Brown SN, Butz EA, Viney JL, et al. 2000. Reversible defects in natural killer and memory CD8 T cell lineages in interleukin 15-deﬁcient mice. J. Exp. Med. 191:771–80 15. Tagaya Y, Burton JD, Miyamoto Y, Waldmann TA. 1996. Identiﬁcation of a novel receptor/signal transduction pathway for IL-15/T in mast cells. EMBO J. 15:4928–39 16. von Freeden-Jeffry U, Vieira P, Lucian LA, McNeil T, Burdach SE, Murray R. 1995. Lymphopenia in interleukin (IL)-7 gene-deleted mice identiﬁes IL-7 as a nonredundant cytokine. J. Exp. Med. 181:1519–26 17. Peschon JJ, Morrissey PJ, Grabstein KH, Ramsdell FJ, Maraskovsky E, et al. 1994. Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deﬁcient mice. J. Exp. Med. 180:1955–60 18. Pandey A, Ozaki K, Baumann H, Levin SD, Puel A, et al. 2000. Cloning of a receptor subunit required for signaling by thymic stromal lymphopoietin. Nat. Immunol. 1:59–64 19. Hanke T, Mitnacht R, Boyd R, Hunig T. 1994. Induction of interleukin 2 receptor β chain expression by self-recognition in the thymus. J. Exp. Med. 180:1629–36 20. Schorle H, Holtschke T, Hunig T, Schimpl A, Horak I. 1991. Development and function of T cells in mice rendered interleukin-2 deﬁcient by gene targeting. Nature 352:621–24 21. Sakaguchi S. 2000. Regulatory T cells: key controllers of immunologic self-tolerance. Cell 101:455–58 22. Malek TR, Yu A, Vincek V, Scibelli P, Kong L. 2002. CD4 regulatory T cells prevent lethal autoimmunity in IL-2Rβ-deﬁcient mice. Implications for the nonredundant function of IL-2. Immunity 17:167–78 23. Ilangumaran S, Ramanathan S, Ning T, La Rose J, Reinhart B, et al. 2003. Suppressor of cytokine signaling 1 attenuates IL-15 receptor signaling in CD8+ thymocytes. Blood 102:4115–22 24. Munitic I, Williams JA, Yang Y, Dong B, Lucas PJ, et al. 2004. Dynamic regulation of IL-7 receptor expression is required for normal thymopoiesis. Blood 104:4165–72 25. Hofmeister R, Khaled AR, Benbernou N, Rajnavolgyi E, Muegge K, Durum SK. 1999. Interleukin-7: physiological roles and mechanisms of action. Cytokine Growth Factor Rev. 10:41–60

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

674

Ma

· · Koka

Burkett

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

26. Maki K, Sunaga S, Komagata Y, Kodaira Y, Mabuchi A, et al. 1996. Interleukin 7 receptordeﬁcient mice lack γδ T cells. Proc. Natl. Acad. Sci. USA 93:7172–77 27. Nakajima H, Noguchi M, Leonard WJ. 2000. Role of the common cytokine receptor γ chain (γc) in thymocyte selection. Immunol. Today 21:88–94 28. Maraskovsky E, O’Reilly LA, Teepe M, Corcoran LM, Peschon JJ, Strasser A. 1997. Bcl-2 can rescue T lymphocyte development in interleukin-7 receptor-deﬁcient mice but not in mutant rag-1−/− mice. Cell 89:1011–19 29. Pallard C, Stegmann AP, van Kleffens T, Smart F, Venkitaraman A, Spits H. 1999. Distinct roles of the phosphatidylinositol 3-kinase and STAT5 pathways in IL-7-mediated development of human thymocyte precursors. Immunity 10:525–35 30. Kang J, Der SD. 2004. Cytokine functions in the formative stages of a lymphocyte’s life. Curr. Opin. Immunol. 16:180–90 31. Opferman JT, Letai A, Beard C, Sorcinelli MD, Ong CC, Korsmeyer SJ. 2003. Development and maintenance of B and T lymphocytes requires antiapoptotic MCL-1. Nature 426:671–76 32. Ma A, Pena JC, Chang B, Margosian E, Davidson L, et al. 1995. Bclx regulates the survival of double-positive thymocytes. Proc. Natl. Acad. Sci. USA 92:4763–67 33. Laouar Y, Crispe IN, Flavell RA. 2004. Overexpression of IL-7Rα provides a competitive advantage during early T-cell development. Blood 103:1985–94 34. Marrack P, Kappler J. 2004. Control of T cell viability. Annu. Rev. Immunol. 22:765–87 35. Schluns KS, Kieper WC, Jameson SC, Lefrancois L. 2000. Interleukin-7 mediates the homeostasis of naive and memory CD8 T cells in vivo. Nat. Immunol. 1:426–32 36. Rathmell JC, Farkash EA, Gao W, Thompson CB. 2001. IL-7 enhances the survival and maintains the size of naive T cells. J. Immunol. 167:6869–76 37. Vivien L, Benoist C, Mathis D. 2001. T lymphocytes need IL-7 but not IL-4 or IL-6 to survive in vivo. Int. Immunol. 13:763–68 38. Vella A, Teague TK, Ihle J, Kappler J, Marrack P. 1997. Interleukin 4 (IL-4) or IL-7 prevents the death of resting T cells: Stat6 is probably not required for the effect of IL-4. J. Exp. Med. 186:325–30 39. Granucci F, Vizzardelli C, Pavelka N, Feau S, Persico M, et al. 2001. Inducible IL-2 production by dendritic cells revealed by global gene expression analysis. Nat. Immunol. 2:882–88 40. Kundig TM, Schorle H, Bachmann MF, Hengartner H, Zinkernagel RM, Horak I. 1993. Immune responses in interleukin-2-deﬁcient mice. Science 262:1059–61 41. Vella AT, Dow S, Potter TA, Kappler J, Marrack P. 1998. Cytokine-induced survival of activated T cells in vitro and in vivo. Proc. Natl. Acad. Sci. USA 95:3810–15 42. D’Souza WN, Lefrancois L. 2003. IL-2 is not required for the initiation of CD8 T cell cycling but sustains expansion. J. Immunol. 171:5727–35 43. Lenardo MJ. 1991. Interleukin-2 programs mouse αβ T lymphocytes for apoptosis. Nature 353:858–61 44. Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, Lenardo MJ. 1995. Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377:348–51 45. Refaeli Y, Van Parijs L, London CA, Tschopp J, Abbas AK. 1998. Biochemical mechanisms of IL-2-regulated Fas-mediated T cell apoptosis. Immunity 8:615–23 46. D’Souza WN, Schluns KS, Masopust D, Lefrancois L. 2002. Essential role for IL-2 in the regulation of antiviral extralymphoid CD8 T cell responses. J. Immunol. 168:5566–72 47. Zhang X, Sun S, Hwang I, Tough DF, Sprent J. 1998. Potent and selective stimulation of memory-phenotype CD8+ T cells in vivo by IL-15. Immunity 8:591–99 www.annualreviews.org • IL-2, IL-15, and IL-7

675

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

55. and 94. These studies demonstrate the dual requirement of memory T cells for both IL-7 and IL-15.

63. This study first suggested that the high-affinity IL-15 receptor, IL-15Rα, can present IL-15 to cells bearing IL-2/15Rβ and γc receptor chains.

65. This study first demonstrated that IL-15Rα functions in a noncellautonomous manner in vivo.

676

16:33

48. Schluns KS, Williams K, Ma A, Zheng XX, Lefrancois L. 2002. Cutting edge: requirement for IL-15 in the generation of primary and memory antigen-speciﬁc CD8 T cells. J. Immunol. 168:4827–31 49. Becker TC, Wherry EJ, Boone D, Murali-Krishna K, Antia R, et al. 2002. Interleukin 15 is required for proliferative renewal of virus-speciﬁc memory CD8 T cells. J. Exp. Med. 195:1541–48 50. Kaech SM, Tan JT, Wherry EJ, Konieczny BT, Surh CD, Ahmed R. 2003. Selective expression of the interleukin 7 receptor identiﬁes effector CD8 T cells that give rise to long-lived memory cells. Nat. Immunol. 4:1191–98 51. Park JH, Yu Q, Erman B, Appelbaum JS, Montoya-Durango D, et al. 2004. Suppression of IL7Rα transcription by IL-7 and other prosurvival cytokines: a novel mechanism for maximizing IL-7-dependent T cell survival. Immunity 21:289–302 52. El Kassar N, Lucas PJ, Klug DB, Zamisch M, Merchant M, et al. 2004. A dose effect of IL-7 on thymocyte development. Blood 104:1419–27 53. Li J, Huston G, Swain SL. 2003. IL-7 promotes the transition of CD4 effectors to persistent memory cells. J. Exp. Med. 198:1807–15 54. Ku CC, Murakami M, Sakamoto A, Kappler J, Marrack P. 2000. Control of homeostasis of CD8+ memory T cells by opposing cytokines. Science 288:675–78 55. Goldrath AW, Sivakumar PV, Glaccum M, Kennedy MK, Bevan MJ, et al. 2002. Cytokine requirements for acute and basal homeostatic proliferation of naive and memory CD8+ T cells. J. Exp. Med. 195:1515–22 56. Kieper WC, Tan JT, Bondi-Boyd B, Gapin L, Sprent J, et al. 2002. Overexpression of interleukin (IL)-7 leads to IL-15-independent generation of memory phenotype CD8+ T cells. J. Exp. Med. 195:1533–39 57. Lantz O, Grandjean I, Matzinger P, Di Santo JP. 2000. γ chain required for naive CD4+ T cell survival but not for antigen proliferation. Nat. Immunol. 1:54–58 58. Kondrack RM, Harbertson J, Tan JT, McBreen ME, Surh CD, Bradley LM. 2003. Interleukin 7 regulates the survival and generation of memory CD4 cells. J. Exp. Med. 198:1797–806 59. Lenz DC, Kurz SK, Lemmens E, Schoenberger SP, Sprent J, et al. 2004. IL-7 regulates basal homeostatic proliferation of antiviral CD4+ T cell memory. Proc. Natl. Acad. Sci. USA 101:9357–62 60. Seddon B, Tomlinson P, Zamoyska R. 2003. Interleukin 7 and T cell receptor signals regulate homeostasis of CD4 memory cells. Nat. Immunol. 4:680–86 61. Burkett PR, Koka R, Chien M, Chai S, Chan F, et al. 2003. IL-15Rα expression on CD8+ T cells is dispensable for T cell memory. Proc. Natl. Acad. Sci. USA 100:4724–29 62. Weninger W, Crowley MA, Manjunath N, von Andrian UH. 2001. Migratory properties of naive, effector, and memory CD8+ T cells. J. Exp. Med. 194:953–66 63. Dubois S, Mariner J, Waldmann TA, Tagaya Y. 2002. IL-15Rα recycles and presents IL-15 in trans to neighboring cells. Immunity 17:537–47 64. Schluns KS, Klonowski KD, Lefrancois L. 2004. Transregulation of memory CD8 T-cell proliferation by IL-15Rα+ bone marrow-derived cells. Blood 103:988–94 65. Lodolce JP, Burkett PR, Boone DL, Chien M, Ma A. 2001. T cell-independent interleukin 15Rα signals are required for bystander proliferation. J. Exp. Med. 194:1187–94 66. Eicher DM, Waldmann TA. 1998. IL-2R α on one cell can present IL-2 to IL-2Rβ/γc on another cell to augment IL-2 signaling. J. Immunol. 161:5430–37 Ma

· · Koka

Burkett

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

67. Burkett PR, Koka R, Chien M, Chai S, Boone DL, Ma A. 2004. Coordinate expression and trans presentation of interleukin (IL)-15Rα and IL-15 supports natural killer cell and memory CD8+ T cell homeostasis. J. Exp. Med. 200:825–34 68. Dubois SP, Waldmann TA, Muller JR. 2005. Survival adjustment of mature dendritic cells by IL-15. Proc. Natl. Acad. Sci. USA 102:8662–67 69. Sandau MM, Schluns KS, Lefrancois L, Jameson SC. 2004. Cutting edge: transpresentation of IL-15 by bone marrow-derived cells necessitates expression of IL-15 and IL-15Rα by the same cells. J. Immunol. 173:6537–41 70. Koka R, Burkett P, Chien M, Chai S, Boone DL, Ma A. 2004. Cutting edge: murine dendritic cells require IL-15Rα to prime NK cells. J. Immunol. 173:3594–98 71. Loza MJ, Peters SP, Zangrilli JG, Perussia B. 2002. Distinction between IL-13+ and IFN-γ+ natural killer cells and regulation of their pool size by IL-4. Eur. J. Immunol. 32:413–23 72. Miller JS, Alley KA, McGlave P. 1994. Differentiation of natural killer (NK) cells from human primitive marrow progenitors in a stroma-based long-term culture system: identiﬁcation of a CD34+ 7+ NK progenitor. Blood 83:2594–601 73. Mrozek E, Anderson P, Caligiuri MA. 1996. Role of interleukin-15 in the development of human CD56+ natural killer cells from CD34+ hematopoietic progenitor cells. Blood 87:2632–40 74. Williams NS, Moore TA, Schatzle JD, Puzanov IJ, Sivakumar PV, et al. 1997. Generation of lytic natural killer 1.1+ , Ly-49− cells from multipotential murine bone marrow progenitors in a stroma-free culture: deﬁnition of cytokine requirements and developmental intermediates. J. Exp. Med. 186:1609–14 75. Vosshenrich CA, Ranson T, Samson SI, Corcuff E, Colucci F, et al. 2005. Roles for common cytokine receptor γ-chain-dependent cytokines in the generation, differentiation, and maturation of NK cell precursors and peripheral NK cells in vivo. J. Immunol. 174:1213–21 76. Kawamura T, Koka R, Ma A, Kumar V. 2003. Differential roles for IL-15Rα-chain in NK cell development and Ly-49 induction. J. Immunol. 171:5085–90 77. Saleh A, Davies GE, Pascal V, Wright PW, Hodge DL, et al. 2004. Identiﬁcation of probabilistic transcriptional switches in the Ly49 gene cluster: a eukaryotic mechanism for selective gene activation. Immunity 21:55–66 78. Arase H, Mocarski ES, Campbell AE, Hill AB, Lanier LL. 2002. Direct recognition of cytomegalovirus by activating and inhibitory NK cell receptors. Science 296:1323–26 79. Cooper MA, Bush JE, Fehniger TA, VanDeusen JB, Waite RE, et al. 2002. In vivo evidence for a dependence on interleukin 15 for survival of natural killer cells. Blood 100:3633–38 80. Koka R, Burkett PR, Chien M, Chai S, Chan F, et al. 2003. Interleukin (IL)-15Rαdeﬁcient natural killer cells survive in normal but not IL-15Rα-deﬁcient mice. J. Exp. Med. 197:977–84 81. Minagawa M, Watanabe H, Miyaji C, Tomiyama K, Shimura H, et al. 2002. Enforced expression of Bcl-2 restores the number of NK cells, but does not rescue the impaired development of NKT cells or intraepithelial lymphocytes, in IL-2/IL-15 receptor βchain-deﬁcient mice. J. Immunol. 169:4153–60 82. Nguyen KB, Salazar-Mather TP, Dalod MY, Van Deusen JB, Wei XQ, et al. 2002. Coordinated and distinct roles for IFN-αβ, IL-12, and IL-15 regulation of NK cell responses to viral infection. J. Immunol. 169:4279–87 83. Fawaz LM, Sharif-Askari E, Menezes J. 1999. Up-regulation of NK cytotoxic activity via IL-15 induction by different viruses: a comparative study. J. Immunol. 163:4473–80 www.annualreviews.org • IL-2, IL-15, and IL-7

67. This study shows that trans-presentation of IL-15 by IL-15Rα to IL-2/15Rβ and γc receptors, rather than other indirect mechanisms, is the physiological mechanism by which IL-15 signals are delivered in vivo. It also shows that IL-15 and IL-15Rα must be coexpressed by the same cells to effect trans-presentation in vivo.

677

ARI

15 February 2006

16:33

84. Gri G, Chiodoni C, Gallo E, Stoppacciaro A, Liew FY, Colombo MP. 2002. Antitumor effect of interleukin (IL)-12 in the absence of endogenous IFN-γ: a role for intrinsic tumor immunogenicity and IL-15. Cancer Res. 62:4390–97 85. Comes A, Di Carlo E, Musiani P, Rosso O, Meazza R, et al. 2002. IFN-γ-independent synergistic effects of IL-12 and IL-15 induce anti-tumor immune responses in syngeneic mice. Eur. J. Immunol. 32:1914–23 86. Biber JL, Jabbour S, Parihar R, Dierksheide J, Hu Y, et al. 2002. Administration of two macrophage-derived interferon-γ-inducing factors (IL-12 and IL-15) induces a lethal systemic inﬂammatory response in mice that is dependent on natural killer cells but does not require interferon-γ. Cell. Immunol. 216:31–42 87. Gerosa F, Baldani-Guerra B, Nisii C, Marchesini V, Carra G, Trinchieri G. 2002. Reciprocal activating interaction between natural killer cells and dendritic cells. J. Exp. Med. 195:327–33 88. Fernandez NC, Lozier A, Flament C, Ricciardi-Castagnoli P, Bellet D, et al. 1999. Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo. Nat. Med. 5:405–11 89. Ferlazzo G, Tsang ML, Moretta L, Melioli G, Steinman RM, Munz C. 2002. Human dendritic cells activate resting natural killer (NK) cells and are recognized via the NKp30 receptor by activated NK cells. J. Exp. Med. 195:343–51 90. Piccioli D, Sbrana S, Melandri E, Valiante NM. 2002. Contact-dependent stimulation and inhibition of dendritic cells by natural killer cells. J. Exp. Med. 195:335–41 91. Andrews DM, Scalzo AA, Yokoyama WM, Smyth MJ, Degli-Esposti MA. 2003. Functional interactions between dendritic cells and NK cells during viral infection. Nat. Immunol. 4:175–81 92. Matsuda JL, Gapin L, Sidobre S, Kieper WC, Tan JT, et al. 2002. Homeostasis of Vα14i NKT cells. Nat. Immunol. 3:966–74 93. Townsend MJ, Weinmann AS, Matsuda JL, Salomon R, Farnham PJ, et al. 2004. Tbet regulates the terminal maturation and homeostasis of NK and Vα14i NKT cells. Immunity 20:477–94 93a. Intlekofer AM, Takemoto N, Wherry EJ, Longworth SA, Northrup JT, et al. 2005. Effector and memory CD8+ T cell fate coupled by T-bet and eomesodermin. Nat. Immunol. 6:1236–44 94. Tan JT, Ernst B, Kieper WC, LeRoy E, Sprent J, Surh CD. 2002. Interleukin (IL)-15 and IL-7 jointly regulate homeostatic proliferation of memory phenotype CD8+ cells but are not required for memory phenotype CD4+ cells. J. Exp. Med. 195:1523–32 95. Schluns KS, Nowak EC, Cabrera-Hernandez A, Puddington L, Lefrancois L, Aguila HL. 2004. Distinct cell types control lymphoid subset development by means of IL-15 and IL-15 receptor α expression. Proc. Natl. Acad. Sci. USA 101:5616–21 96. Meresse B, Chen Z, Ciszewski C, Tretiakova M, Bhagat G, et al. 2004. Coordinated induction by IL15 of a TCR-independent NKG2D signaling pathway converts CTL into lymphokine-activated killer cells in celiac disease. Immunity 21:357–66 97. Hue S, Mention JJ, Monteiro RC, Zhang S, Cellier C, et al. 2004. A direct role for NKG2D/MICA interaction in villous atrophy during celiac disease. Immunity 21:367– 77 98. Becker TC, Coley SM, Wherry EJ, Ahmed R. 2005. Bone marrow is a preferred site for homeostatic proliferation of memory CD8 T cells. J. Immunol. 174:1269–73 99. Bulanova E, Budagian V, Pohl T, Krause H, Durkop H, et al. 2001. The IL-15Rα chain signals through association with Syk in human B cells. J. Immunol. 167:6292–302

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

678

Ma

· · Koka

Burkett

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-21

ARI

15 February 2006

16:33

100. Bulfone-Pau SS, Bulanova E, Pohl T, Budagian V, Durkop H, et al. 1999. Death deﬂected: IL-15 inhibits TNF-α-mediated apoptosis in ﬁbroblasts by TRAF2 recruitment to the IL-15Rα chain. FASEB J. 13:1575–85 101. Budagian V, Bulanova E, Orinska Z, Ludwig A, Rose-John S, et al. 2004. Natural soluble interleukin-15Rα is generated by cleavage that involves the tumor necrosis factor-αconverting enzyme (TACE/ADAM17). J. Biol. Chem. 279:40368–75 102. Mortier E, Bernard J, Plet A, Jacques Y. 2004. Natural, proteolytic release of a soluble form of human IL-15 receptor α-chain that behaves as a speciﬁc, high afﬁnity IL-15 antagonist. J. Immunol. 173:1681–88 103. Mohrs M, Shinkai K, Mohrs K, Locksley RM. 2001. Analysis of type 2 immunity in vivo with a bicistronic IL-4 reporter. Immunity 15:303–11 104. Rickert M, Wang X, Boulanger MJ, Goriatcheva N, Garcia KC. 2005. The structure of interleukin-2 complexed with its α receptor. Science 308:1477–80

www.annualreviews.org • IL-2, IL-15, and IL-7

104. This solution of the structure of IL-2 complexed with IL-2Rα provides important biochemical clues about how IL-2 and IL-15 may interact with their receptors.

679

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:657-679. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefranc¸ois and Lynn Puddington Center for Integrative Immunology and Vaccine Research, Department of Immunology, University of Connecticut Health Center, Farmington, Connecticut 06030-1319; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:681–704 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090650 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0681$20.00

Key Words mucosal, T lymphocytes, lung, intestine

Abstract Mucosal immunity in the lung and intestine is controlled by complex multifaceted systems. While mucosal T cells are essential for protection against invading pathogens owing to their proximity to the outside world, powerful systems must also be in place to harness ongoing inﬂammatory processes. In each site, distinct anatomical structures play key roles in mounting and executing both protective and deleterious mucosal T cell responses. Although analogies can be drawn regarding the immune systems of these two organs, there are substantial dissimilarities necessitated by unique physiologic constraints. Here, we discuss how T cell activation and effector function are generated in the mucosae.

681

ANRV270-IY24-22

ARI

15 February 2006

2:12

INTRODUCTION

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

In terms of total surface area, the major mucosal organs of the body are the lung and the intestine. In essence, a single layer of epithelial cells within each of these tissues separates the inner corpus from the outside world. Although both organs share barrier function as direct portals into the body and are of common endodermal origin, their respiratory or digestive functions engender distinct anatomical and immunological compartments. The anatomical constraints in turn regulate the nature of lymphocyte responses within each organ with regard to lymphocyte subset development, migration, and effector function. The original concept of “the common mucosal immune system” suggests an overall integration of immune responses in anatomically distinct mucosal tissues, particularly with respect to IgA responses, the hallmark of mucosal immunity (1, 2). Although recent data point out major differences in T cell functional capacity within these organ systems, many commonalities also provide potential links between mucosal sites.

ANATOMICAL COMPARTMENTS OF THE MUCOSAL IMMUNE SYSTEM Before discussing the speciﬁcs of mucosal T cell responses, it is necessary to deﬁne the relevant locations of T cells within the mucosal tissues and the associated “plumbing,” which regulate movement of antigens, antigenpresenting cells (APCs), lymphocytes, and soluble mediators involved in control of the adaptive immune response. There are multiple lymphocyte locales associated with the lung and the intestine, some of which promote highly specialized functions while others may provide common functions in each mucosal site.

Immune Inductive Sites of Mucosae Organized secondary lymphoid tissues are associated with each organ system and are 682

Lefran¸cois

·

Puddington

thought to be the primary sites of T cell response initiation. For antigens derived from the gut lumen, T cell encounter with antigen occurs in the gut-draining mesenteric lymph nodes (MLN) and the Peyer’s patches (PP) (Figure 1). Thus, afferent lymph from the intestine carrying antigen or antigen-bearing APCs, such as dendritic cells (DCs), transits to the MLN. In contrast, there is no afferent lymphatic drainage to the PP so antigens are acquired directly from the intestinal lumen, in part through specialized epithelial cells, termed M cells (M for microvilli or microfold, due to their distinct surface structure), that overlie approximately 10% of the PP (3–5). In the intestine a population of villus M cells has been identiﬁed that may also participate in antigen uptake and pathogen entry (6). Similarly, a population of M cells has recently been identiﬁed in the lung (7). Efferent lymphatic drainage from the PP can also potentially carry antigen and cells to the MLN (8, 9). Note that immune responses to antigens or infectious agents introduced systemically may be initiated in these sites because blood may carry antigens to mucosal inductive sites (10). Numerous small isolated lymphoid follicles scattered along the wall of the intestine may also be involved in induction of immune responses (11). In the lung, the mediastinal lymph nodes (MedLN), which drain the lower respiratory tract, are major sites of T cell priming (12–16). (The MedLN are also variously termed the tracheobronchial, parathymic, or hilar LN.) Other LN (e.g., cervical) serve the upper respiratory tract (17). Neolymphoid tissue can also be induced by infection and inﬂammation (18). An example in the lung is the bronchus-associated lymphoid tissue induced by respiratory virus infection (iBALT) (19, 20). Such structures are likely sites of immune response initiation or maintenance. In fact, in mice lacking lymph nodes and spleen, a CD8 T cell response is still initiated but with delayed kinetics following inﬂuenza virus infection. The response is protective and appears to be initiated in iBALT (21). Thus, under certain

ANRV270-IY24-22

ARI

15 February 2006

2:12

Infectious agents Allergens Foods

Mediastinal LN, BALT from lung; PP, mesenteric LN from gut Naive T cell

+

Imprinting of + migratory and functional preferences

Resting DC

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Afferent lymphatics “2nd hit” upon entry tissue-specific functions induced?

DC migration constitutive, enhanced with inflammation Effector and memory T cells

Tblast (effector function?)

Via thoracic duct

(destruct) Tanergic

Treg

Treg

Effector cell

Tmem

Efferent lymphatics

Migration of effector, naive, and memory T cells into bloodstream to lymphoid and nonlymphoid tissues Figure 1 Schematic diagram of mucosal T cell activation and migration. Infectious agents, foodstuffs, or allergens are ingested or inhaled, with subsequent antigen acquisition by APCs resident in mucosal parenchymal tissues. Resting APCs bearing self or innocuous antigens or APCs activated by infection/inﬂammation migrate via afferent lymphatics to draining LN and induce either tolerance or immunity. Effector cells exit the LN via efferent lymphatics and trafﬁc to nonlymphoid tissues. Entry into mucosal effector sites may result in further activational events that tune tissue-speciﬁc effector functions.

circumstances of infection or perhaps chronic inﬂammation, induced secondary lymphoid tissue can participate in immune response initiation and control.

Effector Sites of Mucosae Once activated in secondary lymphoid tissues, T cells are able to leave the inductive sites

Activated DC

Tblast

Naive

Blood

Naive T cell

and migrate to all nonlymphoid tissues (22). Those lymphocytes capable of entering nonlymphoid tissues during a primary immune response are likely to be effector cells at the time of entry, or immediately thereafter, e.g., upon re-encounter with antigen within the effector site. The environment of antigen presentation is likely unique to the effector site,

www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

683

Effector cell

ARI

15 February 2006

2:12

representing the compilation of antigen exposures in that site. For example, the immunological milieu in the gut is inﬂuenced by antigens derived from food or commensal bacteria or in the lung from environmental aeroantigens or respiratory infection. Therefore the capacity of cells of the innate immune system to present antigen within each nonlymphoid tissue is likely to be distinct [because dendritic cell populations are distinct in lymph nodes draining different tissues (23, 24)], and this is an area of intense investigation. The lamina propria of the intestine (gut LP) is a loosely organized connective tissue beneath the basement membrane supporting the overlying epithelial cells of the small and large intestine. The LP contains a variety of cells of the innate and adaptive immune system including IgA-producing plasma cells, DCs, macrophages, and T and B cells. Furthermore, the vast majority of T cells present in the gut LP express an effector/memory phenotype (25–27). An additional population of T lymphocytes (the intraepithelial lymphocytes, IELs) are resident within the intraepithelial spaces above the basement membrane and below the tight junctions of the epithelial cells (28, 29). Although IELs are composed in part of unique T cell subsets including TCRγδ and TCRαβCD8αα cells (28, 30, 31), those CD4 or CD8β+ IELs with conventional phenotypes are also largely of the effector/memory phenotype (32, 33). For the lung, effector cells are located primarily in the lung parenchyma and the airways. Although the lung parenchyma appears analogous to the gut LP, there is no cell population in the intestine topologically equivalent to that in the lung airways [those cells removable by bronchoalveolar lavage (BAL), located above the epithelial cells in the airway lumen]. The parenchyma or interstitium of the lung, the lamina propria or lung LP, underlies the basement membrane supporting the overlying epithelial cells, and it is composed of loosely organized connective tissue containing T cells, APCs, and many other cell types

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

684

Lefran¸cois

·

Puddington

of hematopoietic origin. Moreover, similar to the gut LP, many of the T cells in the lung LP express effector/memory phenotypes. However, in contrast to the gut, the lung LP in normal mice contains a signiﬁcant population of naive T cells. T cell entry into the lung airways, as with entry into the intestinal epithelium, is tightly regulated. These T cells, although closely associated with the epithelium, are essentially outside of the body. Nevertheless, airway CD4 and CD8 T cells can mediate protection against respiratory virus infection (34, 35) as well as participate in deleterious inﬂammatory reactions, such as allergic asthma (36, 37).

INITIATION OF MUCOSAL T CELL RESPONSES Antigen Acquisition The ﬁrst step necessary for mucosal T cell activation is the acquisition of antigen by APCs. The mechanism of antigen capture in the mucosal tissues is dependent in part on the nature of the immunogen. For example, soluble protein antigens (e.g., some allergens) may be passively absorbed across the epithelial surfaces of the mucosal tissues (38). In contrast, particular pathogens, for example, Salmonella sp. and reovirus, preferentially bind and enter through intestinal PP (39, 40). In some cases, speciﬁc receptors have been identiﬁed on epithelial cells (e.g., reovirus binding to M cells) that pathogens have subjugated for adherence and entry into the body. Listeria monocytogenes (LM) utilizes epithelial cell Ecadherin as a receptor for internalin, which promotes entry into the epithelium (41). Similar targeting pathways exist for lung-speciﬁc infections or certain aeroallergens. For example, in inﬂuenza virus infection, only the respiratory epithelial cells express the protease requisite for cleavage of the hemagglutinin protein necessary for production of infectious virus (42). Many allergens contain protease functions that facilitate their entry into the mucosae by cleavage of the tight junction

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

proteins between epithelial cells (43, 44). Thus, the form of the antigen and its biological characteristics determine the mode of access into the host. DCs are believed to be essential for activation of naive T cells in the inductive sites of the mucosae. DCs are located in speciﬁc regions of the PP, form a contiguous network within the LP of intestine and lung, but are rare or absent from the epithelial layers (with the exception of the trachea) under normal conditions (45, 46). At steady state there is a slow rate of constant trafﬁc of immature DCs from the mucosal tissues via the lymph to the draining LN (i.e., MLN, MedLN) (8, 16, 47–49). Inﬂammation in the mucosa results in a dramatic increase in numbers of activated DCs and other APCs in nonlymphoid tissues and migration to the draining LN (14, 45, 50). The increase in DCs in response to inﬂammatory signals is likely the result of recruitment of circulating immature APC precursors from the blood that then undergo development into particular APC subsets as directed by the “inﬂammatory” microenvironment (51–55). APCs may acquire antigen directly through pinocytosis, as for soluble proteins (by DCs), or via phagocytosis of particulate material (by DCs or macrophages). Particulate antigens are thought to be more immunogenic and include intact bacteria or viruses, as well as material derived from dead and dying cells, or immune complexes (56–60). In addition, some microorganisms actively infect DCs or other cells capable of presenting antigen, whereas others are unable to do so (61–66). Antigen-speciﬁc memory B cells are potential APCs, capable of acquiring antigen via cell-surface immunoglobulin. However, their paucity in number and speciﬁc localization in lymphoid and nonlymphoid tissue may restrict their ability to restimulate T cells to those colocalized with antigenspeciﬁc CD4+ cells. In the intestinal mucosa a novel method for antigen sampling has been visualized. DCs in the LP are able to extend dendrites through the basement membrane and between the ep-

ithelial cells into the gut lumen (46, 67). In this way, antigen and bacteria can be obtained and carried into the body. As DCs in MLN contain material derived from effete intestinal epithelial cells (IEC) (9), perhaps their acquisition by DCs is via this process. Under steady-state conditions, this continuous sampling and migration by DCs is likely responsible for induction and maintenance of T cell tolerance to food antigens and normal ﬂora (68). A similar system is thought to be in place in the lung, although this has not been demonstrated directly. Note also that macrophages, although rare in LN, are present at the border between red and white pulp in the spleen, are abundant in the lung airways and LP, and are also present in the gut LP. Macrophages could potentially serve as APCs, perhaps to initiate T cell recall responses (69). “Activated” macrophages may also be competent to prime naive T cells (70), e.g., perhaps in the context of a mucosal inﬂammatory response. However, in the respiratory tract of naive mice, it is generally accepted that the resident “resting” macrophages suppress T cell activation via their effects on colocalized DCs (45, 71).

T Cell-APC Interaction in Mucosal Lymphoid Tissues: Induction of Tolerance Versus Immunity Although direct evidence for DC involvement in mucosal T cell responses exists in only a limited number of cases, the current paradigm favors the DC as the requisite APC for initiation of all primary T cell responses irrespective of whether the immunological outcome is immunity or tolerance. However, deﬁnitive proof of this idea has been limited until recently (72), owing to the lack of systems in which DCs can be selectively depleted. Although the CD11c-DTR model represents a signiﬁcant advance to studies of the role of APCs in T cell responses, use of this model is more limited in studies of immune responses in mucosal sites. This limitation is due to the fact that CD11c is expressed by macrophages in the respiratory tract and spleen, which are

www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

685

ARI

15 February 2006

2:12

depleted by DT treatment of transgenic mice (73, 74), and the potential complication with subsets of mucosal effector T cells that express CD11c (75, 76). Nonetheless, confocal and intravital imaging analysis is beginning to provide views of T cell-DC interactions in situ and in real time (77–79). A large body of evidence from these studies and others points to the DC as the main initiator of mucosal T cell responses (24, 49, 73, 80–83), although in some cases non-DC APCs may participate in T cell activation. T cell encounter with antigen-bearing DCs in the mucosal inductive sites can result in multiple outcomes depending on the nature and form of the antigen (Figure 1). Administration of protein antigens such as ovalbumin (OVA) either orally or as an aerosol induces antigen-speciﬁc T cell tolerance (84– 92). This tolerance is systemic, presumably due to the dissemination of antigen into the blood following absorption in either the lung or the intestine. Tolerance is likely mediated by antigen presented by resting APCs and ultimately results in T cell deletion and/or generation of T cells with immunoregulatory function (93–99). In either case, it has been demonstrated in situations of antigeninduced T cell tolerance in vivo that some level of T cell activation and proliferation occurs prior to deletion (100, 101) or the generation of regulatory T cells (102, 103). Thus, the DC-T cell interaction drives T cell activation but critical signals required to mount a productive immune response are absent. In the absence of Toll-like receptor (TLR) ligands and/or inﬂammatory cytokines, the DC is unable to provide sufﬁcient costimulation to the T cell, which leads to an abortive event ultimately resulting in deletion or perhaps anergy in the case of CD4 T cells (50, 101, 104–110). Early events in induction of CD4 T cell tolerance versus protective immunity in the MLN have recently been visualized by two-photon videomicroscopy (111). After feeding large quantities of OVA with or without adjuvant to induce either immunity or tolerance, respectively, adoptively transferred

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

686

Lefran¸cois

·

Puddington

OVA-speciﬁc CD4 T cells behave quite differently. In the tolerogenic situation, the CD4 T cells form smaller and shorter-lived clusters in the MLN than CD4 T cells in the immunityinducing environment. This series of events can be recapitulated for self antigens as well. When a neoself antigen is expressed in the intestinal epithelium in transgenic mice, systemic tolerance is induced (112). However, low levels of antigen may be expressed in the thymus, as are many “tissue-speciﬁc” proteins (113, 114). Nevertheless, transfer of naive antigen-speciﬁc CD8 T cells into such mice results in T cell accumulation preferentially in the MLN and PP, which indicates that intestinal epithelial cell-derived proteins can gain access to the intestinal inductive sites. Because the model neoself antigen was an engineered OVA protein whose expression was strictly cytoplasmic, this result suggests that DCs in the PP or gut LP are able to acquire epithelial cellderived antigen and present it either directly within the PP or following their migration via the lymphatics to the MLN. Thus, these results are in agreement with the ﬁndings cited above showing continuous trafﬁcking of DCs from the intestinal mucosa to the MLN, and they provide a mechanism for induction and maintenance of self tolerance to intestinal epithelial cell-speciﬁc self proteins, and potentially other innocuous antigens, such as those derived from food acquired from the gut lumen. Maintenance of mucosal tolerance is critical, as evidenced by the inﬂammatory bowel disease (IBD)-like syndromes that occur in various cell or cytokine deﬁciencies affecting immunoregulation (115). However, inﬂammatory mediators and infections can also provide signals that “break” T cell tolerance in the mucosae. For example, the adoptive transfer of naive neoself antigen-speciﬁc CD8 cells to the iFABP (intestinal fatty-acid-binding protein promoter)-OVA transgenic mice described above is necessary and sufﬁcient to initiate destruction of intestinal tissue in response to nonspeciﬁc inﬂammation induced

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

by viral infection or TLR ligand stimulation (116). In the absence of such inﬂammatory mediators, transfer of the same population of neoself antigen-speciﬁc CD8 cells generates large numbers of activated T cells that enter the intestinal epithelium but do not cause tissue damage. Thus, the addition of inﬂammatory signals, even in the absence of added antigen, promotes a greater level of T cell activation that subsequently results in tissue pathology. Because the capacity of APC function within mucosal sites is responsive to environmental stimuli, changes in the outcome of coincident T cell (re)encounter with antigens can be induced (50, 106, 109, 116–118). Similar events may regulate CD4 T cell autoreactivity since normal bacterial ﬂora are required for induction of colitis in models where CD4 T cells are the aggressors (119, 120). In these cases, it is unclear whether the “autoreactive” T cells are speciﬁc for intestinal selfantigens or are reactive with lumenal bacterial antigens, although there is experimental evidence in support of the latter (121). In either scenario, gut cytokine dysregulation (e.g., lack of anti-inﬂammatory cytokines such as IL-10 or TGF-β) may lead to defective control of T cell proliferation and activation with concomitant production of deleterious levels of proinﬂammatory cytokines that mediate bystander tissue damage. Similar to the oral route, administration of innocuous proteins to the lung via an aerosol or intranasally results in the induction of systemic tolerance. In the respiratory tract, it has been demonstrated that anti-inﬂammatory cytokines [e.g., IFNγ (85) or IL-10 (90)], regulatory T cells [including some populations of TCRγδ cells (85)], or speciﬁc populations of APCs [e.g., alveolar macrophages (71) or plasmacytoid DCs (99)] are involved in maintaining tolerance in the lung. All of these mechanisms likely contribute to some aspect of the tolerogenic state of the lung in healthy animals. An additional level of complexity in the lung is the observation that the distribution and function of DCs and other APCs in the respiratory tract is based on anatomical

location (45) and this may also apply to the intestinal mucosa.

T CELL MIGRATION TO MUCOSAL TISSUES Activated T Cell Migration to Mucosal Tissues In general, activation engenders properties to the T cells that promote their migration to nonlymphoid tissues. For example, the great majority of CD8 T cells, regardless of the secondary lymphoid tissue in which they are activated, upregulate a variety of adhesion molecules, such as CD11a and CD44, that are likely to participate in movement of these cells into multiple nonlymphoid tissues (22). Broadly expressed trafﬁcking molecules such as these may be required but are insufﬁcient on their own to promote entry into a particular tissue. For tissue-speciﬁc homing, specialized adhesion and chemokine receptors are required (122). CD8 effector T cell migration to the intestinal LP requires the α4β7 integrin (10, 123, 124), whose major counterreceptor is MadCAM (125). CD8 T cells lacking this integrin and responding to a systemic virus infection fail to enter the LP and consequently do not enter the IEL compartment (10). Expression of the β7 integrin is required for CD4+ regulatory T cells to enter the MLN but not to inhibit intestinal pathology in a murine model of colitis (126). Interruption of P-selectin ligand interactions also inhibits entry of effector CD4 T cells into the LP (124). Although naive β7−/− CD8 T cells enter the MLN, activated β7−/− CD8 T cells do not. This phenomenon may be the result of the use of L-selectin and CCR7 by naive T cells to enter MLN (127) while these molecules are downregulated by T cell activation, thereby increasing the relative importance of α4β7 for MLN entry. In contrast, most naive lymphocytes appear to require α4β7 to enter the PP. The second β7 integrin, αEβ7(CD103), is highly expressed by many mucosal T cells (128–130) and binds to E-cadherin (131,

www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

687

ARI

15 February 2006

2:12

132). This integrin is not required for CD8 T cell migration to the intestine and also does not appear to be involved in retention of virus-speciﬁc effector or memory cells in the epithelium (10), although more rigorous testing is required to solidify this conclusion. However, the αEβ7 integrin is important for epithelial retention of autoreactive T cells in graft-versus-host disease and is involved in tissue destruction (133). Chemokines and their receptors are also important for lymphocyte migration to the intestinal mucosa. In particular, CCR9 (whose ligand TECK is expressed by IEC in the small but not the large intestine) is involved in T cell migration to the mucosa (134–136). Elegant studies have shown deﬁnitively that CCR9 expression by activated CD8 T cells is crucial to their ability to migrate to the intestinal epithelium (136, 137). However, it is not known whether CCR9 is required for activated T cell migration to the intestinal LP and whether CD4 T cells behave similarly. Although less is known regarding antigenspeciﬁc T cell migration to the lung, significant headway in this area has been recently made [for an excellent review see (138)]. Although the role of integrins in activated T cell migration to the lung is not fully understood, it is known that integrins are expressed by leukocytes, particularly those present within the lung airways (139), and that integrins are signiﬁcantly involved during the effector phase of inﬂammatory responses within the respiratory tract (140–142). Similar to the gut, accumulation of lymphocytes in the BAL and lung LP as induced by challenge with aerosolized allergen is inhibited by interruption of α4-containing integrins (α4β1 and/or α4β7) or P-selectin, providing protection from allergic airway disease (142, 143). In these models, because T cells are initially primed by systemic immunization with allergen in adjuvant, the antigen-speciﬁc T cells are likely localized in the lung prior to treatment with the integrin-blocking reagents and challenge with aerosolized allergen. Thus, these integrins may well be necessary for ad-

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

688

Lefran¸cois

·

Puddington

ditional T cell recruitment to the lung during the inﬂammatory response or they participate during the second hit within mucosal sites (see below). Another example is the collagenbinding α1β1 integrin (VLA-1), which is highly expressed by virus-speciﬁc primary and memory CD8 T cells in the lung airways following inﬂuenza virus infection (144). Although this integrin does not appear to be required for migration of CD8 T cells to the lung, VLA-1 blockade inhibits protection against secondary infection by an as-yet undeﬁned mechanism. Similar to many intestinal T cells, effector T cells in the airways responding to respiratory virus infection or allergen challenge in allergic airway disease express the αEβ7 integrin (139, 145). Thus, expression of this adhesion molecule is linked to mucosal tissues where its ligand, E-cadherin, is expressed. The function of αEβ7 in the lung remains unclear, but in the gut may be important for retention of T cells and interaction with epithelial cells when antigen is present (133). The β2 integrin, CD11c, traditionally thought of as a DC marker, is also expressed by subsets of cytotoxic lung and gut LP T cells and IELs, but the impact of the expression of the CD11c molecule itself on T cell function is not known (75, 76). A novel mechanism of lymphocyte migration to the lung has recently been described involving leukotriene and prostaglandin receptors. CD4 T cells utilize the leukotriene B4 receptor BLT1 in early effector cell migration to the inﬂamed lung (146). Prostaglandin D2 is also involved in CD4 T cell migration, especially Th2-type cells, to the lung and the lung airways. PGD2 interacts with its receptor, DP1, expressed by lung epithelium, which in turn induces production of T cell-attracting chemokines. The chemokine receptors involved in CD4 Th2 effector migration to the lung include CCR3, CCR4, and CCR8. Thus, through a multitiered induction of chemoattractants, effector T cells are recruited to the lung. In murine models of asthma, such as allergic airway disease, initiation of mucosal disease requires a population

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

of memory T cells capable of producing proinﬂammatory cytokines upon exposure to recall antigen within the respiratory tract [e.g., IFNγ to upregulate VCAM/ICAM expression by epithelial cells (148, 149)]. It is hypothesized that antigen-speciﬁc T cells enter the lung LP upon differentiation to effector cells during the primary response. In other models, it was shown that a primary immune response within the respiratory tract is induced by introduction of aerosolized antigen in the context of a TLR agonist, such as LPS, or proinﬂammatory cytokine, such as TNF (50, 109, 150). Adoptive transfer of effector T cells generated in vitro in conjunction with aerosolized antigen challenge is also sufﬁcient to initiate the airway inﬂammatory response (151, 152). The overproduction of a wide variety of cytokines by lung epithelial cells, e.g., IL-4, IL-9, or IL-13 in naive mice, results in T cell recruitment to the lung and airways and spontaneous airway inﬂammation typical of asthma (143, 153, 154). Models of allergic intestinal disease have also been developed but the requirements for T cell migration to the intestinal mucosa in these systems have not been elucidated (155–157).

Imprinting of Tissue-Homing Properties Tissue-speciﬁc homing is a long-held concept suggesting that activated T cells preferentially home to the tissue in which they were originally primed (158). Recent results indicate that DCs derived from mucosal lymphoid sites (e.g., MLN and PP) preferentially induce mucosal homing molecules on the responding T cells (Figure 1) (137, 159, 160). Thus, activation of CD4 or CD8 T cells by MLN or PP DCs induces expression of α4β7 and CCR9. In contrast, DCs from the spleen or peripheral LN do not preferentially induce these molecules. Rather, LN draining the skin impart a skin-homing phenotype to the T cells. The vitamin A metabolite, retinoic acid (RA), has been implicated in imprinting intestinal tropism (α4β7 and CCR9 expression) to acti-

vated CD4 T cells, and DCs from MLN and PP express the enzymes required for production of RA from retinol (161). Whether imprinting is a feature of all LNdraining nonlymphoid tissues is unknown, and thus far induction of tissue-speciﬁc tropism of T cells migrating to the lung has not been described. Note also that initiation of an immune response in the secondary lymphoid tissues draining a mucosal site is not absolutely required for T cell migration to that site. For example, low-dose intravenous LM infection results in T cell priming nearly exclusively in the spleen, yet antigen-speciﬁc T cells migrate efﬁciently to the gut LP as well as to the lung (K.D. Klonowski, A.L. Marzo, K.J. Williams, S.-J. Lee, Q.-M. Pham & L. Lefranc¸ois, unpublished results). Inﬂuenza virus or Sendai virus infection of the lung also generates a primary and memory CD8 T cell response, albeit small, in the gut LP, whereas oral rotavirus infection induces primary CD8 T cells capable of migrating to the lung and liver (22). In general, activation of CD8 T cells, regardless of the priming site, imparts a transient ability to migrate to all nonlymphoid tissues. Whether the same is true for CD4 T cells has yet to be rigorously tested. This is not to say that priming in a mucosal site does not impart a preference for migration to mucosal effector sites, which in itself may enhance immune responses and perhaps memory generation within that tissue.

FUNCTIONAL REGULATION OF T CELL RESPONSES IN MUCOSAL EFFECTOR SITES: THE SECOND-HIT HYPOTHESIS Once primary activation of T cells occurs in LN-draining mucosae, the activated cells leave the LN and migrate to mucosal effector sites. Because effector T cells in mucosal tissues may exhibit distinct functional properties, a question arises as to the location where functional differentiation occurs.

www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

689

ARI

15 February 2006

2:12

That is, are functional properties imparted during priming in secondary lymphoid organs or do changes occur as a result of entry into nonlymphoid tissues or residence in that tissue? The function or phenotype of effector or memory T cells entering tissue may be impinged upon by at least three events: (a) traversing endothelium which requires adhesion receptor engagement; (b) interaction with costimulators expressed by APCs or non-APCs (which may include TCR engagement with antigen); or (c) interaction with locally produced cytokines. These phenomena may be considered as a “second hit” that occurs following the initial priming event and results in functional modiﬁcations of the responding cells. Although endothelial cells may in some cases present antigen and express costimulators (162), no data are available that demonstrate a role for endothelium in mucosal T cell activation. In addition, the process of transendothelial migration, which may require engagement of selectins, integrins, and chemokine receptors (163), could functionally alter migrating T cells, although a direct demonstration of such effects on mucosal T cells is lacking. However, there is reasonable evidence that secondary costimulator signaling can occur upon entry of T cells into the lung or intestinal mucosa effector sites. For example, the response of antiviral CD8 T cells migrating into the gut LP is ampliﬁed by CD40/CD40L interactions (164). A recent report also shows that the CD27/CD70 costimulator pair enhances intestinal effector CD8 T cell responses to oral LM infection, apparently without affecting priming, although this was not directly examined (165). A CD70+ LP APC of nonhematopoietic origin is implicated in this process, although other CD70+ APCs could also be involved. Local cytokines can also affect T cells in the intestinal mucosa. Memory CD8 T cells migrating into the gut LP and IEL compartment upregulate αEβ7 and CD69 (166–168), molecules whose expression is upregulated by TGF-β and IFNα, respectively. TGF-β also

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

690

Lefran¸cois

·

Puddington

promotes αEβ7 expression on incoming T cells in a model of graft-versus-host disease, and this integrin is involved in inducing IEC pathology (133). In the lung and the intestine, secondary costimulatory events as well as local cytokines play important roles in regulating T cell responses. For example, antiviral (LCMV)speciﬁc CD8 T cells migrating into the lung can alter their cytokine production from IFNγ to IL-5 when a concomitant Th2 response is under way (169) and allergic airway inﬂammation is reduced by a concomitant virus infection (170). Also, an enteric parasitic infection dampens the allergic response to a dietary allergen (171, 172). Costimulation in the parenchymal tissue following priming in the LN or other secondary lymphoid organs may also occur in the mucosae. For example, following systemic infection with vesicular stomatitis virus (VSV), the frequency of antigen-speciﬁc IL-5-producing CD4 T cells was ﬁvefold greater than the frequency of IL4-producing cells speciﬁcally in the lung. Similar results were observed following sensitization with OVA/Alum in a murine model of allergic airway disease. This was in contrast to the frequency of IL-4 or IL-5 cytokine production by antigen-speciﬁc CD4 T cells in the spleen, which was equivalent in both systems. Production of IL-5, but not of IL-4, by the antiviral CD4 T cells was greatly inhibited in the lung and intestinal LP following blockade of the inducible costimulator (ICOS) (A.L. Marzo, L. Puddington & L. Lefranc¸ois, unpublished results). Moreover, ICOS is involved in regulating Th2 cytokine production during allergic airway disease, although the precise site or cell type that mediates this effect is unknown (173). Thus, there appear to be both positive and negative costimulators that are not necessarily involved in priming in LN or spleen, but that ﬁnetune the T cell response of effector cells upon entry to mucosal tissues. It also remains possible that DCs migrating from mucosal tissues to draining LN, which can be basal and constitutive or accelerated based on the

ANRV270-IY24-22

ARI

15 February 2006

2:12

character of the mucosal inﬂammatory response, express a tissue-speciﬁc array of costimulators capable of inducing a speciﬁc set of functional outcomes.

