Advances in Immunology Volume 62

ADVANCES IN

Immunology VOLUME 62

This Page Intentionally Left Blank

ADVANCES IN

Immunology EDITED BY

FRANK J. DIXON Research Institute of Scripps Clinic La Jolla, California ASSOCIATE EDITORS

Frederick Alt K. Frank Austen Tadamitsu Kisimoto Fritz Melchers Jonathan W. Uhr

VOLUME 62

ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto

Find Us on the Web! http://www.apnet.com

This book is printed on acid-free paper.

@

Copyright 0 1996 by ACADEMIC PRESS, INC, All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Academic Press, Inc. A Division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495 United Kingdom Edition published by Academic Press Limited 24-28 Oval Road. London NW 1 7DX

International Standard Serial Number: 0065-2776 International Standard Book Number: 0-12-022462-3 PRINTED IN THE UNITED STATES OF AMERICA 96 97 9 8 9 9 00 0 1 E B 9 8 7 6 5

4

3 2 1

CONTENTS

CONTHIBUTOHS

ix

Organization of the Human Immunoglobulin Heavy-Chain Locus

FUMIHIKO MATSUDA A N D TASUKU HONJO

I. Introduction 11. Structure and Organization of the Human VH Loci 111. Analysis of Human V,, Sequences IV. VH Segment Usage and Repertoire Formation V. Evolution of the Human V,, Loci References

1 2 12 17 19 26

Analysis of Gene Function in Lymphocytes by RAG-2-Deficient Blastocyst Complementation JIANZHU

CHEN

I. Introduction 11. RAG-2-Deficient Blastocyst Complementation 111. Comparison of Germline Mutational Approach and RAG-2Deficient Blastocyst Complementation IV. Technical Aspects of RAG-2-Deficient Blastocyst Complementation V. Analysis of Gen Function in pymphocytes f VI. Conclusion References

31 31 34 36 41 53 54

Interferon-y: Biology and Role in Pathogenesis

ALFONSBILLIAU

I. 11. 111. IV. V.

Introduction Structure and Structure-Function Relationship Producers and Production of IFN-7 Biochemical Basis of IFN-7 Action Biological Effects on Cells and Tissues Y

61 62 63 66 76

vi

CONTENTS

VI. VII. VIII. IX.

Role of IFN-y in Infection and Cancer Role of IFN-y in Inimunopathology Final Thoughts Notes Added in Proof References

94 96 105 106 108

Role of the CD28-B7 Costimulatory Pathways in T Cell-Dependent B Cell Responses

KARENS. HATHCOCK AND RICHARD J. HODES I. 11. 111. IV. V. VI. VII.

Introduction Two-Signal Model of T Cell Activation CD2WCTLA-4 Receptor Family B7-1/B7-2 Ligand Family Contribution of CD28 and CTLA-4 to Costimulation Contribution of B7-1 and B7-2 to Costimulation Regulation of in Vivo Td Immune Responses by CD28-B7 Costimulation VIII. Cellular Events in Td B Cell Responses IX. Contribution of CD28-B7 Costimulation to Td B Cell Responses X. A Model for the Function of CD28-B7 Interactions in Td B Cell Responses XI. Concluding Remarks References

131 131 132 135 138 140

142 144 148 154 156 157

Prostaglandin Endoperoxide H Synthases-1 and -2

WILLIAM L. SMITHAND DAVID L. DEWITT I. An Overview 11. PGHS-1 and PGHS-2 Structure/Function Interrelationships sis by PGHS-1 and PGHS-2 of Expression of the Genes for PGHS-1 and PGHS-2 Physiological Actions of PGHS-1 and PGHS-2 VI. PGHS-1 and PGHS-2 in Pathophysiologies-Inflammation, Thrombosis, and Colon Cancer VII. Future Work References

167 169 178 188 196 200 201 202

Human Tumor Antigens Are Ready to Fly

ROBERTA. HENDERSON AND OLIVERA J. F I N N 1. 11. 111. IV.

Introduction Tumor Antigens Defined by Antibodies Tumor Antigens Defined by T Cells Reflections and Perspectives References

217 218 227 240 246

CONTENTS

vii

Inflammatory Mediators, Cytokines, and Adhesion Molecules in Pulmonary Inflammation and Injury

NICHOLAS W. LUKACS AND PETER A. WARD I. 11. 111. IV.

Introduction Pathophysiolo ‘c Mechanisms of Inflammatory Injury Aniind Mode s of Lung Injury Therapeutic Interventions V. Conclusions References

P

INDEX CONTENTS OF RECENTVOLUMES

257 260 271 285 290 290 305 315

This Page Intentionally Left Blank

CONTRIBUTORS

Numbers in pnrenthe.srs indicate the p a p s on which the authors’ contribution9 begin.

Alfons Billiau (61), Rega Institute, University of Leuven, Leuven B3000, Belgium Jianzhu Chen (31), Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 David L. DeWitt (167), Department of Biochemistry, Michigan State University, East Lansing, Michigan 48824 Olivera J. Finn (217), Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15260 Karen S. Hathcock (131),Experimental Immunology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892 Robert A. Henderson (217), Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15260 Richard J. Hodes (131), Experimental Immunology Branch, National Cancer Institute and National Institute on Aging, National Institutes of Health, Bethesda, Maryland 20892 Tasuku Honjo ( l ) ,Center for Molecular Biology and Genetics, Kyoto University, Kyoto 60601, Japan Nicholas W. Lukacs (257), Department of Pathology, University of Michigan Medical School, Ann Arbor, Michigan 48109 Fumihiko Matsuda (l),Center for Molecular Biology and Genetics, Kyoto University, Kyoto 60601, Japan William L. Smith (167), Department of Biochemistry, Michigan State University, East Lansing, Michigan 48824 Peter A. Ward (257), Department of Pathology, The University of Michigan Medical School, Ann Arbor, Michigan 48109

This Page Intentionally Left Blank

ADVANCES IN IMMUNOLOGY. VOL 62

Organization of the Human Immunoglobulin Heavy-Chain Locus FUMlHlKO MATSUDA AND TASUKU HONJO Center for Mderukrr Biobgy and Genetics, K p b llniversify, Kyob 60601, Japon

1. Introduction

The number and organization of the immunoglobulin (Ig) variable (V) region genes provide the germline basis of antibody repertoire, which is further amplified by somatic DNA recombination and hypermutation. In the Ig h e a v (H)-chain, the V region is responsible for antigen binding whereas the constant (C) region specifies the isotype of Ig (reviewed by Tonegawa, 1983; Honjo and Habu, 1985). Genes encoding IgH V regions are split into VH,diversity (DH),and joining ( J H ) segments, each of which is composed of multiple copies, and these gene segments are physically linked on a single chromosome. The complete V region gene is created by assembly of each one of the three segments by a somatic genetic event called V-D-J recombination during the ontogeny of B lymphocytes. In human, VH region genes are mapped to chromosome 14 q32.33 (Croce et al., 1979;Kirsch et al., 1982).The human VHlocus consists of approximately 100 copies of VH segments, approximately 30 DH segments, and 6 JH segments, whereas the human CHlocus comprises 11 CH genes. The V H and CH loci are tightly linked on the chromosome in the order of 5’-V,DH-J~-C~-3’. The distance between the 3’-most VH locus segment ( JH) and the 5’-most CHgene (Cp) is approximately 8 kb in man (Ravetch et

al., 1981). The complete elucidation of the organization and structure of the germline VHsegments will allow us to address a number of questions concerning the molecular events underlying the generation of antibody diversity. Comparison between the total germline and expressed VHsequences will provide a clue to estimate the relative contribution of germline VH repertoire, VD-J recombination, somatic hypermutation, and subsequent selection of B lymphocytes to Ig repertoire. Accumulating evidence indicates that the human VHlocus is highly polymorphic. Complete understanding of the VH organization is essential for the accurate characterization of polymorphic variation in the number and structure of germline VH segments and its genetic association with the susceptibility to immune disorders. A number of reports showed the overrepresentation of particular VH segments, suggesting the preferential utilization of VH segments. It is important to test whether there is any association of the germline organization or structure of VH segments with the biased usage. Finally, from the evolutionary point 1 Copyright 0 1996 by Academic Press, Inc All rights of reproductitin 111 any fom reservrd

2

F U M I H I K O MATSUDA AND TASUKU HONJO

of view, the human VH segments, like any other multigene families, are considered to have evolved through repeated duplication and recombination of DNA. Needless to say, diversification of newly generated VH members by such events contributes to the germline VH repertoire. Detailed studies on the VH organization and structure will allow us to elucidate the molecular mechanisms that govern evolution of multigene families. The long-standing effort to understand the overall organization of the entire human VH locus by several research groups resulted in the completion of the physical map of the 1.1-megabase (Mb) DNA (Matsuda et al., 1993; Cook et al., 1994). More than 80 VH segments within the locus were localized along the chromosome and their nucleotide sequence was determined. In addition, human VH and DH segments are shown to be frequently dispersed on other chromosomes (Matsudaet al., 1990; Nagaoka et al., 1994; Tomlinson et al., 1994). Furthermore, isolation of the total human VH segments has played an important role for generation of human Ig in J H segment-disrupted mice carrying human Ig mini loci (Taylor et al., 1994; Green et al., 1994) or in phage-displayed Ig libraries (Griffiths et al., 1994). In this chapter, we overview the organization and structure of the human IgH locus based on the recent findings and discuss its biological significance and implication. II. Structure and Organization of he Human VH Loci

A. PHYSICAL MAPPINGOF THE HUMAN VH Locus O N CHROMOSOME 14 Large-scale isolation and physical mapping of the human VH locus were initiated by cosmid cloning (Kodairaet al., 1986). Distribution of VHfamilies on 23 cosmid clones has shown that members of different Vzr families are interspersed as previously described (Rechavi et al., 1982). Another important conclusion from early studies using cosmid and phage vectors is the presence of abundant pseudogenes (about 40%), many of which are highly conserved with only a few point mutations (Givol et al., 1981; Kodaira et al., 1986; Berman et al., 1988). Physical linkage between the 3’-most VHsegment (V6-1)and the JH segments was completed by extensive analyses of DH and JH regions in cosmid clones. The D3 segment was mapped only 22 kb upstream of the JH cluster (Matsuda et al., 1988; Buluwela et al., 1988) and the V6-1 segment was mapped 20 kb upstream of the D4segment (Buluwela and Rabbits, 1988; Sat0 et al., 1988; Schroeder et al., 1988). Development of the techniques to separate DNA fragments of more than 100 kb by pulse field gel electrophoresis (PFG) made it possible to survey a more general organization of the entire human VH locus. By the analysis using PFG that allows one to examine the VH content on a few

ORGANIZATION OF THE HEAW-CHAIN LOCUS

3

hundred- to 1000-kb DNA fragments, the total size of the human VH locus was estimated to be about 2.5-3.0 Mb (Berman et al., 1988; Matsuda et al., 1988) including VH- and D,-hybridizing fragments that later mapped to chromosomes 15 and 16. Above all, PFG analysis using two-dimensional DNA electrophoresis (2D-DE) has provided a more precise organization of the total VH locus of approximately 1.2 Mb. Altogether, 76 human VH segments were localized in DNA fragments separated by 2D-DE of DNA digested with SjI, BssHII, and Not1 (Walter et al., 1990). A common insertioddeletion polymorphism of 80 kb in length involving at least three VH segments was identified within the JH distal part of the VHlocus (Walter et al., 1990, discussed below). The physical map generated by 2D-DE was further refined using the deletion profile of VH segments associated with V-D-J recombination in human B cell lines (Walter et al., 1991a). These studies provided the overall organization of the human VHlocus of better resolution with a more reliable VHnumber within the locus although the precise location of each VH segment cannot be determined and some VH segments were inevitably missed. Although the initial approach to the physical mapping by isolation of cosmids revealed several unique features on the organization of the human VHlocus as mentioned previously, it was almost impossible to isolate contigs larger than 100 kb with cosmids or phages because of several technical limitations, especially the difficulty in the linkage around duplicated copies and repeated sequences. To circumvent these difficulties, it was essential to isolate large chromosomal regions as a single insert by using different vector-host systems. The yeast artificial chromosome (YAC)system, which can clone up to several hundred kilobases of DNA (Burke et al., 1987), has been a key in the completion of the physical mapping of the human VH locus. Several different human YAC libraries were screened by PCR with VH-specificoligonucleotide primers and more than 50 VH-carrying YACs were isolated (Matsuda et al., 1993). The clones were classified into several contigs based on their physical maps constructed by PFG and the content of VH segments in each clone. For detailed structural analysis, however, the YAC system had several disadvantages including a low copy number of artificial chromosomes and difficulty in manipulation of large insert DNA. These shortcomings were compensated by subcloning the whole YAC insert into cosmids. Many of the YAC clones were subcloned into cosmids, phages, or plasmids, and complete restriction maps of YAC contigs were constructed by detailed analysis of these subclones (for detailed procedure, see Matsuda and Honjo, 1993). The first report using YAC cloning has identified and localized five VH segments in the JH proximity (Shin et nl., 1991). These authors proposed to name all the VH segments by the family number and the order from

4

FUMIHIKO MATSUDA AND TASUKU HONJO

the 3' end of the Vn locus. For example, V3-36P indicates a vH3 family member located at the 36th from the 3' most VH segment, i.e., V6-1. P indicates the pseudogene. An insertional polymorphic VH segment is indicated by a number with decimal point (V7-4.1 is an insertional VH segment between V1-4 and V2-5). This newly proposed nomenclature defined VHsegments only when they were mapped on the chromosome, excluding the confusion due to the misassignment of somatically mutated VHcDNAs as different Vn segments. The physical map was extended to approximately 70% (0.8 Mb) of the human Vn locus covered by more than seven overlapping YAC clones (Fig. 1). All the YACs were subcloned into

17 H

. .

Y103 3-23 3-22P II I.

3-21 II II

3-20 I. I.

3-19P II

I.

3-16P 1-17P 3-15 1-14P

1-18

In

Y91

.. ..

4-39

a .

u

bin

IU

Y20

1-12P 3-9 3-13 3-11 2-1OP

1-8 200

U

3-37P

4-31 3-30 3-29P 7-27P

Y24

3-64 3-82P 360P 1-58P 7-58P 3-54P 3-52P

Y6 Y24

-

551 IP

I----3-76P

--

13.3 YW6 13-p26.1-5 200

150

349

3-48

II

I

II

.

Y21

146 3-47P 1-45 I Y IN

344P 3-43 600 ID

11 I

WH6

CY24-68

.

3-SOP

100

yLiz

Ytll ytLl

polymorphic(ref.3)

; , I , I

-

50

u u 111-

0

kb

FIG.1. Organization of the human IgH locus. The 1.1-Mh DNA is shown by five thick horizontal lines with the 3' end at the top right corner. YAC and cosmid clones covering the region are shown with their names by horizontal lines below. VH segments belonging to vH1, 2 , 3 , 4 , 5 , 6 ,and 7 families are indicated by the symbols m, B, m, m, B, m, and 0, respectively. Exact location of the V,, contig located between V2-70 and V1-67P and relative order of four VHsegments within the contig have not been determined. Nine VHsegments whose sequences have not been determined are underlined. References for polymorphic insertions are: reference 1, Shin et al., 1991; reference 2, Walter et al., 1993; reference 3, F. Matsuda, unpublished results.

ORGANIZATION OF THE HEAVY-CHAIN LOCUS

5

cosmids, phages, or plasmids and the location and the relative order of a total of 64 VH segments were determined by Southern blot analysis. The results show that the intergenic distance is quite variable. The average distance between neighboring VH segments is approximately 12 kb and the longest region (approximately 50 kb) without VH segments is that between V6-1 and V1-2. Two small internal duplications were identified by careful comparison of restriction site maps among different portions within the locus. V3-62P-V4-61 and V3-6OP-V4-59 were described as a part of cluster 71 consisting of seven VH segments (Kodaira et al., 1986). Another tandem homology pair is V3-33-V3-32P-V4-31 and V33O-V3-29P-V4-28. The VH-hybridizingDNA fragments were further subcloned into plasmids and nucleotide sequences of 64 VH segments within the 0.8-Mb region were determined (Matsuda et al., 1993).Transcriptional orientation of 43 V, segments that are located at the 5’-most, middle, or 3’-most part of the contig was determined, and all of them were found to have the same polarity as the JH segments as shown in Fig. 1 (Matsuda et al., 1993;F. Matsuda, unpublished observation). The results indicate that there is no gross inversion within the human VHlocus and that the majority of VH segments rearrange to associate with the DH and JH segments by looping out but not by inversion as observed in the human VKlocus (Lorenz et al., 1987). The distal part of the human VH locus was identified by Cook et al. (1994) using the human telomere activity in yeast and a chromosome translocation that places telomere-proximal VH segments onto chromosome 8. A 200-kb clone (yIgH6) carrying 19 VH segments was isolated and subcloned into cosmids. Physical mapping and nucleotide sequencing analysis revealed that the two 3’-most VHsegments on yIgH6 were virtually identical to V3-63P and V3-64, which are the 5’-most VH segments previously identified by Matsuda et al. (1993),thus linking the entire VHlocus physically by YAC clones (Fig. 1). Notably, the 5’-most V H segment (V781) is located only a few kilobases downstream of the telomeric repeat sequence. Matsuda and colleagues isolated two independent YAC clones (13.3 and 111) from different YAC libraries (F. Matsuda, unpublished data). The clone 13.3 contains the 5’-most VHsegments and 14q telomere. Comparison of physical maps of yIgH6 and 13.3revealed that 13.3covers the region between V2-70 and 14q telomere and showed the identical VH segment content and organization to the corresponding region of yIgH6, confirming the telomeric end of the physical map by Cook et al. (1994) (Fig, 1). On the other hand, the clone 111 contains 18 VH segments and shares the common region between V3-50P and V3-64 with Y6/Y24. The 5’ part of this clone contains three VH segments that correspond to V365P, V3-66, and V1-67P. However, a striking difference between the physi-

6

FUMIHIKO MATSUDA AND TASUKU HONJO

cal maps of yIgH6 and 111 begins at the 5'-flanking region of V1-67P, suggesting that these two clones represent different alleles as reported previously (Walter et al., 1990). Two-color in situ hybridization (FISH) of interphase nuclei showed that the estimated size of the gap between cosmid clone CY24-68 and 13-p26.1-5 (Fig. 1) is either 194 2 27 kb or 109 kb depending on haplotypes. Because the distance between the same clones is approximately 90 kb on the map of yIgH6, this result again indicates the presence of a large deletion polymorphism (about 90 kb) in the region between V1-67P and V2-70 and the clone yIgH6 is likely to represent a haplotype with a large deletion. Additional supporting evidence was obtained by isolation of a 50-kb contig from a lymphoblastoid cell line that does not have this large deletion (F. Matsuda, unpublished observation). One novel vH2 family segment was identified in the newly isolated contig (Fig. 1).The BglII fragment containing the v H 2 segment has the same size (18 kb) as that carrying VH2-2, which serves as a merkmal of this deletion (Walter et al., 1990). The result allowed us to localize the contig within the gap between 13.3 and 111. In addition, three VH1-hybridizing segments were found within this contig (Fig. 1).The direct test by sequencing these VH segments will clarify whether any of them correspond to V168 and V1-69. Isolation of independent YAC or cosmid clones covering this region would give a direct and clear answer to the above issue, excluding the possible misinterpretations that could arise as a cloning artifact due to the deleterious recombination of YAC in yeast cells. The complete physical map of the human VHlocus by the linkage between yIgH6 and Y6/Y24 shows that the size of the human VHlocus is 1.0-1.1 Mb depending on haplotypes.

B. NUMBEROF THE HUMAN VHSEGMENTS Estimation of the total number of the human VH segments has been performed using a variety of methods. In earlier studies, the number of VH segments was roughly estimated to be between 100 and 200 either by counting the number of VH-hybridizingrestriction fragments by Southern blot analysis or by determination of the number of independent VHclones in a single genome equivalent phage library (Kodaira et al., 1986; Lee et al., 1987; Berman et aE., 1988). More reliable and precise calculation using 2D-DE (Walter et al., 1990) or specific amplification of VH segments by PCR (Tomlinson et at., 1992) estimated the total number to be 76 and 74, respectively. Another study using the quantitative hybridization analysis predicted approximately 120 VH segments within the genome including orphon VHsegments (Matsuda et al., 1993).The completion of the physical map of the entire human VH locus settled the issue. The total number of the VH segments in the smallest haplotype is calculated to be 81 with an

7

ORGANIZATION OF THE HEAVY-CHAIN LOCUS

additional 10 or more polymorphic VH segments (Table I), which is in general agreement with the previous estimations, and nucleotide sequences of 82 of 91 mapped VH segments were determined (Matsuda et al., 1993; Cook et al., 1994).Isolation of YAC clones as well as PCR-based VHcloning identified a total of 24 translocated VH segments on chromosomes 15 and 16 (Nagaoka et al., 1994; Tomlinson et al., 1994, see below). Hence, the total number of the human VH segments in the genome would be 115 (Table I). However, many DNA fragments of YAC and cosmid clones in the VH locus weakly hybridize with VH probes, although such fragments are not detectable in Southern hybridization of genomic DNA (F.Matsuda, unpublished observation). If these fragments correspond to VH pseudogenes with extensive divergence or their relicts as identified in the JHproximal part (Buluwela et al.,1988), the number of VH segments would be greater than the number mentioned previously. I N THE HUMAN VH Locus C. POLYMORPHISMS RFLP and DNA sequencing have shown that there are a number of polymorphic VH alleles. Deletion polymorphisms whose exact location on the chromosome was determined are shown in Fig. 1. Polymorphic insertion (or deletion) appears to cluster at three regions; V4-W2-5, V3-30/ V4-31, and V1-67PN2-70 (discussed below). In addition, some of the germline VHsegments mapped to chromosome 14 have not been located in the current map (Fig. l),suggesting that polymorphisms are often accompanied by deletion or insertion of VH segments. One of the most polymorphic VH segments is V1-69, which has 13 known alleles including duplication (Sasso et al., 1993) and is located very close to the region of the large polymorphic deletion. Walter et al. (1993) identified an insertion polymorphism of 50-kb DNA containing five functional VHsegments be-

TABLE I SUMMARY OF THE HUMAN VH SEGMENTS~ V, families Chromosomes

(length)

1

2

3

4

5

6

7

Total

14q32 (1100kb) 15qll (>250 kb) 16pll (>700 kb) Total (>2 Mb)

17 6 4 27

5 0 1 6

48 1 11 60

12 1 0 13

2 0 0

1 0 0 1

6 0 0 6

91 8 16 115

2

a The number of V, segments on chromosome 14 is calculated by the results from Matsuda et nl. (1993) and C w k et al. (1994). Six polymorphic VH segments reported (Shin et al., 1991; Walter et al. 1993) and four novel VH segments are included. Information of V, segments on chromosomes 15 and 16 is taken from Nagaoka et at. (1994) and Tomlinson et al. (1994).

8

FUMIHIKO MATSUDA AND TASUKU HONJO

tween V4-31 and V3-30 that is present in 73% of the Caucasian population. V4-30.1 and V3-30 are duplicates of V4-31 and V3-30.5, respectively. Interestingly, the insertion is located between a tandem homology pair of V,, segments, namely V3-33N3-32PN4-31 and V3-3ON3-29PN4-28, suggesting that this region is highly recombinogenic and that 27% of alleles lost five VH segments by homologous recombination. The Jn-proximal 100-kb region consisting of four VH segments also has extraordinary polymorphism. Comparative analyses of the two contigs isolated from different DNA sources (termed haplotypes A and B) showed that the restriction sites surrounding VH segments are almost identical, and the relative order of VHsegments is the same except for the insertion of one additional VHsegment (V7-4.1) in haplotype B (Shin et al., 1991). Nucleotide and amino acid sequences of corresponding VH segments share 97-98.5 and 95-97% identity, respectively, except for the pair V4-4 and V4-4b (discussedbelow). It is not clear whether corresponding VHsegments have different antigen specificities when used in functional VH regions. However, the affinities of the antibody for its ligand can be affected by this level of amino acid difference because even mutations in F R residues of the Ig have been shown to influence the binding affinity (Foote and Winter, 1992). Expression of particular allelic variants could also influence the efficiency of H-L chain pairing or interaction with B cell superantigens. Later RFLP analysis showed that there are at least four alleles in both European and Japanese populations and the frequency of each allele is remarkably different between the two populations (Shin et al., 1993a). It would be interesting to know whether any particular DNA sequences are responsible for these clustered polymorphic deletions (or insertions). A project to determine the entire nucleotide sequence of the human VH locus is under way by F. Matsuda and colleagues and more than 700-kb DNAs have been sequenced to date (F.Matsuda, unpublished results). The completion of the project will provide an answer to the previous question. It is important to test whether VH polymorphisms are associated with disease susceptibility. Some reports suggested the association of VHpolymorphisms with autoimmune diseases, such as rheumatoid arthritis, systemic lupus erythematosus, and multiple sclerosis (Yanget al., 1990;Walter et al., 1991b), whereas others reported the absence of a clear association (Hashimoto et al., 1993; Shin et aE., 1993a). However, in many cases V, probes are used to look for polymorphisms because only limited numbers of polymorphic probes are available. This makes the analysis complicated and confused because of multiple hybridizing fragments. The best procedure would be to test polymorphisms using a series of intergenic primers or probes distributed at certain intervals across the locus. Fourteen nonrepetitive intergenic probes were isolated from different parts of the locus

ORGANIZATION OF THE HEAW-CHAIN LOCUS

9

to trace the evolution of the VH locus (Matsumura et al., 1994, discussed below). Extensive RFLP analysis using these probes showed that 7 of them detect RFLP in human DNAs, suggesting that these probes can serve as good genetic markers (F. Matsuda, unpublished results). In addition, more than 20 mononucleotide or dinucleotide repeats were found to be scattered throughout the locus by nucleotide sequencing analysis (F. Matsuda, unpublished results). These probes or primers would facilitate the systematic analyses for the haplotype definition, and particularly for susceptibility studies of polymorphisms to immune disorders.

D. ORGANIZATION OF DH A N D JH SEGMENTS The 70-kb region of the 3‘-most part of the human VH locus is occupied by DH and JH gene segments. The human JH cluster contains three pseudo JH segments interspersed among six functional JH segments and is located approximately 8 kb upstream of the C p gene (Ravetch et al., 1981). A human counterpart to the murine DQ52 segment is located about 100 bp 5‘ of the first functional JH (Ravetch et al., 1981). A number of additional human DH segments have been identified in the region between the V61 and JH cluster. The first report of human germline DH segment described the isolation of a family of DH segments (D1-D4 or DLRI-DLRI) using as a probe an aberrantly rearranged DH-JH segment in a CLL cell clone (Siebenlist et al., 1981). The fact that they are positioned at 9-kb regular intervals along the chromosome demonstrates the generation of the human DH region by gene duplication. However, these four DHsegments were rarely found in functional VH genes, predicting the presence of several other DH families. It was not until 1988 that five novel DHsegments (DM, DXP,DA,DK, and DN)were identified by nucleotide sequencing analysis of a 15-kb DNA fragment containing the D1 segment and its surrounding region (Ichihara et al., 1988a,b). Southern blot analysis revealed that each DH cluster (D1-D4) appears to contain a homologous set of six DH segments (Ichihara et al., 1988a, b). This result strongly suggests that all of the five novel DH segments also exist as a family and the human DH region comprises four tandemly arranged D clusters consisting of a set of six DH segments with the order 5’-D~-D~tR,-DXP-DA-DK-DN-3’. Among the six DH families, the DAfamily is highly homologous to the murine DFL16 segments. An additional DHfamily (DIR) with an unusual structure was identified in the 5‘ adjacent portion of DMfamily segments (Ichihara et al., 1988b). The authors pointed out the possible involvement of the DIR family in DD rearrangement because DIR has irregular spacer lengths (23 bp) of recombination signal sequence. Several other DH segments belonging to any of the six DH families were identified and localized within a cosmid contig covering the DH region, further confirming the multiplication of DH

10

FUMIHIKO MATSUDA A N D TASUKU HONIO

segments (Buluwela et al., 1988). Taken together, the estimated total number of DH segments on chromosome 14 would be about 29 and the most proximal DH segment known to date, namely D21/9 (Buluwela et al., 1988), is 20 kb upstream of the DQSUJ, cluster. However, this number is still a rough estimation. By comparing nucleotide sequences between the D4 and D1 clusters, Ichihara et al. (1988b) found rather frequent insertioddeletion of DNA in the intergenic regions and duplication of the DDl segment in the D1 cluster. In addition, some of the DH segments might be polymorphic among individuals.One such example is polymorphic deletion of the 9-kb DNA containing the D1 segment found at a high frequency (48% of alleles are D1negative) among the Japanese population (Zong et al., 1988). Because the DH segment is short, it is difficult to find the coding region by hybridization. Therefore, the exact number of the DH segments would be determined only by determination of the nucleotide sequence of the entire DH region. The complete knowledge of the germline DH segments would be of great help in the examination of the usage of DH segments because only a few of the D regions in functional VH genes correspond to known germline DH segments.

E. ORGANIZATION OF ORPHON VH AND DH LOCION CHROMOSOMES 15 AND 16 The human VHlocus has evolved through repeated duplication, deletion, and translocation, resulting in interspersion of VHsegments belonging to different families and abundant polymorphisms. Another example that testifies to the dynamic reorganization of this multigene family is interchromosomal translocation of VHand DH segments. Such translocation has also been demonstrated in the human VK locus (Lotscher et al., 1986, 1988). The human VH locus is located at the telomeric end of chromosome 14q and most of the isolated VH clones were mapped to 14q32.33. However, several VH and two DH clusters remained unmapped for some time. The first evidence that a DH segment is located on chromosome 15was obtained by in situ hybridization (Chung et al., 1984). Subsequent studies using FISH as well as humadrodent somatic cell hybrid panels identified two VH orphon loci on chromosome 15qll and chromosome 1 6 ~ 1 1 (Cherif and Berger, 1990; Matsuda et al., 1990; Nagaoka et al., 1994; Tomlinson et al., 1994). Three contigs containing human VH or DH segments were generated by isolation of YAC and cosmid clones and physically mapped. A contig of 850 kb in length, which is termed VH-F, contains 7 VH segments (2 VH1, 1VH2, and 4 VH3) within the 160-kb region in its middle part (Fig. 2). FISH studies using cosmid clones confirmed that VH-F is located on chromosome 16pll (Nagaoka et al., 1994; Tomlinson et al., 1994). All 7 VHsegments

11

ORGANIZATION OF THE H E A W - C H A I N LOCUS 100 kb 1 IJ

~

Y22 61Y3 Y3

,

I

1

1

___..--

.._.---

-_.--. m

1

1

-..--.._ --._ --. -"-F1-12P F1-14P F3-13 F3-11

;

.,, '.. ,' I

1

1

1

1

50

l

Mlu I Not I

I

.

I

(' F3-'16P F3-15P

:-.F2-26

0

I

llcl hl

I

1

1

1

1

1

I

1

100

1

1

I

P P

I

II

EC*l

11 I l l &

I

l

150

l

I

l

l

]

200 (kb)

Fic. 2. Organization of the hunian VH-F region on chromosome 16. YACs are shown by horizontal lines. Thick bars indicate the finely mapped regions by subcloning. Restriction sites are indicated by vertical lines. Modified from Nagaoka et al. (1994).

were subcloned into plasmids and sequenced. Surprisingly, 3 of the 7 VH segments (VF2-26,VF3-13, and VF3-11) are apparently functional without any structural defects in the coding region as well as recombination signal sequences. A totally different approach based on PCR using somatic cell hybrid DNAs as templates specifically amplified 16 VHsegments including the above 7 VH segments on chromosome 16, 14 of which make seven pairs of closely related VH sequences (Tomlinson et al., 1994). Because the somatic cell hybrid is unlikely to contain both copies of chromosome 16,the authors pointed out the possibility of intrachromosomal duplication of the DNA involving the 7 VH segments. Taken together, the number of VH segments on chromosome 16 would be 16 (Table I). However, the critical test for the duplication depends on their mapping on YAC clones. The other two contigs (Ds.aand Ds.b) consist of DH clusters (Fig. 3). Although obtained from two independent contigs, both were mapped to chromosome 15qll-q12 by FISH (Nagaoka et al., 1994). Restriction maps of the upstream 40-kb portion of the two D5 regions are almost identical, whereas there is a big difference between those of the downstream. Each of the Ds clusters contains five novel DH segments in the order 5'-DM5D~LR,s-DXPs-DAS-DKS-3' whereas the DNsegment, which is located at the 3'most portion of all the D1-D4 clusters (Ichihara et al., 1988a,b),is absent from both D5 clusters. Nucleotide sequencing analysis revealed that all 10 newly identified D segments have open reading frames, and nucleotide sequence homologies of the corresponding D segments between the D5-, and D5.b clusters are much higher than those between the D5 and any of

12

FUMIHIKO MATSUDA AND TASUKU HONJO

Y105 (330kb)

4/

3G (220kb)

/+

5-a A

w

D5-a

M XPK Ill I

I

101 1

I

I

543A

m

D5-b

Ill

I Ill1

v3

v54 V13C

I I

I I I I ~ U B Q I

M XPK

n

:1

I l l

I;"

IIPPP

I

n

#

l

0

l

l

l

l

50

l

l

l

I Y Hin dlll

IPI

l

l

l

I I.

I.,

, ,I I ,

l

100

l

. 13C (350 kb) l

l

150(kb)

FIG.3. Physical maps of two D5 regions. YAC clones covering the two D5regions are shown symmetrically above and below by horizontal lines with their names. D5exons are shown by closed lozenges. Restriction sites of Hind111 are indicated by vertical lines. Vertical lines with open circles represent the sites whose orders are not determined. Modified from Nagaoka et al. (1994).

the D1-D4 regions (Matsuda et al., 1990).Taken together, it is likely that the duplication of D5 clusters took place very recently after the initial translocation from chromosome 14. Another interesting finding is that the D5.bcluster is flanked by three VH segments, all of which are located 3' to the D5.bcluster (Fig. 3).Among the three VHsegments, V13C has an open reading frame and apparently is functional (Nagaoka et al., 1994). The polarity of one of them (V3) was determined and shown to have the same transcriptional orientation relative to DH(Matsuda et al., 1990; Nagaoka et al., 1994). Estimation by quantitative hybridization analysis showed that the orphon D5 locus on chromosome 15 appears to contain at least four clusters of DH segments (Nagaoka et al., 1994). PCR-based specific amplification identified eight VH segments on chromosome 15 as shown in Table I (Tomlinson et al., 1994). 111. Analysis of Human Vn Sequences

A. HUMAN V, SUBGROUPS AND FAMILIES

The human VH segments have been classified into three different subgroups based on the amino acid sequences obtained mostly from various myeloma proteins (reviewed in Kabat et al., 1991). Earlier studies of isolation and nucleotide sequencing analysis of human VH segments have

ORGANIZATION OF THE HEAW-CHAIN LOCUS

13

shown that nucleotide sequences of VH segments in the same subgroup show 70% or greater homology, whereas the homologies between different subgroups are less than 60%, suggesting that classification of VH segments into VH-I, -11, and -111 subgroups by the amino acid sequence is in agreement with that by the nucleotide sequence (Kodaira et d . , 1986). Thus, VH segments encoding the VH-I, -11, and -111 subgroups are defined as the VH1, vH2, and vH3 families, respectively. Subsequent studies by comparison of nucleotide sequences of VH segments revealed three additional VH families, namely VH4,VH5, and VH6. A group of VH segments that were tentatively classified into the vH2 family was defined as the vH4 family because nucleotide homology within the family is greater than 90%, whereas each inember of the family shows less than 65% identity to the other members of the V112family (Lee et aZ., 1987). The vH4 family members are most strongly conserved, suggesting that vH4 may have evolved most recently (Lee et d., 1987; Haino et d . , 1994; see below). The first VH segment belongmg to the vH5 family was identified in the V-D-J rearrangement of a CLL lymphoblast (Shen et al., 1987). Southern blot analysis showed that this family contains only two (and one polymorphic) members. The VH6family was first identified in the fetal VH repertoire and reported as a inember of a previously undescribed VH family (Schroeder et nl., 1987). Subsequent analysis clarified 6 consists of a single VH segment and is located 3’ most that the v ~ family within the VHlocus (Berman et d.,1988).It is noteworthy that the vH4, VH5, and VF16families have been identified according to nucleotide sequence homology. Taken together, the current status of VH family classification is that VH segments that show 80% or greater similarity to one another are considered to be in the same family, whereas VH segments that have less than 70% similarity to one another form different VH families. By this definition, three protein subgroups are subdivided into six distinct VH families. Such criterion correlates well with the experimental observation that DNA probes representative of each family detect distinct sets of hybridized fragments in human DNA. To examine the evolutionary relationship of the human VH families, a phylogenetic tree of human germline VH segments was constructed by the neighbor joining method based on the alignment of 37 apparently functional VH segments located in the 3’ 0.8-Mb region of the VH locus (Fig. 4) (Haino et al., 1994). This phylogenetic tree shows that human VH segments first diverged into two groups; subgroup I1 and an ancestor of subgroup I and 111, followed by the segregation of VH1NH5 families and V H W H W Hfamilies 6 in subgroup I and 11, respectively. Subgroup 111 consists of the largest number of VH segments but constitutes a single family vH3. Subgroup I contains a member (V7-4.lb) of a unique set of

14

FUMIHIKO MATSUDA AND TASUKU HONJO

V2-26

V6-1

Vk6

V1-2b v1-45 V1-3b

111

L

v3-33

V3-15 v3-49

vH3 V3-23 v3-53

V3-13 V3-11

I

v3-43 V3-20 . v3-35 ~~

FIG.4. Phylogenetic tree of human germline functional VIi segments. Modified from Haino et al. (1994). I, 11, and 111 indicate subgroups.

VH segments that share about 80% overall homology with the VH1 family but much less similarity to vH1 at a cIustered region between framework 2 (FR2) and FR3. This group was also identified from nucleotide sequence homology and has been proposed to be classified as the vH7 family (Schroeder et aE., 1990).According to the previous definition of the VH family, vH7 should be a subfamily of vH1 or a family captured in transition from vH1 to independence (Kirkham and Schroeder, 1994).The phylogenetic tree strongly supports this argument because segregation of V7-4. l b took place probably about the same time as when the VH3 family branched into several subfamilies. However, Southern blot and sequencing analyses

ORGANIZATION OF THE HEAW-CHAIN LOCUS

15

revealed that the VH7family is a small but discrete VH family consisting of five to eight members that are dispersed within the VH locus (van Dijk et al., 19931, indicating that the classification of VH7is practically useful.

B. FAMILY SPECIFICSTRUCTURE OF HUMAN VHSEGMENTS Studies on the structure of the human germline VH segments found several VH family specific conserved regions (Kabat et al., 1991; Tomlinson et al., 1992; Matsuda et al., 1993; Haino et al., 1994). Extensive analysis of amino acid sequences has shown that the codons 1-30 (FRl), 36-49 (FR2),and 66-94 (FR3) are highly conserved among members within the same VHsubgroup with possible amino acid insertions (Kabat et al., 1991). Among them, the regions with higher conservation within the same family are the codons 9-30 in FR1 and the codons 66-85 of FR3, which differ slightly from the regions (6-24 of FR1 and 65-85 of FR3) indicated previously (Schroeder et al., 1990). The codons 1-8, FR2 (codons 38-47), and the codons 86-92, in which the embedded heptamer recombination signal is located, are more or less universally conserved. However, some codons failed to be noticed by the analysis of protein sequences. The codons 60-65, which were previously considered as the 3' portion of complementarity determining region 2, are found to be highly conserved in a family specific manner (Matsuda et al., 1993; Haino et al., 1994). Striking family specific conservation is also observed in the 5'-flanking region that is very important for the regulation of H-chain expression. Highly conserved octamer motif and TATA box are present in all families. However, the locations of these sequences are different among families (Haino et al., 1994). A heptamer sequence with consensus CTCATGA is located 2-22 bp upstream of the octamer motif in mouse VH segments (Eaton and Calame, 1987; Siu et al., 1987) and considered to be required for the full VH promoter activity in mouse lymphoid cells (Ballard and Bothwell, 1986; Eaton and Calame, 1987). Based on this observation, Kemler et al. (1989) proposed the stepwise activation hypothesis of Vregion promoters; the activation by the oct protein occurs first in the Hchain promoter that carries the heptamer element, followed by the activation of the L-chain promoter that does not contain the heptamer motif. Interestingly, however, no heptamer element is detectable around similar places in human VH segments except in VH1family segments where heptamer is located 2 bp upstream of the octamer motif. This finding does not support the hypothesis that the heptamer element is essential for the activation of the H-chain promoter. Another interesting finding is that the 5'-flanking regions of one branch of the VH3family (V3-9, V3-20, and V343) have a common 65-bp deletion in the region at 251/315 bp upstream of the initiation codon. Only one (V3-9) among these three VH segments

16

FUMIHIKO MATSUDA AND TASUKU HONJO

was found in the functional VH regions in the previous database search (Haino et al., 1994). However, the other two VHsegments are found in the functional V-D-J rearrangement in the latest VHdatabase, which consists of a greater number of registered VH sequences, indicating that the deletion would not drastically reduce the promoter activity.

C . FUNCTIONAL A N D PSEUDO VH SEGMENTS As demonstrated in earlier studies, there are abundant VHpseudogenes. At least 36 of 82 sequenced VH segments were found to be pseudogenes (Fig. 1). Detailed features of VH pseudogenes can be found elsewhere (Tomlinson et al., 1992; Haino et al., 1994; Cook et al., 1994). Formerly, VH pseudogenes were classified into two groups; one with a few point mutations (conserved pseudogenes) and the other with rather extensive mutations (diverged pseudogenes). Conserved pseudogenes were considered to serve as a reservoir of VH segments; VH segments with a few point mutations are able to accept some correction mechanisms such as gene conversion, whereas those whose mutation rates exceed a threshold would diverge rapidly without any selection and correction. However, detailed sequence comparison showed that most of those previously classified as diverged pseudogene contained truncations of the 5‘ or 3’ half. Extensively diverged pseudogenes without truncation are surprisingly rare and V3-54P is the only such example identified to date (Haino et al., 1994).As mentioned previously, however, the presence of additional VH pseudogenes with multiple divergence still remains to be examined. Although V3-16P and V3-38P have a complete open reading frame, both of them have quite diverged heptamer sequences (TCCTGTG for V316P and TACACAG for V3-38P) from the consensus heptamer sequence CAC(A!T)GTG.Divergence of the heptamer sequence from the consensus reduces the recombination frequency. In particular, a single base substitution of the third C or the fifth G nucleotide in the heptamer is known to drastically reduce the recombination activity (Akira et al., 1987). Indeed, V3-38P is not found in the expressed VH database, and it is natural to consider that this VH segment would not or very rarely be used for the functional VH gene. The third and fifth nucleotides are conserved in V316P, but this VH segment is not found in the expressed VH database as well, and V3-16P is also likely to be a pseudogene. V1-24P and V1-58P have an abnormal splicing donor signal GC at the 5’ end of the intron and therefore are tentatively classified as pseudogenes. These two VHsegments have not been identified in full-length cDNAs registered to date. However, in vitro studies have shown that this substitution is the only one that still allows the 5’ site to be accurately cleaved, albeit more slowly than the usual GT sequence (Aebi et al., 1987;Jackson, 1991).Although V1-58P has

ORGANIZATION OF THE HEAW-CHAIN LOCUS

17

an additional one-base substitution in the nonamer signal for recombination (Kodaira et al., 1986), the nonamer signal TCAGAAACG of V1-58P is almost identical to the consensus nonamer signal TCAGAAACC of the vH1 family. Hence, the low splice efficiency might be the reason for the difficulty in the identification of cDNAs containing these VH segments. In contrast, it is not easy to define “functional” VH segments only from nucleotide sequences. Obviously, candidates should have a complete open reading frame and no defects in the recombination signal sequence (apparently functional). In the most critical sense, only VH segments translated into protein are functional because even VHsequences found in the V-D-J rearrangement might have a defect in its 5’ regulatory region or difficulty in functional VH polypeptide conformation. Among 46 apparently functional VH segments, 39 VH segments fulfill this requirement because they are known to be utilized as a part of functional H-chains including those in autoantibodies and B cell malignancies (Matsuda et al., 1993; Cook et al., 1994; F. Matsuda, unpublished data). Although sequences of V3-20 and V3-64 have not been identified in functional H-chain polypeptides, both of them had more than 99% identity with full-length cDNAs found in fetal repertoire and are therefore classified as functional, making the total number of sequenced functional VH segments 41. These 41 VH segments are classified into 6 vH1, 3 VH2,20 vH3, 9 VH4,1 VH5, 1 v ~ 6 , and 1 vH7 family segments. Only 5 VH segments (V4-28, V3-35, V1-45, V3-72, and V7-81) were not found in full-length VH cDNA databases, indicating that most of the apparently functional VHsegments on chromosome 14 participate in functional VH region formation. IV. Vn Segment Usage and Repertoire Formation

Compelling evidence indicates that VH, DH,and JH segments are not used equally in both mouse and human. Biased usage of particular VH segments during early phases of ontogeny was first reported in mouse (Yancopoulos et at., 1984; Reth et d.,1986). In human, VH5 and VH6 families are selectively expressed at 7 weeks of gestation when B-lineage development is initiated (Cuisinier et al., 1989), and a rapid expansion of VH repertoire including biased utilization of specific VH segments takes place between 8 and 15 weeks of gestation. The initial observation of the preferential usage of JH-proximalVHsegments in mouse led to the hypothesis that the proximity of VH segments to JH favors biased expression of VH segments in early stages of ontogeny (Yancopoulos et al., 1984; Reth et al., 1986). However, a number of factors might be involved in the biased VHusage that can be grouped into two; (a) those affecting the recombination frequency and (b) those affecting selection of B cells expressing those

18

FUMIHIKO MATSUDA AND TASUKU HONJO

particular VH segments. Group a includes the proximity to DH (or JH), variation in the recombination signal sequence, and locations that favor the recombinase accessibility. Group b includes self-antigens and bacterial superantigens. One study examined 14 and 10 independent VHcDNA sequences isolated from 130- and 104-day human fetal liver cDNA libraries, respectively (Schroeder et al., 1987; Schroeder and Wang, 1990), and found that 6 of them employed an identical VHsegment of the vH3 family (56P1 or V330). V3-15 was second overrepresented (3 of 24 VH cDNAs) and V6-1, V323, V3-53, V4-59, and V1-69 were picked up twice. Cuisinier et al. (1993) performed a similar examination using cDNA libraries from the fetal livers of 8 and 13 weeks. Eleven VHcDNAs employed the V3-30 segment among 43 cDNAs whose germline counterpart was determined, further confirming the overrepresentation of V3-30. V6-1, the single member of the JH-proximal VH6family, appeared at very high frequency (11 of 43 VH cDNAs), and V3-23 was also used often (4 of 43 VH cDNAs). These studies clearly indicate the biased usage of certain VH segments in the early human repertoire. However, the frequency of the other VH segments appears to still be controversial and far from a consensus. The V3-15 segment, which is the second-most used VHsegment in the former study, was found only once in the latter study, and V3-53, V4-59, and V1-69 were not found in either of the 8-week and 13-week libraries. Conversely, V4-4, V3-7N1-8, and V3-9N4-39, none of which appeared in the former study, were isolated four, three, and two times, respectively, in the latter study. When these VHsegments were localized on the complete physical map of the human VHlocus, little association was observed between the biased usage of these VH segments and their JH proximity. The V3-30 segment, which is most often found in the fetal repertoire, is located approximately 470 kb upstream of the JH cluster (Fig. 1).Similarly, V3-23 is localized 395 kb upstream of the JH cluster. One exception is the V6-1 segment that is located 3’ most of the locus. The other VHsegments that were found at least twice in either study are located in a scattered manner within the region between 160 (V4-4) and 900 kb (Vl-69) upstream of the JH cluster (Fig. 1). Furthermore, none of the JH-proximalfunctional VH segments, including V1-2, V1-3, V2-5, and V3-11, were found repeatedly in either examination. The results indicate that VH segments preferentially used in early stages of ontogeny do not necessarily cluster in the JH-proximalregion. Computer-assisted homology analysis revealed that 17 of 41 functional VH segments have greater than 99%homology to VHregions of autoantibodies although it is premature to conclude that only limited VH segments could be used for autoantibodies. The V3-30 segment, which is most frequently used in the fetal repertoire, is widely used for autoantibodies with

ORGANIZATION OF THE HEAW-CHAIN LOCUS

19

diverse specificities including rheumatoid factors, anti-Sm antibodies, and anti-DNA antibodies (Matsuda et al., 1993). Similarly, the V3-23 segment is the germline counterpart of a number of autoantibodies (Matsuda et al., 1993). Notably, this VH segment is most commonly utilized for antithyrotropin receptor autoantibodies in patients with Graves’ disease (Shin et al., 1994). The other eight VH segments that are found in the VH region of a variety of autoantibodies are V3-7 (anti-DNA, anti-T cell receptor pchain, anti-thyrotropin receptor), V3-15 (anti-Sm, rheumatoid factor, antithyrotropin receptor), V4-39 (rheumatoid factor, anti-B cell, anti-platelet), V3-48 (rheumatoid factor), V5-51 (anti-DNA, anti-thyrotropin receptor, anti-actin, anti-acetylcholine receptor), V4-59 (anti-DNA, anti-D antigen, anti-thyrotropin receptor), V4-61 (anti-B cell, anti-platelet), and V1-69 (anti-DNA, rheumatoid factor). Five of them (V3-7, V3-15, V4-39, V4-59, and V1-69) are found more than once in the fetal repertoire. Such correlation between autoantibody VH and early repertoire VH may indicate that the preferred usage of VH segments in early stages of ontogeny is due to positive selection by self-antigens rather than by JH-proximallocation of the VH segments. In any case, the complete physical map of the human VH locus has contributed to the identification of germline origins of autoantibodies. Comparison of the VHusage among polymorphic individuals may also shed light on mechanisms for biased VH usage. V. Evolution of the Human V, Loci

A. STRUCTURAL ANALYSIS OF THE HUMAN VH LOCUSUSING INTERGENIC PROBES Evolution of multigene families, including Ig, has been driven by dynamic reorganization of the gene locus including duplication, deletion, and translocation. Comparative analysis of VHloci among different vertebrates from shark to man revealed dramatic difference of H-chain gene organization among species, but all of them carry multiple VH segments. Even among mammals, between man and mouse for instance, the number and organization of VH, DH, and JH segments are totally different although the relative orders of these segments are essentially the same between two species. From these observations one can easily imagine that duplication of VHsegments must have started quite a long time ago and that reorganization of the VI, locus is still underway on a variety of scales. Recent translocation of VHand DHsegments to chromosomes 15 and 16 is further evidence for dynamic reshuffling of the VHlocus (see above). Nevertheless, examination of the restriction map of the VH locus identified only two small internal duplications, and no large-scale duplication as demonstrated in the human

20

FUMIHIKO MATSUDA AND TASUKU HONJO

VK locus (Straubinger et al., 1988) was found. It is not easy to follow an evolutionary trail of this locus only by comparison of VHsequences because the human VHsegments are reasonably homologous to each other; besides, frequent gene conversion in VHsegments often decreases the nucleotide homology between duplication partners. Matsumura et al. (1994) looked for genetic traits that may allow the tracing of the steps of the reorganization of DNA within the human VH loci using intergenic nonrepetitive probes. Fourteen nonrepetitive probes were isolated from the intergenic regions of the VHlocus. These probes can detect two to seven cross-hybridizing bands within the 3’ 0.8-Mb region of the VHlocus on chromosome 14, and most of them also detected a few bands on chromosomes 15 or 16, further confirming the recent translocation of the orphon loci. Identification of exact locations of crosshybridized bands made it feasible to trace genetic events such as duplication and translocation. Such studies identified several pairs of regions that are hybridized by an identical set of nonrepetitive probes (Fig. 5). Each of five different sets of probes hybridized as a cluster to two or three regons in a dispersed manner. Notably, none of these pairs of translocation and duplication were inverted, excluding the possibility of gross inversion within the locus. The longer set of the two obvious tandem duplications, namely V3-33-V4-31N3-3O-V4-28, is also associated with duplication of nonrepetitive probes (Matsumura et al., 1994). In most of the cases, distantly located VH segments flanked by homologous clusters have much higher homology (90 -93%) than the average (73 -88%) within the same family. Such sets of VH segments are V3-9N3-2ON3-43, V3-7N3-21N3-48, and V3-13N3-47P. A close relationship among these VHsegments was further confirmed by the phylogenetic tree of 37 functional VHsegments in which each of these three VH sets constitutes a clustered branch (Fig. 4). The divergence times of these duplicated VHsegments were calculated using the rate of the silent substitution in the coding region as described (Miyata et al., 1980; Hayashida and Miyata, 1983). The estimated times of the duplication of V3-9N3-2ON3-43, V3-7N3-2 1N3-48, and V3-13N3-47P are 30-38, 22-40, and 55 million years (Myr) ago, respectively (Matsumura et al., 1994). Hence, these duplications must have taken place after mamrnalian radiation (75 Myr ago), but much earlier than the two obvious duplications that occurred 9 Myr ago (Kodaira et al., 1986; Haino et al., 1994). Dispersed appearance of these clusters and close relationship among flanking VHsegments provide further evidence for recent duplication, deletion, and translocation of VH-containingDNA fragments.

B. DISTRIBUTION OF REPETITIVESEQUENCES IN THE HUMAN vHLocus The human genome contains a large fraction of interspersed repetitive sequences such as short interspersed elements ( A h repeats) and long

ORGANIZATION OF THE HEAVY-CHAIN LOCUS

21

FIG.5. Localization of DNA fragments homologous to nonrepetitive probes and distribution of AZu and L1 repetitive sequences in the 730-kb region of the VH locus. Solid boxes represent locations of restriction fragments that were hybridized with probes indicated on the right. Clusters of fragments hybridized with an identical set of probes are enclosed with the same symbols. Closely related VH segments are also marked with the same symbols. Two internal duplications are indicated by brackets above the corresponding VHsegments.

22

FUMIHIKO MATSUDA AND TASUKU HONJO

interspersed elements (L1 repeats). Studies have demonstrated that some of the repetitive sequences could be hot spots for recombination in the genome (Hyrien et al., 1987; Devlin et al., 1990). Therefore, the frequent reorganization of the human VH locus may be associated with the content and distribution of these repetitive sequences. The total numbers of AZu and L1 repeats in the human genome have been estimated to be 3 to 5 X lo5 and lo4to lo5, respectively. If the distribution of these repeats is also random within the VH locus, the numbers of Alu and L1 would be 70-110 and 2-20, respectively, within the 730-kb region between V6-1 and V3-64. Southern blot analysis using these elements as probes detected 44 Alupositive and 11 L1-positive regions in the 730-kb DNA (Fig. 5), both of which do not greatly exceed the average content of these elements which is expected by random distribution. The authors could not exclude the possibility of underestimation due to the multiple repetitive elements within a single restriction fragment. Interestingly, no L1-positive fragments were found in the distal 200-kb region. It must be noted that flanking regions of closely related VHsegments do not have similar distribution ofAZu and L1 repeats and that these VH surrounding regions were not necessarily abundant in repetitive sequences (Fig. 5). The results did not agree with the speculation that the homologous recombination mediated by repetitive sequences might be the main driving force of frequent reorganization of the VH locus. Most of the A h repeats were reported to have amplified within the past 60 million years (Shen et al., 1991). In particular, members of HS subfamily have appeared after the divergence of the chimpanzee and man (within the past 5 million years) (Shen et al., 1991). Because the reorganizations of VH segments were estimated to have taken place 55-22 million years ago (Matsumura et al., 1994), the most likely explanation is that duplications and translocations of VHsegments were followed by recent frequent transposition of many A h repetitive elements into random positions. C. EVOLUTION OF ORPHON VHAND DH LOCI Study on cosmid and YAC clones derived from these orphon loci revealed several striking findings. First, about 40% of VH segments in both loci ( 3 of 7 VH on chromosome 16 and 1of 3 V H on chromosome 15)are apparently functional (Matsuda et al., 1990; Nagaoka et al., 1994). Ten VH segments were shown to be apparently functional among 24 VHsegments specifically amplified by PCR from DNAs of somatic cell hybrid carrying a single human chromosome 15 or 16 (Tomlinson et al., 1994). Several possible explanations for conservation of apparently functional VH orphans include (1)translocation took place rather recently, (2) they are still under selective

23

ORGANIZATION OF THE HEAVY-CHAIN LOCUS

constraint through their usage by transchromosomal rearrangement as happened in the human y and S T cell receptor loci (Tycko et al., 1989), or (3)they might have been kept conserved by some correction mechanisms such as gene conversion (discussed below). Second, putative origins for the orphon VH segments on chromosomes 15 and 16 were found in the 0.43-0.25 Mb JH-proximalVHregion on chromosome 14 (Fig. 6). Comparison of the corresponding VH segments revealed that all of the 7 orphon VH segments have more than 93% identity with the corresponding VH segments on chromosome 14. A most remarkable homology was found between two truncated pseudogenes, VF1-12P and V1-12P, in which the homology extends into the region 3' to the truncation site. Divergence time of the VH-F region from chromosome 14 was calculated by the synonymous site substitution in the coding region between corresponding VH segments. The time of the translocation is estimated to be, at the earliest, 20 Myr ago (Nagaoka et al., 1994). In contrast, homology search analysis could not identify obvious counterparts of the D5 clusters in the DH segments on chromosome 14. The homology between the orphon VH segments on chromosome 15 and the corresponding VHsegments on chromosome 14 is less remarkable except

D5-b

2-26

1-24P

3-22P 3-21

Chr.14

"

I I

VH-F F2-26 I

0

I

I

F3-16P F3-15P I

I

I

50

I

I

I

I

I

I

100

W

F1-14P F1-12P F3-13 F3-11 I

I

I

I

150

I

I

I

I

I

200(kb)

FIG. 6. Comparison of VH segments in the VH-F and D5.b regions with their putative counterparts on chromosome 14.Corresponding VHsegments are indicated with the percentage of homologies of coding and intron sequences by bold lines. Neighboring VH segments of V1-18 were compared with V3 or V13C on the D5.b region and shown with their percentage of homologies by dashed lines. The regions detected by pHael.2 or pAfa550 probes are indicated by bars with the name of the probes. Modified from Nagaoka et al. (1994).

24

FUMIHIKO MATSUDA AND TASUKU HONJO

for one pair. V54 displayed 94.7% identity to V1-18 located within the putative ancestor of the VH-F region on chromosome 14, whereas V3 and V13C could not find any corresponding VH segments with more than 90% homology. However, the 94.7% identity between V54 and V1-18 is significant because no pairs of vH1 segments on chromosome 14 share more than 90% nucleotide homology. This levels of homology is similar to that between the VH-F segments and their putative counterparts (Fig. 6). The date of the segregation of V54 and Vl-18 was estimated to be approximately 13 Myr ago, Furthermore, the region that was detected by two DNA probes (pHae1.2 and pAfa550) flanking the VM segment was found only in the proximity of V1-18 within the 0.8-Mb VH region on chromosome 14 (Fig. 6). These results suggest that a DNA fragment of more than 100 kb might have been translocated simultaneously to chromosomes 15 and 16 approximately 20 M y ago. AND GENECONVERSION D, PSEUDOGENES As shown previously, there are a considerable number (40%) of VH pseudogenes within the VH locus on chromosome 14 and more than 10 orphon VH segments on other chromosomes (Table I). Strikingly, most of them are conserved pseudogenes, and even the VH half of truncated pseudogenes is structurally highly conserved. The presence of these VH pseudogenes raises the question of whether they are of any functional significance. Because 40% of orphon VH segments are apparently functional, they can theoretically recombine with D-J rearrangements on chromosome 14 through interchromosomal recombination, Recombination between a VH-DH fusion on chromosome 15and a JH segment on chromosome 14 is also conceivable because a fusion of VH-DH was found in a human B cell line (Shin et aZ., 1993b) and in B lymphocytes of transgenic mice carrying IgH mini locus (Tuaillon et al., 1995). In addition, germline transcripts of orphon VHhave been identified in human fetal liver (Cuisinier et aZ., 1993),suggesting that orphon VHloci might be targets of recombinase. Unfortunately, however, none of their sequences have been found in any published V-D-J rearrangements and interchromosomal recombination of Ig loci remains to be tested. Conserved pseudogenes already have been shown to serve as sequence donors for gene conversion in other species. Somatic gene conversion generates the V-region repertoire in chicken (Raynaud et al., 1987; Tompson and Neiman, 1987) and rabbit (Becker and Knight, 1990) but not in mouse and man. One attempt to obtain direct evidence for germline gene conversion is based on the theory of molecular evolution (Haino et aE.,1994).It is widely accepted that the introns and the synonymous positions of the coding region evolve at high and remarkably similar rates in different genes, and

ORGANIZATION OF THE HEAW-CHAIN LOCUS

25

base substitutions are accumulated at approximately constant rates with respect to the geological time (Miyata et aZ., 1980; Hayashida and Miyata, 1983).Hence, the recipients ofgene conversion would be found bycomparing the substitution rates in the intron and the synonymous position of pairs of VH segments; a clear difference in substitution rates of the two portions in a given pair O f V H segments suggests some recombination events. V3-62P and V3-60P, one of the recently duplicated VH segment pairs that were chosen for the analysis, displayed a significant difference in the substitution rates in the intron and the synonymous position (Haino et al., 1994). Nucleotide sequence homology of V3-62P and V3-60P is 94%, which is much higher than the homology of other known VHsegments of the vH3 family (Kodaira et al., 1986). In addition, these two VHsegments have the same mutation in the heptamer signal sequence and the same 3bp deletion in the 23-bp spacer of the recombination signal, suggesting that these deleterious mutations were followed by the internal duplication. When V3-62P and V3-60P are compared using the above evolutionary molecular clock, the nucleotide difference (0.1637 ? 0.0461) at the intron (Kci)was significantly greater than that (0.0821 2 0.0339) at the synonymous positions of the coding region (K's), indicating possible segmental change in the intron in either V3-60Por V3-62P (Fig. 7). Sequences ofthese two VHsegments were compared with those of the other vH3 members to search for the putative donor of the segmental transfer. As a result, V343 and V3-62P were found to have significantly smaller P i (0.0820 2 0.0280) than K's (0.4165 2 0.0761) (Fig. 7). Nucleotide sequences of the leader, intron, and FRl of V3-43 and V3-62P are highly homologous, whereas their 3' halves are diverged. Such unusually high homology of the 5'-half region of V3-43 and V3-62P is most likely explained by the segmental transfer of V3-43 sequence to V3-62Pin a unidirectional manner. This example supports the hypothesis that gene conversion contributes to the maintenance of the pseudogene structure. Haino et al. (1994) cannot completely rule out the possibility of double unequal crossing over between V3-43 and V3-60P and subsequent loss of a modified allele of V3-43. The same method was applied to VH segments belonging to the vH4 family, which is the most conserved (>go%)and richest in functional VH segments among seven human VHfamilies (Lee et al., 1987). Rather frequent unidirectional correction was observed between vH4 segments (Fig. 7), demonstrating that the vH4 family members evolved by recent duplication, followed by gene conversion. It must be noted that V4-55P served as a donor of two functional VH segments, V4-4b and V4-28. This might indicate that the high percentage of pseudogenes should also contribute to the generation of the germline VH repertoire.

26

FUMIHIKO' MATSUDA AND TASUKU HONJO

Ki

K:

v3-43

0.4165 t0.0761 0.0821 t0.0339

Internal

aiielic counterpart

V4-61

-

['&'

V4-4b v4-55P

duplication

-......,....,.,,~..,,,...,,~

0.0248 t0.0175 0.051 1 i0.0251 0.0248 *0.0175 0.1781 *0.0466 0.1626 10.0445 0.0246 "0.0174

V4-31 Ludy

lntron

53

32

t 23 33

,

~

FRl

CDRl FR2 .CDR2

0.0362 i0.0206 0.0274 iO.0192 0.1195 i0.0375 0.0491 i0.0242 0.0366 10.0269 0.1011 k0.0347 FR3

FIG.7. Schematic demonstration of gene conversion. Boxes and horizontal lines indicate coding regions and introns of VH segments, respectively. Arrows indicate direction of sequence transfer. The values of K'i and K'S of the V, segments are calculated for introns and synonymous positions of codons 4/92, respectively. Subregions of VH segments are shown below. Modified from Haino et al. (1994).

ACKNOWLEDGMENTS This work was supported in part by grants from Creative Basic Research (Human Genome Program) of the Ministry of Education, Science, Sports and Culture of Japan and from the Science and Technology Agency of Japan.

REFERENCES Aebi, M., Hornig, H., and Weissmann, C. (1987). Cell 50, 237-246. Akira, S., Okazaki, K., and Sakano, H. (1987). Science 238, 1134-1138. Ballard, D.W., and Bothwell, A. (1986). Proc. Ned. Acad. Sci. USA 83,9626-9630. Becker, R. S., and Knight, K. L. (1990). Cell 63,987-997. Berrnan, J. E., Mellis, S. J., Pollock, R., Smith, C. L., Suh, H., Heinke, B., Kowal, C., Surti, U., Chess, L., Cantor, C. R., and Alt, F. W. (1988). EMBO 1. 7, 727-738. Buluwela, L., and Rabbitts, T. H. (1988). Eur. J. Immunol. 18, 1843-1845. Buluwela, L., Albertson, D. G., Shemngton, P., Rabbitts, P. H., Spurr, N., and Rabbitts, T. H. (1988). EMBO 1. 7, 2003-2010. Burke. D. T., Carle, G . F., and Olson, M. V. (1987). Science 236, 806-812. Cherif, D.,and Berger, R. (1990). Genes Chromosomes Cancer 2, 103-108. Chung, J. H., Siebenlist, U., Morton, C. C., and Leder, P. (1984). Fed. Proc. 43, 1486. Cook, G., Tomlinson, I. M., Walter, G., Riethman, H., Carter, N. P., Buluwela, L., Winter, G., and Rabbitts, T. H. (1994). Nature Genet. 7, 162-168.

ORGANIZATION OF THE HEAVY-CHAIN LOCUS

27

Croce, C. M. , Sander, M., Martins, J., Cicurel, L., D’Ancona, C . C., Dolby, T. W., and Koprowski, H. (1979). Proc. Natl. Acad. Sci. USA 76, 3416-3419. Cuisinier, A. M., Guigou, V., Boubli, L., Fougereau, M., and Tonnelle, C. (1989). Scand. J. Immunol. 30,493-497. Cuisinier, A. M., Gauthier, L., Boubli, L., Fougereau, M., and Tonnelle, C. (1993). Eur. J. Immunol. 23, 110-118. Devlin, R. H., Deeb, S . , Brunzell, J., and Hayden, M. R. (1990). Am. J . Hum. Genet. 46, 112-119. Eaton, S . , and Calame, K. (1987). Proc. Nut!. Acad. Sci. USA 84, 7634-7638. Foote, J., and Winter, G. (1992).J. Mol. Biol. 224, 487-499. Givol, D., Zakut, R., Effron, K., Rechavi, C., Ram, D., and Cohen, J. B. (1981). Nature 292, 426-430. Green, L. L., Hardy, M. C., Maynard-Currie, C. E., Tsuda, H., Louie, D. M., Mendez, M. J., Abderrahim, H., Noguchi, M., Smith, D. H., Zeng, Y., David, N. E., Sasai, H., Garza, D., Brenner, D. G., Hales, J. F., McGuiness, R. P., Capon, D. J., Klapholz, S., and Jakobovits, A. (1994). Nature Genet. 7, 13-21. Griffiths, A. D., Williams, S. C., Hartley, O., Tomlinson, I. M., Waterhouse, P., Crosby, W. L., Kontermann, R. E., Jones, P. T., Low, N. M., Allison, T. J., Prospero, T. D., Hoogenboom, H. R., Nissim, A., Cox, J. P. L., Harrison, J. L., Zaccolo, M., Gherardi, E., and Winter, G. (1994). E M B O J 13, 3245-3260. Haino, M., Hayashida, H., Miyata, T., Shin, E. K., Matsuda, F., Nagaoka, H., Matsumura, R., Taka-ishi, S., Fukita, Y., Fujikura, J., and Honjo, T. (1994).J.Biol. Chem. 269,2619-2626. Hashimoto, L. L.. Walter, M. A,, Cox, D. W., and Ebers, G. C. (1993).J. Neumimmunol. 44,77-83. Hayashida, H., and Miyata, T. (1983). Proc. Natl. Acad. Sci. USA 80, 2671-2675. Hofker, M. H., Walter, M. A., and Cox, D. W. (1989).Proc. Natl. Acad. Sci. USA 86,55675571. Honjo, T., and Habu, S. (1985). Annu. Reo. Biochem. 152, 803-830. Hyrien, O., Debatisse, M., Buttin, G., and Robert de Saint Vincent, B. (1987). EMBO J. 6, 2401-2408. Ichihara, Y., Abe, M., Yasui, H., Matsuoka, H., and Kurosawa, Y. (1988a).Eur. J. Immunol. 18, 649-652. Ichihara, Y., Matsuoka, H., and Kurosawa, Y. (198813).EMBO J. 7, 4141-4150. Jackson, I. J. (1991). Nucleic Acids Rex 19, 3795-3798. Kabat, E. A,, Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. (1991). “Sequences of Proteins of Immunological Interest,” 5th ed. NIH publications, Washington DC. Kemler, I., Schreiber, E., Miiller, M. M., Matthias, P., and Schaffner, W. (1989) EMBO J. 8, 2001-2008. Kirkham, P. M., and Schroeder, H. W., Jr. (1994). Semin. Immunol. 6,347-360. Kirsch, I. R., Morton, C. C., Nakahara, K., and Leder, P. (1982). Science 216, 301-303. Kodaira, M., Kinashi, T., Umemura, I., Matsuda, F., Noma, T., Ono, Y., and Honjo, T. (1986).J. Mol. Biol. 190, 529-541. Lee, K. H., Matsuda, F., Kinashi, T., Kodaira, M., and Honjo, T. (1987). J . Mol. B i d . 195, 761-768. Lorenz, W., Straubinger, B., and Zachau, H. G. (1987). Nucleic Acids Res. 15,9667-9676. Liitscher, E,, Grzeschik, K.-H., Bauer, H. G., Pohlenz, H.-D., Straubinger, B., and Zachau, H. G. (1986). Nature 320, 456-458. Lijtscher, E., Zimmer, F.-J., Klopstock, T., Grzeschik, K.-H., Jaenichen, R., Straubinger, B., and Zachau, H. G. (1988). Gene 69, 215-223.