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

MUCOSAL T CELL MEMORY Many of the T cells present in mucosal effector sites phenotypically resemble effector/memory cells. However, especially in the lung, a signiﬁcant population of naive CD4 and CD8 T cells can be found within the parenchyma (S.C. Cose & L. Lefranc¸ois, unpublished results). Such cells are likely to be transient residents of the tissue (174), and this may be a mechanism for inducing or maintaining tolerance to tissue-speciﬁc antigens (175). For memory T cells, there are distinct requirements for migration into the lung LP versus the gut LP and the IEL compartment. In the lung, the presence of memory T cells in the parenchyma does not necessarily indicate their presence in the airways. For example, in normal unimmunized mice, memory phenotype CD4 and CD8 T cells are present in the lung LP but few lymphocytes are present in the airways. In addition, following systemic infection with VSV or LM, although substantial populations of antigen-speciﬁc memory cells are present in the lung parenchyma, few cells are found in the airways. In contrast, following pulmonary infection with inﬂuenza or Sendai virus, or induction of allergic airway disease, memory cells are present in the lung parenchyma as well as in the airways for several weeks after challenge (34, 35, 176). Many of the memory cells in the airways following respiratory virus infection express the early activation antigen CD69 (76, 177), suggesting recent encounter with antigen. Local inﬂammation appears to be needed for recruitment of memory T cells into the airways. Antigennonspeciﬁc memory T cells can be recruited to the airways by heterologous viral infections or intranasal instillation of TLR ligands such as CpG DNA (178, 179). Whether a similar phenomenon occurs with T cell migration into the intestinal epithelium is not known.

There are two general types of memory T cells based on phenotype and location (180, 181). Central memory cells (Tcm) are present primarily in lymph nodes and express homing molecules necessary for entry into that site including the selectin CD62L and the chemokine receptor CCR7. Effector memory cells (Tem) lack these molecules and reside primarily in nonlymphoid tissues. Both populations can be found in the blood and spleen, suggesting that each has the ability to migrate. However, there is signiﬁcant phenotypic heterogeneity within each subtype even within nonlymphoid tissues (182), making function perhaps the most reliable indicator of memory cell subclass. That is, effector memory CD8 T cells in mouse and human express direct ex vivo lytic activity as compared to Tcm, whereas CD4 Tem exhibit increased effector cytokine production over Tcm (180, 183). There has also been signiﬁcant discussion as to the interrelatedness of Tem and Tcm (184– 186), and recent evidence suggests that under normal circumstances Tem and Tcm represent distinct, noninterconvertible memory cell lineages, at least when phenotypic segregation is based on CD62L expression (187). The protective capability of each subset is also controversial, but the location of the infection or tumor is likely to dictate the functional relevance of each subset (188). Thus, whereas protection against the spleen-centric LCMV requires Tcm (185), lung Tem protect against respiratory virus infection (188). It makes teleological sense to link protective capabilities of each subset to its tissue of residence, perhaps as dictated by the tropism of a particular pathogen, rather than to assign protective function to one or the other of the lineages, which would necessarily limit the efﬁcacy of recall responses. As discussed above, the imprinting of tissue-speciﬁc migratory abilities to effector cells occurs in secondary lymphoid tissues, such as the skin-draining LN or the small intestinal PP (159, 160). The resulting memory cells may retain these homing properties. Also, subsets of memory-phenotype

www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

691

ARI

15 February 2006

2:12

human peripheral blood T cells expressing either mucosal or skin-homing molecules have been described. However, there appears to be plasticity in the process of imprinting in that memory cells expressing a nonmucosal pattern of trafﬁcking molecules can be redirected to express a mucosal homing phenotype (189). This dynamic process suggests that migrating memory cells may modify their homing behavior depending on their location at the time of antigen reencounter. A remaining question is whether mucosal memory T cells exit the tissues and enter the blood stream and either home back to their tissue of origin or exhibit a less restricted homing pattern. Memory cells may exit tissues via the draining lymph, resulting in their eventual appearance in the blood. However, since effector T cells require CCR7 expression to exit tissues and move into the afferent lymph vessels (190, 191), it will be of interest to compare the ability of Tem versus Tcm to access draining lymph nodes of lung and intestine. The patterns of memory T cell migration to the mucosae have recently been studied using parabiosis (168). In this system, surgically joined mice with shared blood ﬂow allow analysis of migration of blood-borne lymphocytes. CD8 and CD4 memory T cells raised by systemic infection of one mouse with VSV or LM rapidly equilibrate in the lymphoid tissues as well as in the lung LP and liver of the conjoined partner. In contrast, memory T cell entry into the gut LP and epithelium is highly restricted, with only small numbers of cells entering these sites over several weeks. Memory T cell migration into the intestinal mucosa, including the PP, requires β7 integrin expression by the circulating T cells. Entry of memory T cells into the intestinal mucosa results in upregulation of CD69, even in uninfected partner mice, indicating that CD69 upregulation is not obligately linked to recent encounter with antigen in this tissue. Still unclear is the tissue of origin of the migrating memory cells. Thus, a common pool of mobile memory cells may have the ability to enter multiple tissues. Alternatively, mem-

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

692

Lefran¸cois

·

Puddington

ory T cells entering the lung may have originated in the partner lung, although why cells would enter the bloodstream only to return to their original location is unclear. There could also exist multiple pools of circulating memory cells, with subsets able to access either multiple or single tissues. In the intestine, the altered phenotype of the entering memory cells and the lack of such cells elsewhere suggest that memory T cells in the intestinal LP and epithelium are tissue-bound and do not exit the tissue, but this theory remains unproven. Similarly, airway memory cells appear to remain in that site rather than re-enter the lung LP (192). Finally, it remains to be examined whether memory T cell migration will be differentially affected when memory is raised against distinct pathogens or antigenic moieties. Thus, memory cells generated in response to infections restricted to mucosal tissues such as inﬂuenza virus in the lung or rotavirus in the intestine, or memory cells speciﬁc for chronic pathogens, may exhibit specialized migratory properties due to modiﬁcations at the level of the T cell, the APC, and/or the environmental milieu (e.g., infections or during chronic inﬂammatory disease).

CONCLUSIONS Mucosal T cells are essential to maintaining barrier function of a vast surface area. Consequently, mucosal T cells must be functionally diverse, acquiring the ability to combat a multitude of phylogenetically distinct organisms while also providing homeostatic control over dynamic inﬂammatory processes. Although substantial progress has been made in understanding mucosal T cell responses, many questions remain regarding the molecular events controlling their trafﬁcking and, in particular, the differentiation events leading to the emergence of effector functions. Further analysis of the mucosal immune system is critical to our ability to deliver efﬁcacious mucosal vaccines as well as to intervene to prevent or reverse tissue damage during deleterious autoimmune or other pathogenic conditions.

ANRV270-IY24-22

ARI

15 February 2006

2:12

SUMMARY POINTS 1. The anatomy of the mucosal tissues orchestrates immunity: Distinct inductive sites (lymph nodes, PP, BALT) and effector sites (LP, epithelium, airways) work in concert to provide effective local immunity. Unregulated T cell inﬂammation, however, can induce pathology (e.g., IBD, asthma).

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

2. APCs acquire antigen in parenchymal tissue and migrate to inductive sites via multiple mechanisms dependent on the source of antigen (e.g., characteristics of pathogen, self-antigen, location of antigen/infectious organism). 3. Tolerance or immunity results from APC-T cell interaction in the mucosal LN or PP. The choice depends in part on the status of the APC—resting APCs bearing self or innocuous antigens drive tolerance induction, whereas activated APCs drive immunity. 4. T cell responses in mucosal sites can inﬂuence the environment of antigen presentation within that site. 5. Naive, effector, and memory T cell migration to mucosal tissues is a multitiered event requiring distinct adhesion molecules for entry into lung versus intestine. Imprinting of T cells by mucosal DCs results in acquisition of tissue-speciﬁc homing abilities. 6. Mucosal T cells may be acted upon by secondary costimulatory events upon entry into the tissue, thereby providing tissue-speciﬁc tuning of effector functions. 7. Memory cells comprise a large proportion of mucosal T cells and exhibit functional and phenotypic characteristics unique to their location.

ACKNOWLEDGMENTS We thank the members of our laboratories for their dedication to deciphering the concepts discussed here. Our work is funded by the National Institutes of Health and the American Lung Association of Connecticut. We regret any errors of omission in referencing as a result of space limitations.

LITERATURE CITED 1. Mestecky J. 1987. The common mucosal immune system and current strategies for induction of immune responses in external secretions. J. Clin. Immunol. 7:265–76 2. Brandtzaeg P, Pabst R. 2004. Let’s go mucosal: communication on slippery ground. Trends Immunol. 25:570–77 3. Neutra MR. 1999. M cells in antigen sampling in mucosal tissues. Curr. Top. Microbiol. Immunol. 236:17–32 4. Neutra MR, Mantis NJ, Kraehenbuhl JP. 2001. Collaboration of epithelial cells with organized mucosal lymphoid tissues. Nat. Immunol. 2:1004–9 5. Kato T, Owen RL. 1994. Structure and function of intestinal mucosal epithelium. In Handbook of Mucosal Immunology, ed. PL Ogra, J Mestecky, ME Lamm, W Strober, JR Mcghee, J Bienenstock, pp. 11–26. San Diego: Academic 6. Fujihashi K, Dohi T, Rennert PD, Yamamoto M, Koga T, et al. 2001. Peyer’s patches are required for oral tolerance to proteins. Proc. Natl. Acad. Sci. USA 98:3310–15 www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

693

ARI

15 February 2006

2:12

7. Teitelbaum R, Schubert W, Gunther L, Kress Y, Macaluso F, et al. 1999. The M cell as a portal of entry to the lung for the bacterial pathogen Mycobacterium tuberculosis. Immunity 10:641–50 8. Liu LM, MacPherson GG. 1993. Antigen acquisition by dendritic cells: intestinal dendritic cells acquire antigen administered orally and can prime naive T cells in vivo. J. Exp. Med. 177:1299–307 9. Huang F-P, Platt N, Wykes M, Major JR, Powell TJ, et al. 2000. A discrete subpopulation of dendritic cells transports apoptotic intestinal epithelial cells to T cell areas of mesenteric lymph nodes. J. Exp. Med. 191:435–43 10. Lefranc¸ois L, Parker CM, Olson S, Schon MP, Muller W, et al. 1999. The role of β7 integrins in CD8 T cell trafﬁcking during an anti-viral immune response. J. Exp. Med. 189:1631–38 11. Eberl G. 2005. Inducible lymphoid tissues in the adult gut: recapitulation of a fetal developmental pathway? Nat. Rev. Immunol. 5:413–20 12. Flynn KJ, Belz GT, Altman JD, Ahmed R, Woodland DL, Doherty PC. 1998. Virusspeciﬁc CD8+ T cells in primary and secondary inﬂuenza pneumonia. Immunity 8:683–91 13. Usherwood EJ, Hogg TL, Woodland DL. 1999. Enumeration of antigen-presenting cells in mice infected with Sendai virus. J. Immunol. 162:3350–55 14. Legge KL, Braciale TJ. 2003. Accelerated migration of respiratory dendritic cells to the regional lymph nodes is limited to the early phase of pulmonary infection. Immunity 18:265–77 15. Lawrence CW, Braciale TJ. 2004. Activation, differentiation, and migration of naive virus-speciﬁc CD8+ T cells during pulmonary inﬂuenza virus infection. J. Immunol. 173:1209–18 16. Vermaelen KY, Carro-Muino I, Lambrecht BN, Pauwels RA. 2001. Speciﬁc migratory dendritic cells rapidly transport antigen from the airways to the thoracic lymph nodes. J. Exp. Med. 193:51–60 17. Zuercher AW, Cofﬁn SE, Thurnheer MC, Fundova P, Cebra JJ. 2002. Nasal-associated lymphoid tissue is a mucosal inductive site for virus-speciﬁc humoral and cellular immune responses. J. Immunol. 168:1796–803 18. Drayton DL, Chan K, Lesslauer W, Lee J, Ying XY, Ruddle NH. 2002. Lymphocyte trafﬁc in lymphoid organ neogenesis: differential roles of LTα and LTαβ. Adv. Exp. Med. Biol. 512:43–48 19. Tschernig T, Pabst R. 2000. Bronchus-associated lymphoid tissue (BALT) is not present in the normal adult lung but in different diseases. Pathobiology 68:1–8 20. Delventhal S, Hensel A, Petzoldt K, Pabst R. 1992. Effects of microbial stimulation on the number, size and activity of bronchus-associated lymphoid tissue (BALT) structures in the pig. Int. J. Exp. Pathol. 73:351–57 21. Moyron-Quiroz JE, Rangel-Moreno J, Kusser K, Hartson L, Sprague F, et al. 2004. Role of inducible bronchus associated lymphoid tissue (iBALT) in respiratory immunity. Nat. Med. 10:927–34 22. Masopust D, Vezys V, Usherwood EJ, Cauley LS, Olson S, et al. 2004. Activated primary and memory CD8 T cells migrate to nonlymphoid tissues regardless of site of activation or tissue of origin. J. Immunol. 172:4875–82 23. Henri S, Vremec D, Kamath A, Waithman J, Williams S, et al. 2001. The dendritic cell populations of mouse lymph nodes. J. Immunol. 167:741–48 24. Iwasaki A, Kelsall BL. 2001. Unique functions of CD11b+ , CD8α+ , and double-negative Peyer’s patch dendritic cells. J. Immunol. 166:4884–90

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

694

Lefran¸cois

·

Puddington

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

25. Targan SR, Deem RL, Liu M, Wang S, Nel A. 1995. Deﬁnition of a lamina propria T cell responsive state. Enhanced cytokine responsiveness of T cells stimulated through the CD2 pathway. J. Immunol. 154:664–75 26. Schieferdecker HL, Ullrich R, Hirseland H, Zeitz M. 1992. T cell differentiation antigens on lymphocytes in the human intestinal lamina propria. J. Immunol. 149:2816–22 27. De Maria R, Fais S, Silvestri M, Frati L, Pallone F, et al. 1993. Continuous in vivo activation and transient hyporesponsiveness to TCR/CD3 triggering of human gut lamina propria lymphocytes. Eur. J. Immunol. 23:3104–8 28. Goodman T, Lefranc¸ois L. 1989. Intraepithelial lymphocytes. Anatomical site, not T cell receptor form, dictates phenotype and function. J. Exp. Med. 170:1569–81 29. Lefranc¸ois L. 1991. Phenotypic complexity of intraepithelial lymphocytes of the small intestine. J. Immunol. 147:1746–51 30. Goodman T, Lefranc¸ois L. 1988. Expression of the γ-δ T-cell receptor on intestinal CD8+ intraepithelial lymphocytes. Nature 333:855–58 31. Bandeira A, Mota-Santos T, Itohara S, Degermann S, Heusser C, et al. 1990. Localization of γ/δ T cells to the intestinal epithelium is independent of normal microbial colonization. J. Exp. Med. 172:239–44 32. Parrott DMV, Tait C, MacKenzie S, Mowat AM, Davies MDJ, Micklem HS. 1983. Analysis of the effector functions of different populations of mucosal lymphocytes. Ann. NY Acad. Sci. 409:307–20 33. Wang HC, Zhou Q, Dragoo J, Klein JR. 2002. Most murine CD8+ intestinal intraepithelial lymphocytes are partially but not fully activated T cells. J. Immunol. 169:4717– 22 34. Hogan RJ, Zhong W, Usherwood EJ, Cookenham T, Roberts AD, Woodland DL. 2001. Protection from respiratory virus infections can be mediated by antigen-speciﬁc CD4+ T cells that persist in the lungs. J. Exp. Med. 193:981–86 35. Hogan RJ, Usherwood EJ, Zhong W, Roberts AA, Dutton RW, et al. 2001. Activated antigen-speciﬁc CD8+ T cells persist in the lungs following recovery from respiratory virus infections. J. Immunol. 166:1813–22 36. Hogan SP, Koskinen A, Matthaei KI, Young IG, Foster PS. 1998. Interleukin-5-producing CD4+ T cells play a pivotal role in aeroallergen-induced eosinophilia, bronchial hyperreactivity, and lung damage in mice. Am. J. Respir. Crit. Care Med. 157:210– 18 37. Cohn L, Tepper JS, Bottomly K. 1998. IL-4-independent induction of airway hyperresponsiveness by Th2, but not Th1, cells. J. Immunol. 161:3813–16 38. Mowat AM. 2005. Dendritic cells and immune responses to orally administered antigens. Vaccine 23:1797–99 39. Jones BD, Ghori N, Falkow S. 1994. Salmonella typhimurium initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer’s patches. J. Exp. Med. 180:15–23 40. Wolf JL, Rubin DH, Finberg R, Kauffman RS, Sharpe AH, et al. 1981. Intestinal M cells: a pathway for entry of reovirus into the host. Science 212:471–72 41. Mengaud J, Ohayon H, Gounon P, Mege R-M, Cossart P. 1996. E-cadherin is the receptor for internalin, a surface protein required for entry of L. monocytogenes into epithelial cells. Cell 84:923–32 42. Zhirnov OP, Ikizler MR, Wright PF. 2002. Cleavage of inﬂuenza A virus hemagglutinin in human respiratory epithelium is cell associated and sensitive to exogenous antiproteases. J. Virol. 76:8682–89 www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

695

ARI

15 February 2006

2:12

43. Kheradmand F, Kiss A, Xu J, Lee SH, Kolattukudy PE, Corry DB. 2002. A proteaseactivated pathway underlying Th cell type 2 activation and allergic lung disease. J. Immunol. 169:5904–11 44. Wan H, Winton HL, Soeller C, Tovey ER, Gruenert DC, et al. 1999. Der p 1 facilitates transepithelial allergen delivery by disruption of tight junctions. J. Clin. Invest. 104:123– 33 45. von Garnier C, Filgueira L, Wikstrom M, Smith M, Thomas JA, et al. 2005. Anatomical location determines the distribution and function of dendritic cells and other APCs in the respiratory tract. J. Immunol. 175:1609–18 46. Niess JH, Brand S, Gu X, Landsman L, Jung S, et al. 2005. CX3CR1-mediated dendritic cell access to the intestinal lumen and bacterial clearance. Science 307:254–58 47. Liu LM, MacPherson GG. 1991. Lymph-borne (veiled) dendritic cells can acquire and present intestinally administered antigens. Immunology 73:281–86 48. Liu LM, MacPherson GG. 1994. The role of dendritic cells in the uptake and presentation of oral antigens. Adv. Exp. Med. Biol. 355:81–86 49. Lambrecht BN, De Veerman M, Coyle AJ, Gutierrez-Ramos JC, Thielemans K, Pauwels RA. 2000. Myeloid dendritic cells induce Th2 responses to inhaled antigen, leading to eosinophilic airway inﬂammation. J. Clin. Invest. 106:551–59 50. Oriss TB, Ostroukhova M, Seguin-Devaux C, Dixon-McCarthy B, Stolz DB, et al. 2005. Dynamics of dendritic cell phenotype and interactions with CD4+ T cells in airway inﬂammation and tolerance. J. Immunol. 174:854–63 51. van Furth R, Diesselhoff-den Dulk MC, Mattie H. 1973. Quantitative study on the production and kinetics of mononuclear phagocytes during an acute inﬂammatory reaction. J. Exp. Med. 138:1314–30 52. McWilliam AS, Napoli S, Marsh AM, Pemper FL, Nelson DJ, et al. 1996. Dendritic cells are recruited into the airway epithelium during the inﬂammatory response to a broad spectrum of stimuli. J. Exp. Med. 184:2429–32 53. Randolph GJ, Inaba K, Robbiani DF, Steinman RM, Muller WA. 1999. Differentiation of phagocytic monocytes into lymph node dendritic cells in vivo. Immunity 11:753–61 54. Delneste Y, Charbonnier P, Herbault N, Magistrelli G, Caron G, et al. 2003. Interferon-γ switches monocyte differentiation from dendritic cells to macrophages. Blood 101:143–50 55. Geissmann F, Jung S, Littman DR. 2003. Blood monocytes consist of two principal subsets with distinct migratory properties. Immunity 19:71–82 56. Steinman RM, Hawiger D, Nussenzweig MC. 2003. Tolerogenic dendritic cells. Annu. Rev. Immunol. 21:685–711 57. Aichele P, Zinke J, Grode L, Schwendener RA, Kaufmann SH, Seiler P. 2003. Macrophages of the splenic marginal zone are essential for trapping of blood-borne particulate antigen but dispensable for induction of speciﬁc T cell responses. J. Immunol. 171:1148–55 58. den Haan JM, Bevan MJ. 2002. Constitutive versus activation-dependent crosspresentation of immune complexes by CD8+ and CD8− dendritic cells in vivo. J. Exp. Med. 196:817–27 59. Schuurhuis DH, Ioan-Facsinay A, Nagelkerken B, van Schip JJ, Sedlik C, et al. 2002. Antigen-antibody immune complexes empower dendritic cells to efﬁciently prime speciﬁc CD8+ CTL responses in vivo. J. Immunol. 168:2240–46 60. Zuberi RI, Apgar JR, Chen S-S, Liu F-T. 2000. Role for IgE in airway secretions: IgE immune complexes are more potent inducers than antigen alone of airway inﬂammation in a murine model. J. Immunol. 164:2667–73

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

696

Lefran¸cois

·

Puddington

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

61. Bender A, Albert M, Reddy A, Feldman M, Sauter B, et al. 1998. The distinctive features of inﬂuenza virus infection of dendritic cells. Immunobiology 198:552–67 62. Bhardwaj N. 1997. Interactions of viruses with dendritic cells: a double-edged sword. J. Exp. Med. 186:795–99 63. Bender A, Bui LK, Feldman MAV, Larsson M, Bhardwaj N. 1995. Inactivated inﬂuenza virus, when presented on dendritic cells, elicits human CD8+ cytolytic T cell responses. J. Exp. Med. 182:1663–71 64. Feng H, Zhang D, Palliser D, Zhu P, Cai S, et al. 2005. Listeria-infected myeloid dendritic cells produce IFN-β, priming T cell activation. J. Immunol. 175:421–32 65. Gerna G, Percivalle E, Lilleri D, Lozza L, Fornara C, et al. 2005. Dendritic-cell infection by human cytomegalovirus is restricted to strains carrying functional UL131–128 genes and mediates efﬁcient viral antigen presentation to CD8+ T cells. J. Gen. Virol. 86:275–84 66. MacDonald GH, Johnston RE. 2000. Role of dendritic cell targeting in Venezuelan equine encephalitis virus pathogenesis. J. Virol. 74:914–22 67. Rescigno M, Urbano M, Valzasina B, Francolini M, Rotta G, et al. 2001. Dendritic cells express tight junction proteins and penetrate gut epithelial monolayers to sample bacteria. Nat. Immunol. 2:361–67 68. Mowat AM, Donachie AM, Parker LA, Robson NC, Beacock-Sharp H, et al. 2003. The role of dendritic cells in regulating mucosal immunity and tolerance. Novartis Found. Symp. 252:291–302 69. Crowe SR, Turner SJ, Miller SC, Roberts AD, Rappolo RA, et al. 2003. Differential antigen presentation regulates the changing patterns of CD8+ T cell immunodominance in primary and secondary inﬂuenza virus infections. J. Exp. Med. 198:399–410 70. Pozzi LA, Maciaszek JW, Rock KL. 2005. Both dendritic cells and macrophages can stimulate naive CD8 T cells in vivo to proliferate, develop effector function, and differentiate into memory cells. J. Immunol. 175:2071–81 71. Holt PG. 2000. Antigen presentation in the lung. Am. J. Respir. Crit. Care Med. 162:S151– 56 72. Jung S, Unutmaz D, Wong P, Sano G, De los Santos K, et al. 2002. In vivo depletion of CD11c+ dendritic cells abrogates priming of CD8+ T cells by exogenous cell-associated antigens. Immunity 17:211–20 73. van Rijt LS, Jung S, Kleinjan A, Vos N, Willart M, et al. 2005. In vivo depletion of lung CD11c+ dendritic cells during allergen challenge abrogates the characteristic features of asthma. J. Exp. Med. 201:981–91 74. Probst HC, Tschannen K, Odermatt B, Schwendener R, Zinkernagel RM, van den Broek M. 2005. Histological analysis of CD11c-DTR/GFP mice after in vivo depletion of dendritic cells. Clin. Exp. Immunol. 141:398–404 75. Huleatt JW, Lefranc¸ois L. 1995. Antigen-driven induction of CD11c on intestinal intraepithelial lymphocytes and CD8+ T cells in vivo. J. Immunol. 154:5684–93 76. Beyer M, Wang H, Peters N, Doths S, Koerner-Rettberg C, et al. 2005. The β2 integrin CD11c distinguishes a subset of cytotoxic pulmonary T cells with potent antiviral effects in vitro and in vivo. Respir. Res. 6:70–80 77. Miller MJ, Wei SH, Parker I, Cahalan MD. 2002. Two-photon imaging of lymphocyte motility and antigen response in intact lymph node. Science 296:1869–73 78. Bousso P, Robey E. 2003. Dynamics of CD8+ T cell priming by dendritic cells in intact lymph nodes. Nat. Immunol. 4:579–85 79. Sumen C, Mempel TR, Mazo IB, von Andrian UH. 2004. Intravital microscopy: visualizing immunity in context. Immunity 21:315–29 www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

697

ARI

15 February 2006

2:12

80. Lambrecht BN, Salomon B, Klatzmann D, Pauwels RA. 1998. Dendritic cells are required for the development of chronic eosinophilic airway inﬂammation in response to inhaled antigen in sensitized mice. J. Immunol. 160:4090–97 81. Fleeton M, Contractor N, Leon F, He J, Wetzel D, et al. 2004. Involvement of dendritic cell subsets in the induction of oral tolerance and immunity. Ann. NY Acad. Sci. 1029:60– 65 82. Bellini A, Vittori E, Marini M, Ackerman V, Mattoli S. 1993. Intraepithelial dendritic cells and selective activation of Th2-like lymphocytes in patients with atopic asthma. Chest 103:997–1005 83. van den Heuvel MM, Vanhee DD, Postmus PE, Hoefsmit EC, Beelen RH. 1998. Functional and phenotypic differences of monocyte-derived dendritic cells from allergic and nonallergic patients. J. Allergy Clin. Immunol. 101:90–95 84. Seymour BWP, Gershwin LJ, Coffman RL. 1998. Aerosol-induced immunoglobulin (Ig)-E unresponsiveness to ovalbumin does not require CD8+ or T cell receptor (TCR)γ/δ+ T cells or interferon (IFN)-γ in a murine model of allergen sensitization. J. Exp. Med. 187:721–31 85. McMenamin C, Pimm C, McKersey M, Holt PG. 1994. Regulation of IgE responses to inhaled antigen in mice by antigen-speciﬁc γδ T cells. Science 265:1869–71 86. Maclean JS, Sauty A, Luster AD, Drazen JM, De Sanctis GT. 1999. Antigen-induced airway hyperresponsiveness, pulmonary eosinophilia and chemokine expression in B-cell deﬁcient mice. Am. J. Respir. Cell Mol. Biol. 20:379–87 87. Strobel S, Mowat AM. 1998. Immune responses to dietary antigens: oral tolerance. Immunol. Today 19:173–181 88. Zivny JH, Moldoveanu Z, Vu HL, Russell MW, Mestecky J, Elson CO. 2001. Mechanisms of immune tolerance to food antigens in humans. Clin. Immunol. 101:158–68 89. Tsitoura DC, DeKruyff RH, Lamb JR, Umetsu DT. 1999. Intranasal exposure to protein antigen induces immunological tolerance mediated by functionally disabled CD4+ T cells. J. Immunol. 163:2592–600 90. Akbari O, DeKruyff RH, Umetsu DT. 2001. Pulmonary dendritic cells producing IL-10 mediate tolerance induced by respiratory exposure to antigen. Nat. Immunol. 2:725–31 91. Mowat AM. 2003. Anatomical basis of tolerance and immunity to intestinal antigens. Nat. Rev. Immunol. 3:331–41 92. Chirdo FG, Millington OR, Beacock-Sharp H, Mowat AM. 2005. Immunomodulatory dendritic cells in intestinal lamina propria. Eur. J. Immunol. 35:1831–40 93. Viney JL, Mowat AM, O’Malley JM, Williamson E, Fanger NA. 1998. Expanding dendritic cells in vivo enhances the induction of oral tolerance. J. Immunol. 160:5815–25 94. Haﬂer DA, Kent SC, Pietrusewicz MJ, Khoury SJ, Weiner HL, Fukaura H. 1997. Oral administration of myelin induces antigen-speciﬁc TGF-β1 secreting T cells in patients with multiple sclerosis. Ann. NY Acad. Sci. 835:120–31 95. Chen Y, Inobe J-I, Marks R, Gonnella P, Kuchroo VK, Weiner HL. 1995. Peripheral deletion of antigen-reactive T cells in oral tolerance. Nature 376:177–80 96. Chen YH, Kuchroo VK, Inobe J, Haﬂer DA, Weiner HL. 1994. Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis. Science 265:1237–40 97. Friedman A, Weiner HL. 1994. Induction of anergy or active suppression following oral tolerance is determined by antigen dosage. Proc. Natl. Acad. Sci. USA 91:6688–92 98. Mottet C, Uhlig HH, Powrie F. 2003. Cure of colitis by CD4+ CD25+ regulatory T cells. J. Immunol. 170:3939–43

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

698

Lefran¸cois

·

Puddington

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

99. de Heer HJ, Hammad H, Soullie T, Hijdra D, Vos N, et al. 2004. Essential role of lung plasmacytoid dendritic cells in preventing asthmatic reactions to harmless inhaled antigen. J. Exp. Med. 200:89–98 100. Adler AJ, Huang CT, Yochum GS, Marsh DW, Pardoll DM. 2000. In vivo CD4+ T cell tolerance induction versus priming is independent of the rate and number of cell divisions. J. Immunol. 164:649–55 101. Higgins AD, Mihalyo MA, McGary PW, Adler AJ. 2002. CD4 cell priming and tolerization are differentially programmed by APCs upon initial engagement. J. Immunol. 168:5573–81 102. Bour-Jordan H, Salomon BL, Thompson HL, Szot GL, Bernhard MR, Bluestone JA. 2004. Costimulation controls diabetes by altering the balance of pathogenic and regulatory T cells. J. Clin. Invest. 114:979–87 103. Kubo T, Hatton RD, Oliver J, Liu X, Elson CO, Weaver CT. 2004. Regulatory T cell suppression and anergy are differentially regulated by proinﬂammatory cytokines produced by TLR-activated dendritic cells. J. Immunol. 173:7249–58 104. Kearney ER, Pape KA, Loh DY, Jenkins MK. 1994. Visualization of peptide-speciﬁc T cell immunity and peripheral tolerance induction in vivo. Immunity 1:327–39 105. Pape KA, Merica R, Mondino A, Khoruts A, Jenkins MK. 1998. Direct evidence that functionally impaired CD4+ T cells persist in vivo following induction of peripheral tolerance. J. Immunol. 160:4719–29 106. Hurst SD, Seymour BW, Muchamuel T, Kurup VP, Coffman RL. 2001. Modulation of inhaled antigen-induced IgE tolerance by ongoing Th2 responses in the lung. J. Immunol. 166:4922–30 107. Barton GM, Medzhitov R. 2002. Control of adaptive immune responses by Toll-like receptors. Curr. Opin. Immunol 14:380–83 108. Bendelac A, Medzhitov R. 2002. Adjuvants of immunity: harnessing innate immunity to promote adaptive immunity. J. Exp. Med. 195:F19–23 109. Eisenbarth SC, Piggott DA, Huleatt JW, Visintin I, Herrick CA, Bottomly K. 2002. Lipopolysaccharide-enhanced, toll-like receptor 4-dependent T helper cell type 2 responses to inhaled antigen. J. Exp. Med. 196:1645–51 110. Heath WR, Carbone FR. 2001. Cross-presentation, dendritic cells, tolerance and immunity. Annu. Rev. Immunol. 19:47–64 111. Zinselmeyer BH, Dempster J, Gurney AM, Wokosin D, Miller M, et al. 2005. In situ characterization of CD4+ T cell behavior in mucosal and systemic lymphoid tissues during the induction of oral priming and tolerance. J. Exp. Med. 201:1815–23 112. Vezys V, Olson S, Lefranc¸ois L. 2000. Expression of intestine-speciﬁc antigen reveals novel pathways of CD8 T cell tolerance induction. Immunity 12:505–14 113. Derbinski J, Schulte A, Kyewski B, Klein L. 2001. Promiscuous gene expression in medullary thymic epithelial cells mirrors the peripheral self. Nat. Immunol. 2:1032– 39 114. Kyewski B, Derbinski J, Gotter J, Klein L. 2002. Promiscuous gene expression and central T-cell tolerance: more than meets the eye. Trends Immunol. 23:364–71 115. Strober W, Fuss IJ, Blumberg RS. 2002. The immunology of mucosal models of inﬂammation. Annu. Rev. Immunol. 20:495–549 116. Vezys V, Lefranc¸ois L. 2002. Inﬂammatory signals drive organ-speciﬁc autoimmunity to normally cross-tolerizing endogenous antigen. J. Immunol. 169:6677–80 117. Herrick CA, Bottomly K. 2003. To respond or not to respond: T cells in allergic asthma. Nat. Rev. Immunol. 3:405–12 www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

699

ARI

15 February 2006

2:12

118. Dahl ME, Dabbagh K, Liggitt D, Kim S, Lewis DB. 2004. Viral-induced T helper type 1 responses enhance allergic disease by effects on lung dendritic cells. Nat. Immunol. 5:337–43 119. Sellon RK, Tonkonogy S, Schultz M, Dieleman LA, Grenther W, et al. 1998. Resident enteric bacteria are necessary for development of spontaneous colitis and immune system activation in interleukin-10-deﬁcient mice. Infect. Immun. 66:5224–31 120. Elson CO, Cong Y, McCracken VJ, Dimmitt RA, Lorenz RG, Weaver CT. 2005. Experimental models of inﬂammatory bowel disease reveal innate, adaptive, and regulatory mechanisms of host dialogue with the microbiota. Immunol. Rev. 206:260–76 121. Lodes MJ, Cong Y, Elson CO, Mohamath R, Landers CJ, et al. 2004. Bacterial ﬂagellin is a dominant antigen in Crohn disease. J. Clin. Invest. 113:1296–306 122. Tanaka T, Ebisuno Y, Kanemitsu N, Umemoto E, Yang BG, et al. 2004. Molecular determinants controlling homeostatic recirculation and tissue-speciﬁc trafﬁcking of lymphocytes. Int. Arch. Allergy Immunol. 134:120–34 123. Masopust D, Jiang J, Shen H, Lefranc¸ois L. 2001. Direct analysis of the dynamics of the intestinal mucosa CD8 T cell response to systemic virus infection. J. Immunol. 166:2348– 56 124. Haddad W, Cooper CJ, Zhang Z, Brown JB, Zhu Y, et al. 2003. P-selectin and p-selectin glycoprotein ligand 1 are major determinants for Th1 cell recruitment to nonlymphoid effector sites in the intestinal lamina propria. J. Exp. Med. 198:369–77 125. Berlin C, Berg EL, Briskin MJ, Andrew DP, Kilshaw PJ, et al. 1993. Alpha 4 beta 7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM-1. Cell 74:185–96 126. Denning TL, Kim G, Kronenberg M. 2005. CD4+ CD25+ regulatory T cells impaired for intestinal homing can prevent colitis. J. Immunol. 174:7487–91 127. Wagner N, Lohler J, Tedder TF, Rajewsky K, Muller W, Steeber DA. 1998. L-selectin and β7 integrin synergistically mediate lymphocyte migration to mesenteric lymph nodes. Eur. J. Immunol. 28:3832–39 128. Kilshaw PJ, Murant SJ. 1990. A new surface antigen on intraepithelial lymphocytes in the intestine. Eur. J. Immunol. 20:2201–7 129. Parker CM, Cepek KL, Russell GJ, Shaw SK, Posnett DN, et al. 1992. A family of β7 integrins on human mucosal lymphocytes. Proc. Natl. Acad. Sci. USA 89:1924–28 130. Lefranc¸ois L, Barrett TA, Havran WL, Puddington L. 1994. Developmental expression of the αIEL β7 integrin on T cell receptor γδ and T cell receptor αβ T cells. Eur. J. Immunol. 24:635–40 131. Cepek KL, Shaw SK, Parker CM, Russell GJ, Morrow JS, et al. 1994. Adhesion between epithelial cells and T lymphocytes mediated by E-cadherin and an integrin, αE β7 . Nature 372:190–93 132. Karecla PI, Bowden SJ, Green SJ, Kilshaw PJ. 1995. Recognition of E-cadherin on epithelial cells by the mucosal T cell integrin αM290 β7 (αEβ7). Eur. J. Immunol. 25:852–56 133. El Asady R, Yuan R, Liu K, Wang D, Gress RE, et al. 2005. TGFβ-dependent CD103 expression by CD8+ T cells promotes selective destruction of the host intestinal epithelium during graft-versus-host disease. J. Exp. Med. 201:1647–57 134. Kunkel EJ, Campbell JJ, Haraldsen G, Pan J, Boisvert J, et al. 2000. Lymphocyte CC chemokine receptor 9 and epithelial thymus-expressed chemokine (TECK) expression distinguish the small intestinal immune compartment: epithelial expression of tissue-speciﬁc chemokines as an organizing principle in regional immunity. J. Exp. Med. 192:761–68

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

700

Lefran¸cois

·

Puddington

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

135. Zabel BA, Agace WW, Campbell JJ, Heath HM, Parent D, et al. 1999. Human G protein-coupled receptor GPR-9–6/CC chemokine receptor 9 is selectively expressed on intestinal homing T lymphocytes, mucosal lymphocytes, and thymocytes and is required for thymus-expressed chemokine-mediated chemotaxis. J. Exp. Med. 190:1241– 56 136. Svensson M, Marsal J, Ericsson A, Carramolino L, Broden T, et al. 2002. CCL25 mediates the localization of recently activated CD8αβ+ lymphocytes to the small-intestinal mucosa. J. Clin. Invest 110:1113–21 137. Johansson-Lindbom B, Svensson M, Wurbel MA, Malissen B, Marquez G, Agace W. 2003. Selective generation of gut tropic T cells in gut-associated lymphoid tissue (GALT): requirement for GALT dendritic cells and adjuvant. J. Exp. Med. 198:963–69 138. Luster AD, Tager AM. 2004. T-cell trafﬁcking in asthma: lipid mediators grease the way. Nat. Rev. Immunol. 4:711–24 139. Erle DJ, Brown T, Christian D, Aris R. 1994. Lung epithelial lining ﬂuid T cell subsets deﬁned by distinct patterns of β7 and β1 integrin expression. Am. J. Respir. Cell Mol. Biol. 10:237–44 140. Lazaar AL, Albelda SM, Pilewski JM, Brennan B, Pure E, Panettieri RA Jr. 1994. T lymphocytes adhere to airway smooth muscle cells via integrins and CD44 and induce smooth muscle cell DNA synthesis. J. Exp. Med. 180:807–16 141. Lobb RR, Hemler ME. 1994. The pathophysiologic role of α4 integrins in vivo. J. Clin. Invest 94:1722–28 142. Kanehiro A, Takeda K, Joetham A, Tomkinson A, Ikemura T, et al. 2000. Timing of administration of anti-VLA-4 differentiates airway hyperresponsiveness in the central and peripheral airways in mice. Am. J. Respir. Crit. Care Med. 162:1132–39 143. Borchers MT, Crosby J, Farmer S, Sypek J, Ansay T, et al. 2001. Blockade of CD49d inhibits allergic airway pathologies independent of effects on leukocyte recruitment. Am. J. Physiol. Lung Cell Mol. Physiol. 280:L813–21 144. Ray SJ, Franki SN, Pierce RH, Dimitrova S, Koteliansky V, et al. 2004. The collagen binding α1β1 integrin VLA-1 regulates CD8 T cell-mediated immune protection against heterologous inﬂuenza infection. Immunity 20:167–79 145. Marshall DR, Olivas E, Andreansky S, La Gruta NL, Neale GA, et al. 2005. Effector CD8+ T cells recovered from an inﬂuenza pneumonia differentiate to a state of focused gene expression. Proc. Natl. Acad. Sci. USA 102:6074–79 146. Tager AM, Bromley SK, Medoff BD, Islam SA, Bercury SD, et al. 2003. Leukotriene B4 receptor BLT1 mediates early effector T cell recruitment. Nat. Immunol. 4:982–90 147. Deleted in proof 148. Cohn L, Homer RJ, Niu N, Bottomly K. 1999. T helper 1 cells and interferon gamma regulate allergic airway inﬂammation and mucus production. J. Exp. Med. 190:1309– 18 149. Das J, Chen CH, Yang L, Cohn L, Ray P, Ray A. 2001. A critical role for NF-kappa B in GATA3 expression and Th2 differentiation in allergic airway inﬂammation. Nat. Immunol. 2:45–50 150. Piggott DA, Eisenbarth SC, Xu L, Constant SL, Huleatt JW, et al. 2005. MyD88dependent induction of allergic Th2 responses to intranasal antigen. J. Clin. Invest. 115:459–67 151. Cohn L, Homer RJ, Marinov A, Rankin J, Bottomly K. 1997. Induction of airway mucus production by T helper 2 (Th2) cells: A critical role for interleukin 4 in cell recruitment but not mucus production. J. Exp. Med. 186:1737–47 www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

701

ARI

15 February 2006

2:12

152. Hansen G, Berry G, DeKruyff RH, Umetsu DT. 1999. Allergen-speciﬁc Th1 cells fail to counterbalance Th2 cell-induced airway hyperreactivity but cause severe airway inﬂammation. J. Clin. Invest. 103:175–83 153. Zhu Z, Homer RJ, Wang Z, Chen Q, Geba GP, et al. 1999. Pulmonary expression of interleukin-13 causes inﬂammation, mucus hypersecretion, subepithelial ﬁbrosis, physiologic abnormalities, and eotaxin production. J. Clin. Invest. 103:779–88 154. Temann UA, Ray P, Flavell RA. 2002. Pulmonary overexpression of IL-9 induces Th2 cytokine expression, leading to immune pathology. J. Clin. Invest. 109:29–39 155. Li XM, Serebrisky D, Lee SY, Huang CK, Bardina L, et al. 2000. A murine model of peanut anaphylaxis: T- and B-cell responses to a major peanut allergen mimic human responses. J. Allergy Clin. Immunol. 106:150–58 156. Mishra A, Hogan SP, Brandt EB, Rothenberg ME. 2000. Peyer’s patch eosinophils: identiﬁcation, characterization, and regulation by mucosal allergen exposure, interleukin5, and eotaxin. Blood 96:1538–44 157. Hino A, Kweon MN, Fujihashi K, Mcghee JR, Kiyono H. 2004. Pathological role of large intestinal IL-12p40 for the induction of Th2-type allergic diarrhea. Am. J. Pathol. 164:1327–35 158. Butcher EC, Picker LJ. 1996. Lymphocyte homing and homeostasis. Science 272:60– 66 159. Campbell DJ, Butcher EC. 2002. Rapid acquisition of tissue-speciﬁc homing phenotypes by CD4+ T cells activated in cutaneous or mucosal lymphoid tissues. J. Exp. Med. 195:135–41 160. Mora JR, Bono MR, Manjunath N, Weninger W, Cavanagh LL, et al. 2003. Selective imprinting of gut-homing T cells by Peyer’s patch dendritic cells. Nature 424:88–93 161. Iwata M, Hirakiyama A, Eshima Y, Kagechika H, Kato C, Song SY. 2004. Retinoic acid imprints gut-homing speciﬁcity on T cells. Immunity 21:527–38 162. Marelli-Berg FM, Jarmin SJ. 2004. Antigen presentation by the endothelium: a green light for antigen-speciﬁc T cell trafﬁcking? Immunol. Lett. 93:109–13 163. Springer TA. 1994. Trafﬁc signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 76:301–14 164. Lefranc¸ois L, Olson S, Masopust D. 1999. A critical role for CD40-CD40 ligand interactions in ampliﬁcation of the mucosal CD8 T cell response. J. Exp. Med. 190:1275–84 165. Laouar A, Haridas V, Vargas D, Zhinan X, Chaplin D, et al. 2005. CD70+ antigenpresenting cells control the proliferation and differentiation of T cells in the intestinal mucosa. Nat. Immunol. 6:698–706 166. Kim SK, Reed DS, Heath WR, Carbone F, Lefranc¸ois L. 1997. Activation and migration of CD8 T cells in the intestinal mucosa. J. Immunol. 159:4295–306 167. Kim SK, Schluns KS, Lefranc¸ois L. 1999. Induction and visualization of mucosal memory CD8 T cells following systemic virus infection. J. Immunol. 163:4125–32 168. Klonowski KD, Williams KJ, Marzo AL, Blair DA, Lingenheld EG, Lefrancois L. 2004. Dynamics of blood-borne CD8 memory T cell migration in vivo. Immunity 20:551–62 169. Coyle AJ, Erard F, Bertrand C, Walti S, Pircher H, Le Gros G. 1995. Virus-speciﬁc CD8+ cells can switch to interleukin 5 production and induce airway eosinophilia. J. Exp. Med. 181:1229–33 170. Wu CA, Puddington L, Whiteley HE, Yiamouyiannis CA, Schramm CM, et al. 2001. Murine cytomegalovirus infection alters Th1/Th2 cytokine expression, decreases airway eosinophilia, and enhances mucus production in allergic airway disease. J. Immunol. 167:2798–807

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

702

Lefran¸cois

·

Puddington

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-22

ARI

15 February 2006

2:12

171. Bashir ME, Andersen P, Fuss IJ, Shi HN, Nagler-Anderson C. 2002. An enteric helminth infection protects against an allergic response to dietary antigen. J. Immunol. 169:3284– 92 172. Prioult G, Nagler-Anderson C. 2005. Mucosal immunity and allergic responses: lack of regulation and/or lack of microbial stimulation? Immunol. Rev. 206:204–18 173. Tesciuba AG, Subudhi S, Rother RP, Faas SJ, Frantz AM, et al. 2001. Inducible costimulator regulates Th2-mediated inﬂammation, but not Th2 differentiation, in a model of allergic airway disease. J. Immunol. 167:1996–2003 174. Luettig B, Kaiser M, Bode U, Bell EB, Sparshott SM, et al. 2001. Naive and memory T cells migrate in comparable numbers through the normal rat lung: only effector T cells accumulate and proliferate in the lamina propria of the bronchi. Am. J. Respir. Cell Mol. Biol. 25:69–77 175. Arnold B, Schonrich G, Hammerling GJ. 1993. Multiple levels of peripheral tolerance. Immunol. Today 14:12–14 176. Wiley JA, Hogan RJ, Woodland DL, Harmsen AG. 2001. Antigen-speciﬁc CD8+ T cells persist in the upper respiratory tract following inﬂuenza virus infection. J. Immunol. 167:3293–99 177. Cauley LS, Cookenham T, Miller TB, Adams PS, Vignali KM, et al. 2002. Virus-speciﬁc CD4+ memory T cells in nonlymphoid tissues express a highly activated phenotype. J. Immunol. 169:6655–58 178. Stephens R, Randolph DA, Huang G, Holtzman MJ, Chaplin DD. 2002. Antigennonspeciﬁc recruitment of Th2 cells to the lung as a mechanism for viral infection-induced allergic asthma. J. Immunol. 169:5458–67 179. Ely KH, Cauley LS, Roberts AD, Brennan JW, Cookenham T, Woodland DL. 2003. Nonspeciﬁc recruitment of memory CD8+ T cells to the lung airways during respiratory virus infections. J. Immunol. 170:1423–29 180. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401:708–12 181. Masopust D, Vezys V, Marzo AL, Lefranc¸ois L. 2001. Preferential localization of effector memory cells in nonlymphoid tissue. Science 291:2413–17 182. Unsoeld H, Pircher H. 2005. Complex memory T-cell phenotypes revealed by coexpression of CD62L and CCR7. J. Virol. 79:4510–13 183. Reinhardt RL, Khoruts A, Merica R, Zell T, Jenkins MK. 2001. Visualizing the generation of memory CD4 T cells in the whole body. Nature 410:101–5 184. Baron V, Bouneaud C, Cumano A, Lim A, Arstila TP, et al. 2003. The repertoires of circulating human CD8+ central and effector memory T cell subsets are largely distinct. Immunity 18:193–204 185. Wherry EJ, Teichgraber V, Becker TC, Masopust D, Kaech SM, et al. 2003. Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat. Immunol. 4:225–34 186. Roberts AD, Ely KH, Woodland DL. 2005. Differential contributions of central and effector memory T cells to recall responses. J. Exp. Med. 202:123–33 187. Marzo AL, Klonowski KD, Le Bon A, Borrow P, Tough DF, Lefranc¸ois L. 2005. Initial T cell frequency dictates memory CD8+ T cell lineage commitment. Nat. Immunol. 6:793–99 188. Roberts AD, Woodland DL. 2004. Effector memory CD8+ T cells play a prominent role in recall responses to secondary viral infection in the lung. J. Immunol. 172:6533– 37 www.annualreviews.org • Intestinal and Pulmonary Mucosal T Cells

703

ANRV270-IY24-22

ARI

15 February 2006

2:12

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

189. Mora JR, Cheng G, Picarella D, Briskin M, Buchanan N, von Andrian UH. 2005. Reciprocal and dynamic control of CD8 T cell homing by dendritic cells from skin- and gut-associated lymphoid tissues. J. Exp. Med. 201:303–16 190. Bromley SK, Thomas SY, Luster AD. 2005. Chemokine receptor CCR7 guides T cell exit from peripheral tissues and entry into afferent lymphatics. Nat. Immunol. 6:895–901 191. Debes GF, Arnold CN, Young AJ, Krautwald S, Lipp M, et al. 2005. Chemokine receptor CCR7 required for T lymphocyte exit from peripheral tissues. Nat. Immunol. 6:889–94 192. Harris NL, Watt V, Ronchese F, Le Gros G. 2002. Differential T cell function and fate in lymph node and nonlymphoid tissues. J. Exp. Med. 195:317–26

704

Lefran¸cois

·

Puddington

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:681-704. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf Department of Developmental and Molecular Biology, Albert Einstein College of Medicine, Bronx, New York 10461; email: [email protected]

Annu. Rev. Immunol. 2006. 24:705–38 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090742 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0705$20.00

Key Words hematopoietic stem cell, hematopoietic lineage trees, transcription factors, lineage commitment, lineage priming

Abstract In recent years, investigators have made great progress in delineating developmental pathways of several lymphoid and myeloid lineages and in identifying transcription factors that establish and maintain their fate. However, the developmental branching points between these two large cell compartments are still controversial, and little is known about how their diversiﬁcation is induced. Here, we give an overview of determinants that play a role at lymphoid-myeloid junctures, in particular transcription factors and cytokine receptors. Experiments showing that myeloid lineages can be reversibly reprogrammed into one another by transcription factor network perturbations are used to highlight key principles of lineage commitment. We also discuss experiments showing that lymphoid-to-myeloid but not myeloid-to-lymphoid conversions can be induced by the enforced expression of a single transcription factor. We close by proposing that this asymmetry is related to a higher complexity of transcription factor networks in lymphoid cells compared with myeloid cells, and we suggest that this feature must be considered when searching for mechanisms by which hematopoietic stem cells become committed to lymphoid lineages.

705

ANRV270-IY24-23

ARI

23 February 2006

15:47

INTRODUCTION

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Lymphoid cells: B and T lymphocytes and their precursors; also include natural killer (NK) lineage cells Myeloid cells: all nonlymphoid cell types, including macrophages, granulocytes, erythrocytes, and megakaryocytes (precursors of platelets) Hematopoietic stem cell (HSC): cell capable of differentiation into all hematopoietic cell lineages and of self-renewal CLP: common lymphoid progenitor CMP: common myeloid progenitor GMP: granulocytemacrophage progenitor MEP: megakaryocyteerythrocyte progenitor

Hematopoiesis has traditionally been studied by two separate camps: immunologists predominantly involved with the development of lymphoid cells such as B, T, and natural killer (NK) cells; and hematologists interested in the formation of myeloid cells, which encompass monocytes, macrophages, and different classes of neutrophils as well as red blood cells and platelets. However, these camps have converged to some extent in recent years, and the need for such a merger is perhaps no more evident than when one tries to understand the molecular basis of lineage diversiﬁcation between lymphoid and myeloid cells. The establishment of all hematopoietic lineages during development is mediated by transcription factors that act in sequential and parallel fashions, building lineage-speciﬁc networks or circuits (for reviews, see 1–4). Logically, the lymphoid and myeloid networks intersect at the developmental branching points between these lineages. This review focuses on transcription factors and other parameters, such as cytokine receptor signaling, that have an instructive role at lymphoid-myeloid junctures. In the following section, we give an overview of several current models of hematopoietic lineage trees to pinpoint developmental branching points between the lymphoid and myeloid cell compartments.

LYMPHOID-MYELOID BRANCHING POINTS: TWILIGHT ZONES IN BLOOD CELL FORMATION The stochastic model of hematopoiesis states that a single multipotent progenitor (MPP) has the option to differentiate along more than two pathways. This early model was based on the observation that colony assays, using single myeloid progenitors, yielded highly heterogeneous outcomes (5). As discussed below, most current models imply that hematopoietic differentiation proceeds along an ordered pathway with binary deci706

Laiosa

·

Stadtfeld

·

Graf

sion steps. However, ordered binary choices are not apparent, at least during the earliest stages of differentiation.