28

FUMIHIKO MATSUDA AND TASUKU HONJO

Matsuda, F., Lee, K. H., Nakai, S., Sato, T., Kodaira, M., Zong, S. Q., Ohno, H., Fukuhara, S., and Honjo, T. (1988). EMBO J. 7, 1047-1051. Matsuda, F., Shin, E. K., Hirahayashi, Y., Nagaoka, H., Yoshida, M. C., Zong, S. Q., and Honjo, T. (1990). EMBO J. 9, 2501-2506. Matsuda, F., Shin, E. K., Nagaoka, H., Matsumura, R., Haino, M., Fukita, Y.,Taka-ishi, S., Imai, T., Riley, J. H., Anand, R., Soeda, E., and Honjo, T. (1993).Nature Genet. 3,88-94. Matsuda, F., and Honjo, T. (1993). In “Methods in Molecular Genetics” (K. Adolph, Ed.), Vol. 2, pp. 226-245. Academic Press, Orlando, FL. Matsumura, R., Matsuda, F., Nagaoka, H., Shin, E. K., Fukita, Y., Haino, M ., Fujikura, J., and Honjo, T. (1994).J. Immunol. 152, 660-666. Miyata, T.,Yasunaga, T., and Nishida, T. (1980).Proc. Natl. Acad. Sci. USA 77,7328-7332. Nagaoka, H., Ozawa, K., Matsuda, F., Hayashida, H., Matsumura, R., Haino, M., Shin, E. K., Fukita, Y., Imai, T., Anand, R., Yokoyama, K., Eki, T., Soeda, E. and Honjo, T. (1994). Genomics 22, 189-197. Ravetch, J. V., Siebenlist, U., Korsmeyer, S., Waldmann, T., and Leder, P. (1981). Cell 27,583-591. Raynaud, C.-A,, Anques, V., Grimal, H., and Weill, J, C. (1987). Cell 48,379-388. Rechavi, G., Bientz, B., Ram, D., Ben-Neriah, Y., Cohen, J. B., Zakut, R., and Givol, D. (1982). Proc. Natl. Acad. Sci. USA 79,4405-4409. Reth, M., Jackson, N., and Alt, F. W. (1986). EMBO J. 5, 2131-2138. Sasso, E. H., Willems van Dijk, K., Bull, A. P., and Milner, E. C. B. (1993). Proc. Nutl. Acad. Sci. USA 89, 10430-10434. Sato, T., Matsuda, F., Lee, K. H., Shin, E. K., and Honjo, T. (1988). Biochem. Biophys. Res. Commun. 154,265-271. Schroeder, H. W., Jr., Hillson, J. L., and Perlmutter, R. M. (1987). Science 238,791-793. Schroeder, H. W., Jr., Walter, M. A., Hofker, M. H., Ebens, A., Willem van Dijk, K., Liao, L. C., Cox, D. W., Milner, E. C. B., and Perlmutter, R. M. (1988).Proc. Nutl. Acad. Sci. USA 85,8196-8200. Schroeder, H. W., Jr., Hillson, J. L., and Perlmutter, R. M. (1990).lnt, Immunol. 20,41-50. Schroeder, H. W., Jr., and Wang, J. Y. (1990). Proc. Nutl. Acad. Sci. USA 87, 6146-6150. Shen, A., Humphries, C., Tucker, P., and Blattner, F. (1987). Proc. Natl. Acad. Sci. USA 84, 8563-8567. Shen, M. R., Batzer, M. A., and Deininger, P. L. (1991).]. Mol. Euol. 33, 311-320. Shin, E. K., Matsuda, F., Nagaoka, H., Fukita, Y., Imai, T., Yokoyama, K., Soeda, E., and Honjo, T. (1991). EMBO J. 10, 3641-3645. Shin, E. K., Matsuda, F., Ozaki, S., Kumagai, S, Olerup, O., Strom, H., Melchars, I., and Honjo, T. (19934. lmmunogenetics 38,304-306. Shin, E. K., Matsuda, F., Fujikura, J., Akamizu, T., Sugawa, H., Mori, T., and Honjo, T. (1993b). Eur. J. Immunol. 23,2365-2367. Shin, E. K., Akamizu, T., Matsuda, F., Sugawa, H., Fujikura, J., Mori, T., and Honjo, T. (1994).J. Immunol. 152, 1485-1492. Siebenlist, U., Ravetch, J. V., Korsmeyer, S., Waldmann, T., and Leder, P. (1981). Nature 294, 631-635. Siu, G., Springer, E. A., Huang, H. V., Hood, L. E., and Crews, S. T. (1987).J. lmmunol. 138,4466-4471. Straubinger, B., Huber, E., Lorenz, W., Osterholzer, E., Pargent, W., Pech, M., Pohlenz, H.-D., Zimmer, F.-J., and Zachau, H. G. (1988).J. Mol. Biol. 199, 23-34. Taylor, L. D., Carmack, C. E., Huszar, D., Higgins, K. M., Mashayekh, R., Sequar, G., Schramm, S. R., Kuo, C. C., O’Donnell, S. L., Kay, R. M., Woodhouse, C. S., and Lonberg, N. (1994). lnt. Immunol. 6,579-591.

ORGANIZATION OF THE HEAVY-CHAIN LOCUS

29

Tomlinson, I. M., Walter, G., Marks, J. D., Llewelyn, M. B., and Winter, G. (1992).]. Mol. Biol. 227, 776-798. Tomlinson, I. M., Cook, G. P., Carter, N. P., Elaswarapu, R., Smith, S., Walter, G., Buluwela., L., Rabbitts, T. H., and Winter, G. (1994). Hum. Mol. Genet. 3, 856-860. Tompson, C. B., and Neiman, P. E. (1987). Cell 48, 369-378. Tonegawa, S. (1983). Nature 302, 575-581. Tuaillon, N., Miller, A. B., Tucker, P. W., and Capra, J. D. (1995). In “The 9th International Congress of Immunology Abstract Book,” p. 501. The American Association of Immunologists and the International Union of Immunological Societies. Tycko, B., Palmer, J. D., and Sklar, J. (1989). Science 245, 1242-1246. van Dijk, K. W., Mortari, F., Kirkham, P. M., Schroeder, H. W., and Milner, E. C. B. (1993). Eur. J. lmmunol. 23, 832-839. Walter, G., Tomlinson, I. M., Cook, G. P., Winter, G., Rabbitts, T. H., and Dear, P. H. (1993). Nucleic Acids Res. 21, 4524-4529. Walter, M. A., Surti, U., Hofker, M. H., and Cox, D. W. (1990). EMBO J. 9,3303-3313. Walter, M. A., Dosch, H. M., and Cox, D. W. (1991a).J. Exp. Med. 174, 335-349. Walter, M. A., Gibson, W. T., Ebers, G. C., and Cox, D. W. (1991b). J. Clin. Invest. 87, 1266-1273. Yancopoulos, G . D., Desiderio, S. V., Pasking, M., Kearney, J. F., Baltimore, D., and Alt, F. W. (1984). Nature 311, 727-733. Yang, P.-H., Olsen, N. J., Siminovitch, K. A,, Olee, T., Kozin, F., Carson, D. A,, and Chen, P. P. (1990). Proc. Nutl. Acud. Sci. USA 88, 7907-7911. Zong, S. Q., Nakai, S . , Matsuda, F., Lee, K. H., and Honjo, T. (1988). lmmunol. Lett. 17, 329-334. This article was accepted for publication on 19 October 1995.

This Page Intentionally Left Blank

ADVANCES IN IMMUNOLOGY. VOL 62

Analysis of Gene Function in Lymphocytes by RAG-2-Deficient Blastocyst Complementation JIANZHU CHEN Depaflnmniof Biology, hsachuseth lnsriruh of lechdogy, Cambridge, Mouochusetis 02 139

1. Introduction

Gene-targeted mutation has greatly facilitated the analysis of gene function in normal developmental processes (1-3). By this approach, a specific gene is first disrupted (targeted) in one of the two alleles through homologous recombination in embryonic stein (ES) cells. The heterozygous mutant ES cells are then injected into blastocytes usually from C57BL/6, followed by the implantation of the injected blastocytes into pseudopregnant females. The chimeric mice generated are then bred with C57BL/6 for the transmission of the mutation. Interbreeding of heterozygous mutant mice yields homozygous mutant mice, in which the effect of the mutation is assayed. Because the immune system is disposable when mice are kept in a pathogen-free environment, the germline mutational approach has been widely used to assay gene function in lymphocyte development (4,5). Despite its effectiveness, the approach is limited by practical considerations such as time and expense. In addition, many genes that are involved in lymphocyte development are also required for early embryogenesis (6-9). Homozygous mutation of these genes is embryonic lethal and therefore their function in lymphocyte development cannot be analyzed by the germline mutational approach. For these reasons, the recombination activating gene-2 (RAG-2)-deficient blastocyst complementation system was developed to assay gene function in lymphocytes (10, ll),providing an efficient complementary assay to the germline mutational approach. II. RAG-2-Deficient Biastocyst Complementation

RAG-2 is specifically expressed in precursor lymphocytes for the assembly of T cell receptor (TCR) and immunoglobulin (Ig) variable region genes (12, 13).Inactivation of RAG-2 in mice results in a complete block of V(D)J recombination and therefore complete block of lymphocyte development (14). Mutant mice have no mature T or B cells in the periphery. T cell differentiation is arrested at the CD4-CD8- (double negative, DN) stage in thymus and B cell differentiation at the B220tCD43' pro-B cell 31 Copflght Q 1996 by Academic Press. Inc. AII rights of reproduction in any form reserved.

32

JIANZHU CHEN

stage in the bone marrow. Otherwise, RAG-2-deficient mice are viable and fertile, and blastocysts from them are amenable for further manipulation. Implantation of blastocysts from RAG-2-deficient mice into pseudopregnant females simply gives rise to RAG-2-deficient pups (Fig. 1) (10). However, injection of normal ES cells into RAG-2-deficient blastocysts followed by implantation into pseudopregnant females yields somatic chimeras composed of both the blastocyst-derived cells and ES cell-derived cells. Because RAG-2 deficiency blocks lymphocyte development at the progenitor stage, precursor lymphocytes (DP thymocytes and pre-B cells) and mature T and B cells in the generated chimeras must all be derived from the injected ES cells. Thus, if homozygous mutant ES cells are used to generate chimeras in the RAG-2-deficient blastocysts, depending on the gene mutated, the generated chimeras may or may not have any ES cell-derived precursor and/or mature lymphocytes (Fig. 1) (11).The presence of precursor and mature lymphocytesin the chimeras will indicate that the mutated gene is not absolutely required for lymphocyte differentiation. Functional analysis of mature lymphocytes will then determine whether the mutated gene is required for lymphocyte function. The absence of precursor and/or mature lymphocytes in the chimeras will suggest that the targeted gene is probably required for lymphocyte differentiation. Thus, RAG-2-deficient blastocyst complementation assays for gene function in lymphocyte development in vivo. For example, the effect of Ig JH gene segment deletion on lymphocyte development was assayed by RAG-2-deficient blastocyst complementation

RAG-2-1- blastocyst

Pseudopregnantfemale

injectlonof normal ES cells

Phenotype

somatic chimera

-&iniection of mutant ES cells

ES-derivedT and B cells

somatic chimera

FIG. 1. The RAG-2-deficient blastocyst complementation system (from Zhang et d., 1995).

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

33

(10). Implantation of RAG-2-deficient blastocysts into pseudopregnant females simply yielded RAG-2-deficient mice. In contrast to normal mice, RAG-2-deficient mice have no mature T and B cells in the spleen but rather have DN cells in thymus and pro-B cells in bone marrow (Figs. 2A and 2B). Injection of normal or heterozygous mutant ( JH+’-) ES cells into RAG-2-deficient blastocysts generated somatic chimeras that had mature T and B cells in the spleen, DP and SP thymocytes, and pre-B cells in the bone marrow (Figs. 2C and 2D), indicating the restoration of lymphoid compartment. However, chimeras generated by injection of homozygous

FIG.2. Analysis of Chen et al., 1993).

JH

deletion by RAG-2-deficient blastocyst complementation (from

34

JIANZHU CHEN

mutant (JH-’-) ES cells had only mature T cells in the spleen and DP and SP thymocytes (Fig. 2E). There were no mature B cells in spleen and B lineage cells in the bone marrow were at the pro-B cell stage (Fig. 2E). Thus, deletion of JH gene segments blocks B cell differentiation. To confirm the observation, homozygous mutant ES cells were transfected with a functionally assembled VDJ-Cp expression vector. ES transfectants were then injected into RAG-2-deficient blastocysts. The generated chimeras had restored both T and B cell development (Fig. 2F). As expected, the generated B cells expressed only IgM compared to normal B cells, which expressed both IgM and IgD, because the transfected vector contained only the C p gene. Together, these findings suggest that functional rearrangement and expression of Ig p heavy-chain gene is required for further B cell differentiation. 111. Comparison of Germline Mutational Approach and RAG-2-Deficient Blastocyst Complementation

Experimentally, both the germline mutational approach and the RAG2-deficient blastocyst complementation require the generation of heterozygous mutant ES cells by homologous recombination and the generation of chimeric mice (Fig. 3).However, they differ in generating homozygous mutation for analysis. Using the germline mutational approach, homozyGermline mutation a-

1

homologous recombination

ES+I-

1

injection into C57BU6 biastocyst

Chimera

I

breeding with C57BU6

Heterozygous mutant mouse

1

interbreeding

Homozygous mutant mouse

. lementation RAG-2-deficient blastocvst como

1

homologous recombination

ES+lincreased G418 selection or second targeting

1 1

ES-1-

injection into RAG-2-deficient blastocyst

Chimera

1

Analysis

1 Analysis

FIG.3. Comparison of germline mutational approach and RAG-2-deficient blastocyst complementation.

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

35

gous mutation is achieved through breeding of chimeric mice for germline transmission of the mutant allele and then of heterozygous mutant mice (2,3).In contrast, by RAG-2-deficient blastocyst complementation, homozygous mutation is obtained by manipulating heterozygous mutant ES cells in vitro (10, 11).Because of this difference, RAG-2-deficient blastocyst complementation has some unique advantages. A. POTENTIAL SAVINGS IN TIME AND EXPENSE The germline mutational approach requires extensive breeding of chimeric mice first and then heterozygous mutant mice (3).The process is time consuming and expensive. In contrast, using RAG-2-deficient blastocyst complementation, the effect of the mutation in lymphocyte development is analyzed directly in the chimeras without breeding (10, 11).Although the latter approach requires the maintenance of a RAG-2-deficient mouse colony, the same colony can be used for assaying multiple mutations and can be shared by different investigators. Furthermore, in the latter approach, individual mutant lines are stored as frozen ES cells while homozygous mutant mice are often maintained as a colony. However, when mutant mice are required for long-term studies, it is more cost effective to have germline mutant mice for a steady supply without constantly generating chimeras. In this case, the advantage of generating germline mutant mice can be ascertained in advance.

B. THEABILITYTO ASSAYGENESWHOSEHOMOZYGOUS MUTATION Is EMBRYONIC LETHAL The germline mutational approach depends on the generation of homozygous mutant mice. However, many genes that are important for lymphocyte development are also required for embryogenesis. Homozygous mutation of these genes in mice often results in embryonic lethality and thus the role of these genes in lymphocyte development cannot be analyzed (6-9, 15-17). In contrast, RAG-2-deficient blastocyst complementation assays gene function in lymphocyte development in somatic chimeras (10, 18, 19). The survival of embryos and chimeras is supported by the RAG%deficient blastocysts and does not require the contribution from the injected mutant ES cells. The normal development of RAG-%deficient blastocyst provides a physiological environment for the differentiation of injected mutant ES cells in vivo. Thus, RAG-2-deficient blastocyst complementation is uniquely suited for assaying lymphocyte-specific function of generally expressed genes whose homozygous mutation is embryonic lethal. OF THE MUTATION C. EASYCOMPLEMENTATION Analysis of gene function often requires the complementation of the introduced mutation, To complement germline mutation, mutant mice

36

JIANZHU CHEN

have to be bred with appropriate transgenic mice. If the desired transgenics are not available, they will have to be constructed, which is an expensive and time-consuming process. In contrast, by RAG-2-deficient blastocyst complementation, complementation of the targeted mutation requires only a simple transfection of appropriate expression vectors into mutant ES cells, followed by the generation of chimeric mice in the RAG-2-deficient blastocysts (10,20-22). The expression vectors are analyzed for integration and copy numbers prior to their injection into blastocysts. Because each ES transfectant is equivalent to an independent transgenic line, the latter approach is ideal for studying structure/function in vivo when a large panel of mutations is to be analyzed. D. SHORTFALLS The versatility of RAG-2-deficient blastocyst complementation derives from assaying gene function in chimeric mice, which, unfortunately, also gives rise to its inherent drawback. Each chimera can potentially have different degrees of lymphocyte reconstitution (10, 11, 18, 19). For mutations that affect the quantitative aspects of lymphocyte development, such as the number of lymphocytes, it is sometimes difficult to unequivocally determine whether the phenotype observed is due to the mutation or to the degree of reconstitution. In addition, chimeras are generated individually. For a steady supply of mutant mice, the generation of germline mutant mice is more cost effective if the homozygous mutants are viable. IV. Technical Aspects of RAG-2-Deficient Blasbcyst Complementation

A. GENERATION OF HETEROZYCOUS MUTANTES CELLS As in the germline mutational approach, RAG-2-deficient blastocyst complementation requires the construction of heterozygous mutant ES cells. Mutant ES cells are generated through homologous recombination using standard protocols (1,2). Briefly, a replacement or insertional targeting vector is constructed using neomycin phosphotransferase (neo)gene as a positive selectable marker with or without thymidine kinase ( t k ) gene as a negative selectable marker (1).The neo gene either replaces a part of the gene or is inserted into an critical exon of the gene. The tk gene is placed outside of flanking homologous sequences. Through homologous recombination, the neo gene is targeted into the endogenous locus and the tk gene is lost. The targeting vector is transfected into ES cells and transfectants are enriched for homologous recombinants by G418 selection for the presence of the neo gene (positive selection) and gancyclovir selection for the loss of the tk gene (negative selection). Homologous recombinants (heterozygous mutant ES cells) are identified either by PCR or by

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

37

Southern blot hybridization of genomic DNA isolated from individual ES transfectants ( 2 ) . B. GENERATION OF HOMOZYGOUS MUTANTES CELLS Homozygous mutant ES cells are constructed from heterozygous mutant ES cells by one of the two approaches. The first approach involves a simple selection in increased G418 concentrations (23). The success of the approach relies on the use of a mutant neo gene in the initial targeting vector. A point mutation that changes aspatic acid to glutamic acid at residue 182 substantially decreases the phosphotransferase activity (24). Cells harboring one copy of the mutant gene grow normally in 0.4 mg/ml G418 used in the initial selection. However, they grow very poorly when cultured in higher C418 concentrations such as 8-12 times the initial concentration. Those few clones that grow up in the high G418 concentrations have frequently undergone homologous recombination on the second allele SO that they contain two copies of the neo gene. Although the mechanism is not known, the increased G418 selection readily yields homozygous mutant ES cells. Over 30 homozygous mutant ES cell clones have been derived by this simple selection protocol in Dr. Fred Alt’s laboratory. For an unknown reason, J1 ES cells do not work as efficiently as CCE or E l 4 ES cells when selected in the increased G418 concentration. Because high G418 concentration may not kill heterozygous mutant ES cells, the homozygous mutant ES cells are normally subcloned to avoid the generation of chimeras from a mixture of mutant ES cells. When the normal neo gene is used in the initial targeting vector or the increased G418 selection fails to yield homozygous mutants, a second targeting vector is constructed and transfected into heterozygous mutant ES cells for targeting the second allele (25,26).Usually the second targeting vector is the same as the initial targeting vector except that the neo gene is replaced with a different selectable marker such as puromycin-resistant gene or hygromycin-resistant gene. The targeting vector usually retains the same targeting efficiency. Because both the normal and mutant alleles are targeted at the equal frequency, only half of the homologous recombination events yield homozygous mutant ES cells. C. DELETION OF THE

NEO

GENE

When assaying for the role of cis-regulatory elements in lymphocyte development, the introduced neo gene must be deleted to minimize the effect of the heterologous DNA sequences. Three approaches can be used for this purpose, one utilizes hit-and-run homologous recombination (27) and the other two are based on recombination mediated by either Cre/ loxP or FLP (28, 29). The CreAoxP-mediated recombination has been

38

JIANZHU CHEN

most widely used. Using this approach, the neo gene in the targeting vector is flanked by loxP target sequences. After homologous recombination, Cre is transiently expressed in mutant ES cells to catalyze recombination between the two loxP sequences (30,31). Successful recombination results in the deletion of the neo gene and the retention of only a 34-nucleotide loxP sequence in place of the neo gene. D. GENERATION OF SOMATIC CHIMERAS Injection of homozygous mutant ES cells into blastocysts from RAG-2deficient mice is the same as injection of heterozygous mutant ES cells into C57BL/6 blastocysts (3). Blastocyst donors can be estrus-selected adult females. Because only one-fifth of females are in estrus cycle on a given day, this protocol requires the maintenance of a large pool of females and therefore a large breeding colony. A more economical way is to superovulate 4-to 6-week-old females (3).This method not only eliminates the maintenance of a large pool of females but also reduces the length of time for housing donor females. Both methods work, but the yield of blastocysts vanes probably due to the heterogenous background of RAG-&deficient mice and the physiological state of the females. On average, five injectable blastocysts can be obtained from each donor female. Compared to injected C57BL/6 blastocysts, lower percentages of the injected RAG-2-deficient blastocysts develop into pups and lower percentages of pups are chimeras (F. W. Alt, unpublished data). Particularly, when ES cells harboring a mutation that causes embryonic lethality are injected, the number of pups born and chimerism of the pups are often low. This may result from selection against embryoes that have high levels of ES cell contribution. Injection of fewer ES cells may improve the number of pups born. FOR IDENTIFYING CHIMERAS AND E. USEFULMARKERS ES-DERIVED CELLS To readily identifjl the generated chimeras, the RAG-2 mutation has been backcrossed onto C57BL/6 background (S. Koyasu et aE., unpublished data). Although the mutant mice are still considerably heterogenous, many markers are stable including the expression of black coat color, MHC H2b, and Ly-5.2. Mice of the 129 strain from which most ES cell lines are derived have agouti coat color and express H-2k and Ly-5.1. Generated chimeras are easily identified by the agouti coat color on the black background. Furthermore, all ES-derived cells in the chimeras are readily identifiable by H-2k expression and all ES-derived hematopoietic lineage cells by Ly-5.1 expression. Previously, it was not possible to determine the exact stage at which the JH mutation blocks B cell differentiation in the RAG-2-deficient chimeras. With the new marker Ly-5.1, pro-B cells de-

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

39

rived from JH mutant ES cells were readily detected in the bone marrow (J. Chen, unpublished data). Thus, JH mutation blocks B cell differentiation at the pro-B cells stage, the same as in homozygous mutant mice (32).

F. ANALYSIS Chimeras are usually analyzed by 4-8 weeks of age, initially by flow cytometry of lymphoid cells (11). The presence of mature lymphocytes in the chimeras suggests that the test gene is not absolutely required for lymphocyte differentiation. To confirm that the generated lymphocytes are indeed derived from homozygous mutant ES cells, DNA isolated from lymphocytes is assayed by Southern blot hybridization for the presence of mutation on both alleles (10, 18, 19). This takes advantage of T cell development in the chimeras. During thymocyte differentiation, the transition from DN to DP thymocytes is accompanied by an approximately 100fold expansion of DP thymocytes (33,34). RAG-2-deficient mice have only a few million DN thymocytes (14). If the mutated gene does not affect lymphocyte differentiation, the cellularity of thymus in the chimera is often restored to the normal level containing mostly DP or SP cells (10, 18, 19). These almost pure populations of ES cell-derived thymocytes should have targeted mutations on both alleles. When thymocyte cellularity is not completely restored either due to a partial block of thymocyte differentiation or to the poor quality of ES cells, mature T cells in the periphery are stimulated by Con A to obtain a relatively pure population of T cells for Southern blot analysis (19).Alternatively, the genotype may be verified by cell sorting followed by PCR. In general, the T cell lineage is better reconstituted than the B cell lineage in the chimeras (10, 18, 19). This probably reff ects the differentiation process of T and B cell precursors in their respective organs. Differentiation of DN to DP thymocytes is accompanied by an approximately 100fold increase in cell numbers and the thymus has space for the generated cells. In contrast, pro-B to pre-B cell differentiation is accompanied by a much smaller expansion of pre-B cells (35). Furthermore, the bone marrow is always filled with cells; newly generated pre-B cells must therefore compete for limited space. Alternatively, the difference observed in chimeric mice may suggest that T cells have a greater potential for expansion than B cells in the periphery, which is consistent with the maintenance of T cell numbers in adults after thymus degeneration. When thymocytes are not restored to the normal levels in a generated chimera, the percentage of SP thymocytes is often elevated (19) (F. W. Alt, unpublished data). Most times, this is not due to the introduced mutation because it occurs even when thymocytes are completely restored in other chimeras generated by injecting the same ES cell clone. The level

40

(IANZHU CHEN

of CD3 on SP thymocytes and T cells in periphery in these chimeras is sometimes found to be lower than those in fully reconstituted chimeras or normal mice, suggesting an accelerated process of SP thymocyte differentiation. These chimeras consistently have normal numbers of T cells in the spleen and lymph nodes. If mature lymphocytes are generated in the chimera, the lymphocytes are then assayed for the effect of the mutation on lymphocyte function (11).Functional assays in the chimeras are the same as those in normal mice. However, for in uivo assays, the levels of lymphocyte reconstitution in the chimeras need to be determined before the analysis. This is usually done by flow cytometry of peripheral blood lymphocytes. For in uitro assays using purified lymphocytes, the level of lymphocyte reconstitution in the chimeras should not be a concern. Many types of functional assays have been performed with lymphocytes from chimeric mice including lymphocyte activation, proliferation, signal transduction, lymyhokine and antibody secretion, cytotoxicity, apoptosis, and immune responses. If no mature T and B cells are detected in the chimeras, the mutated gene may be required for lymphocyte differentiation. This will be supported if the injected ES cells can be shown to contribute to nonlymphoid lineages by Southern blot analysis of DNA isolated from different tissues. The exact stage at which lymphocyte differentiation is blocked is determined by flow cytometry using lymphocyte stage-specificmarkers in combination with Ly5.1,which is expressed by all ES-derived hematopoietic cells. Occasionally, mutant ES cells were found not to contribute to any tissues. This could be due to either the introduced mutation or a nonspecific defect that occurred in the ES cells. To distinguish the two possibilities, a normal copy of the mutated gene is transfected into the homozygous mutant ES cells. The expression of the normal gene in the mutant background may complement the mutation and restore lymphocyte differentiation if the observed phenotype is due to the introduced mutation.

G. TRANSFECTION OF MUTANT ES CELLS To transfect an expression vector into mutant ES cells that are already resistant to G418, other selectable markers must be used. Puromycin- or hygromycin-resistant genes have been successfully used (10). Puromycin kills ES cells much more rapidly than hygromycin and therefore is easier to use. ES cells are usually cultured on feeder cells of primary embryonic fibroblasts or established embryonic fibroblast cell lines; feeder cells have to be made resistant to puromycin or hygromycin as well. Established cell lines can be rendered resistant by transfection. However, because primary fibroblasts have limited passages, mice harboring puromycin- or hygromycin-resistant genes will be very useful for obtaining resistant pri-

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

41

mary fibroblasts. ES transfectants are identified by Southern blot analysis. The same assay also establishes the copy numbers of the introduced expression vector. At least three independent transfectants with different copy numbers are used to generate chimeras to eliminate the potential effect of the integration sites on transgene expression. V. Analysis of Gene Function in Lymphocytes

Lymphocyte development involves the differentiation of lymphocytes from hematopoietic stem cells first and then the differentiation of lymphocytes into effector cells during the immune response. Many genes, lymphocyte specific and generally expressed, as well as cis-regulatory elements mediate and control the complex lymphocyte developmental program. A diverse series of genes and cis elements have been analyzed by RAG-2deficient blastocyst complementation for their role in lymphocyte development. Table I summarizes these genes and cis-elements and their effect on lymphocyte development according to their modes of function. In the following section, specific examples are used to illustrate how these studies have advanced our understanding of lymphocyte differentiation and function. When appropriate, the effect of the mutation in homozygous mutant mice is also discussed. A. RECEPTORS

The germline mutational approach has been widely used to assay lymphocyte receptor function. However, many of these receptors are also expressed on nonlymphoid cells that are involved in the same immunological processes. Sometimes, the observed effect in homozygous mutant mice cannot be unequivocally attributed to the mutation in lymphocytes. As illustrated below by targeted mutation of CD40 and CD21, RAG-2deficient blastocyst complementation is able to assay the effect of the mutation in lymphocytes specifically because nonlymphoid cells derived from the blastocysts express the targeted gene normally.

1. CD40 CD40 is a member of the tumor necrosis factor receptor family of molecules (36) and is expressed on B cells and follicular dendritic cells (37,38).Evidence suggests that CD40 provides an essential costimulatory signal for B cell proliferation and antibody response (39). To determine the role of CD40 in B cell development in vivo, ES cells harboring a homozygous mutation at the CD40 locus were assayed by RAG-2-deficient blastocyst complementation (40). Chimeras generated developed phenotypically normal mature B cells, indicating that CD40 is not required for

42

JIANZHU CHEN

TABLE I SUMMARY OF GENESILOCI ASSAYEDBY RAG-%DEFICIENT BUSTOCYST COMPLEMENTATION GenesAoci Receptors Ig J H segments CD40 CD21

Signal transduction csk" VaV Ras Calcineurin A a Btk ZAP-70 Transcription Ets

C-jun"

Rb" GATA-s" PLRLC LH2 N-myc" Cis-reghatory elements IgH 3' enhancer

Major phenotype

Reference

B cell differentiation blocked at pro-B cell stage, low levels of K gene rearrangement Normal B cell maturation, no T cell-dependent antibody response, no germinal center formation Diminished antibody response to suboptimal doses of T cell-dependent antigen, no germinal center formation Diminished DP thymocyte expansion but apparently normal TCRP allelic exclusion

10, 32

40,41 45 21,49

T cell development blocked at fetal DP stage with reduced levels of CD4 expression Thymic atrophy, reduced numbers of mature T cells, no CD5+ B cells, diminished antigen receptor-mediated proliferation Promoting DP thymocyte differentiation in the absence of TCRP expression Refractory to in viuo T cell priming, increased sensitivity to CsA and FK506 No CD5+ B cells, no TI-I1 antibody response, reduced proliferation to anti-IgM and antiCD40 stimulation T cell differentiation blocked at DP stage

52

Reduced DP thymocytes and mature T cells, increased apoptosis of mature T cells, increased numbers of plasma cells and serum IgM Phenotypically normal T and B cells and apparently normal lymphocyte activation and proliferation Phenotypically normal T and B cells and apparently normal lymphocyte activation and proliferation Incomplete block of early hematopoiesis, generation of mature T cells Phenotypically normal T and B cells Phenotypically normal T and B cells Phenotypically normal T and B cells

77, 78

Phenotypicallynormal B cells, diminished switch recombination to IgG2a, IgGeb, IgG3, and IRE

89

56-58 22 66 73, 74 20

19 18 17 h h b

~~

(continues)

43

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

TABLE I-Continued GenesAoci

Major phenotype

Ig

K

intronic enhancer

Ig

K

3' enhancer

TCRP enhancer Iy2b

I& Miscellaneous TdT

Equd numbers of

K~ and A+ B cells, normal levels of K expression, K loci in germline configuration in A t B cells Few K~ B cells, mostly At B cells, little 3' RS rearrangement Diminished TCRP rearrangement, replacable by E p in T cells but not in B cells Blocked switch recombination to IgG2b Required for efficient switch recombination to IgE

No N regon nucleotide addition at VDJ junctions, increased homology-mediated joining Two independent mechanisms for cytotoxicity, perforin mediated and fas mediated Decreased numbers of DP thymocytes and preB cells, apoptosis of DP thymocytes and preB cells in oitro Disappearance of mature T and B cells after 4 weeks of age, increased apoptosis of mature lymphocytes in oitro

Perforin Bcl-x"

Bcl-2

Reference

97 h

98 91 92

101, 102

103 16, 26

104

" Enibyonic lethal. Alt, Xu, Khan, Malynn, Gorinan et al., iinpublished data.

B cell differentiation and survival. In vitro, mutant B cells responded normally to LPS IL-4 by proliferation and Ig secretion; however, they failed to proliferate and undergo class switching to CD40 ligand + IL-4 stimulation. In vivo, chimeric mice failed to generate germinal centers or to produce antibody response to T cell-dependent antigens while they responded normally to T cell-independent antigens. Because follicular dendritic cells derived from RAG-2-deficient blastocysts should express normal levels of CD40, the absence of antibody response in the chimeras must be due to the lack of CD40 expression on B cells. Similar phenotypes were also observed in homozygous mutant mice in which CD40 was absent from both B cells and follicular dendritic cells (41).

+

2. CD21 As with CD40, CD21 (complement receptor) is expressed on both B cells and follicular dendritic cells (42). CD21 forms a complex with CD19, CD81, and Leu-13 and binds to activated complement C3 (43). The crosslinking of antigen receptor and the CD21 complex by C3-antigen complex

44

JIANZHU CHEN

was shown to lower the threshold of B cell activation (44).The role of CD21 in B cells in vivo was assayed specifically by RAG-2-deficient blastocyst complementation (45) because blastocyst-derived follicular dendritic cells express normal levels of CD21. CD21-’-/RAG-’- chimeras produced phenotypically normal mature B cells. However, they failed to produce primary or secondary antibody responses and to generate germinal centers to suboptimal doses of T cell-dependent antigens. These findings were similar to observations in CD19-deficient mice (46,47) and mice lacking complement C3 and C4 (48). Thus, crosslinking of antigen receptor and CD21 on B cells is critical for optimal antibody responses.

3. p T a During T cell differentiation in thymus, functional assembly of the TCR, gene leads to the expression of pre-T cell receptor that consists of a heterodimer of TCR, and pre-T cell receptor a-chain (pTa) and CD3 proteins (34). Interaction of pre-T cell receptor with a putative ligand in the thymic microenvironment is thought to signal the DN to DP thymocyte differentiation, the expansion of DP thymocytes, and TCR, allelic exclusion (33,34).To determine the role ofpTa in these processes, ES cells harboring homozygous mutation of pTa gene were constructed and assayed in RAG%deficient blastocysts (21). Thymocytes representing all stages of T cell differentiation were detected in thymus; however, the number of DP thymocytes was greatly reduced, whereas the relative percentages of DN thymocytes and TCRy6 T cell were increased. Thus, pTa is critical for the expansion of DP thymocytes, consistent with observations in homozygous mutant mice (49). Furthermore, a functionally assembled TCR, gene was introduced into the mutant ES cells and chimeras were generated to determine the effect of the absence of pTa on TCR, allelic exclusion (21). Expression of TCR, transgene in the deficient background inhibited V, to DpJ, rearrangement to the same levels as TCR, expression in the normal background, suggesting that pTa may not be required for TCR, allelic exclusion. B. SIGNAL TRANSDUCTION Lymphocyte differentiation and function depends on the signal transduction from cell surface receptors to nuclei. Many genes involved in lymphocyte signal transduction are also required in other developmental processes, including embryogenesis, and homozygous mutations are embryonic lethal. RAG-2-deficient blastocyst complementation has been used to circumvent these difficulties. In addition, RAG-2-deficient blastocyst complementation has been used to delineate signal transduction pathways by rescuing lym-

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

45

phocyte development through the introduction of genes that likely function downstream into mutant ES cells. 1. Csk Src family protein tyrosine kinases play an essential role in lymphocyte development (50). Csk, which encodes a nonreceptor protein tyrosine kinase, specifically downregulates Src family kinases by phosphorylating the C-terminal regulatory tyrosine (51). To determine whether Csk is part of the regulatory circuit for lymphocyte development, homozygous csk mutant ES cells were assayed in normal as well as RAG-2-deficient blastocysts (52) because the homozygous mutation of csk is embryonic lethal (8, 9). ES cell contribution into lymphoid lineage was determined by Ly-9.1 expression in combination with other lymphocyte markers. In chimeras generated in normal blastocysts, heterozygous mutant ES cells contributed efficiently into lymphoid lineages, but homozygous mutant ES cells did not. Analysis of T cell differentiation in fetal thymus showed that homozygous mutant ES cells can differentiate to DP stage. However, the mutant thymocytes expressed lower levels of CD4 compared to those of normal fetal thymocytes. To determine whether the compromised thymocyte differentiation from homozygous mutant ES cells in normal blastocysts is due to an inefficient competition with wild-type T-lineage cells, chimeras were generated in the RAG-2-deficient blastocysts (52). As in chimeras from normal blastocysts, a low percentage of DP thymocytes was detected in fetal thymus and the level of CD4 expression was reduced. Thus, Csk is required for thymocyte differentiation probably through the regulation of Src family kinases. 2. Vav

Vav consists of structural motifs homologous to SH2, SH3, and pleckstrin domains, and protein kinase C and guanine nucleotide exchange protein (53).It is rapidly phosphorylated on crosslinking of T and B cell antigen receptors (54, 55). To determine the role of Vav in antigen receptor signaling in vivo, homozygous mutant ES cells were constructed and assayed by the RAG-2-deficient blastocyst complementation (56-58). Chimeras generated displayed thymic atrophy and reduced numbers of SP T cells in the periphery. In contrast, chimeras had normal numbers of conventional B cells, but very few CD5+ B cells. Furthermore, mutant T cells failed to proliferate in response to anti-CD3 and Con A stimulation but proliferated normally to TPA and ionomycin stimulation. Similarly, mutant B cells showed much reduced proliferation to anti-IgM or IgD stimulation but proliferated normally in response to LPS and CD40 ligand stimulation.

46

JIANZHU CHEN

Thus, Vav is required for efficient T cell maturation and for mediating antigen receptor signal transduction in mature lymphocytes. 3. Ras A major mode of antigen receptor signaling is through protein phosphorylation and dephosphorylation (50). Ras, a GTPase, is rapidly activated by T cell receptor ligation (59).To determine the role of ras in signaling for T cell development, a genetic complementation approach was employed (22). An activated ras expression vector was introduced into RAG-1deficient ES cells that were then differentiated in vivo using RAG-2deficient blastocyst complementation. Inactivation of RAG-1has identical effects to those of inactivation of RAG-2 and blocks T cell development at the DN stage (60). Injection of RAG-l-deficient ES cells into RAG-2deficient blastocysts simply generated chimeras with only a few millions of DN thymocytes. However, injection of ras-complemented RAG-l-deficient ES cells generated chimeras with normal numbers of DP thymocytes. These thymocytes did not undergo further differentiation into SP thymocytes. Thus, expression of an activated Ras, like the expression of a functionally assembled TCR, gene or expression of the activated Lck (61-63), rescues T cell developmental block due to the lack of V(D)J recombination. These findings support the notion that TCR, expression in DN thymocytes leads to activation of Ras GTPase through Lck and eventually the developmental transition from DN to DP thymocytes. This example illustrates the potential of RAG-2-deficient blastocyst complementation in delineating signal transduction pathways. 4. Calcineurin A a

Calcineurin is a calmodulin-dependent, Ca2+-activatedprotein phosphatase and mediates the immunosuppressive effect of cyclosporine A and FK506 (64, 65). Two isoforms, A a and Ap, are ubiquitously expressed with the former as the dominant isoform in T cells. When differentiated in vivo in RAG-2-deficient blastocysts (66), homozygous A a mutant ES cells readily generated mature T and B cells in the chimeras, suggesting that A a is not absolutely required for lymphocyte differentiation. In vitro stimulation of T cells with Con A, anti-CD3, or anti-CD28 antibodies induced the expression of IL-2Ra, the secretion of IL-2, and the proliferation of the mutant T cells. However, in vivo immunization with T celldependent antigens failed to prime mutant T cells. Because calcineurin is generally expressed, blastocyst-derived cells should express calcineurin A a in the RAG-2-deficient chimeras; thus, the T cell defect observed in vivo is likely due to the calcineurin Aa deficiency in T cells. Furthermore, mutant T cells were more sensitive to CsA and FK506, indicating that

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

47

calcineurin A a is a target of the immunosuppressive drugs and suggesting that AP isoform may compensate for some of the A a function. 5. Bruton’s Tyrosine Kinase (Btk) Mutations in Bruton’s tyrosine kinase (Btk) gene are associated with human X-linked agammaglobulinemia (XLA) and milder murine X-linked immunodeficiency (xid) (67, 68). Btk consists of a pleckstrin homology (PH),Tec homology, SH3, SH2, and SH1 (kinase) domains and is expressed in B cells as well as in myeloid and erythroid cells (69, 70). In human X U , mutations occur in all five domains; in mouse, the xid mutation results from a single amino acid change in the PH domain (71, 72). The effect of mutations in PH and kinase domains on B cell development and function in mice was assayed by RAG-2-deficient blastocyst complementation (73,74).Chimeric mice generated from ES cells harboring mutations in either domains had normal numbers of thymocytes and peripheral T cells. However, they had fewer numbers of B cells in the periphery and almost no CD5+ B cells in the peritoneum. In the bone marrow, there was an increased number of B220+, CD43’ pro-B cells. Similar to xid mice, chimeric mice failed to produce antibody responses to T cellindependent type I1 (TI-11) antigen, whereas they produced a normal response to T cell-dependent antigens. These phenotypes were confirmed by germline mutant mice generated from the same ES cell clones (73). Thus, Btk mediates both pre-B cell receptor and B cell antigen receptor signaling. C. TRANSCRIPTION Lymphocyte development through successive stages is initiated by receptor-ligand interaction, followed by signal transduction from cell surface to nuclei, and ultimately through changes in gene expression. Control of gene expression underlies the developmental program in lymphocytes as well as in other organs. RAG-2-deficient blastocyst complementation has been used to assay the role of transcription factors in lymphocyte development. 1 . Ets-1

In adult mice, transcription factor Ets-1 is expressed predominantly in lymphoid cells in which it has been implicated in regulating gene expression in response to developmental and mitogenic stimuli (75,76).Although the phenotype of ets-1 homozygous mutant mice is not yet known, differentiation of homozygous mutant ES cells in RAG-2-deficient blastocysts has already yielded important information regarding its role in lymphocyte differentiation and function (77, 78). Thymi from ets-l-’- chimeras con-

48

JIANZHU CHEN

tained all types of T lineage cells; however, the relative percentage of DP thymocytes was much reduced and DN thymocytes increased. Consistently, much fewer T cells were present in spleen and Iymph node. Furthermore, thymocytes and mature T cells underwent a high rate of apoptosis in vitro. Thus, Ets-1 is required for efficient T cell maturation in the thymus and probably T cell maintenance in the periphery. In contrast, normal numbers of B cells were present in chimeras; however, a high percentage of B cells exhibited plasma cell phenotype and consequently a much increased IgM level in serum. Together, these findings suggest that Ets-1 may be required for maintaining T and B cells in a quiescent state. 2. C-jun AP-1 transcription factors are heterodimers of Jun and Fos family proteins or homodimers of Jun family proteins. Among Jun family members of C-jun, JunB, and JunD, C-jun is the most widely expressed and transcriptionally potent (79). Homozygous c-jun mutation in mice is embryonic lethal (15).Because AP-1 is part of the nuclear factor of activated T cells (NF-AT) that is required for IL-2 transcription during T cell activation (80),the role of C-jun in lymphocyte differentiation and in T cell activation was assayed by RAG-2-deficient blastocyst complementation (19). Chimeras generated from homozygous mutant ES cells produced phenotypically normal T cells and B cells, indicating that C-jun is not absolutely required for lymphocyte differentiation. Mutant T cells responded normally to antiCD3 antibody and Con A stimulation by expression of IL-2Ra, secretion of IL-2, and proliferation. Thus, C-jun is not required for IL-2 transcription. Analysis of cell lysate from activated T cells by gel retardation assay with specific anti-C-jun, JunB, or JunD antibodies showed that normal T cells express very little C-jun compared with JunB and JunD. Thus, the relatively normal T cell differentiation and activation observed in chimeric mice is probably mediated by JunB and JunD. 3. Retinoblastom Gene (Rb) Rb is involved in the control of cell proliferation and differentiation (81). It is highly expressed in lymphoid cells. However, its role in lymphocyte development is not clear. RAG-2-deficient blastocyst complementation was utilized to assay the effect of Rb deficiency on lymphocyte development in vivo (18) because homozygous Rb mutation is embryonic lethal (7). In Rb-/-iRAG-2-’- chimeras, mature T and B cells were readily generated from homozygous mutant ES cells, indicating that Rb is not required for lymphocyte maturation. In vitro, mutant B cells responded normally to LPS stimulation by proliferation and Ig secretion. Mutant T cells responded normally to Con A and anti-CD3 antibodies by IL-2Ra induction, IL-

ANALYSIS OF G E N E FUNCTION I N LYMPHOCYTES

49

2 secretion, and proliferation. Thus, RB is apparently not required for lymphocyte activation and terminal differentiation. Its function may be compensated for by homologous genes. D. CIS-REGULATORY ELEMENTS The hallmark of lymphocyte development is the assembly and expression of antigen receptor genes. Both processes occur in a lineage and developmental stage-specific manner. By RAG-2-deficient blastocyst complementation, cis-regulatory DNA elements have been shown to play a critical role in the control of V(D)J recombination, class switch recombination, and expression of antigen receptor genes.

1 . l g 3’ Enhancer (3‘ E H ) During early B cell differentiation, Ig genes undergo V(D)J recombination to assemble variable region genes (82). On B cell activation, Ig genes undergo class switch recombination to express Ig classes other than IgM and IgD. IgH locus spans several megabases. Two enhancers have been identified in the locus, one in the JH-Cp intron ( E p ) and the other at the 3’ end of the locus (3’ EH) (83, 84). The intronic enhancer E p has been shown to enhance Ig transcription as well as VDJ recombination (85-87). The 3’ EH was initially defined as consisting of two DNase 1hypersensitive sites (84). Two additional DNase 1 hypersensitive sites, located 11 and 15 kb 3’ downstream, were identified later (88).The upstream hypersensitive sites were replaced by a neo gene and homozygous mutant ES cells were analyzed by RAG-2-deficient blastocyst complementation (89). Mutant ES cells developed phenotypically normal B cells and T cells, indicating that the replaced region in the 3’ EH is required for neither Ig gene rearrangement at the precursor stage nor for IgM and IgD expression on mature B cells. Mutant B cells proliferated normally to LPS stimulation. However, they exhibited a defect in class switch recombination to certain isotypes. For example, LPS failed to induce IgG2b and IgG3 secretion; LPS + IL-4 failed to induce IgE secretion but induced normal IgGl secretion. The lack of IgG2b and IgE secretion correlated with the lack of germline transcription from the loci and switch recombination in the loci. Consistently, chimeric mice had dramatically reduced levels of serum IgG2a, IgG2b, IgG3, and IgE but normal levels of IgM and IgG1. These findings suggest that cis-regulatory elements present at the 3’ EH region control germline transcription, class switching, and Ig isotype expression. 2. ly2b and IE Class switch recombination is mediated by switch region (S) present at 5‘ of exons encoding each Ig isotypes except IgD. Each S region is preceded

50

JIANZHU CHEN

by an I exon from which germline transcription is initiated (90). The processed germline transcript contains the I exon spliced to the constant region gene exons. Because of multiple stop codons in the I exon, germline transcripts apparently do not encode any proteins. The role of germline transcription and transcript were tested by replacing Iy2b exon and promoter with a neo gene (91). Mutant ES cells developed phenotypically normal B cells in the RAG-2-deficient chimeras. However, mutant B cells failed to switch to IgGeb, whereas they switched to upstream and downstream isotypes after appropriate in vitro stimulation. Chimeric mice were deficient of IgG2b in serum. Heterozygous mutant B cells switched to IgGZb on the normal allele but not the mutant allele and chimeric mice had about half the amount of IgGSb, indicating that transcripts from the normal allele do not function in trans for the mutant allele. Although the inserted neo gene was actively transcribed after LPS stimulation, its transcription was clearly not sufficient for inducing class switching to IgG2b. Thus, the transcription process and the specific transcript are both necessary for efficient class switch recombination. To determine the relative contribution of germline transcription and transcripts in mediating class switch recombination, the IE promoter and exon were replaced by the LPS-inducible EpVH promoter (92). Mutant ES cells differentiated into phenotypically normal mature B cells in the RAG-%deficient chimeras. Stimulation of these B cells by LPS alone resulted in high levels of transcription at the S E region from the introduced EpVHpromoter; however, it only induced low levels of switching to IgE. LPS + IL4 stimulation of homozygous mutant B cells resulted in the same low levels of IgE switching. In contrast, LPS + IL4 stimulation of normal and heterozygous mutant B cells induced high levels of IgE switching and secretion. Thus, transcription per se induces low levels of class switch recombination in the absence of I region sequences. Efficient switch recombination requires the presence of the intact I region and/or 1 region promoter in cis. Similarly, the Iyl promoter and exon were found to be necessary for efficient class switch recombination to IgGl in germline mutant mice (31). In addition, deletion of the Iyl exon splicing donor site also abolished switch recombination to IgGl (93), further suggesting that germline transcripts in cis are critical for efficient switch recombination. 3. l g K Enhancers Similar to the IgH locus, the Ig K locus also contains two enhancers, one in the J K - C Kintron (EK)and the other at the 3’ end of the locus (3’EK)(94,95).Although both were identified as transcriptional enhancers, they play critical roles in regulating K gene assembly. The intronic enhancer

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

51

was initially replaced by a neo gene in germline mutant mice (96). Mutant mice did not produce any K+ B cells; all B cells were A+. Because insertion of the neo gene downstream of the E K also blocked K+ B cell differentiation, it was not clear whether the neo gene replacement nonspecifically blocked K gene assembly (96).Thus, mutant ES cells were generated in which EK was replaced by a neo gene flanked by two loxP target sequences. Then, the neo gene was deleted from the targeted allele by transiet expression of Cre (97). Homozygous mutant ES cells developed into K+ B cells in the RAG-2-deficient chimeras. However, the percentage of K+ B cells is approximately the same as that of At B cells, whereas normal mice contain 95% K+ B cells. Because the level of K expression in homozygous mutant B cells is indistinguishable from that of normal B cells, E K is apparently not critical for K transcription. Moreover, the majority of K loci were in the germline configuration in A+ B cells, indicating that EK functions to enhance the probability of K locus to undergo recombination. The 3’ EKwas replaced by a neo gene through homologous recombination in ES cells and analyzed by chimeric as well as germline mutational approaches ( J . Gorman and F. Alt et al., unpublished data). Both assays gave identical results. The replacement resulted in the generation of B cells that mostly bear A light chain. Only 10% of B cells express K at a similar level to that of normal B cells. Furthermore, little 3’ KRS rearrangement was detected in A+ B cells. Thus, similar to E K , the 3’ EK is also required for efficient K rearrangement. The apparent normal K expression in both deletions may reflect overlap activities for transcription by the two enhancers. 4. TCR, Enhancer (ED)

T cell receptor EP enhancer was shown to be required for TCRP gene rearrangement by targeted mutation (98). ES cells harboring a heterozygous deletion of E, readily differentiated into DP thymocytes and mature T cells in RAG-2-deficient chimeras. In both DP thymocytes and mature T cells, the targeted allele remained in the germline configuration as determined by Southern blot hybridization and PCR, indicating that the E, is required for TCR, gene rearrangement in cis. Replacement of the E, with IgH intronic enhancer ( E p ) rescued rearrangement of the targeted TCR, allele in the generated DP thymocytes and mature T cells. However, in B cells, the targeted TCR, allele did not undergo appreciable levels of rearrangement, although the E p enhancer is active in precursor B cells. Together, these findings suggest that Ep can replace E B function for promoting TCR, rearrangement in T cells; but in B cells, negative cisregulatory elements at the TCR, locus suppress rearrangement. This is

52

JIANZHU CHEN

probably one of the mechanisms that controls lineage-specificV( D)J recombination.

E. MISCELLANEOUS 1. Terminal Deoxylnucleotidyl Transferme (TdT) The ability of the immune system to respond to almost all antigenic substances relies on the tremendous diversity of the antigen receptors expressed on lymphocytes. The primary antigen receptor diversity is generated through the random assembly of germline-encoded V, (D), and J gene segments as well as the built-in diversification mechanisms of V( D)J recombination reaction (99). One of these built-in mechanisms is the nontemplated nucleotide addition, referred to as N region, at the DJ and VD junctions. TdT has been postulated to be responsible for the N region addition because of its ability to catalyze nontemplated nucleotide addition at the DNA ends and precursor lymphocyte-specificexpression (100).This hypothesis was tested by sequence analysis of VDJ rearrangements at TCR, and IgH loci in TdT-deficient lymphocytes generated by RAG-2-deficient blastocyst complementation and germline mutant mice (101, 102). Chimeras generated by homozygous TdT mutant ES cells contained phenotypically normal DP thymocytes and mature T and B cells in the periphery, suggesting that TdT is not required for V(D)J recombination and lymphocyte differentiation (101). Sequence analysis of VDJ junctions of TCR, and Ig heavy-chain genes revealed an almost complete lack of N region nucleotides, indicating that TdT is probably solely responsible for N region nucleotide addition. Furthermore, in the absence of TdT, homologous nucleotides are frequently found in the DJ and VD junctions. Thus, although TdT activity is not absolutely required for V( D)J recombination, its presence inhibits the homology-mediated V(D )J recombination process and qualitatively affects the V(D)J recombination products. Similar findings were also obtained from germline TdT-deficient mice (102). The absence of N region diversity may result in a limited antigen receptor repertoire. To determine the effect of TdT deficiency on antigen receptor repertoire and immune response, germline mutant mice are the better choice for providing a steady supply of mutant mice and avoiding potential complications due to the varying levels of lymphocyte reconstitution in chimeras. Nevertheless, the analysis by the RAG-2-deficient blastocyst complementation readily revealed the role of TdT in N region nucleotide addition. 2. Perforin To determine the mechanisms by which cytotoxic T cells lyse antigenspecific target cells, ES cells that have both copies of the perforin gene

ANALYSIS OF GENE FUNCTION IN LYMPHOCYI'ES

53

deleted were used to generate chimeric mice in the RAG-2-deficientblastocysts (103). Chimeras contained phenotypically normal mature T and B cells. CTL lines were then derived from splenic T cells and analyzed for cytotoxicity on different target cells. Perforin-deficient CTLs were capable of inducing antigen-specific target cell lysis and DNA degradation, indicating the existence of other mechanisms for lysing target cells. To determine the contribution of fas-mediated apoptosis in cytotoxicity, fas-negative cells were used as target cells for perforin-deficient CTLs, and there was no lysis. Thus, antigen-specific cytotoxicity is mediated by two mechanisms-one involves perforin and the other requires fas antigen. 3. Bcl-2 and Bcl-x Bcl-2 and Bcl-x are homologous genes whose products can prevent many cell types from undergoing apoptosis. Homeostasis of the lymphocyte compartment involves cell proliferation and apoptosis. Because homozygous bcl-x mutation is embryonic lethal, its role in lymphocyte development and survival was analyzed by chimeric approach either in normal blastocysts or in RAG-2-deficient blastocysts (16,26).Flow cytometry analysis showed the presence of lymphocytes at all stages of T and B cell differentiation in thymus and bone marrow and mature T and B cells in the periphery. In vitro culture of thymocytes and bone marrow cells preferentially resulted in the apoptosis of DP thymocytes and pre-B cells. Similarly, homozygous bcl-2 mutant ES cells developed into mature T and B cells. In contrast to Bcl-x mutant lymphocytes, mature T and B cells in bcl-2 mutant mice or chimeras disappeared after 4 weeks of age (104). In uitro, mutant T cells were more sensitive to steroid- and irradiation-induced apoptosis. These findings suggest that Bcl-x and Bcl-2 function at different stages of lymphocyte development. The former is required for the survival of immature lymphocytes and the latter is required for the survival of mature lymphocytes. VI. Conclusion

RAG-2-deficient blastocyst complementation has been used to assay the role of a diverse group of genes and loci at all aspects of lymphocyte development. It provides an efficient complementary approach to the germline mutation method. Because of its easy complementation, the approach is expected to be utilized more widely for studying structurefunction relationships and elucidating signal transduction pathways in the developing animals.

ACKNOWLEDGMENTS The author thanks Drs. Fred Alt and Dennis Willerford for critical review of the manuscript and Drs. Fred At, Wojciech Swat, Jean-Christophe Bones, Jim Gorman, Yang Xu,

54

JIANZHU CHEN

David Baltimore, and Michael Carroll for making unpublished data available. The author is supported by the Arthritis Investigator Award and the New Investigator Award from Harcourt General Charitable Foundation, Inc.

REFERENCES 1 . Mansour, S. L., Thomas, K. R., and Capecchi, M. R. (1988). Disruption of the protoon-

cogene int-2 in mouse embryo-derived stem cells: A general strategy for targeting mutations to non-selectable genes. Nature 336,348-352. 2. Ramirez-Solis,R., Davis, A. C., Bradley, A. (1992). Gene targeting in embryonic stem cells. Methods Enzymol. 225, 855-878. 3. Hogan, B., Beddington R., Costantini F., and Lacy, E. (1994). “Manipulating the Mouse Embryo: A Laboratory Manual.” Cold Spring Harbor Laboratory Press; Cold Spring Harbor, NY. 4. Gu, H. (1994). Gene targeting and its application to the study of B-cell development. C u m Opin. lmmunol. 6,308-312. 5. Yeung, R. S. M., Penninger, J., and Mak, T. W. (1994).T-cell development and function in gene-knockout mice. C u m Opin. lmmunol. 6,298-307. 6. Charron, J., Malynn, B. A,, Fisher, P., Stewart, V., Jeannotte, L., Gaff, S. P., Robertson, E. J., and Alt, F. W. (1992). Embryonic lethality in mice homozygous for a targeted disruption of the N-myc gene. Gene Deo. 6,2248-2257. 7. Jacks, T., Fazeli, A., Schmitt, E. M., Bronson, R. T., Goodell, M. A., and Weinberg, R. A. (1992). Effects of an Rb mutation in the mouse. Nature 359, 295-300. 8. Nada, S., Yagi, T., Takeda, H., Tohnaga, T., Nakagawa, H., Ikawa, Y., Okada, M., and Aizawa, S. (1993). Constitutive activation of Src family kinases in mouse embryos that lack Csk. Cell 73, 1125-1135. 9. Imamoto, A,, and Soriano, P. (1993). Disruption of the csk gene, encoding a negative regulator of Src family tyrosine kinases, leads to neural tube defects and embryonic lethality in mice. Cell 73,1117-1124. 10. Chen, J., Lansford, R., Stewart, V., Young, F., and Alt, F. W. (1993). RAG-2-deficient blastocyst complementation: An assay of gene function in lymphocyte development. Pmc. Natl. Acad. Sci. USA 90,4528-4532. 11. Chen, J., and Alt, F. W. (1994). RAG-2-deficient blastocyst complementation. In “Transgenesis and Targeted Mutagenesis in Immunology” (H. Bluethmann and P. Ohashi, Eds.), pp. 35. Academic Press, San Diego, CA. 12. Oettinger, M. A., Schatz, D. G., Gorka, C., and Baltimore, D. (1990). RAG-1 and RAG-2, adjacent genes that synergistically activate V( D)J recombination. Science 248, 1517-1523. 13. Schatz, D. G,, Oettinger, M. A., and Schlissel, M. S. (1992). V(D)J recombination: Molecular biology and regulation. Annu. Rw.lmmunol. 10, 359-383. 14. Shinkai, Y., Rathbun, G., Lam, K., Oltz, E. M., Stewart, V., Mendelsohn, M. Charron, J.. Datta, M., Young, F., Stall, A. M., and Alt, F. W. (1992). RAG-2-deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68,855-867. 15. Hilberg, F., Aguzzi, A., Howells, N., and Wagner, E. F. (1993). c-Jun is essential for normal mouse development and hepatogenesis. Nature 365, 179-181. 16. Motoyama, N., Wang, F., Roth, K. A., Sawa, H., Nakayama, K . 4 , Nakayama, K., Senju, S., Zhang, Q . , Fujii, S., and Loh, D. Y. (1995). Massive cell death of immature hematopoietic cells and neurons in Bcl-x-deficient mice. Science 267, 1506-1510. 17. Tsai, F.-Y., Keller, G., Kuo, F. C., Weiss, M., Chen, J., Rosenblatt, M., Alt, F. W., and Orkin, S. H. (1994). An early haematopoietic defects in mice lacking the transcription factor GATA-2. Nature 371,221-226.

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

55

18. Chen, J., Gorman, J. R., Stewart, V., Williams, B., Jacks, T., and Alt, F. W. (1993). Generation of normal lymphocyte populations by &-deficient embryonic stem cells. Curr. Biol. 3,405-413. 19. Chen, J., Stewart, V., Spyrou, G., Hilberg, F., Wagner, E. F., and Alt, F. W. (1994). Generation of normal T and B lymphocytes by c-jun deficient embryonic stem cells. Immunity 1, 65-72. 20. Negishi, I., Motoyama, N., Nakayama, K., Nakayama, K., Senju, S., Hatekeyama, S., Zhang, Q., Chan, A. C., and Loh, D. Y. (1995). Essential role for ZAP-70 in both positive and negative selection of thymocytes. Nature 376, 435-438. 21. Xu, Y., Davidson, L., Alt, F. W., and Baltimore, D. (1996). Function of the pre-Ta chain in the T cell development and the allelic exclusion of the TCRP locus. Submitted for publication. 22. Swat, W., Shinkai, Y., Cheng, H.-L., Davidson, L, and Alt, F. W. (1995).Activated Ras signals differentiation and expansion of CD4+8+thymocytes. Submitted for publication. 23. Mortensen, R. M., Conner, D. A,, Chao, S., Geisterfer-Lowrance, A. A. T., and Seidman, J. G. (1992). Production of homozygous mutant ES cells with a single targeting construct. Mol. Cell Biol. 12, 2391-2394. 24. Yenofsky, R. L., Fine, M., and Pellow, J. W. (1990).A mutant neomycin phosphotransferase I1 gene reduces the resistance of transformants to antibiotic selection pressure. Proc. Natl. Acad. Sci. USA 87, 3435-3439. 25. Cruz, A., Coburn, C. M., and Beverley, S. M. (1991).Double targetedgene replacement for creating null mutants. Proc. Natl. Acad. Sci. USA 88, 7170-7174. 26. Ma, A,, Pena, J. C., Chang, B., Margosian, E., Davidson, L., Alt, F. W., and Thompson, C. B. (1995). Bclx regulates the survival of double-positive thymocytes. Proc. Nutl. Acad. Sci. USA 92,4763-4767. 27. Hasty, P., Ramirez-Solis, R., Krumlauf, R., and Bradley, A. (1991). Introduction of a subtle mutation into the Hox-2.61locus in embryonic stem cells. Nature 350,243-246. 28. O’Gorman, S., Fox, D. T., and Wahl, G. M. (1991). Recombinase-mediated gene activation and site-specific integration in mammalian cells. Science 251, 1351-1355. 29. Sauer, B., and Henderson, N. (1990).Targeted insertion of exogenous DNA into the eukaryotic genome by the Cre recombinase. New Biol. 2, 441-449. 30. Gu, H., Zou, Y.-R., and Rajewsky, K. (1993). Independent control of immunoglobulin switch recombination at individual switch regions evidenced through Cre-loxPmediated gene targeting. Cell 73, 1155-1164. 31. Jung, S., Rajewsky, K., and Radbruch, A. (1993). Shutdown of class switch recombination by deletion of a switch region control element. Science 259, 984-987. 32. Chen, J., Trounstine, M., Alt, F. W., Young, F., Kurahara, C., Loring, J. F., and Huszar, D. (1993). Immunoglobulin gene rearrangement in B cell deficient mice generated by targeted deletion of the J H locus. Int. Immunol. 5, 647-656. 33. Godfrey D. I., and zlotnik, A. (1993). Control points in early T-cell development. Immunol. Today 14,547-553. 34. Groettrup, M., and von Boehmer, H. (1993). A role for a pre-T-cell receptor in T-cell development. Immunol. Today 14, 610-614. 35. Hardy, R. R., Carmack, C. E., Shinton, S. A,, Kemp, J. D., and Hayakawa, K. (1991). Resolution and characterization of pro-B and pre-pro-B cell stages in normal mouse bone marrow. /. Exp. Med. 173, 1213-1225. 36. Armitage, R. J., Fanslow, W. C., Strockbine, L., Sato, T. A., Clifford, K. N., Macduff, B. M., Anderson, D. M., Cimpel, S. D., Davis-Smith, T., Maliszewski, C. R., Clark, E. A,, Smith, C. A., Grabstein, K. H., Cosman, D., and Spriggs, M. K. (1992).Molecular and biological characterization of a murine ligand for CD40. Nature 357, 80-82.

56

JIANZHU CHEN

37. Clark, E. A,, and Ledbetter, J. A. (1986).Activation of human B cells mediated through two distinct cell surface differentiation antigens, Bp35 and Bp50. Proc. N d . Acad. Sci. USA 83,4494-4498. 38. Hart, D. N. J,, and McKenzie, J. L. (1988). Isolation and characterization of human tonsil dendritic cells. J. Exp. Med. 168, 157-170. 39. Banchereau, J., Bazan, F., Briere, F., Calizzi, J. P., van Kooten, C., Liu, Y. J., Rousset, F., and Saeland, S. (1994). The CD40 antigen and its ligand. Annu. Rev. Immunol. 12, 881-922. 40. Castigli, E., At, F. W., Davidson, L., Bottaro, A., Mizoguchi, E., Bhan, A. K., and Geha, R. S. (1994). CD40-deficient mice generated by recombination-activatinggene%deficient blastocyst complementation. Proc. Natl. Acad. Sci. USA 91, 12135-12139. 41. Kawabe, T., Naka, T., Yoshida, K., Tanaka, T., Fujiwara, H., Suematsu, S., Yoshida, N., Kishimoto, T., and Kikutani, H. (1994). The immune responses in CD40-deficient mice: Impaired immunoglobulin class switchingand germinal center formation. Immunity 1, 167-178. 42. Aheam, J. M., and Fearon, D. T. (1989). Structure and function of the complement receptors CR1 (CD35) and CR2 (CD21). Ado. Zmmunol. 46, 183-219. 43. Tedder, T. F., Zhou, L.-J.,and Engel, P. (1994). The CD19/CD21 signal transduction complex of B lymphocytes. lmmunol. Today 15, 437-442. 44. Carter R. H., and Fearon, D. T. (1992). CD19: Lowering the threshold for antigen receptor stimulation of B lymphocytes. Science 256, 105-107. 45. Croix, D. A,, Ahearn, J. M., Rosengard, A. M., Han, S., Kelsoe, G., Minghe, M., and Carroll, M. C. (1996). Antibody response to a T-dependent antigen requires B cell expression of complement receptors. Submitted for publication. 46. Engel, P., Zhou, L.-J., Ord, D. C., Sato, S., Koller, B., and Tedder, T. F. (1995). Abnormal B lymphocyte development, activation, and differentiation in mice that lack or overexpress the CD19 signal transduction molecule. Immunity 3, 39-50. 47. Rickert, R. C., Rajewsky, K., and Roes, J. (1995). Impairment of T-cell-dependent Bcell responses and B-1 cell development in CD19-deficient mice. Nature 376,352-355. 48. Fischer, M. B., Ma, M., Zhou, X., Xia, J., Finco, O., Rothstein, R., Kremmer, E., Rosen, F. S., and Carroll, M. C. (1996). Disruption of complement C3 or C4 loci results in a profound defect in the B cell response to T-dependent antigens. Submitted for publication. 49. Fehling H. J., Krotkova, A. Saint-Ruf, C., and von Boehmer, H. (1995). Crucial role of the pre-T-cell receptor a gene in development of ap but not y8 T cells. Nature 375, 795-798. 50. Perlmutter, R. M., Levin S. D., Appleby, M. W., Anderson, S. J., and Alberola-Ila, J. (1993). Regulation of lymphocyte function by protein phosphorylation. Annu. Rev. lmmunol. 11, 452-499. 51. Nada, S., Okada, M., MacAuley, A., Cooper, J. A., and Nakagawa, H. (1991). Cloning of a complementary DNA for a protein-tyrosine kinase that specificallyphosphorylates a negative regulatory site of ~60’”“.Nature 351, 69-72. 52. Gross, J. A,, Appleby, M. W., Chien, S., Nada, S., Bartelmez, S. H., Okada, M., Aizawa, S., and Perlmutter, R. M. (1995). Control of lymphopoiesis by p5OCsk’, a regulatory protein tyrosine kinase. J. Exp. Med. 181, 463-473. 53. Adams, J. M., Houston, H., Allen, J., Lints, T., and Harvey, R. (1992).The hematopoietically expressed vav proto-oncogene shares homology with the dbl GDP-GTP exchange factor, the bcr gene and a yeast gene (CDC24) involved in cytoskeletal organization. Oncogene 7, 611-618.

ANALYSIS OF GENE FUNCTION IN LYMPHOCYTES

57

54. Bustelo, X. R., and Barbacid, M. (1992). Tyrosine phosphorylation of the vav protooncogene product in activated B cells. Science 256, 1196-1199. 55. Bustelo, X. R., Ledbetter, J. A., and Barbacid, M. (1992). Product of vav protooncogene defines a new class of tyrosine protein kinase substrates. Nature 356,68-71. 56. Tarakhovsky, A., Turner, M., Schaal, S., Mee, P. J., Duddy, L. P., Rajewsky, K., and Tybulewicz, V. L. J. (1995). Defective antigen receptor-mediated proliferation of B and T cells in the absence of Vav. Nature 374,467-470. 57. Zhang, R., Alt, F. W., Davidson, L., Orkin, S. H., and Swat, W. (1995). Defective signalling through the T- and B-cell antigen receptors in lymphoid cells lacking the vav proto-oncogene. Nature 374, 470-473. 58. Fischer, K.-D., Zmuldzinas, A,, Gardner, S., Barbacid, M., Bernstein, A., and Guidos, C . (1995). Defective T-cell receptor signalling and positive selection of Vav-deficient CD4+ CD8+ thymocytes. Nature 374,474-477. 59. Downward, J., Graves, J. D., Warne, P. H., Rayter, S., and Cantrell, D. A. (1990). Stimulation of p2lras upon T-cell activation. Nature 346, 719-723. 60. Mombaerts, P., Iacomini, J., Johnson, R. S., Kerrup, K., Tonegawa, S . , and Papioannou, V. A. (1992). RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68, 869-877. 61. Mombaerts, P., Clarke, A. R.,Rudnicki, M. A., Iacomini, J., Itohara, S., Lafaille, J. J., Wang, L., Ichikawa, Y., Jaenisch, R., Hooper, M. L., and Tonegawa, S. (1992). Mutations in T-cell antigen receptor genes a and /3 block thymocyte development at different stages. Nature 360, 225-231. 62. Mombaerts, P., Anderson, S. J., Perlmutter, R. M., Mak, T. W., and Tonegawa, S. (1994).An activated Ick transgene promotes thymocyte development in RAG-1 mutant mice. Immunity 1, 261-267. 63. Shinkai, Y., Koyasu, S., Nakayama, K., Murphy, K. M., Loh, D. Y., Reinherz, E. L., and Alt, F. W. (1993). Restoration of T cell development in RAG-%deficient mice by functional TCR transgenes. Science 259,822-825. 64. Cohen, P. (1989). The structure and regulation of protein phosphatases. Annu. Rev. Bwchem. 58,453-508, 65. Liu, J., Farmer, K. D., Jr., Lane, W. S., Friedman, J., Weissmann, I., and Schreiber, S. L. (1991). Calcineurin is a common target of cyclophilin-cyclosporin A and FKBPFK506 complexes. Cell 66, 807-815. 66. Zhang, W., Zimmer G., Chen, J., Ladd, D., Li, E., Alt, F. W., Wiederrecht, G., Cryan, J., O’Neill, E. A., Seidman, C. E., Abbas, A. K., and Seidman, J. G. (1996). T cell responses in calcineurin Aa-deficient mice. J E r p Med, in press. 67. Bruton, 0. C. (1952). Agammaglobdinemia. Pediatrics 9, 722-727. 68. Scher, I. (1982). The CBA/N mouse strain: An experimental model illustrating the influence of X chromosome on immunity. Adv. Immunol. 33, 1-71. 69. Tsukada, S., Saffran, D. C., Rawlings, D. J., Parolini, O., Alen, R. C., Klisak, I., Sparkes, R. S., Kubagawa, H., Mohandas, T., Quan, S., Belmont, J. W., Cooper, M. D., Conley, M. E., and Witte, 0. N. (1993). Deficient expression of a B cell cytoplasmic tyrosine kinase in human x-linked agammaglobulinemia. Cell 72, 279-290. 70. Vetrie, D., Vorechovsky, I., Sideras, P., Holland, J., Davies, A., Flinter, F., Hammarstrom, L., Kinnon, C., Levinsky, R., Bobrown, M., Smith, C. I. E., and Bentley, D. R. (1993). The gene involved in X-linked agammaglobulinaemia is a member of the src family of protein-tyrosine kinases. Nature 361, 226-233. 71. Thomas, J. D., Sideras, P., Smith, C . I. E., Vorechovsly, I., Chapman, V., and Paul, W. E. (1993). Colocalization of x-lined agammaglobulinemiaand x-linked immunodeficiency genes. Science 261, 355-358.

58

JIANZHU CHEN

72. Rawlings, D. J . , Saffran, D. C., Tsukada, S., Largaespada, D. A., Grimaldi, J. C., Cohen, L., Mohr, R. N., Bazan, J. F., Howard, M., Copeland, N. G., Jenkins, M. A., and Witte, 0. N. (1993). Mutations of unique region of Bruton’s tyrosine kinase in immunodeficient XID mice. Science 261,358-361. 73. Khan, W. N., Alt, F. W., Gerstein, R. M., Malynn, B. A., Larsson, I., Rathbun, G., Davidson, L., Muller, S., Kantor, A. B., Herzenberg, L. A., Rosen, F. S., and Sideras, P. (1995).Defective B cell development and function in Btk-deficient mice. Zmmunity 3,283-299. 74. Kerner, J . D., Appleby, M. W., Mohr, R. N., Chien, S., Rawlings, D. J., Maliszewski, C. R., Witte, 0. N., and Perlmutter, R. M. (1995). Impaired expansion of mouse B cell progenitors lacking Btk. Immunity 3, 301-312. 75. Ho, I . C., Bhat, N. K., Gottschalk, L. R., Lindsten, T., Thompson, C. B., Papas, T. S., and Leiden, J. M. (1990). Sequence-specific binding of human Ets-1 to the T cell receptor alpha gene enhancer. Science 250, 814-818. 76. Nelsen, B., Tian, G., Erman, B., Gregoire, J., Maki, R., Graves, B., and Sen, R. (1993). Regulation of lymphoid-specific immunoglobulin mu heavy chain gene enhancer by ETS-domain proteins. Science 261, 82-86. 77. Bones, J. C., Willerford, D. M., Crkvin, D., Davidson, L., Camus, A., Martin, P., Stkhelin, D., and Alt, F. W. (1995). Increased T-cell apoptosis and terminal B-cell differentiation induced by inactivation of the Ets-1 proto-oncogene. Nature 377, 635-638. 78. Muthusamy, N., Barton, K., and Leiden, J. M. (1995).Defective activation and survival of T cells lacking the Ets-1 transcription factor. Nuture 377, 639-642. 79. Karin, M., and Smeal, T. (1992).Control of transcription factors by signal transduction pathways: The beginning of the end. TZBS 17, 419-423. 80. Schreiber, S. L., and Crabtree, G. R. (1992).The mechanism of action of cyclosporin A and FK506. Immunol. Today 126, 136-142. 81. Hollingsworth, R. E., Jr., Hensey, C. E., and Lee, W.-H. (1993). Retinoblastoma protein and the cell cycle. Curr. Biol. 3, 55-62. 82. Alt, F., Blackwell, T. K., and Yancopoulos, G. (1987). Development of the primary antibody repertoire. Science 238, 1079- 1087. 83. Gillies, S. D., Morrison, S. L., Oi, V. T., and Tonegawa, S. (1983). A tissue-specific transcription enhancer elements in immunoglobulin heavy chain gene. Cell 33, 717-728. 84. Pettersson, S., Cook, G. P., Bruggemann, M., Williams, G. T., and Neuberger, M. S. (1990). A second B cell-specific enhancer 3’ of the immunoglobulin heavy-chain locus. Nature 344, 165-168. 85. Femer, P., Krippl, B., Blackwell, T. K., Furley, A. J.. Heikyung, S., Winoto, A,, Cook, W., Hood, L., Constantini, F., and Alt, F. (1990). Separate elements control DJ and VDJ rearrangement in a transgenic recombination substrate. EMBO J . 9, 117-125. 86. Chen, J., Young, F., Bottaro, A., Stewart, V., Smith, R., and Alt, F. W. (1993).Mutations of the intronic 1gH enhancer and its flanking sequences differentiallyaffect accessibility of the Jh locus. EMBO J. 12, 4635-4645. 87. Senve, M., and Sablitzb, 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]. 12,2321-2327. 88. Madisen, L., and Groudine, M. (1994). Identification of a locus control region in the immunoglobulin heavy-chain locus that deregulates c-myc expression in plasmacytoma and Burkitt’s lymphoma cells. Gene Dev. 8,2212-2226.

ANALYSIS OF G E N E FUNCTION IN LYMPHOCYTES

59

89. CognC, M., Lansford, R., Bottaro, A,, Zhang, J., Gorman, J,, Young, F., Cheng, H., and Alt, F. W. (1994).A class switch control region at the 3’ end of the immunoglobulin heavy chain locus. Cell 77,737-747. 90. Zhang, J., Alt, F. W., and Honjo, T. (1995). Regulation of class switch recombination of the immunoglobulin heavy chain genes. In “Immunoglobulin Genes” (T. Honjo, F. W. Alt, and T. H. Rahhilett, Eds.), 2nd ed., pp. 235. Academic Press, Inc. 91. Zhang, J., Bottaro, A., Li, S., Stewart, V., and Alt, F. W. (1993). A selective defect in IgG2b switching as a result of targeted mutation of the I gamma 2b promoter and exon. EMBO]. 12, 3529-3537. 92. Bottaro, A., Lansford, R., Xu, L., Zhang, J., Rothman, P., and Alt, F. W. (1994). S region transcription per se promotes basal IgE class switch recombination but additional factors regulate the efficiency of the process. E M B O ] . 13, 665-674. 93. Lorenz, M., Jung, S., and Radbruch, A. (1995). Switch transcripts in immunoglobulin class switching. Science 267, 182551828, 94. Picard, D., and Schaffner, W. (1984). A lymphocyte-specific enhancer in the mouse immunoglobulin K gene. Nature 307, 80-82. 95. Meyer, K. B., and Neuberger, M. S. (1989).The immunoglobulin kappa locus contains a second, stronger B-cell-specificenhancer which is located downstream of the constant region. EMBO]. 8, 1959-1964. 96. Takeda, S., Zou, Y., Bluethmann, H., Kitamura, D., Muller, U., and Rajewsky, K. (1993). Deletion of the immunoglobulin K chain intron enhancer abolishes K chain gene rearrangement in cis hut not A chain gene rearrangement in trans. E M B O ] . 12, 2329-2336. 97. Xu, Y., Davidson, L., Alt, F. W., and Baltimore, D. (1996).Deletion ofimmunoglobulin kappa light chain intron enhancedmatrix attachment region impairs but does not abolish VKJKrearrangement. Submitted for publication. 98. Bones, J.-C., Demongeot, J., Davidson, L., and Alt, F. W. (1996). Deletion of the TCRP enhancer inhibits TCRP chain gene rearrangement. Submitted for publication. 99. Lansford, R., Okada, A,, Chen, J., Oltz, E. M., Blackwell, K., Alt, F. W. and Rathbun, G. (1995). Mechanism and control of immunoglobulin gene rearrangement. Mol. Immunol., in press. 100. Alt, F. W., and Baltimore, D. (1982). Joining of immunoglobulin heavy chain gene segments: Implications from a chromose with evidence of three D-JH fusions. Proc. Natl. Acad. Sci. USA 79,4118-4122. 101. Komori, T., Okada, A,, Stewart, V., and Alt, F. (1993). Lack of N regions in antigen receptor variable regions genes ofTdT-deficient lymphocytes.Science 261,1171-1175. 102. Gilfillan, S., Dierich, A., Lemeur, M., Benoist, C., and Mathis, D. (1993). Mice lacking TdT: Mature animals with an immature lymphocyte repertoire. Science 261, 11751178. 103. Kojima, H., Shinohara, N., Hanaoka, S., Someya-Shirota, Y., Takagaki, Y., Ohno, H., Saito, T., Katayama, T., Yagita, H., Okumura, K., Shinkai, Y., At, F. W., Matsuzawa, A,, Yonehara, S., and Takayama, H. (1994). Two distinct pathways of specific killing revealed by perforin mutant cytotoxic T lymphocytes. Immunity 1, 357-364. 104. Nakayama, K., Negishi, I., Kuida, K., Shinkai, Y., Louie, M. C., Fields, L. E., Lucas, P. J., Stewart, V., Ak, F. W., and Loh, D. Y. (1993). Disappearance of the lymphoid system in Bcl-2 homozygous mutant chimeric mice. Science 261, 1584-1588. This article was accepted for publication on 10 November 1995.

This Page Intentionally Left Blank

ADVANCES IN IMMUNOLOGY, VOL 62

Interferon-y: Biology and Role in Pathogenesis

1. Introduction

The protein that is now called interferon-? (IFN-y) was discovered independently by two groups of investigators and was originally given two different names after the biological activities detected: (a) type I1 or immune interferon, and (b) macrophage-activatingfactor (MAF). Homonymy of IFN-y with IFN-a and -/3 does not imply molecular relationship but merely reflects sharing of the biological property to protect cells against virus infection. The corresponding bioassay is the so-called antiviral assay, which consists in demonstrating that cultured cells exposed to IFN-y resist destruction by a standard challenge virus. Historically, the name interferon refers to the phenomenon of interference, i.e., the fact that cells infected with any one virus species are relatively resistant to infection with viruses of other species. However, whereas IFN-a and -/3 (collectively called type I interferons) do play a role in such interference, IFN-y is not involved. The earliest report dealing with IFN-y is probably that by Wheelock in 1965 (l),demonstrating that an interferon-like antiviral activity appears in supernatants of mononuclear cells exposed to a mitogen. In the early seventies, the terms “type I1 interferon” (2) and “immune interferon” (3) were coined. The term immune interferon remained in use for some time in recognition of the awareness that the activity is associated with a protein physicochemically distinct from that responsible for the then-classical interferon, that production of the factor is the prerogative of immune competent cell types, and that the factor possesses immune regulatory properties distinct from those of classical interferon. In 1980, an international interferon nomenclature committee (4)agreed on a new name for all interferons. Type I interferons, then known as leukocyte and fibroblast interferons, were renamed IFN-a and -/3, respectively; immune interferon was renamed IFN-y. Availability of pure preparations of IFN-y and of monospecific antibodies made it possible to prove that MAF activity in biological fluids is largely, if not entirely, accounted for by IFN-7. The term MAF appeared in the literature around 1980; it refers to the ability of supernatants of mitogen- or antigen-challenged mononuclear cell cultures to augment various biological activities of macrophages. That stimulated lymphocytes’ supernatants have these macrophage-stimulating abilities was first re61 Copyright 8 19% by Academic Press. Inc.

All rights of reproduction in any form reserved.

62

ALFONS BILLIAU

ported in 1966 by Bloom and Bennett (5a). MAF bioassays have variably relied on intracellular killing of parasites or increased oxidative metabolism ( S ) , enhanced expression of class I1 antigens (6), or enhanced tumor cell killing (7). Characterization of MAF with monoclonal antibodies soon revealed its identity with IFN-y ( 5 ) . It is possible that other cytokines besides IFN-y have MAF-like activities, but they seem to be of less importance. Ever since their discovery, interferons have attracted vivid interest from clinical investigators. Medical interest in IFN-7 stems from awareness that a prominent target cell of IFN-y, the macrophage, occupies a central position in the immune system. Adequate function of the IFN-ylmacrophage system is essential for natural as well as acquired resistance to infection and cancer. Malfunctioning of the system is recognized to be instrumental in inflammatory and autoimmune disease. Not surprisingly, therefore, of all known cytokines, IFN-7 belongs to the small group that has already been tested for therapeutical effects in patients. Being among the first cytokines to be discovered, IFN-y is the subject of an extensive and still expanding literature. Surprisingly, only a few reviews have been devoted to IFN-y (8,9).Here, the salient issues and most recent advances concerning IFN-y’s biological function are reviewed so as to provide a framework for understanding its role in disease. II. Structure and Structure-Function Relationship

IFN-y is a glycoprotein the size, amino acid sequence, and glycosylation of which are well conserved among animal species. Most studies on structure have been done on either human or mouse IFN-y. In its biologically active form, IFN-y is a 34-kDa homodimer stabilized by noncovalent forces. The peptide is N-glycosylated on two sites. Natural IFN-y is heterogeneous in size and charge due to enzymatic trimming of the carboy terminus and to variation in degree of glycosylation. This heterogeneity seems to be unimportant for biological activity on cells but might well influence dynamics of tissue distribution. X-ray crystallographic analysis (10)has revealed that the subunits consist of six a-helices, accounting for 62% of the molecule, with no &sheet domains. The two subunits, each of which is elongated in shape, are held together in an antiparallel configuration by intertwining of the helical domains. The amino terminus of each chain is juxtaposed to the carboy terminus of the opposing chain. The symmetry in this structure allows for IFN-y to bind a pair of identical receptor peptides, as has been demonstrated by analysis of the crystal structure of an IFN-yIIFN-yR complex (11). Peptide stretches at both the amino and the carboxy termini of

INF-7: BIOLOGY AND ROLE IN PATHOGENESIS

63

IFN-y seem to have binding properties for the receptor, as demonstrated by competition-type experiments using the corresponding peptides (12). The amino-terminal binding regon has been identified (13). By substituting glutamine residues for asparagine residues at positions 25 and/or 97, four differently glycosylated mutant forms of human interferon could be obtained (14) and studied. These replacements were found to affect both production yields by insect cells as well as correct dimer formation. However, specific biological activity of the correctly formed dimers remained largely unaffected. Studies employing monoclonal antibodies and recombinant mouse/man hybrid IFN-y molecules have provided suggestive but not conclusive evidence that mouse IFN-y is composed of distinct functional domains that induce different biological responses. A carboxy-terminal domain was proposed to be sufficient for macrophage activation, whereas both a carboxy- and a different aminoterminal domain would be required for induction of the antiviral effect in cells (15). Neutralizing monoclonal antibodies against human IFN-y fall into four epitope-specific groups ( E l , E2, EUE2 overlap, and E3) identifjmg three epitopes (El, E2, and E3), each of which is somehow involved in biological activity (16). However, none of these epitopes identifies the receptorbinding site. Therefore, it has been hypothesized that receptor binding and signal transduction functions are separate on the IFN-y molecule and that the neutralization epitopes so far identified represent signal transduction functional domains. Only epitope E l has to date been identified with a particular stretch of amino acids (residues 84-94). Intriguingly, the amino acid sequence at this domain resembles the nuclear localization (targeting) signal (NLS) of many nuclear proteins. 111. Producers and Production of IFN-y

IFN-y is a typical lymphokine, being produced exclusively by NK cells and certain subpopulations of T lymphocytes, namely the TH1 subclass of CD4' lymphocytes and certain CD8' lymphocytes (17) (Fig. 1). In the human system, T cells that express the activation-dependent CD30 membrane antigen have been identified as the principal subset producing IFNy (18).A single literature report describes production of IFN-y by mononuclear phagocytes (19). To produce IFN-y, lymphocytes need to be activated, as a result of which they also secrete other cytokines, e.g., IL-2 in the case of TH1 cells. As a general rule, production of IFN-y by either NK or T cells requires cooperation of accessory cells, mostly mononuclear phagocytes, which also need to be in some state of activation. One aspect of the ancillary role of

64

ALFONS BILIJAU

1st signal

2nd signal

Lectin

3rd signal

1st signal

2nd signal

Lectin

/

c=,IFN-Y

Cell adhesion molecules

3rd signal Cell adhesion molecules 7

FIG.1. Producers of IFN-y. IFN-y is produced by only two types of cells: T lymphocytes (both TH1 and CDW cells) and large granular lymphocytes (NK cells). Constitutive production is minimal or nonexistant. Optimally induced production occurs when the cells receive a combination of three types of signals: (a) a specific or aspecific ligand for the receptor, (b) a balanced assembly of cytokines, and (c) contact with accessory cells through cell adhesion molecules.

these cells is that they need to produce regulatory cytokines, e.g., TNFa, IL-12, or IFN-P, but also that they need to make contact with the lymphocytes through intercellular adhesion molecules. The evidence that has led to this more or less generally applicable schedule is diverse, as is summarized below. Gram-positivebacterial components, e.g., S. aureus preparations, induce NK cells to produce IFN-y in vitro (20). In gram-positive bacterial infections, e.g., in listeriosis, NK cells contribute a large part in IFN-y production, which accounts for resistance of the host to the infection (21). Similarly, in parasitic infections, e.g., in Leishmaniasis models, early production of IFN-y, which is of crucial importance for the further course of the infection, is accounted for by NK cells (22). It should be noted, however, that for exogenous stimuli to induce IFN-y production in NK cells, the presence of cytokines produced by monocytes is mostly required. Thus, inscid mice, induction by Listeria of IFN-7 production by NK cells requires monocytes. The role of the monocyte could be to produce TNF. TNF can indeed substitute for monocytes (23,24)and TNF-receptor knockout mice are highly susceptible to Listeria infection (25),but an additional monocyteproduced cytokine seems to be required for optimal production. This is most likely IL-12, known to strongly upregulate IFN-y production by NK cells (26). A comparison of IFN-y production by murine splenocytes of responsive and nonresponsive strains led to the conclusion that, for gram-negative bacteria to induce IFN-7, the presence of mononuclear phagocytes and production of IFN-P is required (27). IFN-y produced in response to injection of endotoxins of gram-negative bacteria apparently originates from NK cells as well as CD4+ and CD8+ T cells, because the mRNA is detectable in all three populations (28).Again, in endotoxin-injected mice,

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

65

release of IL-12 into the circulation precedes that of IFN-y, and pretreatment with anti-IL-12 antibody inhibits such production of IFN-y (28). Nevertheless, in the lethal Shwartzman reaction caused by two consecutive injections of endotoxin, elimination of NK cells, but not of CD4+ or CD8+ T cells, prevents the toxic manifestations of the reaction (29). However, not only monocyte-derived cytokines, but also IL-2 produced by TH1 cells can stimulate IFN-7 production by NK cells (22). As an exception to the general rule that IFN-y induction in NK cells requires help from accessory cells, the superantigen SEB reportedly can independently induce IFN-y production in human NK cells (30);activation into cytotoxic cells, however, was found to require the presence of cytokines delivered by T cells. The situation may be quite different in mycobacterial infections (31). IL-12 induction by Mycobactem'um bovis (BCG) in murine bone marrowderived macrophages was found to depend on the presence of IFN-y and TNF, as it failed to occur in macrophages of mice lacking functional receptors for either cytokine or was prevented by addition of neutralizing antibodies to either cytokine. Also, in viuo, IL-12 production in the spleen was absent in TNF-a or IFN-y receptor knockout mice. Therefore, in this system IL-12 does not precede IFN-y as seems to be the case in other systems. In mice infected with Toxoplasma, both CD4+ T cells and NK cells contribute to the early IFN-7 response. Monoclonal antibodies against IL12 were found to reduce both IFN-y production and resistance to the infection (32). IFN-y production by T lymphocytes occurs when these cells are activated by antigens, e.g., aviral antigen presented in MHC class I molecule context, to the idiotypic receptor (33). Activation by antigens presented by professional antigen-presenting cells also involves cell adhesion molecules and their corresponding hgands, Thus, enhancement of IFN-y production by signaling through the CD2/LFA-3 pathway has been demonstrated in studies with CD2-blocking antibodies and with cells transfected with LFA3 (34). Moreover, cytokines and other mediators released by the antigenpresenting cells, as well as by the lymphocytes themselves and by bystander lymphocytes, have a profound regulatory effect on the quantities of IFNy that are released. IL-2 is classically said to stimulate production of IFNy. In allergen-driven murine T cells, IFN-y production is strictly dependent on prior production of IL-2, whereas IL-4 production is IL-2 independent (35). IL-10, on the other hand, inhibits IFN-y production (see below). Crucial to the production of IFN-y during antigen-specific immune responses in vivo is the development of naive CD4' T cells into either TH1 or TH2 lymphocyte clones. Total production of IFN-.)I during later

66

ALFONS BILLIAU

phases of immune responses is more elevated if the antigen-specific TH1 cells predominate. Using T cell receptor transgenic CD4' T cells, it has been shown that IL-12 substantially enhances development into cells that produce IFN-y on restimulation; IL-4, by contrast, inhibits such development (36). The stimulatory effect of IL-12 on IFN-y production by activated human T cells is largely mediated by its effect on the CD3O-expressing subset (37). Production of IFN-y by T cells is also under regulatory control of prostaglandins. PGES inhibits production by TH1 cells of IFN-y and IL-2 but not production of IL-4 by TH2 cells. CAMP is involved in this control mechanism, which allows PGES to skew immune responses in the TH2 direction (38). Some information is available on the molecular basis of IFN-y induction. Analysis of the promoter region has yielded evidence for the existence of three distinct response elements upstream of the TATA box. One of these (BE), originally detected by its silencer activity, was shown to be able to interact with two protein complexes, S and E, with silencer and enhancer activity, respectively. The S protein complex recognizes a sequence similar to that recognized by nuclear factor AP2, but does not contain AP2 itself. The E complex contains nuclear factor YY1 (39). IV. Biochemical Basis of IFN-)I Action

A. THERECEPTOR The IFN-y receptor complex on human mononuclear leukocytes consists of at least three distinct proteins (40). One of these, the IFN-y receptor a-chain (IFN-yRa), is the membrane protein that primarily binds IFNy with high affinity and the gene of which is located on human chromosome 6 (41) or mouse chromosome 10. It encompasses three domains (42, 43): the extracellular ligand-binding, the transmembrane, and the intracellular domains. When the human or mouse variants of this protein are expressed in heterologous cells, they do bind the homologous IFN-7 but fail to transmit any signal to the cell. Chimeric mousehuman constructs of the protein, expressed in cells of either species, has indicated that, for signals to be transmitted, the extracellular domain must interact with one or more species-matching proteins (44). Studies with cell hybrids have revealed one of these proteins to be coded for by human chromosome 21 (45) or mouse chromosome 16 (46). Both in murine and in human cells, the species-matching protein has been fulIy characterized by molecular cloning (47, 48). This protein, termed IFN-yRP, has a structure resembling that of IFN-yRa. Interaction of the a-chain extracellular domain with this protein is sufficient for transmission of the signal to express MHC class

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

67

I1 molecules, but insufficient for expression of the antiviral state. Therefore, interaction with a third protein is surmised. The gene for this third factor is also located on human chromosome 21 (49). IFN-y causes dimerization of recombinant soluble receptor (50).Analysis of the crystal structure of the complex revealed that the two IFN-yRa chains in this complex remain separated from each other (11). Studies aimed at locating the extracellular IFN-yRa domains responsible for interaction with ligand and with the @chain have been performed by replacing segments of the a-chain in human receptor with corresponding murine counterparts (51).This has led to a provisional model in which both the species-specific N termini and the conserved C termini of the IFN-y homodimer bind to the extracellular domains of an a-chain receptor pair. The model also predicts that multiple segments of the extracellular a-chain cooperate to bind with p-chain. Another approach to elucidate interaction of IFN-y with its receptor has consisted of studying binding of a-chain peptide fragments with IFN-y (52).These studies have led to the suggestion that IFN-y recognizes a binding site in the cytoplasmic domain of the a-chain. Aside from the ubiquitous high-affinity IFN-y-binding protein, an additional distinct low-affinity receptor has been reported to occur on cells of the mononuclear phagocyte lineage (53). Expression of IFN-y-binding sites on cells of the monocpc lineage is regulated by other cytokines. A twofold increase was recorded after exposure to TNF-a or IL-1 (54). Such upregulation was found to be associated with increased IFN-y-induced expression of MHC class I1Ag (54).Upregulated transcription of the IFN-y receptor gene has also been reported in a human carcinoma cell line treated with TNF (55) and in a monocytic cell line treated with TNF-a or IL-6 (56). Hormonal control of IFN-yR expression is evidenced by studies on mRNA levels in maternal uterine and embryonic hematopoietic cells during gestation in mice (57). The distribution of IFN-yR mRNA among leukocytes in cycling uteri was found to be low and fairly uniform among the cell populations, with macrophages exhibiting a maximal expression level in diestrous uteri. In pregnant uteri, hematopoietic cells were found to proliferate and express IFN-yR mRNA to degrees that vaned among different cell types; concomitantly, these cells expressed activation markers.

B. SIGNAL TRANSDUCTION, TRANSCRIPTION FACTORS, AND RESPONSE ELEMENTS (FIG.2) Binding of IFN-7 to the membrane receptor complex transmits signals to the cytoplasma and nucleus by the Jak-STAT mechanism of signal transduction (58).This mechanism involves a cascade of tyrosine phosphor-

68

ALFONS BILLIAU

SECRETED PROTEINS

CYTOPLASMIC

non-receplor TPKs

FIG.2. Schematic model for ligand-receptor interaction and signal transduction in IFNy-responding cells. The IFN-y homodimer binds to the IFN-yRa chain which, as a result of this binding, also dimerizes and associates to two molecules of the IFN-yRP chain. At the intracellular side, the IFN-yRdFN-yRP chain complex associates with the nonreceptor tyrosine phosphohnases JAK-1 and JAK-2, thereby triggering their (auto)phosphorylation. JAK-2 catalyzes phosphorylation of STAT-1 [formerly STAT-91, 91 kDa E a subunit, or gamma-activator factor (GAF)]. STAT-1 translocates to the nucleus where it binds to interferon-sensitive response elements (ISRE) and gamma-responsive elements (GRR) in the promotor regions of several IFN-y-responsive genes. Derepression of these genes leads to formation of mRNAs and proteins, some of which are located intracellularly (e.g., 2',5'oligoadenyl synthase), others in the membrane (e.g.,class I1 MHC gene products), and still others are secreted extracellularly (e.g., IFN-y-induced cytokines).

ylations on cytoplasmic proteins, the so-called STAT (for signal transducer and activator of transcription) proteins. The STATs were discovered in the context of investigations aimed at elucidating the mode of action of interferons but were later found to play roles in signal transduction of many other cytokines and growth factors (59-62). In essence, the intracellular part of the IFN-.), receptor protein does not have catalytic activity but interacts with nonreceptor-type tyrosin protein kinases belonging to the Jak (for Janus kinase) family. These kinases were originally discovered by low stringency hybridization with probes to domains of known protein kinases and by PCR with primers based on highly conserved sequences of known protein tyrosine kinases. Three such Jaks (TYK-2, JAK-1, and JAK-2) have so far been implicated in receptor triggering by IFNs. JAK2 is required for cells to respond to IFN-7 but not to I F N - d P (63). JAK-1 seems to be associated with the IFN-y receptor before ligand

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

69

binding, whereas JAK-2 is recruited into the ligand-receptor complex within seconds after IFN-y is added to cells (64). The mechanism of phosphorylation of and by Jaks has not completely been defined. A working hypothesis is that Jaks, by binding to the dimerized receptor complex, acquire the ability to phosphorylate each other and the receptor peptides. This phosphorylated receptor/Jak complex may next constitute the catalyst for phosphorylation of the STATs. Tyr,,, on one of the IFN-y receptor chains is essential for activation by IFN-y and a phosphopeptide fragment containing Tyr440binds to STAT-1 (65). The Jaks seem to lack an src homology 2 (SH2) domain (i.e., a phophotyrosine-binding domain), a characteristic of other nonreceptor tyrosine kinases and their substrates. STAT proteins do contain SH2 domains. Experiments in which the SH2 domains between STAT-1 and STAT-2 were swapped demonstrated that these domains specifically direct selection of a particular STAT for phosphorylation at the IFN-y receptor as well as subsequent dimer formation between STAT proteins (66). Members of the STAT family vary in molecular mass from 84 to 113. They possess SH2 and SH3 domains. Activation requires phosphorylation on specific tyrosines followed by formation of dimers or multimers that are translocated to the nucleus where they bind to specific promotor regions (response elements) and act as transcription factors. The system of STATs and response elements involved in IFN-.)Iaction resemble that involved in IFN-alp action (67). The IFN-alp-connected transcription factors encompass ones that are inducible [E (ISGF3) and M (ISGF2 = IRF-l)] and one that is constitutive [C (ISGFl)].These interferon-stimulated gene factors are composed of subunits, e.g., E is composed of E a and Ey. Ey (STAT48) is the actual DNA-binding protein, requiring one or more of the other STATs to be translocated to the nucleus. E a is itself composite, containing three peptides (STAT84 or STAT-la, STAT91 or STAT-lp, and STAT113 or STAT-2). Following interaction of IFN-a with the receptor complex, the three components of E a are phosphorylated and subsequently translocated to the nucleus. In response to IFN-y, only STAT91 [formerly called gamma-activated factor (GAF)] is phosphorylated on tyrosine and translocated. Thus, the two types of interferons trigger phophorylation of different combinations of the same latent complex of cytoplasmictranscription factors, which then interact with different sets of genes (68, 69). Depending on whether the transcription factor complex activated by IFN-a or -p contains the DNA-binding elements IRF-1 or IRF-2 (interferon-regulatory factors), it stimulates or inhibits transcription of those genes that contain interferon-sensitive response elements ( ISREs). IRF-1 and IRF-2 are DNA-binding factors originally identified as regulators of the IFN-a and IFN-P genes (70, 71).They are structurally related,

70

ALFONS BILLIAU

particularly in their amino-terminal regions which confer specificity for DNA binding. Both factors bind to the same region in the promoters of the IFN-a and -/3 genes, but also to ISREs in the promoters of several IFN-inducible genes. Thus, IRF-1 and -2 are involved in regulating the induction as well as the action of IFN-a and -/3. IRF-1 functions as a transcription activator, whereas IRF-2 suppresses the action of IRF-1. In the context of cell growth regulation, IRF-1 has growth-inhibitorywhereas IRF-2 has growth-promoting activity. The genes for IRF-1 and IRF-2 themselves are inducible by infection of the cells with viruses or exposure to cytokines (IFNs, TNF-a, IL-1, IL-6, or LIF) or prolactin. Some IFN-y-responsive genes contain ISREs; it seems that, in these genes, the presence of ISRE is sufficient for responsiveness. Other genes known to be triggered by IFN-y, e.g., class I1 MHC and Fc-receptor genes, do not contain ISREs, but do contain other IFN-y-specific response sequences, e.g., the gamma response region (GRR) in the promotor of FcyRI (72), the yRE-1 element in the promoter of the mig gene (mig is the monokine induced by IFN-y, a member of the interleukn-8 family) (73) and the y-activated sequence (GAS) in the IRF-1 gene (71). Signaling systems other than the ISGFs may be involved in IFN-y action. Many studies have indeed suggested involvement of PKC and/or Ca2+ calmodulin in the action of IFN-y. For instance, induction of MHC antigen expression by IFN-y was found to be associated with rapid increases in PKC activity (74) and to be blocked by H7, an inhibitor of PKC (75). Furthermore, activation of PKC following exposure to interferon was found to require increased [Ca"], and activation of the calmodulin system as well as increased PI breakdown and tyrosine phosphorylation (76). Insofar as PKC is involved, tyrosine phosphorylation is probably a proximal event (77). There is speculation that the IFN-y protein, in addition to using the membrane receptor-activated phosphorylation cascade, also interacts with an intracellular or even intranuclear receptor. Binding of IFN-y to its receptor is followed by endocytosis of the ligand/receptor complex (78, 79). Inhibition of receptor-mediated endocytosis, e.g., by acidification of the medium, inhibits induction of MHC antigen induction by IFN-y, indicating that internalization of the ligandheceptor complex contributes to IFN-y activity (80). Studies with murine macrophages have confirmed uptake of the receptor/ligand complex and have also indicated that the receptor is neither up- nor downregulated by ligand binding, but is being recycled (81).Several observations have given ground for speculation that transportation of IFN-y from the cell membrane to some nuclear receptor system contributes to the establishment of biological effects of IFN-y on

INF-1: BIOLOGY AND ROLE IN PATHOGENESIS

71

cells. Such observations include demonstration of biological activity of human IFN-y in murine cells following microinjection (82) or intracellular production (83). Similar results were obtained in studies involving treatment of macrophages with liposome-encapsulated IFN-y (84). Accumulation of IFN-.)Iin the cell nucleus, as demonstrated by autoradiography and direct immunofluorescence, gives further credibility to this concept (85). Also, the IFN-y protein contains two conserved stretches of basic amino acids that resemble the NLS of large T antigen of simian virus 40 (86). These stretches are exposed on the surface of the IFN-y molecule. Antibodies directed against the amino-terminal stretch were found to neutralize biological activity of IFN-y without preventing its binding to the membrane receptor (86). Removal of the C-terminal NLS-like stretch was likewise found to result in abolishment of biological activity (87).

C. BIOCHEMICAL CHANGES I N CELLS 1. Tryptophan Metabolism One of the proteins induced by IFN-y is the enzyme indoleamine-2,3dioxygenase (IDO). I D 0 is the first enzyme of the kynurenine pathway that links tryptophan to alanine and acetyl CoA. It catalyzes the oxidative cleavage of the pyrrole ring in tryptophan to yield N-formyl-kynurenine. Kynurenine is transformed by enzymes in liver and brain to alanine and metabolites, some of which, such as quinolinic acid, have neuroactive or toxic potential. Induction of I D 0 by IFN-y explains why exposure to endogenous or exogenous IFN-y is often associated with decreased serum tryptophan levels and increased levels of kynurenine in serum and urine. It has been argued that quinolinic acid, released as a result of exposure to endogenous IFN-y, may be responsible for central nervous system manifestations in AIDS patients, in patients receiving cytokine therapy, or in patients with so-called cytokine release syndromes, such as the first dose reaction in renal allograft recipients given anti-CD3 antibody as the immunosuppressant (88, 89). Induction of I D 0 by IFN-y may indirectly function as a scavenger mechanism of the superoxide anion because this anion is used in the IDOcatalyzed conversion of tryptophan to N-formyl-kynurenine. The cell- and tissue-damaging effect of IFN-y may thereby be dampened. IFN-y-induced depletion of tryptophan has also been speculated to contribute to te antiproliferative effect of IFN-y. In a panel of cultured human ovarian tumor cell lines, induction of I D 0 could not be correlated with the sensitivity of the lines to the antiproliferative effect of IFN-y (90). However, IFN-y was found to also induce the synthesis of trypto-

72

ALFONS BILLIAU

phanyl tRNA synthetase, which might in resistant cell lines blunt the consequences of tryptophan depletion for overall protein synthesis. 2. Generation of Reactive Oxygen IFN-y is well known to potentiate respiratory burst responsiveness of macrophages to stimulants, resulting in increased production of highly reactive oxidants such as HzOzand the superoxide anion (0;)( 5 ) . These effects of IFN-7 are believed to be due to regulation of the transcription of genes coding for enzymes of the NADPH oxidase system (91). This membrane-associated system is dormant in resting cells, but becomes activated during phagocytosis or on interaction with certain soluble stimuli. In addition, IFN-7 also stimulates production of nitrogen monoxide (NO), which in turn may react with HeOBto generate reactive oxygen. 3. Pteridin Metabolism: Generation of Tetrahydrobiopterin Tetrahydrobiopterin is a limiting factor in the synthesis of NO (see below). Pteridins are synthesized from GTP, the first step being conversion to 7,8-&hydroneopterin3’-triphosphate by the enzyme GTP cyclohydrolase I (92).The synthesis of this enzyme is upregulated by IFN-y. Further steps leading to production of tetrahydrobiopterin are catalyzed by enzymes that appear not to be under control of cytokines. The tetrahydrobiopterin concentration in cells is therefore mainly dependent on the level of GTP cyclohydrolase, although part of the primary product of this enzyme may also leave the cells after dephosphorylation. The latter escape pathway explains the increased levels of neopterin in urine and tissues of patients with inflammatory and infectious diseases. 4. Generation of Nitrogen Monoxide Various cell synthesize and release endogenous NO as a short-distance and short-lived messenger (93). NO is produced by NO synthases, which convert L-arginine to L-citrulline and NO. NADPH is consumed in the reaction in stoichiometric proportion and tetrahydrobiopterin is required as a cofactor that does not seem to be involved stoichiometrically but is apparently required for each step of the reaction. The NO synthase reaction can be blocked, both in vitro and in vivo, by NG-monomethyl-L-arginine (L-NMMA).The use of L-NMMA has helped a great deal in defining the role of NO in physiological and pathological processes. Production of NO is controlled by IFN-y via two pathways: (a) regulation of the synthesis of the inducible isoform of NO synthase; and (b) control of the production of the cofactor, tetrahydrobiopterin [reviewed in ref. (92)]. NO synthase occurs in at least three isoforms. Two forms are “constitutive,” being produced mainly by endothelial cells and platelets on the one

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

73

hand and by neurons on the other hand. The constitutive NO synthase is tightly associated with calmodulin, which allows for regulation of the enzyme's activity by intracellular Ca'+ ion concentration. In particular, activity of the endothelial constitutive NO synthase is regulated by pulsatile blood flow and shear stress and can be enhanced by neurotransmitters, e.g., acetylcholine, bradykinine, and ATP. Certain neurons contain constitutivetype NO synthase [for review see Ref. (94)]. One location is the myenteric plexus of the gastrointestinal tract; these neurons govern peristaltic activity. Other locations are the nerve plexuses of brain and penile arteries. Regulated NO release by these neurons accounts for relaxation of cerebral arteries and for erection. NO is considered to be a neurotransmitter. However, in contrast to classical neurotransmitters, which are released by exocytosis of synaptosomes, NO is synthesized on demand and diffuses out of the cell.

a. NO Induction by ZFN- y. The inducible NO synthase occurs in vitro in mononuclear phagocytes, granulocytes, fibroblasts, Kupffer cells, hepatocytes, endothelial cells, vascular smooth muscle cells and, probably, in many other cell types. The enzyme is not detectable in uninduced cells and differs from the constitutive type in that its induced release requires protein synthesis, that it can act in the absence of Ca2+,and that its activity is stimulated by flavin adenine dinucleotide and reduced glutathione. In macrophages and fibroblasts, the known natural inducers of the enzyme are cytokines and endotoxin. Maximal activation of the cells to produce NO via this pathway is obtained by their exposure to a combination of IFN-y and endotoxin, IFN-.)Iand TNF, or IFN-.)Iand IL-1. Macrophages of mice with a targeted disruption of the IFN-y receptor gene were expectedly found not to produce NO in response to IFN-y. Other cytokines, TNF and IFN-dP, could only marginally substitute for IFN-.)I,indicating that IFN-y is indeed the major cytokine controlling the NO-generating pathways (95). The promotor of the inducible NO synthase gene contains a response element for IRF-1, a transcription factor affecting IFN-.)I-as well as IFNdp-responsive genes. Peritoneal macrophages of IRF- 1'"' mice were found to produce little or no NO on stimulation with IFNs, and mRNA levels for the synthase remained unchanged. The IRF-1 defect in these mice was associated with a severely reduced resistance to infection with BCG (95). However, whether defective NO formation, rather than defects in other IRF-1-controlled events, explains this reduced antimycobacterial resistance remains to be defined.

b. Occurrence and Eflects of NO in Biological Systems. NO exists in three equilibrating redox states (96)-nitroxyl anion (NO-), neutral nitric

74

ALFONS BILLIAU

oxide (NO', single electron in an anti-binding orbital), and the nitrosonium ion (NO+)-that have different biological activities. NO' reacts with (di)oxOh-, and Oi'-) to generate OONOygen in its various redox forms (02, (peroxynitrite) and NO,. NO' also forms complexes with transition metal ions in metalloproteins, heme-containing ones such as hemoglobin, and proteins containing iron-sulfur centers such as those involved in mitochondrial electron traffic (aconitases and complex I and 11). The formation of iron-dinitrosyl-dithiolate complexes in mitochondrial enzymes accompanies cytotoxicity-inducing effects of IFN-y. NO+ nitrosates organic molecules at s, N, 0, or C centers. In biological systems, this may result in amine deamination (R-NH2 -+ R-H), N-nitrosylation (RR'-NH -+ RR'-N-NO), or thionitrite formation (RS- -+ R-S-NO). In biological systems nitrosothiols predominate over other nitrosated organics. The existence of these different equilibrating pools of NO provides the means for regulation of transport, lifetime, and biological availability of NO. The physiological receptor for NO is the soluble guanylyl cyclase, which generates cGMP. The enzyme is activated by binding of NO to its heme iron. Effects of NO mediated via this signaling pathway are vasodilation, platelet inhibition, cell adhesion, and neurotransmission. By nitrosylating free thiols, NO can regulate the activity of certain enzymes and thereby exert physiological regulatory functions. Thus, the reaction of NO with cell surface thiols has been associated with antimicrobial effects, modulation of ligand-gated receptor (NMDA) activity, and alterations of smooth muscle function. The antimicrobial effect of NO may also be due to loss of iron from infected host cells (97). In uitro, cells producing NO, following exposure to IFN-y, can die a suicide-like death (98, 99), in particular when they have no access to glucose or when glycolysis is blocked, so that the respiratory chain is the the only pathway for ATP generation. However, aside from being cytotoxic, NO can also exert cytostatic activity by causing arrest of DNA synthesis. This inhibition precedes inhibition of mitochondrial respiration and may be due to impairment of the enzyme ribonucleotide reductase, which contains at least three targets for NO: a tyrosyl radical, cysteines, and an iron center. Macrophages in which NO production is induced by IFN-y have cytocidal activity toward other cells in which no NO can be generated. Cell death in this case can be due to interruption of mitochondrial respiration (loo), but generation of peroxynitrite may also be involved. c. The Role of IFN-y-Induced NO in Disease. Production of NO in phagocytic cells is associated with reduced survival of ingested microorganisms. The role of the IFN-y-induced NO synthase is therefore assumed

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

75

to consist in augmented defense against infection with bacteria, molds, or protozoa [for review, see Ref. (lol)]. NO released by activated macrophages is also held responsible for the ability of these cells to kill tumor cells and is therefore considered a defense mechanism against cancer. The usefulness of cytokine-induced NO release by other cells (fibroblasts, Kupffer cells, hepatocytes, etc.) is largely a matter of speculation, The cytotoxic effect of induced NO may also cause undesirable cell and tissue damage. In view of the effects of NO on the vasculature, one may propose that NO produced by IFN-y-activated macrophages is in part responsible for local vasodilation in the inflammatory focus and possibly for the hyperdynamic circulation response associated with inflammation. In animal models, beneficial effects of induced release of NO have been revealed by the administration of L-NMMA, which blocks synthesis of NO [for review see Ref. (93)]. Thus, administration of L-NMMA was found to potentiate endotoxin-induced intestinal damage as well as liver damage in a sepsis model and to abrogate the protective effect of a low endotoxin dose against subsequent challenge with a lethal dose (102). However, as a contrast, hypotension also occurring as part of the septic shock syndrome seems in part to be due to excessive production through the endotoxidcytokineinduced pathway of NO. There is circumstantial evidence that IFN-y is involved in the pathogenesis of multiple sclerosis (see below). Therefore, it is of interest to note that NO release may be involved in demyelinative processes: microglial cells have the ability to kill oligodendrocytes in culture and the lysis was found to inhibited by antagonists of NO (103). Similarly, NO induced by IFN-y or other cytokines may be involved in the pathogenesis of insulin-dependent diabetes in NOD mice (see below). In vitro, IL-lP can kill P cells in certain circumstances. This cytocidal activity was found to depend on generation of NO as it could be prevented

by L-NMMA (104). Finally, NO production has also been found to contribute to the antiviral effects of IFN-y in macrophages infected with ectromelia, vaccinia, or herpes simplex-1 viruses (105). In the macrophage-like cell line, RAW 264.7, IFN-y-induced NO was found to inhibit vaccinia virus DNA replication, late viral protein synthesis, and particle assembly, but not to affect early protein synthesis. Virus replication was inhibited not only in the iNOS-producing macrophages themselves but also in bystander cells cocultured with the IFN-y-pretreated macrophages (106). Remarkably, this apparently paracrine effect was not seen when the cells were separated from each other by a semipermeable membrane, suggesting that, in addi-

76

ALFONS BILLIAU

tion to NO, cell-cell contact is also necessary for transfer of the antiviral state. 5. Other IFN- y-Induced Proteins Aside from the IFN-y-induced proteins and enzymes already mentioned, several others have been reported. Some of these proteins have been found serendipitously or by molecular screening strategies such as subtraction cloning. Often, the function of these proteins has remained unknown for at least some time. One example is the protein IP-10, originally identified by cloning a mRNA from cells treated with IFN-y. At the time the sequence of IP-10 became known, no sequence-homologous proteins were known. Only when interleukin-8 and related molecules were isolated did it become evident that IP-10 belongs to the large family of chemokines (107). Another example is the murine protein Mg21, identified by subtraction cloning of cDNA from peritoneal macrophages treated with IFN-y (108). This protein appears to belong to a family of GTP-binding proteins that also encompasses IRG-47 (log), which is induced by IFN-y in B lymphocytes, and Mx, which is an IFN-dp-induced protein responsible for the antiviral effect against influenza virus. GBP-1 is a GTP-binding protein induced by IFN-y in human fibroblasts (110) and in mouse macrophages (111);however, its sequence is unrelated to those of IRG-47 and Mg21. V. Biological Effects on Cells and Tissues

A. THEANTIPROLIFERATIVE EFFECTOF IFN-y IFN-y exerts a mild antiproliferative effect on most cell types (with the exception of at least some populations of activated T cells or T cell lines). One example of cells that are relatively sensitive to growth inhibition by IFN-y is normal keratinocytes (112). As is the case with many other actions of IFN-y, a synergy with TNF was found. IFN-y is also a potent inducer of differentiation in keratinocytes. These effects are characterized by a dramatic reduction in expression of the mRNA of two growth-regulatory genes, cdc2 and E2F-1, and an increase in the expression of squamous cell-specific genes (113).The cytostatic effect of IFN-y on keratinocytes may pertain to the epithelial atrophy seen in the skin of patients with GVH disease. A dose-dependent and reversible antiproliferative effect of IFN-y was seen to occur in a human submandibular salivary gland epithelial cell line (HSG). Concurrently, the expression of class I1 MHC gene product and ICAM-1 was increased. The effect may play a role in Sjogren’s syndrome, which is characterized by an inflammatory cell infiltrate associated with

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

77

progressive atrophy of acinar secretory epithelium and loss of secretory function ( 114).

B. EXPRESSION OF MEMBRANE PROTEINS One of the best documented actions of IFN-y is the induction of MHC class I1 antigens on many but not all types in culture. IFN-y thus has the ability to enhance or to induce these cells’ ability to present foreign antigens. MHC class I antigen expression can also be enhanced under the influence of IFN-y. Cells in which this effect occurs may thus become better targets for cytotoxic T cells recognizing viral, tumor, or autoantigens present in such cells. In rats given IFN-y injections, MHC class I1 antigen expression is seen to occur in all organs including the brain. This induced expression is rapidly lost after interferon withdrawal except in keratinocytes, in which it lasts for several days (115).Some cell types are completely resistant to induction of class I1 antigens, e.g., capillary endothelium, neurons, and endocrine islet cells. Human islet cells do, however, express class I1 antigens after treatment with IFN-y + TNF (116), an effect that may play a role in the pathogenesis of autoimmune insulin-dependent diabetes. IFN-y regulates the expression on phagocytes of the high-affinity Fcy receptor I (FcyRI). Resting monocytes express both FcyRI and the lowaffinity receptors FcyRII and F c ~ R I I I ;resting polymorphonuclear cells express only the low-affinity receptors. IFN-y induces expression of FcyRI on neutrophils (117) and augments its expression on mononuclear phagocytes (118). Ligand binding to FcyRs is widely recognized to stimulate effector functions of phagocytes, such as phagocytosis, tumor cell killing, and inflammatory mediator release. Therefore, augmented F q R expression is one of the pathways by which IFN-y can act as a proinflammatory cytokine. Isolated rat brain microglia also displays enhanced expression of Fc receptors after treatment with IFN-y, TNF-a, or IL-1. Remarkably, the combination of IFN-y and TNF-a inhibits Fc receptor expression (119). Certain viruses, e.g., Dengue virus, enter host cells via specific antibodies bound to FcyR. As a consequence, IFN-y paradoxically augments virus infection in mononuclear cells in the presence of antibody (120). IFN-y augments expression of FCERon the human mononuclear cell line U937 (121) and on platelets (122). In view of the important role of IgE in resistance to parasitic diseases and in type I allergic reactions, these effects of IFN-y need to be taken into account when considering the role of IFN-y in these diseases (see below). IFN-y is among the cytokines that augment expression of the adhesion molecule ICAM-1 on various cell types, including cultured endothelial cells (see below) and epidermal keratinocytes (123),resulting in increased

78

ALFONS BILLIAU

adhesiveness for leukocytes expressingthe integrin LFA-1. The significance of this effect may be that IFN-y produced early in an aspecific inflammatory focus (e.g., by NK cells) is coresponsible for firm adhesion of granulocytes to endothelial cells in postcapillary veins as a prelude to their spreading and diapedesis. Similarly, this mechanism is believed to promote proximity of dendritic cells, epidermal keratinocytes, and lymphocytes during antigen presentation in the skin (123). Another important membrane protein induced by IFN-y is the B7 antigen, whose ligand on T cells is the CD28 molecule (124). The presence of B7 on antigen-presenting cells is indispensable for them to avoid delivering an anergizing signal (125). IFN-y has also been reported to augment expression and/or shedding of tumor-associated antigens by tumor cells, thereby modulating their targetability for the corresponding antibodies or sensitized T cells (126128). Bone marrow-derived macrophages express a protein with an epitope recognized by a monoclonal antibody specific for a peptide of the mycobacterial heat shock protein, hsp60. Exposure of the cells to IFN-.)Iwas found to result in increased expression of this cross-reactive epitope. Antibodies recognizing heat shock proteins are believed to play a role in autoimmune diseases. Therefore, the augmenting effect of IFN-y on expression of hsp60 cross-reactive proteins may be of relevance to the pathogenesis of such diseases (129). C. EFFECTS ON MONONUCLEAR PHAGOCYTES

IFN-.)I has long been recognized as the foremost important cytokine converting macrophages from a “resting” to an “activated’ state. “Activation” is a rather indiscriminately used term that has meaning only if placed in a context of a well-defined function or functional ability that is considered. The mononuclear phagocytes (MPC) population, to which macrophages belong, comprises cells in different stages of differentiation and maturation, i.e., bone marrow precursors, blood monocytes, and different types of tissue macrophages (e.g., connective tissue histiocytes, alveolar macrophages, Kupffer cells, exsudate macrophages, microglial cells, osteoclasts, etc.). These stages are usually considered as steps in a process that, although regulated by environmental signals, is in essence irreversible. Activation states of tissue macrophages, as a contrast, are mostly seen as reversible changes determined by the temporally changing tissue microenvironment. IFN-y and other cytokines, being constituents of the cellular microenvironment, play an important role both in the differentiation and maturation of MPCs and in activation of tissue macrophages.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

79

Circulating monocytes, when placed in culture, undergo apoptosis unless provided with certain stimuli, e.g., LPS andlor certain cytokmes. Typically, human blood monocytes will survive for more than 7 days in the presence of pure MCSF, whereas in its absence, they will die within 24 hr by apoptosis and secondary necrosis. In the presence of MCSF, the 100% surviving monocytes become progressively resistant against withdrawal of the growth factor and can subsequently be activated by exposure to a stimulus (e.g., serum-activated zymosan) to become biologically active (as evident from adherence, phagocytic, and respiratory activity). In the presence of not only MCSF but also IFN-y, resistance to cytokine withdrawal develops more rapidly. Thus, although IFN-y cannot by itself replace MCSF, it can be seen as a synergist for MCSF to avoid apoptosis and promote maturation into a macrophage. However, remarkably, such monocytes that have avoided apoptotic death in the presence of both MCSF and IFN-y, as opposed to those that have maturated in the presence of only MCSF, do undergo apoptosis when subsequently exposed to an activating stimulus (130). The authors speculate that this yin-yang-type effect of IFN-y fulfills a useful function in host defense against infection in that it allows for rapid development of a microbicidal macrophage, but also for subsequent elimination of the macrophage that, by producing toxic metabolites, might otherwise inflict undue damage on the hosts own tissues. Macrophages activated by IFN-y or other agents differ from resident ones by enhanced endocytic capacity, as manifested by increased pinocytosis and phagocytosis via receptors for complement and IgG2a. Such enhanced endocytic capacity does not, however, apply to all ligands. Expression of Fc receptors for other Ig classes, for instance, has been shown to be reduced in the activated state. Significantly, macrophages activated by IFN-y have been found to have reduced ability to ingest a variety of obligately intracellular microorganisms, e.g., Rickettsiae, Trypanosoma cruzi, and Leishmania amastigotes. In the case of Leishmania in mouse macrophages, reduced binding was demonstrated to be due to reduced expression of lectin-like receptors for the organisms (131).Reduced uptake of intracellular parasites by IFN-y-activated macrophages may represent one of the mechanisms by which IFN-y contributes to both specific and aspecific host resistance to these pathogens. IFN-y also affects accessory cell function of mononuclear phagocytes. In vivo, endogenous IFN-y seems to promote helper and to counteract suppressive circuitry, as evident from experiments in which anti-IFN-y antibody was injected in carrier-primed mice whose spleen cells were subsequently immunized in vitro with carrier-haptene complex ( 132). However, IFN-y may also activate suppression circuits in mixed mononuclear cell populations (133, 134).

80

ALFONS BILLIAU

IFN-y regulates production of chemokines by macrophages. It is apotent inducer of IP-10 mRNA expression in human and murine macrophages (135, 136). IP-10 belongs to the chemokine-a family. Human IP-10 has been shown to exhibit chemoattractant activity for monocytes and activated T cells and to promote T cell adhesion to endothelial cells (107). On the other hand, IFN-y was found to suppress LPS-induced expression of certain other related chemokines, namely MCP-1 and KC (GRO/melanoma growth-stimulating activity) in mouse peritoneal macrophages (137). Similarly, induction of IL-8 by IL-2 or IL-1 in human monocytes was found to be inhibited by IFN-y (138). However, IFN-.)I is not to be regarded as a general inhibitor of IL-8 expression, because it synergizes with TNF to induce IL-8 in other cells. Clearly, IFN-y has the ability to either stimulate or inhibit production of proinflammatory mediators by macrophages. IFN-.)I induces production of C1 inhibitor protein (ClINH), both in mononuclear phagocytes in vitro and in the serum of patients given intravenous infusions in vivo (139,140). ClINH is a 105-kDa glycoprotein considered to be the major inhibitor of several serine proteases including Cls, Clr, kallikrein, and activated Hageman factor. Hepatocytes, activated mononuclear phagocytes, and endothelial cells constitute the major source of ClINH.

D. IFN-.), AND ANTIGENPRESENTATION IFN-7 can affect antigen presentation by augmenting expression of MHC class I1 molecules on membranes of professional antigen-presenting cells (APCs), but also by inducing de no00 expression on the surfaces of various other cell types, thus converting them to nonprofessional APCs. The consequences of these effects are subject of debate. In particular, MHC class II-restricted antigen presentation by nonprofessional APCs that outnumber the professional ones is speculated to result in antigenspecific T cell anergy or induction of suppressor cells. Thus, IFN-y was demonstrated to inhibit the ability of epidermal cells to present tumorassociated antigens for protective immunity, as well as the ability of the epidermal cells to elicit tumor-specific DTH in tumor-immune mice in vivo (141). One possible underlying mechanism for anergy to be induced by nonprofessional APCs may be that these cells lack critical costimulatory cell surface molecules such as ICAM-1 (142), B7 (125), or their ligands. These molecules do occur on professional APCs, and their expression is enhanced by IFN-.)I (124). T cell anergy has also been proposed to occur when antigen fragments occurring in high concentration in tissues are captured and presented in MHC class 11 context by T cells (143). However, there is no evidence that IFN-y would induce expression of MHC class I1 molecules on T cells.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

81

E. EFFECTSON ENDOTHELIAL CELLS Evidence from a large number of studies indicates that IFN-y induces expression of MHC class I1 antigen on endothelial cells in vitro and in viva Increased class I1 antigen expression following IFN-y treatment is also seen in monolayers of cerebral vascular endothelia, which are endowed with unique barrier properties, as they differ from extracerebral endothelia by the presence of interendothelial tight junctions, by a paucity of cytoplasmic vesicles, and the presence of specialized membrane transport systems. Exposure of monolayers of such cells to IFN-y is followed by morphological alterations associated with increased permeability of confluent monolayers to macromolecules (144). IFN-7, as well as IL-1 and TNF-a, augments expression of the adhesion molecule ICAM-1 on cultured endothelid cells of extracerebral (145) as well as cerebral origin (146) resulting in increased adhesiveness for leukocytes expressing the integrin LFA-1 (147). In the rat, an antigen (4A2) has been identified that is induced on cerebral endothelial cells after exposure to IFN-y. Triggering of the antigen with its corresponding monoclonal antibody was found to augment lymphocyte adhesion via LFA-1 and possibly other integrins (148). IFN-y enhances the capacity of endothelial cells for producing IL-1 in response to LPS (149-151). A similar effect has been noted in cultures of synovial cells (152). Chemokine secretion by endothelial cells is under regulatory control by cytokines, in particular IFN-y and TNF-a: barely any RANTES was found to be produced by human vascular endothelial cell cultures treated with IFN-y, TNF-a, or IL-lP. However, the combination, TNFa + IFN-y, was highly effective in inducing RANTES production: pretreatment with IFN-y sensitized the cells to induction with TNF-a. This regulatory control was found to be exerted at the transcriptional level (153).

F. SYNERGY BETWEEN IFN-y AND TNF Many instances have been reported in which IFN-y synergizes with TNF either in vitro or in uiuo: in vitro cytotoxicity for certain tumor cells, induction of microbicidal activity in macrophages, induction of NO release by various cell types, expression of cell surface adhesion molecules, in vivo antitumor effects, in vim induction of other cytokines, systemic toxicity, and lethality (for references, see relevant sections). The subcellular mechanism(s) underlying this synergy are poorly understood. IFN-y has been found to augment expression of TNF receptors by certain cell lines (154-156). However, this is not the general rule and cannot explain all synergistic effects described. In explanted murine perito-

82

ALFONS BILLIAU

neal macrophages, IFN-y, when added at relatively high concentrations, was found to inhibit rather than enhance expression of TNF receptors: in freshly explanted macrophages it prevented the appearance of receptors, and in mature macrophages it downregulated receptors already expressed (157). This inhibitory effect was not considered to be contradictory to the synergy between TNF and IFN-y because such synergy was only seen at lower doses of IFN-y than those inhibiting TNF receptor expression. This dose-dependent difference in interaction between the two cytokines may be of particular importance for the interpretation of often contradictory effects of both cytokines seen in inflammatory models in uiuo. It should also be mentioned that in these studies no distinction was made between effects of IFN-y on the low- versus the high-molecular-weight receptor of TNF. G. NATURAL IFN-y ANTAGONISTS

Several cytokines interact with the IFN-y system in a synergistic fashion either by enhancing its production or its action (Fig. 3). The biological significance of these synergisms is commented on in several sections of

GISTS

I

1 TNF 1 IL-12 1 'w

1 IL-2 1

FIG.3. Synergists of IFN-y. Several cytokines can reinforce the production or the action (or both) of IFN-y. IL-12 and IL-2 stimulate IFN-y production by, respectively, NK cells and T cells. The monokines, TNF and IL-1, reinforce the actions of IFN-y, in particular those exerted on mononuclear phagocytes and endothelial cells, but also those exerted on thyrnocytes and lymphocytes.

83

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

this review. On the other hand, several cytokines act antagonisticallytoward IFN-y and can thereby counterbalance its disadvantageous actions (Fig. 4). 1. Interleukin-4 IL-4 [for review see Ref. (158)] is produced by stimulated T helper cells, in particular the TH2 subset. It affects mainly B and T lymphocytes. Thus, IL-4 stimulates growth and membrane receptor expression on B lymphocytes elicited by other agents. It also acts as an isotype switch factor in favor of IgE and IgGl in the mouse and of IgE in man. IL-4 may also play a role in the development of B cells by promoting differentiation of stem cells into B lymphocyte rather than myeloid lineage. IL-4, in conjunction with other cytokines, favors the growth of TH2 cells, the subset which is also the source of IL-4. It appears to have no effect on proliferation of TH1 cells but inhibits their IFN-y production. IL-4 does promote growth and differentiation of cytotoxic T cells and plays an important role as a growth inhibitor for immature thymocytes and a maturation promoting factor for CD4+lCD8' cells in the thymus (159). In several systems, IL-4 and IFN-y have been found to have opposite effects and to antagonize each other when they are both present. Typical effects of IFN-7 on monocytes and macrophages (induction of IFN-yIFN-Y ANTAGONISTS

u 0 0

0

CTL dilferentiation lsotype switch Thymocyte maturation

/

1

1 I IL-10

--

Killing

1

..I

Antiviral

..........

FIG.4. Antagonists of IFN-y. Both the production and the actions of IFN-y are kept in check by several other cytokines, which have their origin in lymphocytes (IL-4, IL-10) as well as in fibroblasts and mononuclear phagocytes (IFN-dB and TGF-B1).

84

ALFONS BILLIAU

responsive genes, HzOzproduction, and intracellular antimicrobial activity) are counteracted by IL-4 (160, 161), although synergy has also been reported (162). Also, the antiviral affect of IFN-y on L929 cells has been found to be antagonized by IL-4 (163).IL-4 inhibits induction of chemokines by IFN-y (153, 164). In reverse, typical IL-4 actions on B lymphocytes (isotype switch in favor of IgE) and T lymphocytes (inhibition of thymocyte proliferation) are counteracted by IFN-y (159, 165, 166). Finally, IL-4 antagonizes with IFN-y by inhibiting its production by peripheral blood mononuclear cells (167-169). The subcellular mechanism of antagonism between IFN-.)Iand IL-4 in monocytic cells involves inhibition by IL-4 of transcriptional activation of the IFN-y-inducible proteins as was shown to be the case for the chemokine IP-10. The target for inhibition seems to be the ISRE in the promoter of the IP-10 gene (164). However, rather than inhibiting activation of the ISRE-binding transcription factor, ISFGF-3, IL-4 seems to induce a distinct ISRE-binding factor that functions to inhibit IFN-y-driven, ISREdependent transcription (170). The production of IFN-y by antigen-stimulated CD4+ T cells depends on the presence of IL-2 during both the priming and the expression phases. In contrast, production of IL-4 requires IL-2 only during the priming phase. This difference in requirement of IL-4 and IFN-y production may be part of the mechanism directing the immune response toward TH1 or TH2 predominance. 2. ZFN-d/3 and TGF-PI Type I interferon (aor 0) counteracts induction by IFN-y of MHC class I1 antigens in murine macrophages (171,172). The antagonistic effect has also been observed in cultured human astrocytes (173)and astrocytoma cells, but not in human monocytes (174). As documented by nuclear runon experiments, the antagonistic effect is exerted at the transcriptional level (174). TGF-P1 is a 25-kDa, disulfide-linked protein that has either growthenhancing or growth-inhibitory properties depending on the cell type and the presence of other growth factors. TGF-P1 is secreted, as a latent protein complex, by a variety of cells including lymphocytes, platelets, and activated macrophages. Once activated, it binds to a complex set of receptor proteins present on many different cell types. On immune cells, TGF-P exerts a variety of mostly inhibitory effects: it inhibits IL-2-dependent T cell proliferation, expression of IL-2R, B cell proliferation and differentiation, IL-l-induced thymocyte proliferation, and cytotoxic lymphocyte generation. Although TGF-P1 can by itself induce transcription of cytokine

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

85

genes, it mostly inhibits production of cytokines, such as IL-lP, TNF, and IFN-y, induced by other agents such as LPS or PHA (175, 176). The mechanism of this inhibitory effect remains unclear; synthesis of PGE2 or CAMP,which have both been implicated in post-transcriptional control of cytokine expression, do not seem to be involved. TGF-P1 can also counteract IFN-y action as it was found to downregulate constitutive as well as IFN-y-induced expression of MHC class I1 antigens on a human melanoma cell line (177). Another IFN-y-controlled activity that is counteracted by TGF-P is induction of nitrogen oxide synthesis in macrophages (178). In reverse, IFN-y can antagonize actions of TGF-P. Thus, it has been reported that IFN-y reverses the stimulation by TGF-P of the collagen gene but not that of the fibronectin gene expression in normal human fibroblasts (179).

3. ZL-lo Interleukin-10 [for review, see Ref. (180)l was originally described as “cytokine synthesis inhibitory factor” produced by TH2 cells. However, monocytes and B lymphocytes can also produce IL-10. Production by mononuclear phagocytes seems to be induced mainly in an autocrine fashion by TNF-a, and may represent an important negative feedback pathway (181).Indeed, although IL-10, like all cytokines, possesses multiple biological activities, its main effect remains its capacity to inhibit the synthesis of IFN-y. In addition, however, IL-10 also counteracts the action of IFN-y, e.g., as it inhibits IFN-y-induced nitrogen oxide production (182). The powerful inhibitory effect of IL-10 on IFN-y production is illustrated by the observation that IL- 10 administration to mice can inhibit shock induced by SEB in mice (183). The antagonistic effect of IL-10 on IFN-y is reciprocal. IL-10 production by monocytes is inhibited by IFN-y (184). It seems likely, therefore, that the balance between IL-10 and IFN-y in the initial stages of an immune response is of crucial importance to determine the further course of the response. For instance, selective stimulation of IL-10 production has been proposed as a strategy used by certain microorganisms, in particular helminths and protozoans, to evade the IFN-y-mediated host defense. 4. Znterleukin-13

IL-13 is a TH2 cell-produced cytokine whose amino acid sequence is partly homologous to that of IL-4 [for review see Ref. (185)l.IL-13 and IL-4 also share some biological properties, one of them being a modulatory effect on activated macrophages. Treatment of activated macrophages with IL-13 reduces production of inflammatory monokines in response to

86

ALFONS BILLIAU

IFN-y or LPS. In addition, IL-13 also inhibits production of nitrogen monoxide by activated macrophages (186).

5. Tumor Necrosis Factor and Interleukin-6 TNF-a synergizes with IFN-y in many in vitro test systems. However, this is by no means a general rule. With regard to two important activities, enhancement of expression of MHC class I1 molecules and of Fc receptors, the combination of IFN-y and TNF-a has been reported to be less active than either cytokine alone, both in rat peritoneal macrophages and in microglia (119, 187). In fact, whether IFN-.)I and TNF-a! synergize or antagonize with each other may critically depend on the state of differentiation of the cells. Thus, TNF was found to enhance IFN-y-induced MHC class I1 molecule expression in undifferentiated macrophages and to inhibit such enhancement in mature macrophages (188). IL-6 has been found to act as an antagonist for the Toxoplasrnacidal activity of IFN-.)Iin murine peritoneal macrophages in vitro (189).Combining TNF-a with IL-6 pretreatment resulted in restoration of the Toxoplasmacidal activity, whereas addition of anti-TNF-a antibody to this combination resulted in enhancement of the IL-Bmediated impairmant of IFN-y function. It would appear, therefore, that in this system IFN-y and IL-6 interact at the level of TNF-a! triggering. H. IFN-y AND LYMPHOCYTES 1 . Eflects on T Cell Prolqeration and Apoptosis Whereas IFN-y acts as a mild inhibitor of proliferation for most cell types, it stimulates proliferation of mitogen-triggered primary T cells as well as a variety of T cell lines [see Ref. (190)l. Remarkably, the antiviral effect of IFN-y is not expressed in these cells. It has been considered that this exceptional situation is due to modulation of signal transduction in T cells and that enhancement of IL-2 production and IL-2R expression are involved (190). In studies using murine T cell clones, it was found that IFN-y exerts a slight suppressive effect on IL-2- and IL-4-mediated proliferation of TH2 but not TH1 clones (191, 192). In mice, IFN-y was shown to mediate the inhibitory effect of IL-12 on TH2 cell-mediated pathology (193). In fact, a basic difference between TH1 and TH2 cells appears to be the absence of the @chain of the IFN-y receptor in the TH1 population ( 194). One aspect of the regulatory effect of IFN-.)Ion lymphocytes is its ability to promote apoptosis under specific conditions. Thus, blockage of IFN-y inhibits cell death induced in effector T cells by TCR linkage in the absence of accessory cells (195). Furthermore, both in normal and in cultured malignant lymphocytes, IFN-y has been shown to exert contrast-

INF-y: BIOLOGY AND ROLE I N PATHOGENESIS

87

ing effects, i.e., apoptosis or proliferation, depending on the level of expression of IFN-y receptors: high-level expression is associated with an apoptotic response, low-level expression with a proliferative one (196). In vivo apoptosis of thymocytes after treatment with anti-CD3 antibody is more pronounced in mice that are deficient in expression of the IFN-yRa chain than in their wild-type counterparts, indicating that in this system IFN-y triggers an antiapoptotic pathway (197).

2. ZFN-y and the Generation of Cytotoxic Lymphocytes (CTLs) Exogenous IFN-y can augment the development of CTL activity in mixed lymphocyte cultures, and neutralizing antibodies to IFN-y can inhibit the development of CTL activity in antigen- or lectin-stimulated lymphocyte cultures [e.g., Ref (198)l. Anti-IFN-y was also found to inhibit both proliferation and activation of CTL in the primary in vitro mixedlymphocyte reaction ( 199, ZOO), an observation not subsequently confirmed under apparently similar conditions (201). A contribution of IFN-y to the generation of CTL may, in principle, result from stimulatory effects on mononuclear phagocytes, on helper T cells, or on CTL precursors, or from inhibitory effects on suppressor T cells. In human mixed-lymphocyte cultures, the augmentation by IFN-y of CTL generation was found to result mainly from a direct effect on CD8+ T cells (198). In a system suitable for the expansion of single murine CD4-/CD8+ T cells into clones, both IL-2 and IFN-y were found to be required (202). 3. Efiects of ZFN-y on B Cells Reports on the effects of IFN-y on B cells are somewhat contradictory. In early studies, recombinant murine IFN-y was found to possess B cell maturation factor activity for resting splenic B cells and the comparable B cell tumor line WEHI-279.1 (203).However, IFN-y was found to inhibit LPS-induced IgM production in murine spleen-derived B cells by reducing the number of IgM-forming cells and without affecting overall proliferation (204). In fact, in B cells, as opposed to T cells, the stage of differentiation seems to codetermine the type of response to IFN-7. Resting B cells seem to be unaffected by IFN-y, whereas preactivated B cells are inhibited from further differentiation. These cells also show increased expression of IFN-y receptors (204,205). Also, cultured normal murine pre-B cells (206, 207) or human pre-B cell lines (208) that are exposed to IFN-y undergo apoptosis, whereas human CD5+ chronic B lynphocytic leukemia cells are protected from apoptosis by IFN-7 (209). I. IFN-y AND ANTIBODY FORMATION IFN-y affects antigen-presenting cells, T cells, and B cells, each of which intervene in the complex chain of events that leads from antigen exposure

88

ALFONS BILLIAU

to antibody production. No wonder that experiments testing the effect of IFN-y on antibody formation have yielded disparate results depending on the experimental conditions-stimulus (antigen or mitogen), immunization system (invitro or in vivo),immunization schedule, and isotype considered. IFN-y augments expression of MHC class I1 antigen expression and may therefore be expected to facilitate antibody induction in systems in which antigen presentation is the limiting factor. In murine in vitru irnmunization systems testing for primary IgM antibody responses to sheep erythrocytes, blockage of endogenous IFN-y with antibodies that neutralize macrophage activation by IFN-y have indeed been found to reduce the amount of antibody produced (210). IFN-y was also found to be a necessary component of T cell-derived helper factors for antibody induction in in vitro immunization systems (211). In contrast to this immunization-promoting effect of endogenous IFN-y of which only minute amounts are actually detectable during the process, exogenous IFN-y suppresses early antibody formation (210). In cultured human PBMC, addition of IFN-y has been found to promote and anti-IFN-y antibody to inhibit spontaneous late production of IgG2; in PWM-stimulated cultures, exogenous IFN-y inhibited and anti-IFN-y stimulated early production of IgGl (212). In this system, IFN-y did not seem to act as a IgG2 switch factor because the IgG2-promoting effect disappeared when the culture system had been depleted of sIgG2' cells. However, in a culture system in which T cells are eliminated, IFN-y seemed to possess IgG2-switching activity (213). IFN-.)I has also been found to stimulate polyclonal Ig production by resting or activated human B cells (203, 214). IFN-.)I and IL-4 antagonize each other in a variety of systems, and it has become general knowledge that antibody responses depend on the balance between two categories of cytokines, IFN-y and IL-2 belonging to the first one (theTH1 cassette) and IL-4, IL-5, IL-6, and IL-10 belonging to the second one (the TH2 cassette). In this setting, the role of IFN-y consists in suppressing IgGl and IgE antibody formation and stimulating IgG2a antibody formation (215, 216). Following infection with influenza virus (217), the virus-specific IgGl response was found to be significantly higher in IFN-y-deficient than in normal mice, probably reflecting increased production of IL-4. Treatment with neutralizing anti-IFN-y antibody in mice vaccinated with influenza virus antigens resulted in increased levels of antigen-specific IgGl and IgE but reduced levels of IgG2 and IgG3 (218). In IFN-y receptor knockout mice, the IgGl antibody response to ovalbumin was not different from that in normal mice, but the IgG2a isotype was reduced (219).