The Akashi-Kondo-Weissman Scheme of Hematopoietic Differentiation The identiﬁcation of stem and progenitor cells by Weissman and collaborators (6–9, 13, 14) led to the construction of a hematopoietic lineage tree that is characterized by a cascade of binary decisions (Figure 1a). The staining of bone marrow, with a combination of cell surface antigenspeciﬁc antibodies led to the prospective isolation of hematopoietic stem cells (HSCs) as lin−/low Sca-1+ c-kit+ (LSK) cells (6–8). These cells can be further subdivided into long-term HSCs (Thy-1low Flk2/Flt3− ), short-term repopulating HSCs (Thy1low Flt3+ ), and MPPs (Thy-1− Flt3+ ) (9), populations that were also deﬁned by other combinations of markers (10–12). Similar approaches led to the identiﬁcation of progenitors with a more restricted differentiation potential. Thus, in the bone marrow investigators identiﬁed a common progenitor for all lymphoid lineages (CLP) (13), as well as a common myeloid progenitor (CMP) that generates granulocytic-macrophage (GM) and megakaryocytic-erythroid (MegE) lineages (14). CLPs give rise to pro-B and pro-T cells, uncommitted lymphoid progenitors that will differentiate further into mature B and T cells (13). CLPs also produce NK lineage cells (13). CMPs in turn generate two more restricted progenitors: granulocyte-macrophage progenitors (GMPs) and megakaryocyte-erythrocyte progenitors (MEPs), generating GM and MegE cells, respectively (14). The offspring of GMPs also includes neutrophils, eosinophils, and possibly basophils/mast cells (15, 16). The observation that CMPs and CLPs derived from adult bone marrow generate mutually exclusive progeny (13, 14) suggests that their diversiﬁcation represents the

ANRV270-IY24-23

ARI

23 February 2006

15:47

a pro-B

pro-T

CLP

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LT-HSC

ST-HSC

MEP

MPP

GMP

CMP

b B cell pro-B

pre-B NK cell

CLP T cell pro-T

pre-T

dendritic cells ETP macrophage HSC GMP CMP

granulocyte

megakaryocyte

MEP

erythrocyte

Figure 1 Lineage trees of adult hematopoiesis and lymphoid-myeloid branching points. (a) The Akashi-Kondo-Weissman model of adult hematopoiesis (13, 14) with the branching point between lymphoid and myeloid lineages indicated by the gray shaded circle. (b) Revised lineage tree, showing three areas where branching might occur (LT- and ST-HSC, long-term and short-term HSC; MPP, multipotent progenitor; ETP, early T lineage progenitor).

www.annualreviews.org • Lymphoid-Myeloid Development

707

ANRV270-IY24-23

ARI

23 February 2006

15:47

earliest branching point during hematopoietic differentiation. ETP: early T lineage progenitor

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

DN: CD4/CD8 double negative; DN1–DN4: stages of DN cell differentiation Lineage commitment: the process by which cell fates are determined

Adult Lympho-Myeloid Progenitors Lacking Erythroid and Megakaryocytic Potentials? Although the Weissman scheme has proven invaluable for the conceptual understanding of hematopoiesis, more recent experimental data suggest that alternative developmental pathways generating myeloid and lymphoid cells exist (Figure 1b). The recent description of macrophage–/T cell–/B cell– restricted progenitors lacking MegE potential by Jacobsen and colleagues (17) is particularly intriguing. These cells resemble previously described MPPs because they are contained within the LSK population and express high levels of Flt3. The model proposed by Jacobsen and colleagues implies that GM cells can be generated by two distinct pathways: one via classical CMPs and one via lymphomyelomonocytic progenitors without MegE potential. An early branching between MegE and lymphoid-myelomonocytic cells might reﬂect the choices of an ancestral HSC because lymphocytes are evolutionarily younger than GM and MegE cells (18). However, in a reassessment of the differentiation potential of Flt3+ MPPs, a transient MegE potential was detectable both in culture and after transplantation (C.E. Forsberg, T. Serwold, S. Kogan, I.L. Weissman & E. Passegue, personal communication).

Uncertainty About the Physiology and the Lineage Restriction of T Cell Progenitors Unlike all other lineages, which are speciﬁed in the bone marrow, T cells differentiate after migration of early progenitors into the thymus. Bone marrow–derived CLPs can differentiate into T lymphocytes in fetal thymic organ culture (FTOC) and after intrathymic injection (19). Another clonogenic B/T bipotent progenitor termed CLP-2, which coex708

Laiosa

·

Stadtfeld

·

Graf

presses pre-Tα (a T lineage marker) and B220 (a B lineage marker), was later identiﬁed in bone marrow and shown to seed the thymus after intravenous transfer (19a). However, the physiological relevance of the CLP pathway has recently been called into question by the ﬁnding of Bhandoola and colleagues (20) that the earliest T lineage progenitors (ETPs), contained within the CD44+ CD25− (DN1) fraction of CD4/CD8 double-negative (DN) thymocytes, have a more robust T cell reconstitution capacity than CLPs. The same study showed that mice deﬁcient for the Ikaros transcription factor lack CLPs but not ETPs (20), although it cannot be ruled out that in these mice CLPs are present but lack the expression of one of the markers that serve to identify them. Although ETPs also generate B cells with delayed kinetics, they are not lymphoid restricted because they can generate myeloid cells in vitro at low frequencies (20). Later, the same authors also reported that CLPs are absent from peripheral blood and that the sole population with T lineage potential has an HSC phenotype, again suggesting a CLPindependent pathway for T cell differentiation (21). Together, these results suggest that T cells can develop by at least two distinct pathways, CLPs and ETPs (22). If HSCs are the thymus-seeding cells, the question arises as to when they lose their myeloid and B lymphoid potential. That lineage commitment occurs at the pro-T (DN1/DN2) to pre-T (DN3/DN4) transition is well established (23). Several reports indicate that pro-T cells can form NK and dendritic cells at the single-cell level (24, 25), although their B cell potential is more controversial. One report describes the existence of bipotent T/B progenitors in the thymus and conﬁrmed their identity in single-cell assays (25). However, two other studies failed to detect B cell potential in thymic progenitors with robust T lineage potential (26, 27). Using limiting dilution assays, one of these studies reported the generation of Mac-1+ F4/80+ myeloid cells from a thymic progenitor without B cell potential (27), suggesting that a

ANRV270-IY24-23

ARI

23 February 2006

15:47

general lymphoid-myeloid branching point does not exist during the generation of T cells, and that B cell potential is lost before myeloid potential. Thus, the thymic microenvironment appears to suppress alternative fates in an ordered fashion (28).

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Multiple Origins of Dendritic Cells Dendritic cells can be broadly classiﬁed as Mac-1/CD11b+ “myeloid” dendritic cells and CD8α+ “lymphoid” dendritic cells (29). Specialized subtypes of dendritic cells also exist, including Langerhans dendritic cells found in the skin and plasmacytoid dendritic cells (29). Curiously, functionally equivalent and phenotypically indistinguishable myeloid and lymphoid dendritic cells can be derived from either CMPs or CLPs (30, 31). In fact, progenitors downstream of CMPs and CLPs such as GMPs and pro-T cells but not pro-B cells can still give rise to these two classes of dendritic cells (31). Although molecular differences between CMP- and CLP-derived dendritic cells might exist, their dual origin suggests that myeloid and lymphoid progenitors can converge into the same phenotype and therefore do not have mutually exclusive differentiation potentials.

Lineage Trees in the Fetal Liver: Dissent and Consensus To analyze the differentiation of single fetal liver–derived LSK cells, Katsura and colleagues (32–34) used an in vitro culture system consisting of a modiﬁed FTOC, which measured macrophage (M), erythroid (E), B cell, and T cell differentiation potentials simultaneously. They found progenitors with various potentials, including M/E/T/B, M/E, M/T/B, M/T, and M/B, as well as monopotent progenitors (32–34). However, T/B progenitors were never found, even when the fraction corresponding to bone marrow CLPs (lin− c-kitlo IL-7Rα+ ) was examined (34). These observations challenged the notions that CLPs exist in the fetal liver and

that hematopoietic commitment must entail an early myeloid-lymphoid branching point (Figure 2a). However, because researchers could not deﬁne the phenotype of the individual progenitor subtypes, these cells could not be prospectively puriﬁed and tested in other assays. Akashi and colleagues (35, 36) also sought to identify intermediate progenitors in the fetal liver and found populations with similar cell surface characteristics as bone marrow CLPs and CMPs. In addition to the expected progeny of these cells, limiting dilution analysis in vitro showed that 1 in 14 CLPs generated macrophages but not MegE lineage cells (35). Fetal CMPs gave rise to B cells, but not T cells, at a frequency of 1 in 160 cells (36) (Figure 2b). Because the former study did not simultaneously test the B, T, and myeloid potentials of fetal CLPs at the single-cell level, questions as to whether the myeloid potential in this fraction segregates with T or B potential and whether T/B progenitors exist could not be addressed. An open question therefore remains whether the fetal liver M/T/B progenitors (34) are equivalent to the low-ﬁdelity CLPs (35). Nevertheless, there is a consensus that the biological properties of the intermediate fetal progenitors differ from those in the bone marrow, implying that fetal and adult lymphoid-myeloid branching points are not equivalent.

Intermediate progenitor: a cell capable of differentiation into a number of deﬁned lineages but without self-renewal potential Cytokines: growth factors secreted by cells that bind to speciﬁc receptors, induce cell signaling, and result in survival, growth, and differentiation

DETERMINANTS OF CELL FATE DECISIONS IN THE HEMATOPOIETIC SYSTEM Lineage commitment could be induced either by extracellular factors, including cytokines, direct cell-cell interactions, or other environmental cues. Alternatively, it could be induced by intrinsic mechanisms, such as the stochastic upregulation of transcription factors, or other regulatory molecules, such as microRNAs (see MicroRNAs: An Emerging Group of Putative Lineage Determinants). Both extrinsic and intrinsic factors may either have an instructive role and actively induce commitment and differentiation or be merely permissive for www.annualreviews.org • Lymphoid-Myeloid Development

709

ANRV270-IY24-23

ARI

23 February 2006

15:47

a macrophage

M/T

HSC

T cell

macrophage

M/T/B

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B cell M/B macrophage

M/E Figure 2 Lineage trees of fetal hematopoiesis and lymphoid-myeloid branching points. (a) Lineage tree proposed by Katsura (32), showing two separate and late branching points between lymphoid and myeloid lineages. (b) Model based on the work by Akashi and collaborators (35, 36), in which CMPs and CLPs retain plasticity. Branching points are indicated by gray shaded circles.

b B cell NK cell CLP

T cell

macrophage

HSC

granulocyte GMP CMP

megakaryocyte MEP

the outgrowth of precommitted progenitors by promoting cell survival and/or expansion (37, 38). In fact, evidence discussed below suggests that both cytokine receptor signaling and transcription factors have instructive roles in a cell context–dependent manner. 710

erythrocyte

Laiosa

·

Stadtfeld

·

Graf

erythrocyte

CELL CONTEXT–DEPENDENT INSTRUCTIVE ROLES OF CYTOKINE RECEPTOR SIGNALING At ﬁrst glance, the phenotypes of mouse models deﬁcient for different myeloid cytokines and cytokine receptors suggest a permissive

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

role for cytokine signaling, as these mice typically do not display a complete loss of the associated lineage. For example, mice deﬁcient for the genes encoding G-CSF, GMCSF, or EPO display minor to signiﬁcant reductions in numbers of progenitors and mature progeny (39–41). At least in myeloid differentiation, for which multiple receptors have been identiﬁed, this could be attributed in part to overlapping patterns of expression of functionally redundant receptors. However, the phenotype of mice deﬁcient for both G-CSF and GM-CSF or M-CSF and GM-CSF argues against this explanation because these mice did not develop a more severe phenotype than G-CSF- or M-CSFdeﬁcient mice (42, 43). Observations gained from studies in which the EPO receptor (EpoR) and M-CSF/CSF-1 receptor (c-fms) were overexpressed also support the permissive model. Thus, expression of constitutively active EpoR on MPPs derived from bone marrow did not skew differentiation toward the erythroid lineage (44). Likewise, expression of CSF-1R in MPPs was compatible with differentiation into megakaryocytes and erythroid cells (45). Additional experiments tested the instructive capacity of a chimeric receptor consisting of the extracellular domain of the TPO receptor (TpoR or c-mpl) and the intracellular signaling domain of the G-CSF receptor (G-CSFR), expressed from the mpl genetic locus. Expression of this molecule rescued the thrombocytopenic phenotype of mpl-deﬁcient mice without promoting granulopoieisis (46). This result indicates that GCSFR can substitute for mpl signaling and does not speciﬁcally instruct granulocyte differentiation, but rather provides a generic survival function. Using a similar approach, the signaling component of the EpoR in a chimera with the ligand-binding domain of G-CSFR did not redirect differentiation toward the erythroid lineage nor disturb the formation of granulocytes when activated by G-CSF (47). Despite these results supporting the permissive model, ectopic cytokine signaling can act in an instructive fashion and is capable

MicroRNAs: AN EMERGING GROUP OF PUTATIVE LINEAGE DETERMINANTS MicroRNAs are small regulatory RNA molecules that bind target sequences in messenger RNAs and inhibit their expression either by inducing their degradation or by inhibiting their translation (182). The ﬁnding that microRNAs play a role in differentiation and cell growth in plants and in invertebrates (183) suggested similar functions in mammals. Indeed, miR-1 expression has been found to be activated during myocyte differentiation by the transcription factors SRF, MyoD, and MEF2, and this RNA in turn negatively regulates the transcription factor Hand2. As a result, proliferation of cardiogenic precursors is inhibited, reducing the number of ventricular cardiomyocytes (184). About 100 microRNAs have been found so far that are expressed in the hematopoietic system, and their combinations are highly diagnostic for different cell lineages and types of leukemia (185). miR-181 is most abundant in B lineage cells. When multipotent hematopoietic progenitors overexpressing the microRNA were transplanted into irradiated mice, an increase of B cells relative to T cells was observed (186). This ﬁnding raised the possibility that miR-181 targets genes that are critical for the decision making between lymphoid lineages. However, the effects observed may reﬂect an increased survival/proliferation of a subset of precommitted MPPs expressing the microRNA.

of reprogramming restricted progenitors in certain cellular contexts. One study described the effects of expressing human CSF-1R in murine pre-B cell lines. B cell factor– independent sub-lines arose after prolonged culture that spontaneously gave rise to macrophage-like cells. These cells lost all markers characteristic of the B cell lineage and could be expanded in human (but not mouse) M-CSF (48). More recent studies using primary progenitors derived from transgenic mice have extended the instructive role of cytokine signaling to nonimmortalized progenitors (Figure 3). In an attempt to induce NK cell differentiation in early lymphoid progenitors, Kondo et al. (19) analyzed the differentiation potential of bone marrow CLPs derived from transgenic mice expressing the β chain of the human IL-2 receptor (IL-2Rβ). www.annualreviews.org • Lymphoid-Myeloid Development

G-CSF: granulocyte-colony stimulating factor GM-CSF: granulocytemacrophage-colony stimulating factor EPO: erythropoietin M-CSF: macrophage-colony stimulating factor TPO: thrombopoietin

711

ANRV270-IY24-23

ARI

23 February 2006

15:47

IL-2R IL-2

CLP or pro-T

GM cell

GM-CSFR

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

GM-CSF

CLP or pro-T

dendritic cell

Figure 3 Reprogramming of early lymphoid cells by instructive cytokine signaling. Scheme illustrating how exogenous expression of human IL-2R followed by ligand stimulation converts CLPs and pro-T cells into myelomonocytic cells (top). IL-2R signaling leads to upregulation of endogenous GM-CSFRα, which likely contributes to myelomonocytic differentiation in this context (13, 50). As illustrated at the bottom, ectopic GM-CSFR expression and stimulation is sufﬁcient for this conversion and also converts T cell progenitors into myeloid dendritic cells (49).

Surprisingly, a signiﬁcant proportion of IL-2Rβ-expressing CLPs formed GM colonies after culture in medium containing human IL-2, whereas wild-type CLPs formed exclusively B cell colonies. IL-2R signaling did not serve solely as a survival signal for myelomonocytic progenitors because expression of Bcl-2 in CLPs did not induce the formation of GM colonies. The effects of IL2R signaling might be mediated in part by the induction of the GM-CSFRα gene because it was upregulated in stimulated IL-2R transgenic CLPs (19). Ectopic expression of both components of the human GM-CSFR (α and βc chains) in CLPs, followed by human GM-CSF stimulation, induced a similar CLP to GM conversion, as well as the formation of myeloid dendritic cells (19, 49) (Figure 3). Myeloid-instructive effects of IL-2R and GM-CSFR signaling could also be observed with pro-T cells, but not with pre-T cells or pro-B cells (49, 50). Strikingly, the instructive effect of GM-CSFR signaling is cell context dependent, as transgenic MEPs expressing hGM-CSFR were undisturbed by hGM-CSF, 712

Laiosa

·

Stadtfeld

·

Graf

indicating that GM-CSFR signaling is permissive for commitment of the MegE lineages (49). Moreover, GM-CSF signaling could not rescue defects in B and T cell differentiation caused by disruption of lymphoid cytokine IL-7 signaling (49), indicating that it is not a generic survival signal. Together, these experiments show that the IL-2 and GMCSF receptors can have lineage-instructive potential and suggest that early lymphoid and pro-T cell progenitors exhibit a latent myeloid developmental capacity. They also indicate that GM-CSFR, which is known to be expressed on a subset of HSCs, must be downregulated to permit commitment of MPPs into lymphoid precursors. Although multiple myeloid cytokine receptors have been identiﬁed, a smaller repertoire of receptors has been described for lymphoid cells. Flt3 is ﬁrst expressed on two subsets of LSK cells, short-term HSCs and MPPs that have lost the ability to self-renew (51). Flt3-expressing MPPs rapidly differentiate into B and T lineage cells, and mice deﬁcient in Flt3 or its ligand Flt3L have reduced numbers of B and T progenitors (51–53). Importantly, CLPs, but not HSCs or CMPs, are severely decreased in these animals, suggesting that Flt3 signaling fosters formation of downstream lymphoid progenitors (53). The IL-7Rα chain is expressed on CLPs as well as early B cells and T cell progenitors, and IL-7 stimulates their proliferation. Analysis of mice deﬁcient in IL-7 or one subunit of its receptor (IL-7Rα) revealed severe defects in B and T lymphoid differentiation, implicating IL-7 as a nonredundant cytokine (54, 55). IL-7 functions, at least in part, by inducing expression of Bcl-2, thereby promoting survival of lymphoid progenitors (56, 57). However, IL-7 also inﬂuences differentiation of CLPs because CLPs from IL-7-deﬁcient mice differentiate into T and NK cells but not B cells (58). Interestingly, mice lacking both IL-7R and Flt3 show a total absence of B cells in both fetal and adult development (59), raising the possibility that they act in an instructive fashion, perhaps by activating B cell transcription

ANRV270-IY24-23

ARI

23 February 2006

15:47

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

factors (2). In support of this hypothesis, enforced expression of early B cell factor (EBF) in CLPs derived from IL-7-deﬁcient mice restores B cell differentiation potential (58), and IL-7 stimulation of pre-pro-B cells derived from these mice rapidly induces expression of EBF (58a).

TRANSCRIPTION FACTORS INVOLVED IN LYMPHOID AND MYELOID DIFFERENTIATION Whether lineage decisions are induced by extracellular cues, by intrinsic events, or by a combination of both, they always involve changes in gene expression programs. Because these programs are ultimately controlled by transcription factors, new cell fates must be induced by changes in the expression or activity of these factors. Before discussing experiments showing how transcription factors both activate novel gene expression programs and extinguish existing ones in a lineageinstructive fashion, we give a brief overview about transcriptional regulators with known or suspected roles in lineage commitment processes.

Transcriptional Regulators Required for Myeloid and Lymphoid Development Figure 4 compiles a list of regulators whose ablation affects the formation of intermediate hematopoietic progenitors or inhibits differentiation or function of mature cells from diverse hematopoietic lineages. The ﬁgure does not include factors that act primarily on HSCs. Five types of phenotypes, observed in either conventional or conditional knockout studies, are described: (a) complete loss of a progenitor or lineage, (b) maturational or differentiation block, (c) functional defect, and (d ) decreased or (e) increased numbers of cells in a given lineage. The reader is also referred to more detailed reviews on transcription factors in the B cell lineage (2, 60, 61), T cell

lineage (23, 62, 63), and myeloid lineages (4, 64).

Early Lymphoid and Myeloid Progenitors Ikaros and PU.1 are broadly expressed in the hematopoietic system, including HSCs, early lymphoid progenitors such as CLPs, and various myeloid lineages. Loss of either factor disrupts B cell differentiation from uncommitted progenitors (65–68). Ikaros regulates transcription by recruiting co-repressors and chromatin remodeling proteins to target genes (69). Although Ikaros deﬁciency most severely affects the lymphoid lineages, alterations can also be observed in myeloid lineages (66, 69, 70). Ikaros may promote lymphoid cell fates by activation of Flt3 expression and repression of GM-CSFRα expression in HSCs (71). How Ikaros acts in a lineage-speciﬁc fashion is not well understood but might be explained by the expression of different splice isoforms in different lineages and cooperation with the lymphoid-restricted family member Aiolos (72–74). Loss of PU.1 causes a profound inhibition of B lineage and myelomonocytic cell formation as well as defects in T cell and dendritic cell differentiation (67, 68, 75–77). Some of the broad effects of PU.1 deletion can be attributed to a loss of CMPs and CLPs in these mice (78, 79). In addition, although HSCs are formed in PU.1-deﬁcient mice, and MegE cells are still produced, PU.1 deletion in conditional knockouts reduces the competitive repopulation potential of HSCs (78). C/EBPα, whose lineage-associated phenotype is discussed below, has the opposite effect in that its loss enhances their repopulation potential (80).

B Cell Lineage B cell development requires a complex set of transcription factors, namely PU.1, Ikaros, EBF, E2A, and Pax5, and inactivation of any of these factors yields a severe phenotype. The dramatic impairment of B cell development in www.annualreviews.org • Lymphoid-Myeloid Development

713

ANRV270-IY24-23

ARI

23 February 2006

15:47

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

PU.1-deﬁcient mice has been ascribed to lack of expression of the PU.1 target genes IL-7Rα and EBF (81, 82) because B cell differentiation is impaired in mice deﬁcient for either of these genes (83, 84). However, PU.1 is not strictly required for B cell formation because culture of PU.1-deﬁcient fetal liver cells led

to an outgrowth of IL-7-dependent B lineage cells that express both IL-7Rα and EBF and that can mature into IgM+ cells (84a). This outgrowth was delayed compared with that of wild-type cells, suggesting that PU.1 facilitates B cell lineage speciﬁcation in conjunction with other factors such as Ikaros (2). Mice

Regulator

Family

HSC

B

T

NK

GM

MegE

DC

References

Ikaros

ZnF

func

lack

lack/ matur

lack

matur

decr

lack

65, 66, 70, 71, 138

PU.1

ets

func

lack

67, 68, 75, 76, 77, 78, 115

EBF

HLH

matur

83

Pax5

paired

matur

93, 95, 164, 165

E2A

HLH

matur matur

84, 85, 111

HEB

HLH

matur

GATA-3

ZnF

lack

99, 100, 101, 102

Notch1

transmem

lack

98

Id3

HLH

C/EBPα

bZip

C/EBPβ

bZip

GATA-1

ZnF

FOG-1

ZnF

Id2

HLH

RelB

RHD

714

Laiosa

lack/ lack/ decr/ matur matur matur

func

110, 111

matur

118, 119

func decr/ func

matur

decr

func

·

Stadtfeld

·

Graf

lack

lack

80, 122

func

124, 125 lack/ matur

126, 127, 128

lack/ matur

130, 131, 132

decr

116, 117, 137

decr

133, 134, 135, 136

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

in which PU.1 is conditionally inactivated in the bone marrow lack CLPs despite normal expression of IL-7Rα on lin− cells (78). Loss of EBF blocks B cell development at the pro-B cell stage before initiation of B cell receptor (BCR) rearrangements (83). A similar phenotype was described for E2Adeﬁcient mice (85) and for transgenic mice overexpressing Id1, an inhibitor of E2A and other helix-loop-helix (HLH) proteins (86). E2A forms homodimers in B cells and collaborates with EBF at target gene promoters (87, 88). It probably also plays a role in growth control because expression of the HLH inhibitor Id3 in pro-B cells limits their proliferation in vitro (89). Target genes of E2A encode proteins necessary for BCR rearrangement, assembly, and signaling (63), a fact that explains the observed block of differentiation at the pro-B cell stage. EBF appears to be upstream of Pax5 because it binds to the Pax5 promoter and activates its expression in reporter assays (90, 91). However, the regulatory network of B cells clearly involves both the sequential and concerted action of transcription factors. This is exempliﬁed by the mb-1 gene, which is regulated by E2A and EBF, as well as by Pax5 in cooperation with Ets-1 (92). Pax5−/− mice display a block of B cell differentiation at the pro-B cell stage, just after initiation of BCR heavy chain rearrangements

(93). Pax5−/− pro-B cells express both EBF and E2A (94), demonstrating that neither of these factors nor their combination is sufﬁcient for B cell maturation. Most interestingly, these cells also express myeloid genes, such as M-CSFR, G-CSFR, and GM-CSFRα, as well as the T cell transcription factor Notch1, indicating that an important role of Pax5 is to repress lineage-inappropriate gene expression (95–97). E2A−/− pro-B cells also inappropriately express myeloid genes. The relevance of these ﬁndings with regard to lineage commitment is discussed below.

T Cell Lineage In mice lacking either GATA-3 or Notch1, T cell development is arrested at the earliest discernable T cell progenitor stage (98, 99). Chimeras between GATA-3-deﬁcient embryonic stem cells and wild-type blastocysts showed contributions of knockout cells to all hematopoietic lineages except T cells (99, 100). Besides a role in T cell speciﬁcation, conditional knockouts revealed functions of GATA-3 at various stages of T cell maturation (101, 102). Notch1 is a transmembrane receptor that, after ligand binding and cleavage, translocates to the nucleus and regulates transcription by converting its nuclear cofactor CSL into a transcriptional activator. Notch1 inactivation or that of CSL leads

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 4 Hematopoietic cell phenotypes of mice lacking transcriptional regulators. The ﬁgure lists transcription factors with proven or suspected roles in lineage commitment. Shaded boxes indicate altered phenotypes of the cell types indicated above, as observed after conventional (germ line) and conditional (Mx1-Cre or lineage-speciﬁc-Cre) gene inactivation of each regulator. Lack of shading indicates that no phenotype was observed or, in some cases, that defects were not studied. The altered phenotypes were either a complete loss of a lineage (lack), a maturational block (matur), a functional defect ( func), decreased numbers of lineage cells (decr), or increased numbers of lineage cells (incr). For PU.1, results from conventional and conditional knockout differ as to whether the factor is required in B cell development. For both Ikaros and PU.1, null phenotypes completely lack fetal but not adult T cells. FOG-1-deﬁcient embryos lack megakaryocytes and display a maturational defect in erythroid cells. GATA-1-deﬁcient embryos lack erythroid cells and display a maturational defect in megakaryocytes. Relevant references are listed in the right column. Abbreviations: ZnF, zinc ﬁnger domain; HTH, helix-turn-helix domain; HLH, helix-loop-helix domain; transmem, transmembrane; HMG box, high motility group box; bZip, basic leucine zipper; RHD, Rel homology domain. Lineage abbreviations are as used throughout the text, with DC indicating dendritic cells. www.annualreviews.org • Lymphoid-Myeloid Development

715

ARI

23 February 2006

15:47

to an early block in T cell development (98, 103). Conversely, expression of constitutively active Notch1 in bone marrow chimeras induces T cell differentiation in the bone marrow at the expense of B cell differentiation (104). Whether Notch1 signaling specifically instructs T cell lineage commitment or rather promotes survival/proliferation of precommitted progenitors remains a matter of contention. Expression of activated Notch1 in fetal liver–derived hematopoietic progenitors or their culture on OP9 stromal cells that constitutively express Notch ligand delta-like-1 (DL1) induces T cell differentiation and inhibits differentiation into alternative lineages such as B and NK cells (105, 106). High Notch signaling strengths are required to exclude differentiation into NK cells, whereas weak Notch signals are sufﬁcient to inhibit B cell differentiation (106). Nevertheless, when hematopoietic progenitor cells are cultured in T permissive culture conditions and then switched to B permissive conditions, some cells still retain B potential, indicating that Notch is not sufﬁcient to drive irreversible commitment (107). E2A forms heterodimers with HEB at all stages of thymocyte development. The complex activates the transcription of genes associated with T cell receptor (TCR) rearrangement and signaling (108). In the absence of either E2A or HEB, T cell differentiation is partially blocked (109, 110). One HLH factor likely compensates for the other, as mice expressing a dominant-negative form of HEB (which sequesters E2A or HEB into inactive heterodimers) show a more severe T cell phenotype than mice that lack either factor alone (111). Moreover, a role for E2A in commitment is inferred from experiments in which expression of Id3 in progenitors diverts T cell differentiation into the NK lineage (112). Although GATA-3, Notch1, and E2A/ HEB are all expressed in pro-T cells, T cell commitment does not occur until the preT cell stage, when all other developmental options are exhausted (23). This fact suggests that T cell commitment requires an additional

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

716

Laiosa

·

Stadtfeld

·

Graf

event, perhaps involving the downregulation of PU.1, which occurs at the transition between pro-T and pre-T cells (23, 113, 114). The observation that overexpression of PU.1 in hematopoietic progenitors limits their expansion in FTOC and promotes myeloid differentiation at the expense of T cell differentiation supports this suggestion (114). However, additional positive or negative regulators may exist that play a role in this transition.

Natural Killer Cell Lineage Ikaros-, PU.1-, and Id2-deﬁcient mice all possess severe NK lineage defects. In Ikarosdeﬁcient mice, T lymphocytes but not NK cells are produced in adult animals, suggesting an NK-speciﬁc defect (66). PU.1deﬁcient mice display a more subtle phenotype, with reduced numbers of NK cells that also have functional defects (115). In Id2deﬁcient mice, NK cells in the periphery are severely reduced (116, 117). An instructive role of Id proteins in NK cell development is also implied by experiments in which Id3 expression in human CD34+ hematopoietic progenitors promoted NK cell development at the expense of T cell development (112). In this case, Id3 likely mimics the function of Id2 in the physiological context because mice lacking Id3 do not exhibit a NK-speciﬁc phenotype (C. Murre, personal communication). Id3-deﬁcient mice also display defects in T and B cell maturation (118, 119).

Granulocytic-Macrophage Lineages Many macrophage- and granulocyte-restricted promoters are regulated by PU.1 and/or C/EBPα (120). These factors cooperate in the regulation of the genes encoding the myeloid growth factor receptors MCSFR, G-CSFR, and GM-CSFR (64, 120). In PU.1-deﬁcient mice, all myelomonocytic cells are absent. However, as in the B cell lineage, PU.1 is not strictly required for commitment because immature myeloid precursor

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

lines dependent on IL-3 can be established from PU.1-deﬁcient fetal liver (121). Conditional deletion of PU.1 in adult bone marrow, using the Mx1-Cre deleter system, leads to a complete loss of CMPs and GMPs (78, 79). Mice deﬁcient in C/EBPα lack neutrophil and eosinophil granulocytes, and conditional inactivation in the bone marrow shows the speciﬁc absence of GMPs and reduced numbers of CMPs, leading to decreased formation of all downstream lineages (80, 122). In colony assays, the remaining CMPs yielded normal MegE colonies and GM colonies containing cells with an immature phenotype (80). Deletion of C/EBPα by retroviral expression of Cre recombinase in GMPs does not prevent their terminal differentiation (80), showing that this factor is not needed at late stages, perhaps because its function is replaced by that of other C/EBP family members. The functions of C/EBPα in granulocyte formation appear to be redundant with its close relative C/EBPβ because C/EBPβ expressed from the C/EBPα locus rescues neutrophil granulocyte development (123). Macrophages from C/EBPβ-deﬁcient mice develop normally but display defective bacterial cell killing (124). Unlike C/EBPα, C/EBPβ is also expressed in B cells, and C/EBPβ-deﬁcient mice display defects in B cell expansion (125).

Megakaryocytic and Erythroid Lineages In GATA-1-deﬁcient embryos, development of erythroid cells is blocked early in differentiation, leading to a lethal anemia (126, 127). Expression of both GATA-2 and GATA3 transgenes rescued the erythroid lineage defect in GATA-1-deﬁcient mice but also revealed nonredundant roles of GATA-1 in erythroid maturation (128). GATA-1 is also required for the maturation of the megakaryocytic lineage, a ﬁnding that was only made after the gene was speciﬁcally inactivated in megakaryocytes (129). GATA-1 collaborates with FOG-1 (friend of GATA-1) during ery-

thropoiesis, and loss of FOG-1 leads to a partial block in erythroid differentiation and a complete loss of megakaryocytes (130, 131). The more severe megakaryocyte phenotype of FOG-1-deﬁcient compared with GATA1-deﬁcient mice suggests that FOG-1 has GATA-1-independent functions in this lineage. FOG-1 expression is not completely restricted to megakaryocytes and erythroid cells but is also present in late stages of T cell development, when it negatively regulates GATA-3 (132).

Transcription factor network: repertoire of transcription factors that interact both positively and negatively to establish and maintain gene expression programs

Dendritic Cell Lineages In a number of transcription factor knockout mice, dendritic subtypes are decreased in number or completely absent. Besides minor defects in B cell maturation (133), RelB knockouts speciﬁcally lack Mac-1/CD11b+ dendritic cells (134–136). In contrast, Id2deﬁcient mice lack CD8α+ dendritic cells as well as Langerhans cells, and CD11b+ dendritic cells are also reduced in these mice (137). Mice carrying a dominant-negative form of Ikaros lack both types of dendritic cells (138). Finally, PU.1-deﬁcient mice lack Mac-1/CD11b+ dendritic cells, although there are conﬂicting reports as to whether CD8α+ dendritic cells are also affected (76, 77).

MOLECULAR MECHANISMS OF CELL FATE DETERMINATION BY TRANSCRIPTION FACTORS As is evident from the above section, some transcription factors are required in multiple cell types, and more than one transcription factor is required to specify a given lineage. These requirements suggest that each lineage is deﬁned by a speciﬁc transcription factor combination and, as discussed below, by speciﬁc regulatory interactions, comprising the transcription factor network. The ultimate proof for the instructive capacity of a transcription factor is the demonstration

www.annualreviews.org • Lymphoid-Myeloid Development

717

ANRV270-IY24-23

ARI

23 February 2006

15:47

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

THE TRANSCRIPTION FACTOR–CHROMATIN CONNECTION Transcription factors activate genes not only by recruiting coactivators to the basic transcriptional machinery, but also by altering chromatin through the formation of complexes with enzymes that modify and remodel chromatin (187). In addition, they recruit co-repressors and chromatin-modifying factors when they repress genes. In particular, the activating and repressive functions of a number of hematopoietic factors, including Ikaros, PU.1, Runx1/AML1, C/EBPα, Pax5, and GATA-3, have been linked to changes in chromatin architecture (69, 141d, 188–192). Speciﬁcally, at the IgH locus, Pax5 alters histone methylation (188), whereas PU.1 modulates chromatin accessibility of IgH enhancers (189). The actions of these transcription factors at speciﬁc loci are developmentally regulated. A well-studied example is the M-CSFR (c-fms) locus. This gene is expressed in multilineage progenitors, where it is in an active state accessible to nucleases and bound by transcription factors (which include PU.1 and C/EBP). However, in cells where M-CSFR becomes downregulated, such as in mature B cells, the enhancers no longer bind the critical transcription factors, and the locus becomes nuclease insensitive (193). The ability of instructive transcription factors to reprogram cells and effectively make over established gene expression programs likely is limited by the accessibility of promoter/enhancer elements embedded within chromatin.

that it can reprogram a committed cell into another lineage by perturbing its transcription factor network, deconstructing the old one and reconstructing it into a new one. In this logic, transcription factor network perturbations in committed cells to probe mechanisms of cell programming can be compared to inducing mutations to probe a gene’s function. Similar experiments, performed with MPPs, cannot unambiguously demonstrate instructive effects of a molecule because cell selection, rather than instruction, is always a possibility. In the following sections, we discuss mechanisms that enable a single transcription factor to change a cell’s transcription factor network.

718

Laiosa

·

Stadtfeld

·

Graf

Myeloid Cell Reprogramming: The GATA-1:PU.1 Paradigm Several principles of how transcription factors control hematopoietic cell fates have been derived from studies of avian progenitors transformed with the E26 virus, which expresses the Myb-Ets oncogene. These cells, which differentiate into either erythrocytes or thrombocytes (avian megakaryocyte equivalents) when the oncoprotein is inactivated, are termed MEPE26 cells and can be induced to differentiate into myeloblasts (macrophage precursors) by activation of the Ras or protein kinase C pathways (139). Concomitant with this transition into myeloblasts, the cells downregulate the MegE regulator GATA-1. Enforced expression of GATA-1 in myeloblasts induces the formation of MEPE26 cells (140). Conversely, enforced expression of PU.1, a transcription factor whose high-level expression regulates myelomonocytic genes, reprograms MEPE26 cells into myeloblasts (141) (Figure 5). Thus, simply changing the balance of two lineage hematopoietic transcription factors can lead to the reversible reprogramming of committed myeloid cells. A simple scheme that explains the role of transcription factor stoichiometry in hematopoietic lineage decisions is depicted in Figure 6a and 6b. GATA-1 and PU.1 extinguish the alternative cell fate by antagonizing each other via protein-protein interactions (141a– c). The inhibition of GATA-1 by PU.1 entails recruitment of the Rb protein to the GATA-1:PU.1 complex bound to erythroid regulatory sequences, followed by chromatin remodeling and the formation of a repressosome (141d). Other interactions of transcription factors with chromatin are discussed in The Transcription Factor–Chromatin Connection. GATA-1 antagonizes PU.1 by blocking the interaction between PU.1 and its cofactor c-Jun (142). A two- to three-day long exposure of the active form of GATA-1 (or of PU.1) is sufﬁcient to induce the irreversible commitment of myeloblasts into MEPE26 cells (or of MEPE26 cells into myeloblasts)

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

(140, 141). This time frame suggests that overexpressed factors activate their respective endogenous counterparts, establish an autoregulatory loop, and also interrupt autoregulation of the antagonistic factor. A diagram that summarizes the GATA-1:PU.1 antagonism is shown in Figure 6c. Although there is strong evidence for PU.1 autoregulation (143), the situation for GATA-1 is less clear because deletion of a high-afﬁnity GATA-1binding site in the GATA-1 promoter mostly affects eosinophil development, but it only weakly affects erythroid cell numbers (144). Similar changes in cell fates can also be induced in primary, nontransformed cells. For example, enforced expression of GATA1 in GMPs induced the formation of MegE colonies in a high percentage of the cells (145). In addition, expression of GATA-1 in density-selected c-kit+ CD34+ cells that have restricted GM potential induced the formation of MegE colonies and also eosinophil and basophil colonies (146). Importantly, investigators have recently shown with morpholino technology in zebraﬁsh that lowering the dosage of GATA-1 (but not of GATA-2) induces a conversion of erythroid cells into myeloid cells (147) and that reducing PU.1 levels induces a switch of myeloid into erythroid cells (148). These experiments also suggest that the affected progenitors express both transcription factors, whose balance is altered by the antisense experiments. In conclusion, myeloid lineage switching can be induced in predictable ways not only by transcription factor overexpression, but also by their knockdown, in both cases altering the balance between two antagonistic transcription factors.

Simple Transcription Factor Combinations Establish Myeloid Lineage Fates The deceptively simple schemes in Figures 6a, b, and c do not account for the fact that a myeloid fate is determined by combinations of cooperating transcription factors in all lineages studied so far. This

myeloblast

MEP PU.1 (C/EBP)

GATA-1 high FOG-1

PU.1 C/EBP

GATA-1

FOG-1

PU.1

C/EBP

GATA-1

C/EBP GATA-1 low

eosinophil Figure 5 Reversible myeloid lineage reprogramming in the avian MEPE26 system. The scheme represents experiments performed with cultured avian cells transformed by the Myb-Ets oncoprotein (3). The arrows denote direction of induced differentiation. Transcription combinations that are minimally required for the speciﬁcation of each cell type are indicated. MEPE26 cells and eosinophils show approximately a twofold difference in the level of GATA-1 expression. Correspondingly, higher concentrations of GATA-1 are required to induce MEPE26 cell formation than to induce eosinophils (140). PU.1 induces an eosinophil intermediate when temporarily expressed in MEPE26 cells, as indicated by a dashed arrow (141).

was largely deduced from enforced transcription factor experiments. For example, in the avian system, expression of either C/EBP in MEPE26 cells or of low amounts of GATA-1 in myeloblasts induced eosinophil formation. Conversely, expression of FOG-1 in eosinophils converted them into MEPE26 cells (for reviews, see 3, 149). Therefore, the MegE lineage is speciﬁed by high levels of GATA-1 in combination with FOG-1, myelomonocytic cells by high levels of PU.1 together with C/EBP, and eosinophils by moderate levels of GATA-1 and C/EBP (Figure 5). A similar binary code was described for the speciﬁcation of mast cells: Here, GATA-2 cooperates with PU.1 (150). These cooperative interactions appear to be reinforced by feed-forward mechanisms, at www.annualreviews.org • Lymphoid-Myeloid Development

719

ANRV270-IY24-23

ARI

23 February 2006

a

15:47

MegE progenitors

b

PU.1

GM progenitors GATA-1

GATA-1

PU.1

d

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

c

GATA-1

GATA-1

PU.1

PU.1

FOG-1

MegE genes

C/EBP

GM genes

Figure 6 Principles of transcription factor–mediated lineage commitment. (a) and (b) Role of stoichiometry in MegE and GM lineage commitment as illustrated for the GATA-1:PU.1 paradigm. GATA-1 in excess speciﬁes MegE cell fate, and PU.1 in excess speciﬁes GM cell fate. The illustration in (c) shows cross antagonisms and autoregulatory loops of GATA-1 and PU.1, which lead to stabilization of cell phenotypes during commitment. (d) Feed-forward mechanisms whereby GATA-1 activates expression of its cooperating partner FOG-1 (145), and C/EBP activates PU.1 expression (120, 154), shown with solid arrows. The possibility that C/EBP is activated by PU.1 is indicated by the dashed arrow.

least in speciﬁc cell contexts. For example, enforced GATA-1 expression in myeloblasts and in GATA-1-defective erythroid precursors upregulates FOG-1 expression (151, 152). Although never directly addressed, PU.1 may be able to induce C/EBP expression. Some of these mechanisms may work in the opposite direction because C/EBP expression upregulates PU.1 expression in myeloid cell lines as well as in primary B cell precursors (153, 154). Feed-forward mechanisms (illustrated in Figure 6d ) serve to stabilize lineage decisions and to induce high-level expression of lineage-restricted proteins such as beta globin expression in erythroid progenitors [through the cooperation between GATA-1 and FOG-1 (152)] and of Mac-1 in macrophages [through the cooperation between PU.1 720

Laiosa

·

Stadtfeld

·

Graf

and C/EBP (154)]. The cooperation between GATA-1 and FOG-1 strictly requires speciﬁc protein-protein interactions (155). Promoter/enhancer studies generally support the concept that the pairwise combinations of lineage-instructive transcription factors described here, together with other more broadly expressed transcription factors such as Ets1, Runx1 (AML-1), and Myb, cooperate in the regulation of lineage-restricted genes (156–159).

Lymphoid-to-Myeloid Cell Reprogramming Approaches similar to those used to study the instructive effects of transcription factors in myeloid lineages have recently led

ANRV270-IY24-23

ARI

23 February 2006

15:47

Mac-1

Day 0

Day 2

Day 4

105

105

105

104

104

104

103

103

103

0

0

0

0

103

104

105

0

103

104

105

0

103

104

105

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CD19 Figure 7 C/EBPα-induced reprogramming of B cell progenitors over time as visualized by FACS analysis. CD19-expressing B cell progenitors were isolated from the bone marrow, infected with a retrovirus containing C/EBPα, and seeded under culture conditions permissive for both B and myeloid cell development. As indicated by arrows, at day 2 after infection some cells downregulated CD19, others upregulated Mac-1, and many did both. At day four about 60% of the cells converged on a CD19− Mac-1+ population (arrows), also revealing a population of nonresponders. [Figure adapted from Xie et al. (154).]

to the successful reprogramming of lymphoid progenitors and to the elucidation of some of the underlying mechanisms. Enforced expression of GATA-1 in CLPs switched ∼40% of cells seeded singly in colony assays into MegE colonies, and a similar conversion was observed after transplantation into irradiated hosts (145). In contrast, when C/EBP was expressed in CLPs, they generated GM colonies (M. Kondo, personal communication). Developmental plasticity in the lymphoid compartment is not restricted to the earliest lymphoid progenitors. A link between B cell and granulocyte development has been suggested by a mouse line (Max41) in which the integration of a transgene dramatically increased the number of neutrophils at the expense of B cells. The granulocytes of the transgenic line exhibit immunoglobulin rearrangements, indicating that they originated by a trans-differentiation of B cells in vivo (160). How integration of the transgene induces the switch is not clear. Experiments in which the Raf or Ras oncogenes were expressed in pro- or pre-B cell lines showed that the cells switch at low frequencies into functional macrophages (161), suggesting that in these cells a lineage-

instructive myeloid transcription factor had become activated. Reprogramming into inﬂammatory-type macrophages could indeed be induced by enforced expression in B cell precursors by both C/EBPα and C/EBPβ in ∼60% of the cells, whereas PU.1 and other C/EBP factors showed little effect (154). FACS analyses permitted the visualization of the reprogramming process over time, showing that reprogramming was initiated in a stochastic manner, with individual cells upregulating Mac-1 or downregulating CD19 or both. After four to ﬁve days, all reprogrammed cells had acquired a Mac-1+ CD19− phenotype (Figure 7). The switch could also be induced in immunoglobulin-expressing B cells from the spleen, although at somewhat lower frequencies. C/EBP requires PU.1 for expression of at least some myelomonocytic genes because PU.1−/− B lineage cells did not upregulate Mac-1. In a feed-forward mechanism, enforced expression of C/EBP induced the upregulation of endogenous PU.1 at levels comparable to those seen in macrophages. The extinction of CD19 expression was PU.1 independent and was due to the ability of C/EBPα to inhibit the activity of Pax5 on the CD19 promoter (154), www.annualreviews.org • Lymphoid-Myeloid Development

721

ANRV270-IY24-23

ARI

23 February 2006

C/EBPα

15:47

C/EBPα

pro-B

pre-B

C/EBPα

C/EBPα

imm B

Mac-1

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CD19 Pax5

CD19

PU.1

Mac-1 macrophage

Figure 8 Mechanisms of transcription factor–induced B cell reprogramming. The scheme illustrates that different stages of B cell differentiation, including immature B cells (imm B), which express immunoglobulin, can be reprogrammed into macrophages. The insert shows that C/EBPα extinguishes CD19 expression by inhibiting Pax5 activity (red T bar) and induces myeloid gene expression by cooperating with endogenous PU.1 (green arrow), as well as by upregulating PU.1 expression.

probably owing to antagonistic interactions between the two proteins (W. Ci & T. Graf, unpublished observations). The results described are illustrated in Figure 8. The inhibition of Pax5 activity does not explain the ability of C/EBP to induce a complete collapse of the B cell gene expression program. How this is achieved is not yet entirely clear but may involve a secondary antagonism by upregulated PU.1 (H. Xie & T. Graf, unpublished results). In another study, enforced expression of C/EBPα did not cause myeloid reprogramming of B cell precursors, although it did reprogram Pax5−/− B cell precursors (161a). The reason for this apparent discrepancy to the previous study is not obvious because, in an exchange of vectors encoding C/EBPα, the construct of Heavey et al. (161a) reprogrammed wild-type B cell precursors at similar efﬁciencies as the one reported by Xie & Graf (H. Xie & T. 722

Laiosa

·

Stadtfeld

·

Graf

Graf, unpublished observations). Committed T cells can also be reprogrammed by C/EBPα into macrophages after enforced expression, using an approach similar to that used for the B lineage (C.V. Laiosa, H. Xie & T. Graf, manuscript submitted). Furthermore, T cell commitment is also altered by overexpression of PU.1 or SpiB, a transcription factor related to PU.1, either of which converts pre-T (DN3) cells into cells resembling dendritic cells (162, 162a). The ﬁndings that CLPs and committed B and T cell precursors can be switched into myeloid cells beg the questions whether this process is reversible and whether B and T cells can be converted into one another. This was studied using a knockin mouse line in which Pax5 was expressed under the control of the Ikaros locus, which is expressed in all lymphoid and myeloid progenitors (96). Pax5 did not signiﬁcantly disrupt myeloid or erythroid differentiation, suggesting that in this context it is insufﬁcient for B cell speciﬁcation. However, T cell differentiation was disturbed, with decreased thymic cellularity and a differentiation block around the time of T cell commitment. In reconstitution experiments with lethally irradiated mice, bone marrow progenitors expressing Pax5 did not contribute to T cell differentiation. This might be explained by the ability of Pax5 to inhibit Notch1 expression, as shown indirectly by reintroduction of Pax5 into Pax5−/− B cell precursors (96). Moreover, although Pax5 failed to interfere with the expression of the T cell markers Thy-1 and CD3ε in thymocytes, it upregulated the B cell–speciﬁc gene CD19 (96). In an independent study in which Pax5 expression in T cells was driven by elements from the human CD2 promoter, Pax5 induced expression of its known targets as well as heavy chain rearrangements without initiating other aspects of the B cell program (162b). Finally, experiments in which primary macrophage cultures were infected with retroviruses carrying either Pax5, E2A, or EBF and cultured under conditions permissive for B cell development likewise did

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

not induce the formation of nonadherent CD19-expressing lymphoid cells (H. Xie & T. Graf, unpublished observations). In sum, ectopic Pax5 expression in myeloid cells has no detectable effects on their phenotype, whereas it induces a differentiation defect in early T cells and activation of some B cell–speciﬁc target genes. The effect of T cell transcription factors on the phenotype of B cell precursors has not been directly studied. However, seeding fetal liver LSK cells onto OP9-DL1 stroma, conditions under which Notch1 is activated, inhibited their differentiation into the B cell lineage, normally seen on OP9 cells not expressing DL1. Once cells were committed to the B cell lineage, they were no longer impeded by OP9-DL1 stroma (105). This ﬁnding is consistent with the notion that Notch signaling interferes with B cell commitment but not with B cell proliferation and maturation.

Lymphoid Transcription Factors Repress Lineage Inappropriate Gene Expression A serendipitous observation led to the discovery that the inactivation of a lineage commitment factor can awaken a dormant multilineage potential in cells that were ostensibly speciﬁed to the B cell lineage. Thus, when Pax5−/− pro-B cells were cultured for several weeks on ST2 stromal cells (which produce myeloid cytokines) without replenishing IL-7, the formation of macrophages was observed. The Pax5−/− pro-B cells also had the capacity to differentiate into osteoclasts, granulocytes, NK cells, and dendritic cells (95) under appropriate culture conditions and into T cells upon transfer into immunodeﬁcient recipients (95, 163). The multilineage potential (summarized in Figure 9) was not brought about by selection in culture because freshly isolated Pax5−/− pro-B cells also generated T cell progeny after transplantation (163). The B cell origin of

T cell

NK cell

dendritic cell

Pax5-/E2A-/“pro-B” cell

macrophage

granulocyte

erythrocyte Figure 9 Multilineage differentiation potential of Pax5−/− and E2A−/− B cell progenitors. See text for details.

these cells was supported by the detection of immunoglobulin heavy chain DJ rearrangements, although a small percentage of normal thymocytes also possess heavy chain DJ rearrangements (162b). The ﬁnding that IL7 effectively suppresses myeloid differentiation strongly argues for an instructive effect of this cytokine. Pax5 is also required throughout B cell development to maintain commitment. Thus, conditional deletion in B cells caused the downregulation of B cell markers and derepression of myeloid genes even in mature B cells (164); after transfer to appropriate conditions, these cells differentiated into macrophages (165). How Pax5 represses the expression of various myelomonocytic genes, such as that of M-CSFR (166) and Notch1 (96), remains to be determined. Similar observations were made with mice defective in the E2A gene (Figure 9). This gene regulates expression of EBF, which in turn regulates the B cell commitment factor

www.annualreviews.org • Lymphoid-Myeloid Development

723

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

Pax5 (63). B cell development in E2A−/− mice is blocked at the pro-B cell stage (85, 109, 167). Cultured E2A−/− pro-B cells express multiple myeloid genes (168), which likely results from the absence of Pax5 (95). Importantly, like Pax5−/− cells, E2A−/− cells were capable of differentiating in vivo into all hematopoietic lineages tested (with the exception of B cells), as shown by transplantation into irradiated hosts (168). However, whether freshly isolated E2A−/− cells not selected in culture are capable of multilineage differentiation is unclear. That E2A−/− cells also generate T cells is surprising given the deﬁciency in T cell formation from progenitors of E2A−/− mice (109). This raises the possibility that expression of a redundant factor abrogates the requirement for E2A in T cell differentiation. PU.1 is another example of a transcription factor whose ablation induces lineageinappropriate gene expression. Cultured fetal liver–derived B cells lacking PU.1 ac-

GATA-1

CLP

C/EBPα

tivate the expression of a subset of T cell genes, including ZAP-70 and lck (84a). Therefore, the apparent requirement of PU.1 downregulation at the pro-T/pre-T cell transition for T cell commitment (23, 114) might be explained by the ability of PU.1 to repress T cell genes. In addition, inactivation or reduction of PU.1 levels in B lineage cells in vivo induced a dramatic shift from B-2 to B1 cells (84a, 168a). An unsuspected link to leukemia was observed in animals in which alterations or deletions of a regulatory site within the PU.1 enhancer, leading to a reduced expression of the factor, induced acute myeloid and T cell leukemia (168a, 169). PU.1 dosage also inﬂuences lymphoid versus myeloid speciﬁcation because enforced expression of PU.1 in PU.1+/− fetal liver progenitors showed that higher levels of the protein favored myelomonocytic speciﬁcation rather than B cell formation (170). However, as mentioned above, it is unclear whether this is a true dosage effect because under certain culture conditions B cells develop in a PU.1-independent manner, as for fetal liver progenitors. The multiple functions of PU.1 in the differentiation and proliferation of both lymphoid and myeloid lineages show that this transcription factor has a prominent role as a master coordinator of hematopoiesis.

MegE C/EBPα

pro-B pre-B

C/EBPα GM

pro-T pre-T

Figure 10 Asymmetries in lymphoid and myeloid cell plasticity. Scheme summarizing experiments showing reprogramming of CLPs into MegE cells by GATA-1 (145) and into GM cells by C/EBPα expression (M. Kondo, personal communication). In addition, the scheme illustrates reprogramming of pro-B, pre-B, pro-T, and pre-T cells into macrophages by enforced expression of C/EBPα (154; C.V. Laiosa, H. Xie & T. Graf, unpublished results). 724

Laiosa

·

Stadtfeld

·

Graf

TRANSCRIPTION FACTOR NETWORK COMPLEXITY AS A CRITICAL DETERMINANT OF LYMPHOID AND MYELOID CELL PLASTICITY Why can the enforced expression of a single transcription factor reprogram lymphoid-tomyeloid cells, whereas lymphoid-to-myeloid conversions have not been achieved (summarized in Figure 10)? A simple explanation is that the conversions have not been achieved owing to technical inadequacies of the experiments performed so far. Another is that the plasticity bias reﬂects irreversible chromatin changes in myeloid cells that

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

prevent accessibility of lymphoid genes to lymphoid transcription factors. However, the data discussed in this review suggest an alternative explanation. First, a comparison of the myeloid and lymphoid compartments shows striking differences in their transcription factor network complexities. Thus, as described above, at least four transcription factors are required for B and T lineage establishment, whereas two appear to be sufﬁcient to specify myeloid lineages. Second, the B cell transcription factor network is set up sequentially and strengthened by external inputs, such as signaling through Flt3 and IL-7R (2). Although somewhat less clear, the T cell transcription factor network is possibly also set up in a sequential fashion and depends on signaling through the Notch1 receptor (23, 62). In contrast, the simple binary transcription factor code of myeloid cells can be established in one or two steps that might be independent of external stimuli. Third, related to the binary code of myeloid cells, at least some myeloid instructive transcription factors can establish feed-forward loops, upregulating their respective partner. In contrast, lymphoid transcription factors do not appear to do so or, if they do, not in a reciprocal fashion. For example, although EBF can activate Pax5 expression, the reciprocal mechanism, Pax5-activating expression of EBF, is not known to occur. The notion that complex transcription factor combinations, assembled in a sequential manner, specify lymphoid cells, whereas much simpler combinations specify myeloid cells, may reﬂect their more recent evolutionary origin (18). Thus, the basic wiring of myeloid cells may have served as a platform for the design of more complex cell phenotypes. This idea is also supported by the fact that B and T cell lineages appear relatively late during ontogeny (170a) and that they exhibit a number of well-deﬁned developmental stages, not seen in myeloid lineages (171–173). The above speculations predict that to achieve the reprogramming of myeloid into lymphoid lineages or lymphoid into other lymphoid lineages will require the

expression of two or more lymphoid instructive transcription factors, perhaps in a deﬁned sequence.

LINEAGE PRIMING IN HSCs AND INTERMEDIATE PROGENITORS

Lineage priming: expression of lineage-restricted genes in multipotent hematopoietic progenitors

Commitment of HSCs and progenitor cells to a given lineage may simply be caused by the activation of one or several key transcription factors that then establish the gene expression proﬁle characteristic of, for example, a macrophage or a B cell. Thus, HSCs might start out as blank slates that express no genes characteristic of differentiated cells. However, Enver and colleagues (174) showed that single multipotent hematopoietic cells coexpress genes at low levels normally found in mature MegE and GM cells, such as hemoglobin and myeloperoxidase. Subsequent studies have shown that lineage priming or lineage promiscuity occurs in all adult hematopoietic progenitor cells, including long-term HSCs (175, 176) and in MPPs isolated from the early mouse embryo (177). The coexistence of different transcriptional programs in progenitor cells, followed by the stepwise extinction of all but one of them, is therefore a deﬁning feature of the hematopoietic system. The repertoire of primed genes is not the same in different hematopoietic progenitor cells. Thus, single long-term HSCs as well as CMPs coexpress a wide variety of markers of mature GM and MegE cells, including several myeloid cytokine receptors as well as the transcription factors PU.1, C/EBPα, and GATA1. GMPs and MEPs, in turn, only express GM- or MegE-associated genes, respectively. As expected, CLPs prime both T and B cell genes but not myeloid genes (175). Priming therefore appears to be restricted to genes expressed in the physiological progeny of a cell. However, markers of mature B and T cells are essentially absent from long-term HSCs (175), which is also evidenced by the total lack of myeloid cell labeling in a mouse model www.annualreviews.org • Lymphoid-Myeloid Development

725

ANRV270-IY24-23

ARI

23 February 2006

15:47

that traces the progeny of CD19-expressing cells in vivo (176). Of the speciﬁc regulators of lymphoid development, only GATA-3 and Notch1, both implicated in T cell development, are expressed in long-term HSCs, although the Notch1 target gene pTα cannot be detected (175, 178, 179). The gene expression program of lymphocytes is therefore not previewed in HSCs, in contrast to that of myeloid lineages.

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

HOW DO HEMATOPOIETIC STEM CELLS BECOME LYMPHOID? Because HSCs express myeloid markers as well as several of the key myeloid regulators, but appear to express no lymphoid markers and few T cell–associated transcription factors, they resemble more closely myeloid than lymphoid progenitors. The expression of lineage-restricted markers in early progenitors is most likely a direct reﬂection of the transcription factor combinations they express. A summary of lineage-instructive transcription factors expressed in HSCs as well as in selected intermediate and monopotent progenitors is shown in Figure 11. The scheme supports the notion that lineage speciﬁcation involves a stepwise simpliﬁcation of transcription factor networks, with fully committed cells representing stable, low-energy states. That the transcription factor network complexity of CLPs appears to be higher suggests that HSCs have to go uphill to become lymphoid. This notion sharply focuses the question as to how HSCs escape a myeloid fate and instead become lymphoid. In light of the self-renewal and multilineage differentiation capacities of HSCs, one must assume that the expression of myeloidinstructive factors is too low to induce myeloid commitment, at least in the bulk of the population. Small transcription factor ﬂuctuations and imbalances in HSCs might then represent the ﬁrst step in lineage commitment. Transcription factor antagonisms and syner-

726

Laiosa

·

Stadtfeld

·

Graf

gisms, combined with external signals such as cytokine receptor signaling, could amplify such imbalances, making commitment irreversible. Although whether lineage priming in HSCs is an oscillatory process has not yet been determined, RT-PCR analysis suggests that HSCs alternate between different patterns of gene expression (175), some of which might be more conducive for commitment to a certain lineage than others. However, at least for the expression of the myeloid marker lysozyme, a developmental bias based on lineage priming does not exist because CMPs expressing lysozyme M differentiate into MegE cells at similar frequencies (175). Nevertheless, this situation could be different for the stochastic upregulation of cytokine receptors, which then can be triggered by environmental cues. Important players in the loss of selfrenewal, entry into cell cycle, and initiation of lymphoid differentiation of HSCs/MPPs may be the tyrosine kinase receptor Flt3 and the transmembrane receptor Notch1. Thus, on the one hand, expression of Flt3, which causes myeloid leukemias when mutated (180), is progressively upregulated from long-term HSCs, to short-term HSCs, to MPPs, correlating with the loss of self-renewal potential (9). On the other hand, Flt3 is required for the generation of CLPs (53), and its ligandmediated activation in synergy with PU.1 may induce the expression of IL-7Rα which in turn upregulates EBF (58a) in cells destined to become B cells (2). Similarly, Notch1, which is frequently mutated in T cell leukemias (181), is progressively upregulated from long-term HSCs to MPPs (178). This upregulation may be important for the successful seeding of the thymus and T cell speciﬁcation. In contrast to the expression pattern of Flt3 and Notch1, GATA-1 and GATA-2 are expressed at higher levels in long-term HSCs than in short-term HSCs or MPPs (178). The apparent inverse regulation of myeloid- and lymphoid-associated regulators in LSK cells raises the possibility that they inhibit each

ANRV270-IY24-23

ARI

23 February 2006

15:47

GATA-1 PU.1 C/EBPα Notch1 GATA-3

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

HSC

GATA-1 PU.1 C/EBPα

PU.1 E2A EBF Pax5 GATA-3 Notch1

CMP

CLP

GATA-1 FOG-1

PU.1 C/EBPα

PU.1 E2A GATA-3 Notch1

PU.1 E2A EBF Pax5

MEP

GMP

pro-T

pro-B

Figure 11 Transcription factor network complexity of HSCs, intermediate, and monopotent progenitors. The expression of lineage-instructive transcription factors is indicated within each cell type. The levels of a given transcription factor may differ substantially in the various cell types shown, and this may be of biological signiﬁcance. GATA-2, which is expressed in all intermediate myeloid progenitors, is not included because its signiﬁcance in establishing lineage-speciﬁc gene expression programs is unclear. (References used to compile the factors listed: 14, 35, 58, 72, 175, 178, 194.)

other, changing probabilities of lymphoid versus myeloid lineage speciﬁcation.

CONCLUDING REMARKS As highlighted by this review, a great deal has been learned about the determinants that specify lymphoid and myeloid fate, and in particular about the role of transcription factors in these processes. However, many open questions remain, such as whether lymphoidmyeloid branching occurs at invariant positions along differentiation pathways (as proposed by the CMP/CLP model) or occurs

with certain probabilities within the progeny of different early hematopoietic progenitors. Although the reprogramming experiments discussed provide plausible scenarios about how lymphoid cells diverge from myeloid cells and why they retain a certain degree of plasticity, future experiments must address this issue in a physiological context. One problem that was brought into sharp focus is how HSCs, which more closely resemble myeloid than lymphoid precursors, acquire a lymphoid fate. We hope that this interpretation of the available data will help in designing experiments that directly approach this question.

www.annualreviews.org • Lymphoid-Myeloid Development

727

ANRV270-IY24-23

ARI

23 February 2006

15:47

ACKNOWLEDGMENTS We thank Emmanuelle Passegue, Cornelius Murre, Daniel Tenen, Arthur Skoultchi, and Motonari Kondo for providing unpublished data; Ellen Rothenberg, Juan Carlos ZunigaPﬂucker, and Fabio Rossi for comments on the manuscript; and Huafeng Xie for helpful discussions as well as contribution of the data analysis shown in Figure 7. This work was supported by NIH grants R01 CA89590-01 and R01 NS43881-01.

LITERATURE CITED

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

1. Rothenberg EV, Anderson MK. 2002. Elements of transcription factor network design for T-lineage speciﬁcation. Dev. Biol. 246:29–44 2. Singh H, Medina KL, Pongubala JM. 2005. Contingent gene regulatory networks and B cell fate speciﬁcation. Proc. Natl. Acad. Sci. USA 102:4949–53 3. Graf T. 2002. Differentiation plasticity of hematopoietic cells. Blood 99:3089–101 4. Orkin SH. 2000. Diversiﬁcation of haematopoietic stem cells to speciﬁc lineages. Nat. Rev. Genet. 1:57–64 5. Ogawa M. 1993. Differentiation and proliferation of hematopoietic stem cells. Blood 81:2844–53 6. Morrison SJ, Hemmati HD, Wandycz AM, Weissman IL. 1995. The puriﬁcation and characterization of fetal liver hematopoietic stem cells. Proc. Natl. Acad. Sci. USA 92:10302–6 7. Morrison SJ, Weissman IL. 1994. The long-term repopulating subset of hematopoietic stem cells is deterministic and isolatable by phenotype. Immunity 1:661–73 8. Spangrude GJ, Heimfeld S, Weissman IL. 1988. Puriﬁcation and characterization of mouse hematopoietic stem cells. Science 241:58–62 9. Christensen JL, Weissman IL. 2001. Flk-2 is a marker in hematopoietic stem cell differentiation: a simple method to isolate long-term stem cells. Proc. Natl. Acad. Sci. USA 98:14541–46 10. Kiel MJ, Yilmaz OH, Iwashita T, Terhorst C, Morrison SJ. 2005. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell 121:1109–21 11. Yang L, Bryder D, Adolfsson J, Nygren J, Mansson R, et al. 2005. Identiﬁcation of Lin− Sca1+ kit+ CD34+ Flt3− short-term hematopoietic stem cells capable of rapidly reconstituting and rescuing myeloablated transplant recipients. Blood 105:2717– 23 12. Osawa M, Hanada K, Hamada H, Nakauchi H. 1996. Long-term lymphohematopoietic reconstitution by a single CD34-low/negative hematopoietic stem cell. Science 273:242– 45 13. Kondo M, Weissman IL, Akashi K. 1997. Identiﬁcation of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91:661–72 14. Akashi K, Traver D, Miyamoto T, Weissman IL. 2000. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature 404:193–97 15. Denburg JA, Telizyn S, Messner H, Lim B, Jamal N, et al. 1985. Heterogeneity of human peripheral blood eosinophil-type colonies: evidence for a common basophil-eosinophil progenitor. Blood 66:312–18 16. Iwasaki H, Mizuno S, Mayﬁeld R, Shigematsu H, Arinobu Y, et al. 2005. Identiﬁcation of eosinophil lineage-committed progenitors in the murine bone marrow. J. Exp. Med. 201:1891–97 728

Laiosa

·

Stadtfeld

·

Graf

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

17. Adolfsson J, Mansson R, Buza-Vidas N, Hultquist A, Liuba K, et al. 2005. Identiﬁcation of Flt3+ lympho-myeloid stem cells lacking erythro-megakaryocytic potential a revised road map for adult blood lineage commitment. Cell 121:295–306 18. Anderson MK, Pant R, Miracle AL, Sun X, Luer CA, et al. 2004. Evolutionary origins of lymphocytes: ensembles of T cell and B cell transcriptional regulators in a cartilaginous ﬁsh. J. Immunol. 172:5851–60 19. Kondo M, Scherer DC, Miyamoto T, King AG, Akashi K, et al. 2000. Cell-fate conversion of lymphoid-committed progenitors by instructive actions of cytokines. Nature 407:383–86 19a. Martin CH, Aifantis I, Scimone ML, von Andrian UH, Reizis B, et al. 2003. Efﬁcient thymic immigration of B220+ lymphoid-restricted bone marrow cells with T precursor potential. Nat. Immunol. 4:866–73 20. Allman D, Sambandam A, Kim S, Miller JP, Pagan A, et al. 2003. Thymopoiesis independent of common lymphoid progenitors. Nat. Immunol. 4:168–74 21. Schwarz BA, Bhandoola A. 2004. Circulating hematopoietic progenitors with T lineage potential. Nat. Immunol. 5:953–60 22. Petrie HT, Kincade PW. 2005. Many roads, one destination for T cell progenitors. J. Exp. Med. 202:11–13 23. Rothenberg EV, Taghon T. 2005. Molecular genetics of T cell development. Annu. Rev. Immunol. 23:601–49 24. Shen HQ, Lu M, Ikawa T, Masuda K, Ohmura K, et al. 2003. T/NK bipotent progenitors in the thymus retain the potential to generate dendritic cells. J. Immunol. 171:3401–6 25. Benz C, Bleul CC. 2005. A multipotent precursor in the thymus maps to the branching point of the T versus B lineage decision. J. Exp. Med. 202:21–31 26. Porritt HE, Rumfelt LL, Tabrizifard S, Schmitt TM, Zuniga-Pﬂucker JC, Petrie HT. 2004. Heterogeneity among DN1 prothymocytes reveals multiple progenitors with different capacities to generate T cell and non-T cell lineages. Immunity 20:735–45 27. Balciunaite G, Ceredig R, Rolink AG. 2005. The earliest subpopulation of mouse thymocytes contains potent T, signiﬁcant macrophage, and natural killer cell but no Blymphocyte potential. Blood 105:1930–36 28. Zediak VP, Maillard I, Bhandoola A. 2005. Closer to the source: Notch and the nature of thymus-settling cells. Immunity 23:245–48 29. Shortman K, Liu YJ. 2002. Mouse and human dendritic cell subtypes. Nat. Rev. Immunol. 2:151–61 30. Traver D, Akashi K, Manz M, Merad M, Miyamoto T, et al. 2000. Development of CD8α-positive dendritic cells from a common myeloid progenitor. Science 290:2152– 54 31. Manz MG, Traver D, Miyamoto T, Weissman IL, Akashi K. 2001. Dendritic cell potentials of early lymphoid and myeloid progenitors. Blood 97:3333–41 32. Lu M, Kawamoto H, Katsube Y, Ikawa T, Katsura Y. 2002. The common myelolymphoid progenitor: a key intermediate stage in hemopoiesis generating T and B cells. J. Immunol. 169:3519–25 33. Katsura Y. 2002. Redeﬁnition of lymphoid progenitors. Nat. Rev. Immunol. 2:127–32 34. Kawamoto H, Ikawa T, Ohmura K, Fujimoto S, Katsura Y. 2000. T cell progenitors emerge earlier than B cell progenitors in the murine fetal liver. Immunity 12:441–50 35. Mebius RE, Miyamoto T, Christensen J, Domen J, Cupedo T, et al. 2001. The fetal liver counterpart of adult common lymphoid progenitors gives rise to all lymphoid lineages, CD45+ CD4+ CD3− cells, as well as macrophages. J. Immunol. 166:6593–601 www.annualreviews.org • Lymphoid-Myeloid Development

729

ARI

23 February 2006

15:47

36. Traver D, Miyamoto T, Christensen J, Iwasaki-Arai J, Akashi K, Weissman IL. 2001. Fetal liver myelopoiesis occurs through distinct, prospectively isolatable progenitor subsets. Blood 98:627–35 37. Enver T, Heyworth CM, Dexter TM. 1998. Do stem cells play dice? Blood 92:348–51 38. Metcalf D. 1998. Lineage commitment and maturation in hematopoietic cells: the case for extrinsic regulation. Blood 92:345–47 39. Lieschke GJ, Grail D, Hodgson G, Metcalf D, Stanley E, et al. 1994. Mice lacking granulocyte colony-stimulating factor have chronic neutropenia, granulocyte and macrophage progenitor cell deﬁciency, and impaired neutrophil mobilization. Blood 84:1737–46 40. Wu H, Liu X, Jaenisch R, Lodish HF. 1995. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell 83:59–67 41. Stanley E, Lieschke GJ, Grail D, Metcalf D, Hodgson G, et al. 1994. Granulocyte/macrophage colony-stimulating factor-deﬁcient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc. Natl. Acad. Sci. USA 91:5592–96 42. Seymour JF, Lieschke GJ, Grail D, Quilici C, Hodgson G, Dunn AR. 1997. Mice lacking both granulocyte colony-stimulating factor (CSF) and granulocyte-macrophage CSF have impaired reproductive capacity, perturbed neonatal granulopoiesis, lung disease, amyloidosis, and reduced long-term survival. Blood 90:3037–49 43. Lieschke GJ, Stanley E, Grail D, Hodgson G, Sinickas V, et al. 1994. Mice lacking both macrophage- and granulocyte-macrophage colony-stimulating factor have macrophages and coexistent osteopetrosis and severe lung disease. Blood 84:27–35 44. Pharr PN, Ogawa M, Hofbauer A, Longmore GD. 1994. Expression of an activated erythropoietin or a colony-stimulating factor 1 receptor by pluripotent progenitors enhances colony formation but does not induce differentiation. Proc. Natl. Acad. Sci. USA 91:7482–86 45. McArthur GA, Rohrschneider LR, Johnson GR. 1994. Induced expression of c-fms in normal hematopoietic cells shows evidence for both conservation and lineage restriction of signal transduction in response to macrophage colony-stimulating factor. Blood 83:972–81 46. Stoffel R, Ziegler S, Ghilardi N, Ledermann B, de Sauvage FJ, Skoda RC. 1999. Permissive role of thrombopoietin and granulocyte colony-stimulating factor receptors in hematopoietic cell fate decisions in vivo. Proc. Natl. Acad. Sci. USA 96:698–702 47. Semerad CL, Poursine-Laurent J, Liu F, Link DC. 1999. A role for G-CSF receptor signaling in the regulation of hematopoietic cell function but not lineage commitment or differentiation. Immunity 11:153–61 48. Borzillo GV, Ashmun RA, Sherr CJ. 1990. Macrophage lineage switching of murine early pre-B lymphoid cells expressing transduced fms genes. Mol. Cell. Biol. 10:2703–14 49. Iwasaki-Arai J, Iwasaki H, Miyamoto T, Watanabe S, Akashi K. 2003. Enforced granulocyte/macrophage colony-stimulating factor signals do not support lymphopoiesis, but instruct lymphoid to myelomonocytic lineage conversion. J. Exp. Med. 197:1311–22 50. King AG, Kondo M, Scherer DC, Weissman IL. 2002. Lineage inﬁdelity in myeloid cells with TCR gene rearrangement: a latent developmental potential of proT cells revealed by ectopic cytokine receptor signaling. Proc. Natl. Acad. Sci. USA 99:4508–13 51. Adolfsson J, Borge OJ, Bryder D, Theilgaard-Monch K, Astrand-Grundstrom I, et al. 2001. Upregulation of Flt3 expression within the bone marrow Lin− Sca1+ c-kit+ stem cell compartment is accompanied by loss of self-renewal capacity. Immunity 15:659–69

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

730

Laiosa

·

Stadtfeld

·

Graf

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

52. Mackarehtschian K, Hardin JD, Moore KA, Boast S, Goff SP, Lemischka IR. 1995. Targeted disruption of the ﬂk2/ﬂt3 gene leads to deﬁciencies in primitive hematopoietic progenitors. Immunity 3:147–61 53. Sitnicka E, Bryder D, Theilgaard-Monch K, Buza-Vidas N, Adolfsson J, Jacobsen SE. 2002. Key role of ﬂt3 ligand in regulation of the common lymphoid progenitor but not in maintenance of the hematopoietic stem cell pool. Immunity 17:463–72 54. Corcoran AE, Riddell A, Krooshoop D, Venkitaraman AR. 1998. Impaired immunoglobulin gene rearrangement in mice lacking the IL-7 receptor. Nature 391:904–7 55. Peschon JJ, Morrissey PJ, Grabstein KH, Ramsdell FJ, Maraskovsky E, et al. 1994. Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deﬁcient mice. J. Exp. Med. 180:1955–60 56. von Freeden-Jeffry U, Solvason N, Howard M, Murray R. 1997. The earliest T lineagecommitted cells depend on IL-7 for Bcl-2 expression and normal cell cycle progression. Immunity 7:147–54 57. Akashi K, Kondo M, von Freeden-Jeffry U, Murray R, Weissman IL. 1997. Bcl-2 rescues T lymphopoiesis in interleukin-7 receptor-deﬁcient mice. Cell 89:1033–41 58. Dias S, Silva H Jr, Cumano A, Vieira P. 2005. Interleukin-7 is necessary to maintain the B cell potential in common lymphoid progenitors. J. Exp. Med. 201:97179 58a. Kikuchi K, Lai AY, Hsu C-L, Kondo M. 2005. IL-7 receptor signaling is necessary for stage transition in adult B cell development through up-regulation of EBF. J. Exp. Med. 201:1197–203 59. Sitnicka E, Brakebusch C, Martensson IL, Svensson M, Agace WW, et al. 2003. Complementary signaling through ﬂt3 and interleukin-7 receptor α is indispensable for fetal and adult B cell genesis. J. Exp. Med. 198:1495–506 60. Matthias P, Rolink AG. 2005. Transcriptional networks in developing and mature B cells. Nat. Rev. Immunol. 5:497–508 61. Busslinger M. 2004. Transcriptional control of early B cell development. Annu. Rev. Immunol. 22:55–79 62. Maillard I, Fang T, Pear WS. 2005. Regulation of lymphoid development, differentiation, and function by the Notch pathway. Annu. Rev. Immunol. 23:945–74 63. Engel I, Murre C. 2001. The function of E- and Id proteins in lymphocyte development. Nat. Rev. Immunol. 1:193–99 64. Tenen DG. 2003. Disruption of differentiation in human cancer: AML shows the way. Nat. Rev. Cancer 3:89–101 65. Georgopoulos K, Bigby M, Wang JH, Molnar A, Wu P, et al. 1994. The Ikaros gene is required for the development of all lymphoid lineages. Cell 79:143–56 66. Wang JH, Nichogiannopoulou A, Wu L, Sun L, Sharpe AH, et al. 1996. Selective defects in the development of the fetal and adult lymphoid system in mice with an Ikaros null mutation. Immunity 5:537–49 67. McKercher SR, Torbett BE, Anderson KL, Henkel GW, Vestal DJ, et al. 1996. Targeted disruption of the PU.1 gene results in multiple hematopoietic abnormalities. EMBO J. 15:5647–58 68. Scott EW, Simon MC, Anastasi J, Singh H. 1994. Requirement of transcription factor PU.1 in the development of multiple hematopoietic lineages. Science 265:1573–77 69. Georgopoulos K. 2002. Haematopoietic cell-fate decisions, chromatin regulation and Ikaros. Nat. Rev. Immunol. 2:162–74 70. Lopez RA, Schoetz S, DeAngelis K, O’Neill D, Bank A. 2002. Multiple hematopoietic defects and delayed globin switching in Ikaros null mice. Proc. Natl. Acad. Sci. USA 99:602–7 www.annualreviews.org • Lymphoid-Myeloid Development

731

ARI

23 February 2006

15:47

71. Nichogiannopoulou A, Trevisan M, Neben S, Friedrich C, Georgopoulos K. 1999. Defects in hemopoietic stem cell activity in Ikaros mutant mice. J. Exp. Med. 190:1201– 14 72. Klug CA, Morrison SJ, Masek M, Hahm K, Smale ST, Weissman IL. 1998. Hematopoietic stem cells and lymphoid progenitors express different Ikaros isoforms, and Ikaros is localized to heterochromatin in immature lymphocytes. Proc. Natl. Acad. Sci. USA 95:657–62 73. Morgan B, Sun L, Avitahl N, Andrikopoulos K, Ikeda T, et al. 1997. Aiolos, a lymphoid restricted transcription factor that interacts with Ikaros to regulate lymphocyte differentiation. EMBO J. 16:2004–13 74. Georgopoulos K, Winandy S, Avitahl N. 1997. The role of the Ikaros gene in lymphocyte development and homeostasis. Annu. Rev. Immunol. 15:155–76 75. Spain LM, Guerriero A, Kunjibettu S, Scott EW. 1999. T cell development in PU.1deﬁcient mice. J. Immunol. 163:2681–87 76. Guerriero A, Langmuir PB, Spain LM, Scott EW. 2000. PU.1 is required for myeloidderived but not lymphoid-derived dendritic cells. Blood 95:879–85 77. Anderson KL, Perkin H, Surh CD, Venturini S, Maki RA, Torbett BE. 2000. Transcription factor PU.1 is necessary for development of thymic and myeloid progenitor-derived dendritic cells. J. Immunol. 164:1855–61 78. Iwasaki H, Somoza C, Shigematsu H, Duprez EA, Iwasaki-Arai J, et al. 2005. Distinctive and indispensable roles of PU.1 in maintenance of hematopoietic stem cells and their differentiation. Blood 106:1590–600 79. Dakic A, Metcalf D, Di Rago L, Mifsud S, Wu L, Nutt SL. 2005. PU.1 regulates the commitment of adult hematopoietic progenitors and restricts granulopoiesis. J. Exp. Med. 201:1487–502 80. Zhang P, Iwasaki-Arai J, Iwasaki H, Fenyus ML, Dayaram T, et al. 2004. Enhancement of hematopoietic stem cell repopulating capacity and self-renewal in the absence of the transcription factor C/EBPα. Immunity 21:853–63 81. DeKoter RP, Lee HJ, Singh H. 2002. PU.1 regulates expression of the interleukin-7 receptor in lymphoid progenitors. Immunity 16:297–309 82. Medina KL, Pongubala JM, Reddy KL, Lancki DW, Dekoter R, et al. 2004. Assembling a gene regulatory network for speciﬁcation of the B cell fate. Dev. Cell 7:607–17 83. Lin H, Grosschedl R. 1995. Failure of B-cell differentiation in mice lacking the transcription factor EBF. Nature 376:263–67 84. von Freeden-Jeffry U, Vieira P, Lucian LA, McNeil T, Burdach SE, Murray R. 1995. Lymphopenia in interleukin (IL)-7 gene-deleted mice identiﬁes IL-7 as a nonredundant cytokine. J. Exp. Med. 181:1519–26 84a. Ye M, Ermakova O, Graf T. 2005. PU.1 is not strictly required for B cell development and its absence induces a B-2 to B-1 cell switch. J. Exp. Med. 202:1411–22 85. Bain G, Maandag EC, Izon DJ, Amsen D, Kruisbeek AM, et al. 1994. E2A proteins are required for proper B cell development and initiation of immunoglobulin gene rearrangements. Cell 79:885–92 86. Sun XH. 1994. Constitutive expression of the Id1 gene impairs mouse B cell development. Cell 79:893–900 87. Sigvardsson M, O’Riordan M, Grosschedl R. 1997. EBF and E47 collaborate to induce expression of the endogenous immunoglobulin surrogate light chain genes. Immunity 7:25–36 88. Gisler R, Sigvardsson M. 2002. The human V-preB promoter is a target for coordinated activation by early B cell factor and E47. J. Immunol. 168:5130–38

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

732

Laiosa

·

Stadtfeld

·

Graf

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

89. Kee BL, Rivera RR, Murre C. 2001. Id3 inhibits B lymphocyte progenitor growth and survival in response to TGF-β. Nat. Immunol. 2:242–47 90. Zhuang Y, Cheng P, Weintraub H. 1996. B-lymphocyte development is regulated by the combined dosage of three basic helix-loop-helix genes, E2A, E2-2, and HEB. Mol. Cell. Biol. 16:2898–905 91. O’Riordan M, Grosschedl R. 1999. Coordinate regulation of B cell differentiation by the transcription factors EBF and E2A. Immunity 11:21–31 92. Sigvardsson M, Clark DR, Fitzsimmons D, Doyle M, Akerblad P, et al. 2002. Early Bcell factor, E2A, and Pax-5 cooperate to activate the early B cell-speciﬁc mb-1 promoter. Mol. Cell. Biol. 22:8539–51 93. Urbanek P, Wang ZQ, Fetka I, Wagner EF, Busslinger M. 1994. Complete block of early B cell differentiation and altered patterning of the posterior midbrain in mice lacking Pax5/BSAP. Cell 79:901–12 94. Nutt SL, Morrison AM, Dorﬂer P, Rolink A, Busslinger M. 1998. Identiﬁcation of BSAP (Pax-5) target genes in early B-cell development by loss- and gain-of-function experiments. EMBO J. 17:2319–33 95. Nutt SL, Heavey B, Rolink AG, Busslinger M. 1999. Commitment to the B-lymphoid lineage depends on the transcription factor Pax5. Nature 401:556–62 96. Souabni A, Cobaleda C, Schebesta M, Busslinger M. 2002. Pax5 promotes B lymphopoiesis and blocks T cell development by repressing Notch1. Immunity 17:781–93 97. Chiang MY, Monroe JG. 1999. BSAP/Pax5A expression blocks survival and expansion of early myeloid cells implicating its involvement in maintaining commitment to the B-lymphocyte lineage. Blood 94:3621–32 98. Radtke F, Wilson A, Stark G, Bauer M, van Meerwijk J, et al. 1999. Deﬁcient T cell fate speciﬁcation in mice with an induced inactivation of Notch1. Immunity 10:547–58 99. Ting CN, Olson MC, Barton KP, Leiden JM. 1996. Transcription factor GATA-3 is required for development of the T-cell lineage. Nature 384:474–78 100. Hendriks RW, Nawijn MC, Engel JD, van Doorninck H, Grosveld F, Karis A. 1999. Expression of the transcription factor GATA-3 is required for the development of the earliest T cell progenitors and correlates with stages of cellular proliferation in the thymus. Eur. J. Immunol. 29:1912–18 101. Pai SY, Truitt ML, Ting CN, Leiden JM, Glimcher LH, Ho IC. 2003. Critical roles for transcription factor GATA-3 in thymocyte development. Immunity 19:863–75 102. Pai SY, Truitt ML, Ho IC. 2004. GATA-3 deﬁciency abrogates the development and maintenance of T helper type 2 cells. Proc. Natl. Acad. Sci. USA 101:1993–98 103. Han H, Tanigaki K, Yamamoto N, Kuroda K, Yoshimoto M, et al. 2002. Inducible gene knockout of transcription factor recombination signal binding protein-J reveals its essential role in T versus B lineage decision. Int. Immunol. 14:637–45 104. Pui JC, Allman D, Xu L, DeRocco S, Karnell FG, et al. 1999. Notch1 expression in early lymphopoiesis inﬂuences B versus T lineage determination. Immunity 11:299–308 105. Schmitt TM, Zuniga-Pﬂucker JC. 2002. Induction of T cell development from hematopoietic progenitor cells by delta-like-1 in vitro. Immunity 17:749–56 106. Schmitt TM, Ciofani M, Petrie HT, Zuniga-Pﬂucker JC. 2004. Maintenance of T cell speciﬁcation and differentiation requires recurrent Notch receptor-ligand interactions. J. Exp. Med. 200:469–79 107. Taghon TN, David ES, Zuniga-Pﬂucker JC, Rothenberg EV. 2005. Delayed, asynchronous, and reversible T-lineage speciﬁcation induced by Notch/Delta signaling. Genes Dev. 19:965–78 www.annualreviews.org • Lymphoid-Myeloid Development

733

ARI

23 February 2006

15:47

108. Tremblay M, Herblot S, Lecuyer E, Hoang T. 2003. Regulation of pTα gene expression by a dosage of E2A, HEB, and SCL. J. Biol. Chem. 278:12680–87 109. Bain G, Engel I, Robanus Maandag EC, te Riele HP, Voland JR, et al. 1997. E2A deﬁciency leads to abnormalities in αβ T-cell development and to rapid development of T-cell lymphomas. Mol. Cell. Biol. 17:4782–91 110. Barndt R, Dai MF, Zhuang Y. 1999. A novel role for HEB downstream or parallel to the pre-TCR signaling pathway during αβ thymopoiesis. J. Immunol. 163:3331–43 111. Barndt RJ, Dai M, Zhuang Y. 2000. Functions of E2A-HEB heterodimers in T-cell development revealed by a dominant negative mutation of HEB. Mol. Cell. Biol. 20:6677– 85 112. Heemskerk MH, Blom B, Nolan G, Stegmann AP, Bakker AQ, et al. 1997. Inhibition of T cell and promotion of natural killer cell development by the dominant negative helix loop helix factor Id3. J. Exp. Med. 186:1597–602 113. Anderson MK, Hernandez-Hoyos G, Diamond RA, Rothenberg EV. 1999. Precise developmental regulation of Ets family transcription factors during speciﬁcation and commitment to the T cell lineage. Development 126:3131–48 114. Anderson MK, Weiss AH, Hernandez-Hoyos G, Dionne CJ, Rothenberg EV. 2002. Constitutive expression of PU.1 in fetal hematopoietic progenitors blocks T cell development at the pro-T cell stage. Immunity 16:285–96 115. Colucci F, Samson SI, DeKoter RP, Lantz O, Singh H, Di Santo JP. 2001. Differential requirement for the transcription factor PU.1 in the generation of natural killer cells versus B and T cells. Blood 97:2625–32 116. Yokota Y, Mansouri A, Mori S, Sugawara S, Adachi S, et al. 1999. Development of peripheral lymphoid organs and natural killer cells depends on the helix-loop-helix inhibitor Id2. Nature 397:702–6 117. Ikawa T, Fujimoto S, Kawamoto H, Katsura Y, Yokota Y. 2001. Commitment to natural killer cells requires the helix-loop-helix inhibitor Id2. Proc. Natl. Acad. Sci. USA 98:5164– 69 118. Pan L, Sato S, Frederick JP, Sun XH, Zhuang Y. 1999. Impaired immune responses and B-cell proliferation in mice lacking the Id3 gene. Mol. Cell. Biol. 19:5969–80 119. Rivera RR, Johns CP, Quan J, Johnson RS, Murre C. 2000. Thymocyte selection is regulated by the helix-loop-helix inhibitor protein, Id3. Immunity 12:17–26 120. Friedman AD. 2002. Transcriptional regulation of granulocyte and monocyte development. Oncogene 21:3377–90 121. DeKoter RP, Walsh JC, Singh H. 1998. PU.1 regulates both cytokine-dependent proliferation and differentiation of granulocyte/macrophage progenitors. EMBO J. 17:4456– 68 122. Zhang DE, Zhang P, Wang ND, Hetherington CJ, Darlington GJ, Tenen DG. 1997. Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein α-deﬁcient mice. Proc. Natl. Acad. Sci. USA 94:569– 74 123. Jones LC, Lin ML, Chen SS, Krug U, Hofmann WK, et al. 2002. Expression of C/EBPβ from the C/ebpα gene locus is sufﬁcient for normal hematopoiesis in vivo. Blood 99:2032– 36 124. Tanaka T, Akira S, Yoshida K, Umemoto M, Yoneda Y, et al. 1995. Targeted disruption of the NF-IL6 gene discloses its essential role in bacteria killing and tumor cytotoxicity by macrophages. Cell 80:353–61 125. Chen X, Liu W, Ambrosino C, Ruocco MR, Poli V, et al. 1997. Impaired generation of bone marrow B lymphocytes in mice deﬁcient in C/EBPβ. Blood 90:156–64

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

734

Laiosa

·

Stadtfeld

·

Graf

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

126. Fujiwara Y, Browne CP, Cunniff K, Goff SC, Orkin SH. 1996. Arrested development of embryonic red cell precursors in mouse embryos lacking transcription factor GATA-1. Proc. Natl. Acad. Sci. USA 93:12355–58 127. Pevny L, Simon MC, Robertson E, Klein WH, Tsai SF, et al. 1991. Erythroid differentiation in chimaeric mice blocked by a targeted mutation in the gene for transcription factor GATA-1. Nature 349:257–60 128. Takahashi S, Shimizu R, Suwabe N, Kuroha T, Yoh K, et al. 2000. GATA factor transgenes under GATA-1 locus control rescue germline GATA-1 mutant deﬁciencies. Blood 96:910–16 129. Shivdasani RA, Fujiwara Y, McDevitt MA, Orkin SH. 1997. A lineage-selective knockout establishes the critical role of transcription factor GATA-1 in megakaryocyte growth and platelet development. EMBO J. 16:3965–73 130. Tsang AP, Fujiwara Y, Hom DB, Orkin SH. 1998. Failure of megakaryopoiesis and arrested erythropoiesis in mice lacking the GATA-1 transcriptional cofactor FOG. Genes Dev. 12:1176–88 131. Tsang AP, Visvader JE, Turner CA, Fujiwara Y, Yu C, et al. 1997. FOG, a multitype zinc ﬁnger protein, acts as a cofactor for transcription factor GATA-1 in erythroid and megakaryocytic differentiation. Cell 90:109–19 132. Zhou M, Ouyang W, Gong Q, Katz SG, White JM, et al. 2001. Friend of GATA-1 represses GATA-3-dependent activity in CD4+ T cells. J. Exp. Med. 194:1461–71 133. Snapper CM, Rosas FR, Zelazowski P, Moorman MA, Kehry MR, et al. 1996. B cells lacking RelB are defective in proliferative responses, but undergo normal B cell maturation to Ig secretion and Ig class switching. J. Exp. Med. 184:1537–41 134. Weih F, Carrasco D, Durham SK, Barton DS, Rizzo CA, et al. 1995. Multiorgan inﬂammation and hematopoietic abnormalities in mice with a targeted disruption of RelB, a member of the NF-κB/Rel family. Cell 80:331–40 135. Wu L, D’Amico A, Winkel KD, Suter M, Lo D, Shortman K. 1998. RelB is essential for the development of myeloid-related CD8α− dendritic cells but not of lymphoid-related CD8α+ dendritic cells. Immunity 9:839–47 136. Burkly L, Hession C, Ogata L, Reilly C, Marconi LA, et al. 1995. Expression of relB is required for the development of thymic medulla and dendritic cells. Nature 373:531–36 137. Hacker C, Kirsch RD, Ju XS, Hieronymus T, Gust TC, et al. 2003. Transcriptional proﬁling identiﬁes Id2 function in dendritic cell development. Nat. Immunol. 4:380–86 138. Wu L, Nichogiannopoulou A, Shortman K, Georgopoulos K. 1997. Cell-autonomous defects in dendritic cell populations of Ikaros mutant mice point to a developmental relationship with the lymphoid lineage. Immunity 7:483–92 139. Graf T, McNagny K, Brady G, Frampton J. 1992. Chicken “erythroid” cells transformed by the Gag-Myb-Ets-encoding E26 leukemia virus are multipotent. Cell 70:201–13 140. Kulessa H, Frampton J, Graf T. 1995. GATA-1 reprograms avian myelomonocytic cell lines into eosinophils, thromboblasts, and erythroblasts. Genes Dev. 9:1250–62 141. Nerlov C, Graf T. 1998. PU.1 induces myeloid lineage commitment in multipotent hematopoietic progenitors. Genes Dev. 12:2403–12 141a. Rekhtman N, Radparvar F, Evans T, Skoultchi AI. 1999. Direct interaction of hematopoietic transcription factors PU.1 and GATA-1: functional antagonism in erythroid cells. Genes Dev. 13:1398–411 141b. Nerlov C, Querfurth E, Kulessa H, Graf T. 2000. GATA-1 interacts with the myeloid PU.1 transcription factor and represses PU.1-dependent transcription. Blood 95:2543– 51 www.annualreviews.org • Lymphoid-Myeloid Development

735

ARI

23 February 2006

15:47

141c. Zhang P, Zhang X, Iwama A, Yu C, Smith KA, et al. 2000. PU.1 inhibits GATA-1 function and erythroid differentiation by blocking GATA-1 DNA binding. Blood 96:2641–48 141d. Stopka T, Amanatullah DF, Papetti M, Skoultchi AI. 2005. PU.1 inhibits the erythroid program by binding to GATA-1 on DNA and creating a repressive chromatin structure. EMBO J. 24:3712–23 142. Zhang P, Behre G, Pan J, Iwama A, Wara-Aswapati N, et al. 1999. Negative cross-talk between hematopoietic regulators: GATA proteins repress PU.1. Proc. Natl. Acad. Sci. USA 96:8705–10 143. Okuno Y, Huang G, Rosenbauer F, Evans EK, Radomska HS, et al. 2005. Potential autoregulation of transcription factor PU.1 by an upstream regulatory element. Mol. Cell. Biol. 25:2832–45 144. Yu C, Cantor AB, Yang H, Browne C, Wells RA, et al. 2002. Targeted deletion of a high-afﬁnity GATA-binding site in the GATA-1 promoter leads to selective loss of the eosinophil lineage in vivo. J. Exp. Med. 195:1387–95 145. Iwasaki H, Mizuno S, Wells RA, Cantor AB, Watanabe S, Akashi K. 2003. GATA-1 converts lymphoid and myelomonocytic progenitors into the megakaryocyte/erythrocyte lineages. Immunity 19:451–62 146. Heyworth C, Pearson S, May G, Enver T. 2002. Transcription factor-mediated lineage switching reveals plasticity in primary committed progenitor cells. EMBO J. 21:3770–81 147. Galloway JL, Wingert RA, Thisse C, Thisse B, Zon LI. 2005. Loss of Gata1 but not Gata2 converts erythropoiesis to myelopoiesis in zebraﬁsh embryos. Dev. Cell 8:109–16 148. Rhodes J, Hagen A, Hsu K, Deng M, Liu TX, et al. 2005. Interplay of Pu.1 and Gata1 determines myelo-erythroid progenitor cell fate in zebraﬁsh. Dev. Cell 8:97–108 149. Cantor AB, Orkin SH. 2001. Hematopoietic development: a balancing act. Curr. Opin. Genet. Dev. 11:513–19 150. Walsh JC, DeKoter RP, Lee HJ, Smith ED, Lancki DW, et al. 2002. Cooperative and antagonistic interplay between PU.1 and GATA-2 in the speciﬁcation of myeloid cell fates. Immunity 17:665–76 151. Querfurth E, Schuster M, Kulessa H, Crispino JD, Doderlein G, et al. 2000. Antagonism between C/EBPβ and FOG in eosinophil lineage commitment of multipotent hematopoietic progenitors. Genes Dev. 14:2515–25 152. Welch JJ, Watts JA, Vakoc CR, Yao Y, Wang H, et al. 2004. Global regulation of erythroid gene expression by transcription factor GATA-1. Blood 104:3136–47 153. Kummalue T, Friedman AD. 2003. Cross-talk between regulators of myeloid development: C/EBPα binds and activates the promoter of the PU.1 gene. J. Leukoc. Biol. 74:464–70 154. Xie H, Ye M, Feng R, Graf T. 2004. Stepwise reprogramming of B cells into macrophages. Cell 117:663–76 155. Crispino JD, Lodish MB, MacKay JP, Orkin SH. 1999. Use of altered speciﬁcity mutants to probe a speciﬁc protein-protein interaction in differentiation: the GATA-1:FOG complex. Mol. Cell 3:219–28 156. Ness SA, Kowenz-Leutz E, Casini T, Graf T, Leutz A. 1993. Myb and NF-M: combinatorial activators of myeloid genes in heterologous cell types. Genes Dev. 7:749–59 157. McNagny K, Graf T. 2002. Making eosinophils through subtle shifts in transcription factor expression. J. Exp. Med. 195:F43–47 158. Wang X, Crispino JD, Letting DL, Nakazawa M, Poncz M, Blobel GA. 2002. Control of megakaryocyte-speciﬁc gene expression by GATA-1 and FOG-1: role of Ets transcription factors. EMBO J. 21:5225–34

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

736

Laiosa

·

Stadtfeld

·

Graf

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

ARI

23 February 2006

15:47

159. Tenen DG, Hromas R, Licht JD, Zhang DE. 1997. Transcription factors, normal myeloid development, and leukemia. Blood 90:489–519 160. Lindeman GJ, Adams JM, Cory S, Harris AW. 1994. B-lymphoid to granulocytic switch during hematopoiesis in a transgenic mouse strain. Immunity 1:517–27 161. Klinken SP, Alexander WS, Adams JM. 1988. Hemopoietic lineage switch: v-raf oncogene converts Emu-myc transgenic B cells into macrophages. Cell 53:857–67 161a. Heavey B, Charalambous C, Cobaleda C, Busslinger M. 2003. Myeloid lineage switch of Pax5 mutant but not wild-type B cell progenitors by C/EBPα and GATA factors. EMBO J. 22:3887–97 162. Lefebvre JM, Haks MC, Carleton MO, Rhodes M, Sinnathamby G, et al. 2005. Enforced expression of Spi-B reverses T lineage commitment and blocks β-selection. J. Immunol. 174:6184–94 162a. Dionne CJ, Tse KY, Weiss AH, Franco CB, Wiest DL, et al. 2005. Subversion of T lineage commitment by PU.1 in a clonal cell line system. Dev. Biol. 280:448–66 162b. Hsu LY, Liang HE, Johnson K, Kang C, Schlissel MS. 2004. Pax5 activates immunoglobulin heavy chain V to DJ rearrangement in transgenic thymocytes. J. Exp. Med. 199:825– 30 163. Rolink AG, Nutt SL, Melchers F, Busslinger M. 1999. Long-term in vivo reconstitution of T-cell development by Pax5-deﬁcient B-cell progenitors. Nature 401:603–6 164. Horcher M, Souabni A, Busslinger M. 2001. Pax5/BSAP maintains the identity of B cells in late B lymphopoiesis. Immunity 14:779–90 165. Mikkola I, Heavey B, Horcher M, Busslinger M. 2002. Reversion of B cell commitment upon loss of Pax5 expression. Science 297:110–13 166. Nutt SL, Urbanek P, Rolink A, Busslinger M. 1997. Essential functions of Pax5 (BSAP) in pro-B cell development: difference between fetal and adult B lymphopoiesis and reduced V-to-DJ recombination at the IgH locus. Genes Dev. 11:476–91 167. Zhuang Y, Soriano P, Weintraub H. 1994. The helix-loop-helix gene E2A is required for B cell formation. Cell 79:875–84 168. Ikawa T, Kawamoto H, Wright LY, Murre C. 2004. Long-term cultured E2A-deﬁcient hematopoietic progenitor cells are pluripotent. Immunity 20:349–60 168a. Rosenbauer F, Owens BM, Yu L, Tumang JR, Steidl U, et al. 2006. Lymphoid cell growth and transformation are suppressed by a key regulatory element of the gene encoding PU.1. Nat. Genet. 38:27–37 169. Rosenbauer F, Wagner K, Kutok JL, Iwasaki H, Le Beau MM, et al. 2004. Acute myeloid leukemia induced by graded reduction of a lineage-speciﬁc transcription factor, PU.1. Nat. Genet. 36:624–30 170. DeKoter RP, Singh H. 2000. Regulation of B lymphocyte and macrophage development by graded expression of PU.1. Science 288:1439–41 170a. Godin I, Cumano A. 2002. The hare and the tortoise: an embryonic haematopoietic race. Nat. Rev. Immunol. 2:593–604 171. Hardy RR, Hayakawa K. 2001. B cell development pathways. Annu. Rev. Immunol. 19:595–621 172. MacDonald HR, Radtke F, Wilson A. 2001. T cell fate speciﬁcation and αβ/γδ lineage commitment. Curr. Opin. Immunol. 13:219–24 173. Ceredig R, Rolink T. 2002. A positive look at double-negative thymocytes. Nat. Rev. Immunol. 2:888–97 174. Hu M, Krause D, Greaves M, Sharkis S, Dexter M, et al. 1997. Multilineage gene expression precedes commitment in the hemopoietic system. Genes Dev. 11:774–85 www.annualreviews.org • Lymphoid-Myeloid Development

737

ARI

23 February 2006

15:47

175. Miyamoto T, Iwasaki H, Reizis B, Ye M, Graf T, et al. 2002. Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment. Dev. Cell 3:137–47 176. Ye M, Iwasaki H, Laiosa CV, Stadtfeld M, Xie H, et al. 2003. Hematopoietic stem cells expressing the myeloid lysozyme gene retain long-term, multilineage repopulation potential. Immunity 19:689–99 177. Delassus S, Titley I, Enver T. 1999. Functional and molecular analysis of hematopoietic progenitors derived from the aorta-gonad-mesonephros region of the mouse embryo. Blood 94:1495–503 178. Forsberg EC, Prohaska SS, Katzman S, Heffner GC, Stuart JM, Weissman IL. 2005. Differential expression of novel potential regulators in hematopoietic stem cells. PLoS Genet. 1:e28 179. Gounari F, Aifantis I, Martin C, Fehling HJ, Hoeﬂinger S, et al. 2002. Tracing lymphopoiesis with the aid of a pTα-controlled reporter gene. Nat. Immunol. 3:489–96 180. Gilliland DG, Grifﬁn JD. 2002. The roles of FLT3 in hematopoiesis and leukemia. Blood 100:1532–42 181. Weng AP, Ferrando AA, Lee W, Morris JP, Silverman LB, et al. 2004. Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science 306:269– 71 182. He L, Hannon GJ. 2004. MicroRNAs: small RNAs with a big role in gene regulation. Nat. Rev. Genet. 5:522–31 183. Ambros V. 2004. The functions of animal microRNAs. Nature 431:350–55 184. Zhao Y, Samal E, Srivastava D. 2005. Serum response factor regulates a muscle-speciﬁc microRNA that targets Hand2 during cardiogenesis. Nature 436:214–20 185. Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, et al. 2005. MicroRNA expression proﬁles classify human cancers. Nature 435:834–38 186. Chen CZ, Li L, Lodish HF, Bartel DP. 2004. MicroRNAs modulate hematopoietic lineage differentiation. Science 303:83–86 187. Muller C, Leutz A. 2001. Chromatin remodeling in development and differentiation. Curr. Opin. Genet. Dev. 11:167–74 188. Johnson K, Pﬂugh DL, Yu D, Hesslein DG, Lin KI, et al. 2004. B cell-speciﬁc loss of histone 3 lysine 9 methylation in the V(H) locus depends on Pax5. Nat. Immunol. 5:853–61 189. Marecki S, McCarthy KM, Nikolajczyk BS. 2004. PU.1 as a chromatin accessibility factor for immunoglobulin genes. Mol. Immunol. 40:723–31 190. Lutterbach B, Westendorf JJ, Linggi B, Isaac S, Seto E, Hiebert SW. 2000. A mechanism of repression by acute myeloid leukemia-1, the target of multiple chromosomal translocations in acute leukemia. J. Biol. Chem. 275:651–56 191. Lee GR, Fields PE, Flavell RA. 2001. Regulation of IL-4 gene expression by distal regulatory elements and GATA-3 at the chromatin level. Immunity 14:447–59 192. Muller C, Calkhoven CF, Sha X, Leutz A. 2004. The CCAAT enhancer-binding protein α (C/EBPα) requires a SWI/SNF complex for proliferation arrest. J. Biol. Chem. 279:7353–58 193. Tagoh H, Schebesta A, Lefevre P, Wilson N, Hume D, et al. 2004. Epigenetic silencing of the c-fms locus during B-lymphopoiesis occurs in discrete steps and is reversible. EMBO J. 23:4275–85 194. Akashi K, He X, Chen J, Iwasaki H, Niu C, et al. 2003. Transcriptional accessibility for genes of multiple tissues and hematopoietic lineages is hierarchically controlled during early hematopoiesis. Blood 101:383–89

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-23

738

Laiosa

·

Stadtfeld

·

Graf

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:705-738. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet1 and Dennis R. Burton1,2 1

Department of Immunology, 2 Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:739–69 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090557 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0739$20.00

Key Words AIDS, neutralizing antibodies, vaccine design, antigen engineering

Abstract The glycoprotein (gp) 120 subunit is an important part of the envelope spikes that decorate the surface of HIV-1 and a major target for neutralizing antibodies. However, immunization with recombinant gp120 does not elicit neutralizing antibodies against multiple HIV-1 isolates (broadly neutralizing antibodies), and gp120 failed to demonstrate vaccine efﬁcacy in recent clinical trials. Ongoing crystallographic studies of gp120 molecules from HIV-1 and SIV increasingly reveal how conserved regions, which are the targets of broadly neutralizing antibodies, are concealed from immune recognition. Based on this structural insight and that from studies of antibody structures, a number of strategies are being pursued to design immunogens that can elicit broadly neutralizing antibodies to gp120. These include (a) the construction of mimics of the viral envelope spike and (b) the design of antigens speciﬁcally tailored to induce broadly neutralizing antibodies.