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

89

When antigen in adjuvant (oil or alum) is injected, a very early change (Day 3) is the appearance at the injection site of IFN-?-producing NK cells. Only later (Day 7 ) do cytolune-producing T cells appear in draining lymph nodes. At this site, 1L-2 and IL-4 predominate over IFN-y (220). The possibility is considered that early IFN-y production by NK cell is in fact secondary to IL-12 production by macrophages that respond to the antigen and/or the adjuvant. Early IL-12 and IFN-y may therefore play the crucial role in directing the immune response toward TH1 or TH2 predominance. Association of increased IgE levels with imbalance between IL-4 and IFN-y has been suggested to be involved in the pathogenesis of elevated IgE levels observed in patients with hyper-IgE recurrent infection (HIE) syndrome (221), atopic dermatitis, or helminth infections. In HIE, increased IL-4 production has not been observed. Conversely, some but not all studies reported decreased production of IFN-y. Mitogen-driven IL-4 production by PBMNC of atopic subjects was found to be higher and production of IFN-y lower than that of PBMNC of normal subjects (222). In patients with helminth infections, IgE levels were found to correlate with increased IL-4 and decreased IFN--y production by parasite antigenstimulated lymphocytes (223). However, in a placebo-controlled trial, exogenous IFN-y failed to affect clinical parameters or IgE levels in patients with hay fever-type rhinitis due to ragweed allergy (224). In contrast, in a murine model for allergen sensitization, nebulized but not parenteral IFN-y was found to decrease IgE production and to normalize airway function (225). Elevated levels of IgE are among the immunological characteristics of chronic atopic eczema. IFN-y responses have been reported to be defective in these patients (226). However, other changes are diminished delayed hypersensitivity reactions, diminished in vitro responsiveness to mitogens and recall antigens, and a decreased proportion of CD8+ T cells. In a double-blind, placebo-controlled trial (227), replacement therapy with recombinant IFN-y has been found to result in improvement of the clinical parameters. However, therapeutic failures in children with severe refractory disease have also been reported (228). MEDIATOROF IMMUNE SUPPRESSION In a variety of experimental systems, IFN-y has been identified as a mediator of suppression of immune responses. GVH, a major complication of bone marrow transplantation, is associated with suppression of cellular immune responses, as evident from reduced proliferative responses of lymphocytes to mitogens. Addition of monoclonal antibodies against IFN-y has been shown to relieve suppression, implying that endogenous J. IFN-y

AS A

90

ALFONS BlLLlAU

IFN-y is involved (229).The target cell for IFN-y in this system is believed to be a so-called natural suppressor cell (230). Other suppressor circuits induced by IFN-y are those evident from inhibition of antibody production and general immunosuppression associated with parasite infections (see relevant sections). Suppression by IFN-y, as demonstrated by inhibition of proliferative responses of T cells, depends on the presence of mononuclear phagocytes. In fresh adherent mouse spleen macrophages as well as in macrophage hybridomas, the ability to suppress is MHC class I1 antigen-restricted and specifically depends on expression of I-J region-coded antigens, which are stimulated by IFN-7 (133, 134). A possible mechanism of suppression by mononuclear phagocytes is generation of HzOzand prostaglandins because both catalase and indomethacin can alleviate suppression (231). Boraschi et al. (232), on the other hand, found that IFN-7 reduces rather than stimulates murine macrophage suppressive activity by inhibiting prostaglandin E2 release and inducing IL-1 induction. Another pathway used by suppressor macrophages, which can be activated by IFN-y, is the generation of nitric oxide (233, 234). IFN-y also enhances release by mononuclear phagocytes of TGF-P (235),which is generally known as an antiinflammatory cytohne.

K. IFN-y A N D EPITHELIAL BARRIERS Epithelia represent an example of tight but flexible permeability barriers within the body. In an in vitro model, IFN-y was shown to modulate permeability of an epithelial layer. IFN-y was found to be rather unique in this respect, as the effect was not seen with IFN-a. Also, only the basolateral side and not the apical side of the epithelial layer was found to be responsive (236). Most epithelial cells contain cytokeratins, a family of proteins forming cytoskeletal filaments. The expression of at least one of these molecules, the acid cytokeratin K17, was shown to be enhanced in HeLa cells treated with IFN-y (237), and the promoter of this protein was found to contain three putative GAS elements (238). However, the exact mode of action of the promoter remains to be elucidated. L. IFN-y

AND

CONNECTIVE TISSUE

Maintenance of intact and remodeling of traumatized or inflamed connective tissue is increasingly recognized to be controlled in part by the cytokine network. The contribution of IFN-7 to this control mechanism derives at least in part from its ability to inhibit the synthesis of collagen and fibronectin by fibroblasts or chondrocytes in vitro (239, 240). This inhibition is associated with reduction in mRNA levels (241,242).Systemic

INF-7: BIOLOGY AND ROLE IN PATHOGENESIS

91

administration of IFN-y was found to inhibit collagen synthesis in murine models-tissue reaction to a subcutaneous foreign body (243) or to skin wounds (244) and to alleviate pulmonary fibrosis induced by bleomycin (245).

M. IFN-y A N D ADIPOCYTES Dissipation of fat stores during cachexia may in part be due to direct effects of IFN-y (and other cytokines) on adipocytes. For instance, treatment of cultured adipocytes to IFN-y results in a reduction of the amount of lipoprotein lipase activity releasable from the cells by incubation with heparin (246,247).This enzyme is responsible for hydrolysing the triglycerides in circulating lipoproteins and is a major determinant of fat accumulation in adipocytes. IFN-y has also been shown to reduce the rate of fatty acid synthesis (246,247)and to cause increased hydrolysis of endogenous intracellular triglyceride in adipocytes (247, 248). In accord with these effects, IFN-y was found to reduce the level of lipoprotein lipase and fatty acid synthetase mRNAs. However, the level of Ac-CoA carboxylase mRNA was found to remain unaffected and the level of hormone-dependent lipase mRNA was decreased rather than increased. Therefore, posttranscriptional as well as transcriptional regulatory pathways seem to be involved (247). N. IFN-y AND CENTRAL NERVOUSSYSTEM (CNS) CELLS IFN-y influences the expression of membrane molecules in CNS cells, as it does in many other cell types. Class I and I1 MHC antigens are barely detectable in the normal human CNS. By contrast, in active lesions of multiple sclerosis and certain other neurological diseases (e.g., AIDS dementia complex or Alzheimer’s disease), these antigens are prominently expressed on certain brain cells, especially microglia. Another membrane molecule induced by IFN-y is ICAM-1. In MS lesions, small CNS vessels, rnononuclear cells, and some glial elements contain ICAM-1 (249). Active CNS lesions in MS patients also contain cells producing several cytokines including IFN-y (250). The two main types of CNS targets for IFN-y are microglia and astroglia. Microglid cells fulfill such functions as antigen presentation, phagocytosis, cytocidal activity, and production of tissue-degradative activity. Astrocytes act as regulators of the ionic balance in the CNS and neurotransmitter distribution. When cultured in vitro, these cells spontaneously express MHC antigens. Exposure to IFN-y greatly enhances this expression. However, in astrocytes, as opposed to microglia, the effect of IFN-y can be counteracted by various factors including IFN-/3, IL-1, TGF-P, glutamate, and CAMPagonists, as well as contact with neurons. Normal human CNS

92

ALFONS BILLIAU

astrocytes, when put in culture, spontaneously express MHC antigens as well as ICAM-1. However, this expression is greatly increased in the presence of IFN-y (251-256). The enhancing effect of IFN-y on expression of class I1 MHC antigen, but not that on expression of class I MHC antigen or ICAM-1, was found to be counteracted by IFN-P (173, 257). The molecular mechanism of class I1 MHC antigen induction in rat astrocytes and microglia has been studied in some detail. The class I1 MHC gene promoter DRA contains four conserved cis-acting elements-W, XI, X,, and Y-necessary for constitutive and induced transcription. Each of these elements binds to DNA-binding factors that are present in various cells. Also, a non-DNA-binding protein, CIITA, is necessary for expression of class I1 MHC genes. Xz binds to a factor, IFNEX, which is induced by IFN-y in astrocytes (258). A related factor was found to be operational in class I1 MHC gene induction in microglia (259). In a murine oligodendroglioma cell line (MOCH-l), IFN-.)I was found to induce a morphological change from a small round cell with thin branches to a large fibroblast-like cell. IFN-y also stimulated markedly enhanced expression of the astrocyte marker protein GFAP (260). 0. IFN-y IN HEMATOPOIESIS Effects of IFN-y on hematopoiesis have been demonstrated in many studies using different experimental settings, which invariably employ one or several hemopoietic factors and or cytokines. The question as to whether the IFN-y effects depend on the presence of accessory cells (stroma cells, monocytes, or lymphocytes) has been a matter of controversy in early studies (261, 262) and can as yet not be considered resolved. Also, IFN-y induces or enhances production of various other cytokines, and its effects on hematopoiesis may therefore be indirect. For instance, the IFN-y-inducible chemokine IP-10 has been shown to inhibit colony formation from early hematopoietic progenitors, apparently by counteracting r-steel factor (rSLF) (263). 1. Eflect on Hematopoietic Progenitors and Myelopoiesis

In bone marrow cultures prepared from normal mice, IFN-y was found to inhibit granulocyte/macrophage colony growth. However, in cultures from mice pretreated with 5-fluorouracil it promoted colony formation. It was suggested that primitive progenitors require stem cell factor in the early stages and IFN-y for subsequent growth (264).In the human system, IFN-7 has similarlybeen shown by many studies to counteract the proliferative activity of colony-stimulating factors. However, in cultures of pure CD34' progenitor cells, IFN-y has been found to synergize with IL-3 (265). IFN-y by itself did not affect proliferation, and in the presence of

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

93

IL-3, although augmenting the number of colonies, failed to affect their size, indicating that it acted only on early progenitors. In fact, in long-term cultures, once mature cells appeared, IFN-y inhibited further proliferation. 2. E ythropoiesis IFN-y has been found to suppress erythropoiesis both in vitro and in vivo, an effect that may play a role in anemia that accompanies chronic infections or autoimmune diseases. In affecting erythropoiesis in vim, IFN-y undoubtedly interacts with several other hemopoietic factors and cytokines. Studies have been aimed at revealing the most important of these interactions and at defining the most sensitive stages of erythropoiesis. In Epo-containing human CFU-E or BFU-E cultures, addition of IFN-y was found to inhibit colony formation, an effect that could be reversed by increasing Epo concentrations (266). Similarly, in murine macrophagecontaining bone marrow cultures supplemented with Epo, addition of IFN-y was found to suppress formation of both BFUE and CFUE colonies (267). However, the dose required was lower for suppression of BFUE, and the overall effect was more pronounced the earlier IFN-y was added after culture initiation. The effect of IFN-y was not prevented by addition of single or combined antibodies against TNF-a, IL-la, or GM-CSF. Accordingly, increased production of these cytokines in IFN-?-treated cultures or synergy of these cytokines with IFN-y did not seem to play a significant role in these macrophage-containing cultures. Nevertheless, in macrophage-depleted cultures, IFN-.)I was shown to synergize with the antierythropoietic effect of TNF-a. Therefore, and also because cytokines other than those that were examined may be involved, it should not be concluded that IFN-y exerts a direct inhibitory effect on BFU-E colony formation independently of other cells and cytokines.

P. EFFECTSON ENDOCRINE CELLS 1 . Thyroid In cultured human thyrocytes, IFN-y has been found to inhibit expression of thyrotropin receptors (268), production of triiodothyronin (269), and transcription of the thyroglobulin gene (270). It also reduces basal thyroid peroxidase content and inhibits thyrotropin-induced increase of the enzyme (271).Similar effects have been noted to occur in a rat thyroid cell line (FRTLd), although the effects were minor unless IFN-.)I was given in combination with TNF (272). These effects may in part explain the decrease in thyroid function in autoimmune thyroiditis. 2. Hypophysis The effects of IFN-y on the function of the anterior pituitary have been studied in an in vitro system consisting of organotypic anterior pituitary

94

ALFONS BILLIAU

cell aggregates (273, 274). In such aggregates, secretion of prolactin and ACTH in response to hypothalamic stimulatory factor was found to be inhibited by IFN-y. This inhibitory effect was, however, codetermined by the composition of the aggregates. In particular, it required the presence of the folliculostellate (FS) cell component. Secretion of anterior pituitary hormones is known to be regulated by a variety of signals received from distant sites through the blood stream (endocrine) and from the hypothalamus (neuroendocrine). In addition, these signals are integrated into a local anterior pituitary network consisting of paracrine factors and perhaps cellto-cell contacts (275). The cell populations in the anterior pituitary comprise diverse secretory elements, each producing its own set of hormones, but also comprise nonsecretory cells, in particular the FS cells. The stellate appearance of these cells results from their multiple cytoplasmic protrusions that embrace surrounding secretory elements grouping them into follicle-like aggregates. In in vitro systems, the FS cells have been shown to mitigate responses of secretory cells to stimulatory as well as inhibitory hypothalamic signals. One of their in vivo functions may therefore consist of avoiding overly brisk fluctuations in hormone levels. The FS cells also have properties and activities that resemble those of mononuclear phagocytes or glial cells-expression of typical macrophage and dendritic cell markers, phagocytosis, and secretion of IL-6 (276). The apparent kinship of FS cells with other typical IFN-y target cells and the FS cell-dependent inhibitory effect of IFN-y on the anterior pituitary secretory activity invites speculation that IFN-y, produced during inflammatory or immune responses, uses the FS cell to act on the neuroendocrine axis (277). VI. Role of IFN-.)I in Infection and Cancer

The role of IFN-y in infection and cancer has been a subject of many studies that will be reviewed separately. Here, only the most salient facts and points of current discussion are mentioned. IFN-y is generally assumed to play a primordial role in defense against intracellular bacteria and parasites. In fact, many of these pathogens have their habitat in mononuclear phagocytes. They use the intracellular environment as a shield against microbicidal antibodies and specific host defense against them is therefore mainly dependent on cellular immunity mechanisms. Early during infection, IFN-y is produced by NK-like cells; later on, by antigen-specific T cells, in particular CD4' cells that recognize microbial antigen on the class I1 antigen-positive infected phagocytes. This IFN-y then triggers several microbicidal mechanisms in infected phagocytes as well as other cells, e.g., tryptophane oxidase and reactive

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

95

oxygen and nitrogen (see relevant section). It should be noted, however, that this cytokine-dependent pathway is complemented by a cytotoxic T cell pathway, which kills phagocytes or other cells that harbor microbial pathogens (278).Another type of mechanism by which early IFN-y production is conceived to determine outcome of infections is by regulating the balance between TH1- and TH2-type responses (279, 280). Remarkably, there are also distinct exceptions to the general rule that endogenous IFN-y exerts a beneficial effect. Certain agents seem to have adapted so well to the immune system of the host that they have succeeded in perverting the cytokine network so that IFN-y acts to their advantage. Such is the case in Typanosomu infections in which endogenous IFN-y accounts for generalized immunosuppression that accompanies the disease (281). Another mechanism by which endogenous IFN-y acts to the disadvantage of the infected host is by being produced in excessive quantities. Such IFN-y may enhance the production of other cytokines, e.g., TNF-a, which may cause tissue damage and death, or may act synergistically with such cytokines. One example is shock caused in mice by endotoxin, which is aggravated by endogenous as well as exogenous IFN-y (282284). IFN-y has the potential to protect the host against virus infection by virtue of its direct antiviral effect on most types of cells and by its regulatory activity on immunocytes. Not unexpectedly, exogenously administered IFN-y has been found to act prophylacticallyagainst a variety of experimental virus infections, e.g., murine CMV infection in mice (285) or rat CMV infection in rats (286). Of more fundamental importance are the studies analyzing production of IFN-.)I and the effects of its neutralization or ablation in model virus infections in mice and rats (219,287-290). Generally speaking, these studies have indicated that endogenous IFN-y is essential for adequate host defense against virus infection, i.e., for elimination of the virus following primary infection and, in some instances, also for establishment of adequate immunity against reinfection. A striking hallmark of the importance of IFN-y as a defense mechanism against virus infections is the fact that poxviruses of diverse animal species, having coevolved with their host under pressure of endogenous IFN-y, have acquired genetic codes to instruct infected cells to produce soluble decoy receptors for IFN-y (291). A question that remains largely unsolved is whether the in viva antiviral effects of endogenous IFN-.)I are due to its direct antiviral effects on cells or to its immunomodulatory activities such as activation of NK cells, promotion of TH1- over TH2-type responses, or maturation of cytotoxic T lymphocytes. Moreover, as is the case in bacterial infections, endogenous IFN-y can act to the detriment of the host as is evident from the observation that treatment with anti-IFN-y antibody can

96

ALFONS BILLIAU

convert an aggressive into an inapparent infection with LCM virus (287). In the case of HIV, activation of monocytoid cells by IFN-y was also found to stimulate rather than to inhibit virus replication (292, 293). Rejection of some experimental tumors is associated with the presence of IFN-y in the tumor tissue. That this IFN-y contributes to this process is evident from studies with immunogenic autologous or isologous tumors, the rejection of which is abrogated by administration of neutralizing antiIFN-y antibodies (294, 295). Investigators have inserted the IFN-y gene into nonimmunogenic murine tumor cells with high metastasizing potential and found that the IFN-y secreting cells, when injected in syngeneic mice, had less ability than the parental cells to develop into tumors (296, 297); this suppression of tumorigenicity was reversed by the administration of anti-IFN-y or anti-Lyt 2.2 antibodies. Discordant with these observations are several reports describing enhancement by IFN-y of tumor growth or metastatic potential of experimental tumors. A metastasizing murine mammary carcinoma (TS/A-pc), productively transfected with the IFN-y gene, was found to metastasize more extensively than the untransfected tumor line (298). Also, treatment of mice bearing Lewis lung tumors with anti-IFN-y antibodies was found to reduce tumor outgrowth (299). In vitro treatment of carcinoma cells with IFN-y prior to their inoculation in mice has been reported to enhance metastatic potential (300); the mechanism involved in this model appeared to be augmentation by IFN-7 of the tumor cells’ resistance to the cytolytic effect of NK cells. IFN-y may be instrumental in bringing about cachexia associated with tumor development. It is generally recognized that unabated overproduction of cytokines constitutes a crucial pathogenetic element in cachexia. The cytokine that is historically most intimately associated with cachexia is TNF, also once called cachectin. IFN-y, sometimes more so than TNF, is crucially involved in the syndrome of cachexia (301). Neutralizing antibodies against IFN-y were shown to prevent cachexia in mice or rats carrying experimental tumors (299, 302). Moreover, acute cachexia was found to develop in nude mice inoculated with CHO cells productively transfected with the mouse IFN-y gene but not in mice receiving the parent tumor (303). VII. Role of IFN-y in lmmunopathology

A. IFN-y IN DELAYED-TYPE HYPERSENSITIVITY (DTH) IFN-y is among the local humoral factors that were the first to be held responsible for the DTH inflammatory response. Indeed, an early finding in studies on DTH reactions was that T cells respond to antigen by production of factors, one of which was termed MAF and was later found to

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

97

be identical with IFN-y. At present, many well-defined cytokines are recognized to be involved in the DTH inflammatory response. Of particular importance as effector cytokines, are the granulocyte- and monocytechemotactic factors, now designated as chemokines, which are instrumental in causing cellular infiltration. IFN-y is likely to play a role both in the immunization process that leads to the establishment of the DTH state (afferent arm) and in the triggering events of the DTH reaction on reexposure to the antigen (efferent arm). However, experimental systems have not always allowed to clearly make this distinction, and the role of IFN-y may be different in these different types of DTH responses. In a 4-day DTH induction model using sheep erythrocytes as an antigen in mice, exogenous IFN-y was found to reverse inhibition of the response by anti-CD4 or anti-IL-2R antibodies (304), supporting the concept that production of IFN-y by TH1 cells is essential for the reaction. Using a local adoptive transfer assay with Agspecific mouse TH1 clones, administration of anti-IFN-y antibodies was found to inhibit responses in some but not all instances (305),indicating that the positive contribution of IFN-y to the DTH effector phase is real but can vary perhaps with specificity of the clone or with mouse strain. In the rat, administration of IFN-y-neutralizing monoclonal antibodies was found to inhibit lymphocyte recruitment in DTH sites induced with KLH (306).Delayed-type hypersensitivitydoes not normally occur after immunization by injection of antigens in the anterior chamber of the eye. Transgenic mice with ectopic expression of IFN-y in the photoreceptors of the retina, on the contrary, were found not only to have more inflammation in the eye, but also to have developed a DTH state. Thus, intraocular IFN-y production can overcome the so-called anterior chamber-associated immune deviation (307). Whereas these observations support the view that endogenous IFN-y enhances DTH reactions to antigens introduced primarily into tissues, the situation may be more complex in skin contact sensitivity-type reactions. Pure IFN-y introduced into healthy skin induces a moderate perivascular lymphohistiocytic infiltrate, an intense class I1 antigen expression on keratinocytes, and an apparent migration of Langerhans cells from the epidermis to the dermis (308). Similar obserwtions were made when IFN-y was injected in the skin of patients with lepromatous leprosy (309). In a passive transfer model in mice allowing the study of a population of T cells that suppresses DNFB contact sensitivity responsiveness, exogenous IFN-y was found to antagonize suppressive cells, thereby identifying a pathway by which IFN-y favors contact sensitivity (310). However, in a DNFB contact sensitivity model in rats (311),IFN-y was found to act as a counterregulator of the inflammatory changes: systemic administration of anti-

98

ALFONS BILLIAU

IFN-y augmented ear swelling toward DNFB in sensitized animals. This effect was accompanied by reduced MHC class I1 antigen expression on keratinocytes. Administration of IFN-y in this same model inhibited ear swelling if given prior to challenge but not if given later (312). The authors considered that MHC class I1 antigen-expressing keratinocytes are instrumental in the counterregulatory effect of IFN-y. Such cells have indeed been shown to be able to inhibit T cell responses (160). Because exogenous IFN-y inhibited the skin reaction when MHC class I1 antigen expression on keratinocytes had already taken place, it seems that it does not (or not only) act by inducing class I1 expression on keratinocytes, but rather (or also) by stimulating these cells to exert their purported suppressive potential.

B. IFN-y IN ORGANSPECIFIC AUTOIMMUNE DISEASE IFN-y has many ways to to intervene in both the afferent and efferent arms of autoimmune responses. It may favor the emergence of autoreactive lymphocytesby participating in lymphocyte selection in the thymic environment: the thymus contains both cells that produce IFN-y and IFN-y target cells. IFN-y can also regulate the activity of autoreactive lymphocytes by affecting their clonal expansion, differentiation, regulation of cytokine production, cytotoxicity, and helper or suppressor activities (see relevant sections). Of particular potential importance is the apparent ability of IFN-y to critically contribute to the conversion of anergic autoreactive T cells into active effectors, as illustrated by experiments with organ-specific overexpression of IFN-y. Mice that constitutively express IFN-y in the pancreas develop autoimmune pancreatitis (313).When such mice are additionally engineered to also express a viral antigen (incasu the LCMV nucleoprotein or glycoprotein antigen), they develop autoimmune diabetes more rapidly and, significantly, possess viral antigen-specific cytotoxic T lymphocytes that are absent in the single-transgenic mice (314). In experimental animal models for autoimmune disease, administration of IFN-y is mostly found to facilitate induction of disease or to aggravate disease manifestations. In the same or similar models, blockage of endogenous IFN-.)Iby administration of neutralizing antibodies reduces disease incidence or symptoms. Examples are autoimmune thyroiditis in mice (315, 3161, autoimmune insulin-dependent diabetes in mice (317), and experimental autoimmune peripheral neuritis (318,319). An autoimmune disease-promoting role for IFN-y is by no means the general rule as will be evident from analysis of some selected model systems. 1 . Autoimmunity in the CNS Experimental autoimmune encephalomyelitis (EAE) and experimental autoimmune uveitis in mice are reduced in incidence and severity by the

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

99

administration of IFN-y and augmented by treatment with antibodies against IFN-y (320-324). Not only actively induced, but also passively transferred EAE is inhibited by anti-IFN-y antibody (321),suggesting that IFN-y exerts its disease-limiting effect by acting on the efferent rather than the afferent arm of the autoimmune process, i.e., by inhibiting the action rather than the generation of CNS antigen-reactive T cells. During EAE in Lewis rats, IFN-y is produced both in the CNS and peripherally (325).The effector T cells by which the disease can be transferred have been found to produce IFN-y when confronted with MBP. In contrast, the suppressor splenic T cells found during reconvalescence of the disease fail to produce IFN-y and inhibit IFN-y production when added to challenged effector T cells (326). Although these data suggest that, in the early induction phase, IFN-.)I favors initiation of the disease, this interpretation is difficult to reconcile with the aggravating effect of anti-IFN-y antibodies in the mouse model. The disease-preventing effect of IFN-y in murine EAE invokes an immunosuppressive or anti-inflammatory pathway. One such pathway is the induction of NO (see relevant sections). However, evidence as to the effect of NO in EAE is contradictory. In adoptively transferred EAE in SJL mice the NO synthase inhibitor aminoguanidine was found to ameliorate disease parameters (327).In actively induced EAE rats, by contrast, treatment with the NO synthase inhibitor L-NMMAwas found to cause aggravation of disease (S. R. Ruuls, personal communication). 2. Autoimmune Arthritis Locally administered IFN-y was found to promote development of collagen-induced autoimmune arthritis in mice (328),whereas systemically administered IFN-y exerted a protective effect (329).Systemicallyadministered IFN-y was also found to inhibit inflammatoy cell recruitment and disease signs in bacterial cell wall-induced arthritis (330). Blockage of endogenous IFN-y by administration of anti-IFN-y antibodies in experimental arthritis has opposing effects depending on time: early blockage tends to favor disease development, whereas late blockage exerts a protective effect (331). Clearly, IFN-y affects the disease process in opposite directions when acting via different pathways. IFN-y stimulates synovial cell proliferation, an effect by which local IFN-y in joints may aggravate disease. It should be noted, however, that in synoviocytescultured from arthritic joints, IFN-y reportedly antagonizes cell growth-promoting effects of TNF (332) and IL-1 (333).Local production of IL-1 and TNF in the joints is considered to be a pathogenetically important process, as these cytokines induce production by chondrocytes and synovial cells of tissue-destructive enzymes. The production of IL-lP

100

ALFONS BILLIAU

by synovial macrophages in rheumatoid arthritis was found to be inhibited by IFN-y (334),suggesting that local production of IFN-y acts as a diseaselimiting factor. Furthermore, evidence indicates that IFN-.), also opposes the action of locally produced IL-1 and TNF. Thus, IFN-y was found to inhibit metalloproteinase production and glycosaminoglycan release by cultured cartilage fragments (335). In cultured human articular chondrocytes, IFN-y at concentrations as low as 1 unit/ml synergizes with TNF in inducing PGE production, but acts antagonistically to TNF in decreasing TNF-induced caseinase production (336):caseinase levels in culture fluid reflect those of the proteoglycanase stromelysin. In synoviocytes cultured from arthritic joints, IFN-y was found to act antagonistically to TNF (332) and to IL-1 (333), as it inhibited cell growth, PGE2 release, and collagenase production. Another property of IFN-y that may be relevant to its action in autoimmune joint disease is its inhibitory effect on bone resorption in an in vitro system (337) and on the formation of osteoclast-like multinucleated cells in long-term human bone marrow cultures stimulated with vitamin D3 (338). 3. Autoimmune lnsulitis Systemic injection of streptozotocin into mice of certain strains induces insulin-dependent diabetes resembling type I diabetes occurring in man. Mononuclear cell infiltrates occur in the pancreatic islets and treatment with immunosuppressive drugs can retard disease development. Streptozotocin-induced diabetes, as assessed by hyperglycemia and body weight loss, was found to be more severe in mice that also received IFN-y injections (339). IFN-y treatment was also found to augment expression of MHC class I and class I1 antigens in the streptozotocin-challengedmice. Remarkably, IFN-y by itself fails to induce MHC class I1 antigen expression in p cells, whereas it does so in most other cell types and tissues-in uiuo. The cofactor required for class I1 antigen expression on cells of insulitisaffected pancreas may be TNF (116). Another experimental model is insulin-dependent diabetes that occurs spontaneously, but with low incidence, in NODNehi mice. By giving single injections of cyclophosphamide, the incidence can be increased so that in a high proportion of mice p cell destruction and hyperglycemia occur in a matter of weeks. In such mice, T cells disappear from the pancreas within 2 or 3 days after cyclophosphamide injection and then reappear in much greater numbers 1 week later. At that time, a dramatic increase occurs in expression of MHC class I protein on islet cells and on infiltrating inflammatory cells. Pretreatment of such mice with anti-IFNy antibodies was found to reduce the incidence and severity of the syndrome (317,340)and also toprevent overexpressionof MHC class I antigen

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

101

(341).It is considered that class I antigen overexpression is instrumental in targeting cytotoxic T cells to p cells. Counter to expectation, administration of IFN-y in this mouse model of diabetes did not affect blood glucose profiles. In fact, in combination with TNF-a, IFN-y treatment was associated with a reduction in severity of islet inflammation, although this treatment caused moderate to severe pancreatitis and several other pathologic changes (342). The apparent contradiction between results obtained with IFN-y and those with antiIFN-y treatment is reminiscent of a similar paradoxical observation in GVH disease models (see relevant section). Further evidence for IFN-y to serve a role in insulitis has come from studies with IFN-y-transgenic mice. Expression of IFN-y in islets of Langerhans was found to result in their inflammatory destruction (313).Another feature of these IFN-y-transgenic mice is the presence of distended ducts indicating duct cell proliferation and islet neogenesis (343).Treatment of the mice with neutralizing anti-IFN-y antibody was found to halt progression of the disease (344). Because no evidence of circulating IFN-y could be found, the authors considered that the pancreatic lesions result from a local effect of IFN-y produced in the pancreas. Lymphocytes from the transgenic mice were found to be cytotoxic for islets in vitro (313).Therefore, the authors proposed that islet destruction in this model is due to cytocidal effects of infiltrating lymphocytes rather than to a direct cytotoxic effect of locally produced IFN-y. Overexpression of MHC class I1 antigens was noted to occur on exocrine cells. Furthermore, pancreatic endothelia of transgenic mice displayed spontaneous expression of the vascular addressins MadCam-1 (known to be required for lymphocyte homing to Peyer’s patches) and MECA-79 ligand (normally located in peripheral lymph nodes); the ICAM-1/LFA-1 pair was found to be expressed on endothelial cells, pancreatic duct epithelial cells, and lymphocytes, but not on islet cells (345).These changes also occurred in IFN-y-transgenic SCID mice indicating that they resulted directly from local IFN-.)I production and not as an effect secondary to cellular infiltration. Whereas induction of vascular addressins and adhesion molecules can be considered to account for the earliest inflammatory changes in insulitis, loss of tolerance remains unexplained. By its ability to induce nitrogen oxide production, IFN-7, in concert with TNF, may directly contribute to islet cell death. Thus, it was shown that IFN-y and TNF synergistically induce NO in mouse islet cells, and that accompanying cytotoxicity can be prevented by L-NMMA (346).

c. IFN-7 IN ALLOGRAFTREJECTIONA N D GRAFTVERSUSHOSTREACTIONS The rejection of an allogeneic tumor by mice was found to be delayed or abrogated by pretreatment of the recipient hosts with anti-IFN-y antibody,

102

ALFONS BILLIAU

suggesting that endogenous IFN-.)I is crucial to allograft rejection (199). Along this line, treatment of skin allograft recipient mice with anti-IFNy antibody has been found to prolong rejection if the graft is MHC class I1 antigen-incompatible, but not if it is only MHC class I-incompatible (347). Only a small percentage of resting skin epidermal cells, i.e., Langerhans cells, express MHC class I1 gene products. However, after exposure to IFN-y, all keratinocytes can become class I1 antigen-positive. The result was therefore interpreted to mean that, if endogenous IFN-y contributes to the rejection of a skin allograft, it does so because it induces class I1 expression on keratinocytes. The role of IFN-y in skin allograft rejection was also studied in C3H/ HEJ mice rendered hyporesponsive to B10.BR skin allografts by pretreatment with irradiated donor lymphoid cells. In this model, allograft enhancement can be achieved if the irradiated cells are injected via the portal but not via the tail vein. This state of relative anergy was found to be associated with reduced mRNA levels in lymphocytes for IFN-y and IL-2 and increased levels of mRNA for IL-4 and IL-I0. significantly, repeated treatment with anti-IFN-y antibody allowed graft enhancement and altered T H l R H 2 cytokine profile to be achieved with injection of irradiated lymphocytes in the tail vein (348). In spleens of mice with experimentally induced acute GVH disease, production of IFN-y, together with that of other cytokines, has been demonstrated (349, 350). Spleen cells of mice with GVH disease respond to mitogen stimulation with IFN-y production when taken during the early proliferative phase of the disease and with TNF production when taken during the later destructive phase (351). In several independent studies (349, 352), the contribution of this IFN-y to the disease manifestations has been assessed by the use of neutralizing anti-IFN-y antibodies. These studies are unanimous in observing that blockage of IFN-y inhibits disease development, in particular the lesions in the gut mucosa. On the other hand, as an apparent paradox, it has been reported that systemic administration of IFN-y inhibits disease development in much the same way as anti-IFNy antibody does. This inhibition was associated with reduced numbers of IFN-y-producing cells (349). Suppression of endogenous cytokine production may reflect immunosuppressive circuitry triggered by systemic and perhaps less so by locally produced IFN-y. Another model of GVH disease consists of inoculating neonatal mice with semiallogeneic lymphocytes, which allows for the persistence of the donor cells in the host. These cells differentiate into TH2-like cells as evident from predominance of IL-4 production over that of I L 2 and IFN-y. As an apparent consequence, donor B cells differentiate to produce large quantities of IgE and IgGl autoantibodies, resulting in immune deposits

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

103

and SLE-like pathology. In this model, exogenous IFN-y was found to prevent the disease apparently by restoring the ability of the lymphocytes to produce IL-2 and IFN-y (353). A N D GENERALIZED D. IFN-y IN NONSPECIFICINFLAMMATION CYTOKINE RELEASE SYNDROMES Systemic administration of IFN-y in mice can provoke a state of reduced responsiveness of tissue to aspecificinflammatory stimuli. Thus, the footpad swelling reaction to local endotoxin injection is diminished (354) and accumulation of polymorphonuclear cells in response to local injection of IL-1 into the footpad is reduced (244). These apparently anti-inflammatory effects of IFN-y are so far unexplained and, in fact, contrast with established activities of IFN-y in in vitro systems, e.g., induction of increased adhesion between endothelia and leukocytes. In clinical immunology, many situations have become known in which exposure to infection, allergens, or other immune stimulants elicit overreactions that are often so acute and severe that they threaten life. Some of the best known examples are anaphylactic shock due to massive exposure to IgE-reactive allergens, serum sickness due to exposure to xenogeneic serum proteins, circulatory and organ failure due to gram-negative bacterial sepsis (septic shock), and toxic shock syndrome due to a staphylococcal exotoxin entering the general circulation. A recent example is the antiCD3 antibody first-injection syndrome, an anaphylactoid state occurring in organ transplant patients who receive a large dose of anti-CD3 antibody as a means to eliminate or inactivate T cells and thus to prevent transplant rejection. A common central element in the pathogenesis of these syndromes is the sudden and excessive release of immune mediators, such as histamine and histamine-like substances in anaphylaxis, or activated complement components in serum sickness. Normally, these mediators support the immune system in dealing with microbial or other aggressors. However, when overproduced, they cause more harm than good. The same holds true for cytokines:when inappropriately produced, they can trigger complications that have become known as “cytokine release syndromes.” In septic shock, the eliciting factor is endotoxin, which has long been recognized to trigger release of several mediators from various sources and by different pathways, including direct activation of complement and induction of degranulation of platelets. A recently recognized pathway leading to endotoxin-associated generalized reactions is the induction of cytokines. Particularly well studied is the induction by endotoxin of IL-1, TNF, IL-6, IL-12, and IFN-y. There is experimental evidence for a role of each of these cytokines in bringing about or in counteracting the in

104

ALFONS BILLIAU

vivo changes associated with endotoxin-induced shock reactions (283). Experimental animals exposed to endotoxin produce only small quantities of IFN-y in the circulation. Nevertheless, the case for this endogenous IFN-y to play a critical role as a promotor of the pathological signs is very strong, as is discussed in another section. Toxic shock syndrome toxin of Staphylococcus is a superantigen that associates with antigen-presenting cells and then activates a large number of T cell subsets. Such activated T cells release cytokines, including IFN-y, that flood the circulation. A murine model system for the study of the superantigen-induced pathology consists of injecting BALBlc mice with staphylococcal enterotoxin B (SEB). These mice develop acute but transient hypoglycemia and rapid weight loss. The occurrence of these manifestations can be prevented by pretreatment with antibodies against IFN-y (355), but equally so by antibodies against TNF-a (356), indicating that both cytokines have an important role to play. A similar situation occurs in mice injected with anti-CD3 antibody. In this model, the cytokine release syndrome manifests itself mainly by hypothermia. Both anti-TNF (357) and anti-IFN-y antibodies (358)can prevent the symptoms. It has been considered that IFN-y and TNF, and perhaps additional cytokines, synergize in cytokine syndromes by forming a cascade. Support for this concept was obtained in endotoxin-injected mice: blockage of IFN-y not only prevented disease and death but also reduced the production of TNF, suggesting that in this system endogenous IFN-y acts as an enhancer of TNF production (284). In the SEB- or anti-CD3-injected mice, however, blockage of IFN-y did not affect TNF release (355, 358). Although these observations indicate that IFN-y promotes development of cytokine release syndromes, other evidence points at its ability to also trigger a counterregulatory pathway. Thus, IFN-y receptor-deficient strain 129 mice were found to be more sensitive to the anti-CD3-induced syndrome than the wild-type strain (197), a finding that is in apparent conflict with the results obtained with anti-IFN-y antibodies in BALB/c mice. However, from experiments with the NO synthetase blocker, L-NMMA, it was evident that NO induction by IFN-y provides protection against the manifestations of the syndrome. Failure of this pathway to be triggered in the IFN-.)Ireceptor knockout mice probably accounts for their higher sensitivity. It thus appears that in cytokine release syndromes, as in so many other situations, IFN-.)Iproduction plays a Janus role, promoting manifestations via one set of pathways and counteracting them via others. Depending on the test system, or clinical situation, IFN-y may display one or the other of its Janus faces.

INF-7: BIOLOGY AND ROLE IN PATHOGENESIS

105

VIII. Final Thoughts

What are the most salient issues of recent and current IFN-y research? Especially obvious is the progress in our understanding of the cellular and molecular biological aspects of IFN-y action. The once-elusive IFN-y receptor has yielded most of its secrets. Investigation of IFN-y signal transduction has brought to light a new pathway, i.e., the STAT molecules, which is used by many different cytokines. New insight has been generated in the biology of IFN-y production. The NK cell population has been identified as a source of IFN-y that appears to be as important as the T cells. One of the latter-day cytokines, IL-12 or NK cell-stimulating factor, owes much of its importance to its IFN-y-inducing capacity. Integration of IFN-y in the cytokine network has gained more substance by clarification of interactions with other cytokines, especially those that antagonize IFN-y production and/or action: IL-10, IL-4, and TGF-b. Generation of nitrogen oxide has been pinpointed as a crucial mediator of cellular and organ responses to IFN-y, e.g., cytocidal effects, enhanced intracellular microbial killing and inflammation, but also immunosuppressive counterregulation. The fact that nitrogen oxide is involved in a wide spectrum of physiological and pathological processes and that most cells of the body have the potential to respond to IFN-y makes it seem likely that IFN-y will be found to be involved in an even wider range of tissue reactions than is presently recognized. Nevertheless, evidence that IFN-y plays a role in normal physiology is still lacking. Especially promising from this point of view is the availability of laboratory mouse strains that are deficient in either the production of IFN-y or in the expression of IFN-y receptor molecules. To date, however, experiments employing these animals have little more than confirmed what was already known from other approaches, i.e., the important role of IFN-y as a irreplaceable factor in host defense against infection and as a wrongdoer in inflammation and autoimmune disease. Identification of IFN-y as a mediator of tissue reactions in inflammatory and autoimmune disease has generated a cogent quest for usable antagonists, e.g., humanized monoclonal antibodies, soluble receptors, or antagonistic cytokines. It seems likely that such IFN-.)I antagonists, if made available, will find their way into the clinic for diseases with an autoimmune component such as lupus erythematosus, multiple sclerosis, Crohn’s disease, and others. Intriguing, however, is the unexpected observation that IFN-y can, in some animal models, inhibit rather than promote autoimmune-type diseases, e.g., experimental autoimmune encephalomyelitis and graft versus host disease.

106

ALFONS BILLIAU

IX. Notes Added in Proof

Over the few months since submission of this chapter, numerous publications in relation to IFN-y have appeared. Here, a selection representative of the current trend in IFN-y research is given and commented upon. Our conception of the early events following interactions of IFN-y with its receptor have been further refined. With many cytokines, e.g., IL-2, the number of receptors present on cells is downregulated by exposure to the cytokine. In the case of IFN-y an aberrant type of receptor downregulation occurs in some cells (359), in particular murine CD4' T cells, but not in others, e.g., mouse L929 fibroblast-like cells. Downregulation affects only the @-chainwhich plays no part in ligand binding but does transmit the signal to the cytoplasm. This mechanism is held responsible for the virtual absence of @-chainsin cultured murine TH1 cells which produce IFN-y themselves (194,359). However, @-chaindownregulation is not inherent to TH1 cells since both TH1 and TH2 cells do retain the @-chain in IFN-yR knock-out mice. Research on membrane signaling has refined the model for interaction of the receptor with intracellular proteins: JAKl and JAKZ are associated, respectively, with the a- and @-chainsof the receptor. Clustering of the receptor chains results in phosphorylation of both JAKs, in activation of JAKZ activity and in association with STATla (360). Furthermore, a transcription activation factor, yRF-1, which associates with the yRE-1 sequence in the mig gene, has been isolated and characterized (361). The molecular mechanism of induction of class I1 MIIC gene products by IFN-y MHC has been clarified to some extent. The promoter regions of class I1 MHC and invariant chain genes contain several regulatory elements. However, these elements are only indirectly controlled by IFNy , through an IFN-y-induced transactivator protein called CIITA (class I1 transactivator). The CIITA gene has been isolated in the course of studies on the defects in patients suffering from bare lymphocyte syndromes (362). CIITA is required for constitutive expression of class I1 MHC genes in professional antigen-presenting cells, but is also involved in IFN-y-induced expression in cells which are constitutively class II-negative. IFN-y induces expression of CIITA mRNA before induction of class I1 MHC mRNA molecules. Moreover, CIITA mRNA induction is not blocked by cycloheximide, whereas induction of MHC Class I1 mRNA itself does require protein synthesis. Induction of CIITA mRNA does require JAKl activity but is not dependent on expression of IRF-1. There is no evidence that CIITA is a DNA-binding factor (363,364). The important role of NK cells as producers of IFN-y has received further support. IL-15 is a recently characterized monocyte-derived cpo-

INF-y: BIOLOGY AND ROLE I N PATHOGENESIS

107

kine which, like IL-12, activates NK cells and stimulates them to produce IFN-y. In cocultures of monocytes and NK cells, the induction of IFN-y by LPS was found to be abrogated by inclusion of neutralizing anti-IL-15 antibody, indicating that endogenous production of IL-15 is essential for IFN-y production in this system (365). In a study on the mechanism of natural resistance to Typanosoma cruzi infection, it was found that depletion of NK cells from mouse splenocyte cultures drastically reduces production of IFN-.)Isubsequent to parasite infection. In addition, in vivo depletion of N K cells by the use of anti-NK cell monoclonal antibody reduced natural resistance against the parasite (366).The NK cell-stimulating cytokine IL-12 was found to protect normal but not IFN-y-gene knock-out mice against infection with Mycobacterium tuberculosis (367). Similarly, the protective effect of IL- 12 against Histoplasma capsulutum infection in mice was found to be abrogated by treatment of the animals with antiIFN-y antibody (368). Treatment with anti-IFN-y antibody was shown to protect normal mice against a lethal dose of live staphylococci; this effect was associated with increased bacterial counts in organs early after infection, but reduced counts at later times (369). Inoculation of IFN-yR knock-out mice with a TSST- l-producing staphylococcus strain produced more frequent and more severe arthritis than was found in the wild-type counterparts. In addition, severe sepsis with high mortality occurred early after infection; at later time points the wild-type counterparts had higher mortality than did the knock-outs (370). The contribution of IFN-.)Ito overall immunosuppression accompanying certain infections has been further documented. Inhibition of splenocyteproliferative responses in mice acutely infected with Toxoplasma was found to be relieved by addition of anti-IFN-y antibody to the culture medium (371). Concomitantly, anti-IFN-y antibody-treated cultures produced less nitrite, implicating IFN-y-induced NO as a factor in the immunosuppression. A similar situation was found to occur in mice infected with Mycobactem'um avium (372). The detrimental role of IFN-y in EAE has been further documented by the obsewation that the actively induced disease is more severe in IFN-y knock-out mice (373). In collagen-induced arthritis in mice, early or midterm treatment with anti-IFN-y was found to inhibit disease development, while late treatment was either without effect or aggravated the symptoms (374). ACKNOWLEDGMENT Basic studies on IFN-7 in the laboratory of the author are supported by funds from the Regional Government of Flanders (GOA initiative), the Federal Government of Belgium

108

ALFONS BILLIAU

(IUAP initiative), the Belgian National Fund for Scientific Research, and the VTM TV station (Lifeline Initiative). The author thanks Ren6 Conings, Willy Put, and Dominique Brabants for editorial help.

REFERENCES 1. Wheelock, E. F. (1965). Interferon-like virus inhibitor induced in human leukocytes by phytohemagglutinin. Science 141, 30-311. 2. Youngner, J., and Salvin, S. B. (1973).Production and properties of migration inhibitory factor and interferon in the circulation of mice with delayed type hypersensitivity. J. Immunol. 111,1914-1922. 3. Falcoff, R. (1972).Some properties of virus- and immune-induced human lymphocyte interferons. J. Gen. Virol. 16, 251-253. 4 . Stewart, W. E. I., et al. (1980). Interferon nomenclature. Nature 286, 110. 5. Nathan, C., Murray, H. W., Wiebe, M. E., and Rubin, B. Y. (1983). Identification of interferon-? as the lymphokine that activates human macrophage oxidative metabolism and antimicrobial activity. J. Exp. Med. 158, 670-681. 5a. Bloom, B. R., and Bennett, B. (1966). Mechanism of a reaction in uitro associated with delayed-type hypersensitivity. Science 153, 80-82. 6. Steeg, P., Moore, R. N., Johnson, H. M., and Oppenheim, J. J. (1982). Regulation of murine macrophage Ia antigen expression by a lymphokine with immune interferon activity. J Exp. Med. 156, 1780-1793. 7. Weinberg, J. B., Chapman, H. A., and Hibbs, J. B. (1978). Characterization of the effects of endotoxin on macrophage tumor cell killing. J. Immunol. 121, 72-80. 8. Farrar, M. A., and Schreiber, R. D. (1993). The molecular cell biology of interferongamma and its receptor. Annu. Reu. Zmmunol. 11, 571-611. 9. Ijzermans, J., and Marquet, R. L. (1989).Interferon-gamma: A review. Irnmunobiobgy 179, 456-473. 10. Ealick, S. E., Cook, W. J., Vijay Kumar, S., Carson, M., Nagabhushan, T. L., Troth, P. P., and Bugg, C. E. (1991). Three-dimensional structure of recombinant human interferon-y. Science 252, 698-702. 11. Walter, M., Windsor, W. T., Nagabushan, T. L., Lundell, D. J., Lunn, C. A., Zauodny, P. J., and Narula, S. K. (1995). Crystal structure of a complex between interferon-y and its soluble high-affinity receptor. Nature 376, 230-235. 12. Griggs, N. D., Jarpe, M. A., Pace, J. L., Russell, S. W., and Johnson, H. M. (1992). The N-terminus and C-terminus of IFN-gamma are binding domains for cloned soluble IFN-gamma receptor. J Immunol. 149, 517-520. 13. Van Volkenburg, M. A,, Griggs, N. D., Jarpe, M. A,, Pace, J. L., Russell, S. W., and Johnson, H. M. (1993). Binding site on the murine IFN-gamma receptor for IFNgamma has been identified using the synthetic peptide approach. J. Immunol. 151, 6206-6213. 14. Sareneva, T., Pirhonen, J., Cantell, K., Kalkkinen, N., and Julkunen, I. (1994). Role of N-glycosylation in the synthesis, dimerization and secretion of human interferongamma. Biochem. J. 303, 831-840. 15. Schreiber, R. D., Hicks, L. J., Celada, A., Buchmeier, N. A., and Gray, P. W. (1985). Monoclonal antibodies to murine y-interferon which differentially modulate macrophage activation and antiviral activity. J. Immunol. 134, 1609-1618. 16. Kwok, A. Y. C., Zu, X., Yang, C., Alfa, M. J., and Jay, F. T. (1993). Human interferony has three domains associated with its antiviral function: A neutralizing epitope typing scheme for human interferon-y. Immunology 78, 131-137.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

109

17. Fong, T. A., and Mosmann, T. R. (1990). Alloreactive murine CD8 + T cell clones secrete the Thl pattern of cytokines.]. Immunol. 144, 1744-1752. 18. Alzona, M., Jack, H.-M., Fisher, R. I., and Ellis, T. M. (1994). CD30 defines a subset of activated human T cells that produce IFN--y and IL-5 and exhibit enhanced B cell helper activity. J. ZmmunoZ. 153,2861-2867. 19. Fultz, M. J., Barber, S. A., Dieffenbach, C. W., and Vogel, S. N. (1995). Induction of IFN-gamma in macrophages by lipopolysaccharide. Int. Immunol. 5, 1383-1392. 20. Yoshihara, R., Shiozawa, S., Fujita, T., and Chihara, K. (1993). Gamma interferon is produced by human natural killer cells but not T cells during StaphyZococcu.~aurms stimulation. Infect. Immun. 61, 3117-3122. 21. Dunn, P., and North, R. J. (1991). Early gamma interferon production by natural killer cells is important in defense against murine listeriosis. Infect. Immun. 59,2892-2900. 22. Scharton, T. M., and Scott, P. (1993). Natural killer cells are a source of interferon gamma that drives differentiation of CD4' T cell subsets and induces early resistance to Leishmania major in mice. J. Exp. Med. 178, 567-577. 23. Wherry, J. C., Schreiber, R. D., and Unanue, E. R. (1991). Regulation of gamma interferon production by natural killer cells in scid mice: Roles of tumor necrosis factor and bacterial stimuli. Infect. Immun. 59, 1709-1715. 24. Bancroft, G. J., Sheehan, K. C. F., Schreiber, R. D., and Unanue, E. R. (1989). Tumor necrosis factor is involved in the T cell independent pathway of macrophage activation in scid mice. J. ImmunoZ. 143, 127-130. 25. Rothe, J., Lesslauer, W., Liitscher, H., Lang, Y., Koebel, P., Kontgen, F., Althage, A., Zinkernagel, R., Steinmetz, M., and Bluethmann, H. (1993). Mice lacking the tumour necrosis factor receptor 1are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature 364, 798-802. 26. Chan, S. H., Kobayashi, M., Santoli, D., Perussia, B., and Trinchieri, G. (1992). Mechanisms of IFN--y induction by natural killer cell stimulatory factor (NKSFAL12).J. Immunol. 148, 92-98. 27. Yaegashi, Y., Nielsen, P., Sing, A,, Galanos, C., and Freudenberg, M. A. (1995). Interferon /3, a cofactor in the interferon gamma production induced by gram-negative bacteria in mice. J. Exp. Med. 181, 953-960. 28. Heinzel, F. P., Rerko, R. M., Ling, P., Hakimi, J., and Schoenhaut, D. S. (1994). Interleukin 12 is produced in vivo during endotoxemia and stimulates synthesis of gamma interferon. Infect. Immun. 62, 4244-4249. 29. Heremans, H., Dillen, C., Van Damme, J., and Billiau, A. (1994). Essential role for natural killer cells in the lethal lipopolysaccharide-inducedShwartzman-like reaction in mice. Eur. J. lmmunol. 24, 1155-1160. 30. D'Orazio, J. A,, Burke, G. W., and Stein-Streilein, J. (1995). Staphylococcal enterotoxin B activates purified NK cells to secrete IFN-gamma but requires T lymphocytes to augment NK cytotodcity. J. Immunol. 154, 1014-1023. 31. Flesch, I. E. A., Hess, J. H., Huang, S., Aguet, M., Rothe, J., Bluethmann, H., and Kaufmann, S. H. E. (1995). Early interleukin-12 production by macrophages in response to mycobacterial infections depends on interferon y and tumor necrosis factor. J. Exp. Med. 181, 1615-1621. 32. Gazzinelli, R, T., Wysocka, M., Hayashi, S., Denkers, E. Y., Hieny, S., Caspar, P., Trinchieri, G., and Sher, A. (1994). Parasite-induced IL-12 stimulates early IFNgamma synthesis and resistance during acute infection with ToxupZasma gondii. J. Immunol. 153,2533-2543. 33. Moms, A. G., Lin, Y., and Askonas, B. A. (1982). Immune interferon release when a cloned cytotoxic T-cell line meets its correct influenza-infected target cell. Nature 295, 150-152.

110

ALFONS BILLIAU

34. Wingren, A. G., Dahlenborg, K., Bjorkhmd, M., Hedlund, G., Kalland, T., Sjogren, H. O., Ljungdahl, A., Olsson, T., Ekre, H. P., and Sanson, D. (1993). Monocyteregulated IFN-gamma production in human T cells involves CD2 signaling.J. Immunol. 151, 1328-1336. 35. Yang, X., and HayGlass, K. T. (1993).IFN-gamma, but not IL-4, synthesis by antigenprimed murine T cells is IL-2 dependent: Differential requirements in de novo and established responses. J. Immunol. 150,4354-4363. 36. Seder, R. A,, Gazzinelli, R., Sher, A,, and Paul, W. E. (1993). Interleukin 12 acts directly on CD4+ T cells to enhance priming for interferon gamma production and diminishes interleukin 4 inhibition of such priming. Proc. NatL Acud. Sci. USA 90, 10188-10192. 37. Alzona, M., Jack, H.-M., Fisher, R. I., and Ellis, T. M. (1995). IL-12 activates IFNgamma production through the preferential activation of CD30+T cells. J. Immunol. 154,9-16. 38. Snijdewint, F. G. M., Kalinski, P., Wierenga, E. A., Bos, J. D., and Kapsenberg, M. L. (1993). Prostaglandin EBdifferentially modulates cytokine secretion profiles of human T helper lymphocytes.J. Immunol. 150,5321-5329. 39. Ye, J., Ghosh, P., Cippitelli, M., Subleski, J., Hardy, K. J., Ortaldo, J. R., and Young, H. A. (1994).Characterization of a silencer regulatory element in the human interferongamma promoter. J. B i d . Chem. 269, 25728-25734. 40. Finbloom, D. S., Wahl, L. M., andwinestock, K. D. (1991).The receptor for interferony on human peripheral blood monocytes consists of multiple distinct units. J . Biol. Chem. 266, 22545-22548. 41. Rashidbaigi, A., Langer, J. A,, Jung, V., Jones, C., Morse, H. G., Tischfield, J. A., Trill, J. J., Kung, H., and Pestka, S. (1986). The gene for the human immune interferon receptor is located on chromosome 6. Proc. Natl. Acad. Sci. USA 83, 384-388. 42. Hemmi, S., Peghini, P., Metzler, M., Merlin, G., Dembic, Z., and Aguet, M. (1989). Cloning of murine interferon y receptor cDNA: Expression in human cells mediates high-affinity binding but is not sufficient to confer sensitivity to murine interferon y. Proc. Natl. Acad. Sci. USA 86, 9901-9905. 43. Aguet, M., Dembic, Z., and Merlin, G. (1988). Molecular cloning expression of the human interferon-gamma receptor. Cell 55, 273-280. 44. Hibino, Y., Kumar, C. S., Mariano, T. M., and Pestka, S. (1992).Chimeric interferony receptors demonstrate that an accessory factor required for activity interacts with the extracellular domain. J . B i d . Chem. 267, 3741-3749. 45. Jung, V., Jones, C., Kumar, C., Stefanos, S., O’Connell, S., and Pestka, S. (1990). Expression and reconstitution of a biologically active human interferon-? receptor in hamster cells. J. Biol. Chem. 265, 1827-1830. 46. Hibino, Y., Mariano,T. M., Kumar, C. S., Kozak, C. A., and Pestka, S. (1991).Expression and reconstitution of a biologically active mouse interferon-y receptor in hamster cells. J. Biol. Chem. 266, 6948-6951. 47. Soh, J., Donnelly, R. J., Kotenko, S., Mariano, T. M., Cook, J. R., Wang, N., Emanuel, S., Schwartz, B., Miki, T., and Pestka, S. (1994). Identification and sequence of an accessory factor required for activation of the human interferon gamma receptor. Cell 76, 793-802. 48. Hemmi, S., Bohni, R., Stark, G., Di Marco, F., and Aguet, M. (1994).A novel member of the interferon receptor family complements functionality of the murine interferon gamma receptor in human cells. Cell 76, 803-810. 49. Soh, J., Donnelly, R., Mariano, T., Cook, J., Schwartz, B., and Pestka, S. (1993). Identification of a yeast artificial chromosome clone encoding an accessor factor for

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

111

the human interferon gamma receptor: Evidence for multiple accessor factors. Proc. Natl. Acad. Sci. USA 90, 87%37-8741. 50. Greenlund, A. C., Schreiber, R. D., Goeddel, D. V., and Pennica, D. (1993). Interferongamma induces receptor dimerization in solution and on cells. /. Biol. Chem. 268, 18103-18110. 51. Axelrod, A., Gibbs, V. C., and Goeddel, D. V. (1994). The interferon-gamma receptor extracellular domain. Non-identical requirements for ligand binding and signaling. 1.Biol. Chem. 269, 15533-15539. 52. Szente, B. E.. and Johnson, H. M. (1994). Binding of IFNgamma and its C-terminal peptide to a cytoplasmic domain of its receptor that is essential for function. Biochem. Biophys. Res. Commun. 201, 215-221. 53. Finbloom, D. S., and Wahl, L. M. (1989). Characterization of a novel low affinity receptor for IFN-y on adherent human monocytes by radioligand binding studies and chemical cross-linking. /. Immunol. 142, 2314-2320. 54. Krakauer, T., and Oppenheim, J. J. (1993). IL-1 and tumor necrosis factor-a each upregulate both the expression of IFN--y receptors and enhance IFN-y-induced H U DR expression on human monocytes and a human monocytic cell line (THP-1). 1.Inzmunol. 150, 1205-1211. 55. Raitano, A. B., and Korc, M. (1990). Tumor necrosis factor up-regulates y-interferon binding in a human carcinoma cell. 1. B i d . Chem. 265, 10466-10472. 56. Sanceau, J., Merlin, G., and Wietzerbin, J. (1992). Tumor necrosis factor and IL-6 upregulate IFN-y receptor gene expression in human monocytic THP-1 cells by transcriptional and post-transcriptional mechanisms. J. Immunol. 149, 1671-1675. 57. Chen, H.-L., Kamath, R., Pace, J. L., Russell, S. W., and Hunt, J. S. (1994). Gestationrelated expression of the interferon-gamma receptor gene in mouse uterine and embryonic hematopoietic cells. I. Leukocyte Biol. 55, 617-625. 58. Taniguchi, T. (1995). Cytokine signalling through nonreceptor protein tyrosine kinases. Science 268, 255. 59. Lamer, A. C., David, M., Feldman, G. M., Igarashi, K., Hackett, R. H., Webb, D. S. A,, Sweitzer, S. M., Petricoin, E. F., 111, and Finbloom, D. S. (1993). Tyrosine phosphorylation of DNA binding proteins by multiple cytokines. Science 261, 17301733. 60. Ruff-Jamison, S., Chen, K., and Cohen, S. (1993).Induction by EGF and interferongamma of tyrosine phosphorylated DNA binding proteins in mouse liver nuclei. Science 261, 1733-1736. 61. Sadowski, H. B., Shuai, K., Darnel], J. E., Jr., and Gilman, M. Z. (1993). A common nuclear signal transduction pathway activated by growth factor and cytokine receptors. Science 261, 1739-1744. 62. Silvennoinen, O., Schindler, C., Schlessinger,J., and Levy, E. (1993). Ras-independent growth factor signaling by transcription factor tyrosine phosphorylation. Science 261, 1736-1739. 63. Watling, D., Guschin, D., Muller, M., Silvennoinen, O., Witthuhn, B. A,, Queue, F. W., Rogers, N . C., Schindler, C . , Stark, G. R., Ihle, J. N., and Kerr, I. M. (1993). Complementation by the protein tyrosine kinase JAK2 of a mutant cell line defective in the interferon-gamma signal transduction pathway. Nature 366, 166-170. 64. Igarashi, K., Garotta, G., Ozmen, L., Ziemiecki, A,, Wilks, A. F., Harpur, A. G., Lamer, A. C., and Finbloom, D. S. (1994). Interferon-gamma induces tyrosine pbosphorylation of interferon-gamma receptor and regulated association of protein tyrosine kinases, Jakl and Jak2, with its receptor.]. Biol. Chem. 269, 14333-14336.

112

ALFONS BILLIAU

65. Greenlund, A. C., Farrar, M. A., Viviano, B. L., and Schreiber, R. D. (1994). Ligandinduced IFN gamma receptor tyrosine phosphorylation couples the receptor to its signal transduction system (p91). EMBO J. 13, 1591-1600. 66. Heim, M. H., Kerr, I. M. Stark, G. R., and Darnell, J. E., Jr. (1995). Contribution of STAT SH2 groups to specific interferon signaling by the Jak-STAT pathway. Science 267, 1347-1349. 67. Stark, G., and Kerr, I. M. (1992). Interferon-dependent signaling pathways: DNA elements transcription factors, mutations, and effects of viral proteins. J. Interjeron Res. 12, 147-151. 68. Schindler, C., Shuai, K., Precioso, V. R., and Darnell, J. E. J. (1992). Interferondependent tyrosine phosphorylation of a latent cytoplasmictranscription factor. Science 257, 809-813. 69. Shuai, K., Schindler, C., Precioso, V. R., and Darnell, J. E. J. (1992). Activation of transcription by IFN-y: Tyrosine phosphorylation of a 91-kD DNA binding protein. Science 258, 1808-1812. 70. Fujita, T., Sakaldbara, J., Sudo, Y., Miyamoto, M., Kimura, Y., and Taniguchi, T. (1988). Evidence for a nuclear factor(s), IRF-1, mediating induction and silencing properties to human IFN-P gene regulatory elements. EMBO J. 7 , 3397-3405. 71. Harada, H., Takahashi, E.-I., Itoh, S., Harada, K., Hori, T.-A,, and Taniguchi, T. (1994). Structure and regulation of the human interferon regulatory factor 1 (IRF-1) and IRF-2 genes: Implications for a gene network in the interferon system. Mol. Cell. Biol. 14, 1500-1509. 72. Pearse, R. N . , Feinman, R., and Ravetch, J. V. (1991).Characterization of the promotor of the human gene encoding the high-affinity IgC receptor: Transcriptional induction by gamma-interferon is mediated through common DNA response elements. Proc. Natl. Acad. Sci. USA 88, 11305-11309. 73. Wong, P., Severns, C. W., Guyer, N. B., and Wright, T. M. (1994).A unique palindromic element mediates gamma interferon induction of mig gene expression. Mol. Cell. Biol. 14, 914-922. 74. Fan, X.-D., Golberg, M., and Bloom, B. R. (1988).Interferon-y-induced transcriptional activation is mediated by protein Idnase. Proc. Natl. Acad. Sci. USA 85, 5122-5125. 75. Politis, A. D., Sivo, J., and Vogel, S. N. (1993) Multiple pathways of interferon-induced gene expression in murine macrophages. J. Leukocyte Biol. 53, 583-590. 76. Koide, Y., h a , Y., Nezu, N., and Yoshida, T. 0. (1988).Calcium i d u x and the Ca2+calmodulin complex are involved in interferon-y-induced expression of HLA class I1 molecules on HL-60 cells. Proc. NatE. Acad. Sci. USA 85, 3120-3124. 77. Ryu, K., Koide, Y., Yamashita, Y., and Yoshida, T. 0. (1993). Inhibition of tyrosine phosphorylation prevents IFN-y-induced HLA-DR molecule expression.J. Immunol. 150, 1253-1262. 78. Anderson, P., Yip, Y. K., and Vilcek, J. (1983). Human interferon-y is internalized and degraded by cultured fibroblasts. J. Biol. Chem. 258, 6497-6502. 79. Wietzerbin, J., Gaudelet, C. Aguet, M., and Falcoff, E. (1986). Binding and crosslinking of recombinant mouse interferon-y to receptors in mouse leukemic L1210 cells; Interferon-? internalization nd receptor down-regu1ation.J.Immunol. 136,24512455. 80. Bradley, J. R., Johnson, D. R., and Pober, J. S. (1993). Four different classes of inhibitors of receptor-mediated endocytosis decrease tumor necrosis factor-induced gene expression in human endothelial cells. J. Immunol. 150, 5544-5555. 81. Celada, A., and Schreiber, R. D. (1987). Internalization and degradation of receptorbound interferon-y by murine macrophages. Demonstration of receptor recycling. J. Zmmunol. 139, 147-153.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

113

82. Smith, M. R., Muegge, K., Keller, J. R., Kung, H.-F., Young, H. A,, and Durum, S. K. (1990).Direct evidence for an intracellular role for IFN-y.J. Immunol. 144,17771782. 83. Sanceau, J., Sondermeyer, P., Beranger, F., Falcoff, R., and Vaquero, C. (1987). Intracellular human gamma-interferon triggers an antiviral state in transformed murine L cells. Proc. Nutl. Acud. Sci. USA 84, 2906-2910. 84. Fidler, I. J., Fogler, W. E., Kleinermann, E. S., and Saiki, I. (1985). Abrogation of species specificity for activation of tumoricidal properties in macrophages by recombinant mouse or human interferon-gamma encapsulated in liposomes. J. bnmunol. 135,4289-4296. 85. Bader, T., and Wietzerbin, J. (1994).Nuclear accumulation of interferon gamma. Proc. Nutl. Acad. Sci. USA 91, 11831-11835. 86. Zu, X., and Jay, F. T. (1991). The E l functional epitope of the human interferon gamma is a nuclear targeting signal-like element. J. Biol. Chem. 266, 6023-6026. 87. Slodowski, O.,Bohm, J., Schone, B., and Otto, B. (1991). Carboy-terminal truncated rhuIFN-gamma with a substitution of Gln133 or Ser132 to leucine leads to higher biological activity than in the wild type. Eur. J. Biochem. 202, 1133-1140. 88. Holmes, E. W., Russell, P. M., Kinzler, G. J., Reckard, C. G., Flanigan, R. C., Thompson, K. D., and Bermes, E. W. J. (1992). Oxidative tryptophane metabolism in renal allograft recipients: Increased kynurenine synthesis is associated with inflammation and OKT3 therapy. Cytokine 4, 205-213. 89. Saito, K., Markey, S. P., and Heyes, M. P. (1991). Chronic effects of y-interferon on quinolinic acid and indoleamine-2,3-dioxygenasein brain of C57B16 mice. Bruin Res 546, 151-154. 90. Burke, F., Knowles, R. G., East, N., and Balkwill, F. R. (1995). The role of indoleamine 2,3-dioxygenase in the anti-turnour activity of human interferon-gamma in uiuo. Int. J. Cancer 60,115-122. 91. Cassatella, M. A,, Bazzoni, F., Flynn, R. M., Dusi, S., Trinchieri, G., and Rossi, F. (1990). Molecular basis of interferon-y and lipopolysaccharide enhancement of phagocyte respiratory burst capability. Studies on the gene expression of several NADPH oxidase components. /. Biol. Chem. 265,20241-20246. 92. Werner, E. R., Werner-Felmayer, G . , and Wachter, H. (1993). Tetrahydrobiopterin and cytokines. Proc. SOC. Exp. Biol. Med. 203, 1-12. 93. Moncada, S., and H i s s , A. (1993). Mechanisms of disease: The L-arginine-nitric oxide pathway. N . Engl. J. Med. 329, 2002-2012. 94. Snyder, S. H. (1993). Nitric oxide: first in a new class of neurotransmitters? Science 257,494-496. 95. Kamijo, R., Shapiro, D., Le, J., Huang, S., Aguet, M.,andvilcek, J. (1993). Generation of nitric oxide and induction of major histocompatibility complex class I1 antigen in macrophages from mice lacking the interferon gamma receptor. Proc. Nutl. Acud. Sci. USA 90,6626-6630. 96. Stamler, J. S., Singel, D. J.. and Loscalzo, J. (1993). Biochemistry of nitric oxide and its redox froms. Science 258, 1898-1902. 97. Weinberg, E. D. (1993).The iron-withholdingdefense system.ASM News 59,559-562. 98. Dijkmans, R., and Billiau, A. (1991). Interferon-yllipopolysaccharide-treated mouse embryonic fibroblasts are killed by a glycolysisL-arginine-dependent process accompanied by depression of mitochondria1 respiration. Eur. J. Biochem. 202, 151-159. 99. Drapier, J.-C., and Hibbs, J. B. (1988). Differentiation of murine macrophages to express nonspecific cytotoxicity for tumor cells results in L-arginine-dependent inhibi-

114

ALFONS BILLIAU

tion of mitochondrial iron-sulfur enzymes in the macrophage effector cells.]. lmmunol. 140, 2829-2838. 100. Henry, Y., Lepoivre, M., Drapier, J.-C., Ducrocq, C., Boucher, J.-L., and Guissani, A. (1993). EPR characterization of molecular targets for NO in mammalian cells and organelles. FASEB J. 7, 1124-1134. 101. Green, S. J., Nacy, C. A,, and Meltzer, M. S. (1991). Cytokine-induced synthesis of nitrogen oxides in macrophages: A protective host response to Leishmania and other intracellular pathogens. J. Leukocyte Biol. 50, 93- 103. 102. Rajas, A,, Padrbn, J., Caveda, L., Palacios, M., and Moncada, S. (1993). Role of nitric oxide pathway in the protection against lethal endotoxemia afforded by low doses of lipopolysaccharide. Biochem. Biophys. Re.s. Commun. 191, 441-446. 103. Merrill, J. E., Ignarro, L. J., Sherman, M. P., Melinek, J., and Lane, T. E. (1993). Micraglid cell cytotoxicity of oligodendrocytes is mediated through nitric oxide. J. Immunol. 151, 2132-2141. 104. Southern, C., Schulster, D., and Green, I. C. (1990). Inhibition of insulin secretion by interleukin-l/3 and tumour necrosis factor-a via an L-arginine-dependent nitric oxide generating mechanism. FEBS. Lett. 276,42-44. 105. Karupiah, G., Xie, Q., Buller, R. M. L., Nathan, C., Duarte, C., and MacMicking, J. D. (1993).Inhibition ofviral replication by interferon-y-induced nitric oxide synthase. Science 261, 1445-1448. 106. Harris, N., Buller, R. M. L., and Karupiah, G. (1995). Gamma interferon-induced, nitric oxide-mediated inhibition of vaccinia virus replication. J. Virol. 69, 910-915. 107. Taub, D. D., Lloyd, A. R., Conlon, K., Wang, J. M., Ortaldo, J. R., Harada, A., Matsushima, K., Kelvin, J., and Oppenheim, J. J. (1993). Recombinant human interferon-inducible protein 10 is a chemoattractant for human monocytes and T lymphocytes and promotes T cell adhesion to endothelial ce1ls.J. Exp. Med. 177,1809. 108. Lafuse, W. P., Brown, D., Castle, L., and Zwilling, B. S. (1995).Cloning and characterization of a novel cDNA that is IFN-gamma-induced in mouse peritoneal macrophages and encodes a putative GTP-binding protein. J. Leukocyte B i d . 57, 477-483. 109. Gilly, M., and Wall, R. (1992). The IRG-47 gene is IFN-y induced in B cells and encodes a protein with GTP-binding motifs. J. Immunol. 148, 3275-3281. 110. Cheng, Y. S.-E., Patterson, C. E., and Staeheli, P. (1991).Interferon-induced guanylatebinding proteins lack an N(T)KXD consensus motif and bind GMP in addition to GDP and GTP. Mol. Cell. Biol. 11, 4717-4725. 111. Wynn, T. A., Nicolet, C. M., and Paulnock, D. M. (1991).Identification and characterization of a new gene family induced during macrophage activation. J. Immunol. 147,4384-4392. 112. Symington, F. W. (1989). Lymphotoxin tumor necrosis factor, and gainma interferon are cytostatic for normal human keratinocytes. J. Inuest. D e m t o l . 92, 798-805. 113. Saunders, N . A., and Jetten, A. M. (1994).Control of growth regulatoly and differentiation-specific genes in human epidermal keratinocytesby interferon gamma. Antagonism by retinoic acid and transforming growth factor pl. J. Biol. Chem. 269, 2016-2022. 114. Wu, A. J., Kurrasch, R. H., Katz, J., Fox, P. C., Baum, B. J., and Atkinson, J. C. (1994). Effect of tumor necrosis factor-a and interferon-gamma on the growth of a human salivary gland cell line. J. Cell. Physiol. 161, 217-226. 115. Steiniger, B., Falk, P., and Van der Meide, P. (1988). Interferon-y in viva. Induction and loss of Class 11 MHC antigens and immature myelomonocytic cells in rat organs. Eur. J. Immunol. 18,661-669. 116. Pujol-Borrell, R., Todd, I., Doshi, M., Botazzo, G. F., Sutton, D., Gray, D., Adolf, G. R., and Feldman, M. (1987). HLA Class I1 induction in human islet cells by interferon-y plus tumor necrosis factor or lymphotoxin. Nature 326, 304-306.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

115

117. Petroni, K., Shen, L., and Guyre, P. M. (1988). Modulation of human polymorphonuclear leukocyte IgG Fc receptors and Fc receptor-mediated functions by IFN-y and glucocorticoids.1. Immunol. 140,3467-3472. 118. Guyre, P. M., Morganelli,P. M., and Miller, R. (1983). Recombinant immuneinterferon increases immunoglobulin G Fc receptors on cultured human mononuclear phagocytes. 1.Clin. Invest. 72, 393-397. 119. Loughlin, A. J., Woodroofe, M. N., and Cuzner, M. L. (1992). Regulation of Fc receptor and major histocompatibility complex antigen expression on isolated rate microglia by tumour necrosis factor, interleukin-l and lipopolysaccharide: Effects on interferon-gamma induced activation. Immunology, 75, 170-175. 120. Kontny, U., Kurane, I., and Ennis, F. (1988). Gamma interferon augments Fey receptor-mediated dengue virus infection of human monocytic cells.]. Virol.62,39283933. 121. Nardy Fejes-Toth, A,, and Guyre, P. M. (1984). Recombinant human immune interferon induces increased IgE receptor expression on the human monocyte cell line U937.1. Immnunol. 133, 1914-1919. 122. Pancr6, V . , Joseph, M., Capron, A,, Wietzerbin, J., Kusnierz, J.-P., Vorng, H., and Auriauk, C. (1988). Recombinant human interferon-y induces increased IgE receptor expression on human platelets. Eur. I. Zrnmunol. 18, 829-832. 123. Dustin, M. L., Singer, K. H., Tuck, D. T., and Springer, T. A. (1988). Adhesion of T lymphoblasts to epidermal keartinocytes is regulated by interferon y and is mediated by intercellular adhesion molecule 1 (ICAM-1).1.Exp. Med. 167, 1323-1340. 124. Freedman, A. S., Freeman, G. J., Rhynhart, K., and Nadler, L. M. (1991). Selective induction of B7/BB-1 on interferon-gamma stimulated monocytes: A potential mechanism for amplification of T cell activation through the CD28 pathway. Cell Imnmunol. 137,429-437. 125. Harding, F. A., McArthur, J. G., Gross, J. A,, Raulet, D. H., and Allison, J. P. (1992). CD28-mediated signaling co-stimulates murine T cells and prevents induction of anergy in T-cell clones. Nature 356, 607-609. 126. Marth, C., Fuith, L. C., Bock, G., Daxenhichler, G., and Dapunt, 0.(1989).Modulation of ovarian carcinoma tumor marker CA-125 by gamma interferon. Cancer Res. 49, 6538-6542. 127. Greiner, J. W., Guadagni, F., Hand, P. H., Pestka, S., Noguchi, P., and Fisher, P. B. (1990). Augmentation of tumor antigen expression by recombinant human interferons: Enhanced targeting of monoclonal antibodies to carcinomas. Cancer Treatment Res. 51, 413-432. 128. Leon, J. A,, Mesa, T. R., Gutierrez, M. C., Estabrook, A,, Greiner, J. W., Schlom, J., and Fisher, P. B. (1989).Increased surface expression and shedding of tumor associated antigens by human interferons or phorbol ester tumor promoters. Anticancer Res. 9, 1639-1647. 129. Wand-Wurttenberger, A,, Schoel, B., Ivanyi, J., and Kaufmann, S. H. E. (1991).Surface expression by mononuclear phagocytes of an epitope shared with mycobacterial heat shock protein 60. Eur. J. Immunol. 21, 1089-1092. 130. Munn, D. H . , Beall, A. C., Song, D., Wrenn, R. W., and Throckmorton, D. C. (1995). Activation-induced apoptosis in human macrophages: Developmental regulation of a novel cell death pathway by macrophage colony-stimulating factor and interferon gamma.]. Exp. Med. 181, 127-136. 131. Mosser, D. M., and Handman, E. (1992). Treatment of murine macrophages with interferon-gamma inhibits their ability to bind leishmania promastigotes. J. Leukocyte Biol. 52, 369-376.