739

ANRV270-IY24-24

ARI

15 February 2006

3:9

INTRODUCTION

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

HIV: human immunodeﬁciency virus Viral envelope: the membrane surrounding a virus particle, usually containing host cell membrane lipids and proteins as well as virus-encoded (glyco)proteins gp: glycoprotein

The antibody response to HIV-1 in vivo is directed against several viral proteins. However, essentially all neutralizing antibodies are directed toward the viral envelope spike, in particular the surface unit glycoprotein (gp) 120 (1, 2), which is anchored to the viral surface by gp41, the transmembrane unit (2). Because of its surface-exposed location, gp120 seemed a natural ﬁrst choice as a subunit vaccine candidate (3). However, it was soon noted that immunization with recombinant monomeric gp120 elicited antibodies that could neutralize viruses that were neutralization-sensitive following extensive passage in immortalized T cell line cultures, but not viruses grown in limited passage in peripheral blood mononuclear cells (PBMCs) (4). The latter are more representative of circulating “primary” viruses. The results from recently completed phase III

HUMORAL AND CELLULAR IMMUNE RESPONSES IN PROTECTION AGAINST VIRAL INFECTION A common feature of protective viral vaccines is their ability to elicit neutralizing antibodies. Neutralizing antibodies generally reduce the severity of primary infection and thus prevent the onset of disease. Neutralization can be achieved by the binding of antibodies to free virus particles. However, antibodies may also mediate activities against infected cells that display viral antigens such as antibody-dependent cellular cytotoxicity [the killing of antibody-coated target cells by immune cells that bear Fc receptors, e.g., natural killer (NK) cells] or complement-dependent cytotoxicity (lysis of antibody-coated target cells by complement). Antibodies are not always sufﬁcient, and vaccine-induced T cell–mediated immunity may be required as well for protective efﬁcacy. In such cases, the combination of antibody and T cell may protect at a level that is not achievable with a single component. Some viruses, such as HIV, have evolved elaborate mechanisms by which to protect conserved regions from effective recognition by humoral and cellular immune responses. Understanding and overcoming these barriers represent major scientiﬁc challenges for vaccinologists.

740

Pantophlet

·

Burton

clinical trials show that antibodies elicited by recombinant-gp120-based vaccine candidates indeed do not protect vaccine recipients against HIV-1 infection or inﬂuence disease progression (5, 6). The general inability of recombinant gp120 to elicit cross-neutralizing (or broadly neutralizing) antibodies to primary viruses has sharpened interest in understanding the differences between the structure of gp120 in its recombinant form and its structure in the context of the envelope spike. Ample evidence suggests that the gp120 conformation in the context of the envelope spike is the crucial structure recognized by neutralizing antibodies (7–13). Here, we summarize recent insights gained in understanding the structural conformation of gp120, particularly as it relates to exposure of antibody epitopes on the viral spike. The recently solved crystal structure of the unliganded gp120 core of simian immunodeﬁciency virus (SIV) (14) is an important advance in this regard. Although a few caveats are associated with this structure, it nevertheless allows us to appreciate better some of the mechanisms that HIV uses to avoid antibody binding to conserved regions on gp120, which are the targets of broadly neutralizing antibodies. We also highlight two strategies that are currently being pursued to induce broadly neutralizing anti-gp120 antibodies: reconstitution of the viral envelope spike and tailored antigen design. These strategies are based, in part, on our current understanding of interactions between the viral envelope and presently known broadly neutralizing antibodies. This overview represents an update of a previous review on the biological aspects of gp120 structural features (15).

STRUCTURAL ORGANIZATION AND TOPOLOGICAL FEATURES OF GP120 The HIV envelope spike is formed as a complex between gp120 and gp41 (2). The gp120 unit mediates attachment of the virus

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

to the target cell, whereas gp41 is required for the fusion of virus and target cell membranes. During HIV infection, the viral envelope spike is ﬁrst synthesized as a single polypeptide precursor (2). In the Golgi, the protein subsequently oligomerizes and undergoes extensive glycosylation. The glycosylation process, which is required for proper folding and conformational stability of the envelope glycoprotein (16), mainly involves the attachment of N-linked highmannose-type oligosaccharides to the protein backbone. As the glycoprotein is transported through the Golgi, accessible glycan moieties are trimmed and modiﬁed by various cellular enzymes (2). These modiﬁcations generate so-called complex-type oligosaccharides; glycans that are relatively inaccessible to modifying enzymes remain as high-mannosetype glycans (17). The resulting glycoprotein, which has a molecular mass of ∼160 kDa, is cleaved in the trans-Golgi network by furin or equivalent endoproteases into gp120 and gp41 (2). The gp120-gp41 complexes, which remain associated through weak noncovalent interactions, are initially expressed at the surface of infected cells. During the HIV budding process, the gp120-gp41 complexes are then incorporated into the virus envelope and displayed on its surface as viral spikes (2).

HIV-1 gp41 cores resemble the transmembrane proteins of other viruses that have been shown to display trimeric envelope spikes (20–22). There is, however, also evidence that other envelope species may be present on the surface of HIV-1. For example, atomic force microscopy analyses have failed to reveal any uniform trimeric envelope species on the surface of virions (23). Also, in recent studies, it has been shown that viruses can be captured onto ELISA plate wells using antibodies that are unable to neutralize viral particles in solution (12, 24). Taken together, these observations suggest that, although trimers may likely represent the functional envelope spike, both functional and nonfunctional forms of the envelope may be present on the virion surface. These nonfunctional envelope entities may be monomers, dimers, or tetramers and could possibly arise as the result of (a) the dissociation of functional gp120-gp41 complexes, which could perhaps cause gp120 to be shed from the viral surface, or (b) inefﬁcient trimerization of the spike in the Golgi (2, 25, 26).

Subunit vaccine: a vaccine that contains only the portion of the pathogen that is considered necessary to induce protection against infection Neutralization: the loss of infectivity that ensues when antibody molecule(s) bind to a virus particle PBMC: peripheral blood mononuclear cell Primary virus: virus that has not been adapted to grow in laboratory culture cell lines SIV: simian immunodeﬁciency virus

Organization of gp120 on the Viral Surface Knowledge of the oligomeric structure of the gp120-gp41 complex is important for vaccine design strategies, as we highlight below. Experimental evidence suggests that the functional unit of the envelope spike is a heterodimeric trimer complex of gp120 and gp41. For example, a recent electron tomography study revealed structures on the surface of negatively stained virions of SIV and HIV-1 that appear to be tri-lobed envelope glycoproteins (18) (Figure 1). Furthermore, the HIV core matrix that interacts with gp41 is organized in a trimeric conﬁguration (19), and the crystal structures of

Figure 1 3D tomograms of SIV and HIV pseudovirions. (a) Tomogram section from an SIV particle with a truncated cytoplasmic domain and a high level of envelope spike expression on the virion surface. (b) Tomogram of an HIV-1 particle expressing the full-length envelope glycoprotein. Representative tri-lobed structures, presumably of the envelope spikes, are indicated (white arrows). Tomograms courtesy of Drs. Ping Zhu and Ken Roux (Florida State University, Tallahassee, Florida). www.annualreviews.org • GP120-Targeted Vaccine Design

741

ANRV270-IY24-24

ARI

15 February 2006

3:9

Topology of gp120

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Affinity: the strength of binding of one molecule to another at a single site

Based on comparative sequence analyses, gp120 is divided into ﬁve conserved (C1– C5) and ﬁve variable (V1–V5) segments (27, 28). Prior to obtaining the gp120 core crystal structures, investigators deduced many topological features of monomeric and oligomeric gp120 from antibody binding/competition experiments and mutagenesis studies (29–35). Thus, the C1 and C5 regions were reasoned to be the main areas on gp120 for contact with gp41, as these regions are accessible to antibody on monomeric gp120 but not on gp120gp41 complexes (29, 31, 35). Major segments of the C2, C3, and C4 regions were suggested to form a buried, relatively hydrophobic core within the gp120 molecule (31, 32). It was proposed that this gp120 core harbors several discontinuous neutralizing antibody epitopes that overlap the binding sites for CD4, the primary HIV receptor, and the coreceptor (31–34). In contrast to the conserved regions, the variable regions (in particular, V1, V2, and V3) were argued to be well exposed on the surface of monomeric gp120 (31). Deletion of V1/V2 and V3 generally increases the binding afﬁnity of antibodies to epitopes that overlap the binding sites for CD4 and the coreceptor, which suggests that these variable regions may shield conserved epitopes from efﬁcient antibody recognition (36–39). For the V4 and V5 variable regions, no deﬁnitive role has been ascribed; although deletion of the V4 region has been shown to disrupt gp160 folding (32, 38), V4 also seems to tolerate insertion of foreign antibody epitopes (40). Determination of the structures of gp120 molecules from HIV and SIV in recent years has supported many of the interpretations made from these earlier observations.

MOLECULAR STRUCTURE OF GP120 At present, four crystal structures of HIV-1 gp120 and one of SIV gp120 have been reported (14, 41–43). All ﬁve structures are of 742

Pantophlet

·

Burton

the gp120 core; i.e., the structures lack the V1/V2 and V3 variable regions, and the N and C termini are truncated. The HIV-1 gp120 structures were determined in complex with the D1D2 fragment of CD4 or CD4 mimics, whereas the SIV gp120 structure was solved unliganded. Because all crystal structures determined so far are of monomeric gp120, they may not adequately represent the structure of oligomeric gp120 on the virus. Despite these and other caveats that we discuss later, the structures do advance insight into the conformational ﬂexibility of monomeric gp120 as well as the locations of receptor-binding sites and putative antibody epitopes on gp120.

Crystal Structures of HIV-1 and SIV gp120 Cores in Complex with CD4 The structure of the gp120 core from the laboratory-adapted virus HXB2 was the ﬁrst determined in complex with CD4 (Figure 2) (42). HXB2 is highly sensitive to antibody neutralization. The second crystal structure was that of CD4 complexed with the gp120 core of the primary virus YU-2 (41), which exhibits a marked resistance to antibody neutralization. The two gp120 core structures are virtually superimposable (41), which is consistent with earlier predictions that the ability of HIV to resist antibody neutralization is likely to be manifested mainly in the context of the gp120 quaternary structure on the viral surface rather than in the monomeric gp120 form (7). Based on these CD4-bound structures, gp120 is organized into three general areas (Figure 3a): (a) the inner domain, (b) the outer domain, and (c) the bridging sheet (44). Inner domain, outer domain, and bridging sheet. The inner domain is formed mainly by the C1 and C5 regions and is largely devoid of glycans (42, 44), which strongly supports the proposition that C1 and C5 function as the major contact interface with the gp41 transmembrane unit. The outer domain, in contrast, is largely covered by glycans (42,

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

44). Modeling of the gp120 oligomer suggests that these glycans likely cover large sections of the outer surface of the spike to lower its overall immunogenicity (44, 46). The glycans themselves are poor targets for antibodies because of their heterogeneous expression on the virus and because they are produced by the glycosylation machinery of the host cell and, thus, are self molecules. Comparison of the liganded gp120 structure of HIV-1 (Figure 3a) and the unliganded gp120 structure of SIV (Figure 3b) shows that the respective outer domains are highly similar (14). However, the conformation of the inner domain in the unliganded structure deviates signiﬁcantly from the conformation in the liganded structures. This observation suggests that the inner domain may have signiﬁcant conformational ﬂexibility in the absence of CD4. Comparison of the inner domain substructures in the unliganded and liganded core structures suggests that, upon CD4 binding, these substructures are repositioned somewhat independently of each other, rather than a shift of the inner domain as a single unit (14). The large structural rearrangements associated with the repositioning of inner domain substructures seem consistent with the large negative entropy and enthalphy changes measured by isothermal titration microcalorimetry (47). It is noteworthy that the majority of gp120 conformational shifts resulting from CD4 binding are in the portion of gp120 that interacts with gp41. Thus, these conformational changes may be necessary to lock the coreceptor-binding site (CoRbs) into a ﬁxed conformation and also trigger gp41 into initiating the ﬁrst steps in the fusion process (2). The conformational changes that occur within the inner domain also affect the formation of the bridging sheet, which links the inner and outer domains. In the CD4liganded gp120 conformation, the bridging sheet is folded into a compact antiparallel, four-stranded β-sheet (β2-β3 and β20-β21) (42, 44). However, in the unliganded structure, the β-strands that constitute the bridging sheet lie separated in pairs at a distance

Figure 2 Crystal structure of HIV-1 gp120 complexed to CD4 and an antibody antigen-binding fragment (Fab). The gp120 core of HXB2 (Protein Data Bank ID 1G9M) (gray), CD4 (orange), the antibody heavy chain (H) (blue) and light chain (L) (green) are shown. Figure adapted from Kwong et al. (42). All ﬁgures were prepared with RasTop (45).

of approximately 20 A˚ (14); the two β-strands (β2-β3) that constitute the V1/V2 stem are located in the vicinity of the inner domain, whereas the other two strands (β20-β21) are situated near the outer domain in approximately the same location as they are on the liganded structure. Conformational changes that occur within the inner domain upon CD4 binding would result in a 40 A˚ shift of the V1/V2 stem to form the bridging sheet (14). However, molecular modeling of the liganded and unliganded gp120 structures suggests that the unliganded structure may be one of many conformations that gp120 may adopt in the www.annualreviews.org • GP120-Targeted Vaccine Design

CoRbs: coreceptor-binding site

743

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

Figure 3 Comparison of the crystal structures of HIV-1 and SIV gp120 core. (a) Structure of the CD4-liganded HIV-1 gp120 core (HXB2), viewed from the perspective of CD4. The gp120 inner domain (blue), outer domain (yellow), and bridging sheet (orange) are shown. The locations of various gp120 regions are also denoted. (b) Structure of the unliganded SIV gp120 core (Protein Data Bank ID 2BF1), viewed from the perspective of CD4 as in (a). (c) HIV core gp120 in same orientation as (a), depicting CD4 contact residues (orange) and residues that inﬂuence coreceptor binding (green). (d ) SIV gp120 core in the same orientation as in (b), colored according to the scheme for HIV gp120 in (c).

absence of CD4 (48). In fact, the models suggest that the β2-β3 strands may oscillate from the conformation observed in the unliganded structure to a conformation resembling the CD4-bound structure via a series of intermediate conformers (48). Epitope-mapping studies of antibodies to the CD4-binding site (CD4bs) also suggest that gp120 can adopt a conformation resembling the CD4-bound form relative to the conformation of the unliganded structure. We discuss this in more detail in the subsequent sections.

CD4bs: CD4-binding site

The CD4-binding site. The binding site for CD4 on the liganded gp120 structure is 744

Pantophlet

·

Burton

formed by the interface between the inner domain, bridging sheet, and outer domain (42, 44). At the center of this interface lies a hydrophobic cavity that has been dubbed the Phe43 cavity (Figure 3a) (42). However, most of the CD4 contact residues are located on the outer domain of the liganded HIV-1 gp120 structures and form a contiguous binding region (Figure 3c). On the unliganded SIV gp120 structure no such region is discernible (Figure 3d ) (14), assuming that equivalent residues in SIV and HIV-1 contact CD4. On the unliganded SIV gp120 core, many of the residues that are presumed to contact CD4 upon complexation with gp120

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

are located near or within a long cavity that is formed primarily by portions of the inner and outer domains and the β20-β21 segment of the bridging sheet. The location of these conserved residues likely minimizes their immediate recognition by antibodies, while preserving the ability to contact CD4. In this regard, the curved structure of the D1D2 fragment of CD4 is particularly noteworthy (Figure 2); it permits CD4 to curl over the outer domain, so residues located near or within the cavity formed by the inner and outer domains can be reached. Given that the CD4bs is not coherently present on the unliganded structure, it indeed seems likely that gp120 transiently samples conformations that are reﬂective of the liganded structure; upon interaction with CD4, the gp120 structure is locked in the bound conformation. The coreceptor-binding site. The region that is important for the interaction with the β-chemokine receptor CCR5 has been mapped to residues in the bridging sheet and near the V3 stem (49, 50). These residues lie close together on the liganded HIV-1 gp120 structure, but the equivalent residues on the unliganded SIV gp120 structure are separated into two areas (Figure 3c,d ) (14). These differences are consistent with the notion that CD4 binding is required to lock these areas into a contiguous binding site. The fact that the coreceptor site is not presented until after CD4 binding suggests that the site may be susceptible to antibody recognition. Several studies have shown that HIV strains that do not require CD4 for entry are, in fact, highly sensitive to antibody neutralization (51–54). Owing to neutralizing antibody–driven selection pressure in vivo, the prevalence of such viruses during infection is likely low. However, in the absence of circulating neutralizing antibodies, e.g., in the central nervous system, such viruses may occur more frequently (54, 55). Potential caveats regarding the HIV and SIV gp120 structures are pointed out in the previous section. We note here also that the

unliganded SIV gp120 structure is derived from strain SIVmac32H. This simian virus is able to infect CD4-negative target cells in vitro with medium efﬁciency (56). The gp120 from this strain may thus harbor certain structural features that are not observed in the gp120 of HIV-1 or SIV strains that require CD4 for entry. Crystallization of gp120s from further SIV and HIV strains, including gp120s from viruses that are entirely CD4 independent, will provide further insight into potential differences between the gp120 structures of these viruses.

Protomer: subunits from which a larger structure is built; for example, the tubulin heterodimer is the protomer for microtubule assembly

VIRAL DEFENSE MECHANISMS FOR ANTIBODY EVASION: STRUCTURAL CONSIDERATIONS The HIV and SIV structures provide valuable insight into the locations of receptor binding sites on gp120, as we summarize above. The structures also allow for a better understanding of some of HIV’s defense mechanisms, in particular how structural features may reduce antibody recognition of conserved regions on gp120.

Variable Regions Previous studies had suggested that the CoRbs is masked by the V1/V2 variable regions (36, 37); anti-CoRbs antibodies typically require CD4 for high-afﬁnity binding to wild-type gp120, but removal of V1/V2 allows these antibodies to bind their epitopes in the absence of CD4 with the same high afﬁnity as when binding to wild-type gp120 in the presence of CD4. The locations of the V1/V2 stem in the CD4-liganded and -unliganded gp120 structures relative to the coreceptor site suggests that V1/V2 may not be close enough to obstruct the CoRbs within the same gp120 molecule on the virus (14, 41). However, oligomeric modeling of the CD4liganded structure does suggest that V1/V2 from a given gp120 protomer may be in close proximity to the CoRbs on a neighboring www.annualreviews.org • GP120-Targeted Vaccine Design

745

ANRV270-IY24-24

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

mAb: monoclonal antibody

3:9

protomer to mask the coreceptor site (46). Within the gp120 monomer, removal of V1/V2 may allow the conformationally ﬂexible inner domain, in particular the β2-β3 strands, to more easily sample the conformations resembling that of the CD4-liganded structure in which the coreceptor site would be more fully presented (48). The locations of the variable region stems in the unliganded gp120 structure (Figure 3b) suggest that V1 and V2 may cover portions of the inner domain, including part of the long cavity that harbors the CD4-binding residues, and is formed by the inner and outer domains (14). Based on the unliganded structure, the V3 variable region may occlude the β20-β21 portion of the bridging sheet (14). The positions of the stems of V1, V2, and V3 also support the notion that these variable regions most likely do not interact with each other within a gp120 protomer, but do interact cooperatively with the variable regions of adjacent protomers within the context of the trimer on the viral surface (14, 46). Further support for this notion comes from recent studies involving a human monoclonal antibody (mAb), termed 2909, which binds exclusively to oligomeric gp120 on virus (55). We discuss this antibody in the context of its binding epitope in more detail in a section below.

Glycan Shield HIV can also use glycans to occlude antibody epitopes on gp120; 50% of the molecule is covered with carbohydrates that render the underlying protein surface invisible to the immune system. This somewhat static glycan shield has been dubbed the gp120 silent face (44). The virus can also shift the locations of its glycans in vivo (58–60). These observations have led to the proposition of a dynamic or so-called evolving glycan-shield model (59) by which, through the continuous repositioning of its glycans, HIV is able to escape type-speciﬁc neutralizing antibody responses. Thus, the evolved resistance is not a 746

Pantophlet

·

Burton

generalized one but, rather, a speciﬁc adaptation to the particular antibody response in each infected individual (59, 60). The repositioning of glycans may also affect local protein folding and, hence, also indirectly affect neutralizing antibody binding. In addition to being involved in blocking neutralizing antibody responses, glycan repositioning also may compensate for conformational changes in the envelope glycoprotein caused by localized amino acid changes directly related to virus escape from neutralizing antibody.

Entropic Masking It has been postulated that the productive interaction of antibodies with conserved regions on the HIV-1 envelope spike, in particular with the receptor binding sites, may be limited as a result of intrinsic entropic barriers that are imposed upon antibodies within the context of the viral spike (61). Large changes in entropy are generally observed upon binding of nonneutralizing or weakly neutralizing antibodies to receptor sites on monomeric gp120, whereas small changes in binding entropy are typically observed with broadly neutralizing antibodies (47, 61). Consequently, it has been postulated that the failure of nonneutralizing antibodies to bind with high afﬁnity to their respective epitopes on virion-associated oligomeric gp120 may be due, at least in part, to the inability of these antibodies to achieve the necessary entropic changes that are required for their high-afﬁnity interaction (61); neutralizing antibodies, in contrast, do not incur such an entropic penalty because they require relatively less conformational reorganization within the gp120 molecule for efﬁcient binding. Functional oligomeric gp120 is likely more conformationally ﬁxed owing to interaction with neighboring gp120 protomers within the spike (44), which supports the notion of entropic masking. However, whether entropic masking indeed plays a substantial role in limiting efﬁcient binding of nonneutralizing

ANRV270-IY24-24

ARI

15 February 2006

3:9

antibodies to oligomeric gp120 remains to be determined experimentally.

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Substitution of Nonessential Residues Most recently, the structures of two CD4 mimics in complex with the core structure of YU-2 gp120 were solved (43). Although the mimics contact fewer residues on gp120 than does CD4, the conformation of gp120 is essentially the same as when gp120 is bound to CD4. For one mimic in particular, only half of the residues contacted by CD4 are also contacted by the mimic (43). Thus, not all gp120 residues that are in contact with CD4 in the gp120-CD4 complex are required to induce conformational changes in gp120; i.e., the functional CD4-binding epitope is smaller than its structural epitope. Consequently, antibody recognition of the CD4bs may be easily circumvented by the virus if the ensemble of residues that are critical for antibody binding do not closely match those that are essential for CD4 binding, even if the epitope maps of both ligands overlap geometrically (62, 63).

BINDING SITES FOR SMALL MOLECULE INHIBITORS ON GP120 The structure of unliganded gp120 is helpful to vaccinologists, but could also be useful to those who seek to design small molecules that inhibit HIV entry. A small number of such inhibitors have been described, the most prominent of which is compound BMS-378806 (BMS-806) (64). The mechanism by which this compound blocks HIV entry is disputed. Initial studies had indicated that BMS-806 blocks the gp120-CD4 interaction (65, 66) and thus prevents the virus from attaching to target cells. However, other studies have suggested that the compound does not inhibit CD4 binding to gp120 but, instead, blocks steps that are required to initiate the fusion process by gp41 (67, 68). On the unliganded SIV gp120 structure, the putative binding

pocket of BMS-806 overlaps partially with the locations of residues that are presumed to contact CD4 upon binding (Figure 4), which supports the notion that BMS-806 may indeed prevent CD4 from binding to gp120. It is worth noting here that BMS-806 can be competed off gp120 at (low) micromolar concentrations of CD4 (65). This observation may explain, at least in part, why BMS806 did not appear to inhibit CD4 binding in select studies (67, 68), in which micromolar concentrations of CD4 were used in some of the assays. Clearly, further studies are required to address the inhibitory mechanism of compound BMS-806 more thoroughly. Nevertheless, the location of the putative BMS806 binding pocket indicates that the entire long hydrophobic cavity formed by the inner and outer domains is likely to be an attractive target for small-molecule inhibitors that can prevent binding of CD4 to gp120.

ANTIBODY EPITOPES ON HIV-1 GP120 Neutralizing antibody responses to HIV-1 in vivo are generally of limited breadth (2). However, a handful of human mAbs have been isolated that can neutralize broadly (69). Intense biochemical and immunological study of these antibodies, combined with insight gained from their crystal structures, provides important clues as to HIV’s vulnerability to antibody. For gp120, two human mAbs exist that can neutralize a wide range of primary viruses (70). The epitopes of these two antibodies, b12 and 2G12, are, therefore, attractive templates for vaccine design.

Conserved, Cross-Neutralizing gp120 Epitopes CD4bs: the b12 epitope. The antibody b12, which binds to an epitope that overlaps the CD4bs on gp120 (71, 72), has been studied extensively with regard to its antiviral activity in vitro and in vivo. In a recent comprehensive analysis involving a panel of 90 viruses, www.annualreviews.org • GP120-Targeted Vaccine Design

747

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

Figure 4 Putative binding site of the small-molecule inhibitor BMS-378806. Map of the sites of mutations ( yellow) that confer resistance to BMS-806 onto the core structure of SIV gp120. The side chains of residues corresponding to the resistance mutations are shown. The inset shows a close-up of the putative BMS-806 binding pocket. Residues that correspond to resistance mutations are noted, with amino acid substitutions that confer resistance in HIV-1 indicated in parentheses.

IC50 : inhibitory concentration (50%)

748

the antibody reached an inhibitory concentration [50% (IC50 )] with approximately half of the viruses within the test panel (70). As such, b12 is currently the most potently and broadly neutralizing anti-gp120 antibody. In vivo studies using macaque models of HIV infection have shown that b12 can also protect animals against viral challenge (73, 74). To gain insight into how b12 is able to neutralize broadly, its structure was determined and, at the time, docked onto the CD4-liganded structure of gp120 (75). Although b12 likely does not interact with the de facto CD4-liganded structure of HIV gp120, the docking model does suggest that b12 may interact with gp120 from an angle that would be permissive in the supposed orgaPantophlet

·

Burton

nization of gp120 in the context of the viral spike. This would explain how b12, in contrast to weakly neutralizing CD4bs antibodies, can effectively bind oligomeric gp120 on the viral surface as well as monomeric gp120 (10). Alanine substitutions that signiﬁcantly diminish b12 binding afﬁnity form a contiguous binding epitope on the CD4-liganded structure (63) (Figure 5a,b). However, when the equivalent substitutions are mapped onto the unliganded SIV gp120 structure, a dispersed pattern results (Figure 5c,d). Taken together, these observations suggest that b12 likely binds a gp120 conformation that resembles the CD4-bound structure. As discussed in the previous section, the notion that gp120 can adopt conformations resembling the

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

Figure 5 Locations of alanine substitutions that signiﬁcantly reduce b12 binding. Alanine mutations that caused >50% decrease in apparent b12 binding afﬁnity relative to the apparent antibody binding afﬁnity for wild-type gp120 of HIV-1JR-CSF were mapped onto the gp120 core structures of HIV-1 and SIV. (a) Alanine substitutions in the gp120 core structure of HIV-1, depicted as a ribbon diagram. (b) Alanine substitutions in the gp120 core structure of HIV-1, shown as a space-ﬁlling model. (c) Alanine substitutions in the gp120 core structure of SIV, depicted as a ribbon diagram. (d ) Alanine substitutions in the gp120 core structure of SIV, shown as a space-ﬁlling model. The perspectives are the same as in Figure 3.

CD4-bound state is supported by recent molecular dynamics simulations (48). These simulations suggest that the β strands (in particular β2-β3) that comprise the bridging sheet may be highly ﬂexible but that the inner and outer domains of gp120 may largely retain their CD4-bound conformation even when the bridging sheet is unfolded. The seemingly high degree of ﬂexibility of β2-β3 in both the liganded and unliganded structures (48) likely results in movement of V1/V2; this conformational ﬂexibility provides a structural basis for how the V1/V2 regions may generally protect

the CD4bs from efﬁcient antibody recognition (2, 44). Silent face: the 2G12 epitope. A second broadly neutralizing antibody reactive with HIV-1 gp120 is 2G12. This antibody recognizes clustered α1 → 2-linked mannose residues on the distal ends of oligomannose sugars located on the carbohydrate-covered silent face of the gp120 outer domain (76– 78). Although the underlying protein surface appears to have an inﬂuence on the proper presentation of these sugars to the antibody www.annualreviews.org • GP120-Targeted Vaccine Design

749

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

3:9

Figure 6 Structure of the domain-swapped Fab molecules of 2G12 as they assemble in the crystal. The Fab light chains ( green and blue) and the corresponding heavy chains (cyan and yellow) are shown. The VH /VH interface and the CDR (complementarity-determining region) H3 are labeled.

(78), there is no evidence for any direct interaction between 2G12 and the side chains of the underlying residues. Like b12, mAb 2G12 has also been studied extensively for its antiviral activity. In vitro, 2G12 achieved an IC50 against 41 viruses included in a 90-member virus panel (70), and in passive transfer studies the antibody has been shown to protect against viral infection, particularly when administered in combination with other broadly neutralizing antibodies (79–83). The in vivo efﬁcacy of this antibody has been underscored in a recent study in which it was administered to HIV-positive individuals undergoing interrupted antiretroviral therapy (84). The study showed that acutely infected individuals who were given the antibody had a signiﬁcant delay in viral rebound compared with controls or chronically infected individuals; the occurrence of 2G12 viral escape mutants shows that the antibody was indeed effective in vivo (84). The crystal structure of 2G12 (85) reveals that the antibody is able to achieve nanomo-

750

Pantophlet

·

Burton

lar binding afﬁnity to a glycan array owing to the unusual conﬁguration of the antigenbinding fragments (Fabs), which form a domain-swapped Fab dimer (Figure 6): Thus, the variable heavy chains (VH ) from each Fab arm have exchanged positions to interact with the light chain (VL ) of the neighboring Fab. The result is a multivalent platform for the binding interaction with carbohydrate (85). The domain swap preserves the conventional antigen binding sites formed by the VH /VL interfaces of the two Fab arms, while creating a novel interface for the two VH domains. Co-crystallization of 2G12 with Man9 GlcNAc2 (85) indicates that this multivalent platform allows 2G12 to interact with up to four oligomannose sugars simultaneously; two of these glycans are located, as expected, within the antigen binding sites formed by VL and VH , whereas the other two are bound within the newly formed secondary binding site formed by the VH /VH interface. The crystal structure of 2G12 complexed to Man9 GlcNAc2 indicates that the antibody

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

interacts predominantly with the terminal α1 → 2-linked mannose residues that comprise the D1 arm of the oligosaccharide, although more recent studies have shown that the conventional binding sites can also be occupied by the D3 arm (86). These results suggest a lower degree of stringency in the antibody-ligand interaction than previously presumed and are important for the design of synthetic glycoside antigens to elicit 2G12like anticarbohydrate antibodies. The 2G12 structure indicates that the humoral arm of the immune system has the capacity to overcome the glycan-masking strategy that HIV employs to protect itself from antibody recognition. However, the uniqueness of the structure does raise the question of whether such domain-swapped antibodies can be elicited by a vaccine. Of note here also is the fact that 2G12, in contrast to mAb b12, is unable to neutralize viruses from clade C (70). Viruses from this clade cause more infections worldwide than viruses from any of the other HIV clades (87). 2G12 also does not frequently neutralize viruses from clades D and AE [circulating recombinant form (CRF) 01]. The general inability of 2G12 to neutralize these viruses is likely due to the absence of one or more glycans that are required for efﬁcient 2G12 binding to gp120 from these viruses (70, 88), although further studies are needed to address this issue fully.

Neutralization-Restricted gp120 Epitopes V3 region: the 447–52D epitope. The b12 (CD4bs) and 2G12 (glycan) epitopes are presently the most attractive targets for vaccine design owing to their highly conserved nature. Another site on gp120 that has attracted much attention is the epitope in the V3 region recognized by the antibody 44752D (89). The core sequence of the antibody epitope is Gly-Pro-Gly-Arg (GPGR) (89, 90), which is situated at the center of the V3 region. This sequence, which is pre-

dominant in clade B viruses but not in non– clade B viruses (91), causes the V3 region to adopt a marked type-II β-hairpin turn (92, 93); in non–clade B viruses, the Arg residue is generally replaced by Gln, although it appears that the Gln residue preserves the β-hairpin conformation (94). Although the GPGR sequence represents the core epitope for mAb 447-52D, antibody binding is also inﬂuenced by certain amino acid substitutions outside of the GPGR sequence, particularly in the N-terminal segment of the V3 region (95, 96). The X-ray crystal structure of mAb 44752D in complex with a V3 peptide indicates how this antibody may have the capacity to neutralize more clade B viruses than other anti-V3 antibodies described so far. First, most of the binding interaction between the antibody and the peptide is mediated by mainchain contacts (93), which broadens the ability of the antibody to recognize a variety of V3 sequences. The only side-chain interactions are with the Pro and Arg residues in the GPGR sequence; the side chains from both residues form extensive interactions with residues in the antibody combining site. Second, comparison of the conformation of the 447-52Dcomplexed V3 peptide to the conformations of V3 peptides solved in complex with other anti-V3 antibodies shows that the side chain of the Arg residue in the GPGR sequence is oriented in the opposite direction in the 447-52D complex relative to its orientation in the other antibody:peptide complexes (93, 96). Thus, the fact that mAb 447-52D neutralizes more viruses than other anti-V3 antibodies that have been described so far may be the result of its unique mode of interaction with V3. The somewhat broad neutralizing activity of mAb 447-52D (70) makes the antibody epitope a potential vaccine target. However, there are certain caveats associated with targeting the V3 region of HIV-1 (97). For example, some of the main-chain interactions exhibited by 447-52D in binding to the V3 region may be difﬁcult to reproduce upon www.annualreviews.org • GP120-Targeted Vaccine Design

CRF: circulating recombinant form (intermixed co-circulating HIV strains)

751

ANRV270-IY24-24

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

scFv: single-chain variable fragment (of an antibody)

3:9

immunization with antigen. Furthermore, the V3 region on some clade B primary viruses appears to be inaccessible to antibody (70); whether the V3 region in non–clade B viruses is also inaccessible to antibody is uncertain. Occlusion of the V3 region appears to be manifested solely in the context of the gp120 oligomer on the virus (97), and its exposure may be modulated by the presence of glycan moieties (98–104). Given these caveats, the V3 region may be a target that only yields antibodies with restricted neutralizing ability. It is worth noting here that a 447-52D equivalent has not been identiﬁed so far for non–clade B viruses. The V3 region in these viruses contains a GPGQ sequence that is more highly conserved than the GPGR sequence in clade B viruses (91). It is possible, though still largely unproven, that the reduced sequence variability in these viruses may be the result of the V3 region being even less accessible to neutralizing antibodies in non–clade B viruses compared with the V3 region in clade B viruses (97). Experimental support for this notion comes from a recent study that found that although crossreactive anti-V3 antibodies from individuals infected with a clade A virus were able to neutralize a neutralization-sensitive clade B virus, these antibodies were unable to neutralize the homologous clade A virus or a heterologous clade B virus that is somewhat resistant to neutralization by anti-V3 antibodies, including mAb 447-52D (105). The resistance of these viruses to neutralization was associated with the presence of V1/V2. This observation suggests that the V3 region may be masked by V1/V2 in the context of oligomeric gp120. Although these observations do not unequivocally demonstrate that the V3 regions on non– clade B viruses are inaccessible relative to the V3 region in clade B viruses, they do point to potential difﬁculties in attempting to elicit cross-clade neutralizing antibodies based on the V3 region. Coreceptor-binding site. As discussed earlier, the CoRbs is likely not presented until af-

752

Pantophlet

·

Burton

ter CD4 has bound gp120. This is reﬂected in the fact that antibodies to the CoRbs of HIV1 generally neutralize virus weakly or not at all in vitro, although their ability to neutralize HIV is dramatically enhanced if soluble CD4 is added at sub-neutralizing concentrations (54, 106, 107). The inability of antiCoRbs antibodies to strongly neutralize HIV in the absence of soluble CD4 is associated with the limited access of these antibodies to the coreceptor site once the virus has engaged membrane-bound CD4 on the target cell (36, 108). Thus, the neutralizing activity of antibodies to the coreceptor site has been shown to be inversely correlated with their size (108, 109); i.e., the neutralizing activity of the single-chain variable fragment (scFv) (∼25 kDa in mass) is greater than the neutralizing activity of the Fab fragment (∼50 kDa), which is greater than the neutralizing activity of the IgG (∼150 kDa). These observations suggest that the CoRbs may not be a good target for vaccine design. However, the conserved nature of the CoRbs does make it an attractive target for small-molecule inhibitors. For example, the chimeric protein sCD4-17b, which is composed of soluble CD4 covalently linked to the scFv of the anti-CoRbs mAb 17b, has been designed as a potential therapeutic agent (110). V1 and V2 variable regions. Very little is known about the potential of V1 and/or V2 as vaccine targets. However, existing mAbs to epitopes in the V1 or V2 variable regions generally do not exhibit any signiﬁcant broad neutralizing activity against primary viruses (111– 116). Sequence comparison of V1/V2 from different viruses shows that both V1 and V2 exhibit considerable polymorphism in length and sequence (117–120); this polymorphism may explain the inability of antibodies to epitopes in V1/V2 to neutralize broadly. Taken together, these studies indicate that the V1 and V2 regions are likely poor vaccine targets. Recently, a human mAb (mAb 2909) has been described that is able to bind oligomeric gp120 on the virus but not soluble monomeric

ANRV270-IY24-24

ARI

15 February 2006

3:9

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

gp120 (57). The epitope of this antibody involves the CD4bs, as well as the V2 and V3 variable regions. Although this antibody only neutralizes the neutralization-sensitive virus SF162, it does so at picomolar antibody concentrations (57). The occurrence of mAb 2909 suggests the possible existence of additional antibodies that are oligomer-speciﬁc. Such antibodies, in contrast to mAb 2909, may possess broader neutralizing activity.

ANTIGEN DESIGN STRATEGIES TO ELICIT CROSS-NEUTRALIZING ANTI-GP120 ANTIBODIES At present, many antigen design strategies are being pursued to evaluate their potential to elicit broadly neutralizing antibodies (121). Two strategies that we focus on here are (a) the design of antigens that mimic the functional envelope spike and (b) the design of novel, epitope-focused antigens.

Mimicry of the Functional HIV Envelope Spike Stabilization of recombinant gp120-gp41 complexes. An immense hurdle in the design of antigens that mimic the envelope spike is the stabilization of the gp120-gp41 interaction, which is normally mediated by noncovalent interactions and is relatively labile. An often-used stabilization strategy is mutation of the protease cleavage site between gp120 and gp41 so that the two subunits remain covalently linked upon protein expression (122–130). However, uncleaved envelope proteins appear to be antigenically distinct from their cleaved counterparts (131). This observation is supported by more recent extensive analyses that show that uncleaved gp120-gp41 is antigenically substantially different from its cleaved functional form (24, 132). Thus, whereas both nonneutralizing and neutralizing antibodies bind uncleaved envelope glycoproteins, only neutralizing anti-gp120 antibodies efﬁciently

bind oligomeric glycoprotein when the gp160 precursor protein is effectively cleaved by furin proteases (or equivalent proteases). As an alternative approach to this problem, single cysteine residues have been introduced into gp120 and gp41, which results in the formation of an intermolecular disulﬁde bridge upon expression of the proteolytically cleaved oligomer (131, 133, 134). To improve cleavage efﬁciency, optimized cleavage recognition sequences have also been inserted between gp120 and gp41 (135). Oligomeric envelope complexes generated by this approach have better antigenic qualities than uncleaved complexes (134). However, the disulﬁde bridgestabilized complexes still display epitopes that are recognized by nonneutralizing antibodies (134), which suggests that mimicry of the native envelope glycoprotein is not fully achieved with these molecules. Trimerization of recombinant gp120gp41 complexes. A further challenge in designing envelope spike mimics is the preservation of the trimeric state of the antigen complex. Thus, although mutation of the protease cleavage site or the introduction of a disulﬁde bridge allows gp120 and gp41 to remain covalently linked, soluble oligomers derived in this manner often either disassociate into single heterodimers of gp120 and gp41, or tend to assemble into varying oligomerization states, such as dimers, trimers and/or tetramers (123, 126–128, 130, 136, 137). One of the strategies that has been employed to circumvent these problems is covalent linkage of heterologous trimerization domains to the C terminus of recombinant oligomeric proteins (138– 140). These domains indeed improve trimerization, and the corresponding proteins elicit neutralizing antibodies at modestly improved levels compared with preparations that contain a mixture of oligomeric proteins (141). Notably, immunization studies show that puriﬁed trimeric glycoproteins—independent of whether or not they are stabilized by a heterologous trimerization domain—are somewhat superior at eliciting cross-neutralizing www.annualreviews.org • GP120-Targeted Vaccine Design

753

ANRV270-IY24-24

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Immunodominant: the epitope on a molecule that provokes the most intense immune response

3:9

antibodies as compared to simple monomeric gp120 formulations (142–144). Also of note is that the elicited neutralizing antibodies appear to be directed to epitopes other than the immunodominant V3 region. Further strategies to stabilize the trimeric envelope complex involve the introduction of mutations in the gp41 subunit. For example, cysteine residues have been inserted into the N-terminal heptad repeat of gp41 to introduce intramolecular disulﬁde bridges (124, 130). This approach has been shown to increase the stability of oligomeric envelope proteins, in particular oligomers that contain appended trimerization domains (124). However, the antigenic proﬁle of such preparations does not seem to improve with the added disulﬁde bridge (124); immunogenicity studies with these disulﬁde-stabilized glycoproteins have not been reported. Mutations have also been introduced into gp41 to prevent it from adopting a postfusion conformation and stabilize it in a supposed native-like, prefusogenic conformation (145). Immunogenicity studies in rabbits with oligomeric preparations containing cleaved, disulﬁdebridged gp120/gp41, in conjunction with a gp41 prefusion conformation-stabilizing mutation, have recently been reported (146). In a subset of animals, antibodies were elicited that could neutralize the homologous virus that is moderately resistant to neutralizing antibodies. However, direct comparison with neutralizing antibody responses elicited by immunizing with monomeric gp120 was not possible owing to the design of the study (146). Thus, it remains unclear to what extent the introduced modiﬁcations are able to enhance the ability of oligomers to elicit cross-neutralizing antibodies. Presentation of the gp120-gp41 complex on particles. Other antigen design strategies have attempted to present the oligomeric envelope spike in a more native environment, for example, through mild chemical inactivation of virus particles (which appears to preserve the oligomerization state and func-

754

Pantophlet

·

Burton

tionality of the envelope spike) (147, 148) or, alternatively, through the incorporation of uncleaved oligomeric envelope glycoprotein complexes into proteoliposome preparations (149). Though promising, both of these two approaches have so far failed to elicit neutralizing antibodies of the desired breadth and potency (142, 150, 151). Another attractive approach to presenting the envelope spike in a native-like context is the generation of virus-like particles (VLPs) (152). VLPs represent a unique class of subunit antigens, as they may allow closer mimicry of structures on the surface of virions compared with other types of antigen presentation. VLPs also have the advantage that, in contrast to soluble subunit antigens, they are able to stimulate both cellular and humoral responses, although, unlike live or inactivated virus, they do not contain viral genetic material (152). So far, however, the VLP approach has not achieved the desired success in eliciting high levels of cross-neutralizing antibodies (153–155). A possible reason for the inability of current VLP preparations to elicit more broadly neutralizing antibodies may be due to the expression of both nonfunctional as well as functional forms of the spike on the surface of the virus particle (12, 24). The elicitation of antibodies to cellular proteins that are incorporated into the virus particle may also limit the efﬁcacy of VLP preparations. Overcoming low spike immunogenicity and genetic diversity. The relatively low accessibility of conserved antibody epitopes on the envelope spike poses another strategic hurdle to the pursuit of envelope spike mimicry as a vaccine approach. To enhance the elicitation of broadly neutralizing antibodies, variant trimeric envelope glycoproteins have been designed with truncated or altered variable regions (156–159). In some cases, alterations have also been made in gp41. Although some of these alterations have resulted in an improvement in the neutralizing capacity of the immune sera (156), additional modiﬁcations and other improvements will

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

be necessary to signiﬁcantly boost the ability of these envelope spike mimics to elicit antibodies that are capable of neutralizing a wide range of primary viral isolates. Among the approaches that are currently being evaluated to overcome HIV’s genetic diversity are multivalent antigen cocktails (160–163) and the use of antigenic formulations based on consensus or ancestral HIV sequences (164–167). The latter are based on computer-generated “centralized” HIV gene sequences (165, 167). These artiﬁcial sequences are genetically equidistant to the sequences of circulating viruses at a given time point, as are the proteins derived from these sequences. It is postulated that immunization with these centralized sequences will yield antibodies to multiple conserved regions across many different viral subtypes (164, 165). So far, however, the multivalent strategies as well as the consensus/ancestral envelope approaches have failed to signiﬁcantly improve the breadth of the neutralizing antibody response (161, 166). In the multivalent approach, incorporation of additional envelope glycoproteins does result in an increase in the number of viruses that are neutralized, but the observed increase seems to stem from the sum of neutralizing antibodies elicited by the individual components rather than from a signiﬁcant increase in the overall quality and breadth of the neutralizing antibody response (161). The notion that the observed neutralizing antibody responses are additive and not qualitatively enhanced likely limits the practical application of this approach, as it would probably require the screening and incorporation of numerous envelope glycoproteins to cover all current and future circulating viruses. Similarly, serum antibodies elicited by current consensus envelope glycoproteins do not seem to neutralize a substantially broader range of viral isolates than antibodies elicited by wild-type envelope proteins. These observations suggest that the current synthetic molecules do not display the anticipated conserved regions, or do so in a manner that is not replicated on the functional spike (165, 166).

Thus, like wild-type proteins, these synthetic proteins will likely also have to be modiﬁed to increase their potential to elicit the desired cross-neutralizing antibodies (166).

Epitope-Focused Antigen Design Despite incremental progress in the ability of current oligomeric glycoprotein antigens to elicit broadly neutralizing anti-HIV antibodies, type-speciﬁc neutralizing antibodies remain the predominant fraction among antibodies that are contained in immune sera. Epitope-focused antigen design may provide a means by which to limit the induction of such antibodies, while preserving the elicitation of antibodies to more conserved epitopes (168). So far, this approach has been aimed mainly at attempts to elicit b12- and 2G12like antibodies.

N-glycan: a carbohydrate chain covalently linked to an Asn residue in the consensus sequence -Asn-Xaa-Ser/Thr (Xaa: any amino acid except Pro)

Directing antibody responses to the b12 epitope. Antibody responses to gp120 can be inﬂuenced by placing glycans at certain positions in the molecule (169). To direct antibody responses to the b12 epitope, monomeric gp120 has been engineered to contain a range of additional N-glycan sites at various locations throughout the molecule to hide gp120 epitopes that could elicit non- or weakly neutralizing antibodies, without altering the accessibility of the b12 epitope (170, 171). This so-called hyperglycosylation strategy has led to the design of a panel of novel glycoproteins that exhibit superior antigenic qualities relative to unmodiﬁed gp120, judging from the inability of these modiﬁed antigens to be recognized by numerous nonand weakly neutralizing antibodies (170, 171). A recently completed study in which rabbits were immunized with a ﬁrst-generation of modiﬁed antigens did not reveal any improvement in the breadth of serum neutralization compared with sera from animals immunized with wild-type gp120 (172); the induction of antibodies to undesired epitopes was reduced, but no increase in neutralizing antibody responses to the CD4bs was www.annualreviews.org • GP120-Targeted Vaccine Design

755

ANRV270-IY24-24

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CDR: complementaritydetermining region (of an antibody) Immunity: a natural or acquired resistance to a speciﬁc disease; immunity may be partial or complete, long lasting or temporary

756

3:9

observed in return. Clearly, further iterative modiﬁcations to the design of these antigens will be necessary to boost their ability to elicit neutralizing antibodies, also taking into account factors that may limit the immunogenicity of these antigens (173). Generation of 2G12-like antibodies. Efforts are currently also underway to design an antigen that will elicit antibodies that target the carbohydrate epitope recognized by mAb 2G12. The structure of 2G12 in complex with various carbohydrate ligands shows that the antibody interacts mainly via residues in CDR L3 and CDR H1–H3 with the α1 → 2 terminal portion of oligomannose sugars and synthetic glycosides (85, 86); synthetic glycosides and glycans derived from natural sources that contain such α1 → 2-linked mannose residues indeed are best at inhibiting 2G12 binding to monomeric gp120 (174–178). However, inhibition of 2G12 binding with monovalent sugars requires concentrations in the millimolar range (174, 176, 178); in contrast, 2G12 binds monomeric gp120 with nanomolar afﬁnity (76), which likely results from its multivalent interaction with gp120. In an attempt to reproduce the high-mannose clustering on gp120, synthetic glycosides and Man9 sugars have been conjugated to protein carrier molecules or to synthetic scaffolds (174–178; H.-K. Lee, C.-Y. Huang, R. Astronomo, C.S. Scanlan, R. Pantophlet, I.A. Wilson, C.-H. Wong & D.R. Burton, unpublished data). Multivalent display of these sugar molecules improves their ability to inhibit 2G12 binding to gp120 ∼1000-fold into the micromolar range. Although the improvement in binding afﬁnity by multivalent display is appreciable, these studies clearly show that additional ﬁne-tuning of such multivalently displayed platforms will be required to further improve 2G12 binding afﬁnity, which, by inference, should improve the ability of these antigenic formulations to elicit 2G12like antibodies. Thus, it is likely that the glycan clusters on current glyco-mimics do not Pantophlet

·

Burton

adequately resemble the cluster of glycans on gp120. Improvements to the design of these glyco-antigens will likely require de novo design of synthetic scaffolds that are able to display a high density of oligomannose sugars or synthetic glycosides in spatial orientations that better mimic the 2G12 epitope.