116

ALFONS BILLIAU

132. Frasca, D., Adorini, L., Landolfo, S., and Dona, G. (1988). Enhancement of suppressor T cell activity by injection of anti-IFN-y monoclonal antibody.]. Immunol. 140,41034107. 133. Ishikura, H., Jayaraman, S., Kuchroo, V., Diamond, B., Saito, S., and Dorf, M. E. (1989). Functional analysis of cloned macrophage hybridomas. J. Immunol. 143, 414-419. 134. Noma, T., and Dorf, M. E. (1985). Modulation of suppressor T cell induction with y-interferon. I. lmmunol. 135, 3655-3660. 135. Luster, A. D., Unkeless, J. C., and Ravetch, J. V. (1985). y-Interferon transcriptionally regulates an early-response gene containing homology to platelet proteins. Nature 315, 672. 136. Hamilton, T. A,, Bredon, N., Ohmori, Y., and Tannenbaum, C. S. (1989). IFN-y and IFN-P independently stimulate the expression of lipopolysaccharide-inducible genes in murine peritoneal macrophages. 1.lmmunol. 142, 2325-2331, 137. Ohmori, Y., and Hamilton, T. A. (1994). IFN-gamma selectively inhibits lipopolysaccharide-inducible JEImonocyte chemoattractant protein-1 and KC/GRO/melanoma growth-stimulatingactivity gene expression in mouse peritoneal macrophages.]. Immunol. 153, 2204-2212. 138. Gusella, G . L., Musso, T., Bosco, M. C., Espinoza-Delgado, I., Matsushima, K., and Varesio, L. (1993). IL-2 up-regulates but IFN-y suppresses IL-8 expression in human monocytes. J. Immunol. 151, 2725-2732. 139. Heda, G. D., Mardente, S., Weiner, L., and Schmaier, A. H. (1990). Interferon gamma increases in vitro and in vivo expression of C1 inhibitor. Blood 75, 2401-2407. 140. Lotz, M., and Zuraw, B. L. (1987). Interferon-y is a major regulator of C1-inhibitor synthesis by human blood monocytes. 1.Immunol. 139,3382-3387. 141. Grabbe, S., Bruvers, S., Beissert, S., and Cranstein, R. D. (1994). Interferon-gamma inhibits tumor antigen presentation by epidermal antigen-presenting cells.]. Leukocyte Bi01. 55, 695-701. 142. Simon, J. C., Cruz, P. D., Bergstresser, P. R., Davis, L. S., and Tigelaar, R. E. (1991). Phorbol myristate acetate-activated keratinocytes stimulate proliferation of resting peripheral blood mononuclear lymphocytesvia a MHC-independent, but protein kinase C- and intercellular adhesion molecule-1-dependent mechanism. 1. Immunol. 146, 476-484. 143. LaSalle, J. M., and Hafier, D. A. (1994). T cell anergy. FASEB]. 8, 601-608. 144. Huynh, H. K., and Dorovini-Zis, K. (1993). Effects of interferon-y on primary cultures of human brain microvessel endothelial cells. Am. 1.Puthol. 142, 1265-1278. 145. Dustin, M. L., Rothlein, R., Bhan, A. K., Dinarello, C. E., and Springer, T. A. (1993). Tissue distribution, biochemistry and function of a natural adherence molecule (ICAM1). Induction by IL-la and interferon-y. 1.lmmunol. 137, 245-254. 146. McCarron, R. M., Wang, L., Racke, M. K., McFarlin, D. E., and Spatz, M. (1993). Cytokine-regulated adhesion between encephalitogenic T lymphocytes and cerebrovascular endothelid cells. 1.Nacroimmunol. 43, 23-30. 147. Yu, C. L., Haskard, D. O., Cavender, D., Johnson, A., and Ziff, M. (1985). Human gamma interferon increases the binding of T lymphocytes to endothelial cells. Clin Exp. lmmunol. 62, 554-560. 148. Male, D., Rahman, J., Linke, A., Zhao, W., and Hickey, W. (1995). An interferoninducible molecular on brain endothelium which controls lymphocyte adhesion mediated by integrins. Immunology 84, 453-460. 149. Arenzana-Seisdedos, F., Virelizier, J. L., and Fiers, W. (1985). Interferons as macrophage activating factors. 111. Preferential effects of interferon-y on the interleukin-1 secretory potential of fresh or aged human monocytes. I. Immunol. 134,2444-2448.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

117

150. Haq, A. U., Rinehart, J. J., and Maca, R. D. (1985). The effect of gamma interferon on IL-1 secretion of in vitro differentiated human niacrophages. 1. Leukocyte Biol. 38, 735-746. 151. Miossec, P., and Ziff, M. (1986). Immune interferon enhances the production of interleukin 1 by human endothelial cells stimulated with lipopolysaccharide.J. Immunol. 137, 2848-2852. 152. Johnson, W. J., Breton, J., Newman-Tarr, T., Connor, J. R., Meunier, P. C., and Dalton, B. J. (1990).Interleukin-1 release by rat synovial cells is dependent on sequential treatment with y-interferon and lipopolysaccharide.Arthritis Rheum. 33,261-267. 153. Marfaing-Koka, A,, Devergne, O., Corgone, C., Portier, A,, Schall, T. J., Galanaud, P., and Eniilie, D. (1995). Regulation of the production of the RANTES chemokine by endothelial cells: Synergistic induction by IFN-gamma plus TNF-a and inhibition by IL-4 and IL-13.1. Immunol. 154, 1870-1878. 154. A g g a r d , B. B., Eessalu, T. E.. and Hass, P. E. (1985). Characterization of receptors for human tumor necrosis factor and their regulation by y-interferon. Nature 318, 665-667. 155. Ruggiero,V., Tavemier, J., Fiers, W., and Baghoni, C. (1986).Induction ofthe synthesis of tumor necrosis factor receptors by interferon-y. J. Biol. Chem. 136, 2445-2450. 156. Tsujimoto, M., and Vilcek, J. (1986). Tumor necrosis factor receptors in HeLa cells and their regulation by interferon-y. J . B i d . Chem. 261, 5384-5388. 157. Drapier, J.-C., and Wietzerbin, J. (1991). IFN-y reduces specific binding of tumor necrosis factor on murine macrophages. J. Immunol. 146, 1198-1203. 158. Paul, W. (1991). Interleukin-4: A prototypic immunoregulatory lymphokine. Blood 77, 1859-1870. 159. Plum, J., De Smedt, M., Billiau, A,, and Heremans, H. (1991). IFN-c reverses IL-4 inhibition of fetal thymus growth in organ culture. J. Immunol. 147, 50-54. 160. Caspari, A. A., Jenkins, M. K., and Katz, S. I. (1988).Class I1 MHC-bearing keratinoytes induce antigen-specific unresponsiveness in hapten-specific Thl clones. J. Immunol. 141,2216-2220. 161. Cautam, S., Tebo, J. M., and Hamilton, T. A. (1992). IL-4 suppresses cytokine gene expression induced by IFN-y andlor IL-2 in murine peritoneal macrophages.J . Immunol. 148, 1725-1730. 162. Belosevic, M., Davis, C. E., Meltzer, M. S., and Nacy, C. A. (1988). Regulation of activated macrophage antimicrobial activities. Identification of lymphokines that cooperate with IFN-gamma for induction of resistance to infection. J. Zmmunol. 141, 890-896. 163. Lohoff, M., Marsig, E., and Rollinghof, M. (1990). Murine IL-4 antagonizes the protective effects of IFN on virus-mediated lysis of murine L929 fibroblast cells. J. Immunol. 144, 960-963. 164. Lamer, A. C., Petricoin, E. F., Nakagawa, Y., and Finbloom, D. S. (1993). IL-4 attenuates the transcriptional activation of both IFN-cy- and IFN-y-induced cellular gene expression in monocytes and monocytic cell lines. J. Immunol. 150, 1944-1950. 165. Pbne, J., Rousset, F., Brikre, F., Chrbtien, I., Palliard, X., Banchereau, J., Spits, H., and De Vries, J. E. (1988). IgE production by normal human B cells induced by doreactive T cell clones is mediated by IL-4 and suppressed by IFN-y. J. Immunol. 141, 1218-1224. 166. Pkne, J., Rousset, F., BriBre, F., Chrktien, I., Bonnefoy, J.-Y., Spits, H., Yokota, T., Arai, N., Arai, K.-I., Banchereau, J., and De Vries, J. E. (1988). IgE production by normal human lymphocytes is induced by interleukin 4 and suppressed by interferons y and a and prostaglandin E2. Proc. Natl. Acud. Sci. USA 85, 6880-6884.

118

ALFONS BILLIAU

167. Peleman, R., Wu, J., Fargeas, C., and Delespesse, G. (1989). Recombinant interleukin 4 suppresses the production of interferon gamma by the human mononuclear cells. J. Exp. Med. 170, 1751-1756. 168. Vercelli, D., Jabara, H. H., Lauener, R. P., and Geha, R. S. (1990). IL-4 inhibits the synthesis of IFN-y and induces the synthesis of IgE in human mixed lymphocyte cultures. 1.lmmunol. 144, 570-573. 169. Wagner, F., Fischer, N., Lersch, C., Hart, R., and Dancygier, H. (1989). Interleukin 4 inhibits interleukin %induced production of its functional antagonist, interferon gamma. lmmunol. Lett. 21, 237-242. 170. Deng, W., Ohmori, Y., and Hamilton, T. A. (1994). Mechanisms of IL-4-mediated suppression of IP-10 gene expression in murine macrophages.J. lmmunol. 153,21302136. 171. Fertsch, D., Schoenberg, D. R., Germain, R. N., Tou, J. Y., and Vogel, S. N. (1987). Induction of macophage Ia antigen expression by rIFN-gamma and down-regulation by IFN-alphdbeta and dexamethasone are mediated by changes in steady-state levels of Ia mRNA. J. lmmunol. 139, 244-249. 172. Ling, P. D., Warren, M. K., and Vogel, S. N. (1985). Antagonistic effect of interferoncU/p on the interferon-y-induced expression of Ia antigen in murine macrophages. J. Immunol. 135, 1857-1863. 173. Bama, B. P., Chou, S. M., Jacobs, B., Yen-Lieberman, B., and Ransohoff, R. M. (1989). Interferon-p impairs induction of HLA-DR antigen expression in cultured adult human astrocytes. J. Neuroimmunol. 23, 45-53. 174. Ransohoff, R. M., Devajyothi, C., Estes, M. L., Babcock, G., Rudick, R. A,, Frohman, E. M., and Barna, B. P. (1991). Interferon-beta specificallyinhibits interferon-gammainduced class 11 major histocompatibility complex gene transcription in a human astrocytoma cell line. /. Neuroimmunol. 33, 103-112. 175. Chantry, D., Turner, M., Abney, A,, and Feldman, M. (1989). Modulation of cytokine production by transforming growth factor p. J. Immunol. 142, 4295-4300. 176. Espevik, T., Figari, I. S., Shalaby, M. R., Lewis, G . D., Shepard, H. M., and Palladino, M. S . J. (1987). Inhibition of cytokine production by cyclosporin A and transforming growth factor. J. Exp. Med. 166, 571-576. 177. Czamiecki, C. W., Chiu, H. H., Wong, G. H. W., McCabe, S. M., and Palladino, M. A. (1988). Transforming growth factor P modulates the expression of Class 11 histocompatibility antigens on human cells. J. Immunol. 140, 4217-4223. 178. Ding, A., Nathan, C. F., Graycar, J., Derynck, R., Stuehr, D. J., and Srimal, S. (1990). Macrophage deactivating factor and transforming growth factors-pl, -p2, and -p3 inhibit induction of macrophage nitrogen oldde synthesis by IFN-y. J. Immunol. 145,940-944. 179. Varga, J., Olsen, A., Herhal, J., Constantine, G., Rosenbloom, J., and Jiminez, S. A. (1990). Interferon-y reverses the stimulation of collagen but not fibronectin gene expression by transforming growth factor-p in normal human fibroblasts. Eur. J. Clin. Invest. 20, 487-493. 180. Moore, K. W., O’Garra, A., De Wad Male@, R., Vieira, P., and Mosmann, T. R. (1993). Interleukin-10. Annu. Reu. lmmunol. 11, 165-190. 181. Wanidworanun, C., and Strober, W. (1993). Predominant role of tumor necrosis factora in human monocyte IL-10 synthesis. ]. lmmunol. 151, 6853-6861. 182. Gazinelli, R., Oswald, I. P., James, S. L., and Sher, A. (1992). IL-10 inhibits parasite killing and nitrogen oxide production by IFN-y-activated macrophages. J. Immunol. 148, 1792-1796.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

119

183. Florquin, S., Amraoui, Z., Abramowicz, D., and Goldman, M. (1994). Systemic release and protective role of IL-10 in staphylococcal entertoxin B-induced shock in mice. J. linmunol. 153, 2618-2623. 184. Chomarat, P., Rissoan, M.-C., Banchereau, J., and Miossec, P. (1993). Interferon gamma inhibits interleuldn 10 production by monocytes. J. Exp. Med. 177,523-527. 185. Zurawski, G., and De Vries, J. (1994). Interleukin 13, an interleukin 4-like cytoldne that acts on monocytes and B cells, but not on T cells. lmmunol. To&y 15, 19-26. 186. Doherty, T. M., Kastelein, R., Menon, S., Andrade, S., and Coffman, R. L. (1993). Modulation of murine macrophage function by IL-13. J. lmmunol. 151, 7151-7160. 187. Zimmer, T., and Jones, P. P. (1990). Combined effects of tumor necrosis factor-a, prostaglandin E2, and corticosterone on induced Ia expression on murine macrophages. J linmunol. 145, 1167. 188. Watanabe, Y., and Jacob, C. 0.(1991).Regulation of MHC Class I1 antigen expression. Opposing effects of tumor necrosis factor-a on IFN-y-induced HLA-DR and Ia expression depends on the maturation and differentiation stage of the cell. J. lmmunol. 146,899. 189. Beaman, M. H., Hunter, C. A,, and Remington, J. S. (1994). Enhancement of intracellular replication of Toxoplusmu gondii by IL-6: Interactions with IFN-gamma and TNFa. J. lmmunol. 153, 4583-4587. 190. Landolfo, S., Gariglio, M., Gribaudo, G., Jemma, C., Giovarelli, M., and Cavallo, G. (1988). Interferon-y is not an antiviral, but a growth-promoting factor for T lymphocytes. Eur. J. lmmunol. 18, 503-509. 191. Fernandez-Botran, R., Sanders, V. M., Mosmann, T. R., and Vitetta, E. S. (1988). Lymphokine-mediated regulation of the proliferative response of clones of T helper 1 and T helper 2 cells. J. Exp. Med. 168, 543-558. 192. Gajewsld, T. F., and Fitch, F. W. (1988).Anti-proliferative effect of IFN-y in immune regulation. I. IFN-y inhibits the proliferation of Th2 but not Thl murine helper T lymphocyte clones. J. lmmunol. 140, 4252. 193. Wynn, T. A,, Jankovic, D., Hieny, S., Zioncheck, K., Jardieu, P., Cheever, A. W., and Sher, A. (1995). IL-12 exacerbates rather than suppresses T helper 2-dependent pathology in the absence of endogenous IFN-7. J. lmmunol. 154,3999-4009. 194. Pernis, A,, Gupta, Gollob, K. J., Garfein, E., Coffman, R. L., Schindler, C., and Rothman, P. (1995). Lack of interferon y receptor p chain and the prevention of interferon y signalling in TH1 cells. Science 269, 245-247. 195. Liu, Y., and Janeway, C. A. J. (1990). Interferon y plays a critical role in induced cell death of effector T cell: A possible third mechanism of self-tolerance. J . Exp. Med. 172, 1735-1739. 196. Novelli, F., Di Pierro, F., Di Celle, P. F., Bertini, S., Affaticati, P., Garotta, G., and Forni, G. (1994). Environmental signals influencing expression of the IFN- gamma receptor on human T cells control whether IFN-gamma promotes proliferation or apoptosis. J. lmmunol. 152, 496404. 197. Matthys, P., Froyen, G., Huang, S.,Sobis, H., Van Damme, J., Aguet, M., and Billiau. A. (1995). Interferon-y receptor-deficient mice are hypersensitive to the anti-CD3induced cytoldne release syndrome and thymocyte apoptosis: Protective role of endogenous nitric oxide. J. lmmunol. 155, 3823-3829. 198. Siegel, J. P. (1988). Effects on interferon-y on the activation of human T lymphocytes. Cell. lmmunol. 111, 461-472. 199. Landolfo, S., Cofano, F., Giovarelli, M., Prat, M., Cavallo, G., and Forni, G. (1985). Inhibition of interferon-gamma may suppress allograft reactivity by T lymphocytes in vitro and in viva. Science 229, 176-179.

120

ALFONS BILLIAU

200. Simon, M. M., Hochgeschwender, U., Brugger, U., and Landolfo, S. (1986). Monoclonal antibodies to interferon y inhibit interleukin 2-dependent induction of growth and maturation in lectidantigen-reactive cytolytic T lymphocyte precursors. 1.Immunol. 136,2755-2762. 201. Bucy, P., Hanto, D. W., Berens, E., and Schreiber, R. D. (1988). Lack of an obligate role for IFN-y in the primary in vitro mixed lymphocyte response. J. Immunol. 140,1148-1152. 202. Maraskowsky, E., Chen, W.-F., and Shortman, K. (1989). IL-2 and IFN-y are two necessary mymphokinesin the development of cytolyticT cells.]. Immunol. 143,12101214. 203. Sidman, C. L., Marshall, J. D., Schulz, L. D., Gray, P. W., and Johnson, H. M. (1984). ?-Interferon is one of several direct B cell-maturing lymphokines. Nature 309,801-803. 204. Abed, N. A,, Chace, J. H., Fleming, A. L., and Cowdery, J. S. (1994). Interferon-y regulation of B lymphocyte differentiation: Activation of B cells is a prerequisite for IFN-y-mediated inhibition of B cell differentiation. Cell Immunol. 153, 356-366. 205. Abed, N. S., Chace, J. H., and Cowdery, J. S. (1994). T cell-independent and T celldependent B cell activation increases IFN-gammaR expression and renders B cells sensitive to IFN-gamma-mediated inhibition. 1.Immunol. 153, 3369-3377. 206. G a y , B. A,, and Riley, R. L. (1994).IFN-gamma abrogates IL-7-dependent proliferation in pre-B cells, coinciding with onset of apoptosis. Immunology 81, 381-388. 207. Grawunder, U., Melchers, F., and Rolink, A. (1993).Interferon-y arrests proliferation and causes apoptosis in stromal cell/interleukin-7-dependentnormal murine pre-B cell lines and clones in vitro, but does not induce differentiation to surface immunoglobulinpositive B cells. Eur. ]. Immunol. 23, 544-551. 208. Trubiani, O., Bosco, D., and Di Primio, R. (1994). Interferon-gamma (IFN-gamma) induces programmed cell death in differentiated human leukemic B cell lines. Exp. Cell Res. 215, 23-27. 209. Buschle, M., Campana, D., Carding, S. R., Richard, C., Hoffbrand, A. V., and Brenner, M. K. (1993), Interferon y inhibits apoptotic cell death in B cell chronic lymphocytic leukemia. 1.Exp. Med. 177, 213-218. 210. White-Helman, S., and Wallace, J. H. (1989). Differential regulation of the immune response to SRBC by monoclonal antibodies to interferon-y. Proc. SOC. Exp. Biol. Med. 191,55-59. 211. Brunswick, M., and Lake, P. (1985). Obligatory role of gamma interferon in T-cell replacing factor-dependent, antigen-specific murine B cell responses. 1. Exp. Med. 161,953-971. 212. Kawano,Y., Noma, T., and Yata, J. (1994).Regulation of human IgG subclassproduction by cytokines: IFN-gamma and IL-6 act antagonistically in the induction of human IgCl but additively in the induction of IgG2.J Immunol. 153, 4948-4958. 213. Kitani, A., and Strober, W. (1993). Regulation of Cy subclass germ-line transcripts in human peripheral blood B cells. 1.Immunol. 151, 3478-3488. 214. Nakagawa, T., Hirano, T., Nakagawa, N., Yooshizaki, K., and Kishimoto, T. (1985). Effect of recombinant IL-2 and y-IFN on proliferation and differentiation of human B cells. J. Immunol. 134,959-966. 215. Finkelman, F. D., Katona, I. M., Mosmann, T. R., and Coffman, R. L. (1988). IFNy regulates the isotypes of Ig secreted during in vivo humoral immune responses. ]. Immunol. 140,1022-1027. 216. Snapper, C. M., Peschel, C., and Paul, W. E. (1993).IFN-y stimulates IgG2a secretion by murine B cells stimulated with bacterial lipopolysaccharide.]. Immunol. 140,21212127.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

121

21 7. Graham, M. B., Dalton, D. K., Giltinan, D., Braciale, V. L., Stewart, T. A,, and Braciale, T. J. (1993). Response to influenza infection in mice with a targeted disruption in the interferon gamma gene. 1.Exp. Med. 178, 1725-1732. 218. Dobber, R., Tielemans, M., and Nagelkerken, L. (1995).The in uiuo effects of neutralizing antibodies against IFN-gamma, IL-4, or IL-10 on the humoral immune response in young and aged mice. Cell Immunol. 160, 185-192. 219. Huang, S., Hendriks, W., Althage, A., Hemmi, S., Bluethmann, H., Kamijo, R., Vilcek, J., Zinkernagel, R. M., and Aguet, M. (1993). Immune response in mice that lack the interferon-gamma receptor. Science 259, 1742- 1745. 220. Bogen, S . A., Fogelman, I., and Abbas, A. K. (1993). Analysis of IL-2, IL-4, and IFNgamma-producing cells in situ during immune responses to protein antigens. 1.Immunol. 150,4197-4205. 221. King, C. L., Gallin, J. I., Malech, H. L., and Abramson, S. L. (1989). Regulation of immunoglobulin production in hyperimmunoglobulin E recurrent-infection syndrome by interferon y. Proc. Natl. Acad. Sci. USA 86, 10085-10089. 222. Rousset, F., Robert, J., Andary, M., Bonnin, J. P., Souillet, G., ChrCtien, I., Brikre, F., Pkne, J., and De Vries, J. (1991). Shifts in interleukin-4 and interferon-gamma production by T cells of patients with elevated IgE levels and the modulatory effects of these lymphokines on spontaneous IgE synthesis.]. Allergy Clin. Immunol. 87,58-69. 223. King, C. L., Low, C. C., and Nutman, T. B. (1993). IgE production in human helminth infection: Reciprocal interrelationship between IL-4 and IFN-y. 1. Immunol. 150, 1873-1880. 224. Li, J. T. C., Yunginger, J. W., Jaffe, H. S., Nelson, D. R., and Gleich, G . J. (1990). Lack of suppression of IgE production by recombinant interferon gamma: A controlled trial in patients with allergic rhinitis. I. Allergy Clin. Immunol. 85, 934-940. 225. Lack, G., Renz, H., Saloga, J., Bradley, K. L., Loader, J., Leung, D. Y. M., Larsen, G., and Gelfand, E. W. (1994). Nebulized but not parenteral IFN-y decreases IgE production and normalizes airways function in a murine model of allergen sensitization. 1. Immunol. 152, 2546-2554. 226. Reinhold, U., Wehrman, W., Kukel, S . , and Kreysel, H. W. (1990). Evidence that defective interferon-gamma production in atopic dermatitis patients is due to intrinsic abnormalities. Clin. Exp. Immunol. 79, 374-380. 227. Hanifin, J. M., Schneider, L. C., Leung, D. Y. M., Ellis, C. N., Jaffe, H. S., Izu, A. E., Bucalo, L. R., Hirabayashi, S. E., Tofte, S. J., Cantu-Gonzales, G., Milgrom, H., Boguniewicz, M., and Cooper, K. D. (1993). Recombinant interferon gamma therapy for atopic dermatitis. /. Am. Acad. Dermutol. 28, 189-197. 228. Horneff, G., Dirksen, U., and Wahn, V. (1994). Interferon-gamma for treatment of severe atopic eczema in two children. Clin. Inwest. 72, 400-403. 229. Wall, D. A,, Hamberg, S. D., Burakoff, S. J., Reynolds, D. S., Abbas, A. K., and Ferrara, L. M. (1988).Immunodeficiency in garft-versus-host disease. I. Mechanism of immune suppression. 1.Immunol. 140, 2970-2976. 230. Huchet, R., Bruley-Rosset, M., Mathiot, C., Grandjon, D., and Halle-Pannenko, 0. (1993). Involvement of IFN-gamma and transforming growth factor-/3 in graft-vs-host reaction-associated immunosuppression. I. Immunol. 150, 2517-2524. 231. Metzger, Z., Hoffeld, J. T., and Oppenheim, J. J. (1980). Macrophage-mediated suppression. I. evidence for participation of both hydrogen peroxide and prostaglandin in suppression of murine lymphocyte proliferation. J. Immunol. 124, 983-988. 232. Boraschi, D., Censini, S., and Tagliabue, A. (1984). Interferon-y reduces macrophagesuppressive activity by inhibiting prostaglandin E2 release and inducing interleukin 1 production. 1.Immunol. 133,764-768.

122

ALFONS BILLIAU

233. Albina, J. E., Abate, J. A,, and Henry, W. L. J. (1991). Nitric oxide production is required for murine resident peritoneal macrophages to suppress mitogen-stimulated T cell proliferation. Role of IFN-y in the induction of nitric oxide-synthesizingpathway. 1. Immunol. 147, 144-148. 234. Mills, C. D. (1991). Molecular basis of suppressor macrophages. Arginine metabolism via the nitric oxide synthetase pathway. 1.Immunol. 146, 2719-2723. 235. Twardzik, D. R., Mikovits, J. A,, Ranchalis, J. E., Purchio, A. F., Ellingswaorth, L., and Ruscetti, F. W. (1990). c-Interferon-induced activation of latent transforming growth-factor-/3 by human monocytes. Ann. NY Acud. Sci. 593, 276-284. 236. Adams, R. B., Planchon, S. M., and Roche, J. K. (1993).IFN-y modulation of epithelial barrier function. Time course, reversibility, and site of cytokine binding. 1.Immunol. 150, 2353-2363. 237. Flohr, T., Buwitt, U., Bonnekoh, B., Decker, T., and Bottger, E. C. (1992).Interferony regulates expression of a novel keratin class I gene. Eur. ]. Immunol. 22,975-979. 238. Vogel, U., Denecke, B., Troyanovsky, S. M., Leube, R. E., and Bottger, E. C. (1995). Transcriptional activation of psoriasis-associated cytokeratin K17 by interferongamma-Analysis of gamma-interferon activation sites. Eur.]. Biochem. 227,143-149. 239. Amento, E. P., Bhan, A. K., McCullagh, K. G., and Krane, S. M. (1985). Influence of gamma interferon on synovial fibroblast-like cells: Ia induction and inhibition of collagen synthesis,1. Clin. Inuest. 76, 836-848. 240. Jiminez, S. A,, Freundlich, B., and Rosenbloom, J. (1984).Selectiveinhibition ofhuman diploid fibroblast collagen synthesis by interferons. J. Clin. Inuest. 74, 1112-1 116. 241. Coldring, M. B., Sandell, L. J., Stephenson, M. L., and Krane, S. M. (1986). Immune interferon suppresses levels of procollagen mRNA and type 111 collagen synthesis in cultured human articular and costal chondrocytes. 1.Biol. Chern. 261, 9049-9056. 242. Rosenbloom, J., Feldman, G., Freundlich, B., and Jiminez, S. A. (1984).Transcriptional control of human diploid fibroblast collagen synthesis by gamma interferon. Biochem. Biophys. Res. Commun. 123, 365-372. 243. Granstein, R. D., Murphy, G. F., Margolis, R. J., Byme, M. H., and Amento, E. P. (1987). Gamma-interferon inhibits collagen synthesis in viva in the mouse. 1. Clin. Invest. 79, 1254-1258. 244. Granstein, R. D., Deak, M.-R., Jacques, S. L., Margolis, R. J,, Flotte, T. J., Whitaker, D., Long, F. H., and Amento, E. P. (1989). The systemic administration of gamma interferon inhibits collagen synthesis and acute inflammationin a murine skin wounding model. I. Inuest. Dermutol. 93, 18-27. 245. Gin, S. N., Hyde, D. M., and Marafino, B. J. J. (1986). Amelioration effect of murine interferon gamma on bleomycin-induced lung collagen fibrosis in mice. Biochem. Med. Metabol. Biol. 36, 194-197. 246. Patton, J. S., Shepard, H. M., Wilking, H., Lewis, G . , Aganval, B. B., Eessalu, T. E., Gavin, L. A., and Grunfeld, C. (1986). Interferons and tumor necrosis factor have similar catabolic effects on 3T3-Ll cells. Proc. Nutl. Acud. Sci. USA 83,8313-8317. 247. Doerrler, W., Feingold, K. R., and Grunfeld, C. (1994). Cytokines induce catabolic effects in cultured adipocytes by multiple mechanisms. Cytokine 6, 478-484. 248. Feingold, K. R., Doerrler, W., Dinarello, C. A,, Fiers, W., and Grunfeld, C. (1992). Stimulation of lipolysis in cultured fat cells by tumor necrosis factor. interleukin-1. and the interferon is blocked by inhibition of prostaglandin synthesis. Endocrinology 130. 10-16. 249. Sobel, R. A,, Mitchell, M. E., and Fondren, C. (1990).Intercellular adhesion molecule1 (ICAM-1) in cellular immune reactions in the human central nervous system. Am. 1.Puthol. 136, 1309-1316.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

123

250. Traugott, U., and Lebon, P. (1988). Multiple sclerosis: Involvement of interferons in lesion pathogenesis. Ann. Neurol. 24, 243-251. 251. Satoh, J., Kastrukoff, L. F., and Kim, S . U. (1991). Cytokine-induced expression of intercellular adhesion molecule-1 (ICAM-1) in cultured human oligodendrocytes and astrocytes. J. Neuropathol. Exp. Neurol. 50, 215-226. 252. Pulver, M., Carrel, S., Mach, J. P., and de Tribolet, N. (1987). Cultured human fetal astrocytes can be induced by interferon-y to express HLA-DR. J. Neuroimmunol. 14, 123-133. 253. Mauerhoff, T., Pujol-Borrell, R., Mirakian, R., and Botazzo, G. F. (1988). Differential expression and regulation of major histocompatibility complex (MHC) products in neural and glial cells of the human fetal brain. /. Neuroirnmunol. 18,271-289. 254. Frohman, E. M., Frohman, T. C., and Dustin, M. L. (1989).The induction of intercellular adhesion molecule-1 (ICAM-I) expression on human fetal astrocytes by interferony, tumor necrosis factor a, lymphotoxin, and interleukin-1: Relevance to intracerebral antigen presentation. J. Neuroimmunol. 23, 117-124. 255. Yong, V .W., Yong, F. P., Ruijs, T. C. G., Antel, J. P., and Kim, S. U. (1991).Expression and modulation of HLA-DR on cultured human adult astrocytes./. Neuropathol. Exp. Newol. 50, 16-28. 256. Aloisi, F., Borsellino, G., and Samoggia, P. (1992). Astrocyte cultures from human embryonic brain: Characterization and modulation of surface inolecules by inflammatory cytokines. J . Neurosci. Res. 32, 494-506. 257. Satoh, J., Paty, D. W., and Kim, S. U. (1995). Differential effects of beta and gamma interferons on expression of major histocompatibilitycomplex antigens and intercellular adhesion molecule-1 in cultured fetal human astrocytes. Neurology 45, 367-373. 258. Panek, R. B., Lee, Y.-J,, Lindstrom-Itoh, Y., Ting, J. P.-Y., and Benveniste, E. N . (1994). Characterization of astrocyte nuclear proteins involved in IFN-y and TNF-(r mediated class I1 MHC gene expression./. lmmunol. 153, 4555. 259. Panek, R. B., and Benveniste, E. N . (199Fj).Class I1 MHC gene expression in microglia. Regulation by the cytokines IFN-y, TNF-a and TCF-P. J. l~nmunol.154,2846-2854. 260. Li, Y . ,Atashi, J., Hayes, C., Reap, E., Hunt, S., 111, and Popko, B. (1995).Morphological and molecular response of the MOCH-1 oligodendrocyte cell line to serum and interferon-gamma: Possible implications for demyelinating disorders. J. Neurosci. Res. 40, 189-198. 261. Mamus, S . W., Beck Schroeder, S., and Zanjani, E. D. (1985). Suppression of normal human erythropoiesis by gamma interferon in vitro. Role of monocytes and T lymphocytes. J. Clin. Inue.st. 75, 1496-1503. 262. Broxmeyer, H. E . , Lu, L., Platzer, E., Feit, C., Juliano, L., and Rubin, B. Y. (1983). Cornparative analysis of the influences of human gamma, alpha and beta interferons on human multipotential (CFU-GEMM), erythroid (BFU-E) and granulocytemacrophage (CFU-GM) progenitor cells. J. Zmtnunol. 131, 1300-1305. 263. Sarris, A. H . , Broxmeyer, H. E., Wirthmueller, U., Karasawas, N., Cooper, S., Lu, L., Krueger, J., and Ravetch, J. V. (1993). Human interferon-inducible protein 10: Expression and purification of recombinant protein demonstrate inhibition of early human hematopoietic progenitors. J . Exp. Med. 178, 1127-1132. 264. Shiohara, M., Koike, K., and Nakahata, T. (1993). Synergism of interferon-gamma and stem cell factor on the development of murine hematopoietic progenitors in serum-free cdture. Blood 81, 1435-1441. 265. Caux, C., Moreau, I., Saeland, S., and Banchereau, J. (1992). Interferon-gamma enhances factor-dependent myeloid proliferation of human CD34+ hematopoietic progenitor cells. Blood 79,2628-2635.

124

ALFONS BILLIAU

266. Means, R. T. J., and Krantz, S. B. (1991). Inhibition of human erythroid colonyforming units by interferon-gamma can be corrected by recombinant human erythropoietin. Blood 78, 2564-2567. 267. Wang, C. Q., Udupa, K. B., and Lipschitz, D. A. (1995). Interferon-gamma exerts its negative regulatory effect primarily on the earliest stages of murine erythroid progenitor cell development. J. Cell. Physiol. 162, 134-138. 268. Nishikawa, T., Yamashita, S., Namba, H., Usa, T., Tominaga, T., Kimura, H., Izumi, M., and Nagataki, S. (1993). Interferon-y inhibition of human thyrotropin receptor gene expression.J. Clin. Endocrinol. Metab. 77, 1084-1089. 269. Kraiem, Z., Sobel, E., Sadeh, O., Kinarty, A,, and Lahat, N. (1990). Effects of yinterferon on DR antigen expression, growth, 3,5,3’-triiodothyronine secretion, iodide uptake and cyclic adenosine 3’,5’-monophosphate accumulation in cultured human thyroid cells. J. Clin. Endocrinol. Metab. 71, 817-824. 270. Kung, A. W. C., and Lau, K. S. (1990). Interferon-y inhibits thyrotropin-induced thyroglobulin gene transcription in cultured human thyrocytes. J. Clin. Endocrinol. Metab. 70, 1512-1517. 271. Asakawa, H., Hanafusa, T., Kobayashi, T., Takai, S.-I., Kono, N., and Tarui, S. (1992). Interferon-y reduces the thyroid peroxidase content of cultured human thyrocytes and inhibits its increase induced by thyrotropin.]. Clin. Endocrinol. Metab. 74,1331-1335. 272. Tang, K.-T., Braverman, L. E., and DeVito, W. J. (1995). Tumor necrosis factor-a and interferon-gamma modulate gene expression of type I 5‘-deiodinase, thyroid peroxidase, and thyroglobulin in FRTL-5 rat thyroid cells. Endocrinology 136, 881-888. 273. Vankelecom, H., Carmeliet, P., Heremans, H., Van Damme, J., Dijkmans, R., Billiau, A,, and Denef, C. (1990). Interferon-y inhibits stimulated adrenocorticotropin, prolactin, and growth hormone secretion in normal rat anterior pituitary cell cultures. Endocrinology 126,2919-2926. 274. Vankelecom,H., Andries, M., Billiau,A,, and Denef, C. (1992).Evidence that folliculostellate cells mediate the inhibitory effect of interferon-y on hormone secretion in rat anterior pituitary cell cultures. Endocrinology 130, 3537-3546. 275. Allaerts, W., Carmeliet, P., and Denef, C. (1990). New perspectives in the function of pituitary folliculo-stellate cells. Mol. Cell. Endocrinol. 71, 73-81. 276. Vankelecom, H., Carmeliet, P., Van Damme, J., Billiau, A., and Denef, C. (1989). Production of interleukin-6 by folliculo-stellate cells of the anterior pituitary gland in a histotypic cell aggregate culture system. Neuroendocrinology 49, 102-106. 277. Vankelecom, H., and Billiau,A. (1992).Interferon-y in neuroimmunology and endocrinology. Adv. Neuroimmuml. 2, 139-161. 278. Ottenhof, T. H. M., and Mutis, T. (1995). Role of cytotoxic cells in the protective immunity against and immunopathology of intracellular infections. Eur. J. Clin. Inuest. 25,371-377. 279. Belosevic, M., Finbloom, D. S., Van der Meide, P. H., Slayter, M. V., and Nacy, C. A. (1989). Administration of monoclonal anti-IFN-y antibodies in vivo abrogates natural resistance of C3HIHeN mice to infection with Leishmania major. 1.Immunol. 143,266-274. 280. Scott, P. (1991). IFN-gamma modulates the early development of Thl and Th2 responses in a murine model of cutaneous leishmaniasis.]. Immunol. 147,3149-3155. 281. Sileghem, M., Hamers, R., and De Baetselier, P. (1987). Experimental Trypanosoma brucei infections selectively suppress both interleukin 2 production and interleukin 2 receptor expression. Eur. J. Zmmunol. 17, 1417-1421.

I N F - ~ BIOLOGY : AND ROLE IN PATHOGENESIS

125

282. Billiau, A., Heremans, H., Vandekerckhove, F., and Dillen, C. (1987). Anti-interferonc antibody protects mice against the generalized Shwartzman reaction. Eur.]. lmmunol. 17, 1851-1854. 283. Billiau, A., and Vandekerckhove, F. (1991).Cytokines and their interactions with other inflammatory mediators in the pathogenesis of sepsis and septic shock. Eur. 1.Clin. Invest. 21, 559-573. 284. Heremans, H., Van Damme, J., Dillen, C., Dijkmans, R., and Billiau, A. (1990). Interferon-c, a mediator of lethal lipopolysaccharide-induced Shwartzman-like shock reactions in mice. J . Exp. Med. 171, 1853-1869. 285. Fennie, E. H., Lie, Y. S., Low, M.-A. L., Gribling, P., and Anderson, K. P. (1988). Reduced mortality in murine cytomegalovirus infected mice following prophylactic murine interferon-y treatment. Antiviral Res. 10, 27-39. 286. Haagmans, B. L., Van der Meide, P. H., Stals, F. S., Van den Eertwegh, A. J. M,, Claassen, E., Bruggeman, C. A., Horzinek, M. C., and Schijns, V. E. C. J. (1994). Suppression of rat cytomegalovirus replication by antibodies against gamma interferon. I. Virol. 68, 2305-2312. 287. Leist, T. P., Eppler, M., and Zinkernagel, R. M. (1989). Enhanced virus replication and inhibition of lymphocyhc choriomeningitis virus disease in anti-gamma interferontreated mice. J. Virol. 63, 2813-2819. 288. Heise, M. T., and Virgin, H. W. (1995). The T-cell-independent role of gamma interferon and tumor necrosis factor alpha in macrophage activation during murine cytomegalovirus and herpes simplex virus infections. J. Virol. 69, 904-909. 289. Lucchiari, M. A., Modolell, M., Eichmann, K., and Pereira, C. A. (1992). In vivo depletion of interferon-gamma leads to susceptibility of A/J mice to mouse hepatitis virus 3 infection. lmmunobiology 185, 475-482. 290. Stanton, G. J., Jordan, C., Hart, A,, Heard, H., Langford, M. P., and Baron, S. (1995). Nondetectable levels of interferon gamma is acritical host defense during the first day of herpes simplex virus infection. Microbial Pathogenesis 3, 179-183. 291. Alcami, A,, and Smith, G. L. (1995). Cytokine receptors encoded by poxviruses: A lesson in cytokine biology. lmmunol. Today 16, 474-478. 292. Biswas, P., Poli, G., Kintner, A. L., Justement, J . S., Stanley, S. K., Maury, W. J., Bressler, P., Orenstein, J. M., and Fauci, A. S. (1992). Interferon y modulates the expression of huan immunodeficiency virus in persistently infected promonocyhc cells by redirecting the production of virions to intracytoplasmic vacuoles. J. Exp. Med. 176, 739-750. 293. Koyanagi, Y., O’Brien, W. A,, Zhao, J. Q., Golde, Gasson, J. C., and Chen, I. S. Y. (1988). Cytokines alter production of HIV-1 from primary mononuclear phagocytes. Science 241, 1673-1675. 294. Jarpe, M. A., Hayes, M. P., Russell, J. K., Johnson, H. M., and Russell, S. W. (1989). Causal association of interferon-y with tumor regression.]. Interferon Res. 9,239-244. 295. Prat, M., Bretti, S., Amedeo, M., Landolfo, S., and Comoglio, P. M. (1987). Monoclonal antibodies against murine I F N y abrogate in vivo tumor immunity against RSV-induced cytotoxic T lymphocyte differentiation. 1.lmmunol. 138,4530-4533. 296. Gansbacher, B., Bannerji, R., Daniels, B., Zier, K., Cronin, K., and Cilboa, E. (1990). Retroviral vector-mediated gamma-interferon gene transfer to tumor cells generates potent and long-lasting antitumor immunity. Cancer Res. 50, 7820-7825. 297. Watanabe, Y., Kuribayashi, K., Miyatake, S., Nishihara, E., Nakayama, E., Taniyama, T., and Sakata, T. A. (1989). Endogenous expression of mouse interferon y cDNA in mouse neuroblastoma C1300 cells results in reduced tumorigenicity and augmented anti-tumor immunity. Proc. Natl. Acad. Sci. USA 86, 9456-9460.

126

ALFONS BILLIAU

298. Ferrantini, M., Giovarelli, M., Modesti, A., Musiani, P., Modica, A,, Venditti, M., Peretti, E., Lollini, P.-L., Nanni, P., Forni, G., and Belardelli, F. (1994).IFN-a1 gene expression into a metastatic murine adenocarcinoma (TS/A) results in CD8+ T cellmediated tumor rejection and development of antitumor immunity: Comparative studies with IFN-gamma-producing TS/A cells. ]. Immunol. 153, 4604-4615. 299. Matthys, P., Heremans, H., Opdenakker, G., and Billiau, A. (1991). Anti-interferonc-antibody treatment, growth of Lewis lung tumors in mice and tumor-associated cachexia. Eur. J. Cancer 27, 182-187. 300. Ramani, P., and Balkwill, F. R. (1987). Enhanced metastasis of a mouse carcinoma after in vitro treatment with interferon-y. Int. J. Cancer 40, 830-834. 301. Billiau, A,, and Matthys, P. (1992). Interferon-y, more of a cachectin than tumor necrosis factor. Cytokine 4, 259-263. 302. Langstein, H. N., Doherty, G. M., Fraker, D. L., Buresh, C. M., and Norton, J. A. (1991). The roles of y-interferon and tumor necrosis factor a in an experimental rat model of cancer cachexia. Cancer Res. 51,2302-2306. 303. Matthys, P., Dijkmans, R., Proost, P., Van Damme, J., Heremans, H., Sobis, H., and Billiau, A. (1991). Severe cachexia in mice inoculated with interferon-y-producing tumor cells. Int. 1. Cancer 49, 77-82. 304. Diamantstein, T., Eckert, R., Volk, H.-D., and Kupier-Weglinski,J.-W. (1988).Reversal by interferon-y of inhibition of delayed-type hypersensitivity induction by anti-CD4 or antiinterleukin 2 receptor (CD25) monoclonal antibodies. Evidence for the physiological role of the CD4+TH1+ subset in mice. Eur. 1. lmmunol. 18, 2101-2103. 305. Fong, T. A., and Mosmann, T. R. (1989).The role of IFN-y in delayed-type hypersensitivity mediated by T h l clones. I. Immunol. 143, 2887-2893. 306. Issekutz, T. B., Stoltz, J. M., and Van der Meide, P. (1988). Lymphocyte recruitment in delayed-type hypersensitivity. The role of IFN-y. 1. Immunol. 140, 2889-2893. 307. Geiger, K., and Sarvetnick, N. (1994). Local production of IFN-gamma abrogates the intraocular immune privilege in transgenic mice and prevents the induction of ACAID. 1.Immunol. 153,5239-5246. 308. Barker, J. N., Goodlad, J. R., Ross, E. L., Yu, C. C., Groves, R. W., and MacDonald, D. M. (1993). Increased epidermal cell proliferation in human skin in vivo following local administration of interferon-gamma. Am. 1.Pathol. 142, 1091-1097. 309. Nathan, C. F., Kaplan, G., Levis, W. R., Nusrat, A,, Witmer, M. D., Sherwin, S. A., Job, C. K., Path, F. R. C., Horowitz, C. R., Steinman, R. M., and Cohn, Z. A. (1986). Local and systemic effects of intradermal recombinant interferon-y in patients with lepromatous leprosy. N . Engl. 1.Med. 315, 6-15. 310. Taborski, U., Freitag, W., Heremans, H., and Knop, J. (1986). Inhibitory effects of interferon-y on the suppressor cell circuit in contact sensitivity. lmmunobiology 171,329-338. 311. Skoglund,C., Scheynius, A., Holmdahl, R., andvan der Meide, P. H. (1988).Enhancement of DTH reactions and inhibition of the expression of class I1 transplantation antigens by in vivo treatment with antibodies against y-interferon. Clin. Exp. Immunol. 71, 428-430. 312. Skoglund, C., and Scheynius, A. (1990). Effects of interferon-y treatment on the cutaneous DTH reaction in rats. Arch. Dermutol. Res. 282, 318-324. 313. Sarvetnick, N., Shizuru, J., Liggitt, L., McIntyre, B., Gregory, A., Parslow, T., and Stewart, T. (1990). Loss of pancreatic islet tolerance induced by beta-cell expression of interferon-gamma. Nature 346, 844-847. 314. Lee, M.-S., Von Herrath, M., Reiser, H., Oldstone, M. B. A., and Sarvetnick, N. (1995). Sensitization to self (virus) antigen by in situ expression of murine interferongamma. ]. Clin. Invest. 95, 486-492.

INF-y: BIOLOGY AND ROLE IN PATHOGENESIS

127

315. Kawakami, Y., Kuzuya, N., Watanabe, T., Uchiyama, Y., and Yamashita, K. (1990). Induction of experimental thyroiditis in mice by recombinant interferon y administration. Acta Endocrinol. 122, 41-48. 316. Tang, H., Mignon-Godefroy,K., Meroni, P. L., Garotta, G., Charreire, J., and Nicoletti, F. (1993).The effects of a monoclonal antibody to interferon-gamma on experimental autoimmune thyroiditis (EAT): Prevention of disease and decrease of EAT-specific T cells. Eur. J. lmmunol. 23, 275-278. 317. Campbell, I. L., Kay, T. W. H., Oxbrow, L., and Harrison, L. C. (1991). Essential role for interferon-? and interleukin-6 in autoimmune insulin-dependent diabetes in NODiWehi mice. J. Clin. Znuest. 87, 739-742. 318. Hartung, H.-P., Schafer, B., Van der Meide, P. H., Fierz, W., Heininger, K., and Toyka, K. V. (1990).The role of interferon-gamma in the pathogenesis of experimental autoimmune disease of the peripheral nervous system. Ann Neurol. 27, 247-257. 310. Strigard, K., Holmdahl, R., Vuri der Meide, P. H . , Klareskog, L., and Olsson, T. (1989). In vivo treatment of rats with monoclonal antibodies against gamma interferon: Effects on experimental allergic neuritis. Acta Neurol. Scand. 80, 201-207. 320. Billiau, A,, Heremans, H., Vandekerckhove, F., Dijkmans, R., Sobis, H., Meulepas, E., and Carton, H. (1988). Enhancement of experimental allergic encephalomyelitis in mice by antibodies against IFN-gamma. J. Zmmunol. 140, 1506-1510. 321. Duong, T. T., St. Louis, J., Gilbert, J. J., Finkelman, F. D., and Strejan, G. H. (1992). Effect of anti-interferon-y and anti-interleukin-2 monoclonal antibody treatment on the development of actively and passively induced experimental allergic encephalomyelitis in the SJLJJ mouse. J. Nercroirnmunol. 36, 105-115. 322. Caspi, R. R., Chan, C.-C., Grubbs, B. G., Silver, P. B., Wiggert, B., Parsa, C. F., Bahmanyar, S., Billiau, A., and Heremans, H. (1994). Endogenous systemic IFNgamma has a protective role against ocular autoimmunity in mice. J. Zmmunol. 152, 890-899. 323. Voorthuis, J. A. C., Uitdehaag, B. M. J., De Groot, C. J. A,, Goede, P. H., Van der Meide, P. H., and Dijkstra, C. D. (1990). Suppression of experimental allergic encephalomyelitis by intraventricular administration of interferon-gamma in rats. Clin. Exp. Zmmunol. 81, 183-188. 324. Duong, T. T., Finkelman, F. D., Singh, B., and Strejan, G. H. (1994). Effect of antiinterferon-gamma monoclonal antibody treatment on the development of experimental allergic encephalomyelitis in resistant mouse strains. J. Neuroimmunol. 53, 101-107. 325. Mustafa, M., Diener, P., Hojeberg, B., Van der Meide, P., and Olsson, T. (1991).T cell immunity and interferon-y secretion during experimental allergic encephalomyelitis in Lewis rats. J. Neuroimmunol. 31, 165-177. 326. McDonald, A. H., and Swanborg, R. H. (1988). Antigen-specific inhibition of immune interferon production by suppressor cells of autoimmune encephalomyelitis.J. Zrnmunol. 140, 1132-1138. 327. Cross, A. H., Misko, T. P., Lin, R. F., Hickey, W. F., Trotter J. L., and Tilton, R. G. (1994). Aminoguanidine, an inhibitor of inducible nitric oxide synthase, ameliorates experimental autoimmune encephalomyelitis in SJL mice. J. Clin. lnuest. 93, 26842690. 328. Mauritz, N. J. Holmdal, R, Jonsson, R., Van der Meide, P. H., Scheynius, A.. and Klareskog, L. (1988). Treatment with gamma-interferon triggers the onset of collagen arthritis in mice. Arthritis Rheum. 31, 1297-1304. 329. Nakajima, H., Takamori, H., Hiyama, T., and Tsukuda, W. (1990). The effect of treatment with interferon-gamma on type 11 collegen-induced arthritis. Clin. Exp. Immunol. 81,441-445.

128

ALFONS BILLIAU

330. Wahl, S. M., Allen, J. B. Ohura, K., Chenoweth, D. W., and Hand, A. R. (1991). IFNy inhibits inflammatorycell recruitment and the evolution of bacterial cell wall-induced arthritis. 1.lmmunol. 146,95-100 331. Matthys, P., Heremans, H., Dillen, C., and Bilhau, A. (1989). Effects of interferony and anti-interferon-y antibodies on experimental arthritis in mice.]. Interferon Rex 9S, 13O.[Abstract] 332. Alvaro-Gracia, J. M., Zvaifler, N. J., and Firestein, G. A. (1990). Cytokines in chronic inflammatory arthritis. V. Mutual antagonism between interferon-gamma and tumor necrosis actor-alpha on HLA-DR expression, proliferation, collagenaseproduction, and granulocyte macrophage colony stimulating factor production by rheumatoid arthritis synoviocytes.J. Clin. Invest. 86, 1790-1798. 333. Nakajima, H., Hiyama, T., Tsukuda, W., Warabi, H., Uchida, S., and Hirose, S. (1990). Effects of interferon gamma on cultured synovial cell from patients with rheumatoid arthritis: Inhibition of cell growth, prostaglandin E2, and collagenase release. Ann. Rheum. Dis. 49,512-516. 334. Ruschen, S., Lemm, C., and Warnatz, H. (1989). Spontaneous and LPS-stimulated production of intracellular IL-1j3 by synovial macrophages in rheumatoid arthritis is inhibited by IFN-y. Clin. Exp. Immunol. 76, 246-251. 335. Andrews, H. J., Bunning, R. A. D., Dinarello, C. A., and Russell, R. G. G . (1989). Modulation of human chondrocyte metabolism by recombinant human interferongamma: In vitro effects on basal and IL-1-stimulated proteinase production, cartilage degradation and DNA synthesis. Biochim. Biophys. Acta 1012, 128-134. 336. Bunning, R. A. D., and Russell, C . G . (1989). The effect of tumor necrosis factor-a and y-interferon on the resorption of human articular cartilage and on the production of prostaglandin E and of caseinase activity by human articular chondrocytes. Arthritis Rheum. 32, 780-784. 337. Gowen, M., and Mundy, G. R. (1986).Action of recombinant interleukin 1, interleukin 2 and interferon-y on bone resorption in vitro. J. lmmunol. 136, 2478-2482. 338. Takahashi, N., Mundy, G., and Roodman, G. D. (1986). Recombinant human interferon-? inhibits formation of human osteoclast-like cells.]. Immunol. 137,35443549. 339. Campbell, I. L., Oxbrow, L., Koulmanda, M., and Harrison, L. C. (1988).IFN--yinduces islet MHC antigens and enhances autoimmune, streptozotocin-induced diabetes in the mouse. J. lmmunol. 140,1111-1116. 340. Debraye-Sachs, M., Carnaud, C., Boitard, C., Cohen, H., Gresser, I., Bedossa, P., and Bach, J. (1991).Prevention of diabetes in NOD mice treated with antibody to murine IFNy. J Autoimmunity 4, 237-248. 341. Kay, T. W. H., Campbell, 1. L., Oxbrow, L., and Harrison, L. C. (1991).Overexpression of class I major histocompatibility complex accompanies insulitis in the non-obese diabetic mouse and is prevented by anti-interferon-y antibody. Diabetologia 34, 779-785. 342. Campbell, I. L., Oxbrow, L., and Harrison, L. C. (1991).Reduction in insulitis following administration of IFN-y and TNF-a in the NOD mouse.]. Autoimmunity 4,249-262. 343. Gu, D., and Sarvetnick, N. (1993). Epithelial cell proliferation and islet neogenesis in IFN-7 transgenic mice. Development 118, 33-46. 344. Wogensen, L., Molony, L., Gu, D., Krahl, T., Zhu, S., and Sarvetnick, N. (1994). Postnatal anti-interferon-y treatment prevents pancreatic inflammation in transgenic mice with &cell expression of interferon-y.]. lnterferon Res. 14, 111-116. 345. Lee, M.-S., and Sarvetnick, N. (1994). Induction of vascular addressins and adhesion molecules in the pancreas of IFN-gamma transgenic mice. 1.lmmunol. 152, 45974603.

INF-1: BIOLOGY AND ROLE IN PATHOGENESIS

129

346. Yamada, K., Otabe, S., Inada, C., Takane, N., and Nonaka, K. (1993). Nitric oxide and nitric oxide synthase mRNA induction in mouse islet cells by inteferon-y plus tumor necrosis factor-a. Biochem. Biophys. Rex Commun. 197, 22-27. 347. Rosenberg, A. S., Finbloom, D. S., Maniero, T. G.,Van der Meide, P. H., and Singer, A. (1990). Specific prolongation of MHC Class I1 disparate skin allografts by in vivo administration of anti-IFN-y monoclonal antibody. 1. Z?nmunoE. 144, 4648-4650. 348. Gorczynski, R. M. (1995). Regulation of IFN-gamma and IL-10 synthesis in oioo, as well as continuous antigen exposure, is associated with tolerance to murine skin allografts. Cell. Zmmunol. 160, 224-231. 349. Brok, H. P. M., Heidt, P. J., Van der Meide, P. H., Zurcher, C., and Vossen, J. (1993). Interferon-y prevents graft versus host disease after allogeneic bone marrow transplantation in mice. J . Zmmunol. 151, 6451-6459. 350. Cleveland, M. G . , Annable, C. R., and Klimpel, G. R. (1988). In vivo and in vitro production of IFN-/3 and IFN-y during graft vs host disease. J. Zmmunol. 141,33493356. 351. Garside, P., Reid, S., Steel, M., and Mowat, A. M. (1994). Differential cytokine production associated with distinct phases of murine graft-versus-host reaction. Zmmunology 82, 211-214. 352. Mowat, A. M. I. (1989). Antibodies to IFN-y prevent immunologically mediated intestinal damage in murine graft-versus-host reaction. Immunology 68, 18-23. 353. Donckier, V., Abramowicz, D., Bruyns, C., Florquin, S., Vanderhaeghen, M.-L., Amraoui, Z., Dubois, C.,Vandenabeele, p., and Goldman, M. (1994). IFN-gamma prevents TH2 cell-mediated pathology after neonatal injection of semiallogenic spleen cells in mice. J. lmmunol. 153, 2361-2368. 354. Heremans, H., Dijkmans, R., Sobis, H., Vandekerckhove, F., and Billiau, A. (1987). Regulation by interferons of the local inflammatory response to bacterial lipopolysaccharide. J . Zmmunol. 138, 4175-4179. 355. Matthys, P., Mitera, T., Heremans, H., Van Damme, J., and Bilhau, A. (1995). AntiIFN-y and anti-IL-6 antibodies affect staphylococcal enterotoxin B-induced weight loss, hypoglycemia and cytokine release in D-galactosamine-sensitizedand unsensitized mice. Infect. Zmmun. 63, 1158-1164. 356. Miethke, T., Wahl, C., Heeg, K., Echtenacher, B., Krammer, P. H., and Wagner, H. (1992).T cell-mediated lethal shock triggered in mice by the superantigen staphylococcal enterotoxin B: Critical role of tumor necrosis factor. J. Exp. Med. 175, 91-98. 357. Ferran, C . , Dy, M., Sheehan, K., Schreiber, R. Grau, G., Bluestone, J., Bach, J., and Chatenoud, L. (1991). Cascade modulation by anti-tumor necrosis factor monoclonal antibody of interferon-y, interleukin-3 and interleukin-6 release after triggering of the CD3R cell receptor activation pathway. Eur. J. lmmunol. 21, 2349-2353. 358. Matthys, P., Dillen, C., Proost, P., Heremans, H., Van Damme, J., and Balliau, A. (1993). Modification of the anti-CD3-induced cytokine release syndrome by antiinterferon-y or anti-interleukin-6 antibody treatment: Protective effects and biphasic changes in blood cytokine levels. Eur. J, Immunol. 23, 2209-2216. 359. Bach, E. A., Szabo, S. J., Dighe, A. S., Ashkenazi, A,, Aguet, M., Murphy, K. M., and Schreiber, R. D. (1995). Ligand-induced autoregulation of IFN-.)I receptor @-chain expression in T helper cell subsets. Science 270, 1215-1218. 360. Sakatsume, M., Igarashi, K., Winestock, K. D., Carotta, G., Lamer, A. C., and Finbloom, D. S. (1995). The Jak kinases differentially associate with the a and /3 (accessoryfactor) chains of the interferon gamma receptor to form a functional receptor unit capable of activating STAT transcription fact0rs.J. Bid. Chem. 270,17528-17534.

130

ALFONS BILLIAU

361. Guyer, N. B., Sevens, C. W., Wong, P., Feghali, C. A,, and Wright, T. M. (1995). IFN-gamma induces a pgl/Statla-reIated transcription factor with distinct activation and binding properties. 1. Immunol. 155,3472-3480. 362. Reith, W., Steimle, V., and Mach, B. (1995). Molecular defects in the bare lymphocyte syndrome and regulation of MHC Class I1 genes. lmmunol. Today 16, 539-546. 363. Chang, C. H., Pontes, J. D., Peterlin, M., andFlavell, R. A. (1994).Class I1 transactivator (CIITA) is sufficient for the inducible expression of major histocompatibility complex class I1 genes. 1.Exp. Med. 180, 1367-1374. 364. Steimle, V., Siegrist, C. A,, Mottet, A., Lisowska-Grospierre,B., and Mach, B. (1994). Regulation of MHC class I1 expression by interferon gamma mediated by transactivator CIITA. Science 265, 106-109. 365. Carson, W. E., Ross, M. E. Baiocchi, R. A,, Marica, M. J., Bolani, N., Grabstein, K., and Caligiuri, M. A. (1995). Endogenous production of interleukin 15 by activated human monocytes is critical for optimal production of interferon-gamma by natural killer cells in uitro. 1. Clin Inuest. 96, 2578-2582. 366. CardiIlo, F., Voltarelli, J. C., Reed, S. G., and Silva, J. S. (1996) Regulation of Trypanosoma cruzi infection in mice by gamma interferon and interleukin 10: Role of NK cells Infect. Immun. 64, 128-134. 367. Flynn, J. L., Goldstein, M. M., Triebold, K. J.. Sypek, J., Wolf, S., and Bloom, B. R. (1995). IL-12 increases resistance of BALB/c mice to Mycobacterium tuberculosis infection. 1.Immunol. 155, 2515-2524. 368. Zhou, P., Sieve, M. C., Bennett, J., Kwon-Chung, K. J., Tewari, R. P., Gazzinelli, R. T., Sher, A., and Seder, R. A. (1995). IL-12 prevents mortality in mice infected with Histoplusma capsulatum through induction of IFN-gamma.]. lmmunol. 155,785-795. 369. Nakane, A., Okamoto, M., Asano, M., Kohanawa, M., and Minagawa, T. (1995). Endogenous gamma interferon, tumor necrosis factor, and interleukin-6 in Staphylococcus aureus infection in mice. Infect. Immun. 63, 1165-1172. 370. Zhao, Y.-X., and Tarkowski, A. (1995). Impact of interferon-gamma receptor deficiency on experimental Staphylococcus aureus septicemia and arthritis. J. Immunol. 155, 5736-5742. 371. Candolfi, E., Hunter, C. A,, and Remington, J. S. (1995).Roles of gamma interferon and other cytokines in suppression of the spleen cell response to concanavalin A and toxoplasm antigen during acute toxoplasmosis. Infect. Immun. 63, 751-756. 372. Tomioka, H., Sato, K., Maw, W. W., and Saito, H. (1995). The role of tumor necrosis factor, interferon-y transforming growth factor p. and nitric oxide in the expression of immunosuppressive functions of splenic macrophages induced by Mycobacterium aoium complex infection. J . Leukoe. Biol. 58, 704-712. 373. Ferber, I. A. Brocke, S., Taylor-Edwards, C., Ridgway, W. Dinisco, C., Steinman, L., Dalton, D., and Fathman, C. G. (1996). Mice with a disrupted IFN-gamma gene are susceptible to the induction of experimental autoimmune encephalolmyclitis (EAE). 1.lmmunol. 156, 5-7. 374. Boissier, M.-C., Chiocchia, G., Bessis, N., Hajnd, J., Garotta, G., Nicoletti, F., and Fournier, C. (1995). Biphasic effect of interferon-gamma in murine collagen-induced arthritis. Eur. J. Immunol. 25, 1184-1190.

This article was accepted for publication on 19 October 1995.

ADVAYCES IN IMMUNOLOGY VOL 62

Role of the CD28-B7 Costimulatory Pathways in T Cell-Dependent B Cell Responses KAREN S. HATHCOCK' AND RICHARD J. HODES',t 'Experimental Immunology Bmnch, National Cancer Institute and tNationa1 Instibte on Aging, National Institutes of Heahh, Berheda, Maryland 20892

1. Introduction

The cellular interactions neccessary for T cell-dependent (Td) B cell activation and generation of a humoral immune response involve the interaction of a complex array of cell surface molecules and soluble mediators. As a result of these interactions, biochemical signals are delivered to both the antigen-specific Th cell and to the antigen-specific B cell leading to the subsequent proliferation and differentiation of both cells. The Td B cell immune response requires antigen recognition by both Th cells and B cells, but T cell receptor (TcR) and B cell receptor (BcR) occupancy, although necessary, is insufficient to generate all of the signals required for the induction of a primary antibody response. Additional signals, costimulatory signals, are also required. Costimulatory signals are neither antigen-specific nor MHC restricted and arise from the coligation of counterreceptors on the cell surface of T cells and B cells or professional antigen-presenting cells (APCs). Costimulatory signals are discrete from but complementary to TcR- or BcR-derived signals. In certain situations, T cells or B cells that encounter their antigens in the absence of costimulation may be anergized or deleted. II. Two-Signal Model of T Cell Activation

The two-signal model of lymphocyte activation was initially proposed by Bretscher and Cohn (1970). This model proposed that activation of lymphocytes requires delivery of two distinct signals, neither one of which is sufficient to activate fully. In its current form, the two-signal model specifies that antigen-specificinduction of proliferation and full differentiation of T or B lymphocytes requires a first signal delivered through the TcR or BcR by antigen and a second, or costimulatory signal, delivered through receptors that are distinct from the antigen-specific TcR and BcR. This model predicts that at least three different outcomes can result from the interaction of a T cell with an APC: TcR engagement by specific antigen in the presence of costimulatory ligands on the APC leads to 131

132

KAHEN S . HATHCOCK AND RICHARD J. HODES

T cell activation, proliferation, and the differentiation of effector functions. In contrast, T cell encounter with cognate antigen in the absence of costimulatory ligands results in unresponsiveness (anergy) to subsequent challenge with antigen or leads to T cell death. Finally, interaction of the T cell with costimulatory ligands in the absence of peptide-MHC complexes is predicted to be a neutral event, resulting in neither activation nor anergy. Thus, costimulatory signals are predicted to play a pivotal role in determining the fate of a T cell when it encounters antigen. Experimental observations from many laboratories have supported the legitimacy of the two-signal model of lymphocyte activation and have identified a number of receptor-ligand interactions that can provide costimulatory second signals for lymphocyte activation. The array of reported costimulatory pairs includes CD2 (LFA-2)-CD58 (LFA-3), CD1ldCD18 (LFA-l)-CD54 (ICAM-l), CD4-MHC class 11, CD40L(gp39)-CD40, and CD28 or CTLA-4-CD80 (B7-1) or CD86 (B7-2). Among these, costimulatory signals delivered through CD28 when it encounters its ligands, B7-1 and B7-2, have been shown to be of considerable importance. The CD28-B7 costirnulatory pathway has been shown to regulate TcRmediated T cell proliferation and cytokine production in vitro as well as the generation of in vivo Td immune responses, including graft rejection, antitumor responses, and responses to infectious agents and autoantigens (Linsley and Ledbetter, 1993; Guinan et al., 1994; June et al., 1994; Harlan et al., 1995). However, only limited information exists as to the contribution of the CD28-B7 costimulatory pathway to Td activation of B cells. The CD28-B7 costimulatory pathway is more complex than was initially appreciated, involving the potential interactions of at least two receptors expressed by T cells, CD28 and its homologue CTLA-4, with at least two ligands or counterreceptors, CD80 (B7-1) and/or CD86 (B7-2), on B cells and other APCs. This review will briefly summarize our understanding of the contribution of CD28-B7 costimulation to immune responses and will present in more detail our emerging but still incomplete understanding of the role that this costimulatory pathway may play in Td B cell activation. 111. CD28/CTLA-4 Receptor Family

CD28 and CTLA-4 are members of the immunoglobulin (Ig) gene superfamily and are structurally homologous molecules encoded by genes with similar introdexon structure. Both molecules are composed of a single Ig V-like extracellular domain, a single transmembrane domain, and a single intracellular domain containing a consensus sequence that predicts

133

ROLE OF CD28-B'i COSTIMULATORY PATHWAYS

binding by phosphoinositide (PI)-3 kinase (Harper et al., 1991). Despite considerable structural homology between these molecules, overall the genes encoding CD28 and CTLA-4 show only limited sequence homology to one another. Sequence similarity between CD28 and CTLA-4 is greatest in the proximal extracellular domain and in the intracytoplasmic domain, On the other hand, mouse and human CD28 are highly conserved with an overall sequence homology of 70%, and mouse and human CTLA-4 are also highly conserved with an overall sequence homology of 74% and a 100% amino acid sequence identity in the intracytoplasmic domain (Table 1). Taken together, these results suggest that (i) signals generated by CD28 and CTLA-4, each of which exhibits a high degree of species conservation, particularly in the cytoplasmic domains, may be functionally important; and that (ii) the signals delivered by ligation of CD28 or CTLA-4, which share only limited sequence homology, may be different. CD28 was originally identified as a 44-kDa homodimeric protein expressed on 80% of human peripheral blood T cells. In human peripheral blood, approximateIy 95% of CD4' T cells and 50% of CD8' T cells express CD28 (Damle et al., 1983). In contrast, essentially all murine T cells isolated from spleen, lymph node, and blood express CD28, although cell surface expression on CD4' T cells is slightly higher (Gross et al., 1992).Additionally, extensive studies from many laboratories have demonstrated that although the majority of resting human and murine CD4' T cells express CD28, surface density of expression can be increased by both TABLE I AMINOACID CONSERVATIOU BETWEEN MEMBERSOF THE CD28 AND B7 RECEPTORFAMILIES' Conservation (% amino acid identity) Receptor hCD28 mCD28 hCTLA-4 mCTLA-4 hCD28 hB7-2 mB7-2 hB7-1 mB7-1 hB7-2

Overall

Signal

IgV

69

72

74 31

IgC

Tm

Cytoplasmic

66

67

80

65 16

67 30

83 42

100 34

50

56

66

59

17

10

45 25

24 17

46 24

59 34

20 23

23 6

Note Abbrevlations h, human, in, mouse

' Adapted from June et a1 (1994, Table 1. p 323)

134

KAREN S. HATHCOCK AND RICHARD J. HODES

antigen-specific stimuli and mitogens such as concanavalin A. Originally, CD28 was regarded as a T cell lineage-specific marker; however, studies have detected CD28 on plasma cells and natural killer (NK) cells (Kozbor et al., 1987; Nandi et al., 1994). Signaling consequences of CD28 engagement have been reviewed and have suggested that more than one signal-transduction pathway may be coupled to CD28, depending on the amount of receptor ligation and the activation state of the T cell (Ledbetter et al., 1990; Linsley and Ledbetter, 1993; June et al., 1994). When T cells have been activated by PMA or TcR crosslinking, ligation of CD28 with mAb or transfected cell lines expressing B7-1 can lead to increased tyrosine phosphorylation of multiple intracellular substrates and activation of PI-3 kinase (Lu et al., 1992; Vandenberghe et al., 1992; August et al., 1994; Hutchcroft and Bierer, 1994). Several studies have identified an activation-induced association between CD28 and PI-3 kinase and have reported that mutation of the PI-3 kinase binding motif found in the cytoplasmic domain of CD28 destroys CD28 costimulatory activity (August and Dupont, 1993; Prasad et al., 1994;Truitt et al., 1994; Stein et al., 1994; Pages et al., 1994). It has been proposed that binding of PI-3 kinase to CD28 activates essential second messenger pathways leading to the accumulation of phosphoinositides, phosphorylation of phospholipase Cyl, and increased cytoplasmic-free calcium that are observed in T cells stimulated via CD28 costimulation (Weiss et al., 1986; Ledbetter et al., 1986, 1990; Ledbetter and Linsley, 1992; Nunes et al., 1991, 1993; Ward et al., 1993). The gene encoding CTLA-4 was originally isolated from a murine cDNA library during a search for genes that were specifically expressed by cytotoxic T lymphocytes (CTLs) (Brunet et al., 1987). Antibodies specific for human and murine CTLA-4 have been isolated, and experiments employing these antibodies have demonstrated that although the majority of resting peripheral T cells constitutively express CD28, CTLA-4 surface expression is not observed on resting T cells. Surface expression of CTLA4 on T cells was readily detected after 24-48 hr of in vitro stimulation with anti-CD3 mAb and returned to resting levels by 72-96 hr of culture. CTLA-4 and CD28 are coexpressed on activated T cells, with activated CD8' T cells expressing a higher surface density of CTLA-4 than activated CD4+ T cells (Linsley et al., 1992a; Damle et al., 1994; Walunas et al., 1994; Krummel and Allison, 1995). Both monomeric and disulfide-bonded homodimeric forms of CTLA-4 have been identified on the surface of activated T cells, but to date there is no evidence for the existence of disulfide-linked heterodimers of CD28 and CTLA-4 (Lindsten et al., 1993; Linsley et al., 1992a;Walunas et al., 1994). CTLA-4 mRNA has been found in both Thl- and Th2-like clones, but there has as yet been no analysis of

ROLE OF CD2S-B7 COSTIMULATORY PATHWAYS

135

CTLA-4 surface expression on such populations (Brunet et al., 1987; Freeman et d., 1992). In the case of activated human T cells and human T cell clones, CTLA-4 mRNA expression is generally restricted to the CD28+ subset of cells (Freeman et al., 1992; Linsley et al., 1992a; Lindsten et al., 1993). CD28, however, is not absolutely required for the expression of CTLA-4 because activated T cells derived from mice in which the CD28 gene has been inactivated by homologous recombination can express CTLA-4 (Walunas et al., 1994). Kuiper et nl. (1995) demonstrated that stimulation of human B cells with activated T cell membranes can induce transient surface expression of CTLA-4 on B cells. The biochemical signals generated through CTLA-4 are as yet unknown; however, it is of interest that the intracellular domain of CTLA-4 contains the same PI-3 kinase-binding motif found in CD28 (Harper et al., 1991). Schneider et al. (1995) reported coprecipitation of PI-3 lanase with CTLA4; however, Stein et al. (1994) failed to detect PI-3 kinase binding to a chimeric protein containing the cytoplasmic domain of CTLA-4. These results raise the possibility that CD28 and CTLA-4 may compete for PI3 kinase during T cell activation. IV. B7-1/B7-2 Ligand Family

Like their receptors, B7-1 and B7-2 are both members of the Ig gene superfamily. The genes encoding murine and human B7-1 and B7-2 have been identified (Freeman et al., 1989, 1991, 1993b, 1993c; Azuma et al., 1993b; Chen et ul., 1994b) and encode structurally homologous proteins that are composed of an extracellular region containing both an Ig V-like and a C2-like domain, a single transmembrane domain, and a single short intracellular domain. These molecules possess only limited sequence homology, however, with the extracellular region showing the greatest homology and the intracellular domain being the most divergent. On the basis of their nucleotide sequences, both B7-1 and B7-2 are predicted to be glycoproteins; both contain multiple sites for N-linked glycosylation in their extracellular domains. B7-2 immunoprecipitated from LPSstimulated murine B cells and from a human NK cell line has been reported to be a glycoprotein containing N-linked carbohydrates (Hathcock et al., 1993; Azuma et al., 1993b). The cytoplasmic tail of B7-2 is predicted to be longer than that of B7-1 and contains three potential sites of phosphorylation by protein kinase C that are absent from B7-1. The lack of extensive species conservation between B7-1 and B7-2, especially in the cytoplasmic domains, suggests that these molecules may transduce different biochemical signals or that these molecules may not be signaling molecules but rather may predominantly function through their role as ligands. The

136

KAREN S. HATHCOCK AND RICHARD J. HODES

existence of a third ligand for CTLA-4, B7-3, has also been described, but the gene for this ligand has not been cloned (Boussiotis et al., 1993). B7-1 was originally described as a cell surface antigen expressed on selected B cell lymphoblast lines and activated human B cells (Yokochi et al., 1982; Freedman et al., 1987). Subsequent studies have demonstrated that, although B7-1 is not expressed on unstimulated B cells, its expression on B cells can be induced by a variety of stimuli including mitogens (RaziWolf et al., 1992; Hathcock et al., 1993, 1994; Lenschow et al., 1993, 1994) or crosslinking of MHC I1 (Nabavi et al., 1992), surface Ig (Hathcock et al., 1994; Lenschow et al., 1994), or CD40 (Ranheim and Kipps, 1993; Roy et al., 1995). Expression of B7-1 on B cells can be augmented by cytokines including IL-2 or IL-4 (Valle et al., 1991; Stack et al., 1994). B7-1 expression is not restricted to B cells; monocytes (Freedman et al., 1991), macrophages (Razi-Wolf et al., 1992; Hathcock et al., 1994), and dendritic cells (Larsen et al., 1992; Inaba et al., 1994) have also been shown to express low levels of B7-1, and expression of €37-1can be regulated on these cells by activation, both positively and negatively. For example, it has been shown that IFN-y treatment of monocytes induces increased expression of B7-1, whereas IFN-y treatment of thioglycollate-induced peritoneal macrophages actually suppresses the expression of B7-1 (Freedman et al., 1991; Hathcock et al., 1994). Although B7-1 expression is not detected on freshly explanted T cells, repeatedly stimulated T cell lines and clones as well as HTLV-l-infected T cells express B7-1 (Valle et al., 1990; Haffar et al., 1993; Sansom and Hall, 1993; Azuma et al., 1993a; Das et al., 1995). B7-2 is expressed at low density on freshly explanted B cells, T cells, dendritic cells, macrophages, and monocytes, and its expression is dramatically and rapidly increased following activation of each of these cell types. Increased surface expression of B7-2 has been reported on B cells stimulated with a variety of stimuli including crosslinking of surface Ig with antibodies (Hathcock et al., 1994; Lenschow et al., 1994) or antigen (Lenschow et d.,1994), CD40 ligation (Azuma et d.,1993b; Roy et d., 1995), or mitogen stimulation (Hathcock et al., 1993, 1994; Lenschow et al., 1993, 1994). Cytokines, including IL-5, IL-4, IFN-y, and, to a lesser extent, IL2 (Hathcock et al., 1994; Stack et al., 1994), have also proven to be potent inducers of B7-2 expression on B cells. IFN-y treatment of dendritic cells and thioglycollate-induced peritoneal macrophages has been reported to affect differentially the expression of B7-1 and B7-2. That is, IFN-y treatment increased the expression of B7-2 on both APC populations, whereas it did not increase B7-1 expression on dendritic cells and actually inhibited the expression of B7-1 on peritoneal macrophages (Hathcock et al., 1994;

ROLE OF CD28-87 COSTIMULATORY PATHWAYS

137

Larsen et al., 1994). Taken together, these experiments suggest that the contributions of B7-1 and B7-2 to CD28-mediated costimulation may be regulated in more complex ways than was previously appreciated. Like B7-1, the expression of B7-2 is not restricted to B cells and other professional APC; rather, B7-2 expression is also detected on NK cell clones as well as on activated T cells (Azuma et al., 199313; Hathcock et al., 1994) and T cell clones (Das et al., 1995). Both CD4+ and CD8+ murine T cells express low levels of B7-2, and it has been reported that expression of B72 but not B7-1 is rapidly increased after TcR stimulation (Hathcock et al., 1994).Unlike freshly explanted murine splenic T cells, it was reported that human peripheral blood T cells do not express B7-2, although expression of both B7-1 and B7-2 was induced after 10 days of stimulation in vitro with immobilized anti-CD3 mAb and IL-2 (Azuma et al., 1993b). In response to stimuli, including LPS, anti-IgD-dextran, or IL-4, increased surface expression of B7-2 is detected on B cells earlier and at higher levels than B7-1 (Lenschow et al., 1993; Hathcock et al., 1994; Stack et al., 1994). B7-2 is also expressed at higher levels than B7-1 on murine dendritic cells (Inaba et al., 1994). The binding of CTLA4Ig to activated B cells and dendritic cells is almost completely inhibited by antiB7-2 mAb, whereas little or no inhibition is observed with B7-1-specific mAb, suggesting that B7-2 is the predominant costimulatory ligand expressed by these populations. Based on temporal differences in the expression of B7-1 and B7-2 on B cells it was proposed that B7-2, which is expressed early after B cell or dendritic cell activation, might preferentially bind to constitutively expressed CD28 on T cells, whereas B7-1 might preferentially bind to activation-dependent CTLA-4 (June et al., 1994; Freeman et al., 1993b,c).However, Linsley et al. (1994) found no evidence for selective pairing, but rather observed that both B7-1 and B7-2 bound to CTLA-4 with higher avidity than to CD28, although B7-1 showed slower disassociation kinetics than B7-2. A number of studies have also examined in situ expression of B7-1 and B7-2 on human or murine tissue sections by immunohistochemical techniques. B7-1 expression was reported on interdigitating dendritic cells present in both normal human spleen and “chronically inflamed’ lymph nodes; B7-2 expression was not analyzed in this study (Vandenberghe et al., 1993). In sections of nonimmunized murine spleen and lymph node, strong B7-2 staining was observed in regions rich in macrophages and dendritic cells, and low B7-2 staining was also seen in B cell follicles; however, no detectable staining was revealed by two B7-1 mAbs (Inaba et al., 1994). Whether this reflects a species difference in the expression of B7-1 or a difference in the mAb used remains to be resolved. These

138

KAREN S. HATHCOCK A N D RICHARD J , HODES

results demonstrate that B7 molecules are broadly expressed on a variety of cell populations with APC function. The signaling consequences of B7-1 and B7-2 engagement have not yet been identified. In a number of other receptor-ligand systems, it has been demonstrated that interaction between counterreceptors on two cells can deliver bidirectional signals to both interacting cells. It remains to be determined whether bidirectional signaling occurs in CD28-B7-mediated interactions between T cells and APCs such as B cells. The observation that the nucleotide sequences of the intracellular domains of B7-1 and B7-2 are highly divergent suggests that these molecules may transduce distinct signals. As has been observed for CD28 signaling in T cells, the amount of ligation and the activation state of the B cell or APC expressing these molecules may result in different biochemical signals as well. V. Contribution of CD28 and CTLA-4 to Costimulation

CD28-mediated costimulation has been shown to facilitate a variety of in vitro T cell responses (Linsley and Ledbetter, 1993; Guinan et al., 1994; June et al., 1994; Harlan et al., 1995). Intact bivalent anti-CD28 mAb can augment T cell activation by a variety of stimuli, including PMA, mitogens, or anti-TcR mAb, resulting in enhanced proliferation and cytokine secretion (Ledbetter et al., 1985; Weiss et al., 1986; Thompson et al., 1989; Jenkins et aE., 1991; Hardinget aE., 1992).In contrast, CTLA4Ig or monovalent F(ab)fragments of anti-CD28 mAb profoundly inhibit T cell activation by Ag-bearing APC (Linsley et al., 1991b; Harding et al., 1992; Tan et al., 1993). CTLA4Ig is a chimeric fusion protein composed of the extracellular domain of CTLA-4 and the Fcy portion of IgG and binds to both B7-1 and B7-2 with a higher affinity than does CD28 (Linsley et al., 1991b). CTLA4Ig inhibits APC-dependent T cell activation, apparently by competing for costimulatory ligands, thus preventing CD28 ligation and costimulation. Together, these results suggest that (i) the interaction of CD28 with its ligands on the APC provides a predominant costimulatory pathway utilized by T cells in vitro, and (ii) the costimulatory signals generated through CD28 are at least in part dependent on crosslinking of CD28 molecules. In certain situations, antigen activation of T cells in the absence of CD28 signals can result in tolerance or anergy (Schwartz, 1992). For example, stimulation of Thl clones by fixed antigen-pulsed APCs or by normal antigen-pulsed APCs in the presence of F(ab) fragments of CD28-specific mAb inhibits proliferation and IL-2 secretion and results in anergy. The induction of anergy by fixed Ag-bearing APCs can be prevented by the addition of IL-2 or CD28 costimulation (Jenkins and Schwartz, 1987;

ROLE OF CD28-Bi COSTIMULATORY PATHWAYS

139

Jenkins et al., 1988; Harding et al., 1992). Tan et al., (1993) reported similar results for normal human T cells activated by allogeneic cells in the absence of CD28 costimulation. CD28 costimulation has been shown to profoundly affect the production of multiple cytokines by activated T cells (reviewed by Harlan et aE., (1995). CD28 signals can synergize with TcR signals to augment the production of Thl cytokines including IL-2, IFN-y, GM-CSF, and TNF-a. CD28 costimulation appears to regulate cytokine production by both transcriptional and post-transcriptional mechanisms (June et al., 1990).CD28 costimulation has also been reported to augment the production of Th2 cytokines, including IL-la, IL-4, IL-5, and IL-6, by mechanism(s) that are as yet unclear. CD28 signaling not only influences the production of cytokines but can also affect the responsiveness of T cells to cytokines. For example, CD28 costimulation has been reported to enhance the responsiveness of Th2 T cell clones to IL-4 (McArthur and Raulet, 1993). Cytokines secreted by CD4+ cells may in turn control the expansion and differentiation of antigen-specific T cells into Thl- or Th2-like cells. For example, although IL-2 is necessary for the generation of both Thl and Th2 T cells, cytokines such as IL-4 and IFN-y are instrumental in selectivelydirecting the differentiation of Th2 and T h l T cells, respectively (Seder and Paul, 1994). In addition, many of the effector and regulatory functions of Th cells are mediated by the cytohnes that they produce. Thus, cytokines controlled by CD28 signaling may regulate immune responses by acting on T cells as well as on other cells participating in the response. CTLA-4 is a second receptor expressed by T cells that binds to both B7-1 and B7-2. In contrast to CD28, CTLA-4 is not expressed on resting T cells; its expression can be induced by TcR ligation and augmented by CD28 costimulation. Currently, the contribution of CTLA-4 to costimulation is less well defined than that of CD28. As discussed previously, a soluble fusion protein of CTLA-4, CTLA4Ig, has been widely reported to modulate costimulation-dependent in vitro and in vivo immune responses. Because CTLA4Ig has a higher avidity for both B7-1 and B7-2 than does CD28Ig, soluble CTLA4Ig may competitively inhibit CD28 signaling by competing for shared ligands. The physiologic role of cell surface CTLA4 could similarly be mediated, at least in part, by such competion with CD28 for the binding of common ligands. Alternatively, CTLA-4 may deliver its own signal, inhibitory or costimulatory, to T cells. A number of studies have utilized CTLA-4-specific mAb to examine the role of CTLA4 in T cell activation. These studies have yielded apparently conflicting results. It was initially reported that the costimulatory effects of CTLA-4 and CD28 were additive or synergistic (Linsley et al., 1992a; Damle et al., 1994); however, subsequent evidence has suggested that this may not be

140

KAREN S. HATHCOCK AND RICHARD J. HODES

the case. For example, although CTLA-4 expression can be induced in CD28-deficient mice, there is no B7-dependent proliferation observed for T cells from these mice, suggesting that CTLA-4 itself cannot replace CD28 costimulatory signals (Green et al., 1994; Walunas et al., 1994). Consistent with this observation, Walunas et al. (1994) demonstrated that the addition of either intact molecules or F(ab) fragments of anti-CTLA4 mAb to an allogeneic mixed lymphocyte reaction enhanced T cell proliferation. In contrast, when T cells were suboptimally stimulated with antiCD3 mAb and optimally costimulated with intact CD28 mAb, intact but not F(ab) fragments of CTLA-4 mAb inhibited T cell proliferation. An interpretation consistent with these results is that signaling through CTLA4 negatively regulates T cell activation and that Fab fragments of CTLA4 mAb or the intact CTLA-4 mAb, under conditions of limiting FcR crosslinking, augment the response by blocking the inhibitory interaction of CTLA-4 with its ligand. Krummel and Allison (1995) demonstrated that anti-CTLA-4 mAb augmented the proliferation and IL-2 production of highly purified human T cells stimulated with anti-CD3 mAb and antiCD28 mAb; addition of anti-B7-l mAb together with anti-B7-2 mAb had a similar effect, suggesting that the augmenting effect of anti-CTLA-4 was the result of blocking inhibitory B7-CTLA-4 interactions rather than a direct augmenting effect of CTLA-4 signaling. These experiments suggest that CTLA-4 and CD28 may mediate distinct and even opposing effects on T cell activation and that the interplay between costimulatory signals delivered through CD28 and inhibitory signals delivered through CTLA4 may be pivotal in regulating the outcome of an immune response (Robey and Allison, 1995; Linsley, 1995). VI. Contribution of 87-1 and B7-2 to Costimulation

Linsley et al. first demonstrated that B7-1 is a ligand for both CD28 (Linsley et al., 1990) and CTLA-4 (Linsley et al., 1991b). Subsequent studies demonstrated that transfected cell lines expressing high IeveIs of B7-1 could costimulate T cells activated via the TcR and mitogenic or pharmacologic agents (Linsley et al., 1991a; Reiser et al., 1992; Galvin et al., 1992). In vivo experiments have demonstrated that induced expression of B7-1 on cells that are normally B7-1 negative can render those cells susceptible to immune rejection. For example, transfection and expression of €37-1on tumor cells that normally grow progressively and are not rejected in vivo resulted in their rejection (Chen et al., 1992; Townsend and Allison, 1993). Similarly, tolerance to self-antigens could be broken in tripletransgenic mice that coexpressed B7-1 and peptide-H-2D" molecules on

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

141

pancreatic islet cells and a transgene-encoded TcR specific for the peptide-H-2Db complex (Harlan et al., 1994). As mentioned previously, CTLA4Ig profoundly inhibited APCdependent T cell activation in uitro, and it was generally presumed, although in many cases not directly demonstrated, that the inhibitory effects of CTLA4Ig treatment resulted from inhibiting interactions between CD28 and what was initially its only known ligand, B7-1. In fact, although B7-1 can costimulate T cell activation when expressed at high levels on transfected cell lines, the direct identification of a costimulatory role for B7-1 was more elusive when normal APCs were used. Anti-B7-1 mAb was reported to inhibit in vitro T cell proliferative responses to allogeneic B cell lines (Koulova et al., 1991; Van Goo1 et al., 1994)or to concanavalin A presented by activated peritoneal exudate macrophages (Razi-Wolfet al., 1992). However, in many situations in which CTLA4Ig was inhibitory, anti-B7-1 mAb were ineffective at inhibiting costimulatory signals provided by normal murine or human APC. The observation that CTLA4Ig could modulate immune responses in situations in which anti-B7-l mAb were ineffective raised the possibility that additional (non-B7-1) CTLA-4-binding costimulatoiy ligands might exist. In fact, when mice were generated that were unable to express B71 as a result of gene disruption by homologous recombination, these mice had surprisingly normal serum Ig levels and proliferative responses to T cell mitogens, suggesting that adequate costimulatory signals might be provided by non-B7-l ligands. When LPS/dextran-stimulated B cells from these mice were examined, it was found that these activated B cells bound high amounts of CTLA4Ig and mediated CTLA4Ig-inhibitable costimulatory activity (Freeman et al., 1993a). Simultaneous with this study, genes encoding murine (Freeman et al., 1993c; Chen et al., 1994b) and human (Freeman et al., 1993b; Azuma et al., 1993b) B7-2/CD86 molecules were identified, and antibodies specific for these molecules were isolated (Azuma et al., 1993b; Hathcock et al., 1993; Nozawa et al., 1993; Chen et al., 1994a). Subsequent stndies directly demonstrated that B7-2 is a ligand for CD28 as well as for CTLA-4 (Chen et al., 1994b; Linsley et al., 1994). Considerable interest has focused on identifylng the individual contributions of B7-1 and B7-2 to T cell activation. The results of in vitro experiments examining T cell activation mediated by normal APCs that are potentially capable of expressing both B7-1 and B7-2 are consistent with the notion that B7-2 is the predominant costimulatory ligand expressed and utilized by B cells and dendritic cells. For example, addition of antiB7-2 mAb substantially inhibited APC-dependent T cell proliferation and IL-2 production in response to anti-CD3 mAb or alloantigen stimulation. In contrast, B7-l-specific mAb generally failed to inhibit these responses

142

KAREN S . HATHCOCK AND RICHARD J. HODES

but in some instances augmented inhibition when added together with anti-B7-2 mAb (Caux et al., 1994; Chen and Nabavi, 1994; Hathcock et al., 1994; Inaba et al., 1994; Muraille et al., 1995). However, given that these APCs express higher levels of cell surface B7-2 than B7-1, these experiments could not distinguish quantitative from qualitative differences in B7-1 and B7-2 costimulation. One approach to comparing the functional costimulatory capacities of B7-1 and B7-2 for T cell activation has utilized transfected cells expressing high amounts of either B7-1 or B7-2. These studies have generally failed to identify functional differences between these molecules. For example, transfected cell lines expressing either B7-1 or B7-2 were reported to costimulate in vitro T cell proliferation, IL-2 secretion, and CTL generation (Linsley et al., 1991a; Gimmi et al., 1991; Galvin et al., 1992; Freeman et al., 1993b,c; Chen et al., 1994b; Lanier et al., 1995). In addition, Levine et al. (1995) reported that long-term human T cell lines stimulated with anti-CD3 mAb and costimulated with transfected cell lines expressing either B7-1 or B7-2 proliferated equivalently and produced a similar array of Thl and Th2 cytokines. Although Freeman et al. (1995) reported that B7-2 preferentially costimulated IL-4 production, Seder and Paul ( 1994) reported that B7-1 expressing cells could also induce IL-4-producing T cells. In addition, it has been reported that transfection of immunogenic tumor cells with either B7-1 or B7-2 resulted in in vivo tumor rejection and protective immunity (Chen et al., 1992; Townsend and Allison, 1993; Yang et al., 1995). Taken together, these experiments suggest that when APCs that express quantitatively similar densities of B7-1 or B7-2 are compared, there is no clear difference in the ability of these molecules to costimulate T cell proliferation and differentiation. VII. Regulation of in Vivo Td Immune Responses by CD28-B7 Costimulation

Animal studies have demonstrated that limiting CD28 costimulation can profoundly affect in vivo immune respones. For example, treatment of recipient mice with CTLA4Ig following transplantation of human islet cells resulted in long-term xenograft acceptance (Lenschow et al., 1992). Chronic treatment of mice with CTLA4Ig also inhibited rejection of fully vascularized cardiac allografts (Pearson et al., 1994). In rats, chronic treatment with CTLA4Ig prolonged cardiac allograft survival but did not completely prevent graft rejection (Turka et al., 1992). However, long-term acceptance of cardiac allografts was achieved if rats received donor splenocytes on the day of cardiac transplantation followed by a single injection of CTLA4Ig 2 days later (Lin et al., 1993).

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

143

Inhibiting CD28 costimulation has also been reported to affect the progression of autoimmune disease in mice. For example, chronic treatment of mice with CTLA4Ig has been shown to inhibit the development of experimental autoimmune encephalomyelitis (EAE) (Cross et al., 1995) in response to immunization with myelin basic protein. In another study, treatment of NZB/NZW F, mice with CTLA4Ig inhibited autoantibody production and prolonged life (Fink et al., 1994). Initial descriptions of costimulatory influences on T cell responses emphasized the overall quantitative inhibition of response that resulted from the absence or insufficiency of costimulus. However, it has become clear the costimuli can qualitatively alter the character of an immune response by differentially affecting specific response parameters. For example, Sayegh et al. (1995) examined the mechanisms mediating CTLA4Ig-induced, antigen-specifictolerance in a rat renal allograft model. Immunohistochemical analysis of grafts obtained from both control and CTLA4Ig-treated rats found comparable numbers of infiltrating CD4+ T cells in the grafts of both groups; however, IL-2- and IFN-y-producing cells were observed in the grafts of control rats, whereas IL-4- and IFN-y-producing cells predominated in the grafts from CTLA4Ig-treated rats. Furthermore, the grafts from CTLA4Ig-treated rats also showed an increase in staining for IL-4-dependent IgGl antibody. In another study, Corry et al. (1994) examined the effect of CTLA4Ig treatment on the progression of disease in response to Leishmania major infection. The immune response to L. major is a well-characterized one in which the in vivo differentiation of CD4' effector T cells into Thl and Th2 subsets correlates with disease progression. C57BL/6 mice, which are resistant to infection, normally generate T h l effector cells, and BALB/c mice, which are susceptible to infection, predominantly generate Th2 effector cells. Treatment of BALB/c mice with a single injection of CTLA4Ig abrogated progressive disease in these mice, whereas CTLA4Ig treatment had no effect on the ability of C57BL/6 mice to contain infection. The ability of CTLA4Igtreated BALB/c mice to contain infection correlated with a decrease in IL-4 mRNA levels and the inhibition of IgE production. These studies support the critical importance of CD28 and/or CTLA-4 in responses to diverse antigenic challenge in vivo and suggest that inhibition of the CD28 costimulatory pathway may not uniformly inhibit all aspects of immune responses but rather may alter the response generated, for example, by changmg the profile of expressed cytokines. Experiments have suggested that differential inhibition of B7-1 and B7-2 can have distinct effects on Td in vivo immune responses. Kuchroo et al. (1995) studied the effect of treating mice with either anti-B7-l or anti-B7-2 mAb on the induction of EAE in response to injection of proteo-

144

KAREN S . HATHCOCK AND RICHARD J. HODES

lipid protein peptide. Treatment of mice with anti-B7-l mAb resulted in decreased disease severity and in the generation of T cells with a Th2 phenotype. Furthermore, injection of these cloned Th2 cells inhibited the induction of EAE and abrogated ongoing disease. In contrast, treatment of mice with anti-B7-2 resulted in increased disease severity and the generation of T cells with a T h l phenotype. Lenshow et al. (1995)examined the effects of B7-1 or B7-2 mAb treatment on the generation of spontaneous autoimmune diabetes in NOD mice. Anti-B7-l mAb treatment resulted in increased severity and incidence of diabetes, whereas treatment with anti-B7-2 mAb prevented disease onset. Thus, these studies suggest that blockade of B7-1 or B7-2 may affect the outcome of an immune response by differentially altering variables such as the differentiation of Thl and Th2 cells. The mechanism by which B7-1 and B7-2 might mediate different effects on immune responses is not clear. Distinct roles for B7-1 and B7-2 might result from differences in the timing and/or quantitative expression of these molecules on distinct APC populations encountered during an immune response (Bluestone, 1995; Thompson, 1995). Alternatively, B7-1 and B7-2 might also differentially regulate immune respones by transducing different signals to the APCs on ligation by CD28 or CTLA-4. VIII. Cellular Events in Td B Cell Responses

The ability of foreign antigen to elicit an immune respone in vivo depends on triggering a number of complex cellular interactions and signaling pathways in appropriate sequence and in discrete microenvirunments. In the case of humoral immune responses, antigen-specific B cells must be activated to proliferate and to differentiate into antibody-secreting plasma cells or into memory B cells (Fig. 1). Antigen-specific B cell activation is initiated by the binding of antigen to specific surface Ig receptors: however, although BcR signaling is necessary for Ag-specific B cell activation, BcR signals alone are often not sufficient to elicit full B cell activation. B cell activation can be further regulated by the interaction of the B cell with a variety of cell types including antigen-specific CD4' helper T (Th) cells (for Td antigens), macrophages, and specialized populations of dendritic cells. Activation and differentiation of T cells into competent Th cell populations generally requires two signals: one delivered through the TcR, and a second, costimulatory signal that can be provided by CD28. In turn, helper function of these T cells is mediated by both surface molecules and secreted cytokines. Signals delivered by the cognate interaction of antigenspecific T cells and B cells are bidirectional and mediated through multiple molecular interactions, including those between the TcR, CD28, and CD40 ligand (CD40L) on the T cell and peptide-MHC class 11, B7-1 and

HOLE OF CD28-B7 COSTIMULATORY PATHWAYS

145

Td antigen activation of B cells

1

Proliferation Development into

~

plasmacel's

Ig isotype switch

CD40L TcR

1

1 1

Proliferation Hypermutation IgM-lgG isotype classswitch Development of

1

CD40 peptide-MHC class II

BcR BcR-antigen complex FIG.1. B cell differentiation in a primary murine Td response.

B7-2, and CD40 on the B cell (Clark and Ledbetter, 1994; Kelsoe, 1994).Cytokines secreted by T cells not only regulate the expansion and differentiation of Th cells but also provide essential signals for the expansion, survival, and differentiation of antigen-activated B cells. The expression of B7 costimulatory ligands, as well as the functional role of costimulatory signals, can be assessed in the context of a model proposed by Jacob et al. (1992) to describe the cellular events involved in primary antibody responses and induction of B cell memory to Td antigens. In what are believed to be early events in the primary in vivo response to Td antigens, antigen-specific CD4+ T cells in the periarteriolar sheaths (PALS) of the spleen are activated to proliferate and to secrete IL-2 (followed by other cytokines) when they encounter peptide presented in the context of MHC class I1 molecules on cells such as dendritic cells, activated B cells, or other APCs that possess costimulatory function. As a result of both TcR- and CD28-costimulatory signals, T cells undergo a complex process of expansion and differentiation that can ultimately lead to functional Th cells expressing new or altered cell surface molecules, such as CD40L (gp39), and secreting multiple cytokines with diverse effector functions (Van den Eertwegh et al., 1994). Multiple factors have

146

KAREN S. HATHCOCK AND RICHARD J. HODES

been shown to influence this differentiation process, including antigen (type, route of immunization, and dose), adjuvants, costimulatory ligands expressed by APCs, and cytokines themselves. Costimulatory signals delivered through CD28 can direct Th cell function by regulating both the amount and the types of cytokines produced by T cells. Recent findings suggest that IL-4 is necessary to generate Th2 cells, which secrete IL-4, IL-5, and IL-10, whereas IFN--y and IL-12 direct the differentiation of Thl cells, which secrete predominantly IL-2, IFN-7, and TNF-/3. The pattern of cytokines produced by these Th cell populations in turn regulates the survival and expansion of antigen-reactive B cells and the profile of isotypes produced by antibody-secreting B cells. In mice, for example, IL-4 stimulates IgGl and IgE production and inhibits production of IgG2a. IFN-y, on the other hand, promotes the production of IgG2a and inhibits the production of IgGl and IgE (Snapper and Finkelman, 1993). It should be emphasized again, however, that survival and differentiation of B cells requires not only Th cell-derived cytokines but also requires cognate interactions between B cells and T cells mediated by receptorligand interactions including that between CD40 on B cells and CD40L on Th cells. Crosslinking of CD40 on B cells with either antibody or recombinant CD40L can induce a variety of effects including B cell proliferation, and, in combination with certain cytokines, Ig secretion and isotype switching (Snow and Noelle, 1993; Banchereau et al., 1994). Although considerable experimental evidence has demonstrated that CD28 can regulate the cytokine-dependent aspects of Th cell effector function, much less is known about the effects of costimulation on the cognate effector functions of T cells mediated by cell surface molecules such as CD40L. B cell involvement in Td antibody responses also appears to be initiated in the PALS region of the spleen when antigen binds to specific Ig receptors (Jacob et al., 1991b; Jacob and Kelsoe, 1992; Liu et al., 1991); the antigen is internalized and processed, and peptide fragments are presented in association with MHC class I1 molecules to antigen-specific Th cells. Antigen binding to the B cell can induce the expression of costimulatory ligands that are necessary for effective antigen presentation to Th cells. In this phase of the response, the relative contributions of B cells or other professional APCs to the initial activation of T cells is unclear; however, once naive B cells have been activated by antigen and later by T cells, they can express costimulatory ligands and can present antigen to T cells efficiently. Th cells provide signals to B cells, via MHC class 11, CD40, and cytokine receptors, that are necessary for subsequent B cell expansion and differentiation either into antibody-producing plasma cells or into germinal center (GC)-derived memory cells.

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

147

In the first few days after primary immunization, proliferating B cells are initially found in the PALS region of the spleen. Thereafter, these activated B cells appear to enter into either of two differentiation pathways: one leading to development of foci of plasmacytes and plasmablasts adjacent to the PALS, and the other involving migration into lymphoid follicles that subsequently differentiate to form GCs (Jacob and Kelsoe, 1992). Plasmacytes in foci are derived from oligoclonal populations of B cells and are thought to be the predominant source of serum antibodies observed early after a primary immunization. B cells that enter this pathway can undergo Ig class switching, but both IgM and IgG antibodies produced by these cells are generally without somatic mutation and are of relatively low affinity for antigen (Jacob et nl., 1991b; Jacob and Kelsoe, 1992a; McHeyzer-Williams et al., 1993). GCs are morphologically and phenotypically distinct from foci and can be identified on the basis of their ability to bind the lectin, peanut agglutinin (PNA) (MacLennan, 1994). Analysis of V(D)J rearrangements in B cells derived from foci and adjacent GCs has demonstrated that these B cell populations have a common clonal origin (Jacob and Kelsoe, 1992). GCs are first observed 4 to 6 days after primary immunization and reach maximum size during the second week of the response, at a time when PALSassociated foci have already begun to &sappear. Thereafter, GCs begin to decrease in size and number and disappear about 3 or 4 weeks after immunization. As demonstrated by the isolation and sequencing of rearranged Ig heavy-chain genes from individual GC B cells, GCs are derived from oligoclonal populations of antigen-specific B cells. Two morphologically distinct zones, best visualized in rat, human, and rabbit, can be distinguished in the mature GCs: (i) a dark zone proximal to the PALS, containing surface Ig-, rapidly proliferating centroblasts; and (ii) a light zone, containing Ig+, nonproliferating centrocytes ( MacLennan, 1994). In addition to B cells, the light zone also contains follicular dendritic cells (FDC) (Tew et aE., 1990) and antigen-specific T cells (Fuller et al., 1993; Zheng et al., 1994). The dark zone is the proposed site where somatic hypermutation of Ig V region genes occurs. During this process, nucleotide substitutions are introduced into the rearranged Ig V genes and immediate flanking sequences while leaving the C regions largely unmodified (Berek and Zieger, 1993).The accumulation of somatic mutations in GCs has been demonstrated by microdissection and PCR-facilitated sequence analysis (Jacob et al., 1991a), and it is thought that this process occurs uniquely or predominantly within the GC microenvironment. Mutations are first detected at Days 7 or 8 of the response, and the frequency of mutations increases progressively for at least another week. The selection and subsequent differentiation of memory B cells in the GC microenvironment is

148

KAREN S. HATHCOCK AND RICHARD J. HODES

thought to involve both antigen-driven selection by FDCs, and cognate and cytokine signals provided by Th cells. These interactions are thought to occur within the light zone of the GC. B cells that possess higher-affinity Ig receptors for antigen as a result of somatic mutation are thought to be selected on the basis of their improved ability to bind unprocessed antigen-antibody complexes present on the FDCs in the GC light zone. The identity of additional signals delivered by cognate T-B cell interactions occurring in the light zone is incompletely understood, but several studies have suggested that the interaction of CD40L on the T cell with CD40 on the B cell is essential for the survival of GC B cells and maintenance of the GC reaction. Crosslinking of CD40 has been reported to retard apoptosis of GC B cells in vitro and to induce B cell proliferation and isotype switching in the presence of T cell cytokines such as IL-4 (Liu et al., 1991; Arpin et aE., 1995). Treatment of mice with anti-CD40L mAb at the time of antigen immunization not only inhibits antibody production but also inhibits GC formation (Han et al., 1995). IX. Contribution of CD28-B7 Costimulation to Td B Cell Responses

A role for CD28-B7 costimulatory signals in formation and function of GCs was initially suggested by immunohistochemical studies that identified B7-expressing cells in both human and murine GCs. Analysis of normal human spleen sections revealed a subset of GC B cells that stained positively for both B7-1 and B7-2 expression (Nozawa et al., 1993). Han et al. (1995) subsequently examined the kinetics of B7-2 expression in splenic GCs generated in mice immunized with either a Td antigen, (nitrophenyliacetyl-chicken y-globulin (NP-GC), or with a type I1 T-independent (Ti) antigen, Streptococcus pneumoniae vaccine (Pn). In spleens from NP-CG immunized mice, B7-2 staining was observed on a subset of GC B cells that appeared to correspond to the light zone and on lymphocytes in the follicular mantle surrounding GCs. Initial B7-2 expression did not occur simultaneouslywith GC formation (4-6 days postimmunization) but rather was delayed until Day 8 and was temporally correlated with Ig class switching and somatic hypermutation. In contrast to GCs induced by Td antigen, GCs that formed in response to Pn did not stain strongly for B7-2, and only occasional B7-2' follicular lymphocytes were observed late in the response. These studies demonstrated that Td GC cells express costimulatory ligands and raised the possibility that costimulatory signals may function within the GC microenvironment to direct survival and/or differentiation of GC B cells. Results of early in vitro studies suggested that CD28 costimulation can regulate Td antibody production. Lum et al. (1982) demonstrated that

ROLE OF CD28-B7 COSTIMULATORY PATlIWAYS

149

CD28+, but not CD28-, human T cells were enriched in their capacity to provide help to normal human B cells stimulated with pokeweed mitogen, a Td activator of human B cells. Damle et al. (1991) subsequently demonstrated that anti-CD28 mAb inhibited Th cell-induced differentiation of EBV-transformed allogeneic B cells into Ig-secreting cells. Taken together, these results suggested that CD28 costimulation could play a role in regulating the Td differentiation of B cells into antibody-secreting cells. The observation that basal levels of serum Ig in CD28-deficient mice are substantially lower than those of wild-type littermates further strengthened the concept that CD28 costimulation is important for the regulation of antibody production. Analysis of serum Ig isotypes revealed that concentrations of IgM and IgG3 were essentially normal in CD28-deficient mice, whereas concentrations of IgGl and IgG2b were substantially reduced and IgG2a was increased in CD28-deficient animals (Shahinian et al., 1993). When immunized with the Td antigen, NP-CG, antigen-specific IgGl production and GC formation were dramatically reduced in CD28deficient mice (K. Hathcock, unpublished data). These findings suggest that CD28 costimulation influences antibody responses in an isotypespecific manner and, furthermore, that the costimulatory signals mediated by CD28 are crucial for B cell responsiveness to Td antigens. To investigate the role of CD28-B7 interactions in humoral immune responses, a number of studies have assessed the effect of CTLA4Ig as an inhibitor of these interactions. Linsley et al. (199213) examined mice treated with CTLA4Ig at the time of antigen immunization for their subsequent ability to generate primary and secondary antibody responses to particulate and soluble Td antigens. Compared to control-treated mice, treatment of mice with CTLA4Ig inhibited the generation of Td antibody responses to both sheep erythrocytes (SRBC) and keyhole limpet hemocyanin (KLH). Maximal inhibition of the primary antibody response to KLH was observed even when CTLA4Ig treatment was delayed for up to 2 days after immunization. In addition, mice treated with high doses of CTLA4Ig at the time of primary immunization and subsequently rechallenged with the same antigen in the absence of additional CTLA4Ig generated substantially inhibited secondary antibody responses. However, after additional immunizations, antibody production by CTLA4Ig-treated mice was equivalent to that of control mice. Administration of high doses of CTLA4Ig at the time of secondary immunization with KLH also inhibited the secondary antibody response, although to a lesser extent. These results demonstrated that inhibition of CD28-B7 interactions early in the course of a humoral immune response had profound effects on subsequent antibody production and further that the antigen-specific immunosupression observed in these

150

KAREN S. HATHCOCK AND RICHARD J. HODES

mice, although long-lived, was not permanent because it could be largely overcome by multiple exposures to antigen. A similar approach for studying the contribution of CD28 costimulation to humoral immune responses was reported by Lane et al. (1994) who generated soluble CTLA4Ig transgenic mice. In these mice, CTLA4Ig is secreted primarily by B cells and plasma cells, is detected in the serum, and is present throughout the course of the immune response. Compared to nontransgenic littermates, CTLA4Ig transgenic mice generated negligible primary antibody response to Td antigen, and secondary responses were also inhibited, although much less completely. Analysis of antigen-specific antibodies generated after primary immunization revealed that IgG isotypes IgG1, IgG2b, and IgG3 were all substantially inhibited but that IgM and IgA antibody levels were essentially unaffected by the transgene. The defective response of CTLA4Ig transgenic mice was also marked by an absence of GCs in spleen and lymph nodes but not in gut-associated tissues. Furthermore, defective GC formation correlated with a reduction in somatic mutations in hybridomas derived from immunized transgenic mice. It is likely that the inhibition of secondary responses in the transgenic mice is, at least in part, due to reductions in GC-derived memory B cells. Given the observation that CTLA4Ig could inhibit in wivo antibody responses to Td antigens, presumably as a consequence of its ability to bind to both B7-1 and B7-2, subsequent experiments attempted to identify the individual contributions of B7-1 and B7-2 to these responses. Hathcock et al. (1993) reported that treatment of mice with anti-B7-2 mAb at the time of primary immunization with a Td antigen, trinitrophenyl-KLH (TNP-KLH), in adjuvant, substantially inhibited the anti-TNP IgG response with little or no effect on IgM antibodies. When antigen was administered in the absence of adjuvant, anti-B7-2 mAb completely inhibited the IgG antibody response. Mice treated with anti-B7-2 mAb at the time of initial antigen administration also failed to generate a secondary response to subsequent rechallenge with the same antigen. In contrast, administration of anti-B7-1 mAb at the time of antigen immunization had no detectable effect on antibody production (K. Hathcock, unpublished data). These results suggest that B7-2 is the predominant costimulatory ligand involved in generating Td antibody responses in this system. Han et al. (1995) examined the effect of anti-B7-2 mAb treatment on antibody responses and GC formation in mice immunized with a Td antigen, NP-CG in alum. These investigators reported that administration of anti-B7-2 mAb at the time of antigen immunization significantly inhibited both the Td IgG and the IgM anti-NP antibody responses. In addition, this treatment completely inhibited the formation of antigen-specific GCs. These investigators also tested the effect of anti-B7-2 mAb administration

151

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

when it was delayed until Day 6 after immunization. Mice treated late with anti-B7-2 mAb generated NP-specific antibody responses that were not inhibited; however, GCs were reduced in size and number relative to controls. Moreover, as assessed by PCR analysis, this late treatment with anti-B7-2 mAb profoundly inhibited the accumulation of somatic mutations in GC B cells. Compared to control-treated mice, secondary antibody responses in mice that were treated late with anti-B7-2 mAb were inhibited. These results suggest that B7-2 regulates Td B cell responses at least two points during the response. B7-2 appears to function during initial events that are required both for antibody generation and for GC formation, and also controls later events that are not essential for the primary antibody response but are necessary for the development of the GC reaction and generation of memory B cells expressing mutated, higher-affinity Ig (Fig. 2). It is not clear whether the costimulatory signals mediated by B7-2 play a role in directly regulating hypermutation or whether B7-2 functions to promote the survival or clonal expansion of hypermutated B cells. Interestingly, the correlation between B7-2 expression and hypermutation is further supported by immunohistochemical studies that examined B7-2 expression NO TREATMENT

EARLY TREATMENT

PALS-foci

LATE TREATMENT

PALS-foci Germinal Centers

no GC formation IgG inhibited IgM variable effects

a

normal

GC formation impaired mutation inhibited

peptide-MHC class II

Frc. 2. Effect of early and late treatment with anti-B7-2 mAb on Td B cell responses. CD2fSB7-2 interactions regulate B cell responses at at least two points during the response. During early events, B7-2-mediated interactions are required for antibody generation and CC formation. Later in the response, B7-%mediated interactions are required for GC function.

152

KAREN S. HATHCOCK AND RICHARD J. HODES

in GCs generated to type I1 Ti antigens or to Td antigens in aged mice. The GCs generated in aged mice were smaller but otherwise phenotypically similar to GCs of young adult mice. However, the GC B cells in aged animals were deficient in B7-2 expression, and they did not appear to support somatic Ig hypermutation (Han et al., 1995; Miller et al., 1994; Miller and Kelsoe, 1995). The most straightforward interpretation of the inhibitory effects of treating mice with anti-B7-2 mAb is that B7-2 is the predominant costimulatory ligand necessary for the generation of Td B cell responses to these antigens, and that treatment with anti-B7-2 mAb inhibits these responses by preventing the essential CD2tbB7-2 recognition event. Alternatively, blocking Th cell interactions with B7-2 might also inhibit B cell responses by altering the balance of functionally distinct signals provided by different costimulatory ligands, for example, by unmasking a negative regulatory effect of B7-1. Consistent with the possibility that B7-1 can mediate an inhibitory effect on Td antibody production, it was shown that transgenic mice whose peripheral B cells overexpress B7-1 on their surface generate negligible IgM and IgG antibody responses to Td antigens (Sethna et al., 1994). Furthermore, to determine whether this impairment in antibody response was due to B7-1 expression, mice were treated with anti-B7-l mAb at the time of antigen immunization. This treatment restored the ability of these mice to generate TNP-specific IgGl and IgG2a antibodies. Although overexpression of B7-1 on B cells apparently reveals a negative regulatory function of this molecule in Td antibody responses, it is not clear what function(s) B7-1 provides when it is expressed at normal density and under physiologic regulation. The mechanism(s) by which CDSS-dependent costimulatory signals contribute to the generation of Td B cell responses is as yet unclear. One way in which costimulatory signals may influence B cell responses is by affecting the expansion and differentiation of Th cells (reviewed by Thompson, 1995). The observation that primary Td B cell responses are generally dependent on CD28-B7 interactions, whereas Ti B cell responses are not (K. Hathcock, unpublished results; Sethna et al., 1994), is consistent with this hypothesis. Only a limited number of studies have directly examined T cell function in circumstances in which interfering with CD28-B7 costimulation resulted in inhibition of Td B cell responses. The results of these studies have not been consistent. For example, although CTLA4Igtreated mice, CTLA4Ig transgenic mice, and B7-1 transgenic mice all showed reduced Td antibody responses, they differed in the capacity of their T cells to respond in vitro to the immunizing antigen. When T cell proliferative responses to Td-immunizing antigens were examined, CTLA4Ig-treated mice failed to proliferate to the immunizing antigen,

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

153

whereas T cell proliferative responses by CTLA4Ig transgenic mice and B7-1 transgenic mice were unaffected (Linsley et al., 1992b; Ronchese et al., 1994; Sethna et al., 1994). In addition to affecting the expansion of antigen-reactive Th cells, CD28 signaling may also influence Th cell function by regulating cytokine production; however, different results were reported when T cell cytokine production was analyzed either directly after in vivo immunization or after reexposure to Ag in vitro. When cytokine mRNA expression was analyzed at various times after in vivo SRBC immunization, it was reported that CTLA4Ig treatment inhibited the induction of IL-2 mRNA by splenic T cells and further that this treatment also inhibited the induction of IL-4 mRNA by both T and non-T cells (Wallace et al., 1995). Thus, inhibition of Td antibody production observed in this antigen system might be the direct consequence of defective IL-2 and IL4 production. Ronchese et al. (1994) compared cytokine production by T cells from CTLA4Ig transgenic and normal mice after reexposure to antigen in vitro and observed different effects on cytokine production than those reported previously. Compared to control mice, T cells isolated from immunized transgenic mice secreted reduced amounts of IL-4 relative to IFNy but secreted comparable amounts of IL-2 and IL-3 when reexposed to the immunizing antigen in vitro. In addition, adoptive-transfer experiments revealed that T cells isolated from antigen-immunized CTLA4Ig-treated or CTLA4Ig transgenic mice were not anergized or deleted, because they could provide help when transferred with B cells and antigen to irradiated hosts. These results suggest that interfering with CD28-B7 interactions does not completely or universally prevent antigen priming of T cells but rather influences their optimal differentiation into functional cytokinesecreting Th cells. Moreover, these results also suggest that memory Th cell function as assessed by cytokine production may be less dependent on CD28-B7 interactions than is naive T ceU function. Th cell effector function for B cell activation to Td antigens is not mediated only by secreted cytokines but is also affected by cell contactdependent signals such as those delivered through the interaction of CD40L on the Th cell with CD40 on the B cell (Snow and Noelle, 1993; Banchereau et al., 1994). Studies have revealed an interdependence between the expression and function of CD40L and B7 in in vitro Td B cell activation. Signals delivered by crosslinking CD40 on the B cell can induce the expression of both B7-1 and B7-2 and thus may augment the costimulatory signals mediated by the interaction of these molecules with CD28 (Ranheim and Kipps, 1993; Roy et al., 1995). In addition, CD28 signaling can augment CD40L expression and function on TcR-stimulated T cells. For example, when TcR-stimulated T cells were activated in the presence or absence of CD28 costimulation and then compared for their ability to

154

KAREN S. HATHCOCK AND RICHARD

J.

HODES

induce B cell proliferation and/or antibody production in uitro, it was reported that CD28 signaling augmented the induction of Th cell activity and the expression of CD40L. Addition of CD40Ig or blocking anti-CD4O mAb substantially inhibited the CD28-dependent augmentation of B cell proliferation and antibody secretion, suggesting that enhanced Th cell function was, at least in part, mediated through CD40-CD40L interactions (de Boer et al., 1993; Klaus et aE., 1994). In contrast, Roy et al. (1995) examined the expression of CD40L on antigen-activated T cells under APC-dependent conditions and reported that blocking CD28 costimulation with a mixture of anti-B7-1 and anti-B7-2 mAb inhibited T cell proliferation but did not inhibit the induction of CD40L on T cells, implying that induction of CD40L expression is not dependent on CD28 costimulation. Taken together, these experiments suggest that CD28 costimulation can, in some circumstances, regulate Th cell function for in vitro B cell activation via a CD40-CD40L-dependent pathway and that interfering with CD28-B7 interactions in uiuo may inhibit Th cell function by interfering with this pathway. X. A Madel for the Fundion of CD28-67 interactions in Td 6 Cell Responses

A model for the role of CD28-B7 interactions during primary murine Td B cell responses can be constructed on the basis of current understanding of CD28 and B7 function superimposed on a model of B cell activation (Fig. 3) (Jacob et al., 1992; Kelsoe, 1994). The response to foreign antigen is initiated by antigen capture and processing by antigen-specific B cells or other professional APCs, such as dendritic cells (DC),present in the T cell-rich PALS. DCs can constitutively express both B7-1 and B7-2, whereas naive B cells express only low levels of B7-2. During this initial phase of the response, crosslinking of BcR with antigen can rapidly upregulate B7-2 expression on antigen-reactive B cells. In Td responses in the mouse, B7-2 appears to be functionallypredominant, but the relative physioIogic importance of B7-1, B7-2, or other B7 family members in human or other species has not been studied. Crosslinking of both TcR and CD28 on T cells that encounter peptide-MHC class I1 complexes on B7-expressing costimulatory-competent antigen-activated B cells or DCs results in the activation and differentiation of Th cells to secrete cytokines and express CD40L. Cytokines secreted by T cells in response to TcR and CD28 crosslinking may then act (i) in an autocrine fashion to promote T cell proliferation and differentiation into Thl or Th2 cells, or (ii) in a paracrine fashion to promote subsequent B cell activation (proliferation and differentiation) and augmented expression of costimulatory B7 ligands on B cells and DCs.

155

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

?

FIG.3. Model for the role of CD28-B7-2 interactions during a primary Td B cell response.

In cognate B cell responses to Td antigens, B cells must interact directly with T cells, and this interaction is mediated through multiple cell surface molecules on the B cell including MHC class 11, CD40, and B7 family members. Signals mediated by the interaction of B7-2 on the B cell with CD28 on the T cell may augment CD40L expression on the T cell and direct the differentiation of T cells into Th cells. Moreover, Th2 cytokines, such as IL-4 and IL-5, as well as CD40 crosslinking may further potentiate the costimulatory function of B cells by enhancing B7-2 expression. Alternatively, crosslinking of B7-2 may directly signal the B cell. At this phase of the response, B cells that have received appropriate Th cell signals are induced to enter into one of two differentiation pathways: one leading to the development of foci of antibody-secreting cells and the other leading to the formation of GCs. It is not known whether B cells that differentiate into antibody-secreting cells in foci require additional T cell signals within this microenvironment. However, the observation that Ig isotype switching occurs in foci and that

156

KAREN S. HATIICOCK AND RICHARD J . HODES

such isotype switching is generally dependent on Th-derived cytokines suggests that Th cells are playing a role within the foci. Some antigen-activated B and T cells migrate from the PALS into nearby primary follicles and subsequently proliferate and differentiate within the FDC reticulum to form PNA+ GCs. Inhibition of CD28-B7-2 interactions during the initial phase of the response to Td antigens inhibits GC formation and, consequently, Ig hypermutation and generation of memory B cells. In a mature GC, proliferating Ig- centroblasts can be identified in the dark zone in which somatic mutation of Ig V region genes is thought to occur. Although PNA+ GCs can be detected as early as Day 4 after immunization, accumulation of mutations does not appear to begin until Days 7 or 8. Proliferating centroblasts differentiate into nonproliferating Ig+ centrocytes that populate the light zone, where B cells expressing mutated Ig are probably selected on the basis of their ability to bind to unprocessed antigen presented by FDCs as immune complexes. BcR crosslinking allows the survival of some B cells expressing mutated Ig of higher affinity, whereas many B cells die by apoptosis. In turn, antigen presented by FDCs may be processed and presented by B cells to antigenspecific T cells in the GC light zone. At this phase of the response, T-B interactions are thought to be regulated both by cytokines and by cognate signals delivered through the interaction of CD40L-CD40 and CD28-B72. B cells that have received appropriate cognate and cytokine signals from T cells are rescued from apoptosis and may either recirculate into the dark zone where they may undergo further mutation as centroblasts or may differentiate into memory B cells and exit the GCs. CD28-B7-2 interactions within the GC itself play a critical role in GC function. Expression of B7-2 is delayed until approximately Day 8 of the primary response, correlating with the onset of somatic mutation and isotype switching. The interaction of CD28 and B7-2 occurs within the GC itself as evidenced by the fact that blocking GD28-B7-2 interactions in Td GC between Days 6 and 8 inhibits somatic mutation and/or the selection and survival of mutated B cells into the memory compartment. XI. Concluding Remarks

A number of critical questions remain with respect to the role of CD28-B7 interactions in Td B cell responses. It remains unclear what distinct functional roles may be played by the multiple members of the CD28 and B7 families of costimulatory molecules; CD28 and CTLA-4 expressed by T cells appear to have highly similar ligand specificities for B7-1 and B7-2. Studies indicate that crosslinking of CD28 and CTLA-4 may, nonetheless, have markedly distinct functional consequences. Simi-

ROLE OF CDZR-B7 COSTIMULATORY PATHWAYS

157

larly, some findings suggest that B7-1 and B7-2 can mediate similar outcomes, whereas other observations suggest that B7-1 and B7-2 mediate distinct in vivo effects. The basis for such differences, in the B7 molecules themselves or in the cellular and temporal patterns of their expression, is unclear and will require further analysis of complex in vivo responses. The availabilityof in vivo genetic manipulations of expression of these molecules may be informative in these systems. A second area in need of further elucidation is the characterization of the signal transduction events mediated by CD28/CTLA-4-B7 interactions. Considerable information has been obtained describing the signal pathways that are coupled to CD28 and the potential interactions of these signal transduction pathways with TcR-linked signaling events. In contrast, less information is available concerning the signaling consequences and functional role of CTLA-4 engagement. Finally, little or no information currently exists concerning the ability of B7-1 or B7-2 to mediate active signal transduction. Most emphasis has to date been placed on the function of B7 in engaging or crosslinking CD28 on T celIs, and it remains to be determined whether B7-1 and B7-2 function solely as ligands for CD28 and CTLA-4, or whether they also deliver information to B7-expressing cells including B cells. Finally, it should be noted that manipulation of costimulatory events has substantial potential for quantitatively or qualitatively altering complex immune responses including Td B cell responses. Such interventions are likely to be relevant to such diverse circumstances as those involved in suppressing unwanted autoimmune or allergic responses, in augmenting desireable responses to infectious or malignant challenges, or in altering the qualitative nature of immune responses to emphasize desireable effector outcomes. Expanded understanding of the cellular and molecular basis of costimulation will be critical to success in these manipulations.

ACKNOWLEDGMENTS The authors thank Drs. David Allman, Ronald Cress, and Garnett Kelsoe for their critical reviews of the manuscript.

REFERENCES Arpin, C., Dechanet, J., Van Kooten, C., Merville, P., Grouard, G., Briere, F., Banchereau, J., and Liu, Y. J. (1995). Generation of memory B cells and plasma cells in vitro. Science 268, 720-722. August, A,, and Dupont, B. (1993). CD28 of T lymphocytes associateswith phosphatidylinosito1 3-kinase. lnt. lmmunol. 6, 769-774. August, A,, Gibson, S., Kawakami,Y., Mills, G. B., and Dupont, B. (1994).CD28 is associated with and induces the immediate tyrosine phosphorylation and activation of the Tec family

158

KAREN S . HATHCOCK AND RICHARD J. HODES

kinase ITK/EMT in the human Jurkat leukemic T-cell line. Proc. Natl. Acad. Sci. USA 91,9347-9351. Azuma, M., Yssel, H., Phillips, J. H., Spits, H., and Lanier, L. L. (19934. Functional expression of B7/BB1 on activated T lymphocytes.J. Exp. Med. 177, 845-850. Azuma, M., Ito, D., Yagita, H., Ohmura, K., Phillips, J. H., Lanier, L. L., and Somoza, C. (199313). B70 antigen is a second ligand for CTLA-4 and CD28. Nature 366, 76-79. Banchereau, J., Bazan, F., Blanchard, D., Briere, F., Galizzi, J. P., van Kooten, C., Liu, Y. J., Rousset, F., and Saeland, S. (1994). The CD40 antigen and its ligand. Annu. Rev. Immunol. 12, 881-922. Berek, C., and Zieger, M. (1993). The maturation of the immune response. Immunol. Today 14,400-404. Bluestone, J. A. (1995). New perspectives of CD28-B7-mediated T cell costimulation. Immunity 2, 555-559. Boussiotis, V. A,, Freeman, G. J., Gribben, J. G., Daley, J., Gray, G., and Nadler, L. M. (1993). Activated human B lymphocytes express three CTLA4 binding counter-receptors which costimulate T cell activation. Proc. Natl. Acad. Sci. USA 90, 11059-11063. Bretscher, P. A,, and Cohn, M. (1970). A theory of self discrimination. Science 169, 10421049. Brunet, J . F., Denizot, F., Luciani, M. F., Roux-Dosseto, M., Suzan, M., Mattei, M. G., and Golstein, P. (1987).A new member of the immunoglobulin supergene family-CTLA4. Nature 328, 267-270. Cam, C., Vanhervliet, B., Massacrier, C., Azuma, M., Okumura, K., Lanier, L. L., and Banchereau, J. (1994). B70/B7-2 is identical to CD86 and is the major functional ligand for CD28 expressed on human dendritic cells. J Exp. Med. 180, 1841-1847. Chen, C., Faherty, D. A., Gault, A., Connaughton, S. E., Powers, G. D., Godfrey, D. I., and Nabavi, N. (1994a). Monoclonal antibody 2D10 recognizes a novel costimulatory molecule on activated murine B lymphocytes. J. Immunol. 152, 2105-2114. Chen, C., Gault, A., Shen, L., and Nabavi, N. (1994b). Molecular cloning and expression of early T cell costimulatory molecule-1 and its characterization as B7-2. J. Immunol. 152,4929-4936. Chen, C., and Nabavi, N. (1994). In vitro induction of T cell anergy by blocking B7 and early cell costimulatory molecule ETC/B7-2. Immunity 1, 147- 154. Chen, L., Ashe, S., Brady, W. A., Hellstrom, I., Hellstrom, K. E., Ledbetter, J. A,, McGowan, and Linsley, P. S. (1992).Costimulation of antitumor immunity by the B7 counterreceptor for the T lymphocyte molecules CD28 and CTLA-4. Cell 71, 1093-1102. Clark, E. A,, and Ledbetter, J. A. (1994). How B and T cells talk to each other. Nature 367,425-428. Carry, D. B., Reiner, S. L., Linsley, P. S., and Locksley, R. M. (1994). Differential effects of blockade of CD28-B7 on the development of Thl or Th2 effector cells in experimental Leishmaniasis. J. Immunol. 153, 4142-4148. Cross, A. H., Girard, T. J., Giacoletto, K. S., Evans, R. J., Keeling, R. M., Lin, R. F., Trotter, J. L., and Karr, R. W. (1995). Long-term inhibition of murine experimental autoimmune encephalomyelitis using CTLA-4-Fc supports a key role for CD28 costimulation.J. Clin. Invest. 95, 2783-2789. Damle, N. K., Mohagheghpour, N., Hansen, J. A., and Engleman, E. G. (1983). Alloantigenspecific cytotoxic and suppressor T lymphocytes are derived from phenotypically distinct precursors. J. Immunol. 131, 2296-3300. Damle, N. K., Linsley, P. S., and Ledbetter, J. A. (1991). Direct helper T cell-induced B cell differentiation involves interaction beiween T cell antigen CD28 and B cell activation antigen B7. Eur. J. Immunol. 21, 1277-1282.

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

159

Damle, N . K., Klussman, K., Leytze, G., Myrdal, S., Aruffo, A., Ledbetter, J. A,, and Linsley, P. S. (1994). Costimulation of T lymphocytes with integrin ligands intercellular adhesion molecule-1 or vascular cell adhesion molecule-l induces functional expression of CTLA-4, a second receptor for B7. /. lmmunol. 152, 2686-2697. Das, M. R. P., Zamvil, S. S., Borriello, F., Weiner, H. L., Sharpe, A. H., and Kuchroo, V. K. (1995).Reciprocal expression of co-stimulatory molecules, B7-1 and B7-2, on murine T cells foll.owing activation. Eur. 1. lrnmunol. 25, 207-211. de Boer, M., Kasran, A., Kwekkeboom, J., Walter, H., Vandenberghe, P., and Ceuppens, J. L. (1993). Ligation of B7 with CD28/CTLA-4 on T cells results in CD40 ligand expression, interleukin-4 secretion and efficient help for antibody production by B cells. Eur. ] lmmunol. 23,3120-3125. Fink, B. K., Linsley, P. S., and Wofsy, D. (1994).Treatment of murine lupus with CTLA4Ig. Science 265, 1225-1227. Freedman, A. S., Freeman, G., Horowitz, J. C., Daley, J., and Nadler, L. M. (1987). B7, a B cell-restricted antigen that identifies preactivated B cells.]. lmmunol. 139,3260-3267. Freedman, A. S., Freeman, G. J., Rhynhart, K., and Nadler, L. M. (1991).Selective induction of B7/BB-1 on interferon-y stimulated monocytes:A potential mechanism for amplification of T cell activation. Cell. lmmunol. 137, 429-437. Freeman, G. J,, Freedman, A. S., Segil, J. M., Lee, G., Whitman, J. F., and Nadler, L. M. (1989). B7, a new member of the Ig superfamily with unique expression on activated and neoplasic B cells. 1.lmmunol. 143, 2714-2722. Freeman, G. J., Gray, G. S., Gimmi, C. D., Lombard, D. B., Zhou, L. J., White, M., Fingeroth, J. D., Gribben, J. G., and Nadler, L. M. (1991). Structure, expression, and T cell costirnulatory activity of the murine homologue of the human B lymphocyte activation antigen B7. /. Exp. Med. 174, 625-631. Freeman, G. J., Lombard, D. B., Gimmi, C. D., Brod, S. A., Lee, K., Laning, J. C., Haller, D. A,, Dorf, M. E., Reiser, H., June, C. H., Thompson, C. B., and Nadler, L. M. (1992). CTLA-4 and CD28 mRNA are coexpressed in most T cells after activation. Expression of CTLA-4 an CD28 mRNA does not correlate with the pattern oflymphokine production. 1. lmmunol. 149,3795-3801. Freeman, G. J., Borriello, F., Hodes, R. J., Reiser, H., Hathcock, K. S., Laszlo, G. L., McKnight, A. J., Kim, J., Du, L., Lombard, D. B., Gray, G. S . , Nadler, L. M., and Sharpe, A. H. (1993a). Uncovering of functional alternative CTLA-4 counter-receptor in B7deficient mice. Science 262, 907-909. Freeman, G. J,, Gribben, J. G., Boussiotis, V. A,, Ng, J. W., Restivo, V. A., Jr., Lombard, L. A,, Gray, G. S., and Nadler, L. M. (1993b). Cloning of B7-2: A CTLA-4 counterreceptor that costimulates human T cell proliferation. Science 262, 909-91 1. Freeman, G. J., Borriello, F., Hodes, R. J., Reiser, H., Gribben, J. G., Ng, J. W., Kim, J., Goldberg, J. M., Hathcock, K., Laszlo, G., Lombard, L. A., Wang, S., Gray, G. S., Nadler, L. M., and Sharpe, A. H. (1993~).Murine B7-2, an alternative CTLA-4 counter-receptor that costimulates T cell proliferation and interleukin 2 production.]. Exp. Med. 178,21852192. Freeman, G. J., Boussiotis, V. A,, Anumanthan, A,, Bernstein, G. M., Ke, X. Y., Rennert, P. D., Gray, G. S., Gribben, J. G., and Nadler, L. M. (1995). B7-1 and B7-2 do not deliver identical costimulatory signals, since B7-2 but not B7-1 preferentially costimulates the initial production of IL-4. Zmmunity 2, 523-532. Fuller, K. A,, Kanagawa, O., and Nahm, M. H. (1993). T cells within germinal centers are specific for the immunizing antigen. J. Zmmunol. 151,4505-4512. Calvin, F., Freeman, G. J., Razi-Wolf, Z., Hall, W., Jr., Benacerraf, B., Nadler, L., and Reiser, H. (1992).Murine B7 antigen provides a sufficient costimulatorysignal for antigenspecific and MHC-restricted T cell activation. ]. lmmunol. 149, 3802-3808.

160

KAREN S. HATHCOCK AND RICHARD J. HODES

Gimmi, C., Freeman, G., Gribben, J,, Sugita, K., Freedman, A. S., Morimoto, C., and Nadler, L. (1991). B cell surface antigen B7 provides a costimulatory signal that induces T cells to proliferate and secrete interleuldn 2. Proc. Natl. Acad. Sci. USA 88,6575-6579. Green, J. M., Noel, P. J., Sperling, A. I., Walunas, T. L., Gray, G., Bluestone, J. A., and Thompson, C. B. (1994). Absence of B7-dependent responses in CD28-deficient mice. Immunity 1, 501-508. Gross, J. A,, Callas, E., and Allison, J. P. (1992). Identification and distribution of the costimulatory receptor CD28 in the mouse. 1.Immunol. 149, 380-388. Guinan, E. C., Gribben, J. G., Boussiotis, V. A., Freeman, G. J., and Nadler, L. M. (1994). Pivotal role of the B7:CD28 pathway in transplantation tolerance and tumor immunity. Blood 84, 3261-3282. Haffar, 0. K., Smithgall, M. D., Bradshaw, J., Brady, B., Damle, N. K., and Linsley, P. S. (1993). Costimulation of T-cell activation and virus production by B7 antigen on activated CD4' T cells from human immunodeficiency virus type I-infected donors. Proc. Natl. Acad. Sci. USA 90, 11094-11098. Han, S., Hathcock, K., Zheng, B., Kepler, T. B., Hodes, R., and Kelsoe, G. (1995). Cellular interaction in germinal centers. Roles of CD40 ligand and B7-2 in established germinal centers. 1.Immunol. 155, 556-567. Harding, F. A,, McArthur, J. G., Grossi, J. A., Raulet, D. H., and Allison, J. P. (1992). CD28-mediated signalling co-stimulates murine T cells and prevents induction of energy in T-cell clones. Nature 356, 607-609. Harlan, D. M., Hengartner, H., Huang, M. L., Kang, Y. H., Abe, R., Moreadith, R. W., Pircher, H., Gray, G. S., Ohashi, P. S., Freeman, G. J., Nadler, L. M., June, C. H., and Aichele, P. (1994). Transgenic mice expressing both B7-1 and viral glycoprotein on pancreatic beta cells along with glycoprotein-specifictransgenic T cells develop diabetes due to a breakdown of T lymphocyte unresponsiveness. Proc. Natl. Acad. Sci. USA 91,3137-3141. Harlan, D. M., Abe, R., Lee, K. P., and June, C. H. (1995). Potential roles of the B7 and CD28 receptor families in autoimmunity and immune invasion. Clin. Immunol. Immunopathol. 75, 99-111. Harper, K., Balzano, C., Rouvier, E., Mattei, M. G., Luciani, M. F., and 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, 1037-1044. Hathcock, K. S., Laszlo, G. L., Dickler, H. B., Bradshaw, J., Linsley, P., and Hodes, R. J. (1993). Identification of an alternative CTLA-4 ligand costimulatory for T cell activation. Science 262,905-907. Hathcock, K. S., Laszlo, G.'L.,Pucillo, C. P., Linsley, P., and Hodes, R. J. (1994).Comparative analysis of B7-1 and B7-2 costimulatory ligands: Expression and function. 1.Erp. Med. 180, 631-640. Hutchcroft, J. E., and Bierer, B. E. (1994). Activation-dependent phosphorylation of the T-lymphocyte surface receptor CD28 and associated proteins. Proc. Natl. Acad. Sci. USA 91,3260-3264. Inaba, K., Witmer-Pack, M., Inaba, M., Hathcock, K. S., Sakuta, H., Azuma, M., Yagita, H., Okumura, K., Linsley, P. S., Ikehara, S., Muramatsu, S., Hodes, R. J., and Steinman, R. M. (1994).The tissue distribution of the B7-2 costimulator in mice: Abundant expression on dendritic cells in situ and during maturation in vitro. 1.Exp. Med. 180, 1849-1860. Jacob, J., Kelsoe, G., Rajewsky, K., and Weiss, U. (1991a). Intraclonal generation of antibody mutants in germinal centers. Nature 354, 389-392.