CONCLUDING REMARKS Advances in HIV therapy have been successful in prolonging the lives of HIV-infected individuals, particularly those who live in the U.S. and Europe. However, in Asian and African nations, the spread of HIV remains a major health concern with dramatic socio-economic consequences. Development of a vaccine offers the best hope of containing the AIDS pandemic. The failure of recent phase III clinical trials that used monomeric gp120 as a candidate vaccine underscores the inability of simple subunit HIV antigens to induce broad humoral immunity, as well as the need for novel scientiﬁc approaches to obtain an AIDS vaccine. In this review, we summarize some of the strategies that are currently being pursued to generate an AIDS vaccine component that elicits broadly neutralizing antibodies. These strategies have proﬁted from the elucidation of the crystal structures of various broadly neutralizing mAbs and of the core gp120 structures of HIV and (most recently) of SIV. This structural insight has allowed vaccinologists to understand better the defense mechanisms of the virus, while highlighting chinks in HIV’s defensive armor. Efﬁcient exploitation of these chinks will require substantial advances in antigen design, combined with further insight from X-ray crystallography and the identiﬁcation of additional neutralizing antibody epitopes (see Future Issues To Be Resolved, below). By correlating the antigenic properties of novel antigens to their immunogenicity, it should then be possible to optimize promising antigenic formulations and forward them as potential vaccine candidates into human clinical and efﬁcacy trials.

ANRV270-IY24-24

ARI

15 February 2006

3:9

SUMMARY POINTS 1. The crystal structures of the gp120 cores from HIV-1 and SIV reveal that the viruses are able to minimize effective antibody recognition of conserved regions by restricting access to these regions by bulky carbohydrate chains or variable regions and by only fully assembling functional sites once the interaction between virus and target cell has been initiated.

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

2. Although oligomeric envelope glycoprotein antigens seem superior at eliciting neutralizing antibodies relative to simple monomeric gp120 formulations, further modiﬁcations will be required to optimize the ability of these antigens to elicit broadly neutralizing antibodies. 3. Three epitopes deﬁned by three different mAbs (two broadly neutralizing and one lesser cross-neutralizing) have been identiﬁed so far on HIV-1 gp120; speciﬁc targeting of these epitopes remains a major challenge in HIV vaccine design.

FUTURE ISSUES TO BE RESOLVED 1. The crystal structures of additional envelope glycoproteins need to be elucidated. The solution of the structures of full-length monomeric gp120 and of the gp120gp41 trimer complex will provide both (a) further insight into how HIV protects itself from antibody and (b) possible ways to circumvent these barriers. 2. The broad-neutralizing antibody activity in the sera of long-term HIV-infected individuals needs to be dissected to allow additional neutralizing antibody epitopes on the virus to be identiﬁed; these epitopes could then become targets for vaccine design. 3. A comprehensive analysis of antibody responses in sera from immunized animals is required. Delineation of the epitopes recognized by antibodies induced by potential vaccine antigens may reveal correlations between antigenicity and immunogenicity. This information would allow further optimization of the antigens for the elicitation of broadly neutralizing antibodies. 4. An evaluation of the inﬂuence of molecular adjuvants on the levels of elicited neutralizing antibodies is also required. The combination of immunostimulatory molecules or equivalent agents with potential vaccine antigens may allow for the improvement of neutralizing antibody titers upon immunization.

NOTE ADDED IN PROOF Very recently, the crystal structure of a gp120 core of HIV-1 containing the V3 variable region was reported (179). In the structure, which was determined in complex with CD4 and the Fab fragment of an antibody to the CoRbs, the V3 region extends approximately 30 A˚ away from the core. The extended conformation of the V3 region may explain why anti-V3 antibodies are readily elicited upon immunization with monomeric gp120. Whether V3 is similarly extended in the absence of CD4 and, more importantly, in the context of the viral spike, remains to be determined. www.annualreviews.org • GP120-Targeted Vaccine Design

757

ANRV270-IY24-24

ARI

15 February 2006

3:9

ACKNOWLEDGMENTS We gratefully acknowledge support from the American Foundation for AIDS Research (fellowship 106436–34-RFVA to R.P.) and grants from the National Institutes of Health and the International AIDS Vaccine Initiative (to D.R.B.). We also thank R. Wyatt and S. Harrison for their critique of the manuscript and valuable comments.

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LITERATURE CITED 1. Parren PWHI, Gauduin MC, Koup RA, Poignard P, Sattentau QJ, et al. 1997. Relevance of the antibody response against human immunodeﬁciency virus type 1 envelope to vaccine design. Immunol. Lett. 57:105–12. Erratum. 1997. Immunol. Lett. 58:125– 32 2. This review provides a comprehensive summary of the function, antigenicity and immunogenicity of HIV envelope glycoproteins.

758

2. Wyatt R, Sodroski J. 1998. The HIV-1 envelope glycoproteins: fusogens, antigens, and immunogens. Science 280:1884–88 3. Graham BS, Wright PF. 1995. Candidate AIDS vaccines. N. Engl. J. Med. 333:1331–39 4. Schultz AM, Bradac JA. 2001. The HIV vaccine pipeline, from preclinical to phase III. AIDS 15(Suppl. 5):S147–58 5. Flynn NM, Forthal DN, Harro CD, Judson FN, Mayer KH, Para MF. 2005. Placebocontrolled phase 3 trial of a recombinant glycoprotein 120 vaccine to prevent HIV-1 infection. J. Infect. Dis. 191:654–65 6. Gilbert PB, Peterson ML, Follmann D, Hudgens MG, Francis DP, et al. 2005. Correlation between immunologic responses to a recombinant glycoprotein 120 vaccine and incidence of HIV-1 infection in a phase 3 HIV-1 preventive vaccine trial. J. Infect. Dis. 191:666–77 7. Moore JP, Cao Y, Qing L, Sattentau QJ, Pyati J, et al. 1995. Primary isolates of human immunodeﬁciency virus type 1 are relatively resistant to neutralization by monoclonal antibodies to gp120, and their neutralization is not predicted by studies with monomeric gp120. J. Virol. 69:101–9 8. Sattentau QJ, Moore JP. 1995. Human immunodeﬁciency virus type 1 neutralization is determined by epitope exposure on the gp120 oligomer. J. Exp. Med. 182:185–96 9. Fouts TR, Binley JM, Trkola A, Robinson JE, Moore JP. 1997. Neutralization of the human immunodeﬁciency virus type 1 primary isolate JR-FL by human monoclonal antibodies correlates with antibody binding to the oligomeric form of the envelope glycoprotein complex. J. Virol. 71:2779–85 10. Parren PWHI, Mondor I, Naniche D, Ditzel HJ, Klasse PJ, et al. 1998. Neutralization of human immunodeﬁciency virus type 1 by antibody to gp120 is determined primarily by occupancy of sites on the virion irrespective of epitope speciﬁcity. J. Virol. 72:3512–19 11. Schonning K, Bolmstedt A, Novotny J, Lund OS, Olofsson S, Hansen JE. 1998. Induction of antibodies against epitopes inaccessible on the HIV type 1 envelope oligomer by immunization with recombinant monomeric glycoprotein 120. AIDS Res. Hum. Retrovir. 14:1451–56 12. Poignard P, Moulard M, Golez E, Vivona V, Franti M, et al. 2003. Heterogeneity of envelope molecules expressed on primary human immunodeﬁciency virus type 1 particles as probed by the binding of neutralizing and nonneutralizing antibodies. J. Virol. 77:353– 65 13. Herrera C, Spenlehauer C, Fung MS, Burton DR, Beddows S, Moore JP. 2003. Nonneutralizing antibodies to the CD4-binding site on the gp120 subunit of human Pantophlet

·

Burton

ANRV270-IY24-24

14.

15. 16.

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

17.

18.

19.

20. 21. 22. 23.

24.

25. 26. 27.

28.

29.

30.

ARI

15 February 2006

3:9

immunodeﬁciency virus type 1 do not interfere with the activity of a neutralizing antibody against the same site. J. Virol. 77:1084–91 Chen B, Vogan EM, Gong H, Skehel JJ, Wiley DC, Harrison SC. 2005. Structure of an unliganded simian immunodeficiency virus gp120 core. Nature 433:834– 41 Poignard P, Saphire EO, Parren PWHI, Burton DR. 2001. gp120: Biologic aspects of structural features. Annu. Rev. Immunol. 19:253–74 Fenouillet E, Gluckman JC, Jones IM. 1994. Functions of HIV envelope glycans. Trends Biochem. Sci. 19:65–70 Leonard CK, Spellman MW, Riddle L, Harris RJ, Thomas JN, Gregory TJ. 1990. Assignment of intrachain disulﬁde bonds and characterization of potential glycosylation sites of the type 1 recombinant human immunodeﬁciency virus envelope glycoprotein (gp120) expressed in Chinese hamster ovary cells. J. Biol. Chem. 265:10373–82 Zhu P, Chertova E, Bess J Jr, Lifson JD, Arthur LO, et al. 2003. Electron tomography analysis of envelope glycoprotein trimers on HIV and simian immunodeﬁciency virus virions. Proc. Natl. Acad. Sci. USA 100:15812–17 Hill CP, Worthylake D, Bancroft DP, Christensen AM, Sundquist WI. 1996. Crystal structures of the trimeric human immunodeﬁciency virus type 1 matrix protein: implications for membrane association and assembly. Proc. Natl. Acad. Sci. USA 93:3099–104 Chan DC, Fass D, Berger JM, Kim PS. 1997. Core structure of gp41 from the HIV envelope glycoprotein. Cell 89:263–73 Tan K, Liu J, Wang J, Shen S, Lu M. 1997. Atomic structure of a thermostable subdomain of HIV-1 gp41. Proc. Natl. Acad. Sci. USA 94:12303–8 Weissenhorn W, Dessen A, Harrison SC, Skehel JJ, Wiley DC. 1997. Atomic structure of the ectodomain from HIV-1 gp41. Nature 387:426–30 Kuznetsov YG, Victoria JG, Robinson WE Jr, McPherson A. 2003. Atomic force microscopy investigation of human immunodeﬁciency virus (HIV) and HIV-infected lymphocytes. J. Virol. 77:11896–909 Herrera C, Klasse PJ, Michael E, Kake S, Barnes K, et al. 2005. The impact of envelope glycoprotein cleavage on the antigenicity, infectivity, and neutralization sensitivity of Env-pseudotyped human immunodeﬁciency virus type 1 particles. Virology 338:154–72 Parren PWHI, Burton DR, Sattentau QJ. 1997. HIV-1 antibody–debris or virion? Nat. Med. 3:366–67 Burton DR, Monteﬁori DC. 1997. The antibody response in HIV-1 infection. AIDS 11(Suppl. A):S87–98 Modrow S, Hahn BH, Shaw GM, Gallo RC, Wong-Staal F, Wolf H. 1987. Computerassisted analysis of envelope protein sequences of seven human immunodeﬁciency virus isolates: prediction of antigenic epitopes in conserved and variable regions. J. Virol. 61:570–78 Willey RL, Rutledge RA, Dias S, Folks T, Theodore T, et al. 1986. Identiﬁcation of conserved and divergent domains within the envelope gene of the acquired immunodeﬁciency syndrome retrovirus. Proc. Natl. Acad. Sci. USA 83:5038–42 Helseth E, Olshevsky U, Furman C, Sodroski J. 1991. Human immunodeﬁciency virus type 1 gp120 envelope glycoprotein regions important for association with the gp41 transmembrane glycoprotein. J. Virol. 65:2119–23 Kowalski M, Potz J, Basiripour L, Dorfman T, Goh WC, et al. 1987. Functional regions of the envelope glycoprotein of human immunodeﬁciency virus type 1. Science 237:1351– 55 www.annualreviews.org • GP120-Targeted Vaccine Design

14. This report describes the first crystal structure of an unliganded gp120 core molecule.

759

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

33. A comprehensive study detailing the topology of HIV gp120 based on the inhibition profiles of anti-gp120 antibodies of defined specificity.

42. This report describes the first crystal structure of a CD4-liganded gp120 core molecule.

46. This study provides the first structure-based model of an HIV envelope spike.

760

3:9

31. Moore JP, Sattentau QJ, Wyatt R, Sodroski J. 1994. Probing the structure of the human immunodeﬁciency virus surface glycoprotein gp120 with a panel of monoclonal antibodies. J. Virol. 68:469–84 32. Pollard SR, Rosa MD, Rosa JJ, Wiley DC. 1992. Truncated variants of gp120 bind CD4 with high afﬁnity and suggest a minimum CD4 binding region. EMBO J. 11:585–91 33. Moore JP, Sodroski J. 1996. Antibody cross-competition analysis of the human immunodeficiency virus type 1 gp120 exterior envelope glycoprotein. J. Virol. 70:1863–72 34. Olshevsky U, Helseth E, Furman C, Li J, Haseltine W, Sodroski J. 1990. Identiﬁcation of individual human immunodeﬁciency virus type 1 gp120 amino acids important for CD4 receptor binding. J. Virol. 64:5701–7 35. Moore JP, Willey RL, Lewis GK, Robinson J, Sodroski J. 1994. Immunological evidence for interactions between the ﬁrst, second, and ﬁfth conserved domains of the gp120 surface glycoprotein of human immunodeﬁciency virus type 1. J. Virol. 68:6836–47 36. Sullivan N, Sun Y, Sattentau Q, Thali M, Wu D, et al. 1998. CD4-induced conformational changes in the human immunodeﬁciency virus type 1 gp120 glycoprotein: consequences for virus entry and neutralization. J. Virol. 72:4694–703 37. Wyatt R, Moore J, Accola M, Desjardin E, Robinson J, Sodroski J. 1995. Involvement of the V1/V2 variable loop structure in the exposure of human immunodeﬁciency virus type 1 gp120 epitopes induced by receptor binding. J. Virol. 69:5723–33 38. Wyatt R, Sullivan N, Thali M, Repke H, Ho D, et al. 1993. Functional and immunologic characterization of human immunodeﬁciency virus type 1 envelope glycoproteins containing deletions of the major variable regions. J. Virol. 67:4557–65 39. Cao J, Sullivan N, Desjardin E, Parolin C, Robinson J, et al. 1997. Replication and neutralization of human immunodeﬁciency virus type 1 lacking the V1 and V2 variable loops of the gp120 envelope glycoprotein. J. Virol. 71:9808–12 40. Ren XP, Sodroski J, Yang XZ. 2005. An unrelated monoclonal antibody neutralizes human immunodeﬁciency virus type 1 by binding to an artiﬁcial epitope engineered in a functionally neutral region of the viral envelope glycoproteins. J. Virol. 79:5616–24 41. Kwong PD, Wyatt R, Majeed S, Robinson J, Sweet RW, et al. 2000. Structures of HIV-1 gp120 envelope glycoproteins from laboratory-adapted and primary isolates. Struct. Fold. Des. 8:1329–39 42. Kwong PD, Wyatt R, Robinson J, Sweet RW, Sodroski J, Hendrickson WA. 1998. Structure of an HIV gp120 envelope glycoprotein in complex with the CD4 receptor and a neutralizing human antibody. Nature 393:648–59 43. Huang CC, Stricher F, Martin L, Decker JM, Majeed S, et al. 2005. Scorpion-toxin mimics of CD4 in complex with human immunodeﬁciency virus gp120 crystal structures, molecular mimicry, and neutralization breadth. Structure 13:755–68 44. Wyatt R, Kwong PD, Desjardins E, Sweet RW, Robinson J, et al. 1998. The antigenic structure of the HIV gp120 envelope glycoprotein. Nature 393:705–11 45. Valadon P. 2004. RasTop 2.1. http://www.geneinfinity.org/rastop 46. Kwong PD, Wyatt R, Sattentau QJ, Sodroski J, Hendrickson WA. 2000. Oligomeric modeling and electrostatic analysis of the gp120 envelope glycoprotein of human immunodeficiency virus. J. Virol. 74:1961–72 47. Myszka DG, Sweet RW, Hensley P, Brigham-Burke M, Kwong PD, et al. 2000. Energetics of the HIV gp120-CD4 binding reaction. Proc. Natl. Acad. Sci. USA 97:9026–31 48. Pan YP, Ma BY, Nussinov R. 2005. CD4 binding partially locks the bridging sheet in gp120 but leaves the beta 2/3 strands ﬂexible. J. Mol. Biol. 350:514–27 Pantophlet

·

Burton

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

49. Rizzuto CD, Wyatt R, Hernandez-Ramos N, Sun Y, Kwong PD, et al. 1998. A conserved HIV gp120 glycoprotein structure involved in chemokine receptor binding. Science 280:1949–53 50. Rizzuto C, Sodroski J. 2000. Fine deﬁnition of a conserved CCR5-binding region on the human immunodeﬁciency virus type 1 glycoprotein 120. AIDS Res. Hum. Retrovir. 16:741–49 51. Hoffman TL, LaBranche CC, Zhang W, Canziani G, Robinson J, et al. 1999. Stable exposure of the coreceptor-binding site in a CD4-independent HIV-1 envelope protein. Proc. Natl. Acad. Sci. USA 96:6359–64 52. Edwards TG, Hoffman TL, Baribaud F, Wyss S, LaBranche CC, et al. 2001. Relationships between CD4 independence, neutralization sensitivity, and exposure of a CD4induced epitope in a human immunodeﬁciency virus type 1 envelope protein. J. Virol. 75:5230–39 53. Kolchinsky P, Kiprilov E, Sodroski J. 2001. Increased neutralization sensitivity of CD4independent human immunodeﬁciency virus variants. J. Virol. 75:2041–50 54. Decker JM, Bibollet-Ruche F, Wei XP, Wang SY, Levy DN, et al. 2005. Antigenic conservation and immunogenicity of the HIV coreceptor binding site. J. Exp. Med. 201:1407–19 55. Martin J, LaBranche CC, Gonzalez-Scarano F. 2001. Differential CD4/CCR5 utilization, gp120 conformation, and neutralization sensitivity between envelopes from a microgliaadapted human immunodeﬁciency virus type 1 and its parental isolate. J. Virol. 75:3568– 80 56. Reeves JD, Hibbitts S, Simmons G, McKnight A, Azevedo-Pereira JM, et al. 1999. Primary human immunodeﬁciency virus type 2 (HIV-2) isolates infect CD4-negative cells via CCR5 and CXCR4: comparison with HIV-1 and simian immunodeﬁciency virus and relevance to cell tropism in vivo. J. Virol. 73:7795–804 57. Gorny MK, Stamatatos L, Volsky B, Revesz K, Williams C, et al. 2005. Identiﬁcation of a new quaternary neutralizing epitope on human immunodeﬁciency virus type 1 virus particles. J. Virol. 79:5232–37 58. Cheng-Mayer C, Brown A, Harouse J, Luciw PA, Mayer AJ. 1999. Selection for neutralization resistance of the simian/human immunodeﬁciency virus SHIVSF33A variant in vivo by virtue of sequence changes in the extracellular envelope glycoprotein that modify N-linked glycosylation. J. Virol. 73:5294–300 59. Wei XP, Decker JM, Wang SY, Hui HX, Kappes JC, et al. 2003. Antibody neutralization and escape by HIV-1. Nature 422:307–12 60. Dacheux L, Moreau A, Ataman-Onal Y, Biron F, Verrier B, Barin F. 2004. Evolutionary dynamics of the glycan shield of the human immunodeﬁciency virus envelope during natural infection and implications for exposure of the 2G12 epitope. J. Virol. 78:12625– 37 61. Kwong PD, Doyle ML, Casper DJ, Cicala C, Leavitt SA, et al. 2002. HIV-1 evades antibody-mediated neutralization through conformational masking of receptor-binding sites. Nature 420:678–82 62. Thali M, Furman C, Ho DD, Robinson J, Tilley S, et al. 1992. Discontinuous, conserved neutralization epitopes overlapping the CD4-binding region of human immunodeﬁciency virus type 1 gp120 envelope glycoprotein. J. Virol. 66:5635–41 63. Pantophlet R, Saphire EO, Poignard P, Parren PWHI, Wilson IA, Burton DR. 2003. Fine mapping of the interaction of neutralizing and nonneutralizing monoclonal antibodies with the CD4 binding site of human immunodeﬁciency virus type 1 gp120. J. Virol. 77:642–58 www.annualreviews.org • GP120-Targeted Vaccine Design

761

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

70. The first extensive study to examine in vitro neutralization breadth and potency of broadly neutralizing antibodies using a high-throughput neutralization assay system.

75. This report describes the first crystal structure of a fully intact human mAb and is also the first structure of a broadly neutralizing antibody to HIV-1.

762

3:9

64. Wang T, Zhang ZX, Wallace OB, Deshpande M, Fang HQ, et al. 2003. Discovery of 4-benzoyl-1-[(4-methoxy-1H-pyrrolo[2,3-b]pyridin-3-yl)oxoacetyl]-2-(R)methylpiperazine (BMS-378806): a novel HIV-1 attachment inhibitor that interferes with CD4-gp120 interactions. J. Med. Chem. 46:4236–39 65. Guo Q, Ho HT, Dicker I, Fan L, Zhou NN, et al. 2003. Biochemical and genetic characterizations of a novel human immunodeﬁciency virus type 1 inhibitor that blocks gp120-CD4 interactions. J. Virol. 77:10528–36 66. Lin PF, Blair W, Wang T, Spicer T, Guo Q, et al. 2003. A small molecule HIV-1 inhibitor that targets the HIV-1 envelope and inhibits CD4 receptor binding. Proc. Natl. Acad. Sci. USA 100:11013–18 67. Madani N, Perdigoto AL, Srinivasan K, Cox JM, Chruma JJ, et al. 2004. Localized changes in the gp120 envelope glycoprotein confer resistance to human immunodeﬁciency virus entry inhibitors BMS-806 and #155. J. Virol. 78:3742–52 68. Si ZH, Madani N, Cox JM, Chruma JJ, Klein JC, et al. 2004. Small-molecule inhibitors of HIV-1 entry block receptor-induced conformational changes in the viral envelope glycoproteins. Proc. Natl. Acad. Sci. USA 101:5036–41 69. Burton DR, Desrosiers RC, Doms RW, Koff WC, Kwong PD, et al. 2004. HIV vaccine design and the neutralizing antibody problem. Nat. Immunol. 5:233–36 70. Binley JM, Wrin T, Korber B, Zwick MB, Wang M, et al. 2004. Comprehensive cross-clade neutralization analysis of a panel of anti-human immunodeficiency virus type 1 monoclonal antibodies. J. Virol. 78:13232–52 71. Barbas CF 3rd, Bjorling E, Chiodi F, Dunlop N, Cababa D, et al. 1992. Recombinant human Fab fragments neutralize human type 1 immunodeﬁciency virus in vitro. Proc. Natl. Acad. Sci. USA 89:9339–43 72. Roben P, Moore JP, Thali M, Sodroski J, Barbas CF 3rd, Burton DR. 1994. Recognition properties of a panel of human recombinant Fab fragments to the CD4 binding site of gp120 that show differing abilities to neutralize human immunodeﬁciency virus type 1. J. Virol. 68:4821–28 73. Parren PWHI, Marx PA, Hessell AJ, Luckay A, Harouse J, et al. 2001. Antibody protects macaques against vaginal challenge with a pathogenic R5 simian/human immunodeﬁciency virus at serum levels giving complete neutralization in vitro. J. Virol. 75:8340– 47 74. Veazey RS, Shattock RJ, Pope M, Kirijan JC, Jones J, et al. 2003. Prevention of virus transmission to macaque monkeys by a vaginally applied monoclonal antibody to HIV-1 gp120. Nat. Med. 9:343–46 75. Saphire EO, Parren PWHI, Pantophlet R, Zwick MB, Morris GM, et al. 2001. Crystal structure of a neutralizing human IGG against HIV-1: a template for vaccine design. Science 293:1155–59 76. Trkola A, Purtscher M, Muster T, Ballaun C, Buchacher A, et al. 1996. Human monoclonal antibody 2G12 deﬁnes a distinctive neutralization epitope on the gp120 glycoprotein of human immunodeﬁciency virus type 1. J. Virol. 70:1100–8 77. Sanders RW, Venturi M, Schiffner L, Kalyanaraman R, Katinger H, et al. 2002. The mannose-dependent epitope for neutralizing antibody 2G12 on human immunodeﬁciency virus type 1 glycoprotein gp120. J. Virol. 76:7293–305 78. Scanlan CN, Pantophlet R, Wormald MR, Saphire EO, Stanﬁeld R, et al. 2002. The broadly neutralizing anti-human immunodeﬁciency virus type 1 antibody 2G12 recognizes a cluster of alpha1–>2 mannose residues on the outer face of gp120. J. Virol. 76:7306–21 Pantophlet

·

Burton

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

79. Baba TW, Liska V, Hofmann-Lehmann R, Vlasak J, Xu WD, et al. 2000. Human neutralizing monoclonal antibodies of the IgG1 subtype protect against mucosal simian-human immunodeﬁciency virus infection. Nat. Med. 6:200–6 80. Ferrantelli F, Rasmussen RA, Buckley KA, Li PL, Wang T, et al. 2004. Complete protection of neonatal rhesus macaques against oral exposure to pathogenic simian-human immunodeﬁciency virus by human anti-HIV monoclonal antibodies. J. Infect. Dis. 189:2167– 73 81. Hofmann-Lehmann R, Vlasak J, Rasmussen RA, Smith BA, Baba TW, et al. 2001. Postnatal passive immunization of neonatal macaques with a triple combination of human monoclonal antibodies against oral simian-human immunodeﬁciency virus challenge. J. Virol. 75:7470–80 82. Mascola JR, Lewis MG, Stiegler G, Harris D, VanCott TC, et al. 1999. Protection of macaques against pathogenic simian/human immunodeﬁciency virus 89.6PD by passive transfer of neutralizing antibodies. J. Virol. 73:4009–18 83. Mascola JR, Stiegler G, VanCott TC, Katinger H, Carpenter CB, et al. 2000. Protection of macaques against vaginal transmission of a pathogenic HIV-1/SIV chimeric virus by passive infusion of neutralizing antibodies. Nat. Med. 6:207–10 84. Trkola A, Kuster H, Rusert P, Joos B, Fischer M, et al. 2005. Delay of HIV-1 rebound after cessation of antiretroviral therapy through passive transfer of human neutralizing antibodies. Nat. Med. 11:615–22 85. Calarese DA, Scanlan CN, Zwick MB, Deechongkit S, Mimura Y, et al. 2003. Antibody domain exchange is an immunological solution to carbohydrate cluster recognition. Science 300:2065–71 86. Calarese DA, Lee HK, Huang CY, Best MD, Astronomo RD, et al. 2005. Dissection of the carbohydrate speciﬁcity of the broadly neutralizing anti-HIV-1 antibody 2G12. Proc. Natl. Acad. Sci. USA 102:13372–77 87. McCutchan FE. 2000. Understanding the genetic diversity of HIV-1. AIDS 14(Suppl. 3):S31–44 88. Chen HY, Xu XD, Bishop A, Jones IM. 2005. Reintroduction of the 2G12 epitope in an HIV-1 clade C gp120. AIDS 19:833–35 89. Gorny MK, Conley AJ, Karwowska S, Buchbinder A, Xu JY, et al. 1992. Neutralization of diverse human immunodeﬁciency virus type 1 variants by an anti-V3 human monoclonal antibody. J. Virol. 66:7538–42 90. Gorny MK, Xu JY, Karwowska S, Buchbinder A, Zolla-Pazner S. 1993. Repertoire of neutralizing human monoclonal antibodies speciﬁc for the V3 domain of HIV-1 gp120. J. Immunol. 150:635–43 91. Gaschen B, Korber B, Foley DT. 1999. Global variation in the HIV-1 V3 region. In Human Retroviruses and AIDS 1999, ed. C. Kuiken, B. Foley, B. Hahn, P. Marx, F. McCutchan, J.W. Mellors, J. Mullins, S. Wolinsky, B. Korber, pp. 594–602. Los Alamos: Los Alamos Natl. Lab. 92. Stanﬁeld RL, Ghiara JB, Saphire EO, Profy AT, Wilson IA. 2003. Recurring conformation of the human immunodeﬁciency virus type 1 gp120 V3 loop. Virology 315:159– 73 93. Stanﬁeld RL, Gorny MK, Williams C, Zolla-Pazner S, Wilson IA. 2004. Structural rationale for the broad neutralization of HIV-1 by human monoclonal antibody 447-52D. Structure 12:193–204 94. Gupta G, Anantharamaiah GM, Scott DR, Eldridge JH, Myers G. 1993. Solution structure of the V3 loop of a Thailand HIV isolate. J. Biomol. Struct. Dyn. 11:345–66 www.annualreviews.org • GP120-Targeted Vaccine Design

85. This detailed report describes the first crystal structure of a domain-swapped anticarbohydrate antibody.

763

ARI

15 February 2006

3:9

95. Keller PM, Arnold BA, Shaw AR, Tolman RL, Van Middlesworth F, et al. 1993. Identiﬁcation of HIV vaccine candidate peptides by screening random phage epitope libraries. Virology 193:709–16 96. Rosen O, Chill J, Sharon M, Kessler N, Mester B, et al. 2005. Induced ﬁt in HIVneutralizing antibody complexes: evidence for alternative conformations of the gp120 V3 loop and the molecular basis for broad neutralization. Biochemistry 44:7250–58 97. Hartley O, Klasse PJ, Sattentau QJ, Moore JP. 2005. V3: HIV’s switch-hitter. AIDS Res. Hum. Retrovir. 21:171–89 98. Back NK, Smit L, De Jong JJ, Keulen W, Schutten M, et al. 1994. An N-glycan within the human immunodeﬁciency virus type 1 gp120 V3 loop affects virus neutralization. Virology 199:431–38 99. Malenbaum SE, Yang D, Cavacini L, Posner M, Robinson J, Cheng-Mayer C. 2000. The N-terminal V3 loop glycan modulates the interaction of clade A and B human immunodeﬁciency virus type 1 envelopes with CD4 and chemokine receptors. J. Virol. 74:11008–16 100. Koch M, Pancera M, Kwong PD, Kolchinsky P, Grundner C, et al. 2003. Structurebased, targeted deglycosylation of HIV-1 gp120 and effects on neutralization sensitivity and antibody recognition. Virology 313:387–400 101. McCaffrey RA, Saunders C, Hensel M, Stamatatos L. 2004. N-linked glycosylation of the V3 loop and the immunologically silent face of gp120 protects human immunodeﬁciency virus type 1 SF162 from neutralization by anti-gp120 and anti-gp41 antibodies. J. Virol. 78:3279–95 102. Gram GJ, Hemming A, Bolmstedt A, Jansson B, Olofsson S, et al. 1994. Identiﬁcation of an N-linked glycan in the V1-loop of HIV-1 gp120 inﬂuencing neutralization by anti-V3 antibodies and soluble CD4. Arch. Virol. 139:253–61 103. Schonning K, Jansson B, Olofsson S, Hansen JE. 1996. Rapid selection for an N-linked oligosaccharide by monoclonal antibodies directed against the V3 loop of human immunodeﬁciency virus type 1. J. Gen. Virol. 77(Pt. 4):753–58 104. Ly A, Stamatatos L. 2000. V2 loop glycosylation of the human immunodeﬁciency virus type 1 SF162 envelope facilitates interaction of this protein with CD4 and CCR5 receptors and protects the virus from neutralization by anti-V3 loop and anti-CD4 binding site antibodies. J. Virol. 74:6769–76 105. Krachmarov C, Pinter A, Honnen WJ, Gorny MK, Nyambi PN, et al. 2005. Antibodies that are cross-reactive for human immunodeﬁciency virus type 1 clade A and clade B V3 domains are common in patient sera from Cameroon, but their neutralization activity is usually restricted by epitope masking. J. Virol. 79:780–90 106. Salzwedel K, Smith ED, Dey B, Berger EA. 2000. Sequential CD4-coreceptor interactions in human immunodeﬁciency virus type 1 Env function: soluble CD4 activates Env for coreceptor-dependent fusion and reveals blocking activities of antibodies against cryptic conserved epitopes on gp120. J. Virol. 74:326–33 107. Xiang SH, Doka N, Choudhary RK, Sodroski J, Robinson JE. 2002. Characterization of CD4-induced epitopes on the HIV type 1 gp120 envelope glycoprotein recognized by neutralizing human monoclonal antibodies. AIDS Res. Hum. Retrovir. 18:1207– 17 108. Labrijn AF, Poignard P, Raja A, Zwick MB, Delgado K, et al. 2003. Access of antibody molecules to the conserved coreceptor binding site on glycoprotein gp120 is sterically restricted on primary human immunodeﬁciency virus type 1. J. Virol. 77:10557– 65

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

764

Pantophlet

·

Burton

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

109. Choe H, Li W, Wright PL, Vasilieva N, Venturi M, et al. 2003. Tyrosine sulfation of human antibodies contributes to recognition of the CCR5 binding region of HIV-1 gp120. Cell 114:161–70 110. Dey B, Del Castillo CS, Berger EA. 2003. Neutralization of human immunodeﬁciency virus type 1 by sCD4-17b, a single-chain chimeric protein, based on sequential interaction of gp120 with CD4 and coreceptor. J. Virol. 77:2859–65 111. Ditzel HJ, Binley JM, Moore JP, Sodroski J, Sullivan N, et al. 1995. Neutralizing recombinant human antibodies to a conformational V2- and CD4-binding site-sensitive epitope of HIV-1 gp120 isolated by using an epitope-masking procedure. J. Immunol. 154:893–906 112. He Y, Honnen WJ, Krachmarov CP, Burkhart M, Kayman SC, et al. 2002. Efﬁcient isolation of novel human monoclonal antibodies with neutralizing activity against HIV-1 from transgenic mice expressing human Ig loci. J. Immunol. 169:595–605 113. Ho DD, Fung MS, Cao YZ, Li XL, Sun C, et al. 1991. Another discontinuous epitope on glycoprotein gp120 that is important in human immunodeﬁciency virus type 1 neutralization is identiﬁed by a monoclonal antibody. Proc. Natl. Acad. Sci. USA 88:8949– 52 114. McKeating JA, Shotton C, Cordell J, Graham S, Balfe P, et al. 1993. Characterization of neutralizing monoclonal antibodies to linear and conformation-dependent epitopes within the ﬁrst and second variable domains of human immunodeﬁciency virus type 1 gp120. J. Virol. 67:4932–44 115. Shotton C, Arnold C, Sattentau Q, Sodroski J, McKeating JA. 1995. Identiﬁcation and characterization of monoclonal antibodies speciﬁc for polymorphic antigenic determinants within the V2 region of the human immunodeﬁciency virus type 1 envelope glycoprotein. J. Virol. 69:222–30 116. Gorny MK, Moore JP, Conley AJ, Karwowska S, Sodroski J, et al. 1994. Human antiV2 monoclonal antibody that neutralizes primary but not laboratory isolates of human immunodeﬁciency virus type 1. J. Virol. 68:8312–20 117. Wang N, Zhu T, Ho DD. 1995. Sequence diversity of V1 and V2 domains of gp120 from human immunodeﬁciency virus type 1: lack of correlation with viral phenotype. J. Virol. 69:2708–15 118. Simmonds P, Balfe P, Ludlam CA, Bishop JO, Brown AJ. 1990. Analysis of sequence diversity in hypervariable regions of the external glycoprotein of human immunodeﬁciency virus type 1. J. Virol. 64:5840–50 119. Cornelissen M, Hogervorst E, Zorgdrager F, Hartman S, Goudsmit J. 1995. Maintenance of syncytium-inducing phenotype of HIV type 1 is associated with positively charged residues in the HIV type 1 gp120 V2 domain without ﬁxed positions, elongation, or relocated N-linked glycosylation sites. AIDS Res. Hum. Retrovir. 11:1169–75 120. Groenink M, Fouchier RA, Broersen S, Baker CH, Koot M, et al. 1993. Relation of phenotype evolution of HIV-1 to envelope V2 conﬁguration. Science 260:1513–16 121. Srivastava IK, Ulmer JB, Barnett SW. 2004. Neutralizing antibody responses to HIV: role in protective immunity and challenges for vaccine design. Expert Rev. Vaccines 3:S33–52 122. Center RJ, Lebowitz J, Leapman RD, Moss B. 2004. Promoting trimerization of soluble human immunodeﬁciency virus type 1 (HIV-1) Env through the use of HIV-1/simian immunodeﬁciency virus chimeras. J. Virol. 78:2265–76 123. Earl PL, Broder CC, Long D, Lee SA, Peterson J, et al. 1994. Native oligomeric human immunodeﬁciency virus type 1 envelope glycoprotein elicits diverse monoclonal antibody reactivities. J. Virol. 68:3015–26 www.annualreviews.org • GP120-Targeted Vaccine Design

765

ARI

15 February 2006

3:9

124. Yang XZ, Farzan M, Wyatt R, Sodroski J. 2000. Characterization of stable, soluble trimers containing complete ectodomains of human immunodeﬁciency virus type 1 envelope glycoproteins. J. Virol. 74:5716–25 125. Srivastava IK, Stamatatos L, Legg H, Kan E, Fong A, et al. 2002. Puriﬁcation and characterization of oligomeric envelope glycoprotein from a primary R5 subtype B human immunodeﬁciency virus. J. Virol. 76:2835–47 126. VanCott TC, Veit SC, Kalyanaraman V, Earl P, Birx DL. 1995. Characterization of a soluble, oligomeric HIV-1 gp160 protein as a potential immunogen. J. Immunol. Methods 183:103–17 127. Zhang CW, Chishti Y, Hussey RE, Reinherz EL. 2001. Expression, puriﬁcation, and characterization of recombinant HIV gp140. The gp41 ectodomain of HIV or simian immunodeﬁciency virus is sufﬁcient to maintain the retroviral envelope glycoprotein as a trimer. J. Biol. Chem. 276:39577–85 128. Chakrabarti BK, Kong WP, Wu BY, Yang ZY, Friborg J, et al. 2002. Modiﬁcations of the human immunodeﬁciency virus envelope glycoprotein enhance immunogenicity for genetic immunization. J. Virol. 76:5357–68 129. Chen B, Zhou GF, Kim MY, Chishti Y, Hussey RE, et al. 2000. Expression, puriﬁcation, and characterization of gp160e, the soluble, trimeric ectodomain of the simian immunodeﬁciency virus envelope glycoprotein, gp160. J. Biol. Chem. 275:34946–53 130. Farzan M, Choe H, Desjardins E, Sun Y, Kuhn J, et al. 1998. Stabilization of human immunodeﬁciency virus type 1 envelope glycoprotein trimers by disulﬁde bonds introduced into the gp41 glycoprotein ectodomain. J. Virol. 72:7620–25 131. Binley JM, Sanders RW, Clas B, Schuelke N, Master A, et al. 2000. A recombinant human immunodeﬁciency virus type 1 envelope glycoprotein complex stabilized by an intermolecular disulﬁde bond between the gp120 and gp41 subunits is an antigenic mimic of the trimeric virion-associated structure. J. Virol. 74:627–43 132. Pancera M, Wyatt R. 2005. Selective recognition of oligomeric HIV-1 primary isolate envelope glycoproteins by potently neutralizing ligands requires efﬁcient precursor cleavage. Virology 332:145–56 133. Sanders RW, Schiffner L, Master A, Kajumo F, Guo Y, et al. 2000. Variable-loop-deleted variants of the human immunodeﬁciency virus type 1 envelope glycoprotein can be stabilized by an intermolecular disulﬁde bond between the gp120 and gp41 subunits. J. Virol. 74:5091–100 134. Schulke N, Vesanen MS, Sanders RW, Zhu P, Lu M, et al. 2002. Oligomeric and conformational properties of a proteolytically mature, disulﬁde-stabilized human immunodeﬁciency virus type 1 gp140 envelope glycoprotein. J. Virol. 76:7760–76 135. Binley JM, Sanders RW, Master A, Cayanan CS, Wiley CL, et al. 2002. Enhancing the proteolytic maturation of human immunodeﬁciency virus type 1 envelope glycoproteins. J. Virol. 76:2606–16 136. Jeffs SA, Goriup S, Kebble B, Crane D, Bolgiano B, et al. 2004. Expression and characterisation of recombinant oligomeric envelope glycoproteins derived from primary isolates of HIV-1. Vaccine 22:1032–46 137. Staropoli I, Chanel C, Girard M, Altmeyer R. 2000. Processing, stability, and receptor binding properties of oligomeric envelope glycoprotein from a primary HIV-1 isolate. J. Biol. Chem. 275:35137–45 138. Yang XZ, Florin L, Farzan M, Kolchinsky P, Kwong PD, et al. 2000. Modiﬁcations that stabilize human immunodeﬁciency virus envelope glycoprotein trimers in solution. J. Virol. 74:4746–54

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

766

Pantophlet

·

Burton

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

139. Yang XZ, Lee J, Mahony EM, Kwong PD, Wyatt R, Sodroski J. 2002. Highly stable trimers formed by human immunodeﬁciency virus type 1 envelope glycoproteins fused with the trimeric motif of T4 bacteriophage ﬁbritin. J. Virol. 76:4634–42 140. Chen B, Cheng YF, Calder L, Harrison SC, Reinherz EL, et al. 2004. A chimeric protein of simian immunodeﬁciency virus envelope glycoprotein gp140 and Escherichia coli aspartate transcarbamoylase. J. Virol. 78:4508–16 141. Yang XZ, Wyatt R, Sodroski J. 2001. Improved elicitation of neutralizing antibodies against primary human immunodeﬁciency viruses by soluble stabilized envelope glycoprotein trimers. J. Virol. 75:1165–71 142. Grundner C, Li YX, Louder M, Mascola J, Yang XZ, et al. 2005. Analysis of the neutralizing antibody response elicited in rabbits by repeated inoculation with trimeric HIV-1 envelope glycoproteins. Virology 331:33–46 143. Earl PL, Sugiura W, Monteﬁori DC, Broder CC, Lee SA, et al. 2001. Immunogenicity and protective efﬁcacy of oligomeric human immunodeﬁciency virus type 1 gp140. J. Virol. 75:645–53 144. Kim M, Qiao ZS, Monteﬁori DC, Haynes BF, Reinherz EL, Liao HX. 2005. Comparison of HIV type 1 ADA gp120 monomers versus gp140 trimers as immunogens for the induction of neutralizing antibodies. AIDS Res. Hum. Retrovir. 21:58–67 145. Sanders RW, Vesanen M, Schuelke N, Master A, Schiffner L, et al. 2002. Stabilization of the soluble, cleaved, trimeric form of the envelope glycoprotein complex of human immunodeﬁciency virus type 1. J. Virol. 76:8875–89 146. Beddows S, Schulke N, Kirschner M, Barnes K, Franti M, et al. 2005. Evaluating the immunogenicity of a disulﬁde-stabilized, cleaved, trimeric form of the envelope glycoprotein complex of human immunodeﬁciency virus type 1. J. Virol. 79:8812–27 147. Rossio JL, Esser MT, Suryanarayana K, Schneider DK, Bess JW Jr, et al. 1998. Inactivation of human immunodeﬁciency virus type 1 infectivity with preservation of conformational and functional integrity of virion surface proteins. J. Virol. 72:7992–8001 148. Arthur LO, Bess JW Jr, Chertova EN, Rossio JL, Esser MT, et al. 1998. Chemical inactivation of retroviral infectivity by targeting nucleocapsid protein zinc ﬁngers: a candidate SIV vaccine. AIDS Res. Hum. Retrovir. 14(Suppl. 3):S311–19 149. Grundner C, Mirzabekov T, Sodroski J, Wyatt R. 2002. Solid-phase proteoliposomes containing human immunodeﬁciency virus envelope glycoproteins. J. Virol. 76:3511– 21 150. Lifson JD, Rossio JL, Piatak M Jr, Bess J Jr, Chertova E, et al. 2004. Evaluation of the safety, immunogenicity, and protective efﬁcacy of whole inactivated simian immunodeﬁciency virus (SIV) vaccines with conformationally and functionally intact envelope glycoproteins. AIDS Res. Hum. Retrovir. 20:772–87 151. Poon B, Safrit JT, McClure H, Kitchen C, Hsu JF, et al. 2005. Induction of humoral immune responses following vaccination with envelope-containing, formaldehyde-treated, thermally inactivated human immunodeﬁciency virus type 1. J. Virol. 79:4927–35 152. Doan LX, Li M, Chen CY, Yao QZ. 2005. Virus-like particles as HIV-1 vaccines. Rev. Med. Virol. 15:75–88 153. Monteﬁori DC, Safrit JT, Lydy SL, Barry AP, Bilska M, et al. 2001. Induction of neutralizing antibodies and gag-speciﬁc cellular immune responses to an R5 primary isolate of human immunodeﬁciency virus type 1 in rhesus macaques. J. Virol. 75:5879–90 154. Dale CJ, Liu XS, De Rose R, Purcell DF, Anderson J, et al. 2002. Chimeric human papilloma virus-simian/human immunodeﬁciency virus virus-like-particle vaccines: immunogenicity and protective efﬁcacy in macaques. Virology 301:176–87 www.annualreviews.org • GP120-Targeted Vaccine Design

767

ARI

15 February 2006

3:9

155. Buonaguro L, Racioppi L, Tornesello ML, Arra C, Visciano ML, et al. 2002. Induction of neutralizing antibodies and cytotoxic T lymphocytes in BALB/c mice immunized with virus-like particles presenting a gp120 molecule from a HIV-1 isolate of clade A. Antivir. Res. 54:189–201 156. Barnett SW, Lu S, Srivastava I, Cherpelis S, Gettie A, et al. 2001. The ability of an oligomeric human immunodeﬁciency virus type 1 (HIV-1) envelope antigen to elicit neutralizing antibodies against primary HIV-1 isolates is improved following partial deletion of the second hypervariable region. J. Virol. 75:5526–40 157. Lu S, Wyatt R, Richmond JF, Mustafa F, Wang S, et al. 1998. Immunogenicity of DNA vaccines expressing human immunodeﬁciency virus type 1 envelope glycoprotein with and without deletions in the V1/2 and V3 regions. AIDS Res. Hum. Retrovir. 14:151– 55 158. Kang SM, Quan FS, Huang CZ, Guo LH, Ye L, et al. 2005. Modiﬁed HIV envelope proteins with enhanced binding to neutralizing monoclonal antibodies. Virology 331:20– 32 159. Kim YB, Han DP, Cao C, Cho MW. 2003. Immunogenicity and ability of variable loop-deleted human immunodeﬁciency virus type 1 envelope glycoproteins to elicit neutralizing antibodies. Virology 305:124–37 160. Cho MW, Kim YB, Lee MK, Gupta KC, Ross W, et al. 2001. Polyvalent envelope glycoprotein vaccine elicits a broader neutralizing antibody response but is unable to provide sterilizing protection against heterologous simian/human immunodeﬁciency virus infection in pigtailed macaques. J. Virol. 75:2224–34 161. Chakrabarti BK, Ling X, Yang ZY, Monteﬁori DC, Panet A, et al. 2005. Expanded breadth of virus neutralization after immunization with a multiclade envelope HIV vaccine candidate. Vaccine 23:3434–45 162. Kim JH, Pitisuttithum P, Kamboonruang C, Chuenchitra T, Mascola J, et al. 2003. Speciﬁc antibody responses to vaccination with bivalent CM235/SF2 gp120: detection of homologous and heterologous neutralizing antibody to subtype E (CRF01.AE) HIV type 1. AIDS Res. Hum. Retrovir. 19:807–16 163. Lemiale F, Brand D, Lebigot S, Verrier B, Buzelay L, et al. 2001. Immunogenicity of recombinant envelope glycoproteins derived from T-cell line-adapted isolates or primary HIV isolates: a comparative study using multivalent vaccine approaches. J. Acquir. Immune Deﬁc. Syndr. 26:413–22 164. Gaschen B, Taylor J, Yusim K, Foley B, Gao F, et al. 2002. Diversity considerations in HIV-1 vaccine selection. Science 296:2354–60 165. Gao F, Korber BT, Weaver E, Liao HX, Hahn BH, Haynes BF. 2004. Centralized immunogens as a vaccine strategy to overcome HIV-1 diversity. Expert Rev. Vaccines 3:S161– 68 166. Gao F, Weaver EA, Lu ZJ, Li YY, Liao HX, et al. 2005. Antigenicity and immunogenicity of a synthetic human immunodeﬁciency virus type 1 group M consensus envelope glycoprotein. J. Virol. 79:1154–63 167. Mullins JI, Nickle DC, Heath L, Rodrigo AG, Learn GH. 2004. Immunogen sequence: the fourth tier of AIDS vaccine design. Expert Rev. Vaccines 3:S151–59 168. Pantophlet R, Burton DR. 2003. Immunofocusing: antigen engineering to promote the induction of HIV-neutralizing antibodies. Trends Mol. Med. 9:468–73 169. Garrity RR, Rimmelzwaan G, Minassian A, Tsai WP, Lin G, et al. 1997. Refocusing neutralizing antibody response by targeted dampening of an immunodominant epitope. J. Immunol. 159:279–89

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

768

Pantophlet

·

Burton

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-24

ARI

15 February 2006

3:9

170. Pantophlet R, Wilson IA, Burton DR. 2003. Hyperglycosylated mutants of human immunodeﬁciency virus (HIV) type 1 monomeric gp120 as novel antigens for HIV vaccine design. J. Virol. 77:5889–901 171. Pantophlet R, Wilson IA, Burton DR. 2004. Improved design of an antigen with enhanced speciﬁcity for the broadly HIV-neutralizing antibody b12. Protein Eng. Des. Sel. 17:749–58 172. Selvarajah S, Puffer B, Pantophlet R, Law M, Doms RW, Burton DR. 2005. Comparing antigenicity and immunogenicity of engineered gp120. J. Virol. 79:12148–63 173. Grundner C, Pancera M, Kang JM, Koch M, Sodroski J, Wyatt R. 2004. Factors limiting the immunogenicity of HIV-1 gp120 envelope glycoproteins. Virology 330:233–48 174. Lee HK, Scanlan CN, Huang CY, Chang AY, Calarese DA, et al. 2004. Reactivity-based one-pot synthesis of oligomannoses: deﬁning antigens recognized by 2G12, a broadly neutralizing anti-HIV-1 antibody. Angew. Chem. Int. Ed. Engl. 43:1000–3 175. Li HG, Wang LX. 2004. Design and synthesis of a template-assembled oligomannose cluster as an epitope mimic for human HIV-neutralizing antibody 2G12. Org. Biomol. Chem. 2:483–88 176. Wang LX, Ni JH, Singh S, Li HG. 2004. Binding of high-mannose-type oligosaccharides and synthetic oligomannose clusters to human antibody 2G12: implications for HIV-1 vaccine design. Chem. Biol. 11:127–34 177. Geng XD, Dudkin VY, Mandal M, Danishefsky SJ. 2004. In pursuit of carbohydratebased HIV vaccines, part 2: The total synthesis of high-mannose-type gp120 fragments– evaluation of strategies directed to maximal convergence. Angew. Chem. Int. Ed. Engl. 43:2562–65 178. Dudkin VY, Orlova M, Geng XD, Mandal M, Olson WC, Danishefsky SJ. 2004. Toward fully synthetic carbohydrate-based HIV antigen design: on the critical role of bivalency. J. Am. Chem. Soc. 126:9560–62 179. Huang CC, Tang M, Zhang MY, Majeed S, Montabana E, et al. 2005. Structure of a V3-containing HIV-1 gp120 core. Science 310:1025–28

RELATED RESOURCES Zinkernagel RM. 2003. On natural and artiﬁcial vaccinations. Annu. Rev. Immunol. 21:515–46 Letvin NL, Barouch DH, Monteﬁori DC. 2002. Prospects for vaccine protection against HIV-1 infection and AIDS. Annu. Rev. Immunol. 20:73–99 Graham BS. 2002. Clinical trials of HIV vaccines. Annu. Rev. Med. 53:207–21

www.annualreviews.org • GP120-Targeted Vaccine Design

769

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:739-769. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents

ANRV270-IY24-25

ARI

15 February 2006

3:17

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips Departments of Medicine, Cell Biology, and Pharmacology, New York University Medical Center, New York, NY 10016–6402; email: [email protected], [email protected]

Annu. Rev. Immunol. 2006. 24:771–800 First published online as a Review in Advance on January 16, 2006 The Annual Review of Immunology is online at immunol.annualreviews.org This article’s doi: 10.1146/annurev.immunol.24.021605.090723 c 2006 by Copyright Annual Reviews. All rights reserved 0732-0582/06/0423-0771$20.00

Key Words MAP kinase, GTPases, cell signaling, oncogenes

Abstract Signal transduction down the Ras/MAPK pathway, including that critical to T cell activation, proliferation, and differentiation, has been generally considered to occur at the plasma membrane. It is now clear that the plasma membrane does not represent the only platform for Ras/MAPK signaling. Moreover, the plasma membrane itself is no longer considered a uniform structure but rather a patchwork of microdomains that can compartmentalize signaling. Signaling on internal membranes was ﬁrst recognized on endosomes. Genetically encoded ﬂuorescent probes for signaling events such as GTP/GDP exchange on Ras have revealed signaling on a variety of intracellular membranes, including the Golgi apparatus. In ﬁbroblasts, Ras is activated on the plasma membrane and Golgi with distinct kinetics. The pathway by which Golgi-associated Ras becomes activated involves PLCγ and RasGRP1 and may also require retrograde trafﬁcking of Ras from the plasma membrane to the Golgi as a consequence of depalmitoylation. Thus, the Ras/MAPK pathway represents a clear example of compartmentalized signaling.