ROLE OF CDPA-Bi COSTIMULATORY PATHWAYS

161

Jacob, J., Kassir, R., and Kelsoe, G. (1991b). In situ studies of the primary immune response I. The architecture and dynamics of responding cell to (4-hydroxy-3-nitropheny1)acetyl. populations. J. Exp. Med. 173, 1165-1175. Jacob, J., and Kelsoe, G. (1992). In situ studies of the primary immune response to (4hydroxy-3-nitrophenyl)-acetyl. 11. A common clonal origen for PALS-associated foci and germinal centers. J. Exp. Med. 176, 679-687. Jacob, J., Miller, C., and Kelsoe, G. (1992). In situ studies of the antigen-driven somatic hypermutation of immunoglobulin genes. Immunol. Cell B i d . 70, 145-5152. Jenkins, M. K., and Schwartz, R. H. (1987). Antigen presentation by chemically modified splenocytes induces antigen-specific T cell unresponsiveness in vitro and in uivo. J. Exp. Med. 165,302-319. Jenkins, M. K., Ashwell, J. D., and Schwartz, R. H. (1988). Allogeneic non-T spleen cells stimulated with antigen and chemically modified antigen-presenting cells. /. Immunol. 140,3324-3330. Jenkins, M. K., Taylor, S. P., Norton, S. D., and Urdahl, K. B. (1991). CDZ8 delivers a costimulatory signal involved in antigen-specific IL-2 production by human T cells. J . Immunol. 147, 2461-2466. June, C. H., Ledbetter, J. A., Linsley, P. S., and Thompson, C. B. (1990). Role of the CD28 receptor in T cell activation. Immunol. Toduy 11, 211-216. June, C. H., Bluestone, J. A,, Nadler, L. M., and Thompson, C. B. (1994). The B7 and CD28 receptor families. Zmnzunol. Tocluy 15, 321 -331. Kelsoe, G. (1994). B cell diversification and differentiation in the periphery. J. Exp. Med. 180, 5-6. Klaus, S. J., Pinchuk, L. M., Ochs, H. D., Law, C. L., Fanslow, W. C., Armitage, R. J., and Clark, E. A. (1994). Coshmulation through CD28 enhances T cell-dependent B cell activation via CD40-CD40L interaction. J . Immunol. 152, 5643-5652. Koulova, L., Clarke, E. A., Shu, G., and Dupont, B. (1991). The CD28 ligand B7/BB1 provides costimulatory signal for alloactivation of CD4+T cel1s.J. Exp. Med. 173,759-762. Kozbor, D., Moretta, A., Messner, H. A., Moretta, L., and Croce, C. M. (1987). Tp44 molecules involved in antigen-independent T cell activation are expressed on human plasma cells. J. Immunol. 138, 4128-4132. Krummel. M. F., and Allison, J. P. (1995). CD28 and CTLA-4 have opposing effects on the response of T cells to stimulation. J . Exp. Med. 182, 459-465. Kuchroo, V. K., Das, M. P., Brown, J. A,, Ranger, A. M., Zamvil, S. S., Sobel, R. A., Weiner, H. L., Nabavi, N., and Glimcher, L. H. (1995). 87-1 and B7-2 costimulatory molecules activate differentially the Thl/Th2 developmental pathways: Application to autoimmune disease therapy. Cell 80, 707-718. Kuiper, H. M., Brouwer, M., Linsley, P. S., and van Lier, R. A. W. (1995). Activated T cells can induce high levels of CTLA-4 expression on B cells.]. Immunol. 155,1776-1783. Lane, P., Burdet, C., Hubele, S,, Scheidegger, D., Muller, U., McConnell, and KoscoVilbois, M. (1994). B cell function in mice transgenic for mCTLA4-Hyl: Lack of germinal centers correlated with poor affinity maturation and class switching despite normal primung of CD4’ T cells. J . Exp. Med. 179, 819-830. Lanier, L. L., O’Fallon, S., Somoza, C., Phillips, J. H., Linsley, P. S., Okumura, K., Ito, D., and Azuma, M. (1995). CD80 (B7) and CD86 (B70) provide similar costimulatory signals for T cell proliferation, cytokiue production, and generation of CTL. J. Immunol. 154, 97-105. Larsen, C. P., Ritchie, S. C., Pearson, T. C., Linsky, P. S., and Lowry, R. P. (1992). Functional expression of the costimulatory molecule, B7/BB1 on murine dendritic cell populations. J Exp Med. 176, 1215-1220.

162

KAREN S . HATHCOCK AND RICHARD J. HODES

Larsen, C. P., Ritchie, S. C., Hendrix, R., Linsley, P. S., Hathcock, K. S., Hodes, R. J., Lowry, R. P., and Pearson, T. C. (1994). Regulation of immunostimulatory function and costimulatory function molecule (B7-1 and B7-2) expression on murine dendritic cells. 1.Immunol. 152,5208-5219. Ledbetter, J. A., Martin, P. J., Spooner, C. E., Wofsy, D., Tsu, T. T., Beatty, P. G., and Gladstone, P. (1985). Antibodies to Tp67 and Tp44 augment and sustain proliferative responses of activated T cells. J. Immunol. 135, 2331-2336. Ledbetter, J. A,, Parsons, M., Martin, P. J,, Hansen, J. A., Rabinovitch, P. S., and June, C. H. (1986). Antibody binding to CD5 (Tp67) and Tp44 T cell surface molecules: Effects on cyclic nucleotides, cytoplasmic free calcium, and CAMP-mediated suppression. J. Immunol. 137, 3299-3305. Ledbetter, J. A., Imboden, J. B., Schieven, G. L., Grosmaire, L. S., Rabinovitch, P. S., Lindsten, T., Thompson, C. B., and June, C. H. (1990).CD28 ligation in T-cell activation: Evidence for two signal transduction pathways. Blood 75, 1531-1538. Ledbetter, J. A., and Linsley, P. S. (1992). CD28 receptor crosslinking induces tyrosine phosphorylation of PLCTI.In “Advances in Experimental Medical Biology” (S. Gupta and T. A. Waldman, Eds.), Vol. 323, pp. 23-27. Plenum Press, New York. Lenschow, D. J., Zeng, Y., Thistlethwaite, J. R., Montag, A., Brady, W., Gibson, M. G., Linsley, P. S., and Bluestone, J. A. (1992). Long-term survival of xenogeneic pancreatic islet grafts induced by CTLA4Ig. Science 257, 789-795. Lenschow, D. J., Su, G. H. T., Zuckerman, L. A,, Nabavi, N., Jellis, C. L., Gray, G. S., Miller, J., and Bluestone, J. A. (1993). Expression and functional significance of an additional ligand for CTLA-4. Proc. Natl. Acad. Sci. U S A 90, 11054-11058. Lenschow, D. J., Sperling, A. I., Cooke, M. P., Freeman, G., Rhee, L., Decker, D. C., Gray, G., Nadler, L. M., Goodnow, C. C., and Bluestone, J. A. (1994). Differential upregulation of B7-1 and B7-2 costimulatory molecules after Ig receptor engagement by antigen. J. Immunol. 153, 1990-1997. Lenschow, D. J., Ho., S. C., Sattar, H., Rhee, L., Gray, G., Nabavi, N., Herold, K. C., and Bluestone, J. A. (1995). Differential effects ofanti-B7-l and anti-B7-2monoclonal antibody treatment on the development of diabetes in the nonobese diabetic mouse. I. Exp. Med. 181, 1145-1155. Levine, B. L., Ueda, Y., Craighead, N,, Huang, M. L., and June, C. H. (1995). CD28 ligands CD80 (B7-1) and CD86 (B7-2) induce long-term autocrine growth of CD4’ T cells and induce similar patterns of cytokine secretion in uitro. Int. Immunol. 7, 891-904. Lin, H., Bolling, S. F., Linsley, P. S., Wei, R. Q.. Gordon, D., Thompson, C. B., andTurka, L. A. (1993). Long-term acceptance of major histocompatibility complex mismatched cardiac allografts induced by CTLA4Ig plus donor-specific transfusion. J. Exp. Med. 178, 1801-1806. Lindsten, T., Lee, K. P., Harris, E. S., Petryniak, B., Craighead, N., Reynolds, P. J., Lombard, D. B., Freeman, G. J., Nadler, L. M., Gray, G. S., Thompson, C. B., and June, C. H. (1993). Characterization of CTLA-4 structure and expression on human T cells.]. Immunol. 151, 3489-3499. Linsley, P. S., Clarke, E. A,, and Ledbetter, J. A. (1990).The cell antigen, CD28, mediates adhesion with B cells by interacting with the activation antigen, B7/BB-1. Proc. Natl. Acud. Sci. USA 87,5031-5035. Linsley, P. S., Brady, W., Grosmaire, L., Aruffo, A., Damle, N. K., and Ledbetter, J. A. (1991a).Binding of the B cell activation antigen B7 to CD28 costimulates T cell proliferation and interleukin-2 mRNA accumulation. J. Exp. Med. 173, 721-730. Linsley, P. S., Brady, W., Urnes, M., Grosmaire, N. K., and Ledbetter, J. A. (1991b). CTLA4 is a second receptor for the B cell activation antigen B7. J. Exp. Med. 174, 561-569.

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

163

Linsley, P. S., Green, J. L., Tan, P., Bradshaw, J., Ledbetter, J. A,, Anasetti, C., and Damle, N. K. (1992a). Coexpressionand functional cooperation of CTLA-4 and CD28 on activated T lymphocytes.J. Erp. Med. 176, 1595-1604. Linsley, P. S., Wallace, P. M., Johnson, J., Gibson, M. G., Greene, J. L., Ledbetter, J. A., Singh, C., and Tepper, M. A. (1992b). Immunosuppression in vivo by a soluble form of the CTLA-4 T cell activation molecule. Science 257, 792-795. Linsley, P. S., Greene, J. L., Brady, W., Bajorath, J., Ledbetter, J. A., and 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. Linsley, P. S., and Ledbetter, J. A. (1993). The role of the CD28 receptor during T cell responses to antigen. Annu. Rev. Immunol. 11, 191-212. Linsley, P. S. (1995). Distinct roles for CD28 and cytotoxic T lymphocyte-associated molecule-4 receptors during T cell activation. J. Exp. Med. 182, 289-292. Liu, Y. J., Zhang, J., Lane, P. J., Chan, E. Y., and MacLennan, I. C. (1991). Sites of specific B cell activation in primary and secondary responses to T cell-dependent and T cellindependent antigens. Em-. ]. Immunol. 21,2951-2962. Lu, Y., Granelli-Piperno, A,, Bjomdahl, J. M., Phillips, C. A,, and Trevillyan, J. M. (1992). CD28-induced T cell activation. Evidence for a protein-tyrosine kinase signal transduction pathway. J. Immunol. 149, 24-29. Lum, L. G., Orcutt-Thordarson, N., Seigneuret, M. C., and Hansen, J. A. (1982). In vitro regulation of immunoglobulin synthesis by T-cell subpopulations defined by a new human T-cell antigen (9.3). Cell. Immunol. 72, 122-129. MacLennan, I. C. M. (1994). Germinal Centers. Annu. Reo. Immunol. 12, 117-139. McArthur, J. G., and Raulet, D. H. (1993). CD28-induced costimulation of T helper type 2 cells mediated by induction of responsiveness to interleukin 4. J. Erp. Med. 178,16451653. McHeyzer-Williams, M. G., McLean, M. J., Lalor, P. A,, and Nossel, G. J. V. (1993). Antigen-driven B cell differentiation in vivo. J. Exp. Med. 175, 295-307. Miller, C., Kelsoe, G., and Han, S. (1994). Lack of B7-2 expression in the germinal centers of aged mice. Aging lmmunol. Infect. Dis. 5, 249-257. Miller, C., and Kelsoe, G. (1995). Ig VHhypermutation is absent in the germinal centers of aged mice. J Immunol. 155, 3377-3384. Muraille, E., De Smedt, T., Thielemans, K., Urbain, J., Mosier, M., and Leo, 0. (1995). Activation of murine T cells by bacterial superantigens requires B7-mediated costimulation. Cell. Immunol. 162, 315-320. Nabavi, N., Freeman, G. J., Gault, A,, Godfry, D., Nadler, L. M., and Climcher, L. M. (1992). Signalling through the MHC class I1 cytoplasmic domain is required for antigen presentation and induces B7 expression. Nature 360, 266-268. Nandi, D., Gross, J. A,, and Allison, J. P. (1994). CD2a-mediated costimulation is necessary for optimal proliferation of murine NK cells. J. Immunol. 152,3361-3369. Nozawa, Y.,Wachi, E., Tominaga, K., Abe, M., and Wakasa, H. (1993).A novel monoclonal antibody (FUN-1) identifies an activation antigen in cells of the B-cell lineage and ReedSternberg cells. J. Puthot. 169, 309-315. Nunes, J., Bagnasco, M., Lopez, M., Lipcey,C., Mawas, C., andolive, D. (1991).Dissociation between early and late events in T cell activation mediated through CD28 surface molecule. Mol. Immunol. 28, 427-435. Nunes, J., Klasen, S., Franco, M. D., Lipcey, C., Mawas, C., Bagnasco, M., and Olive, D. (1993). Signalling through CD28 T-cell activation pathway involves an inositol phospholipid-specific phospholipase C activity. Biuchem. J. 293, 835-842.

164

KAREN S. HATHCOCK AND RICHARD J. HODES

Pages, F., Ragueneau, M., Rottapel, R., Truneh, A,, Nunes, J., Imbert, J., and Olive, D. (1994). Binding of phosphatidyl-inositol-3-OH kinase to CD28 is required for T-cell signalling. Nature (London) 369, 327-329. Pearson, T. C., Alexander, D. Z., Winn, K. J., Linsley, P. S., Lowry, R. P., and Larsen, C. P. (1994).Transplantation tolerance induced by CTLA4-Ig. Transplantation 57,1701-1706. Prasad, K. V., Cai, Y. C., Raab, M., Duckworth, B., Cantley, L., Shoelson, S. E., and Rudd, C. E. (1994). T-cell antigen CD28 interacts with the lipid kinase phosphatidylinositol 3kinase by a cytoplasmic Tyr(P)-Met-Xaa-Met motif. Proc. N d . Acad. Sci. USA 91,28342838. Ranheim, E. A,, and Kipps, T. J. (1993). Activated T cells induce expression of B7/BB1 on normal or leukemic B cells through a CD40-dependent signal.]. Exp. Med. 177,925-935. Razi-Wolf, Z., Freeman, G. J., Galvin, F., Benacerraf, B., Nadler, L., and Reiser, H. (1992). Expression and function of the murine B7 antigen, the major costimulatory molecule expressed by peritoneal exudate cells. Proc. Natl. Acad. Sci. USA 89, 4210-4214. Reiser, H., Freeman, G. J., Razi-Wolf, Z., Gimmi, C. D., Benacerraf, B., and Nadler, L. M. (1992). Murine B7 antigen provides an efficient costimulatory signal for activation of murine lymphocytes via the T-cell receptor/CD3 complex. Proc. Natl. Acad. Sci. USA 89, 271-275. Robey, E., and Allison, J. P. (1995).T-cell activation: Integration of signals from the antigen receptor and costimulatory molecules. Immunol. Today 16, 306-310. Ronchese, F., Hauseman, B., Hubele, S., and Lane, P. (1994). Mice transgenic for a soluble form of murine CTLA-4 show enhanced expansion of antigen-specific CD4+ T cells and defective antibody production in vivo. J . Exp. Med. 179, 809-817. Roy, M., Aruffo, A,, Ledbetter, J., Linsley, P., Kehry, M., and Noelle, R. (1995). Studies on the interdependence of gp39 and B7 expression and function during antigen-specific immune responses. Eur. J. Immunol. 25, 596-603. Sansom, D. M., and Hall, N. D. (1993). B7/BB1, the ligand for CD28, is expressed on repeatedly activated human T cells in vitro. Eor. ]. lrnmunol. 23, 295-298. Sayegh, M. H., Akalin, E., Hancock, W. W., Russell, M. E., Carpenter, C. B., Linsley, P. S., and Turka, L. A. (1995). CD28-B7 blockade after alloantigenic challenge in vivo inhibits Thl cytokines but spares Th2. J. Exp. Med. 181, 1869-1874. Schneider, H., Prasad, V. S., Shoelson, S. E., and Rudd, C. E. (1995). CTLA-4 binding to the lipid kinase phosphatidyl-inositol 3-kinase in T cells. J. Exp. Med. 181, 351-355. Schwartz, R. H. (1992). Costimulation of T lymphocytes: The role of CD28, CTLA-4, and B7/BB1 in interleukin-2 production and immunotherapy. Cell 71, 1065-1068. Seder, R. A,, and Paul, W. E. (1994). Acquisition of lymphokine-producing phenotype by CD4+ T cells. Annu. Rev. Immunol. 12, 635-673. Sethna, M. P., van Parijs, L., Shave, A. H., Abbas, A. K., and Freeman, G. J. (1994). A negative regulatory function of B7 revealed in B7-1 transgenic mice. Immunity 1,415-421. Shahinian, A,, Pfeffer, K., Lee, K. P., Kundig, T. M., Kishihara, K., Wakeham, A,, Kawai, K., Ohashi, P. S.,Thompson, C. B., and Mak, T. W. (1993). Differential Tcell costimulatory requirements in CD28-deficient mice. Science 261, 609-612. Snapper, C. M., and Finkelman, F. D. (1993). Immunoglobulin class switching. In “Fundamental Immunology, Third Edition” (W. E. Paul, ed.), pp. 837-863. Raven Press, New York. Snow, E. C., and Noelle, R. J. (1993). The role of direct cellular communication during the development of a humoral immune response. Adv. Cancer Rex 62, 241-266. Stack, R. M., Lenschow, D. J., Gray, G . S., Bluestone, J. A,, and Fitch, F. W. (1994). IL-4 treatment of small splenic B cells induces costimulatory molecules B7-1 and B7-2. J. Immunol. 152, 5723-5733.

ROLE OF CD28-B7 COSTIMULATORY PATHWAYS

165

Stein, P. H., Fraser, J. D., and 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, 3392-3402. Tan, P., Anasetti, C., Hansen, J. A., Mekose, J.. Brunvand, M., Bradshaw, J., Ledbetter, J. A,, and Linsley, P. S. (1993). Induction of alloantigen-specific hypresonsiveness in human T lymphocytes by blocking interaction of CD28 with its natural ligand B7/BBl. J. Exp. Med. 177, 165-173. Tew, J. G., Kosco, M . H., Burton, G. F., and Szakal, A. K. (1990). Follicular dendritic cells as accessory cells. Immunol. Rev. 117, 185-211. 'Thompson, C. B., Leiden, J. M., and June, C. J. (1989). CD28 activation pathway regulates the production of multiple T-cell-derived lymphokines/cytokines.Proc. Natl. Acad. Sci. USA 86, 1333-1337. Thompson, C . B. (1995). Distinct roles for the costimulatory ligands B7-1 and B7-2 in T helper cell differentiation. Cell 81, 979-982. Townsend, S. E., and Allison, J. P. (1993). Tumor rejection after direct costimulation of CD8+ T cells by B7-transfected melanoma cells. Science 259, 368-370. Truitt, K. E., Hicks, C. M., and Imboden, J. B. (1994). Stiniulation of CD28 triggers an association between CD28 and phophatidylinositol 3-kinase in Jurkat cells. J . Exp. Med. 179, 1071-1076. Turka, L. A., Linsley, P. S., Lin, H., Brady, W., Leiden, J. M., Wei, R. Q., Gibson, M. L., Zheng, X. G., Myrdal, S., Gordon, D., Bailey, T., Bolling, S. F., and Thompson, C. B. (1992). T-cell activation by the CD28 ligand B7 is required for cardiac allograft rejection in uiuo. Proc. Natl. Acad Sci. USA 89, 11102-11105. Valle, A,, Garrone, P., Yssel, H., Bonnefoy, J. Y., Freedman, A. S., Freeinan, G., Nadler, L. M., and Banchereau, J. (1990). Mab 104 a new monoclonal antibody, recognizes the B7 antigen which is expressed on activated B cells and HTLV-1 transformed T cells. lmmunology 69,531-535. Valle, A,, Aubry, J.-P., Durand, I., and Banchereau, J. (1991). IL-4 and IL-2 upregulate the expression of antigen B7, the B cell counterstructure to T cell CD28: An amplification mechanism for T-B cell interactions. Znt. Immunol. 3, 229-235. Vandenberghe, P., Freeman, G. J., Nadler, L. M., Fletcher, M. C., Lamoun, M., Turka, L. A,, Ledbetter, J. A,, Thompson, C. B., and June C. H. (1992). Antibody and B7/BB1mediated ligation of the CD28 receptor induces tyrosine phosphorylation in human T cells. J. Exp. Med. 175, 951-960. Vandenberghe, P., Delable, J., de Boer, M., De Wolf-Peeters, C., and Ceuppens, J. L. (1993). In situ expression of B7-BB1 on antigen presenting cells and activated B cells: An immunohistochemical study. lnt. lmmunol. 5, 317-321. Van den Eertwegh, A. J. M., Noelle, R. J., Roy, M., Shepherd, D. M., Aruffo, A,, Ledbetter, J. A,, Boersma, W. J. A,, and Claassen, E. (1994). In viva CD40-gp39 interactions are essential for thymus-dependent humoral immunity. I. In vivo expression of CD40 hgand, cytokines, and antibody production delineates sites of cognate T-B cell interactions. J Exp. Med. 178,1555-1565. \'an Cool, S. W., de Boer, M., and Ceuppens, J. L. (1994). The combination of anti-B7 monoclonal antibody and Cyclosporin A induces alloantigen-specific anergy during a primary mixed lymphocyte reaction. J . Exp. Med. 179, 715-720. Wallace, P. M., Rodgers, J. N., Leytze, G. M., Johnson, J. S., and Linsley, P. S. (1995). Induction and reversal of long-lived specific unresponsiveness to a T-dependent antigen following CTLA4Ig treatment. J. Immunol. 154, 5885-5895. Walunas, T. L., Lenschow, D. J., Baker, C. Y., Linsley, P. S., Freeman, G. J., Green, J. M., Thompson, C. B., and Bluestone, J. A. (1994). CTLA-4 can function as a negative regulator of T cell activation. lmmunity 1, 405-413.

166

KAREN S. HATHCOCK AND RICHARD J. HODES

Ward, S. G., Westwick, J., Hall, N. D., and Sansom, D. M. (1993). Ligation of CD28 receptor by B7 induces formation of D-3 phosphotides in T lymphocytes independently of T cell receptor activation. Eur. J. lmmunol. 23, 2572-2577. Weiss, A., Manger, B., and Imboden, J. B. (1986). Synergy between the TYantigen receptor complex and Tp44 in the activation of human T cells. J. Immunol. 137, 819-825. Yang, G., Helfstrom, K. E., Hellstrom, I., and Chen, L. (1995). Antitumor immunity elicited by tumor cells transfected with B7-2, a second ligand for CD28ICTLA-4 costiinulatory molecules. J . lmmunol. 154, 2794-2880. Yokochi, T., Holly, R. D., and Clark, E. A. (1982). B lymphoblast antigen (BB-1) expressed on Epstein-Barr virus-activated B cell blasts, B lymphoblastoid cell lines, and Burkitt’s lymphomas. J. lmmunol. 128,823-827. Zheng, B., Xue, W., and Kelsoe, G. (1994).Locus-specific somatic hypermutation in germinal centre T cells. Nature 372, 556-559. This article was accepted for publication on 28 September 1995.

ADVANCES IN IMMUNOLOGY, VOL. 62

Prostaglandin Endoperoxide H Synthases- 1 and -2 WllLlAM 1. SMITH AND DAVID 1. D E W Dapifment of Eiachemishy, Michigan State Uniumify, East lnnsing, Michigan 48824

1. An Overview

Prostaglandin endoperoxide H synthase (PGHS)catalyzes the committed step in the formation of prostaglandins and thromboxanes. There are two PGHS isozymes called PGHS-1 and PGHS-2 or COX-1 and COX-2 (for cyclooxygenase). Both enzymes catalyze the formation of PGHz from arachidonic acid (Fig. 1).Structurally, the enzymes appear to be similar, and they have very similar kinetic properties. Both enzymes are inhibited by nonsteroidal anti-inflammatory drugs (NSAIDS) although inhibitors relatively specific for each isozyme have been developed. PGHzformed through the action of PGHS is converted via distinct synthases to PGDz, PGEz, PGF2,, PGIz, or thromboxane A2 (TxAJ (Fig. 1).These compounds act, at least in part, through specific G protein-linked receptors to modulate the levels of second messengers. PGHS-1 and PGHS-2 are encoded by separate genes. In general, PGHS1 can be viewed as a constitutive enzyme whose expression appears to be regulated developmentally. Prostanoids formed in the endoplasmic reticulum through the action of PGHS-1 e i t cells and function through cell surface receptors to mediate so-called “housekeeping” functions such as regulation of renal HzO and Na+ metabolism, stomach acid secretion, and hemostasis. In contrast, PGHS-2 is an inducible enzyme that is normally absent from cells but is expressed transiently in response to growth factors, tumor promoters, or cytokines. PGHS-2 appears to synthesize prostanoids on the nuclear envelope and these products may function at the level of the nucleus during cell replication or differentiation. PGHSs are of considerable therapeutic interest because of their established or alleged involvement in inflammation, coronary thrombosis, colon cancer, and Alzheimer’s disease. Under Sections I1 and I11 of this chapter we compare and contrast the structural and kinetic properties of PGHS-1 and -2. Included in this discussion is a description of the interactions of the two isozymes with NSAIDs. Under Section IV, we describe the structures and regulation of expression of the PGHS-1 and -2 genes. In the two final, more speculative 167 Copynght Q 1996 by Academic Press, Inc All nghts of reproduction in any form resewed

168

WILLIAM L. SMITH AND DAVID L. DEWITT

HORMONE

J CELL MEMBRANE

PHOSPHOLIPID (PI,PC,PE), TRIGLYCERIDE, OR CHOLESTEROL ESTER /

$-

LIPASE(S)

'

PEROXIDASE

pG@

bOH

ACTIVATION

I

ARACHIDONIC ACID CYCLOOXYGENASE

.-

f

PGH SYNTHASE

COOH

I

FIG. 1. Pathways for the biosynthesis of various prostanoids involving prostaglandin endoperoxide H synthase (PGHS).

sections, Sections V and VI, we discuss how the two isozymes may act independently in intact cells to mediate the formation of prostanoids destined to act on cell surface and/or nuclear targets to mediate different biological and pathobiological events.

CYCLOOXYGENASE ISOZYMES

169

II. PGHS-1 and PGHS-2 Struciure/Function Interrelationships

A. SEQUENCES AND STRUCTURE Shown in Fig. 2 are the deduced amino acid sequences of PGHS-1 and PGHS-2 from various species. The numbering is shown beginning with the methionine at the translational start site for ovine PGHS-1; because the lengths of the signal peptides vary in several cases, the numbering can be confusing. The mature processed forms of PGHSs-1 all have 576 amino acids, and the mature forms of PGHSs-2 have 587 amino acids. The primary sequences of PGHS-1 and -2 are 60% identical within a species. The most signficant sequence differences are in (a) the signal peptides, (b) the membrane-binding domains (MBDs) and (c)the C termini (Fig. 3 ) .Particularly notable is the 18-amino-acidinsert that is located near the C terminus of PGHS-2 and absent from PGHS-1. Illustrated in Fig. 3 are defined functional domains and some key amino acids of ovine PGHS-1. This latter enzyme was the first PGHS purified, and it has been used for most of the biochemical studies of PGHSs. The crystal structure of ovine PGHS-1 was reported by Garavito and co-workers (Picot et al., 1994) (Fig. 4).The enzyme exists as a head-to-tail homodimer. Each subunit is made up of three domains including a small N-terminal domain composed of an EGF module, a membrane-binding domain having four short helices, and a large, globular C-terminal catalytic domain.

B. SIGNAL PEPTIDES The N terminus of purified ovine PGHS-1 has the sequence 25ADPGAPAPVNPCCYYP4"(DeWitt and Smith, 1988). 18ANPCCSNPCQNRGE3' is the N-terminal sequence of isolated rat PGHS-2 (Sirois and Richards, 1992). Because both PGHS-1 and PGHS-2 are membrane proteins and because the sequences near the N termini of the unprocessed isozymes have sequences characteristic of signal peptides, it can be assumed that these peptides represent signal sequences. The signal peptides of PGHSs1 have 24-26 amino acids, and those of PGHS-2 have 17 amino acids. The signal recognition particle of the ER is apparently capable of recognizing both peptides so there is no a priori reason to think that the differences in the structures of the PGHS-1 and PGHS-2 signal peptides are particularly signficant in terms of targeting or processing. However, as is discussed below, the first 3 amino acids of PGHS-1 (MSR) are encoded by an exon of the PGHS-1 gene that is absent from the PGHS-2 gene.

FACTOR DOMAIN C. EPIDERMAL GROWTH Both PGHS-1 and PGHS-2 contain an epidermal growth factor (EGF)like sequence near their N termini. In ovine PGHS-1, the EGF-like sequence includes amino acids 32-79 (Fig. 3) (Toh, 1989). Examination of

-3

TSRSHLIDSPPTYNSDYSYKSVEAYSNLSYYTRSLPPVGHDCPTPMG~GKKELPDSKLI TSRSHLIDSPPTYNADYGYKSUEAFSNLSYYTRALPPVPDDCPTPLGVKGKKPLPDSNEI TSRSHLIDSPPTYNVHYGYKSUEAFSWLSYVTRALPPVADDCPTPMGVKGNKELPDSKEV TSRSYLlDSPPTYNVHYGYKSUEAFSwLSrrTRALPPVADDCPTPMGVKCNKELPDSKEV TVRSNL PSPPTYNSAHDYISVESFSYSYYTR LPSVPKDCPTPMGTKGKKPLPDML TVRSNL!PSPPTYNTAHDY SVESFSWVSYYTR~LPSVPKCPTPMGTKGKKPLPDIHLk TVRSNLlPSPPTYNSAHDY~SMSFSYSYYTRlLPSVPKDCPTPMGTKGKKPLPDAPLL T R NLIPSPPI J I I A H D Y I S ~ ~ F S Y S Y Y ~ ~ ~ L P S V P R DP PMGTKGK a PDAEFL

-5

VEKFLLRRKFIWPPGTNVnFTFFAPHFTHPFFKTDHKRGPGFTKAYGHGWLNHIYGET VGKLLLRRKFIPDPPGSNMMFAFFAPHFTHPFFKTDHKRGPAFTNGLGHGWLNHIYGET LEKVLLRREFIPDPPGTNMMFAFFAPHFTHPFFKTDPKRGPGFTRGLGHGWLNHVYGET LEKVLLRREFIPDPPGSN~FAFFAPHFTHPFFKTDHKRGPGFTRGLGHGWLNHIVGET APRLLLRREFIPAPPGTNILFAFFAPHFTHPFFKTSGKMGPGFTKALGHGWLGHIYWN APRLLLRREFIPAPPGTNVLFAFFAQHFTHFFKTSTKMGPGFTKALGHGWLGHIYWS ARRFLLRRKFIPDPPGTNLMFAFFAPHFTHPFFKTSGKMGPGFTKALGHGWLGHIYGDN GHIYGDN SRR LLRRKFI PPGTNLMF F APHFTHQ FKTSGKMGP KALGHG

ChickHuna

!Mous %!- -f kk!Sheep-] ChickH m

&%!--4 Rat-! llpana:l MWS

y2i

Chick-$

1 x9Mouse-f M W

ld

768

!lo

$06

$10

458

f7b

"Pfo

LERPLKLRLRKDGKL~PMIDGEMYPPTV~TPAEMIYPPHVPEHLPFSVGPEVFGL~G LARPRKlRLFKDGKMKYPIlDGEnYPPTVKDTPAEMlYPPPVPEHLRFAVGPEVFGLVPG LDRPHKLRLFPDGKLKYPVIGGEWPPTVKDTPWMIYPPHVPEHLRFAVGPEVFGLVPG LORPHKLRLFKDGKLKYPVlGGEVYPPTVKDTPVEMlYPPHlPENLPFAVGPEVFGLVPG LERPYHLRLFKDGKLKYPVLDGEWPPSVEPASVLMRYPPGVPPERPMAVWEVFGLLPG LERPYHLRLFKDGKLKYPVLDGELYPPSVEQASVKMRYPPGVPPEKP~VAPEVFGLLPG ~E~~~~~~~~KDGKLKYPVLDGEHYPPSVEEAPVLMHYPRGIPPPSPMAVGPEVFGLLPG %40 ICq~~LKYPHLNq~gYPPSVEE~~gLMHYPRG PPPSWVGO FGLLPC is0 2 H

:re$] Rat-!-

Chick-2

&9Mouse-f MOlB

Rat-!-

llpe;:j Chick-

Hunan-3

L~YA~~ULREHNRVCDVLKPEHPEH)DEPLFPTTRLlLlGETlKlVlDDY~HLSGYHF LHMYA ULREHNRVCDVLKPEHPEUGDEPLFPlSRLlLlGETlKlVlDDYVPHLSGYHF LMHYAT IULREHNRVCD I LKPEHPEWDERLFPTSRL I L I GET I KIVIEDYVPHLRGYHF LMMYATlULREHNRVCDlLKPEHPEUGDEPLFPTSRLlLlGETlKlVlDDYVPHLSGYHF LMLFSTIULREHNRVCDLLKEEHPTH)DEPLFPTTRLlLlGETlKlVlEEYVPHLSGYFL LMLFSTlULREHNRVCDLLKEEHPTH)DEPLFPTTRLlLlGETlElllEEY~HLSGYFL LMLYATLULREHNRVCDLLKAEHPTUGDEPLFPTTRLILIGETIKIVIEEYVPPLSGYFL L~~~ATIULRE~~~VWLLKA~~~TUGDEPL~~~ARLILIG~~~KIVIEEY~~~LSGYFL KLKFDPELLFNPRFPYPNRIMEFNTLYHUIIPLLPDTFPIHNPEYTFPPFLYNNSIMLEH KLKFDPELLFNKPFPYPNRlAAEFNTLYHUHPLLPDTFPlNDP~NYWFlYNNSlLLEH

fl~r:? Rat-2

!$&:I Chick-

Hmn-$ Rat-2 Mouse:? MOW

I&!Sheep-!

Chick-

Hunan-t

Rst-2 Mouse-? Mous Rat-!-

Chick-

Hunan-t Rat-2

Hze:f Rat-!

PLKFDPDLLFNPPFPYPNRlASEFKTLYHUHPLLPDTFNlEDPEYTFKPFLYNNSlLLEH KLKFDPEL FNPPFPYPNRlASEFNTLYHUHPLLPDTFNlEDPEYSFKPFLYNNSlLLEH PLKFDPELkFRAQFPYRNRIAMEFNHLYHUHPLHPNSFPVGSPEYSYEPFLFNTSMLWY PLKFDPELLFRAPFPYRNRIAMEFNHLYHUHPFMPDSFPVGSPEYSYEPFLFNTSMLWY PLKFDPELLFGVPFPYRNRlATEFNHLYHUHPLMPDSFKVGSPEYSYEPFLFNTSMLWY P KFDPELLF PFPYRNRIA NPLYHUH MPDSFRVGPWYSYEPFLFW SMLWY b60 970 !bD 400 418

%6

GLSHmKSSKRPIAGRVAGGKNVPMVPKVAKASIDPSRPMRYPSLNEYRKRFMLKPFKS GITPFVESFTRPIAGRVAGGRNVPPAVPKVSPASIDPSRPMKYPSFNEYRI(RFMLKPYES GLAHFVESFTRQIAGRVAGGRNVPIAVPAVAKASIDPSREMKYPSLNEYRKRFSLKPYTS GLTPFMSFTRPIAGRVAGGRNVPIAVPAVAKASIDPSREMKYPSLNEYRKRFSLKPYTS GVEALWAFSRPRAGRIGGGRNFDYHVLHVA VIKESREHRLPPFNEYRKRFGLKPYTS GVEALWAFSRPRAGRIGGGRNFDVHVLHVA~VIKESREMRLPSFNEYRKRFGLKPYTS GVEALMAFSRQIAGRIGGCRNMDHHILHVAWVIRESREMRLPPFNEYRKRFGMKPYTS G ALWAFSRQ AGRIGGGRN HWILHVA IKESRVLR PPFNEYRKR MKPYTS

150

411

4!0

tK;:]

4b0

458

TICSPEYVKPSTFGGKVGFEIINTASLPSLICNNVKGSPFTAFHVLNPEPTETATI~ST VlCSPAYUKPSTFGGEVGFPIlNTASlPSLICNNVKGCPFTSFSVPDPELlKTVTl~ P CSPPYVKPSTFGGEVGFRIINTAS PSL CNNVKGCPFASFNVPDPP TKTATIWKS* P~CSPPYUKPSTFGGEVGFKIlNTAS!PSLlCNNVKGCPFTSFNVPDPP~TKTATlPICSPEYWKPSTFGGDVGFNLVNTASLKK V C L N T K T C P Y V S F R V P D Y P G D D G S V I ~ P CSPEYUKPSTFGGDVGFNIVNTASL K VCLNTKTCPYVSFRVPDYPGDDGS~V--P CSPEYUKPSTFGGEVGFNIVKTATLEK VCLNTKTCPYVSFRVPDASPDDGPAVE--PJ~~PEVUKAS~JS~GEE'GFNL~~~ATLKKLV~~~TKTCPYV~~~VPDPRPE~~~GVE---

lJr;:l f

Rat-!-

8#

FEELTGEKE~ELEELYWlDAMELYPGLLVEKPRPGAlFGET~EIGAPFSLKGLWGN FEE~TGEKEMSAE~EA~~GD~~A~EL~PAL~~EKPRPDA~FGE~M~E~GAPF~~KG~~GN FEELTGEKEMMELKA YHD IOAMELYPALLVEKPRPDAIFGETMVELCAPFSLKGLMGN FEELTGEKEMMELKA~YSDIDV~ELYPALLVEKPRPDAIFGETMVE~~APFS~KGL~GN FPELTGEKEMM LEELYGDIDALEFYPGLLLEKCPPNSIFGESMIEMGAPFSLKGLLGN FQEFTGEKEMM~LEELYGDIDALEFYPGLMLEKCPPNSLFGESMIEMGAPFSLKGLLGN

PSAEL RSTEL RSTEL RSTEL RSTEL PSTEL RPSTEL

:::::::::::::::

RppTSbO

171

CYCLOOXYGENASE ISOZYMES

I

AIa25

I

I

I+\?

\A

/krgI2O EGF MBD

I

I

Leu600

iz

I

His388

(heme)

dr530 (aspirin)

\1 PTEL

FIG.3. Domain structure and locations of various important amino acids in ovine PGHS1. Shown ar the N and C termini, the Asn N-glycosylation sites, the epidermal growth factor (EGF) domain, the membrane-binding domain (MBD), various amino acids important in catalysis, and the PTEL endoplasmic reticulum retention signal.

this sequence in the crystal structure of ovine PGHS-1 indicates that it indeed forms an EGF-like structural module having three loops resulting from the formation of three intrachain disulfide bonds (Fig. 4). (Picot and Garavito, 1994; Picot et al., 1994). The EGF module of ovine PGHS-1 is linked to the globular catalybc domain by a disulfide bond between Cys37 and Cys159 (Picot and Garavito, 1994). EGF domains are found at the interface between PGHS-1 monomers in the holoenzyme; the EFG domain of each monomer is associated with a different region of the opposite subunit, giving the dimer a head-to-tail arrangement (Picot and Garavito, 1994; Picot et al., 1994). Because it has not been possible to prepare active monomers of PGHS, and because the crystal structure of ovine PGHS-1 indicates that there are specific interactions between enzyme subunits, it is assumed that the PGHS isozymes do function as dimers in membranes. Mutation of Cys69 to a serine in the EGF domain of ovine PGHS-1 yields a catalytically inactive protein ( J . C. Otto and W. L. Smith, unpublished results) consistent with the concept that this domain is important in maintaining the overall structure of the enzyme.

FIG. 2. Comparisons of deduced amino acid sequences PGHS-1 and PGHS-2 from various species. The numbering is shown for ovine PGH synthase-1 and begins with the metbionine at the translation start site. Underlined are the signal peptides, the putative membrane-binding domains (residues 70-117), and the C-terminal 18-amino acid cassettes near the C termini of PGHSs-2. Bold letters are asparagine residues that are N-glycosylated, catalyticallyessential histidine and tyrosine residues, and the serine residue that is acetylated by aspirin. PGHS-1 sequences are shown for ovine (DeWitt and Smith, 1988; Merlie et al., 1988; Yokoyama et al., 1988), human (Yokoyama and Tanabe, 1989; Funk et al., 1991), mouse (DeWitt et a[., 1990), and rat (Feng et al., 1993); PGHS-2 sequences are shown for chicken (Xie et al., 1991),human (Hla and Neilson, 1992;Jones et al., 1993), mouse (Kujubu et al., 1991), and rat (Feng et al., 1993).

172

WILLIAM L. SMITH AND DAVID L. DEWI’IT

Arg 277 POX

FIG.4. Structure of ovine PGHS-1. Ribbon diagram of the crystal structure of the ovine PGHS-1 dimer (Picot et al., 1994). Indicated in this diagram are the EGF domains, the peroxidase (POX) active site near the heme group, the trypsin cleavage site at Arg277, and the membrane-binding domain helices (A-D). The asterisk is in the cyclooxygenase active site where a sidecain of Tyr385 is shown above the asterisk and flurbiprofen is shown bound to this site just below the asterisk. This figure was kindly provided by Dr. R. Michael Garavito.

The C-terminal end of the EGF module in PGHS-1 extends into the putative membrane-binding domain involving residues 70-1 17. Therefore, it has been proposed that the EGF module may play an ancillary role in the association of PGHSs with lipid membranes (Picot and Garavito, 1994). Although the crystal structure of PGHS-2 has not yet been reported, sequence similarities between PGHS-1 and PGHS-2 suggest that PGHS2 also contains a functional EGF structural domain. In fact, we have prepared chimeras of human PGHS-1 and PGHS-2 in which the first 200 amino acids of each protein were “swapped,” and the chimeras are catalytically active (I. Song and W. L. Smith, unpublished results).

D. MEMBRANE-BINDING DOMAIN As discussed in more detail below, PGHS-1 and PGHS-2 are associated with the endoplasmic reticulum and the nuclear envelope, but not the plasma membrane (Rollins and Smith, 1980; Regier et al., 1993; Otto and

CYCLOOXYGENASE ISOZYMES

173

Smith, 1994; Morita et al., 1995). Empirically, ovine PGHS-1 is an integral membrane protein in that it can only be extracted from membrane preparations by the use of detergents (van der Ouderaa and Buytenhek, 1982). Consistent with this prediction, purified PGHS-1 can be reconstituted into pure phospholipid liposomes, indicating that the enzyme can associate directly with the phospholipid component of membranes (Strittmatter et al., 1982). Initially, this suggested that the protein contained one or more transmembrane domains, but efforts to define transmembrane domains yielded confusing results. Indeed, when the crystal structure of the enzyme was resolved, the protein was found to be essentially globular with no obvious transmembrane region. Garavito and co-workers proposed that the enzyme associated with one leaflet of the lipid bilayer through a MBD encompassing residues 70-117 of ovine PGHS-1 and consisting of four amphipathic helics (Figs. 3 and A-D in Fig. 4)(Picot and Garavito, 1994; Picot et al., 1994). Side chains of hydrophobic amino acids in these helices emanate from one surface of the protein and were proposed to be embedded monotopically in the bilayer. Biochemical studies using ['251]TID,a reagent which can be used to specifically photolabel membrane-associated domains of proteins (Blanton and Cohen, 1992), have indicated that a region (i.e., residues 25-166) encompassing the MBD is selectivelylabeled (Otto and Smith, unpublished results). As noted previously, the PGHS-1/ PGHS-2 chimeras in which the MBDs are interchanged are catalytically active indicating that PGHS-2 contains a MBD analogous to the MBD of PGHS-1. As discussed in more detail below, helix D of the MBD forms part of a hydrophobic channel that forms part of the cyclooxygenase active site. This positions the cyclooxygenase active site so that it can acquire arachidonic acid directly from the membrane.

E. N-GLYCOSYLATION SITES Ovine PGHS-1 migrates as a single band on SDS-PAGE with an apparent molecular mass of 72 kDa (Hemler et al., 1976; Miyamoto et at., 1976; van der Ouderaa et al., 1977); however, the molecular mass calculated from the amino acid sequence is 66 kDa (DeWitt and Smith, 1988; Merlie et al., 1988; Yokoyama et al., 1988). This difference is attributable to the presence of three N-linked, high-mannose oligosaccharides-one Man7(NAcGh)pand two Many(NAcGh)2groups (Mutsaers et al., 1985). The three sites of N-glycosylation (Asn-X-Thr/Ser) are Asn68, Asn144, and Asn410 (Otto et ul., 1993; Otto and Smith, 1994); the consensus Nglycosylation site at Asnl04 is not glycosylated in PGHS-1. Importantly, N-glycosylation at Asn410 is essential for the enzyme to fold into its active conformation (Otto et al., 1993).

174

WILLIAM L. SMITH AND DAVID L. DEWITI

In contrast to PGHS-1, PGHS-2 migrates as two (and sometimes three) bands on SDS-PAGE with apparent molecular masses of 72 and 74 kDa (Sirois and Richards, 1992; Habib et al., 1993; Otto et al., 1993; Phillips et al., 1993; Lecomte et al., 1994), compared to a calculated molecular mass of about 67 kDa. The 72-kDa species appears to contain three Nlinked oligosaccarides, whereas the 74-kDa species has four N-linked oligosaccharides. Asparagine residues homologous to those that are glycosylated in PGHS-1 are present in PGHS-2 and thus, are apparently glycosylated in this second isozyme. There is also a fourth asparagine Nglycosylation site, Asn580, in PGHS-2 located near the C terminus and adjoining the 18-amino-acid insert unique to this latter isozyme (Figs. 2 and 3).Asn580 is glycosylated in about 50%of PGHS-2 molecules accounting for both 72- and 74-kDa species (Otto et al., 1993). Glycosylation at this position is not required for the activity of PGHS-2, but it could serve as a marker for a unique subset of PGHS-2 molecules having some specific function. N-glycosylation of PGHS-2 is, however, important for enzyme activity because expression of PGHS-2 in the presence of tunicamycin yields an unglycosylated, inactive enzyme (Otto et al., 1993).

F. TRYPSIN CLEAVAGE SITES When either isolated ovine apo-PGHS-1 or sheep vesicular gland microsomes from which heme is depleted are incubated with trypsin, PGHS-1 is cleaved at a single site, Arg277, into 33- and 38-kDa fragments (Fig. 3) (Marnett et al., 1988; Kulmacz, 1989).Trypsin does not cleave ovine PGHS1 in the presence of heme and/or various NSAIDs. In fact, a differential sensitivity to trypsin cleavage in the presence of heme or NSAIDs provides an index of the structural integrity of catalytically inactive mutants of ovine PGHS-1 (Shimokawa and Smith, 1991, 1992). Curiously, Arg277 appears to be exposed in the crystal structure of the flurbiprofenholo-PGHS-1 complex (Picot and Garavito, 1994) (Fig. 4), so it is not obvious why heme and NSAIDs protect the enzyme from trypsin cleavage. The sequence around the trypsin cleavage site identified for ovine PGHS-1 is present in human but not mouse or rat PGHS-1 (Fig. 2). Murine PGHS-2 is not cleaved by trypsin in this region but is cleaved at Lys576 (Sirois and Richards, 1992; Otto and Smith, 1994). ACTIVESITE G. CYCLOOXYGENASE The cyclooxygenase activity of PGHSs oxygenates and isomerizes arachidonate converting it to PGGz (Fig. 1);PGGz is then released from the cyclooxygenase active site and travels to the spatially distinct peroxidase site on the enzyme where PGGz is reduced to PGHz (Eling et al., 1991). Arachidonic acid acylated at the sn-2 position of glycerophospholipids can

175

CYCLOOXYGENASE ISOZYMES

be viewed as the storage form of prostanoids. Arachidonate is mobilized in response to various hormonal stimuli (Fig. 1).In the case of PGHS-1, the arachidonate can be derived from phospholipids on the cytoplasmic face of the ER and the nuclear envelope. Cytosolic (c) cPLAz is activated and associates with these organelles in a Ca2+-dependent manner in hormone-stimulated cells (Fig. 5) (Glover et al., 1995; Schievella et al., 1996). Arachidonate released by c P U Z is oriented with its methyl end toward the lumenal half of the bilayer and moves into the mouth of a foursided pore formed by the four amphipathic helices of the membranebinding domain of PGHS-1. Newly released arachidonate traverses this opening and enters the hydrophobic channel that extends into the core of the enzyme and forms the cyclooxygenase active site (Picot and Garavito, 1994; Picot et al., 1994) (Figs. 4 and 6). At the innermost reaches of the channel there is an alcove (Picot and Garavito, 1994; Picot et al., 1994). A heme group is ligated at its proximal position by His388 that neighbors the channel near its upper reaches (Fig. 6); also located in this region are Tyr385 and Ser530, which reside on opposite sides of the channel. At the outermost reaches of the cyclooxygenase active site near the opening of the hydrophobic channel, the arginino group of Argl2O serves as a counter

A25 LUMEN

T T

3r

t LI

CYTOPLASM

FIG.5 . Relationship between cytosolic 85-kDa PLAz and lumenal ovine PGHS-1 in the endoplasmic reticulum. Arachidonate is cleaved from phospholipids having polar head groups on the cytoplasmic side of the ER and then traverses one plane of the bilayer through the membrane-binding domain (MBD) helices and into the hydrophobic channel comprising the cyclooxygenase active site.

176

WILLIAM L. SMITH AND DAVID L. DEWI'lT

FIG.6. Model of the cyclooxygenase and peroxidase active sites of the ovine PGHS-1. From Smith and DeWitt (1995) with permission.

ion for the carboxylate group of fatty acid substrates and at least certain NSAIDs (e.g., flurbiprofen and ibuprofen; Picot et al., 1994; Bhattacharyya et al., 1996) (Fig, 6). The roles of the various amino acids found in the active site and involved in catalysis or inhibitor action (e.g., Tyr385 and Ser530) are described in a subsequent section.

H. PEROXIDASE ACTIVESITE As discussed in more detail below, both PGHS-1 (Smith and Marnett, 1994) and PGHS-2 (Fletcher et al., 1992; Meade et al., 1993) have very active peroxidase activities and cycle through typical heme peroxidase spectral intermediates (Smith and Marnett, 1994). Indeed, the crystal structure of ovine PGHS-1 is quite similar to the structures of myeloperoxidase and cytochrome c peroxidase (Picot et al., 1994), suggesting that PGHS isozymes evolved from an ancestral peroxidase which acquired domains involved in MBD and dimerization (EGF domain). The peroxidase active site of PGHS-1 lies near the surface of the protein on the opposite side of the heme group from the cyclooxygenase active site (Figs. 4 and 6). In the crystal structure of the flurbiprofen/PGHS-1 complex, the distal position of the heme group is bordered by Gln203 and His207, although neither residue formally binds the heme (Picot et al., 1994). Gln203 and His207 are shielded from solvent by residues 289-295, which partially enclose the peroxidase active site. PGHS-1 shows at least

CYCLOOXYGENASE ISOZYMES

177

some specificity toward alkyl hydroperoxides (Ohki et nl., 1979), but it is not clear which amino acid residues govern this specificity. Sequences corresponding to residues 289-295 and to Gln203 and His207 of ovine PGHS-1 are conserved in all PGHSs. Thus, it is presumed that the structures of the peroxidase active sites of PGHS-1 and PGHS-2 are quite similar. I. C-TERMINAL P/STEL SEQUENCES All of the known PGHSs have a PTEL, STEL, or SAEL sequence at their C termini (Figs. 2 and 3 ) . This sequence is similar to the KDEL sequence that functions in the retention of some resident ER proteins in the lumen of the ER (Pelham, 1990; Tang et nl., 1992; Nothwehr and Stevens, 1994). Because PGHS-1 and PGHS-2 are found on the lumen side of the ER (Otto and Smith, 1994), it was anticipated that the P/STEL sequences would serve as ER retention signals for these enzymes. The first studies of this topic indicated that modification or deletion of this sequence yielded somewhat less active mutant enzymes, but surprisingly showed that there was no effect on subcellular targeting (Regier et al., 1995; Ren et al., 1995).These latter experiments were performed with transfected cos cells in which ovine PGHS-1 and C-terminal mutants of PGHS-1 were overexpressed for 40 hr. Recent studies, in which shorter expression times (e.g., 18 hr) were used, indicate that ovine PGHS-1 from which the -PTEL region has been deleted becomes concentrated in the Golgi (I. Song and W. L. Smith, unpublished results); this is the result expected if PTEL indeed serves as an ER targeting signal for PGHS-1 and PGHS-2 (Munro and Pelham, 1987). J. C-TERMINAL AMINO ACID INSERT OF PGHS-2

PGHS-2 contains an 18-amino-acid cassette near the C terminus that is absent from PGHS-1 (Fig. 2). We prepared a deletion mutant of murine

PGHS-2 lacking the 18-amino-acid cassette (M. A. Reger, D. L. DeWitt, and W. L. Smith, unpublished results). Membranes from the transfected cos- 1cells expressing this mutant exhibited cyclooxygenase activity comparable to that obtained with native PGHS-2; moreover, the K,, values for the native PGHS-2 and the deletion mutant were the same (3.2 and 2.5 p M , respectively). Although it was anticipated that the deletion mutant might have a subcellular distribution different from that of native PGHS2, preliminary results have indicated that deletion of the 18-amino-acid cassette does not affect targeting. Turnover of PGHS-2 is more rapid than that of PGHS-1 (DeWitt and Meade, 1993; Evett et al., 1993), and it is possible that the 18-amino-acid cassette of PGHS-2 influences the rate of enzyme degradation.

178

WILLIAM L. SMITH AND DAVID L. DEWITT

K. ANTIGENICSITES A number of monoclonal antibodies have been prepared that are reactive with PGHS-1 but not with PGHS-2 (DeWitt et al., 1981; Holtzman et al., 1992). Anti-peptide antibodies selective for PGHS-1 or PGHS-2 and reactive with the native enzymes have been prepared (Habib et al., 1993; Kujubu et al., 1993; Otto et al., 1993; Kargman et al., 1994; Otto and Smith, 1994). Table I provides examples of isozyme-selective antibodies prepared in our laboratories. There are at least four regions of the protein against which antibodies reactive with native enzyme can be prepared; these include the N terminus, the trypsin cleavage region, residues immediately downstream of the N-glycosylation site at Asn410, and near the C terminus. Antibodies selective for PCHS-2 have been prepared using peptides corresponding to the 18-amino-acidinsert unique to this isozyme (Habib et al., 1993; Kujubu et al., 1993; Otto et al., 1993; Kargman et al., 1994; Otto and Smith, 1994). 111. Catalysis by PGHS-1 and PGHS-2

A. CYCLOOXYGENASE CATALYSIS 1. Positioning of Fatty Acid Substrate in the Cycloorygenase Active Site As depicted in Fig. 7 (Smith and Marnett, 1994), arachidonate becomes oriented in the hydrophobic channel comprising the cyclooxygenase active site with a kink in the carbon chain between C-9 and C-10. Abstraction of the 13-pro-S hydrogen and subsequent isomerization leads to a carboncentered radical at C-11. Attack of molecular oxygen occurs at C-11 from the side opposite that of hydrogen abstraction. The resulting 11hydroperoxyl radical adds to the double bond at C-9, leading to intramolecular rearrangement and formation of another carbon-centered radical at C-15. This radical then reacts with another molecule of oxygen.

2. Tyr385 and Tyrosyl Radical Formation The initial step in the cyclooxygenase reaction is the stereospecific removal of the 13-pro-S hydrogen from arachidonate (Hamberg and Samuelsson, 1967). This hydrogen is abstracted as a hydrogen atom (Mason et al., 1980; Schreiber et al., 1986; Kwok et al., 1987; Tsai et al., 1995) generating an arachidonyl radical. This finding indicates a requirement for a protein radical species as the active agent in catalysis. Although the issue is somewhat controversial, most of the evidence suggests that the active principle is a tyrosyl radical formed from Tyr385. The observations that support this concept are as follows: (a) alkyl hydroperoxides are required for cyclooxygenase activity (Smith and Lands, 1972; Hemler et al., 1978; Hemler and

TABLE I CHARACTERISTICS OF PEPTIDE-DIRECTED ANTIBODIES AGAINSTOVINEPCHS-1 Peptide

Source

Domain

=ADPGAPAVNPC% "72LMHWRGIPPQmC CmPDPRQEDRPGVE594 Z"LMRWPGVPPERQMAmC C41zNTSMLVDYGVEALVDAFS4s Cy5h4SHSRLDDINFlVLIK5a C5A3QDPQDTKTATIN594

PGHS,,-1 PGHS,-1 PGHS,-l PGHS,,-l PGHS-1 PGHS,.B PGHS,,-&

Amino terminus Arg277 trypsin cleavage site Carboxyl terminus Analogous to ovine 272-283 As11410 dycosylation site Carboxyl terminus Upstream of 18 aa cassette

AND

MURINEPGHS-2"

Western Blot

Immunofluoroscent Staining

____

+ Ovine-1, human-1

+ Ovine-1, human-1 + Ovine-l + Murine-1 + All PCHS-1 species + Human-2, murine-2 -

Murine-1

Enzymes tested included ovine, murine, and human PGHS-1 and murine and human PGHS-2. Positive staining,

+ Ovine-1, human-1

+ Ovine-1, human-1 + Ovine-l + Murine-1 + All PGHS-1

+ Human-2, murine-2 -

Murine-1

+; negative staining, -

180

WILLIAM L. SMITII AND DAVID L. DEWITT

0

10

-

\O

1

COOH

-H .

+O 2

COOH

FIG.7 . Mechanism of the cyclooxygenase reaction. From Smith and Mamett (1994) with permission.

Lands, 1980) and incubation of hydroperoxides with ovine PGHS-1 leads to the sequential formation of two different visible heme intermediates (intermediates I and I1 in Fig. 8) with characteristics closely resembling those of peroxidase compound I and compound ES intermediates, respectively (Lambeir et al., 1985; Kulmacz et al., 1987; Karthein et al., 1988); (b) protein tyrosyl radicals are formed rapidly in up to 80% yield, based on the heme content of ovine PGHS-1 (Karthein et al., 1988; Kulmacz et al., 1990), and tyrosyl radical formation is coincident with formation of intermediate I1 (Dietz et al., 1988; Karthein et al., 1988); and (c) Tyr385 is essential for enzyme activity (Shimokawa et al., 1990) and is found neighboring the heme group and in a position expected to neighbor the 13-pro-S hydrogen of arachidonate bound to the cycloolojgenase activity site in the crystal structure of ovine PGHS-1 (Picot et al., 1994). A model proposed by Ruf and co-workers (Dietz et al., 1988) integrates the various mechanistic and structural information related to the tyrosyl radical species (Fig. 8). According to the model, a two-electron oxidation

181

CYCLOOXYGENASE ISOZYMES

CYCLOOXYGENASE

2oL

[PPIX-Fe4+=O] Tyr

[PPIX-Fe4+=O] Tyr

, "

/e* [PPIX-Fe4+=Ol

z

\

e-

PEROXIDASE

FIG.8, Ruf model for peroxide-dependent activation of the cyclooxygenase activity of PGH synthase-1 by formlttion of an intermediate tyrosyl radical. Intermediates I and I1 are designated. PPIX-Fe3+,heme; AA, arachidonic acid. From Smith and Marnett (1994) with permission.

of the heme group of PGHS-1 by a hydroperoxide causes formation of an intermediate I having a two-electron oxidized heme in which the iron is in the 4+ state and the porphyrin group is oxidized to a radical cation. Intermediate I can abstract a hydrogen from the phenolic side chain of a protein tyrosine residue to produce an intermediate I1 containing a neutral porphyrin group, an iron in its 4+ state, and a protein tyrosyl radical. The tyrosyl radical is the species that abstracts the 13-pro-S hydrogen from arachidonate thereby initiating cyclooxygenase catalysis. Studies suggesting that incubation of PGHS-1 with arachidonate and hydroperoxide anaerobically leads to the sequential formation of tyrosyl and arachidonate radicals provide additional evidence for the importance of a tyrosyl radical in catalysis (Tsai et al., 1992, 1995). Also consistent with this model is the finding that the cyclooxygenaseactivity, once initiated by a hydroperoxide, can cycle in the absence of active peroxidase turnover (Wei et al., 1995).This indicates that the cyclooxygenase reaction involves a nonheme rddicd species not involved in peroxidase turnover (Wei et al., 1995). Finally, PGHS-2, as well as PGHS-1, has been found to form a tyrosyl radical in the presence of a peroxide (Hsi et al., 1994). In fact, the only important piece of evidence that has yet to be reconciled with the Ruf model is the finding that a H386A mutant of ovine PGHS-1, which catalyzes a cyclooxygenase reaction but lacks peroxidase activity, does not

182

WILLIAM L. SMITH AND DAVID L. DEWI’lT

form a detectable tyrosyl radical species (Shimokawa and Smith, 1991; Tsai et al., 1994a). An important feature of this mutant enzyme and the corresponding H372A PGHS-2 analog (L. Hsi and W. L. Smith, unpublished results) is that both of these enzymes fail to undergo reactionbased “suicide” inactivation typical of native PGHS-1 and PGHS-2. This observation raises the possibility that the tyrosyl radical species is involved in suicide inactivation. 3. Cyclooxygenase Substrates The two best fatty acid substrates for the cyclooxygenase activities of both PGHS-1 and PGHS-2 are arachidonic acid and dihomo-y-linolenic acid (Table 11). The K , values for both enzymes with arachidonate are about 5 p M (Meade et al., 1993; Barnett et al., 1994; Laneuville et al., 1994, 1995). PGHS-1 and PGHS-2 are also capable of catalyzing the oxygenation of 5,8,11,14,17-eicosapentaenoic acid (EPA), y-linolenic acid, a-linolenic acid, and linoleic acid (Kulmacz et al., 1994; Percival et al., 1994; Laneuville et al., 1995). EPA is converted to PGH3, whereas the 18carbon fatty acids are converted to monohydroxy acids (Laneuville et al., 1995). In general, 18-carbon polyunsaturated fatty acids are more efficiently oxygenated by PGHS-2 than by PGHS-1. For example, the ratio of the kc,&,, value for PGHS-2 with a a-linolenic acid is about 40 times that for PGHS-1 (Table TI). Although 03 and 09 polyunsaturated fatty acids containing 18-22 carbons are poor substrates for PGHS-1, they are efficient competitive inhibitors of the oxygenation of arachidonic acid by PGHS-1 (Lands et al., 1973). Docosahexaenoic acid (22 : 6 w3) is a competitive inhibitor of both PGHS-1 and PGHS-2 without being a substrate for either enzyme (Marshall and Kulmacz, 1988; Meade et al., 1993). Overall, the studies on the substrate specificities of PGHS-1 and PGHS-2 suggest that PGHS-2 has a somewhat larger, more accommodating cyclooxygenase active site (Laneuville et al., 1995). As discussed in the following section, this concept is consistent with the results of numerous studies on the interactions of NSAIDs with the two isozymes. 4. Nitric Oxide as an Activator of Cyclooxygenase Activity Factors that modulate the expression of the inducible form of nitric oxide synthase also regulate the expression of PGHS-2 (Marletta, 1994; Nathan and Xie, 1994). Interestingly, nitric oxide itself has been reported to activate both PGHS-1 and PGHS-2 both in vitro (Salvemini et al., 1993) and in animal models (Salvemini et al., 1994). However, the mechanism for this activation is not understood. Although nitric oxide does bind the Fe3+of the heme group of PGHS-1, it does not appear to activate the purified enzyme (Tsai et al., 1994b).

TABLE I1 CATALYTIC CONSTANTS FOR HUMAN PGHS-1 A N D -2 USINGC 2 0C ~1 8~F ~ ACIDS” m

hPGHS-1 Fatty Acid Arachidonic acid (20:4, n-6) Eicosatrienoic acid (20:3, n-6) Linoleic acid (18:2, n-6) a-Linolenic acid (18:3. n-3) a

& (pM) 5.4

Relative V,,

105

(%)

hPGHS-2 Approximate kJK,,, 19

(pM)

5.6

Relative V,

106

(%)

Approximate kc,/& 19

14

74

5.3

28

12

11

0.92

27

76.2

2.8

48

74

1.5

200

From Laneuville et al. (1995) with permission

7.6

.038

128

7.1

184

WILLIAM L. SMITH A N D DAVID L. DEWITI

5. Inhibition of PGHS-1 and PGHS-2 by Nonsteroidal Anti-inflammatory Drugs In this section, we will provide a general overview of the biochemcial pharmacology of NSAIDs and a description of inhibitors selective for either PGHS-1 or PGHS-2. In a later section, we discuss the actions of these drugs in the context of the roles of PGHS-1 and PGHS-2 in various pathop hysiological states. The cyclooxygenase activities of both PGHS-1 and PGHS-2 can be inhibited by most NSAIDs, which compete with arachidonate for binding to the cyclooxygenase active site (Fig. 6); in general, NSAIDs do not affect peroxidase activity. There are significant functional differences among these drugs in the manner with which they interact with the cyclooxygenase site after binding to this site. Accordingly, it is convenient to group NSAIDs that inhibit cyclooxygenase activity into classes 1-111 (Fig. 9 and Table 111) (Smith and DeWitt, 1995).Class I inhibitors (e.g., ibuprofen, mefenamic acid, flufenamic acid, piroxicam, sulindac sulfide, naproxen, and 6-methoxynaphthyl-%acetic acid) are simple competitive cyclooxygenase inhibitors that rapidly but reversibly form an EI complex only (i.e.,k, = 0; Fig. 9). Class I1 inhibitors (e.g., indomethacin, meclofenamic acid, flurbiprofen, and diclofenac) are competitive, time-dependent, reversible inhibitors. They form an EI complex but then secondarily change the structure of the protein to produce an EI" complex without convalently modifymg the

CLASS I:

COMPETITIVE, REVERSIBLE (IBUPROFEN)

k2 = O

CLASS 11: COMPETITIVE, TIME-DEPENDENT, SLOWLY REVERSIBLE (INDOMETHACIN) k;! > O

k.2>0

CLASS 111: COMPETITIVE, IRREVERSIBLE (ASPIRIN) kZ>O

k.2~0

FIG. 9. Interaction of PGHSs with different classes of nonsteroidal antiinflammatory drugs (NSAIDs). Class I inhibitors form EI complexes reversibly. Class 11 inhibitors form EI complexes that rearrange to form El" complexes; I can dissociate from EI" but dissociation is normally quite slow. Class 111 inhibitors form EI" complexes via covalent modification of the protein and k 2= 0.

CYCLOOXYGENAS E I SOZYME S

185

protein (Stanford et al., 1977) (Fig. 9); EI" complex formation is relatively slow-occurring in seconds to minutes; EI" slowly reverts to EI. Class I11 inhibitors (aspirin, propinonylsalicylate, and valerylsalicylate) convert EI to an EI" complex by covalent modification (acylation)of the protein; with class I11 inhibitors, once an EI" complex is formed, it is not possible for the protein to revert to EI. Thus, a cell treated with a class I11 inhibitor must synthesize new PGHS to regain cyclooxygenase activity. Empirically, the relative affinities of NSAIDs for PGHS-1 and PGHS2 can be estimated by measuring "instantaneous" inhibition in vitro using an O2 electrode assay (Laneuville et al., 1994; Smith and DeWitt, 1995). Arachidonate (and 0,) and inhibitor are added to an assay chamber and the rate of O2 consumption is monitored immediately on the addition of enzyme. Because there is no preincubation of inhibitor with enzyme and because EI" complex formation is slow (i.e., k2 4 k,; Fig. 9), one can circumvent the issue of time-dependent inhibition using this protocol. Furthermore, because the Ic, values of PGHS-1 and PGHS-2 for arachidonate are the same (Meade et al., 1993; Barnett et al., 1994; Laneuville et al., 1994), ICSOvalues obtained for each isozyme with an NSAID can be compared directly and provide an estimate of the relative affinity of an NSAID for each isozyme. To measure "time-dependent" inhibition, PGHS is preincubated with inhibitor for various times and then added to an assay chamber containing substrate (Copeland et al., 1994; Laneuville et al., 1994; Smith and DeWitt, 1995). With class I1 and class I11 NSAIDs, a gradual diminution of activity is observed with time of preincubation. This inhibition occurs with first-order kinetics and thus a plot of the logarithm of activity versus time yields a t4 that can be used to determine k, (Fig. 9). Table I11 provides a summary of IC9"values obtained for instantaneous inhibition of human PGHS-1 and PGHS-2 with a variety of NSAIDs and also indicates the class into which different inhibitors fall. In general, common NSAIDs have higher affinities for PGHS-1 than for PGHS-2. This is consistent with determinations of substrate specificitiesin suggesting that the active site of PGHS-2 is somewhat larger and more accommodating than that of PGHS-1. The relationship between potencies in cyclooxygenase inhibition and nonsteroidal anti-inflammatory activities are discussed in the following section. 6. Class I 1 NSAIDs

The biochemical basis for the ability of a class I1 NSAID to cause formation of a semistable EI" complex is not understood. When compared to closely related class I inhibitors, class I1 are seen to have halogenated phenyl rings, and thus possess intrinsic dipoles within their aromatic substituents (Rome and Lands, 1975; Laneuville et al., 1994). Furthermore, it is

186

WILLIAM L. SMITH AND DAVID L. DEWI'IT

TABLE I11 INHIBITIONOF HUMAN PGHS-1 AND PGHS-2 NSAID

Class

Indomethacin Sulindac sulfide Piroxicam Diclofenac Flurbiprofen Meclofenamate Phenylbutazone Naproxen Ibuprofen Ketorolac tromethamine DHA (22:6) 6-MNA Etodolac Salicylic acid

I1 I I I1 I1 I1 ? I I ? I I ? -

BY

NSAIDs"

ICm for PGHS-1 (/AM) ICx for PGHS-2 ( p M )

13.5 2 3.5 1.3 5 1.0 17.7 t 3.3 2.7 5 1.0 0.5 t 0.1 1.5 2 0.6 16.0 4.8 2 1.8 4.0 2 1.0 31.5 5 9.2 25.6 ? 6.6 64.0 t 8.5 74.4 2 7.6 >1000.0

>1000.0 50.7 5 6.6 >500.0 20.5 t 6.4 3.2 2 1.6 9.7 2 1.8 >100.0 28.4 t 7.6 12.5 ? 2.1 60.5 t 14.8 41.0 2 15.6 93.5 2 23.3 60.0 t 8.4 >1000.0

"Values are for instantaneous inhibition of cyclooxygenase activity of hPGHS-1 and hPGHS-2 (Laneuville et al., 1994). NSAID class is indicated if known.

clear that both a free carboxyl group and binding of this carboxyl group to Argl20 of PGHS-1 is required because (a)conversion of class I1 NSAIDs to their corresponding methyl esters converts them to less potent class I inhibitors (Rome and Lands, 1975) and (b) a PGHS-1 mutant in which a neutral glutamine is substituted for Argl2O does not undergo timedependent inhibition in the presence of class I1 inhibitors (D. K. Bhattacharrya, C. J. Reike, and W. L. Smith, unpublished results). Indomethacin forms an EI" complex wtih 0.5 mol of indomethacin bound per dimer (Kulmacz and Lands, 1985), suggesting that the binding of this agent to one molecule of monomer renders the entire dimer inactive. Not surprisingly, there are significant differences in values for both k2 (Rome and Lands, 1975; Copeland et al., 1994; Laneuville et al., 1994) and k - , (Rome and Lands, 1975; Walenga et al., 1986; Copeland et al., 1994; Laneuville et al., 1994) for different inhibitors for a given isozyme; for example, for human PGHS-1, k-2 values decrease in the order of flurbiprofen meclofenamate S indomethacin. ,Furthermore and importantly, some agents such as DuP697 (Copeland et al., 1994) cause timedependent inhibition of PGHS-2 but not PGHS-1. Indeed, this appears also to be the case for at least two other compounds that show selectivity for PGHS-2-NS398 (Futaki et al., 1994) and SC58125 (Masferrer et al., 1994; Munroe and Lau, 1995). In contrast, all inhibitors that are class I1 inhibitors of PGHS-1 (e.g., indomethacin, flurbiprofen, diclofenac, and meclofenamate) are also time-dependent inhibitors of PGHS-2.