771

ANRV270-IY24-25

ARI

15 February 2006

Ras: rat sarcoma viral oncogene MAPK: mitogen-activated protein kinase ER: endoplasmic reticulum

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

GEF: guanine nucleotide-exchange factor GAP: GTPase-activating protein

772

3:17

INTRODUCTION

Ras BIOLOGY

Signal transduction involves the transmission of biochemical information from one part of the cell to another. The best-studied signaling paradigms involve conveying information from the extracellular environment into the cell, often to the nucleus. Because many of the receptors that sense extracellular ligands are transmembrane proteins that traverse the plasma membrane, this compartment has been considered to be the primary platform upon which signaling complexes are assembled. This emphasis on the plasma membrane has certainly been true for T lymphocyte signaling, for which the antigen receptor initiates a nexus of signaling involving kinases, adapter proteins, scaffolds, GTPases, and phospholipases, all of which are generally considered to function at the inner leaﬂet of the plasma membrane. The advent of genetically encoded ﬂuorescent probes for signaling events have enabled spatiotemporal analysis of signaling in living cells. Using these methods, signaling events previously presumed to be restricted to the plasma membrane have been observed on intracellular membranes. Among the signaling pathways that have provided the biggest surprises when analyzed spatially in living cells is the Ras/MAPK pathway. In addition to the plasma membrane, Ras and/or MAPK signaling has now been observed on endosomes, the endoplasmic reticulum (ER), the Golgi apparatus, and mitochondria. Ras signaling in lymphocytes has provided perhaps the biggest surprise of all because the primary platform for signaling appears to be the Golgi apparatus. Subcellular compartmentalization of signaling, such as that regulated by Ras, provides one explanation for the apparent complexity of signaling outputs elaborated by individual signaling molecules. In this review, we examine the current evidence for compartmentalized signaling, focusing on the Ras/MAPK pathway and emphasizing whenever possible evidence provided by the study of lymphocytes.

The GTPase

Mor

·

Philips

The great fascination of cell biologists with Ras stems from the early association of the ras gene with cancer (1). Once the function of Ras as a GTPase was discovered, it became the prototypical monomeric GTPase. Much of the biochemistry that applies to the large superfamily of small GTPases was ﬁrst elucidated by the intensive study of Ras. Ras proteins are molecular switches that cycle between inactive, GDP-bound, and active, GTP-bound, forms. Signal-induced conversion of the inactive to active state is mediated by guanine nucleotide-exchange factors (GEFs) that stimulate the exchange of GDP for GTP. This is accomplished by catalyzing the release of GDP from the guanine nucleotide binding pocket. Once it is nucleotide free, Ras will bind GTP because GTP is tenfold more abundant in cellular cytosol than is GDP. GTP binding induces Ras activation by causing a marked conformational change of the so-called switch I and switch II regions (2). These regions contribute to the effectorbinding domain that engages downstream signaling elements only when the protein is in the GTP-bound state. The activation state of Ras is self-limited by the intrinsic GTPase activity of the protein. However, Ras, like most signaling GTPases, is a poor enzyme, with a Kcat of 2.3 × 10−4 sec−1 (3). The catalytic activity of Ras is greatly accelerated by a class of accessory proteins known as a GTPase activating proteins (GAPs). GAPs afford a critical level of regulation, allowing signaling to proceed for a relatively brief interval. The study of the biological function of Ras has been greatly facilitated by the availability of constitutively active as well as dominant-negative forms. The former are GTP-bound and oncogenic (see below), and the latter mimic the nucleotide-free transition state such that they have a high afﬁnity for GEFs and stoichiometrically sequester these regulatory molecules (4).

ANRV270-IY24-25

ARI

15 February 2006

3:17

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

The Oncogene Ras genes encoded by rat sarcoma viruses, v-H-ras and v-K-ras, were among the ﬁrst oncogenes to be recognized (5). These viral genes proved to be mutant forms of cellular protooncogenes (6–8). The commonly occurring mutations that render Ras oncogenic (at codons 12, 13 and 61) are those that make the GTPase insensitive to the action of GAPs and thereby lock it in the GTP-bound, active state (1, 9). Activated Ras alleles constitute the oncogene most frequently associated with human carcinomas (1), highlighting the great impact of Ras on human health. Mammalian genomes encode three ras genes that give rise to four gene products, N-Ras, H-Ras, KRas4A, and K-Ras4B, that are expressed ubiquitously, although isoform ratios vary from tissue to tissue. K-Ras4A and K-Ras4B are splice variants of the kras gene that use alternative fourth exons. Thirty percent of all human cancers harbor mutant ras genes. Tumors differ both in the isoforms associated with the disease and in the incidence of mutations of that isoform. For example, whereas 90% of pancreatic adenocarcinomas are associated with an oncogenic Ras mutation that is invariably in the kras gene, only 10% of bladder carcinomas harbor Ras mutations, and these occur in the hras gene (10).

The Protooncogene: Ras and T Cell Function As an oncogene, Ras regulates pathways that lead to cellular proliferation and survival. However, proliferation is but one of many cellular functions that can now be ascribed to protooncogenic Ras. Among the cells in which the role of protooncogenic Ras has been most extensively studied are lymphocytes (11). Indeed, engagement of the T cell receptor (TCR) was the ﬁrst physiologic stimulus shown to activate Ras (12). Subsequent studies have demonstrated that Ras is required for thymocyte development, T cell proliferation, and IL-2 production (13). Transgenic mice that express a dominant-negative form

of Ras in the thymus have a defect in positive but not negative selection of thymocytes (14). Conversely, constitutively active Ras expressed in RAG-2-deﬁcient thymocytes promotes the expansion of double-negative cells and their transition to double-positive cells (15). Constitutively active Ras expressed in T cells also results in overexpression of the early activating antigen, CD69, on the cell surface as well as activation of AP-1 in the nucleus. Targeted disruption of the Ras exchange factor Ras guanine nucleotidereleasing protein 1 (RasGRP1) results in a defect in thymocyte differentiation and decreased mature CD4 and CD8 cells, supporting the idea that Ras signaling downstream of the TCR is critical for thymocyte development (16). Autoimmunity that develops in adult RasGRP1-deﬁcient mice suggests that Ras signaling may also be required to maintain peripheral T cell tolerance (17). Recent experiments with a RasGRP1-deﬁcient Jurkat T cell line and RasGRP1 siRNA in wild-type cells conﬁrmed that this GEF is required for optimal antigen receptor–triggered Ras-Erk activation and that this pathway is augmented by phosphorylation by atypical protein kinase Cs (PKCs) of threonine residue 184 of RasGRP1 (18).

RasGRP1: Ras guananine nucleotide-releasing protein 1

Ras PROCESSING AND TRAFFICKING Ras proteins are localized on the cytosolic leaﬂet of cellular membranes (19, 20), and this localization is believed to be absolutely required for biological activity (21). Ras proteins are not intrinsic membrane proteins in that they lack signal sequences and hydrophobic membrane-spanning domains. Rather they are synthesized as hydrophilic proteins on free polysomes in the cytosol and targeted posttranslationally to cellular membranes by virtue of a series of modiﬁcations that include prenylation, proteolysis, and carboxyl methylation (22). Ras GTPases are the founding members of a large class of proteins that terminate in a CAAX motif, in

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

773

ARI

15 February 2006

3:17

which C is cysteine, A is usually but not always an aliphatic amino acid, and X is any amino acid (23). The CAAX sequence is recognized by one of two cytosolic prenyltransferases (24), farnesyl transferase (FTase) or geranylgeranyltransferase type I (GGTase I). The amino acid in the X position determines which prenyltransferase modiﬁes the protein (S, M, A or Q for FTase and L for GGTase I). FTase catalyzes the addition of a 15-carbon farnesyl isoprenoid, whereas GGTase I catalyzes the addition of a 20-carbon geranylgeranyl isoprenoid to the CAAX cysteine via a stable thioether linkage (25). Once prenylated, the S-isoprenyl CAAX moiety becomes a substrate for Rce1, a protease that cleaves the AAX sequence (22, 26). The newly Cterminal prenylcysteine is then recognized by a third enzyme, isoprenylcysteine carboxyl methyltransferase (Icmt), that methylesteriﬁes the carboxyl group (22, 23). Of these three modiﬁcations, only carboxyl methylation is reversible under physiologic conditions (27). The end result of these three modiﬁcations is to create a hydrophobic domain at the C-terminus that mediates membrane association. N-Ras, H-Ras, and Kras4A, but not K-Ras4B, are further modiﬁed by one or two palmitic acids just upstream of the farnesylcysteine. The Ras palmitoyltransferase was recently identiﬁed in a genetic screen of Saccharomyces cerevisiae as Erf2/4 (28). CAAX processing was originally conceived of as a process that targeted nascent cytosolic Ras proteins directly to the plasma membrane. The discovery that Icmt is restricted to the ER (29), and the subsequent characterization of the other postfarnesylation processing enzymes Rce1 (30) and Erf2/4 (31) as similarly restricted, led to a reevaluation of Ras trafﬁcking (Figure 1). CAAX processing alone is insufﬁcient for association with the plasma membrane but is instead a mechanism for directing proteins to the ER, where they encounter Rce1, Icmt, and Erf2/4, the processing enzymes that further modify prenylated proteins (20). The subsequent transfer of Ras proteins from the endomembrane to

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

774

Mor

·

Philips

the plasma membrane requires a second signal found in the C-terminal hypervariable regions of Ras proteins immediately adjacent to their CAAX motifs. For N-Ras, H-Ras, and K-Ras4A (the minor splice variant of the kras gene), this signal consists of one or two cysteine residues that serve as sites of palmitoylation. For K-Ras4B, the ubiquitous and predominant splice variant henceforth referred to as K-Ras, the signal consists of a polybasic region rich in lysine residues (32, 33). Whereas prenylation alone serves as a relatively weak membrane tether, palmitoylation markedly increases afﬁnity for membranes and traps N-Ras and H-Ras in the membrane compartment. Palmitoylated CAAX proteins such as N-Ras and H-Ras are translocated by vesicular transport (20) that may involve the classical secretory pathway or a nonclassical pathway (34). In contrast, CAAX proteins with a polybasic second signal such as K-Ras take an alternate, poorly deﬁned route to the plasma membrane. Recently two groups independently discovered a retrograde pathway of trafﬁcking of palmitoylated Ras isoforms (35, 36). In this pathway, mature N-Ras and HRas that have gained access to the plasma membrane are depalmitoylated and thereby lose sufﬁcient afﬁnity for membranes such that they partition into the cytosol. From there they regain access to the endomembrane, where they can be repalmitoylated. Thus, N-Ras and H-Ras undergo a palmitoylation/depalmitoylation cycle that regulates trafﬁcking from the plasma membrane to the Golgi and back again. This plasma membrane/Golgi cycle may be generalized to a wide range of palmitoylated signaling molecules, which has led to speculation that the bidirectional trafﬁc has a regulatory role in signaling (36). Recently, K-Ras has also been shown to trafﬁc in a retrograde fashion to the Golgi apparatus in hippocampal neurons stimulated with glutamate (37) and after phosphorylation by PKC in the polybasic region to the outer mitochondrial membrane in a variety of cells, including Jurkat T cells

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

Figure 1 Ras processing and trafﬁcking. Ras and other CAAX proteins are translated in the cytosol on free polysomes. Immediately posttranslationally, they become substrates for one of two cytosolic prenyltransferases. Prenylation targets the proteins to the ER, where they encounter the subsequent processing enzymes, Rce1 and Icmt. Once CAAX processing is complete, the pathways used by the various isoforms diverge. K-Ras is sent to the plasma membrane via an uncharacterized pathway and can be returned to the endomembrane following phosphorylation of the hypervariable region or through the action of calcium/calmodulin. In contrast, N-Ras and H-Ras are further processed on the Golgi by a palmitoyltransferase and then sent to the plasma membrane via vesicular transport. Retrograde trafﬁc of N-Ras and H-Ras back to the Golgi occurs following depalmitoylation.

(M.R. Philips, unpublished data). Thus, although until recently mature Ras proteins were thought to be stationary, it is now clear that they trafﬁc between cellular compartments. This revelation has made all the more prescient the question of spatiotemporal Ras signaling.

Ras SIGNALING The Erk Pathway The best characterized of the signaling pathways regulated by Ras is the MAPK pathway that proceeds through the MAPKs Erk1 and Erk2. This pathway is activated when

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

775

ANRV270-IY24-25

ARI

15 February 2006

3:17

any of several growth factors engage their cognate protein tyrosine kinase receptors (PTKRs). Ligation of PTKRs promotes their dimerization, which in turn allows crossphosphorylation of tyrosine residues in their cytosolic domains catalyzed by the kinase that is intrinsic to that domain (38). These phosphotyrosine residues serve as docking sites for signaling molecules and adapter proteins that contain SH2 (Src homology region 2) or PTB (phosphotyrosine-binding) domains. Among these is the adapter protein Grb2 that binds constitutively to SOS through its SH3 domain. SOS (son of sevenless) is a GEF for Ras proteins. Thus, phosphorylation of PTKRs leads to the recruitment of SOS to the plasma membrane, where it can encounter Ras. Once Ras is activated at the membrane, it recruits Raf-1, a serine/threonine kinase, to this compartment. The kinase activity of Raf-1 is activated when it associates with membranes through a complex set of regulatory events that are poorly understood (39). Once active, Raf-1 phosphorylates and activates MEK (MAPK/Erk kinase), a dual speciﬁcity tyrosine/threonine kinase, that in turn phosphorylates and activates Erk1 and Erk2. The Erk proteins are serine/threonine kinases that have numerous substrates, including cytosolic proteins like p90 ribosomal S6 kinase. Phospho-Erk forms dimers that are transported into the nucleus, where they phosphorylate the Ets family of transcription factors, including Elk-1. In this way the signal emanating from an extracellular growth factor is transmitted from the cell surface to the nucleus, and a transcriptional program is set in motion.

PTKR: protein tyrosine kinases receptor GPCR: G protein–coupled receptor

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

CCV: clathrin-coated vesicles

Other Pathways The Raf-1/Erk pathway is but one of many that are regulated by Ras. Ras effectors are deﬁned as proteins that bind Ras with a strong preference for the GTP-bound state. A more stringent deﬁnition requires that the function of the effector is modulated by the binding to GTP-Ras. The precise number of known 776

Mor

·

Philips

Ras effectors is somewhat uncertain because it depends how rigorously the more stringent deﬁnition is applied, but it appears that there are at least ten types of proteins that are putative effectors (40). Besides Raf-1 and its homologs, the best-characterized effectors are phosphatidylinositol 3-kinase (PI3K) and members of a family of exchange factors for the small GTPase Ral, e.g., RalGDS. PI3K activates Akt and thereby promotes cell survival. Although the Raf-1/MAPK pathway has been considered to be the most important with regard to Ras-mediated transformation of rodent ﬁbroblasts, recent studies have suggested that in human cells the RalGDS pathway is the most important in oncogenesis (41).

SIGNALING FROM ENDOSOMES PTKRs and Associated Proteins As described above, Ras links receptors such as PTKRs that traverse the plasma membrane with cytoplasmic effector pathways that regulate cellular function, including transcription. As such, it is intuitive to think of Ras signaling from the plasma membrane. Indeed, for the ﬁrst two decades of the study of Ras, the plasma membrane was considered the only platform from which Ras could signal. The ﬁrst indication that the plasma membrane may not be a unique platform for Ras/MAPK signaling came from the discovery that signaling complexes could be detected on endosomes (42, 43). Investigators have appreciated for decades that receptors such as PTKRs, G protein– coupled receptors (GPCRs) and TCRs are subject to removal from the cell surface by endocytosis mediated by clathrin-coated vesicles (CCVs). This process was originally viewed as one of the major mechanisms for receptor desensitization. Internalized receptors are either further trafﬁcked to lysosomes for degradation or recycled back to the plasma membrane (44). Among the proteins that regulate receptor internalization is c-Cbl. The Cbl family of proteins consists of c-Cbl, Cbl-b,

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

and Cbl-c, and these proteins function as E3 ubiquitin ligases (45, 46). c-Cbl monoubiquitinates a variety of receptors, including epidermal growth factor receptor (EGFR), in a ligand-dependent fashion and thereby targets the receptor for endocytosis and degradation via the proteosome and/or lysosome (47). An oncogenic, viral form of this protein, v-Cbl, interferes with the receptor degradation pathway and shunts receptors toward the recycling pathway (47), offering compelling evidence for the physiologic importance of endocytosis in downregulation of growth factor receptor signaling. Cbl is also a binding partner of Grb2, raising the possibility that, in addition to promoting receptor internalization, it inﬂuences Ras signaling (48). In the case of GPCRs, the β-arrrestins play a critical role in endocytosis by linking the receptor to elements of the clathrin coat (49). Among the GPCRs that are regulated by β-arrestinmediated endocytosis are chemokine receptors on lymphocytes that regulate a wide array of lymphocyte function and serve as coreceptors for HIV (50). T cell activation by speciﬁc antigen results in a rapid and profound internalization and degradation of the TCR (51). Since cCbl is required for T cell function, it has been suggested that, like PTKRs, TCRs may be downregulated by Cbl-induced endocytosis and degradation. Consistent with this idea, TCR surface expression is enhanced in mice deﬁcient in c-Cbl, and this is associated with altered positive selection in the thymus (52). Mice lacking Cbl-b develop autoimmunity, suggesting that in vivo immune tolerance is controlled by signaling pathways in T cells that are constitutively suppressed by Cbl-b (53). These mice also exhibit reduced TCR downregulation, increased cytokine production, and uncontrolled proliferation in response to anti-CD3 antibodies. Cbl ubiquitination has been visualized at the immunological synapse, and Cbl recruitment was found to be antigen dependent (53). As discussed above, the evidence for endocytosis as a mechanism for downregulating re-

ceptor signaling is strong. But there is another side to the story of endosomes and signaling. The ﬁrst evidence for signaling on endosomes came from subcellular fractionation studies in which Shc, Grb2, mSOS, and phospho-Raf1 were differentially observed on endosomes following stimulation with EGF or with insulin (42). Subsequently, the discovery that cells deﬁcient in clathrin-mediated endocytosis were also deﬁcient in Erk activation downstream of PTKRs provided compelling evidence that endocytosis could drive MAPK signaling rather than simply downregulate the pathway at the level of the receptor (54, 55). The most widely used method of blocking clathrin-mediated endocytosis has been to express the K44A dominant-negative form of the large GTPase dynamin that is believed to regulate scission of endosomes (54). Such studies have not been conﬁned to PTKRs but have also been applied to GPCRs. Interestingly, K44A blocks signaling from GPCRs to the MAPK pathway but does not block signaling to other pathways such as adenylcyclase. Moreover, Erk activation was blocked, but GTP loading of Ras was not (56). These studies have prompted a reevaluation of the role of endocytosis in signaling and have led to the idea of signaling complexes on endosomes. Several methods have been used to demonstrate that PTKRs remain active after internalization on endosomes. Unlike TGF-α, which dissociates from EGFR in early endosomes, EGF remains bound to its receptor, suggesting that ligands differ in their ability to sustain signaling on internalized receptors (57). Antiphosphotyrosine antibodies were used to reveal persistent phosphorylation of internalized EGFR (58). Sorkin and coworkers (59) demonstrated internalization of activated, CFP-tagged EGFRs by FRET using YFP-tagged Grb2 as a phosphotyrosine sensor. Importantly, neither the kinase activity nor the trafﬁcking of EGFR was altered by fusion with GFP (60). Using ﬂuorescence lifetime imaging as a sensitive read out of FRET, Wouters & Bastiaens (61) obtained

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

EGFR: epidermal growth factor receptor FRET: ﬂuorescence resonance energy transfer YFP: yellow ﬂuorescent protein GFP: green ﬂuorescent protein

777

ANRV270-IY24-25

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

RBD: Ras-binding domain

778

3:17

results similar to those of Sorkin et al. (59) by using GFP-tagged EGFR and microinjected antiphosphotyrosine antibodies conjugated with a FRET acceptor. In addition to phosphorylated, active PTKRs, other upstream components of Ras signaling have been localized on endosomes, including Shc, Grb2, and SOS (42, 58, 59, 62–65). PLCγ1 has also been localized on endosomes (66, 67), an observation that may have implications for T cells because Ras activation in these cells is downstream of this enzyme (11). Ras itself was observed on endosomes using subcellular fractionation (68). Using GFP fusion proteins and live cell imaging, Ras has been localized to vesicles including endosomes (20, 62). Importantly, Sorkin and colleagues (62) have used the Ras-binding domain (RBD) of Raf-1 fused to YFP to reveal GTP-bound Ras on vesicles decorated with internalized EGF. Hancock and colleagues (69) used dominant-negative dynamin K44A to show that H-Ras but not K-Ras signaling was dependent on endocytosis. Activated GFP-H-Ras12V was associated with Rab5GTP-induced macroendosomes, whereas activated GFP-K-Ras12V was not. Interestingly, GFP-Raf-1 was not observed on these structures. Intersectin, an interesting adapter protein in clathrin-mediated endocytosis, binds mSOS, providing a direct link between the endocytic machinery and Ras activation. Intersectin contains two Eps15 homology domains with which it binds epsin that in turn interacts with AP-2 and clathrin and thereby recruits intersectin to clathrin-coated pits (70). Intersectin also has ﬁve tandem SH3 domains, two of which bind dynamim and synaptojanin, other components of the endocytic machinery (71). One of the remaining SH3 domains binds to mSOS (72), which can activate Ras on endosomes (73). Functional studies reveal that intersectin is involved both in the formation of endosomes (74) and in mitogenic signaling (75). Thus, intersectin lives up to its name as a molecule at the intersection of clathrinmediated endocytosis and Ras signaling (76). Mor

·

Philips

Because endosomes derive from the plasma membrane, and because signaling from the latter compartment does not cease once endosomes have formed, demonstrating unambiguously that signals emanating from endosomes drive pathways that are also driven from the plasma membrane has been difﬁcult. Wang and colleagues (77) have taken a clever approach to this problem. They stimulated cells with EGF in the presence of a readily reversible EGFR kinase inhibitor and in the presence of monemsin, which blocks receptor recycling to the surface. After EGFR internalization, extracellular EGF was washed away, and signaling was subsequently permitted by washing away the EGFR kinase inhibitor. Under these conditions, Erk was phosphorylated and the cells received a survival signal, demonstrating effective signaling from endosomes.

Rap1 Whereas endosomes are not the primary compartment to which Ras proteins are targeted, these organelles are the primary location of Rap1, a closely related small GTPase (78). Rap1 is the GTPase most closely related to Ras and shares 100% of the effector domain amino acids known to contact the RBD of effectors (79). Rap1 has been implicated in a wide variety of cellular functions, including growth control and cell polarity (79). Although Rap1, like Ras, can bind to Raf-1, the ability of Rap1 to substitute for Ras in activating the MAPK pathway has been controversial (80). A recent genetic study in Drosophila indicated that Rap1 can indeed activate Erk in a Ras-independent fashion (81). Among the biological functions of Rap1 is the regulation of cellular adhesion, and within this category the best-studied pathway in which Rap1 is implicated is the regulation of integrins through “inside-out signaling” (82). Activated Rap1 stimulates lymphocyte function– associated antigen (LFA)-1-mediated T cell adhesion (83). The relevant effector for this pathway is RapL, a member of the RASSF family of tumor suppressors (84). RapL binds

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

to the cytoplasmic tail of the α chain of LFA1, suggesting a direct link to integrin afﬁnity modulation (85). As with most Ras family GTPases, the greatest divergence in sequence between Ras and Rap occurs in the membrane-targeting regions of the C-terminus. Unlike Ras proteins that are farnesylated, Rap1 is geranylgeranylated. In both ﬁbroblasts and T cells Rap1 is expressed predominantly on endosomes (78). Matsuda and colleagues (86) were the ﬁrst to study the subcellular localization of Rap1 activation in living cells. They used an innovative FRET probe called RAICHU and concluded that Rap1 was activated in the perinuclear region of the cell, although the limited spatial resolution of their method did not reveal a particular organelle as the site of activation (86). Using a reporter for active Rap1 based on the RBD of the exchange factor RalGDS that has a 50-fold higher afﬁnity for Rap1 than for Ras, Bivona et al. (78) observed Rap1 activation in ﬁbroblasts stimulated with growth factor and Jurkat T cells stimulated through the antigen receptor only on the plasma membrane (Figure 2). The discrepancy in the location of activation reported by these two methods likely stems from the fact that the RAICHU Rap1 reporter was ectopically targeted with a K-Ras sequence, whereas the RalGDS-based probe was untargeted and therefore spatially unbiased. The activation of Rap1 at the plasma membrane of T cells was recently independently conﬁrmed (87). Because Rap1 regulates the adhesive state of integrins that function at the plasma membrane, this is likely the site of activation of this GTPase. But if Rap1 functions at the plasma membrane, why is the bulk of the protein found on endosomes? Part of the answer is that Rap1 has multiple functions, and conditions under which Rap1 becomes active on endosomes may yet be found (88). Another possibility is that Rap1 on endosomes may serve as a storage pool analogous to the pool of integrins that is stored in the membranes of leukocyte granules. Consistent with this idea, the activation of Rap1 at the plasma

membrane of Jurkat cells is dependent on endosome recycling (78).

Special Cases: Neural Cells

NGF: nerve growth factor

What is the advantage to the cell in assembling or maintaining signaling complexes in the Ras/MAPK pathway on endosomes? Aside from the obvious answer that multiple signaling platforms increase the complexity of signaling outputs, another possible answer is that signaling endosomes allow messages to be delivered over great distances within cells. Nowhere is the need for such longdistance delivery more evident than in neuronal cells. Neurotropin receptors at nerve terminals must signal for events that take place in the neuronal cell body and nucleus that can be up to a meter from the terminal. Simple diffusion through the cytosol of, for example, phosphorylated Erk does not seem to be a reasonable solution for such long-range signaling. We now clearly know that, in neuronal cells, signaling endosomes are transported in a retrograde fashion via microtubules and function to convey activated neurotropin PTKR from the site of activation in the nerve terminal to the perinuclear region of the cell body (89). Indeed, the term signaling endosome was coined in the context of neurotropin signaling (90). Early efforts at isolating signaling endosomes from PC12 cells indicated that nerve growth factor (NGF)-binding enhanced the internalization of phosphorylated TrkA into CCVs (90) and that these vesicles contained signaling complexes that included Shc, Ras, Raf-1, and Erk (91). Increased levels of activated Ras were detected on these CCVs, and isolated CVVs were capable of stimulating the phosphorylation of Elk-1, an endogenous substrate of Erk. More recently, experiments using a cell body chamber that isolates the cell body from the neuron extensions showed that phosphorylated TrkA (a neurotropin PTKR) accumulates in the neuron cell bodies (92). Within the immune system, dendritic cells have the longest distances to cover in cell signaling because in some anatomical locations

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

779

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

3:17

Figure 2 Compartment-speciﬁc signaling of Ras and Rap1 in lymphocytes. The Ras-binding domains (RBDs) of various Ras family effectors can be used to construct genetically encoded ﬂuorescent probes of GTPase activity (schematic on left) that accumulate upon membranes where the cognate GTPase is activated (confocal micrographs on right). The RBD of Raf-1 is speciﬁc for active Ras whereas the RBD of RalGDS prefers GTP-bound Rap1. Stimulating Jurkat T cells by cross-linking the TCR results in N-Ras activation on the Golgi but not on the plasma membrane. Conversely, Rap1 is activated only at the plasma membrane.

such as the gut and skin the processes that ﬁrst encounter antigen can be quite far from the cell body (93). Whether efﬁcient antigen presentation depends on signaling endosomes in these cells remains to be determined. In addition to providing the most cogent example of long-distance signaling, neuronal cells also provide a well-established model in which the signaling output from endosomes differs from that emanating from the plasma membrane. PC12 cells derived from a rat pheochromocytoma have been used extensively to study neurotropin signaling because the various signaling outputs that include proliferation, cell survival, and differentiation can be separated (94). Whereas short-lived (minutes) MAPK signaling is associated with proliferation, long-term MAPK activity (hours) appears to be required for differentiation (95). Blocking endocytosis with the K44A mutant of dynamin blocked neurite outgrowth (dif780

Mor

·

Philips

ferentiation) in response to NGF but augmented PI3K signaling and cell survival (96). Monodansylcadaverine, a relatively crude inhibitor of endocytosis, blocked Rap1 but not Ras activation in PC12 cells, suggesting a differential requirement for endocytosis in the activation of these related GTPases (97). Interestingly, endocytosis may be differentially required for signaling from the same neurotropin receptor, depending on ligand. Whereas NGF induced internalization of the TrkA receptor and retrograde transport of signaling endosomes, NT-3, which also signals through TrkA, did not (98). Because NGF and NT-3 are required at different stages of neuronal development, this observation represents a dramatic example of compartmentspeciﬁc signaling at the level of a single receptor. Compartmentalized signaling on endosomes is not restricted to neuronal cells. On gastrointestinal epithelial cells, the PAR2

ANRV270-IY24-25

ARI

15 February 2006

3:17

GPCR signals from β-arrestin-induced endosomes to Erk via a Ras-independent pathway, but a mutant receptor that cannot associate with β-arrestin signals down an alternative, Ras-dependent pathway (99).

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Other Systems: TLRs, Smads, and Notch MAPK pathways are not the only ones that utilize endosome membranes as signaling platforms. Among the other pathways that require endosomes are those involved in antigen processing and innate immunity. For example, Toll-like receptor (TLR) 9 is associated with endosomes that move to the center of the cell upon ligation (100, 101). Whereas signaling of lipopolysaccharide via TLR4 does not require endocytosis, CpG-DNA signaling to TLR9 does (101). TLR signaling from endosomes illustrates another way that compartmentalized signaling on this organelle inﬂuences signaling outcome. Plasmacytoid dendritic cells are unique in their ability to mount a robust type I interferon response to TLR9 signaling through MyD88-IRF-7. Whereas ligated TLR9 is relatively rapidly delivered from endosomes to lysosomes in other cell types, the ligand receptor pair lingers for a much longer time in the endosomal compartment of plasmacytoid dendritic cells. Elegant studies designed to alter the endosomal retention time demonstrated that this was the critical variable in mounting an interferon response (102). TGF-β signaling also uses endosomes. Smad proteins are transcription factors that mediate TGF-β signaling (103). Some Smads, such as Smad2, are tethered to the endosome membrane in association with a protein known as Smad anchor for receptor activation (104). Upon activation, TGF-β receptors internalize into endosomes and phosphorylate Smads (105). Phosphorylated Smads translocate from the endosome to the nucleus to activate transcriptional factor– speciﬁc genes. Notch signaling is a conserved mechanism that transmits signaling between cells that are

in direct contact. Signaling is triggered when the Notch receptor at the cell surface binds to ligands of the DSL (Delta, Serrate, Lag2) family, such as Delta, that are presented on the surface of neighboring cells. Ligand binding leads to proteolysis of Notch at two sites, including an intramembranous site cleaved by γ-secretase. Notch cleavage releases an intracellular signaling fragment that regulates gene transcription. Genetic studies in Drosophila revealed a critical role for endocytosis in Notch signaling (106). Interestingly, signaling involved endocytosis of the Delta receptor while it was bound to the fragment of Notch from an adjacent cell in a process referred to as transendocytosis. New studies reveal that activation of Notch by γ-secretase cleavage requires prior monoubiquitination and endocytosis of the receptor (107).

KSR: kinase suppressor of Ras MP1: MEK partner 1

MAPK Scaffolds Scaffolding molecules in MAPK pathways were ﬁrst conceptualized as a means of organizing the various MAPK modules, e.g., Erk, Jnk, and p38 (108). The most dramatic demonstration of this concept was in budding yeast, in which Lim and colleagues (109) were able to rewire the Fus3 (Erk-like) and Hog1 (p38-like) pathways by expressing chimeric versions of the respective scaffolds, Ste5 and Pbs2. An emerging feature of mammalian MAPK scaffolds is that they appear to have speciﬁc subcellular localizations and are therefore now recognized as integral to compartmentalized signaling (Figure 3). The best-studied of the Erk scaffolds is kinase suppressor of Ras (KSR), ﬁrst identiﬁed in genetic screens in ﬂies (110) and worms (111, 112) as a positive regulator of the Ras/MAPK pathway (KSR is an unfortunate misnomer). KSR is a multidomain protein that binds Raf-1, MEK, and Erk, as well as several other proteins. In resting cells, it is sequestered in the cytosol, like Raf-1, by 14-3-3 proteins. In response to mitogenic signals, KSR becomes dephosphorylated at S392, loses afﬁnity for 14-3-3, and translocates to the plasma membrane (113,

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

781

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

3:17

Figure 3 Compartmentalization of MAPK scaffolds. The various MEK/Erk scaffolds have distinct subcellular localizations. Kinase suppressor of Ras (KSR) translocates from the cytosol to the plasma membrane, where it organizes Raf-1/MEK/Erk signaling. MEK partner 1 (MP1) is restricted to early endosomes by p14 and facilitates MEK1/Erk1 signaling on this compartment downstream of growth factor receptors. β-arrestin also localizes to endosomes and serves as a Raf-1/MEK/Erk scaffold downstream of GPCRs. Finally, Sef localizes MEK/Erk complexes to the cytosolic face of the Golgi apparatus and selectively allows phosphorylation of cytosolic substrates such as RSK2.

114) by virtue of a cysteine-rich domain (115). Thus, KSR represents an inducible scaffold for Ras/MAPK signaling with speciﬁcity for the plasma membrane. There are at least two ksr genes in mammals. Targeted disruption of ksr1 yielded mice that were relatively normal, although T cell activation was impaired (116). MEK partner 1 (MP1) was ﬁrst identiﬁed in a yeast two-hybrid screen as a binding partner of MEK1 (117). Further analysis revealed that it bound MEK1 and Erk1 but 782

Mor

·

Philips

not MEK2 and Erk2 and that binding facilitated the phosphorylation of Erk1 by MEK1, thus fulﬁlling the criteria for a MEK/Erk scaffold (117). Interestingly, MP1 was also found to bind p14, a highly conserved protein that resides on the cytoplasmic face of early endosomes (118). Overexpression studies revealed that MP1 could augment Erk signaling but only when coexpressed with p14. Importantly, ectopic targeting of the p14/MP1 complex to the plasma membrane using the C-terminus

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

of K-Ras failed to augment Erk activation, indicating that the endosomal location is essential for the scaffolding function of p14/MP1. Thus, MP1 is indeed an endosome-speciﬁc scaffold for MEK and Erk. By tracking pErk and employing p14 siRNA, Huber and colleagues (118) showed that whereas p14/MP1 was not required for early activation at the plasma membrane, it was required for the activation seen on endosomes 10-30 min after EGF stimulation. This illustrates an important feature of compartmentalized signaling facilitated by localized scaffolds: It allows for the same pathway to be activated in a given cell with different kinetics, thus increasing the complexity of signaling. MP1 is not the only MAPK scaffold on endosomes; β-arrestin also performs this function in the context of GPCR signaling. Originally characterized as a mediator of GPCR desensitization (119), β-arrestin is now recognized as a multifunctional protein that regulates internalization of GPCRs into CCVs (120, 121) and serves as a scaffold for both the Erk (122) and Jnk (123) MAPK modules. Several lines of evidence implicate β-arrestin functionally in the transmission of signals from Raf-1 to MEK and Erk on endosomes. Raf-1 overexpression increased MEK and Erk binding to β-arrestin (122), and a dominantnegative form of β-arrestin blocked Erk activation downstream of a GPCR (56, 99). Thus, endosomes serve as an organelle that supports both PTKR and GPCR signaling to Erk, and each system uses its own scaffold. Yet another MEK/Erk scaffold with unique subcellular targeting properties is Sef, a protein originally identiﬁed as a negative regulator of ﬁbroblast growth factor signaling (124, 125). Sef was subsequently shown to be a MEK/Erk scaffold that resides on the Golgi apparatus (126). Sef binds only activated MEK. In other systems, Erk dissociates from MEK once it is phosphorylated, and phospho-Erk can then enter the nucleus. However, activated Erk remains associated with MEK on Sef, preventing Erk’s entry into the nucleus and sequestering it from its nu-

clear substrates such as Elk-1. However, the active Erk that remains associated with Sef on the Golgi is capable of phosphorylating cytosolic substrates such as RSK2 (126). When Elk-1 was artiﬁcially targeted to the cytoplasm with a nuclear export signal, it became a substrate for Sef-associated phospho-Erk (126). Thus, Sef is a compartment-speciﬁc scaffold that steers Erk activity toward one set of substrates over another (127). Another example of shunting toward one set of Erk substrates over another is in GPCR signaling to Erk via β-arrestin, which acts to retain Erk in the cytosol such that cell proliferation is not stimulated (99). Similarly, β-arrestin 2 acts to sequester Jnk3 in the cytosol (123). These examples illustrate another central feature of compartmentalized signaling: The location of the signaling complex can determine which set of downstream effectors is most efﬁciently activated.

Ras SIGNALING FROM OTHER ENDOMEMBRANES Signaling on Golgi Because endosomes derive from the plasma membrane and take with them surface receptors and their ligands, the idea that signaling initiated by extracellular molecules continues on the cytoplasmic surface of these organelles takes no great leap of imagination. This is not the case for other organelles such as the Golgi apparatus, ER, and mitochondria, which are topologically removed from the plasma membrane. Accordingly, until the last few years the membrane platforms for receptor-mediated signaling were thought to be restricted to the plasma membrane and endosomes. The surprising observation that at steady-state a signiﬁcant pool of N-Ras and H-Ras resides on the Golgi apparatus (20, 128) raised the possibility that Ras/MAPK signaling might also take place on this organelle. This hypothesis was tested by studying Ras activation in living cells using genetically encoded ﬂuorescent probes. The

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

783

ARI

15 February 2006

3:17

probe that has been the most informative is relatively simple, consisting of the RBD of Raf-1 fused to GFP (63). This probe reveals Ras activation in a spatiotemporal fashion by recruitment from the cytosol and nucleoplasm to membrane-bound structures. Using this probe Chiu et al. (63) showed that upon stimulation with growth factors Ras became activated at both the plasma membrane and the Golgi. Interestingly, the kinetics of activation were different on the two compartments; activation at the plasma membrane was rapid and transient (1–10 min), whereas activation at the Golgi was delayed (>20 min) and sustained. As discussed above for endosome signaling, this observation suggested that subcellular location of signaling underlies the kinetically distinct peaks of Ras/MAPK signaling that had been observed in the past by biochemical methods (95). Because retrograde vesicular trafﬁc from the endosomal compartment to the Golgi has been well documented (129), this pathway presented one possibility of how a signal can be delivered from a surface receptor to Ras on the Golgi. However, the signal transmission was too rapid to be explained by vesicular trafﬁcking. Moreover, neither molecular inhibitors of endocytosis nor low temperature affected Ras activation on the Golgi (63). Instead calcium, a diffusible second messenger, and the Ras exchange factor RasGRP1 proved to be responsible for Ras activation on the Golgi (130). RasGRP1 is a member of a family of Ras/Rap1 exchange factors that are activated by diacylglycerol (DAG) and calcium in a fashion analogous to PKCs and other C1 domain–containing proteins (131). Among the four RasGRP proteins, RasGRP1 has the highest degree of speciﬁcity for Ras over Rap1 (131) and also proved to have intrinsic afﬁnity for Golgi membranes in activated cells (130, 132–134). Thus, one pathway for Ras activation on the Golgi downstream of PTKRs is via activation of PLCγ, resulting in production of DAG and an increase in intracellular calcium, followed by activation and translocation to the Golgi of RasGRP1,

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

784

Mor

·

Philips

where it acts on Ras (130). This pathway also appears to be dependent on Src, probably at the level of activation of PLCγ (130). Golgi-associated Ras was initially conceived of as a pool that had not trafﬁcked to the plasma membrane (20). The recent discovery of a plasma membrane/Golgi cycle that functions as a consequence of a palmitoylation/depalmitoylation cycle (35, 36) presents an alternative mechanism for the accrual of active Ras on the Golgi. In this pathway, prenylated but depalmitoylated N-Ras and H-Ras appear to trafﬁc between membrane compartments though the cytosol as ﬂuid phase protein, perhaps bound to a chaperone. In this model, Ras could be activated on one compartment, for example the plasma membrane, and then trafﬁc in a GTP-bound state to another compartment, for example the Golgi. Rocks et al. (36) interrupted the palmitoylation/depalmitoylation cycle by applying the protein palmitoyltransferase inhibitor 2-bromopalmitate and observed a decrease in Ras activation on the Golgi, suggesting that the acylation/deacylation cycle is required for Ras activation on this organelle. Such a model for Ras activation on internal membranes requires an activation-dependent cytosolic pool of GTP-bound Ras that has not as yet been demonstrated. It should be emphasized that a calcium/RasGRP1-dependent and a Ras depalmitoylation–dependent pathway for Ras activation on the Golgi are not mutually exclusive, and both pathways may operate. As discussed above, very recent work has revealed that the palmitoylated forms of Ras are not the only isoforms that can trafﬁc in a retrograde fashion from the plasma membrane to the Golgi. GFP-tagged K-Ras was seen in rat hippocampal neurons stimulated with glutamate to dissociate from the plasma membrane and associate with intracellular membranes, including Golgi, in a calcium/calmodulin-dependent fashion (37). The physiologic signiﬁcance of this translocation and whether this is a neuronal-speciﬁc process have not been determined.

ANRV270-IY24-25

ARI

15 February 2006

3:17

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Signaling on the ER In addition to evidence of signaling on the Golgi, evidence has been presented for Ras signaling on the ER. When a palmitoylationdeﬁcient form of N-Ras or H-Ras is expressed in cells, it accumulates on the ER and in the cytosol (20). These constructs can be viewed as surrogates either for the biosynthetic intermediates in the trafﬁcking pathway of nascent Ras or for mature Ras that has reached the plasma membrane and then become depalmitoylated. When this construct was expressed along with the GFP-RBD spatiotemporal reporter of Ras activation, GTP-bound Ras accumulated on the ER following growth factor stimulation in less than one minute (63). One interpretation for this result is that GEFs such as RasGRP1 can rapidly activate ER-associated Ras in situ. Evidence to support this idea has been presented by Arozarena et al. (134a) who showed that Ras guanine nucleotide-releasing factors (RasGRFs), calcium-regulated Ras GEFs that are highly expressed in the neurons, translocate to the ER and activate Ras on this compartment. Given the recent evidence for the highly dynamic nature of the membrane binding of Ras proteins that are modiﬁed only with a prenyl group (35, 36), it seems plausible that under these conditions Ras becomes active on another compartment such as the plasma membrane. Whereas the steady-state localization of palmitoylated Ras proteins clearly includes the Golgi, GFP-Ras is only observed on the ER if palmitoylation is blocked (20). This observation calls into question the physiologic signiﬁcance of signaling from the ER. Perhaps the strongest evidence for Ras signaling from the ER comes from the discovery of a Ras effector that is restricted to that compartment. Sobering et al. (135) described a protein in Saccharomyces cerevisiae designated ER-associated Ras inhibitor (Eri1) that behaved genetically like an inhibitor of Ras, bound preferentially to GTPbound Ras via its effector domain, and was localized to the ER. Further analysis re-

vealed Eri1 was an integral component of GPI (glycosylphosphatidylinositol)-GlcNAc transferase, an ER-restricted enzyme that catalyzes the transfer of UDP-GlcNAc to an acceptor phosphatidylinositol, the ﬁrst and rate-limiting step in the biosynthesis of GPI anchors. Interestingly, GTP-bound Ras acted to inhibit the enzyme, making GPI-GlcNAc transferase the ﬁrst Ras effector that is negatively regulated by active Ras (136). The physiologic signiﬁcance of the regulation of GPI anchor synthesis by Ras is not clear, but given the critical roles of GPI-anchored proteins in many systems, including the immune system, and given the fact that somatic mutations in a component of GPI-GlcNAc transferase in hematopoetic stem cells leads to a human disease known as paroxysmal nocturnal hemaglobinuria, it is an area of Ras biology that will attract considerable attention.

Signaling in or on Mitochondria Mitochondria have long been neglected as potential sites of signal transduction. The recognition over the past decade of this organelle as the master gate keeper of programmed cell death has caused cell biologists to be more open-minded regarding signaling and mitochondria. Several signaling molecules have been found in mitochondria. For example, c-Src was found to regulate cytochrome c oxidase in the mitochondria of osteoclasts (137). Recently, EGFR was found to translocate to mitochondria and associate with subunit II of cytochrome c oxidase via pY845, which is a Src phosphorylation site (138). In addition to these tyrosine kinases, a tyrosine phosphatase has been described that is targeted to the inner mitochondrial membrane (139). Components of the Ras/MAPK pathway have been reported on mitochondria. Using subcellular fractionation, Rebollo et al. (140) reported all three Ras isoforms in mitochondrial fractions of an IL-2-dependent murine T cell line and showed that the Ras proteins could be coprecipitated with Bcl-2

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

785

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

3:17

from these fractions. Interestingly, whereas the association of K-Ras with mitochondria was IL-2 dependent, H-Ras association was only seen after IL-2 withdrawal. Surprisingly, mitochondrial association was not dependent on CAAX processing, calling into question the biological relevance of the observation. Moreover, these results are at odds with the extensive literature on Ras localization by microscopy in which mitochondrial localization has not been reported. The discrepancy may relate to the nonspeciﬁc stickiness of Ras proteins for membrane fractions ex vivo. Nevertheless, the idea of Ras/MAPK signaling from mitochondria is intriguing. Wang et al. (141) showed that Raf-1 could be recruited to mitochondria by Bcl-2, and from that compartment it could phosphorylate and inactivate the proapoptotic protein BAD. In these studies, evidence for the association of endogenous Raf-1 with mitochondria was lacking, and, as for Ras, an extensive literature on Raf-1 localization by microscopy does not contain evidence of mitochondrial association. Recently, we have shown that K-Ras phosphorylated by PKC in its polybasic region loses afﬁnity for the plasma membrane and translocates to intracellular membranes, including the outer mitochondrial membrane. Using immunogold electron microscopy, we observed PKCdependent translocation of endogenous Ras to mitochondria in Jurkat T cells. Interestingly, phospho-K-Ras promoted apoptosis in a Bcl-XL-dependent fashion (141a).

gle regulatory molecule such as Ras can control such a plethora of cellular responses. Evidence in support of the compartmentalized signaling model has come from experiments in which transmembrane tethers were used to artiﬁcially and stringently target Ras proteins to various membrane compartments (63). When oncogenic Ras was targeted to the ER or Golgi with a transmembrane tether, it retained full transforming activity, indicating that all the signaling events required for the complex cellular phenotype of transformation can be set into motion from internal membranes (63, 133). This might suggest that Ras signaling from internal membranes is no different than from the plasma membrane. However, quantitative differences in signal output could be detected. Whereas Golgi-associated Ras activated Erk and PI3K with a potency equal to that of natively targeted Ras, the Jnk pathway was poorly activated. Conversely, ER-tethered Ras was a potent activator of Jnk but a relatively poor activator of Erk and PI3K (63). The most compelling evidence for compartmentalized Ras signaling has come from the study of ﬁssion yeast. In Schizosaccharomyces pombe, Ras1 controls both the mating pathway via a MAPK cascade and elongated cellular morphology through an exchange factor for Cdc42 (142). ER-restricted Ras1 can support morphology but not mating and the converse is true for Ras1 restricted to the plasma membrane (B. Onken, M.R. Philips & E.C. Chang, manuscript submitted).

Role of Compartment-Specific Signals

PLASMA MEMBRANE MICRODOMAINS AND Ras SIGNALING

What is the physiologic signiﬁcance of Ras signaling from intracellular membranes? As discussed above in the context of endosomes, compartmentalized signaling, in theory, can increase the complexity of signaling by adding kinetically distinct outputs down a single pathway and/or by allowing for activation of distinct downstream pathways. This increase in complexity may help explain how a sin-

786

Mor

·

Philips

As discussed above, the concept of compartmentalized signaling relates to different organelles. However, compartmentalized signaling can also apply to subcompartments within a given organelle. The best-studied example of this relates to the plasma membrane, where microdomains and their relationship to signaling events have been

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

extensively studied for a decade. The plasma membrane microdomain that has received by far the most attention is the lipid raft (143). Lipid rafts are microdomains rich in cholesterol and sphingolipids that are thought to maintain a liquid-ordered state and thereby partition into patches within the disordered glycerophospholipids of the bulk membrane. In cells that express caveolins, one type of lipid raft forms structures known as caveolae that are readily observed by electron microscopy (144). Glycophosphatidylinositol-anchored proteins partition into the outer leaﬂet of lipid rafts, whereas some lipidated cytosolic proteins such as Src family kinases partition into the cytosolic leaﬂet. The observation that many of the proteins that partition into lipid rafts are involved in signaling has led to the hypothesis that lipid rafts are hot spots for signal transduction (145). Because lipid rafts have been observed in living cells only with indirect methods, such as resistance to detergent extraction, their biological signiﬁcance has yet to be proven (146). The association of components of the Ras/MAPK pathway with lipid rafts is somewhat controversial because the list depends on the methods used to isolate the microdomains. Among the Ras/MAPK signaling components that have been associated with lipid rafts are growth factor receptors, including EGFR, PDGFR, and the insulin receptor. Interestingly, EGFR partitions in and out of caveolae in a ligand-dependent fashion (147). The biophysical basis for the partitioning into lipid rafts of a transmembrane protein lacking lipid modiﬁcation such as PTKRs is not understood. In contrast, the partitioning of acylated proteins such as Src family kinases into lipid rafts can be understood as a function of the ability of acyl chains to insert into phospholipid bilayers (148). Unlike acylated proteins, prenylated proteins are generally excluded from detergent-resistant membrane fractions (149). Nevertheless, Ras proteins have been reported in caveolin-enriched fractions made without detergent (150). For N-Ras and

H-Ras that are both acylated and prenylated, the acyl chain is apparently dominant and can cause these Ras isoforms to partition into lipid rafts. Ras also interacts directly with caveolin1 (151), providing an additional mechanism for recruitment to these specialized membrane microdomains. With regard to the role of lipid microdomains in Ras signaling, the most compelling analysis has come from the work of Hancock and Parton and their colleagues. Roy et al. (69) showed that a dominant-negative form of caveolin inhibited H-Ras but not KRas signaling and that this differential effect could be mimicked by cholesterol depletion. These data represent the ﬁrst demonstration that Ras isoforms operate in functionally distinct plasma membrane microdomains. Using multiple methods, including detergent-free subcellular fractionation, immunogold electron microscopy, and immunoﬂuorescence, Prior et al. (152) showed that H-Ras was associated with lipid rafts, whereas K-Ras was not. Moreover, access of H-Ras to lipid rafts was dynamic; GDP-bound H-Ras was favored, suggesting that activation takes place in rafts, but effector engagement occurs in nonraft domains (152). These investigators went on to develop an elegant method for direct visualization of membrane protein clustering using immunogold electron microscopy staining of exposed membrane sheets, followed by statistical point pattern analysis (153). These studies conﬁrmed that whereas GDPbound H-Ras clustered in domains with a mean diameter of 44 nm that were sensitive to cholesterol depletion, GTP-bound H-Ras and all forms of K-Ras clustered in domains that were insensitive to cholesterol depletion (153). Combining this new method with FRAP (ﬂuorescence recovery after photobleaching) studies and more extensive mutational analysis revealed that the partitioning of H-Ras into lipid rafts is governed by multiple factors that include the acyl chains, the adjacent hypervariable region of the C-terminus, and the GTP-binding domain (154).

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

787

ARI

15 February 2006

3:17

Microdomain studies have recently been complemented with single-particle tracking studies of Ras. Murakoshi et al. (155) have tracked YFP-tagged Ras at the level of single molecules and observed random motion in the plane of the membrane interspersed with periods of immobility that last <1 s. Using FRET between YFP-Ras and Bodipy TR-GTP, these investigators have shown that the periods of immobility correspond to GTP loading. Growth factor stimulation increased the proportion of Ras molecules that are immobilized. Interestingly, these investigators did not observe isoform differences in these behaviors (155). These observations have been interpreted to suggest that nanoclusters of activated Ras and other components of the Ras signaling pathway exist. The mechanism of the nanoclusters’ formation may relate to a corral-like effect imposed on active Ras by cytoskeletal elements, combined with transmembrane proteins that act as pickets (156) or membrane microdomains containing disordered lipids (157). Either way, the plasma membrane clearly can no longer be considered a homogeneous platform for cellular signaling. Similar to organelle-speciﬁc signaling, the heterogeneity of plasma membrane microdomains may contribute to the complexity of signaling outputs.

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

protein using the cholera toxin B subunit (CTB). When rafts were capped, so too were TCR, Lck, and LAT (linker-activated T cell). Conversely, activating the TCR by capping with cross-linking antibodies also patched CT-B, Lck, and LAT (161). The earliest events in TCR signaling involve the Src family kinases Lyn and Fyn that are doubly acylated, explaining their afﬁnity for lipid rafts. Interestingly, LAT, a transmembrane protein that serves as a scaffold for the phosphotyrosine docking sites of several signaling molecules (162), is palmitoylated, and this modiﬁcation likely explains its constitutive afﬁnity for lipid rafts (163). Co-stimulation from CD28 has also been explained in the context of lipid rafts (164). Some accessory molecules in TCR activation, such as CD48, are GPI-linked, and their partition into lipid rafts has been implicated in augmenting TCR signaling (165). Several lines of evidence have implicated lipid rafts in the formation of the immunological synapse (158, 159). However a very recent study that used single-molecule tracking showed that rafts were not required for the clustering of CD2, Lck, and LAT, calling into question the role of lipid microdomains in immunological-synapse formation and TCR signaling (166).

Compartmentalized Ras Signaling in T Cells COMPARTMENTALIZED SIGNALING IN LYMPHOCYTES T Cell Signaling and Lipid Rafts Much of the evidence for lipid rafts serving as signaling platforms comes from the study of T lymphocytes (158, 159). TCR and associated signaling molecules were enriched in detergent-resistant microdomains following TCR activation, and TCR signaling could be inhibiting by disrupting lipid rafts (160). Magee and colleagues (161) developed a method for visualizing, at the resolution of the light microscope, lipid rafts on the surface of lymphocytes by capping a GPI-anchored

788

Mor

·

Philips

Although all three Ras isoforms are expressed in T lymphocytes, the levels of H-Ras are much lower than those of N-Ras and K-Ras. There are several reasons to conclude that NRas has particular signiﬁcance in T cell signaling. First, in contrast to carcinomas where kras mutations predominate, when lymphoid malignancies are associated with oncogenic Ras mutations (30% of cases) they are almost always in the nras gene (10). Second, mice deﬁcient in N-Ras have a defect in T cell function and are extremely sensitive to infection with inﬂuenza virus (167). Finally, although K-Ras can be activated by robust TCR signaling, under conditions of low-level

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

cross-linking of TCR with anti-CD3 antibodies, N-Ras is preferentially activated (132). Surprisingly, when TCRs were capped on Jurkat T cells with anti-CD3 antibodies such that lipid raft clusters were formed as indicated by cocapping of CT-B, GFP-H-Ras was recruited to the clusters but GFP-NRas was not (132). Less surprising was the absence of GFP-K-Ras from the clusters, because this isoform lacks an acyl chain. Interestingly, constitutively active GFP-H-Ras12V was excluded from the clusters (132), consistent with the ﬁndings of Prior et al. (152) indicating that activated H-Ras moves out of lipid rafts. Thus, if the hypothesis that N-Ras signaling is critical to T cell function is correct, this Ras isoform does not appear to colocalize with the receptor under conditions of T cell signaling. Another surprise came when GFP-RBD, the in vivo probe for GTP-bound Ras described above, was applied to Jurkat T cells. Concordant with the earlier studies of ﬁbroblasts, N-Ras and H-Ras were activated on the Golgi apparatus following TCR activation. However, this was the only membrane compartment upon which GTP-Ras could be detected; the plasma membrane was entirely devoid of activity. Unlike ﬁbroblasts, where activation of Ras on the Golgi was delayed, activated Ras could be detected on the Golgi within minutes of stimulation. RasGRP1 is much more highly expressed in lymphocytes than in ﬁbroblasts (168), and the difference in kinetics of Ras activation on the Golgi may simply be a reﬂection of this difference. Recently, these observations in Jurkat T cells were extended to primary murine T cells where GFP-RBD reports Ras activation only on the Golgi apparatus following TCR activation by cross-linking CD3 and CD28 (Figure 4) (A. Mor & M.R. Philips, unpublished observation). In addition, the GFPRBD probe revealed constitutive Ras activation in Raji B cells that occurred only on the Golgi suggesting that, as in T cells, Ras signaling in B cells occurs on the Golgi (A. Mor & M.R. Philips, unpublished observation).

Endogenous N-Ras was localized to the Golgi of Jurkat T cells (132), and YFPRasGRP1 translocated to this organelle in response to TCR signaling or stimulation with phorbol myristate acetate and ionomycin (130, 132). Unlike RasGRP2 and RasGRP4, both RasGRP1 and RasGRP3 have afﬁnity for the Golgi in Jurkat T cells. However, overexpression only of RasGRP1 leads to Ras activation on the Golgi (132). A dominant interfering mutant of RasGRP1 inhibited activation on this compartment (130). Silencing of the RasGRP1 gene with siRNA abolished all Ras signaling (i.e., recruitment of GFP-RBD to the Golgi), demonstrating that this exchange factor is responsible for activating Ras on the Golgi of T cells (130). This result is concordant with studies of RasGRP1deﬁcient mice in which Ras signaling, as determined biochemically, was eliminated (16). Also consistent with RasGRP1 being responsible for the activation of N-Ras on the Golgi of Jurkat T cells was the absence of any activation observed in T cells deﬁcient in PLCγ1 (130, 132). Low concentrations of anti-CD3 antibodies induced N-Ras activation on the Golgi but did not activate K-Ras. However, high concentrations of anti-CD3 antibody activated K-Ras, and the GFP-RBD probe indicated that this activation occurred on the plasma membrane (132). Thus, K-Ras can be activated on the plasma membrane, presumably via signaling through Grb2/SOS, provided that the stimulus is strong (132). N-Ras is expressed on both the plasma membrane and Golgi of Jurkat T cells. What accounts for activation on only one of these compartments when exchange factors are present on both? One explanation is that there may be a Ras GAP that inactivates Ras specifically at the plasma membrane. CAPRI is a Ras GAP that operates at the plasma membrane and is an attractive candidate for a component of Ras signaling in lymphocytes because it is calcium activated (169). Knockdown of CAPRI in Jurkat T cells with siRNA resulted in H-Ras activation on the plasma membrane following TCR stimulation (130).

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

CAPRI: calcium-promoted Ras inactivator

789

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

3:17

Figure 4 Ras activation only on the Golgi of primary T cells. CD4+ splenocytes from Balb/C mice were transfected with CFP-N-Ras and YFP-RBD. N-Ras is expressed in resting cells on both the plasma membrane and Golgi. Following cross-linking of the TCR with anti-CD3/CD28 antibodies, YFP-RBD rapidly (<5 min) accumulated on the Golgi apparatus but not on the plasma membrane.

Thus, CAPRI appears to keep Ras signaling on the plasma membrane of lymphocytes in check while signaling proceeds on the Golgi. Interestingly, because both CAPRI and RasGRP1 are activated by calcium, this single second messenger controls Ras signaling in opposite directions simultaneously on different subcellular compartments (Figure 5), a stark example of compartmentalized signaling. The physiological reason why Ras signaling in T cells occurs primarily on the Golgi remains to be elucidated.

CONCLUSION The evidence for Ras/MAPK signaling on intracellular membranes is strong. The association of PTKRs and downstream elements of the Ras/MAPK pathway with endosomes was ﬁrst demonstrated by subcellular fractionation and conformed to the paradigm of desensitization through endocytosis. Functional studies using molecular inhibitors of endocy-

790

Mor

·

Philips

tosis revealed that endocytosis had to be more than a mechanism to limit receptor signaling because in many cellular contexts it was required for signal propagation. GFP-tagged probes conﬁrmed that active Ras/MAPK signaling occurs on the cytosolic surface of endosomes. These probes also revealed signaling on other endomembranes, including the Golgi apparatus. The characterization of various MAPK scaffolds that are restricted to different membrane compartments, including the plasma membrane, endosomes, and Golgi, further explained the mechanisms of compartmentalized signaling. With regard to Ras activation on internal membranes, a spatiotemporal ﬂuorescent probe has revealed that the pathway of activation on the Golgi involves PLCγ and RasGRP1 and thereby differs from the Grb2/SOS-mediated pathway for Ras activation on the plasma membrane. Thus, compartmentalized signaling can be accomplished by using distinct upstream pathways. The

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

Figure 5 PLCγ and RasGRP1 mediate Ras activation on the Golgi of T lymphocytes, and CAPRI limits activation at the plasma membrane. Activation of the TCR results in tyrosine phosphorylation of the ζ chain of the receptor by Src family kinases and the resulting phosphotyrosines serve to recruit ZAP-70, which in turn phosphorylates the scaffold protein LAT at multiple sites. Among the signaling molecules recruited to phosphorylated LAT is PLCγ, which acts on phosphatidylinositol-4,5-bisphosphate in the plasma membrane to produce diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3 ). Calcium liberated from internal stores by IP3 acts on the calcium- and DAG-sensitive Ras exchange factor RasGRP1 and causes it to translocate to the Golgi, where DAG levels are relatively high. RasGRP1 activates Golgi-associated Ras on this compartment. Meanwhile, calcium also activates the Ras GAP CAPRI that translocates to the plasma membrane and downregulates any Ras that is activated on this compartment by the exchange factor SOS.

PLCγ/RasGRP1 pathway appears to be dominant in T lymphocytes in which all Ras signaling downstream of the TCR was observed on the Golgi apparatus. More difﬁcult than establishing Ras/ MAPK signaling on endomembranes will be understanding the physiologic signiﬁcance of compartment-speciﬁc signaling. The most obvious purpose of compartmentalized signaling is to increase the complexity of signal output by extending and seg-

regating the signaling repertoire of critical regulatory molecules such as Ras. However, evidence for differential signal outputs from different compartments is sparse and more or less conﬁned to overexpression studies of signaling molecules that are rerouted to ectopic compartments. Elucidation of the spatial complexity of signaling networks in a physiologic context represents one of the next frontiers in signal transduction research.

www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

791

ANRV270-IY24-25

ARI

15 February 2006

3:17

ACKNOWLEDGMENTS This work was supported by grants from the National Institutes of Health, the Burroughs Wellcome Fund, the Arthritis Foundation, New York Chapter, and the Arthritis National Research Foundation.

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

LITERATURE CITED

12. This study performed in lymphocytes was the first to show activation of protooncogenic Ras in any cell type downstream of receptor stimulation.

792

1. Barbacid M. 1987. ras genes. Annu. Rev. Biochem. 56:779–827 2. Vetter IR, Wittinghofer A. 2001. The guanine nucleotide-binding switch in three dimensions. Science 294:1299–304 3. Chung HH, Benson DR, Cornish VW, Schultz PG. 1993. Probing the role of loop 2 in Ras function with unnatural amino acids. Proc. Natl. Acad. Sci. USA 90:10145– 49 4. Feig LA. 1999. Tools of the trade: use of dominant-inhibitory mutants of Ras-family GTPases. Nat. Cell Biol. 1:E25–27 5. Harvey JJ. 1964. An unidentiﬁed virus which causes the rapid production of tumours in mice. Nature 204:1104–5 6. Ruta M, Wolford R, Dhar R, DeFeo-Jones D, Ellis RW, Scolnick EM. 1986. Nucleotide sequence of the two rat cellular rasH genes. Mol. Cell. Biol. 6:1706–10 7. DeFeo D, Gonda MA, Young HA, Chang EH, Lowy DR, et al. 1981. Analysis of two divergent rat genomic clones homologous to the transforming gene of Harvey murine sarcoma virus. Proc. Natl. Acad. Sci. USA 78:3328–32 8. Ellis RW, DeFeo D, Shih TY, Gonda MA, Young HA, et al. 1981. The p21 src genes of Harvey and Kirsten sarcoma viruses originate from divergent members of a family of normal vertebrate genes. Nature 292:506–11 9. Lowy DR, Willumsen BM. 1993. Function and regulation of RAS. Annu. Rev. Biochem. 62:851–91 10. Bos JL. 1989. Ras oncogenes in human cancer: a review. Cancer Res. 49:4682–89 11. Genot E, Cantrell DA. 2000. Ras regulation and function in lymphocytes. Curr. Opin. Immunol. 12:289–94 12. Downward J, Graves JD, Warne PH, Rayter S, Cantrell DA. 1990. Stimulation of p21ras upon T-cell activation. Nature 346:719–23 13. Cantrell DA. 2002. Transgenic analysis of thymocyte signal transduction. Nat. Rev. Immunol. 2:20–27 14. Swan KA, Alberola-Ila J, Gross JA, Appleby MW, Forbush KA, et al. 1995. Involvement of p21ras distinguishes positive and negative selection in thymocytes. EMBO J. 14:276– 85 15. Gartner F, Alt FW, Monroe R, Chu M, Sleckman BP, et al. 1999. Immature thymocytes employ distinct signaling pathways for allelic exclusion versus differentiation and expansion. Immunity 10:537–46 16. Dower NA, Stang SL, Bottorff DA, Ebinu JO, Dickie P, et al. 2000. RasGRP is essential for mouse thymocyte differentiation and TCR signaling. Nat. Immunol. 1:317–21 17. Layer K, Lin G, Nencioni A, Hu W, Schmucker A, et al. 2003. Autoimmunity as the consequence of a spontaneous mutation in Rasgrp1. Immunity 19:243–55 18. Roose JP, Mollenauer M, Gupta VA, Stone J, Weiss A. 2005. A diacylglycerol-protein kinase C-RasGRP1 pathway directs Ras activation upon antigen receptor stimulation of T cells. Mol. Cell. Biol. 25:4426–41 Mor

·

Philips

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

19. Willingham MC, Pastan I, Shih TY, Scolnick EM. 1980. Localization of the src gene product of the Harvey strain of MSV to plasma membrane of transformed cells by electron microscopic immunocytochemistry. Cell 19:1005–14 20. Choy E, Chiu VK, Silletti J, Feoktistov M, Morimoto T, et al. 1999. Endomembrane trafficking of ras: the CAAX motif targets proteins to the ER and Golgi. Cell 98:69–80 21. Willumsen BM, Norris K, Papageorge AG, Hubbert NL, Lowy DR. 1984. Harvey murine sarcoma virus p21 ras protein: biological and biochemical signiﬁcance of the cysteine nearest the carboxy terminus. EMBO J. 3:2581–85 22. Clarke S. 1992. Protein isoprenylation and methylation at carboxyl terminal cysteine residues. Annu. Rev. Biochem. 61:355–86 23. Clarke S, Vogel JP, Deschenes RJ, Stock JB. 1988. Post-translational modiﬁcation of the Ha-ras oncogene protein: evidience for a third class of protein carboxyl methyltransferases. Proc. Natl. Acad. Sci. USA 85:4643–47 24. Casey PJ, Seabra MC. 1996. Protein prenyltransferases. J. Biol. Chem. 271:5289–92 25. Zhang FL, Casey PJ. 1996. Protein prenylation: molecular mechanisms and functional consequences. Annu. Rev. Biochem. 65:241–69 26. Kim E, Ambroziak P, Otto JC, Taylor B, Ashby M, et al. 1999. Disruption of the mouse Rce1 gene results in defective Ras processing and mislocalization of Ras within cells. J. Biol. Chem. 274:8383–90 ˜ 27. P`erez-Sala D, Tan EW, Canada FJ, Rando RR. 1991. Methylation and demethylation reactions of guanine nucleotide-binding proteins of rod outer segments. Proc. Natl. Acad. Sci. USA 88:3043–46 28. Lobo S, Greentree WK, Linder ME, Deschenes RJ. 2002. Identiﬁcation of a Ras palmitoyltransferase in Saccharomyces cerevisiae. J. Biol. Chem. 277:41268–73 29. Dai Q, Choy E, Chiu V, Romano J, Slivka S, et al. 1998. Mammalian prenylcysteine carboxyl methyltransferase is in the endoplasmic reticulum. J. Biol. Chem. 273:15030–34 30. Boyartchuk VL, Ashby MN, Rine J. 1997. Modulation of Ras and a-factor function by carboxyl-terminal proteolysis. Science 275:1796–800 31. Zhao L, Lobo S, Dong X, Ault AD, Deschenes RJ. 2002. Erf4p and Erf2p form an endoplasmic reticulum-associated complex involved in the plasma membrane localization of yeast Ras proteins. J. Biol. Chem. 277:49352–59 32. Hancock JF, Paterson H, Marshall CJ. 1990. A polybasic domain or palmitoylation is required in addition to the CAAX motif to localize p21ras to the plasma membrane. Cell 63:133–39 33. Hancock JF, Cadwallader K, Paterson H, Marshall CJ. 1991. A CAAX or a CAAL motif and a second signal are sufﬁcient for plasma membrane targeting of ras proteins. EMBO J. 10:4033–39 34. Dong X, Mitchell DA, Lobo S, Zhao L, Bartels DJ, Deschenes RJ. 2003. Palmitoylation and plasma membrane localization of Ras2p by a nonclassical trafﬁcking pathway in Saccharomyces cerevisiae. Mol. Cell. Biol. 23:6574–84 35. Goodwin JS, Drake KR, Rogers C, Wright L, Lippincott-Schwartz J, et al. 2005. Depalmitoylated Ras traffics to and from the Golgi complex via a nonvesicular pathway. J. Cell Biol. 170:261–72 36. Rocks O, Peyker A, Kahms M, Verveer PJ, Koerner C, et al. 2005. An acylation cycle regulates localization and activity of palmitoylated Ras isoforms. Science 307:1746–52 37. Fivaz M, Meyer T. 2005. Reversible intracellular translocation of KRas but not HRas in hippocampal neurons regulated by Ca2+ /calmodulin. J. Cell Biol. 170:429–41 www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

20. This study established the endomembrane as the compartment upon which post-prenylation CAAX processing takes place.

32. This paper established the importance of sequences upstream of the CAAX motif in membranetargeting of Ras proteins. References 35 and 36 defined a retrograde pathway of H-Ras and N-Ras trafficking from the plasma membrane to the Golgi that is initiated by depalmitoylation.

793

ARI

15 February 2006

3:17

38. Schlessinger J. 2000. Cell signaling by receptor tyrosine kinases. Cell 103:211–25 39. Morrison DK, Cutler RE. 1997. The complexity of Raf-1 regulation. Curr. Opin. Cell Biol. 9:174–79 40. Repasky GA, Chenette EJ, Der CJ. 2004. Renewing the conspiracy theory debate: Does Raf function alone to mediate Ras oncogenesis? Trends Cell Biol. 14:639–47 41. Hamad NM, Elconin JH, Karnoub AE, Bai W, Rich JN, et al. 2002. Distinct requirements for Ras oncogenesis in human versus mouse cells. Genes Dev. 16:2045–57 42. Di Guglielmo GM, Baass PC, Ou WJ, Posner BI, Bergeron JJ. 1994. Compartmentalization of SHC, GRB2 and mSOS, and hyperphosphorylation of Raf-1 by EGF but not insulin in liver parenchyma. EMBO J. 13:4269–77 43. Baass PC, Di Guglielmo GM, Authier F, Posner BI, Bergeron JJ. 1995. Compartmentalized signal transduction by receptor tyrosine kinases. Trends Cell Biol. 5:465–70 44. Carpenter G. 2000. The EGF receptor: a nexus for trafﬁcking and signaling. BioEssays 22:697–707 45. Duan L, Reddi AL, Ghosh A, Dimri M, Band H. 2004. The Cbl family and other ubiquitin ligases: destructive forces in control of antigen receptor signaling. Immunity 21:7–17 46. Stang E, Blystad FD, Kazazic M, Bertelsen V, Brodahl T, et al. 2004. Cbl-dependent ubiquitination is required for progression of EGF receptors into clathrin-coated pits. Mol. Biol. Cell 15:3591–604 47. Levkowitz G, Waterman H, Zamir E, Kam Z, Oved S, et al. 1998. c-Cbl/Sli-1 regulates endocytic sorting and ubiquitination of the epidermal growth factor receptor. Genes Dev. 12:3663–74 48. Buday L, Khwaja A, Sipeki S, Farago A, Downward J. 1996. Interactions of Cbl with two adapter proteins, Grb2 and Crk, upon T cell activation. J. Biol. Chem. 271:6159–63 49. Miller WE, Lefkowitz RJ. 2001. Expanding roles for β-arrestins as scaffolds and adapters in GPCR signaling and trafﬁcking. Curr. Opin. Cell Biol. 13:139–45 50. Aramori I, Ferguson SS, Bieniasz PD, Zhang J, Cullen B, Cullen MG. 1997. Molecular mechanism of desensitization of the chemokine receptor CCR-5: receptor signaling and internalization are dissociable from its role as an HIV-1 co-receptor. EMBO J. 16:4606–16 51. Fomina AF, Deerinck TJ, Ellisman MH, Cahalan MD. 2003. Regulation of membrane trafﬁcking and subcellular organization of endocytic compartments revealed with FM143 in resting and activated human T cells. Exp. Cell Res. 291:150–66 52. Naramura M, Kole HK, Hu RJ, Gu H. 1998. Altered thymic positive selection and intracellular signals in Cbl-deﬁcient mice. Proc. Natl. Acad. Sci. USA 95:15547–52 53. Wiedemann A, Muller S, Favier B, Penna D, Guiraud M, et al. 2005. T-cell activation is accompanied by an ubiquitination process occurring at the immunological synapse. Immunol. Lett. 98:57–61 54. Vieira AV, Lamaze C, Schmid SL. 1996. Control of EGF receptor signaling by clathrinmediated endocytosis. Science 274:2086–89 55. Kranenburg O, Verlaan I, Moolenaar WH. 1999. Dynamin is required for the activation of mitogen-activated protein (MAP) kinase by MAP kinase kinase. J. Biol. Chem. 274:35301–4 56. Daaka Y, Luttrell LM, Ahn S, Della Rocca GJ, Ferguson SS. 1998. Essential role for G protein-coupled receptor endocytosis in the activation of mitogen-activated protein kinase. J. Biol. Chem. 273:685–88 57. Stahl PD, Barbieri MA. 2002. Multivesicular bodies and multivesicular endosomes: the “ins and outs” of endosomal trafﬁc. Science STKE 2002:PE32

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

794

Mor

·

Philips

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

58. Burke P, Schooler K, Wiley HS. 2001. Regulation of epidermal growth factor receptor signaling by endocytosis and intracellular trafﬁcking. Mol. Biol. Cell 12:1897–910 59. Sorkin A, McClure M, Huang F, Carter R. 2000. Interaction of EGF receptor and grb2 in living cells visualized by fluorescence resonance energy transfer (FRET) microscopy. Curr. Biol. 10:1395–98 60. Carter RE, Sorkin A. 1998. Endocytosis of functional epidermal growth factor receptorgreen ﬂuorescent protein chimera. J. Biol. Chem. 273:35000–7 61. Wouters FS, Bastiaens PI. 1999. Fluorescence lifetime imaging of receptor tyrosine kinase activity in cells. Curr. Biol. 9:1127–30 62. Jiang X, Sorkin A. 2002. Coordinated trafﬁc of Grb2 and Ras during epidermal growth factor receptor endocytosis visualized in living cells. Mol. Biol. Cell 13:1522–35 63. Chiu VK, Bivona T, Hach A, Sajous JB, Silletti J, et al. 2002. Ras signalling on the endoplasmic reticulum and the Golgi. Nat. Cell Biol. 4:343–50 64. Sorkin A, Von Zastrow M. 2002. Signal transduction and endocytosis: close encounters of many kinds. Nat. Rev. Mol. Cell Biol. 3:600–14 65. Oksvold MP, Skarpen E, Wierod L, Paulsen RE, Huitfeldt HS. 2001. Re-localization of activated EGF receptor and its signal transducers to multivesicular compartments downstream of early endosomes in response to EGF. Eur. J. Cell Biol. 80:285–94 66. Wang XJ, Liao HJ, Chattopadhyay A, Carpenter G. 2001. EGF-dependent translocation of green ﬂuorescent protein-tagged PLC-gamma1 to the plasma membrane and endosomes. Exp. Cell Res. 267:28–36 67. Matsuda M, Paterson HF, Rodriguez R, Fensome AC, Ellis MV, et al. 2001. Real time ﬂuorescence imaging of PLC gamma translocation and its interaction with the epidermal growth factor receptor. J. Cell Biol. 153:599–612 68. Pol A, Calvo M, Enrich C. 1998. Isolated endosomes from quiescent rat liver contain the signal transduction machinery. Differential distribution of activated Raf-1 and Mek in the endocytic compartment. FEBS Lett. 441:34–38 69. Roy S, Wyse B, Hancock JF. 2002. H-Ras signaling and K-Ras signaling are differentially dependent on endocytosis. Mol. Cell. Biol. 22:5128–40 70. Yamabhai M, Hoffman NG, Hardison NL, McPherson PS, Castagnoli L, et al. 1998. Intersectin, a novel adaptor protein with two Eps15 homology and ﬁve Src homology 3 domains. J. Biol. Chem. 273:31401–7 71. Simpson F, Hussain NK, Qualmann B, Kelly RB, Kay BK, et al. 1999. SH3-domaincontaining proteins function at distinct steps in clathrin-coated vesicle formation. Nat. Cell Biol. 1:119–24 72. Tong XK, Hussain NK, Heuvel E, Kurakin A, Abi-Jaoude E, et al. 2000. The endocytic protein intersectin is a major binding partner for the Ras exchange factor mSos1 in rat brain. EMBO J. 19:1263–71 73. Mohney RP, Das M, Bivona TG, Hanes R, Adams AG, et al. 2003. Intersectin activates Ras but stimulates transcription through an independent pathway involving JNK. J. Biol. Chem. 278:47038–45 74. Sengar AS, Wang W, Bishay J, Cohen S, Egan SE. 1999. The EH and SH3 domain Ese proteins regulate endocytosis by linking to dynamin and Eps15. EMBO J. 18:1159–71 75. Adams A, Thorn JM, Yamabhai M, Kay BK, O’Bryan JP. 2000. Intersectin, an adaptor protein involved in clathrin-mediated endocytosis, activates mitogenic signaling pathways. J. Biol. Chem. 275:27414–20 76. Teis D, Huber LA. 2003. The odd couple: signal transduction and endocytosis. Cell. Mol. Life Sci. 60:2020–33 www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

59. These investigators used FRET measurements in living cells to show EGFR signaling on endosomes.

63. This was the first study to employ GFP-RBD to analyze Ras activation spatiotemporally in live cells.

795

ARI

15 February 2006

3:17

77. Wang Y, Pennock S, Chen X, Wang Z. 2002. Endosomal signaling of epidermal growth factor receptor stimulates signal transduction pathways leading to cell survival. Mol. Cell. Biol. 22:7279–90 78. Bivona TG, Wiener HH, Ahearn IM, Silletti J, Chiu VK, Philips MR. 2004. Rap1 upregulation and activation on plasma membrane regulates T cell adhesion. J. Cell Biol. 164:461–70 79. Bos JL, de Rooij J, Reedquist KA. 2001. Rap1 signalling: adhering to new models. Nat. Rev. Mol. Cell Biol. 2:369–77 80. Bos JL. 1998. All in the family? New insights and questions regarding interconnectivity of Ras, Rap1 and Ral. EMBO J. 17:6776–82 81. Mishra S, Smolik SM, Forte MA, Stork PJ. 2005. Ras-independent activation of ERK signaling via the torso receptor tyrosine kinase is mediated by Rap1. Curr. Biol. 15:366–70 82. Stork PJ, Dillon TJ. 2005. Multiple roles of Rap1 in hematopoietic cells: complementary versus antagonistic functions. Blood 106:2952–61 83. Reedquist KA, Ross E, Koop EA, Wolthuis RM, Zwartkruis FJ, et al. 2000. The small GTPase, Rap1, mediates CD31-induced integrin adhesion. J. Cell Biol. 148:1151–58 84. Katagiri K, Maeda A, Shimonaka M, Kinashi T. 2003. RAPL, a Rap1-binding molecule that mediates Rap1-induced adhesion through spatial regulation of LFA-1. Nat. Immunol. 4:741–48 85. Dustin ML, Bivona TG, Philips MR. 2004. Membranes as messengers in T cell adhesion signaling. Nat. Immunol. 5:363–72 86. Mochizuki N, Yamashita S, Kurokawa K, Ohba Y, Nagai T, et al. 2001. Spatio-temporal images of growth-factor-induced activation of Ras and Rap1. Nature 411:1065–68 87. Medeiros RB, Dickey DM, Chung H, Quale AC, Nagarajan LR, et al. 2005. Protein kinase D1 and the β1 integrin cytoplasmic domain control β1 integrin function via regulation of Rap1 activation. Immunity 23:213–26 88. Wu C, Lai CF, Mobley WC. 2001. Nerve growth factor activates persistent Rap1 signaling in endosomes. J. Neurosci. 21:5406–16 89. Howe CL, Mobley WC. 2004. Signaling endosome hypothesis: a cellular mechanism for long distance communication. J. Neurobiol. 58:207–16 90. Grimes ML, Zhou J, Beattie EC, Yuen EC, Hall DE, et al. 1996. Endocytosis of activated TrkA: evidence that nerve growth factor induces formation of signaling endosomes. J. Neurosci. 16:7950–64 91. Howe CL, Valletta JS, Rusnak AS, Mobley WC. 2001. NGF signaling from clathrincoated vesicles: evidence that signaling endosomes serve as a platform for the Ras-MAPK pathway. Neuron 32:801–14 92. Howe CL, Mobley WC. 2005. Long-distance retrograde neurotrophic signaling. Curr. Opin. Neurobiol. 15:40–48 93. Guermonprez P, Valladeau J, Zitvogel L, Thery C, Amigorena S. 2002. Antigen presentation and T cell stimulation by dendritic cells. Annu. Rev. Immunol. 20:621–67 94. Kaplan DR, Miller FD. 2000. Neurotrophin signal transduction in the nervous system. Curr. Opin. Neurobiol. 10:381–91 95. Marshall CJ. 1995. Speciﬁcity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell 80:179–85 96. Zhang Y, Moheban DB, Conway BR, Bhattacharyya A, Segal RA. 2000. Cell surface Trk receptors mediate NGF-induced survival while internalized receptors regulate NGFinduced differentiation. J. Neurosci. 20:5671–78

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

796

Mor

·

Philips

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

97. York RD, Molliver DC, Grewal SS, Stenberg PE, McCleskey EW, Stork PJ. 2000. Role of phosphoinositide 3-kinase and endocytosis in nerve growth factor-induced extracellular signal-regulated kinase activation via Ras and Rap1. Mol. Cell. Biol. 20:8069–83 98. Kuruvilla R, Zweifel LS, Glebova NO, Lonze BE, Valdez G, et al. 2004. A neurotrophin signaling cascade coordinates sympathetic neuron development through differential control of TrkA trafﬁcking and retrograde signaling. Cell 118:243–55 99. DeFea KA, Zalevsky J, Thoma MS, Dery O, Mullins RD, Bunnett NW. 2000. βarrestin-dependent endocytosis of proteinase-activated receptor 2 is required for intracellular targeting of activated ERK1/2. J. Cell Biol. 148:1267–81 100. Takeshita F, Gursel I, Ishii KJ, Suzuki K, Gursel M, Klinman DM. 2004. Signal transduction pathways mediated by the interaction of CpG DNA with Toll-like receptor 9. Semin. Immunol. 16:17–22 101. Ahmad-Nejad P, Hacker H, Rutz M, Bauer S, Vabulas RM, Wagner H. 2002. Bacterial CpG-DNA and lipopolysaccharides activate Toll-like receptors at distinct cellular compartments. Eur. J. Immunol. 32:1958–68 102. Honda K, Ohba Y, Yanai H, Negishi H, Mizutani T, et al. 2005. Spatiotemporal regulation of MyD88-IRF-7 signalling for robust type-I interferon induction. Nature 434:1035–40 103. Massague J. 1998. TGF-β signal transduction. Annu. Rev. Biochem. 67:753–91 104. Hayes S, Chawla A, Corvera S. 2002. TGF β receptor internalization into EEA1enriched early endosomes: role in signaling to Smad2. J. Cell Biol. 158:1239–49 105. Di Guglielmo GM, Le Roy C, Goodfellow AF, Wrana JL. 2003. Distinct endocytic pathways regulate TGF-β receptor signalling and turnover. Nat. Cell Biol. 5:410–21 106. Parks AL, Klueg KM, Stout JR, Muskavitch MA. 2000. Ligand endocytosis drives receptor dissociation and activation in the Notch pathway. Development 127:1373–85 107. Gupta-Rossi N, Six E, LeBail O, Logeat F, Chastagner P, et al. 2004. Monoubiquitination and endocytosis direct gamma-secretase cleavage of activated Notch receptor. J. Cell Biol. 166:73–83 108. Karandikar M, Cobb MH. 1999. Scaffolding and protein interactions in MAP kinase modules. Cell Calcium 26:219–26 109. Park SH, Zarrinpar A, Lim WA. 2003. Rewiring MAP kinase pathways using alternative scaffold assembly mechanisms. Science 299:1061–64 110. Therrien M, Chang HC, Solomon NM, Karim FD, Wassarman DA, Rubin GM. 1995. KSR, a novel protein kinase required for RAS signal transduction. Cell 83:879–88 111. Sundaram M, Han M. 1995. The C. elegans ksr-1 gene encodes a novel Raf-related kinase involved in Ras-mediated signal transduction. Cell 83:889–901 112. Kornfeld K, Hom DB, Horvitz HR. 1995. The ksr-1 gene encodes a novel protein kinase involved in Ras-mediated signaling in C. elegans. Cell 83:903–13 113. Cacace AM, Michaud NR, Therrien M, Mathes K, Copeland T, et al. 1999. Identiﬁcation of constitutive and ras-inducible phosphorylation sites of KSR: implications for 14-3-3 binding, mitogen-activated protein kinase binding, and KSR overexpression. Mol. Cell. Biol. 19:229–40 114. Muller J, Ory S, Copeland T, Piwnica-Worms H, Morrison DK. 2001. C-TAK1 regulates Ras signaling by phosphorylating the MAPK scaffold, KSR1. Mol. Cell 8:983–93 115. Zhou M, Horita DA, Waugh DS, Byrd RA, Morrison DK. 2002. Solution structure and functional analysis of the cysteine-rich C1 domain of kinase suppressor of Ras (KSR). J. Mol. Biol. 315:435–46 www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

797

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

118. This study showed that MEK1’s partner, MP1, is a compartmentspecific MAPK scaffold. p14 bound constitutively to MP1 and restricted the protein to the cytosolic face of early endosomes.

126. This study showed that Sef is a Golgi-localized MAPK scaffold that activates MEK/ERK complexes and permits phosphorylation of cytosolic but not nuclear substrates.

130. GFP-RBD was used to map the signaling pathway for in situ activation of Ras on the Golgi, which involves Src, PLCγ, calcium, and RasGRP1. 132. This study establishes N-Ras as the only Ras isoform activated in lymphocytes following low-grade TCR stimulation and that the site of activation is the Golgi apparatus.

798

3:17

116. Nguyen A, Burack WR, Stock JL, Kortum R, Chaika OV, et al. 2002. Kinase suppressor of Ras (KSR) is a scaffold which facilitates mitogen-activated protein kinase activation in vivo. Mol. Cell. Biol. 22:3035–45 117. Schaeffer HJ, Catling AD, Eblen ST, Collier LS, Krauss A, Weber MJ. 1998. MP1: a MEK binding partner that enhances enzymatic activation of the MAP kinase cascade. Science 281:1668–71 118. Teis D, Wunderlich W, Huber LA. 2002. Localization of the MP1-MAPK scaffold complex to endosomes is mediated by p14 and required for signal transduction. Dev. Cell 3:803–14 119. Luttrell LM, Lefkowitz RJ. 2002. The role of β-arrestins in the termination and transduction of G-protein-coupled receptor signals. J. Cell Sci. 115:455–65 120. Laporte SA, Oakley RH, Zhang J, Holt JA, Ferguson SS, et al. 1999. The β2-adrenergic receptor/βarrestin complex recruits the clathrin adaptor AP-2 during endocytosis. Proc. Natl. Acad. Sci. USA 96:3712–17 121. Goodman OB Jr, Krupnick JG, Santini F, Gurevich VV, Penn RB, et al. 1996. βarrestin acts as a clathrin adaptor in endocytosis of the β2-adrenergic receptor. Nature 383:447–50 122. Luttrell LM, Roudabush FL, Choy EW, Miller WE, Field ME, et al. 2001. Activation and targeting of extracellular signal-regulated kinases by β-arrestin scaffolds. Proc. Natl. Acad. Sci. USA 98:2449–54 123. McDonald PH, Chow CW, Miller WE, Laporte SA, Field ME, et al. 2000. β-arrestin 2: a receptor-regulated MAPK scaffold for the activation of JNK3. Science 290:1574–77 124. Furthauer M, Lin W, Ang SL, Thisse B, Thisse C. 2002. Sef is a feedback-induced antagonist of Ras/MAPK-mediated FGF signalling. Nat. Cell Biol. 4:170–74 125. Tsang M, Friesel R, Kudoh T, Dawid IB. 2002. Identiﬁcation of Sef, a novel modulator of FGF signalling. Nat. Cell Biol. 4:165–69 126. Torii S, Kusakabe M, Yamamoto T, Maekawa M, Nishida E. 2004. Sef is a spatial regulator for Ras/MAP kinase signaling. Dev. Cell 7:33–44 127. Philips MR. 2004. Sef: a MEK/ERK catcher on the Golgi. Mol. Cell 15:168–69 128. Apolloni A, Prior IA, Lindsay M, Parton RG, Hancock JF. 2000. H-ras but not K-ras trafﬁcs to the plasma membrane through the exocytic pathway. Mol. Cell. Biol. 20:2475– 87 129. Mallard F, Antony C, Tenza D, Salamero J, Goud B, Johannes L. 1998. Direct pathway from early/recycling endosomes to the Golgi apparatus revealed through the study of shiga toxin B-fragment transport. J. Cell Biol. 143:973–90 ´ 130. Bivona TG, Perez de Castro I, Ahearn IM, Grana TM, Chiu VK, et al. 2003. Phospholipase C gamma activates Ras on the Golgi apparatus by means of RasGRP1. Nature 424:694–98 131. Cullen PJ, Lockyer PJ. 2002. Integration of calcium and Ras signalling. Nat. Rev. Mol. Cell Biol. 3:339–48 ´ 132. Perez de Castro I, Bivona T, Philips M, Pellicer A. 2004. Ras activation in Jurkat T cells following low-grade stimulation of the T-cell receptor is specific to N-Ras and occurs only on the Golgi. Mol. Cell. Biol. 24:3485–96 133. Caloca MJ, Zugaza JL, Bustelo XR. 2003. Exchange factors of the RasGRP family mediate Ras activation in the Golgi. J. Biol. Chem. 278:33465–73 134. Carrasco S, Merida I. 2004. Diacylglycerol-dependent binding recruits PKCtheta and RasGRP1 C1 domains to speciﬁc subcellular localizations in living T lymphocytes. Mol. Biol. Cell 15:2932–42 Mor

·

Philips

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

ANRV270-IY24-25

ARI

15 February 2006

3:17

134a. Arozarena I, Matallanas D, Berciano MT, Sanz-Moreno V, Calvo F, et al. 2004. Activation of H-Ras in the endoplasmic reticulum by the RasGRF family guanine nucleotide exchange factors. Mol. Cell. Biol. 24(4):1516–30 135. Sobering AK, Romeo MJ, Vay HA, Levin DE. 2003. A novel Ras inhibitor, Eri1, engages yeast Ras at the endoplasmic reticulum. Mol. Cell. Biol. 23:4983–90 136. Sobering AK, Watanabe R, Romeo MJ, Yan BC, Specht CA, et al. 2004. Yeast Ras regulates the complex that catalyzes the first step in GPI-anchor biosynthesis at the ER. Cell 117:637–48 137. Miyazaki T, Neff L, Tanaka S, Horne WC, Baron R. 2003. Regulation of cytochrome c oxidase activity by c-Src in osteoclasts. J. Cell Biol. 160:709–18 138. Boerner JL, Demory ML, Silva C, Parsons SJ. 2004. Phosphorylation of Y845 on the epidermal growth factor receptor mediates binding to the mitochondrial protein cytochrome c oxidase subunit II. Mol. Cell. Biol. 24:7059–71 139. Pagliarini DJ, Wiley SE, Kimple ME, Dixon JR, Kelly P, et al. 2005. Involvement of a mitochondrial phosphatase in the regulation of ATP production and insulin secretion in pancreatic beta cells. Mol. Cell 19:197–207 140. Rebollo A, Perez-Sala D, Martinez AC. 1999. Bcl-2 differentially targets K-, N-, and HRas to mitochondria in IL-2 supplemented or deprived cells: implications in prevention of apoptosis. Oncogene 18:4930–39 141. Wang HG, Rapp UR, Reed JC. 1996. Bcl-2 targets the protein kinase Raf-1 to mitochondria. Cell 87:629–38 141a. Bivona TG, Quatela SE, Bodemann BO, Ahearn IM, Soskis MJ, et al. 2006. Phosphorylation of K-Ras by PKC regulates a farnesyl-electrostatic switch that promotes association with Bcl-Xl on mitochondria and induces apoptosis. Mol. Cell. In press 142. Papadaki P, Pizon V, Onken B, Chang EC. 2002. Two ras pathways in ﬁssion yeast are differentially regulated by two ras guanine nucleotide exchange factors. Mol. Cell. Biol. 22:4598–606 143. Simons K, Ikonen E. 1997. Functional rafts in cell membranes. Nature 387:569–72 144. Anderson RG. 1998. The caveolae membrane system. Annu. Rev. Biochem. 67:199–225 145. Simons K, Toomre D. 2000. Lipid rafts and signal transduction. Nat. Rev. Mol. Cell Biol. 1:31–39 146. Munro S. 2003. Lipid rafts: elusive or illusive? Cell 115:377–88 147. Mineo C, Gill GN, Anderson RG. 1999. Regulated migration of epidermal growth factor receptor from caveolae. J. Biol. Chem. 274:30636–43 148. Huster D, Vogel A, Katzka C, Scheidt HA, Binder H, et al. 2003. Membrane insertion of a lipidated ras peptide studied by FTIR, solid-state NMR, and neutron diffraction spectroscopy. J. Am. Chem. Soc. 125:4070–79 149. Melkonian KA, Ostermeyer AG, Chen JZ, Roth MG, Brown DA. 1999. Role of lipid modiﬁcations in targeting proteins to detergent-resistant membrane rafts. Many raft proteins are acylated, while few are prenylated. J. Biol. Chem. 274:3910–17 150. Furuchi T, Anderson RG. 1998. Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-related kinase (ERK). J. Biol. Chem. 273:21099–104 151. Song KS, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP. 1996. Copuriﬁcation and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains. Detergent-free puriﬁcation of caveolae microdomains. J. Biol. Chem. 271:9690–97 152. Prior IA, Harding A, Yan J, Sluimer J, Parton RG, Hancock JF. 2001. GTP-dependent segregation of H-ras from lipid rafts is required for biological activity. Nat. Cell Biol. 3:368–75 www.annualreviews.org • Compartmentalized Ras/MAPK Signaling

136. This paper reports that Eri1, a component of the enzyme that initiates GPI-anchor synthesis, is not only the first ER-associated Ras effector described but also the first effector that is negatively regulated by GTP-bound Ras.

799

ANRV270-IY24-25

ARI

15 February 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

153. This study presents compelling evidence for compartmentalized Ras signaling within a single organelle, the plasma membrane.

800

3:17

153. Prior IA, Muncke C, Parton RG, Hancock JF. 2003. Direct visualization of Ras proteins in spatially distinct cell surface microdomains. J. Cell Biol. 160:165–70 154. Rotblat B, Prior IA, Muncke C, Parton RG, Kloog Y, et al. 2004. Three separable domains regulate GTP-dependent association of H-ras with the plasma membrane. Mol. Cell. Biol. 24:6799–810 155. Murakoshi H, Iino R, Kobayashi T, Fujiwara T, Ohshima C, et al. 2004. Single-molecule imaging analysis of Ras activation in living cells. Proc. Natl. Acad. Sci. USA 101:7317–22 156. Fujiwara T, Ritchie K, Murakoshi H, Jacobson K, Kusumi A. 2002. Phospholipids undergo hop diffusion in compartmentalized cell membrane. J. Cell Biol. 157:1071–81 157. Hancock JF, Parton RG. 2005. Ras plasma membrane signalling platforms. Biochem. J. 389:1–11 158. Janes PW, Ley SC, Magee AI, Kabouridis PS. 2000. The role of lipid rafts in T cell antigen receptor (TCR) signalling. Semin. Immunol. 12:23–34 159. Langlet C, Bernard AM, Drevot P, He HT. 2000. Membrane rafts and signaling by the multichain immune recognition receptors. Curr. Opin. Immunol. 12:250–55 160. Xavier R, Brennan T, Li Q, McCormack C, Seed B. 1998. Membrane compartmentation is required for efﬁcient T cell activation. Immunity 8:723–32 161. Janes PW, Ley SC, Magee AI. 1999. Aggregation of lipid rafts accompanies signaling via the T cell antigen receptor. J. Cell Biol. 147:447–61 162. Zhang W, Sloan-Lancaster J, Kitchen J, Trible RP, Samelson LE. 1998. LAT: the ZAP70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell 92:83–92 163. Zhang W, Trible RP, Samelson LE. 1998. LAT palmitoylation: its essential role in membrane microdomain targeting and tyrosine phosphorylation during T cell activation. Immunity 9:239–46 164. Viola A, Schroeder S, Sakakibara Y, Lanzavecchia A. 1999. T lymphocyte costimulation mediated by reorganization of membrane microdomains. Science 283:680–82 165. Moran M, Miceli MC. 1998. Engagement of GPI-linked CD48 contributes to TCR signals and cytoskeletal reorganization: a role for lipid rafts in T cell activation. Immunity 9:787–96 166. Douglass AD, Vale RD. 2005. Single-molecule microscopy reveals plasma membrane microdomains created by protein-protein networks that exclude or trap signaling molecules in T cells. Cell 121:937–50 167. P´erez de Castro I, Diaz R, Malumbres M, Hernandez MI, Jagirdar J, et al. 2003. Mice deﬁcient for N-ras: impaired antiviral immune response and T-cell function. Cancer Res. 63:1615–22 168. Ebinu JO, Stang SL, Teixeira C, Bottorff DA, Hooton J, et al. 2000. RasGRP links T-cell receptor signaling to Ras. Blood 95:3199–203 169. Lockyer PJ, Kupzig S, Cullen PJ. 2001. CAPRI regulates Ca2+ -dependent inactivation of the Ras-MAPK pathway. Curr. Biol. 11:981–86

Mor

·

Philips

Contents

ARI

7 February 2006

11:41

Contents

Annual Review of Immunology Volume 24, 2006

Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

Frontispiece Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p x The Tortuous Journey of a Biochemist to Immunoland and What He Found There Jack L. Strominger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Osteoimmunology: Interplay Between the Immune System and Bone Metabolism Matthew C. Walsh, Nacksung Kim, Yuho Kadono, Jaerang Rho, Soo Young Lee, Joseph Lorenzo, and Yongwon Choi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33 A Molecular Perspective of CTLA-4 Function Wendy A. Teft, Mark G. Kirchhof, and Joaquín Madrenas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p65 Transforming Growth Factor-β Regulation of Immune Responses Ming O. Li, Yisong Y. Wan, Shomyseh Sanjabi, Anna-Karin L. Robertson, and Richard A. Flavell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p99 The Eosinophil Marc E. Rothenberg and Simon P. Hogan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 147 Human T Cell Responses Against Melanoma Thierry Boon, Pierre G. Coulie, Benoît J. Van den Eynde, and Pierre van der Bruggen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 175 FOXP3: Of Mice and Men Steven F. Ziegler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209 HIV Vaccines Andrew J. McMichael p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Natural Killer Cell Developmental Pathways: A Question of Balance James P. Di Santo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 257 Development of Human Lymphoid Cells Bianca Blom and Hergen Spits p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 287 Genetic Disorders of Programmed Cell Death in the Immune System Nicolas Bidère, Helen C. Su, and Michael J. Lenardo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 321

v

Contents

ARI

7 February 2006

11:41

Genetic Analysis of Host Resistance: Toll-Like Receptor Signaling and Immunity at Large Bruce Beutler, Zhengfan Jiang, Philippe Georgel, Karine Crozat, Ben Croker, Sophie Rutschmann, Xin Du, and Kasper Hoebe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353 Multiplexed Protein Array Platforms for Analysis of Autoimmune Diseases Imelda Balboni, Steven M. Chan, Michael Kattah, Jessica D. Tenenbaum, Atul J. Butte, and Paul J. Utz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 391 How TCRs Bind MHCs, Peptides, and Coreceptors Markus G. Rudolph, Robyn L. Stanﬁeld, and Ian A. Wilson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 419 Annu. Rev. Immunol. 2006.24:771-800. Downloaded from arjournals.annualreviews.org by HINARI on 08/26/07. For personal use only.

B Cell Immunobiology in Disease: Evolving Concepts from the Clinic Flavius Martin and Andrew C. Chan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467 The Evolution of Adaptive Immunity Zeev Pancer and Max D. Cooper p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 497 Cooperation Between CD4+ and CD8+ T Cells: When, Where, and How Flora Castellino and Ronald N. Germain p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Mechanism and Control of V(D)J Recombination at the Immunoglobulin Heavy Chain Locus David Jung, Cosmas Giallourakis, Raul Mostoslavsky, and Frederick W. Alt p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541 A Central Role for Central Tolerance Bruno Kyewski and Ludger Klein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 571 Regulation of Th2 Differentiation and Il4 Locus Accessibility K. Mark Ansel, Ivana Djuretic, Bogdan Tanasa, and Anjana Rao p p p p p p p p p p p p p p p p p p p p p p p 607 Diverse Functions of IL-2, IL-15, and IL-7 in Lymphoid Homeostasis Averil Ma, Rima Koka, and Patrick Burkett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 657 Intestinal and Pulmonary Mucosal T Cells: Local Heroes Fight to Maintain the Status Quo Leo Lefrançois and Lynn Puddington p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 681 Determinants of Lymphoid-Myeloid Lineage Diversiﬁcation Catherine V. Laiosa, Matthias Stadtfeld, and Thomas Graf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 705 GP120: Target for Neutralizing HIV-1 Antibodies Ralph Pantophlet and Dennis R. Burton p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 739 Compartmentalized Ras/MAPK Signaling Adam Mor and Mark R. Philips p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 771

vi

Contents