-

CYCLOOXYGENASE ISOZYMES

187

7. Aspirin and Other Acyl Sakicylates Aspirin is the only NSAID that covalently modifies PGHS-1 and PGHS2. Aspirin binds to the cyclooxygenase active site of PGHS with a very low affinity (K-20mM) (DeWitt et al., 1990; LaneuviUe et al., 1994).However, on binding to the cyclooxygenase active site, aspirin transfers its acetyl group from salicylate to a specific “active site” serine residue (Fig. 6) (Roth et al., 1983). In ovine PGHS-1, this aspirin-acetylated serine is Ser530 (DeWitt et al., 1990);the homologous serine in human PGHS-2 (Ser 516) is also acetylated (Lecomte et al., 1994; Mancini et al., 1994). However, the effects of aspirin on the activities of PGHS-1 and PGHS-2 are subtly different. Acetylation of PGHS- 1by aspirin completely inhibits the cyclooxygenase activity but does not affect peroxidase activity (Fig. 6) (van der Ouderaa et ak., 1980; Mizuno et al., 1982). Acetylation of PGHS-2 by low concentrations of aspirin (ca. 0.5 mM for 30 min at 37°C) converts this isozyme to a form that still oxygenates arachidonate, but at C-15 instead of C- 11 yielding 15R-hydroxyeicosatetraenoicacid (15R-HETE) instead of PGGz (Fig. 6) (Holtzman et al., 1992; Meade et al., 1993; Lecomte et al., 1994; O’Neill et al., 1994). It should be noted that the so-called active site serine residues of PGHS1 (Shimokawa and Smith, 1992) and PGHS-2 (Lecomte et al., 1994) are not required for enzyme activities or arachidonate binding. Indeed, replacement of these residues with alanines yield mutant enzymes that are catalpcally active and have K,, values with arachidonate that are very close to those of the native enzyme. Aspirin acetylation of Ser530 appears to interfere sterically with arachidonate binding (DeWitt et al., 1990; Lecomte et al., 1994; Loll et al., 1995). Various acyl salicylates other than aspirin are also cyclooxygenase inhibitors (Bhattacharyya et al., 1995). Propionylsalicylate is about equally effective with PGHS-1 and PGHS-2. Valeryl (pentanoyl) salicylate, for reasons that are still not clear, is a relatively selective inhibitor of PGHS-1. Like aspirin, valerylsalicylate acylates the active site serine of PGHS-1 (Bhattacharyya et al., 1995). B. PEROXIDASE CATALYSIS

The peroxidase activity of PGHS catalyzes the two-electron reduction of PGGe to PGHz with concomitant oxidation of electron donors (Figs. 1 and 8). PGHS-1 contains one heme per subunit. During the peroxidase reaction, the heme group of PGHS initially undergoes a two-electron oxidation to a ferryl-oxo intermediate, which is reconverted to the native heme by two sequential one-electron reductions (Fig. 8) (Smith and Marnett, 1994). Peroxidase catalysis by PGHS-1 and PGHS-2 occurs by

188

WILLIAM L. SMITH AND DAVID L. DEWI?T

the same mechanism as that of other heme peroxidases (Smith and Marnett, 1994). In fact, as noted earlier, the crystal structure of PGHS-1 is remarkably similar to that of myeloperoxidase (Picot et al., 1994). Interestingly, the heme group of PGHS-1 (and presumably PGHS-2) can be replaced with M$+-protoporphyrin IX. The mangano-heme PGHS1has less than 5%of the peroxidase activity of the native enzyme but about 40% of the native cyclooxygenase activity (Ogino et al., 1978; Odenwaller et al., 1992; Strieder et al., 1992). The peroxidase of PGHS can function independent of the cyclooxygenase activity, and there is a convenient spectral assay for PGHS peroxidase activity employing HzOz and guaiacol as oxidizing and reducing cosubstrates, respectively (Marnett et al., 1988). Purified PGHS-1 is capable of reducing exogenous hydroperoxides other than PGG2. The PGHS-1 peroxidase is particularly active with primary and secondary alkyl hydroperoxides, less active with HzO2,and inactive with tertiary hydroperoxides (Ogino et al., 1978; Kulmacz and Lands, 1983; Markey et al., 1987). Furthermore, a variety of compounds can serve as reducing cosubstrates for PGHS-1 peroxidase (Ogino et al., 1978; Markey et al., 1987). The identity of the physiological reducing cosubstrate( s) is not known. Possible reducing cosubstrates in vivo include epinephrine, uric acid, and reduced glutathione (Smith and Marnett, 1994).The peroxidase activity of PGHS-1 can catalyze peroxidatic cooxidations of xenobiotics, such as aromatic amines, polycyclic hydrocarbons, nitrofurans, mycotoxins, synthetic estrogens, bisulfite, phenols, heterocyclic amines, hydantoins, and indoles, leading to intermediates that undergo secondary chemical reactions depending on their solution chemistry (Eling et al., 1990; Marnett and Maddipati, 1991). These processes have been suggested to be important in xenobiotic metabolism in tissues having low levels of cytochrome P450. PGHS-2 exhibits a peroxidase activity that is quantitatively similar to that of PGHS-1 (Fletcher et al., 1992; Meade et al., 1993; Barnett et al., 1994; Laneuville et al., 1994). The substrate specificities of this isozyme have not been examined. However, PGHS-2 may bind hydroperoxides with a 5- to 10-foldgreater affinitythan PGHS-1 because the concentrations of hydroperoxides required to activate the cyclooxygenase activity of PGHS2 are lower than those required to activate PGHS-1 (Capdevila et al., 1995; Kulmacz and Wang, 1995).

IV. Regulation of Expression of the Genes for PGHS-1 and PGHS-2

The most obvious distinction between PGHS-1 and PGHS-2 involves their differential expression. PGHS-1 is expressed constitutively in most

CYCLOOXYGENASE ISOZYMES

189

tissues, whereas PGHS-2 is synthesized and expressed only transiently in selected cells and tissues and only following stimulation by cytokines, growth factors, hormones, or tumor promoters.

A. REGULATION OF PGHS-1 GENEEXPRESSION Although PGHS-1 is present constitutively in most cells and tissues, the relatively high levels of this enzyme in some highly differentiated and specialized cell types, such as endothelium, macrophages, and renal collecting tubules (DeWitt et al., 1983; Smith, 1986), suggests that PGHS-1 expression is developmentally controlled. This view has been borne out by quantitation of PGHS-1 in developing ovine vasculature that indicate that the level of this isoenzyme increases during the 4 weeks immediately following birth (Brannon et al., 1994). A dramatic demonstration of developmental regulation of PGHS-1 expression has also been observed in ovine seminal vesicles (Silvia et al., 1994). Following castration, the seminal vesicles of young weathers do not fully develop, remaining about one-third the size of those from agematched rams. Seminal vesicles from castrated rams express low levels of PGHS-1 protein; however, when these animals are treated with testosterone, the seminal vesicles do develop and elevated levels of PGHS-1 in seminal vesicle tissue are observed. These findings suggest that PGHS-1 expression in ovine seminal vesicles increases during puberty in response to increasing levels of androgens (Silvia et al., 1994). Experiments with several model cell lines also support the concept of developmental regulation of PGHS-1 expression. In two promonocytic cell lines, THP-1 and U937, PGHS-1 expression increases during experimentally induced differentiation to a macrophage phenotype (Hoff et al., 1993; Smith et al., 1993). Stem cell factor, a cytokine essential for mast cell maturation, stimulates the synthesis of PGHS-1 in bone marrow-derived mast cells (Samet et al., 1995), as does the combination of kit ligand and IL-10 (Murakami et al., 1994). In each of these studies, it should be noted that expression of PGHS-1 is sustained and likely becomes a permanent characteristic of the differentiated cell. As discussed in the following section, this aspect of the regulation of PGHS-1 gene expression contrasts with that observed for PGHS-2, a protein that is usually expressed only transiently. Although it seems likely that PGHS-1 expression is regulated developmentally, it does not appear that PGHS-1 is required for mammalian development because knockout mice, which do not express PGHS-l, mature and reproduce normally (Langenbach et al., 1995).

B. REGULATION OF PGHS-2 GENEEXPRESSION PGHS-2 protein is only expressed in a few restricted tissues in the absence of inflammatory or other stimuli. The most notable exception is

190

WILLIAM L. SMITH A N D DAVID L. DEWIlT

the brain (Simmons et al., 1991; O’Neill and Ford-Hutchinson, 1993; Kargman et al., 1994) in which PGHS-2 is expressed in dendrites and cell bodies of neurons, primarily in discreet regions of the cortex, hippocampus, and amygdala (Breder et al., 1995). PGHS-2 expression can also be induced by synaptic stimulation, suggesting that prostaglandins produced by this isozyme may also be involved in neuronal plasticity (Yamagata et al., 1993). The full significance of expression of PGHS-2 in brain is not yet understood. Another resting tissue in which PGHS-2 is expressed is the kidney. Immunochemical localization has demonstrated that PGHS-2 is constitutively expressed in the macula densa of the juxtaglomerular apparatus and in adjacent epithelial cells of the cortical thick ascending limb (Harris et al., 1994). In salt-restricted animals, PGHS-2 levels increase in the macula densa up to three-fold (Harris et al., 1994). Prostanoids produced in the collecting tubule by PGHS-1 have long been known to be important regulators of water and sodium resorption (Smith, 1992). It now seems likely that regulation of expression of PGHS-2, and subsequent prostaglandins formation by this enzyme in the JG apparatus, may also be important in regulating renal blood flow and salt and volume homeostasis. PGHS-2, unlike PGHS-1, is required for proper mammalian development, and transgenic mice that do not express this gene develop nephropathy severe enough to limit their life span (Morham et al., 1995). PGHS-2 expression also occurs in transformed or cancerous cells. The first PGHS-2 cDNA described, that for chicken PGHS-2, was initially characterized on the basis of its elevated expression in Rous sarcoma virustransformed chicken embryo fibroblasts (Simmons et al., 1989). Constitutive overexpression of PGHS-2 has been observed in colon carcinomas (Eberhart et al., 1994; Kargman et al., 1995) and in an experimentally induced mouse epidermal papillomas and carcinomas (Muller-Decker et al., 1995). It is not known whether sustained expression of PGHS-2 is a common feature of transformed cells. However, the observation that regular aspirin ingestion can reduce the occurrence of colon cancer (Thun et al., 1991) has stimulated research in this area. Regulation of PGHS-2 expression has been investigated more extensively than that of PGHS-1 in part because the high level of inducibility of PGHS2 results in easily measurable changes in protein and mRNA levels. Most commonly, PGHS-2 is expressed followingmitogenic or inflammatorystimulation. Western blotting experiments with PGHS-2-specific antiserum are unable to detect PGHS-2 protein in unstimulated rat tissue (Kargman et al., 1994), but PGHS-2 is rapidly induced following an injection of carrageenan into rat paws (Seibert et al., 1994). PGHS-2 is not found in normal joints but is found in articular tissue during staphylococcal cell wall- or adjuvant-induced arthritis in rat (Sano et al., 1992) and in joints of humans

CYCLOOXYGENASE ISOmMES

191

with rheumatoid arthritis (Sane et al., 1992;Crofford et al., 1994; Hulkower et al., 1994). Regulated expression of PGHS-2 is also involved in several stages of mammalian reproduction. The best characterized example is the gonadotrophin-stimulated induction of PGHS-2 in rat and bovine follicles preceding ovulation. Treatment of rats with ovulatory doses of chorionic gonadotrophin leads to a rapid but transient induction of PGHS-2 in granulosa cells (Sirois et aZ., 1992); a similar induction is seen in bovine follicles (Sirois, 1994). Induction of PGHS-2 appears to be a prerequisite for ovulation, at least in rodents. PGHS-2 is also induced at sites of implantation in differentiating uterine stromal cells during blastocyst attachment in mice, suggesting that PGHS-2 participates in the cell growth and neovascularization that accompany this process (Jacobs et al., 1994). Increased prostaglandin synthesis mediated by PGHS-2 may also be required for parturition. PGHS-2 levels from amnion of women undergoing spontaneous labor are twice that of women who deliver by selective cesarian section (Hirst et al., 1995). Increased PGHS-2 expression in amnion correlates well with the observed increase in prostaglandin production, suggesting that prostanoid synthesis at parturition may be entirely through PGHS-2. In vitro studies of PGHS-2 expression have been conducted in a number of cell types including fibroblasts, endothelial cells, purified monocytel macrophages and macrophage-like cell lines, epithelial cells (DuBois et al., 1994; Muller-Decker et d., 1995),mesothelial cells (Topley et al., 1994), osteoblasts (Pilbeam et d., 1993; Kawaguchi et al., 1994), mesangial cells (Kester et nl., 1994; Rzymkiewicz et al., 1994; Stoebel and Goppelt-Struebe, 1994),and mast cells (Murakami et al., 1994). In fibroblasts and endothelial cells, PGHS-2 can be induced rapidly (1-3 hr) and dramatically (20- to 80-fold) by growth factors, phorbol esters, and IL-lP (Han et al., 1990; Kujubu and Herschman, 1992; DeWitt and Meade, 1993; Evett et al., 1993; Jones et al., 1993; Kujubu et al., 1993; Pilbeam et al., 1993). Lipopolysaccharide, IL-1, and TNF-a stimulate PGHS-2 expression ex vivo in monocytes and macrophages (Lee et al., 1992; G. M. O’Sullivan et al., 1992; M. G. O’Sullivan et al., 1992; Riese et al., 1994).Several unique features differentiate the expression of PGHS-2 in monocytelmacrophages from that of other cells and tissues. First, the time required for induction of PGHS-2 is much longer-8-12 hr in monocytes/macrophages versus 2-4 hr in most other systems. This may be explained by the observation that IL-1 antagonists inhibit LPS-stimulated PGHS-2 expression in monocytes (Glaser and Lock, 1995). One possibility is that PGHS-2 expression is delayed because it is secondary to LPS-stimulated synthesis of IL-1 and resulting autocrine stimulation by this cytohne. In addition, PGHS-2 mRNA and protein levels to remain elevated in monocyteslmacrophages.

192

WILLIAM L. SMlTH AND DAVID L. DEWITI

This may reflect sustained autocrine self-activation of these cells or may be a feature of the terminally differentiated state of macrophages. One distinguishing characteristic of PGHS-2 not shared by PGHS-1 is that PGHS-2 expression can be completely inhibited by anti-inflammatory glucocorticoids such as dexamethasone (Kujubu and Herschman, 1992; Lee et al., 1992; O'Banion et al., 1992; Sano et al., 1992; DeWitt and Meade, 1993; Evett et al., 1993). Similarly, anti-inflammatory cytokines, such as IL-10, can also selectively inhibit PGHS-2 induction (Mertz et al., 1994). These observations provide additional evidence that PGHS-2 produces prostaglandins involved in inflammation. C. PGHS GENESTRUCTURES PGHS-1 and PGHS-2 are encoded by separate genes located on different chromosomes. PGHS-1 is on human chromosome 9 (Funk et al., 1991), whereas PGHS-2 is located on human chromosome 1 (Jones et al., 1993; Kosaka et al., 1994). A schematic of these genes is presented in Fig. 10 (Yokoyama and Tanabe, 1989; Fletcher et al., 1992; Kraemer et al., 1992; Appleby et al., 1994). The first two exons of PGHS-1, which contain the translational start site and signal peptides, are condensed to form a single exon in PGHS-2, but the remaining introdexon arrangements of the two genes are identical. The sizes of the introns are, however, very different. The PGHS-2 gene is about 8 kb, whereas the PGHS-1 gene is about 22 kb. The small size of the PGHS-2 gene is consistent with its characterization as an immediate-early gene (Herschman, 1991).

D. PROMOTER STRUCTURES OF PGHS-1 AND PGHS-2 GENES The structures of the promoters of the two PGHS genes are predictive of their mode of regulation. PGHS-1 has no TATA box (Kraemer et al., PGH SYNTHASE-1

EXON-

AB

CDE

FG

c

H

I

J

22.5kb

K Dl

PGH SYNTHASEQ:

EXON- A BCD E FGH I 8kb-

J

U

1KB

Fic. 10. Introdexon structures of the genes for murine PCHS-1 and PCHS-2.

193

CYCLOOXYGENASE ISOZYMES

1992), a promoter element commonly laclang in constitutively expressed housekeeping genes. Reporter plasmids constructed with the 5’-upstream region of the PGHS-1 gene have failed to show any significant inducible transcription from this promoter (S. A. Kraemer and D. L. DeWitt, unpublished results). These findings are consistent with the concept that changes in PGHS-1 expression occur only developmentally (Brannon et al., 1994; Silvia et al., 1994). The PGHS-2 promoter (Fig. ll),on the other hand, contains a TATA box, and experiments with reporter plasmids containing the PGHS-2 upstream 5’-flanking sequences have identified several specific regulatory DNA sequences that can activate PGHS-2 gene transcription. Transcriptional activation of the PGHS-2 gene appears to be the major mechanism for increasing PGHS-2 expression (DeWitt and Meade, 1993; Evett et al., 1993).Multiple signalingpathways have been linked to stimulation of PGHS-2 gene transcription including the protein kinase A pathway (Kujubu et al., 1991; Sirois et al., 1992, 1993),the protein kinase C pathway (phorbol esters) (Hla and Maciag, 1991; Kujubu et al., 1991; Hla and Neilson, 1992; DeWitt and Meade, 1993; Pilbeam et al., 1993; Crofford et al., 1994), viral transformation (srcc) (Evett et al., 1993) and tyrosine hnase (Kester et al., 1994) as well as phosphatase signaling pathways (Stroebel and Goppelt-Struebe, 1994). Bacterial endotoxin (LPS) (Hla and Neilson, 1992; Lee et al., 1992; G. M. O’Sullivan et al., 1992; M. G.

wkB

Mouse

I I

-401

TATA -30 ATFERE

m

CEBP

I I I I

I

I

I

I

SP1

-238 -190

-138

0

-57 TATA -27

wkB

I

Rat

I -

-404

wkB

Human

I

I ~

-448

m I I I I

SP1

-238 -194 SP1

ETs

I I

I I -270 -191

CEBP

I

I

0

-140 CEBP

I

I -133

TATA -31 AmRE

I I

0

-59

FIG. 11. Promoter regions of the PGHS-2 gene. Schematic of the promoter regions of the mouse (Fletcher et al., 1992), rat (Sirois and Richards, 1993), and human promoters (Appleby et al., 1994; Kosaka et al., 1994; Tazawa et al., 1994). Transcriptional regulatory elements and their positions relative to the transcriptional start site are indicated. Shown only are those regulatory elements demonstrated to be involved in transcriptional regulation of the gene (CEBP and CRE) or sites (NfKB, SP1, ETS) that are conserved in sequence and location between the three promoters and that have been shown to regulate transcription of other inflammatory proteins. This list is not meant to be exhaustive, nor are all the sites for the above regulatory elements necessarily indicated.

194

WILLIAM L. SMITH AND DAVID L. DEWITI

O’SuIlivan et al., 1992; Feng et al., 1993; Jones et al., 1993; Riese et al., 1994) and inflammatory cytokines (IL-1) (Feng et al., 1993; Jones et al., 1993; Crofford et al., 1994; Hulkower et al., 1994) also induce PGHS-2, but the signaling pathways utilized by these effectors are not as well understood. Although the primary structures of the 5‘-flanking regulatory regions of the human (Hla and Neilson, 1992; Kosaka et al., 1994; Tazawa et al., 1994), mouse (Fletcher et al., 1992), and rat (Sirois and Richards, 1993) PGHS-2 genes have been determined, our understanding of how multiple effector and signaling pathways converge on and regulate this single gene are sketchy (Kujubu and Henchman, 1992; Sirois and Richards, 1993; Sirois et al., 1993). The transcriptional control elements necessary for activation of the mouse PGHS-2 gene by PDGF, serum, serum and growth factor are located within the first 371 nucleotides upstream of the mouse PGHS-2 transcription start site (Fletcher et al., 1992). An NF-ILGICIEGP. regulatory element in the rat promoter centered at position -140 (Fig. 11)can mediate, at least in part, the increased PGHS-2 gene transcription in rat follicles following exposure to follicle-stimulating hormone, luteinizing hormone, or forskolin (Sirois and Richards, 1993; Sirois et al., 1993). However, only about 50% of the transcription activated by these effectors can be attributed to the NF-IL-G/CIEBP enhancer, indicating that additional DNA regulatory elements also cooperate in the CAMPresponse pathway. NF-IL-GKYEBP response elements are also found in the mouse and human promoters at almost the identical locations (Fig. 11)and likely function in an analogous manner in these species. Functional CAMP response elements (CRE) have also been identified in the mouse and human promoters about 60 nucleotides upstream of the transcription initiation sites (Fig. 11).In tnouse cells, the CRE element mediates the v-src-induced expression of PGHS-2 (Xie and Henchman, 1995). It is likely that activation of c-src or other members of this tyrosine kinase family may be one mechanism whereby growth factors, such as PDGF and CSF, stimulate PGHS-2 synthesis. Experiments with kinase inhibitors also indicate that tyrosine kinase signaling pathway may also be involved in endothelin- and serotonin-stimulated expression of PGHS-2 in rat mesangial cells (Kester et al., 1994; Stroebel and GoppeltStruebe, 1994). The human promoter CRE sequence has also been shown to regulate PGHS-2 expression in U937 cells. Differentiation of U937 to a macrophage-like phenotype leads to stable induction of PGHS-2 mRNA and protein. Transcription from reporter gene plasmids containing the human 5-flanking sequence are elevated when transfected into phorbol ester-differentiated U937 cells compared to untreated cells. Mutation of the CRE sequence in this reporter gene plasmid abolishes transcriptional

CYCLOOXYGENASE ISO%YMES

195

upregulation in the differentiated cells (Inoue et al., 1994). Phorbol estertreated U937 cells contain elevated levels of protein that bind specifically the CRE element, indicating that phorbols esters stimulate the synthesis of transcription factors in U937 cells that stimulate transcription of PGHS2. One interesting curiosity is that no CRE element is present in the reported sequence of the rat promoter (Sirois et al., 1993), even though this sequence is conserved between the mouse and human promoters. The location and identity of a number of cis-acting DNA elements that are conserved between the mouse, rat, and human PGHS-2 promoters have yet to be tested for functionality (Fig. 11) (Fletcher et al., 1992; Sirois et d., 1993; Kosaka et al., 1994; Ristimaki et al., 1994; Tazawa et al., 1994). These include NF-KB, SP1, and ETS elements. Because of the involvement of these sites and their cognate transcription factors in the regulation of numerous other proteins related to inflammation (Libermann and Baltimore, 1990; Grove and Plumb, 1993; Muller et nl., 1993), it seems likely that they may also participate in regulation of PGHS-2 expression. Although the primary level of regulation of PGHS-2 synthesis appear to be transcriptional, post-transcriptional regulation of PGHS-2 may also occur (Evett et al., 1993; Ristiinaki et al., 1994).PGHS-2 mRNA is unstable compared to PGHS-1 mRNA, a feature that probably results from the multiple RNA instability sequences (AUUUA) present in the PGHS-2 3’-UTR. Factors that increase or decrease the half-life of PGHS-2 mRNA can also presumably increase or decrease the efficiency of its turnover during translation. IL-1 appears to regulate PGHS-2 by this mechanism. In the human endothelial cell line ECV204 (Ristimaki et al., 1994), IL-1 not only increases PGHS-2 gene transcription, but also PGHS-2 mRNA stability. Dexamethasone inhibits transcription of PGHS-2 (DeWitt and Meade, 1993) and reduces PGHS-2 mRNA stability (Evett et al., 1993). PGHS-2 expression may also be regulated directly at the level of translation. Dexamethasone only partially inhibits transcription and PGHS-2 mRNA accumulation in serum-stimulated fibroblasts (DeWitt and Meade, 1993) and in IL-l-stimulated mesangial cells (Rzymkiewicz et al., 1994), but dexamethasone completely inhibits PGHS-2 protein expression. The presence of PGHS-2 mRNA in these cells with no protein synthesis suggests that dexamethasone can act directly to inhibit translation of PGHS-2 mRNA. Although it has been clearly demonstrated that dexamethasone and antiinflammatory cytohnes, such as IL-10, can reduce PGHS-2 expression by inhibiting transcription of the gene (DeWitt and Meade, 1993; Mertz et al., 1994), the biochemical mechanism is not known. The two models that have been described for the negative transcriptional regulation by steroids require either AP-1 or nGRE enhancer binding sites to be present in the

196

WILLIAM L. SMITH AND DAVID L. DEWITT

promoters of the regulated genes (Drouin et al., 1989; Jonat et al., 1990). Neither of these sites are present in PGHS-2 gene promoters. V. Cellular and Physiological Actions of PGHS- 1 and PGHS-2

A. PGHS-1 AND PGHS-2 FUNCTION AS PART OF INDEPENDENT PROSTANOID BIOSYNTHETIC SYSTEMS

The generalized constitutive expression of PGHS-1 suggests that this enzyme is part of an immediate acute signaling system. In contrast, because of the time lag required for PGHS-2 induction, prostanoids formed by this enzyme are likely employed only in the secondary elaboration of physiological events. Induction of PGHS-2 occurs during and is assumed to play a role in inflammation (Futald et al., 1993; 1994; Masferrer et al., 1994; Seibert and Masferrer, 1994; Seibert et al., 1994), mitogenesis (Herschman, 1991, 1994), and ovulation (Hedin et al., 1987; Wong et al., 1989). Prostanoid production by PGHS-1 and PGHS-2 appears to be initiated through distinct signaling pathways (Murakami et al., 1994; Langenback et al., 1995) that may rely on the activation of different phospholipases (Reddy and Herschman, 1994). In addition, products of PGHS2, but not PGHS-1, localize to the nucleus (Morita et al., 1995), which suggests that PGHS-2 may independently signal directly to the nucleus. Thus, these two pathways of prostaglandin synthesis are both temporally and spatially separated, which may allow identical prostanoid products to play at least two different roles in signal transduction within a single cell. Transgenic mice in which either the PGHS-1 and PGHS-2 genes have been inactivated display uniquely different phenotypes, further supporting the concept that each enzyme, and prostaglandin pathway, has separate physiological functions (Langenbach et al., 1995; Morham et al., 1995). In this section, we develop two concepts that are presented diagrammatically in Fig. 12. The first is that PGHS-1 is part of an ER prostanoid biosynthetic system that forms products that act extracellularly to mediate acute “housekeeping” responses to circulating hormones, and the second is that PGHS-2 functions as part of a nuclear prostanoid biosynthetic system that forms products that act within the nucleus or on the nucleoplasmic surface of the nuclear envelope in association with cell differentiation or replication.

B. THEPGHS-1 SYSTEM Cells that express only PGHS-1 produce prostanoids rapidly in response to various hormonal stimuli, and the newly formed prostanoids act extracellularly as intercellular mediators. Human platelets (Patrignani et al., 1994) and rabbit renal collecting epithelia (Smith and Bell, 1978; D. L. DeWitt

CYCLOOXYGENASE ISOZYMES

197

FIG. 12. Model for the subcellular cornpartmenation of independent PGHS-1 and PGHS-2 systems. Abbreviations Used: E, effector; G, G protein; R, receptor; C, carrier, TXA-S, thromboxane synthase; PGI-S, prostacyclin synthase; sPLA2,secretory PLA2;sPLA2, cytosolic PLA2; R,, nuclear receptor; and MAPK, MAP kinase.

and W. L. Smith, unpublished results) are examples of cells that express PGHS-1 but not PGHS-2. Platelets and collecting tubule epithelia form TxAe and PGE2, respectively, as their major prostanoid products when incubated with either exogenous arachidonate or relevant hormonal stimuli (e.g., thrombin or bradykinin, respectively) that mobilize endogenous arachidonate from lipid precursors (Marcus et al., 1980; Grenier et al., 1981).

198

WILLIAM L. SMITH AND DAVID I,. DEW177

Two Ca2+-dependentphospholipase Azs-an 85-kDa cytosolic (c) PLA2 and a nonpancreatic, type 11, secretory (s) PLA2-have been implicated in various experimental settings in prostanoid formation (see Bonventre, 1992, and Dennis, 1994, for recent reviews). Figure 12 shows cPLA2mediate arachidonate release coupled to the PGHS-1 synthetic system. cPLAz does function in conjunction with PGHS-1 during platelet TxAz formation (Bartoli et al., 1994; Riendeau et al., 1994) and probably during endothelial cell PGIz synthesis (Murakami et al., 1993); moreover, cPLAz binds specifically to the cytoplasmic surfaces of both the ER and nuclear envelope in response to Ca2+mobilization (Figs. 6 and 12) (Nalefski et al., 1994; Glover et al., 1995; Schievella et al., 1996) thus becoming associated topologically with PGHS-1, which is located on the lumenal surfke of the ER and NE (Morita et al., 1995). sPLA2may also be involved in releasing arachidonate for the PGHS-1 system in some cells (Herschman, 1994); in those cases, sPLA2is secreted and cleaves arachidonate from phospholipids located on the cell surface, and the arachidonate can then enter the cell and subsequently the ER by simple diffusion (Kamp and Hamilton, 1993). PGHS-1 embedded in the lumenal surfaces of both the ER and the outer membrane of the nuclear envelope (Rollins and Smith, 1980; Regier et al., 1993; Otto and Smith, 1994; Morita et al., 1995) converts the newly released arachidonate to PGHz. Once PGH2 is formed by the action of PGHS-1 on the lumenal surface of the ER, PGH2traverses the ER [to which it is freely permeable (Marcus et al., 1980)] and is converted to a biological active prostanoid by enzymes such as TxA synthase (Ruan et at., 1993) or PGE synthase (Tanaka et al., 1987; Smith, 1992) located on the cytoplasmic face of the ER. Newly formed prostanoids (e.g., TxA2 or PGEJ must exit, but cell membranes are impermeant to prostanoids other than PGHz (Bito and Baroody, 1974; Marcus et al., 1980; Garcia-Perez and Smith, 1984). A prostaglandin transporter, such as that cloned by Schuster and co-workers (Kanai et al., 1995), is most likely involved in prostanoid export. Both TxA2 formed by platelets and PGEz synthesized by collecting tubules are found extracellularly within 15-60 sec following a hormonal stimulus (Marcus et al., 1980; Grenier et al., 1981). When TxAz is added exogenously to platelets it causes aggregation (Gorman et al., 1978). Similarly, PGEz added to collecting tubule epithelia inhibits vasopressininduced tubular water reabsorption (Grantham and Orloff, 1968). Thus, TxAz and PGEz elicit responses associated with their known physiological actions when added to the outside of cells. These responses are mediated by G protein-linked receptors (see Coleman et al., 1994, for a recent review) that are probably present on the plasma membrane.

CYCLOOXYGENASE ISOZYMES

199

Overall, one can conclude from studies of the synthesis and actions of prostanoids formed by cells such as platelets and renal collecting tubules, which express only PGHS-1, that prostanoids are formed rapidly via PGHS1 systems within cells exposed to the appropriate hormonal stimulus, that the newly formed prostanoids exit the cells probably via a carrier(s), and that these prostanoids then act via specific G protein-linked receptors located on the plasma membrane. C. THEPGHS-2 SYSTEM There are some general reasons for believing that PGHS-2, as illustrated in Fig. 12, functions independently of PGHS-1 and is involved in nuclear events. First, unlike PGHS-1, expression of PGHS-2 occurs in response to stimuli that cause cell differentiation or replication-events requiring participation of cell nucleus. Second, induction of PGHS-2 in fibroblasts or monocytes expressing PGHS-1 constitutively causes only nominal 1.5to 2-fold increases in the cellular PGHz biosynthetic activity (Habenicht et al., 1985; Raz et al., 1988; Lin et al., 1989; Lee et al., 1992; DeWitt and Meade, 1993; Evett et al., 1993); thus, in those cells that already express PGHS-1, there is no dramatic increase in the total PGHz biosynthetic capacity following the induction of PGHS-2. Coupled with the fact that PGHS-1 and PGHS-2 have very similar kinetic properties, it is difficult to argue that PGHS-2 is induced solely to augment the biosynthetic capacity of PGHS-1. Third, the two PGHS isozymes utilize different lipid stores of arachidonate for PGHz synthesis and perhaps different phospholipase systems to mobilize this arachidonate (Murakami et al., 1994; Reddy and Henchman, 1994). The data that have accumulated thus far suggest that there is no general rule regarding which lipase system is coupled to PGHS-2 (or PGHS-1) (Bonventre, 1992; Smith, 1992; Barbour and Dennis, 1993; Murakami et al., 1993; Balsinde et al., 1994; Nalefski et al., 1994). In some cases, cPLAz expression is controlled by the same factors that regulate PGHS-2 (Hoeck et al., 1993; Angel et al., 1994; Chepenik et al., 1994; Doerfler et al., 1994; Hulkower et al., 1994; Marshall et al., 1994; Roshak et al., 1994),suggesting that cPLAzcan couple to the PGHS-2 biosynthetic system. However, there is also evidence suggesting that sPLA2 is coupled to PGHS-2 in Madin Darby canine kidney cells (J. H. Schaefers and M. Goppelt-Struebe; unpublished results). Immunocytochemical and histochemical studies indicate that PGHS-1 and PGHS-2 are both present on the lumenal surface of the ER and the outer membrane of the NE, but that PGHS-2 is uniquely more concentrated than PGHS-1 in the NE (Morita et al., 1995). One explanation for these results is that a subset of PGHS-2 molecules is present on the inner membrane of the NE.

200

WILLIAM L. SMITH AND DAVID L. DEWITI

TxA synthase is located on the cytoplasmic surface of the ER (Ruan et al., 1993)and on the NE of retinoic acid-treated HL60 cells (K. Matsumoto, I. Morita, and S. Murota, unpublished results). PGIz synthase has been localized to the ER and NE of vascular smooth muscle and endothelial cells (Smith et al., 1983).TxA synthase in platelets and PGIz synthase in quiescent endothelial cells must be coupled to the PGHS-1 system. However, a nuclear TxA synthase activity is induced by treatment of HL60 cells with retinoic acid (K. Matsumoto, I. Morita, and S. Murota, unpublished results). Induction of PGHS-2, part of which is on the NE, is also associated with increases in PG12formation by human umbilical cord endothelial cells (Jones et al., 1993). Thus, under certain conditions both TxA synthase and PGIz synthase appear to be coupled to PGHS-2 and could be on the nucleoplasmic surface of the NE (Fig. 12). There is no direct evidence for functional prostanoid receptors associated with the cell nucleus, but prostanoid binding activities have been described in nuclear fractions (Rao and Mitra, 1982; I. Morita, unpublished results). It is conceivable that some of the known splice variants of prostanoid receptors (Namba et al., 1993) are targeted to the NE. VI. PGHS-1 and PGHS-2 in PathophysiologiesInflammation, Thrombosis, and Colon Cancer

PGHSs are of considerable therapeutic interest because they appear to be the major therapeutic targets of NSAIDs. NSAIDs are known to influence inflammation (Futaki et al., 1993, 1994; Masferrer et al., 1994), hemostasis (Patron0 et al., 1990; Willard et al., 1992), colon cancer (Marnett, 1992; Eberhart et al., 1994), and perhaps Alzheimer’s disease (Schnabel, 1993). A. PGHS-1 AND PLATELET AGGREGATION One “baby” aspirin daily or one regular aspirin every 3 days is a useful antiplatelet cardiovascular therapy (Willard et al., 1992). Platelet PGHS1is the therapeutic target (Funk et al., 1991; Patrignani et al., 1994). The low-dosage regimen leads to selective inhibition of platelet thromboxane formation (and platelet aggregation) without appreciably affecting the synthesis of other prostanoids in other cells. Circulating blood platelets, unlike most other cells, lack nuclei and are unable to synthesize new protein. Exposure of the PGHS-1 of platelets to circulating aspirin causes acetylation and irreversible inactivation of the platelet enzyme. Of course, PGHS1 (and PGHS-2) inactivation also occurs in other cell types, but cell types other than platelets can resynthesize PGHSs relatively quickly. For new PGHS-1 activity to appear in platelets, new platelets must be formed.

CYCLOOXYGENASE ISOZYMES

20 1

Because the replacement time for platelets is 5-10 days, it takes time for the circulating platelet pool to regain its original complement of active PGHS-1. Neither class I nor class I1 NSAIDs cause irreversible inhibition of PGHS-1 and thus are ineffective as antiplatelet agents.

B. PGHS-2 AND INFLAMMATION It now appears that PGHS-2 is the relevant enzyme in inflammation { Futaki et aE., 1993, 1994; Masferrer et at., 1994). Studies with agents that are relatively selective inhibitors of PGHS-2 indicate that these compounds have anti-inflammatory activities. However, to date no corresponding studies have been reported with PGHS-l-selective inhibitors, and it is still perplexing why class I NSAIDs such as piroxicam are anti-inflammatory but have very low affinities for PGHS-2. One possibility is that inhibition of PGHS-1 has the net effect of suppressing CAMPformation and that in some inflammatory situations CAMP promotes the expression of PGHS2. Importantly, PGHS-2-selective NSAIDs appear to lack the ulcerogenic effects of commonly available NSAIDs, which inhibit both isozymes (Futah et al., 1993, 1994; Masferrer et al., 1994). C. PGHS-2 A N D COLONCANCER There has been considerable interest in the role of prostanoids in colon cancer (see Marnett, 1992, for a recent review). Aspirin as well as other NSAIDs, such as sulindac and indomethacin, reduce the relative risk of colon cancer in human populations (Thun et al., 1991; Heath et al., 1994). Studies (Eberhart et al., 1994; Kargman et al., 1995) indicate that PGHS2 upregulation occurs in colonic adenomas. This finding, along with studies cited earlier indicating that PGH S-2 is an immediate-early gene associated with cell replication, are consistent with a role for this isozyme in colon cancer. On possible mechanism for the reduction in colon cancer observed with NSAID use may be the ability of these drugs to promote apoptosis in transformed cells (Lu et al., 1995). Increased synthesis of PGHS-2 products may allow continued survival of colonic epithelium early in the multistep transformation process; NSAID inhibition of the formation of these products could limit proliferation and thus prevent the transition of carcinomas. VII. Future Work

There are three general areas of study important to understanding more about PGHS isozymes-mechanisms of catalysis, regulation of gene expression, and subcellular functional independence. The first major area of study requires examination of (a) the mechanisms involved in hydrogen

202

WILLIAM L SMITl1 AND DAVID L. DEWI'IT

abstraction from arachidonate and the identity of the protein radical involved, (b)the chemical nature of the suicide inactivation, (c)the structural basis for time-dependent inhibition by class I1 NSAIDs, and (d)the nature of the O2binding site. Studies on the regulation of expression of the two PGHS genes will require (a) identification of what factors in cells activate expression of the PGHS-1 gene during differentiation and, correspondingly, what regions of the PGHS-1 gene are important for its activation; and (b) determination of the important transcriptional regulatory elements in the PGHS-2 gene. Finally, the concepts that PGHS-1 and PGHS-2 represent two separate prostaglandin biosynthetic pathways and two separate prostaglandin signaling pathways need to be tested. In this regard, it will be important to determine (a) if PGHS-1 and PGHS-2 are located in distinct subcellular compartments; (b) which lipases are involved in mobilizing arachidoiiate for each PGHS; (c) which PGHl metabolizing enzymes are coupled to each isozyme; and (d) the nature, mechanisms of actions, and subcellular locations of the receptors subserving of each PGHS pathway. Some questions arising in this area include (a) how is arachidonic acid specifically channeled to PGHS-2 following mobilization in cells in response to TPA (and presumably other mitogens)? (b) is the localization of PGHS-2 in the nuclear envelope important for PGHS-2 functioning (such as in the channeling of arachidonic acid) and, if so, what amino acid sequences in the enzyme are involved in targeting PGHS-2 to the nuclear envelope? and (c) because it appears that PGHS-2 releases PGHz into the nucleus of cells, do prostaglandins act in the nucleus directly, or are they released from cells to act at G protein-linked receptors at the cell membrane?

ACKNOWLEDGEMENTS Studies reported in this chapter were supported in part by NIH Grants DK42509 and DK22042 to W.L.S. and GM40713 to D.L.D. We thank Dr. R. Michael Garavito for Fig. 4 and for a careful readmg of a part of the manuscript. We also thank Drs. Oliver Smithies, Dr. Lawrence Marnett, Dr. Richard Kulmacz, Dr. Ahn Tsai, and Dr. Harvey Herschman for making studies from their laboratories available to us prior to publication.

REFERENCES Angel, I., Berenbaum, F., Le-Denmat. C., Nevdainen, T., Masliah, J., and Fournier, C. (1994). Interleukin-1-induced prostaglandin Ee biosynthesis in human synovial cells involves the activation of cytosolic phospholipase Ae and cyclooxygenase-2.Etrr. J. Biochem. 226, 125-131. Appleby, S. B., Ristimaki, A., Neilson, K., Narko, K., and Hla, T. (1994). Structure of the human cyclooxygenase-2 gene. Biochetn. 1. 302, 723-727. Balsinde, J., Barbour, S. E., Bianco, I. D., and Dennis. E. A. (1994). Arachidonic acid mobilization in P338D1 inacrophages is controlled by two distinct Ca(2+)-dependent phospholipase Az enzymes. Proc. Natl. Acad. Sci. USA 91, 11060-11064.

CYCLOOXYGENASE ISOZYMES

203

Barbour, S. E., and Dennis, E. A. (1993). Antisense inhibition of Group I1 phospholipase AL expression blocks the production of prostaglandin EZby P388D1 cells. J. Biol. Cheni. 268, 21875-21882. Barnett, J., Chow. J., Ives, D., Chiou, M., Mackenzie, R., Osen, E., Nguyen, B., Tsing, S., Bach, C., Freire, J., et al. (1994). Purification, characterization and selective inhibition of human prostaglandin G/H synthase 1 and 2 expressed in the baculovirus system. Biochini. Biophys. Acta 1209, 130-139. Bartoli, F., Lin, H.-K., Ghomashchi, F., Gelb, M. H., Jain, M. K., and Apitz-Castro, R. (1994).Tight binding inhibitors of 85 kDa phospholipase ALbut not 14-kDa phospholipase Ar inhibit release of free arachidonate in thrombin-stimulated human platelets. J. Biol. Cheni. 269, 15625-15630. Bhattacharyya, D. K., Lecointe, M., Dunn, J., Morgans, D. J., and Smith, W. L. (1995). Selective inhibiton of prostaglandin endoperoxide synthase-1 (cyclooxygenase-1)by valerylsalicylic acid. Arch. Biochern. Biophys. 317, 19-24. Bhattachatyya, D. K., Lecomte, M., Rieke, C. J., Garavito, R. M., and Smith, W. L. (1996). Involvement of arginine 120, glutamate 524, and tyrosine 355 in the binding of arachidonate and 2-phenylpropionic acid inhibitors to the cyclooxygenase active site of ovine prostaglandin endoperoxide synthase. Submitted for publication. Bito, L. Z., and Baroody, R. A. (1974). Impermeability of rabbit erythrocytes to prostaglandins. Alner. J. Physiol. 229, 1580-1586. Blanton, M. P., and Cohen, J. B. (1992). Mapping the lipid-exposed regions in the Torpedo californica nicotinic acetylcholine receptor. Bioclzeinistry 31, 3738-3750. Bonventre, J. A. (1992). Phospholipase Az and signal transduction. J. Ani. Soc. Nephrol. 3, 128-143. Brannon, T. S., North, A. J., Wells, L. B., and Shaul, P. W. (1994). Prostacyclin synthesis in ovine pumonary artery is developmentally regulated by changes in cycloxygenase-1 gene expression. J. Clin. Invest. 93, 2230-2235. Breder, C. D., DeWitt, D. L., and Kraig, R. P. (1995). Characterization of inducible cyclooxygenase in rat brain. J. Conip. Neurol. 355, 296-415. Capdevila, J. H., Morrow, J. D., Belosludtsev, Y. Y., Beauchamp, D. R., DuBois, R. N., and Falck, J. R. (1995). The catalytic outcomes of the constitutive and the mitogen inducible isoforins of prostaglandin He synthase are markedly affected by glutathione and glutathione peroxidase. Biochemistry 34, 3325-3337. Chepenik, K. P., Diaz, A,, and Jinienez, S. A. (1994). Epidermal growth factor coordinately regulates the expression of prostaglandin G/H synthase and cytosolic phospholipase AL genes in embryonic mouse cells. J. Biol. Cheni. 269, 21786-21792. Coleman, R. A,, Smith, W. L., and Narutniya, S. (1994).Classificationofprostanoid receptors: Properties, distribution and structure of the receptors and their subtpes. P h a m c o l . Rev. 46, 205-229. Copeland, R. A,, Williams, J. M.. Giannaras, J., Nurnberg, S., Covington, M., PInto, D., Pick, S., and Trzaskos, J. M. (1994). Mechanism of selective inhibition of the inducible isoforni of prostaglandin G/H synthase. Proc. Natl. Acad. Sci. USA 91, 11202-11206. Crofford, L. J., Wilder, R. L., Ristimaki, A. P., Sano, H.. Remmers, E. F., Epps, H. R., and Hla, T. (1994). Cyclooxygenase-I and -2 expression in rheumatoid synovid tissues. Effects of interleukin-1 beta. phorbol ester, and corticosteroids.J. Clin. Invest. 93, 10951101. Dennis, E. A. (1994). Diversity of group types, regulation and function of phospholipase A2. J. B i d Chem. 269, 1057-1060. DeWitt, D. L., and Meade, E. A. (1993). Senlm and ghcocorticoid regulation of gene transcription and expression of the prostaglandin H synthase-1 and prostaglandin H synthase-2 isozymes. Arch. Biocheni. Biophys. 306, 94-102.

204

WILLIAM L. SMITH AND DAVID L. DEWITI

DeWitt, D. L., and Smith, W. L. (1988). Primary structure of prostaglandin G/H synthase from sheep vesicular gland determined from the complementary DNA sequence. Proc. Natl. Acad. Sci. USA 85, 1212-1416. DeWitt, D. L., Rollins, T. E., Day, J. S., Gauger, J. A., and Smith, W. L. (1981). Orientation of the active site and antigenic determinants of prostaglandin endoperoxide synthase in the endoplasmic reticulum. J. Biol. Chem. 256, 10375- 10382. DeWitt, D. L., Day, J. S., Sonnenburg, W. K., and Smith, W. L. (1983). Concentrations of prostaglandin endoperoxide synthase and prostaglandin I2 synthase in the endothelium and smooth muscle of bovine aorta. J. Clin. Znuest. 72, 1882-1888. DeWitt, D.L., El-Harith, E. A,, Kraemer, S. A., Andrews, M. J., Yao, E. F., Armstrong, R. L., and Smith, W. L. (1990). The aspirin and heme-binding sites of ovine and murine prostaglandin endoperoxide synthases.J. Biol. Chem. 265,5192-5198. Dietz, R.,Nastainczyk, W., and Ruf, H. H. (1988). Higher oxidation states of prostaglandin H synthase. Rapid electronic spectroscopy detected two spectral intermediates during the peroxidase reaction with prostaglandin G2. Eur. J. Biochem. 171,321-328. Doeder, M. E., Weiss, J,, Clark, J. D., and Elsbach, P. (1994). Bacterial lipopolysaccharide primes human neutrophils for enhanced release of arachidonic acid and causes phosphorylation of an 85-kD cytosolic phospholipase AZ. J. Clin. Invest. 93, 1583-1591. Drouin, J.,Tifiro, M., Plante, R., Nemer, M., Eriksson, P., and Wrange, 0.(1989). Clucocorticoid receptor binding to a specific DNA seqeunce is required for hormone-dependent repression of Pro-opiomelanocortin gene transcription. Mol. Cell. Biol. 9, 5305-5314. DuBois, R. N., Awad, J., Morrow, J., Roberts, L. J., 11, and Bishop, P. R. (1994). Regulation of eicosanoid production and mitogenesis in rat intestinal epithelial cells by transforming growth factor- and phorbol ester. J Clin. Inuest. 93, 493-498. Eberhart, C. E., Coffey, R. J., Radhika, A., Giardiello, F. M., Ferrenbach, S . , and DuBois, R. N. (1994). Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 107, 1183- 1188. Eling, T. E., Thompson, D. C., Foureman, G. L., Curtis, J. F., and Hughes, M. F. (1990). Prostaglandin H synthase and xenobiotic oxidation. Annu. Rev. Phamnacol. Toxicol. 30, 1-45. Eling, T.E., Glasgow, W. C., Curtis, J. F., Hubbard, W. C., and Handler, J. A. (1991). Studies on the reduction of endogenously generated prostaglandin G2 by prostaglandin H synthase. J. Biol. Chem. 266, 12348-12355. Evett, G. E., Xie, W., Chipman, J. G., Robertson, D. L., and Simmons, D. L. (1993). Prostaglandin G/H isoenzyme 2 expression in fibroblasts: Regulation by dexamethasone, mitogens, and oncogenes. Arch. Biochem. Biophys. 306, 169-177. Feng, L.,Sun, W., Xia, Y., Tang, W. W., Chanmugam, P., Soyoola, E., Wilson, C. B., and Hwang, D. (1993). Cloning two isoforms of rat cyclooxygenase: Differential regulation of their expression. Arch. Biochem. Biophys. 307, 361-368. Fletcher, B. S., Kujubu, D. A,, Perrin, D. M., and Henchman, H. R. (1992). Structure of the mitogen-inducible TISlO gene and demonstration that the TIS10-encoded protein is a functional prostaglandin G/H synthase. J. Biol. Chem. 267, 4338-4344. Funk, C. D., Funk, L. B., Kennedy, M. E., Pong, A. S., and Fitzgerald, G. A. (1991). Human plateletlerythroleukemia cell prostaglandin G/H synthase: cDNA cloning, expression, and gene chromosomal assignment. FASEB J. 5, 2304-2312. Futaki, N., Yoshikawa, K., Hamasaka, Y., Arai, I., Higuchi, S., Iizuka, H., and Otomo, S. (1993). NS-398, a novel non-steroidal anti-inflammatory drug with potent analgesic and antipyetic effects, which causes minimal stomach lesions. Gen. Phamcol. 24, 105-1 10. Futaki, N., Takahashi, S., Yokoyania, M., Arai, I., Higuchi, S., and Otomo, S. (1994). NS398, a new anti-inflammatory agent, selectively inhibits prostaglandin G/H synthase/ cyclooxygenase (COX-2) activity in vitro. Prostaglandins 47,55-59.

CYCLOOXYGENASE ISOZYMES

205

Garcia-Perez, A., and Smith, W. L. (1984). Apical-basolateral membrane asymmetry in canine cortical collecting tubule cells: Bradykinin, AW, PCEp interrelationships, /. Clin. Invest. 744, 63-74. Glaser, K. B., and Lock, Y. W. (1995). Regulation of prostaglandin H synthase 2 expression in human monocytes by the marine natural products inanoalide and scalaradial: Novel effects independent of inhibition of lipid mediator production. Biochem. Phamlacol. 50,913-922. Glover, S., Bayburt, T., Mechthild, J., Chi, E., and Gelb, M. H. (1995). Translocation of the 85-kDa phospholipase A2 from cytosol to the nuclear envelope in rat basophilic leukemia cells stimulated with calcium ionophore or IgElantigen. J. Biol. Chem. 270, 15359-15367. Gorrnan, R. R., Fitzpatrick, F. A., and Miller, 0.V. (1978). Reciprocal regulation of human platelet CAMPlevels by throinboxane A2 and prostacyclin. In “Advances in Cyclic Nucleotide Research” (W. W. George and L. J. Ignarro, Eds.), Vol. 9, pp. 597-609. Raven Press, New York. Grantham, J. J., and Orloff, J. (1968). Effect of prostaglandin E l on the permeability response of the isolated collecting tubule to vasopressin, adenosine 3’,5‘-monophosphate, and theophylline. J. Clin. Invest. 47, 1154-1161. Grenier, F. C., Rollins, T. E., and Smith, W. L. (1981).Kinin-induced prostaglandin synthesis by renal papillary collecting tubule cells. Am. J . Physiol. 241, F94-F101. Grove, M., and Plumb, M. (1993). C/EBP, NF-kappa B, and c-Ets family members and transcriptional regulation of the cell-specific and inducible macrophage inflammatory protein I immediate-early gene. Mul. Cell. Biol. 13, 5276-5289. Habenicht, A. J. R., Goerig, M., Grulich, J., Rothe, D., Gronwald, R., Loth, U., Schettler, G., Kommerell, B., and Ross, R. (1985). Human platelet-derived growth factor stimulates prostaglandin synthesis by activation and by rapid de novo synthesis of cyclooxygenase. J. Clin. Invest. 75, 1381-1387. Habib, A,, Creminon, C., Frobert, Y., Grassi, J,, Pradelles, P., and Maclouf, J. (1993). Demonstration of an inducible cyclooxygenasein human endothelial cells using antibodies raised against the carboxyl-terminal region of the cyclooxygenase-2. J. Biol. Chem. 268,23448-23454. Hamberg, M., and Samuelsson, B. (1967). On the mechanism of the biosynthesis of prostaglandins E, and Fle.1. Bid. Chem. 242, 5336-5343. Han, J,, Sadowski, H., Young, D. A., and Macara, I. G. (1990). Persistant induction of cyclooxygenase in p60v-src-transforined 3T3 fibroblasts. Proc. Natl. Acad. Sci. USA 87, 3373-3377. Harris, R. C., McKanna, J. A,, Akai, Y., Jacobson, H. R., Dubois, R. N., and Breyer, M. D. (1994). Cyclooxygenase-2 is associated with the macula densa of rat kidney and increases with salt restriction. J. Clin. Znoest. 94, 2504-2510. Heath, C . W., Jr., Thun, M. J., Greenberg, E. R., Levin, B., and Marnett, L. J. (1994). NSAIDs and human cancer: Report of an interdisciplinary research workshop. Cancer 74, 2885-2888. Hedin, L., Caddy-Kurten, D., Kurten, R., DeWitt, D. L., Smith, W. L., and Richards, J. s. (1987). Prostaglandin endoperoxide synthetase in rat ovarian follicles: Content, cellular distribution and evidence for hormonal induction preceding ovulation. Endrocrinology 121,722-731. Hemler, M. E., and Lands, W. E. M. (1980). Evidence for a peroxide-initiated free radical mechanism of prostaglandin biosynthesis. I. Biol. Chem. 255, 6253-6261. Hemler, M., Lands, W. E. M., and Smith, W. L. (1976). Purification of the cyclooxygenase that forms prostaglandins. Demonstration of two forms of iron in the holoenzyrne.J. Biol. Chem. 251, 5575-5581.

206

WILLIAM L. SMITH AND DAVID L. D E W I T

Hemler, M. E., Crawford, C. G., and Lands, W. E. M. (1978). Lipoxygeiiation activity of purified prostaglandin-forming cyclooxygenase.Biochemistnj 17, 1772- 1779. Herschman, H. R. (1991). Primary response genes induced by growth factors and tumor promoters. In “Annual Review of Biochemistry” (C. C. Richardson, J. N. Abelson, A. Meister, and C. T. W. Walsh, Eds.),Vo!. 60, pp. 281-319. Annual Reviews, Palo Alto, CA. Herschman, H. R. (1994). Regulation of prostaglandin synthase-1 and prostaglandin synthase-2. Can. Metastasis Reu. 13, 241-256. Hirst, J. J., Teixeira, F. J., Zakar, T., and Olson, D. M. (1995). Prostaglandin endoperoxideH synthase-1 and -2 messenger ribonucleic acid levels in human amnion with spontaneous labor onset. 1.Clin. Endocrinol. Metab. 80, 517-523. Hla, T., and Maciag, T. (1991). Cyclooxygenase gene expression is down-regulated by heparin-binding (acidic fibroblast) growth factor-1 in human endothelial cells. J. Biol. Chem. 266,24059-24063. Hla, T., and Neilson, K. (1992). Human cyclooxygenase-2 cDNA. Proc. Nutl. Acad. Sci. USA 89, 7384-7388. Hoeck, W. G., Ramesha, C. S., Chang, D. J., Fan, N., and Heller, R. A. (1993). Cytoplasmic phospholipase Ap activity and gene expression are stimulated by tumor necrosis factor: Dexamethasone blocks the induced synthesis. Proc. Nutl. Acad. Sci. USA 90,4475-4479. Hoff, T., DeWitt, D., Kaever, V., Resch, K., and Goppelt-Struebe, M. (1993).Differentiationassociated expression of prostaglandin G/H synthase in monocytic cells. FEBS Lett. 320,38-42, Holtzman, M. J., Turk, J., and Shornick, L. P. (1992). Identification of a pharmacologically distinct prostaglandin H synthase in cultured epithelial cells.]. Bid. Chem. 267, 2143821445. Hsi, L. C., Hoganson, C. W., Babcock, G. T., and Smith, W. L. (1994). Characterization of a tyrosyl radical in prostaglandin endoperoxide synthase-2. Biochem. Biophys. Res. Comniun. 202, 1592-1598. Hulkower, K. I., Wertheimer, S. J., Levin, W., Coffey, J. W., Anderson, C. M., Chen, T., DeWitt, D. L., Crowl, R. M., Hope, W. C., and Morgan, D. W. (1994). Interleukin-1 beta induces cytosolic phospholipase Ae and prostaglandin H synthase in rheumatoid synovial fibroblasts.Evidence for their roles in the production of prostaglandin Ep.Arthritis Rheum.37,653-661. Inoue, H., Nanayama, T., Hara, S., Yokoyarna, C., and Tanabe, T. (1994). The cyclic AMP response element plays an essential role in the expression of the human prostaglandinendoperoxide synthase 2 gene in differentiated U937 monocytic cells. FEBS Lett. 350, 51-54. Jacobs, A. L., Hwang, D., Julian, J., and Carson, D. D. (1994). Regulated expression of prostaglandin endoperoxide synthase-2by uterine stroma. Endocrinology 135,1807- 1815. Jonat, C., Rahmsdorf, H. J., Park, K., Cato, A. C. B., Gebel, S., Ponta, H., and Herrlich, P. (1990). Antitumor promotion and anti-inflammation: Down-modulation of AP-1 (Fad Jun) activity by glucocorticoid hormone. Cell 62, 1189-1204. Jones, D. A., Carlton, D. P., McIntyre, T. M., Zirnmerman, G. A,, and Prescott, S. M. (1993). Molecular cloning of human prostaglandin endoperoxide synthase type I1 and demonstration of expression in response to cytokines.J. Biol. Chem. 268, 9049-9054. Kamp, F., and Hamilton, J. A. (1993). Movement of fatty acids, fatty acid analogues, and bile acids across phospholipid bilayers. Biochemistry 32, 11074-11085. Kanai, N., Run, L., Satriano, J. A., Bao, Y., Wolkoff, A. W., and Schuster, V. L. (1995). Identification and characterization of a prostaglandin transporter. Science 268,866-869. Kargman, S., Chan, S., Evans, J., Vickers, P., and ONeill, G. (1994). Tissue distribution of prostaglandin G/H synthase-1 and -2 (PGHS-1 and PGHS-2) using specific and antipeptide antibodies. /. Cell. Biochern. Suppl. 18B, 319. [Abstract 01091

CYC LOOXYG E N AS E I SOZYM E S

207

Kargman, S. L., O’Neill, C . P., Vickers, P. J., Evans, J. F., Mancini, J. A,, and Serge, J. (1995).Expression of prostaglandin G/H synthase-l and -2 protein in human colon cancer. Cancer Res. 55, 2556-2559. Karthein, R., Dietz, R., Nastainczyk, W., and Ruf, H. H. (1988). Higher oxidation states of prostaglandin H synthase. EPR study of a transient tyrosyl radical in the enzyme during the peroxidase reaction. Eur. J. Biochetn. 171, 313-320. Kawaguchi, H., Raisz, L. G., Voznesensky, 0. S., Alander, C . B., Hakeda, Y., and Pilbeam, C. C . (1994). Regulation of the two prostaglandin G/H synthases by parathyroid hormone, interleukin-1, cortisol, and prostaglandin EL in cultured neonatal mouse calvariae.Etatlocrinology 135, 1157-1164. Kester, M., Coroneos, E., Thomas, P. J., and Dunn, M. J. (1994). Endothelin stimulates prostaglandin endoperoxide synthase-2 inRNA expression and protein synthesis through atyrosine kinase-signalingpathway in rat mesangial cells.]. B i d . Chem. 269,22574-22580. Kosaka, T., Miyata, A,, Ihara, H., Hard, S., Suginloto, T., Takeda, O., Takahashi, E., and Tanabe, T. (1994).Characterization of the human gene (PTGS2) encoding prostaglandinendoperoxide synthase 2. Eur. J. Biochem. 221, 889-897. Kraenier, S. A., Meade, S. A., and DeWitt, D. L. (1992). Prostaglandin endoperoxide synthase gene structure: Identification of the transcriptional start site and 5’-flanking regulatory sequences. Arch. Biochem. Biophys. 293, 391-400. Kujubu, D. A,, and Herschman, H. R. (1992). Dexainethasone inhibits mitogen induction of the TISlO prostaglandin synthase/cyclooxygenasegene. J. Biol. Chern. 267,7991-7994. Kujubu, D. A,, Fletcher, B. S., Varnum, B. C., Lim, R. W., and Herschman, H. R. (1991). TIS10, a phorbol ester tumor promoter inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthasekyclooxygenase honiologue. J . B i d . Chern. 266, 1286612872. Kujubu, D. A,, Reddy, S. T., Fletcher, B. S.. arid Herschman, H. R. (1993). Expression of the protein product of the prostaglandin synthase-Z/TIS10 gene in mitogen-stimulated Swiss 3T3 cells. J. B i d . Chem 268, 5425-5430. Kulmacz, R. J. (1989). Topography of prostaglandin H synthase. Antiinflammatory agents arid the protease-sensitive arginine 253 region. J . Biol. Chetn. 264, 14136-14144. Kulmacz, R. J., and Lands, W. E. (1983). Requirements for hydroperoxide by the cyclooxygenase and peroxidase activities of prostaglandin €3 synthase. Prostaglandins 25,531-540. Kulmacz. R. J., and Lands, W. E. M. (1985). Stoichioinetry and kinetics of the interaction of prostaglandin H hynthase with anti-inflammatory agents. J. B i d . Chern. 260, 1257212578. Kulmacz, R. J., and Wang, L.-H. (1995).Comparison of hydroperoxideinitiator requirements for the cyclooxygenase activities of prostaglandin H synthase-1 and -2. J. Biol. Cheni. 270, 24019-24023. Kulmacz, R. J., Tsai, A. L., and Palmer, G. (1987). Heme spin states and peroxide-induced radical species in prostaglandin H synthase. J. B i d . Chem. 262, 10524-10531. Kulmacz, R. J.. Ren, Y., Tsai, A. L., and Palmer, G. (1990). Prostaglandin H synthase: Spectroscopic studies of the interaction with hydroperoxides and with indomethacin. Biochemistry 29, 8760-8771. Kulmacz, R. J., Pendleton, R. B.. and Lands, W. E. (1994). Interaction between peroxidase and cyclooxygenase activities in prostaglandin-endoperoxide synthase. Interpretation of reaction kinetics. J . B i d . Chern. 269, 5527-5536. Kwok, P.-Y., Muellner, F. W., and Fried, J. (1987). Enzymatic conversion of 10,lOdifluoroarachidonic acid with PGH synthase and soybean lipoxygenase.J. A m . Chem. SOC. 109,3692-3698.

208

WILLIAM L. SMITH AND DAVID L. DEWI??

Lambeir, A. M., Markey, C. M., Dunford, H. B., and Marnett, L. J. (1985). Spectral properties of the higher oxidation states of prostaglandin H synthase. 1. Biol. Chem. 260, 14894-14896. Lands, W. E. M., LeTellier, P. R., Rome, L. H., and Vanderhoek, J. Y. (1973). Inhibition of prostaglandin biosynthesis. Adv. Biosci. 9, 15-28. Laneuville, O., Breuer, D. K., DeWitt, D. L., Hla, T., Funk, C. D., and Smith, W. L. (1994). Differential inhibition of human prostaglandin endoperoxide H synthases-1 and -2 by nonsteroidal anti-inflammatory drugs. J. Pharmucol. Exp. Ther. 271, 927-934. Laneuville, O., Breuer, D. K., Xu, N., Huang, Z. H., Gage, D. A., Watson, J. T., Lagarde, M., DeWitt, D. L., and Smith, W. L. (1995). Fatty acid substrate specificities of human prostaglandin endoperoxide H synthases-1 and -2. Formation of 12 hydroxy-(9Z,13E/ Z,lSZ)-Octadecatrienoic acids from a-linolenic acid. 1.Biol. Chem. 270, 19330-19336. Langenbach, R., Morhan, S. G., Tiano, H. F., Loftin, C. D., Ghanayem, B. I., Chulada, P. C., Mahler, J. F., Goulding, E. H., Klluckman, K. D., Kim, H. S., and Smithies, 0. (1995).Prostaglandin synthase I gene disruption in mice reduces arachidonic acid induced inflammation and iiidomethacin induced gastric ulceration. Cell 83, 483-493. Leconite, M., Laneuville, O., Ji, C., DeWitt, D. L., and Smith, W. L. (1994). AcetyIation of human prostaglandin endoperoxide synthase-2 (cyclooxygenase-2)by aspirin. J. Biol. Chem. 269, 13207-13215. Lee, S. H., Soyoola, E., Chanmugam, P., Hart, S., Sun, W., Zhong, H., Liou, S., Simmons, D., and Hwang, D. (1992). Selective expression of mitogen-inducible cyclooxygenase in macrophages stimulated with lipopolysaccharide./. Bid. Chem. 267, 25934-25938. Libermann, T. A., and Baltimore, D. (1990). Activation of interleukin-6 gene expression through the NF-kappa B transcription factor. Mol. Cell. B i d . 10, 2327-2334. Lin, A. H., Bienkowski, M. J,, and Gorman, R. R. (1989). Regulation of prostaglandin H synthase mRNA levels and prostaglandin biosynthesis by platelet-derived growth factor. 1. Biol. Chem. 264, 17379-17383. Loll, P. J., Picot, D., and Garavito, R. M. (1995). The structural basis of aspirin activity inferred from the crystal structure of inactivated prostaglandin H2synthase. Nature Structural Bid. 2, 637-643. Lu, X., Xie, W., Reed, D., Bradshaw, W. S., and Simmons, D. L. (1995). Nonsteroidal antiinflammatory drugs cause apoptosis and induce cyclooxygenases in chicken embryo fibroblasts. Proc. Natl. Acad. Sci. USA 92, 7961-7965. Mancini, J. A,, O’Neill, G. P., Bayly, C., and Vickers, P. J. (1994). Mutation of serine-516 in human prostaglandin G/H synthase-2 to methionine or aspirin acetylation of this residue stimulates 15-R-HETE synthesis. F E B S Lett. 342, 33-37. Marcus, A. J., Weksler, B. B., Jaffe, E. A,, and Broekman, J. M. (1980). Synthesis of prostacyclin from platelet-derived endoperoxides by cultured human endothelial cells. I. Clin. Invest. 66, 979-986. Markey, C. M., Alward, A,, Weller, P. E.. and Marnett, L. J. (1987). Quantitative studies of hydroperoxide reduction by prostaglandin H synthase. Reducing substrate specificity and the relationship of peroxidase to cyclooxygenaseactivities. J. Biol. Chem. 262,62666279. Marletta, M. A. (1994). Nitric oxide synthase: Aspects concerning structure and catalysis. Cell 78,9127-9130. Marnett, L. J. (1992). Aspirin and the potential role of prostaglandins in colon cancer. Cancer 52,5575-5589. Marnett, L. J., and Maddipati, K. R. (1991). Prostaglandin H synthase. In “Peroxidases; Chemistry and Biology” (J. Everse, K. Everse, and M. Grisham, Eds.), pp. 293-334. CRC Press, Boca Raton, FL.

CYCLOOXYGENASE ISOZYMES

209

Mamett, L. J., Chen, Y. N., Maddipati, K. R., Ple, P., and Labeque, R. (1988). Functional differentiation of cyclooxygenase and peroxidase activities of prostaglandin synthase by tiypsin treatment. Possible location of a prosthetic heme binding site. J. B i d Chem. 263, 16532-16535. Marshall, L. A., Bolognese, B., and Roshak, A. (1994). Characterization of phospholipase A2 release by elicited-peritoneal macrophage and its relationship to eicosanoid production. J. Lipid Med. 10, 295-313. Marshall, P. J., and Kulmacz, R. J. (1988). Prostaglandin H synthase: distinct binding sites for cyclooxygenase and peroxidase substrates. Arch. Biochem. Biophys. 266, 162170. Masferrer, J. L., Zweifel, B. S., Manning, P. T., Hauser, S. D., Leahy, K. M., Smith, W. G., Isakson, P. C., and Seibert, K . (1994).Selective inhibition of inducible cyclooxygenase 2 in vivo is anti-inflammatoryand nonulcerogenic. Proc. Natl. Acud. Sci. USA 91,32283232. Mason, R. P., Kalyanaraman, B., Tainer, B. E., and Eling, T. E. (1980).A carbon-centered free radical intermediate in the prostaglandin synthetase oxidation of arachidonic acid. Spin trapping and oxygen uptake studies. J. B i d . Chem. 255, 5019-5022. Meade, E. A., Smith, W. L., and DeWitt, D. L. (1993).Differential inhibition ofprostaglandin endoperoxide synthase (cyclooxygenase)isozylnes by aspirin and other non-steroidal antiinflammatory drugs. J . Biol. Chem. 268, 6610-6614. Merlie, J. P., Fagan, D., Mudd, J., and Needeman, P. (1988).Isolation and characterization of the complementary DNA for sheep seminal vesicle prostaglandin endoperoxide synthase (cyclooxygenase).J. B i d . Chem. 263, 3550-3553. Mertz, P. M., DeWitt, D. L., Stetler-Stevenson, W. G., and Wahl, L. M. (1994). Interleukin 10 suppression of monocyte prostaglandin H synthase-2. Mechanism of inhibition of prostaglandin-dependent matrix metalloproteinase production.J. Biol. Chem. 269,2132221329. Miyamoto, T., Ogino, N., Yamamoto, S., and Hayaishi, 0. (1976).Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J . Biol. Chem. 251, 2629-2636. Mizuno, K., Yamamoto, S., and Lands, W. E. M. (1982). Effects of non-steroidal antiinflammatory drugs on fatty acid cyclooxygenaseand prostaglandin hydroperoxidase activities. Prostaglandins 23, 743-757. Morham, S. G., Langenbach, R., Loftin, C. D., Tiano, H. F., Voulounianos, N., Jennette, J. C., Mahler, J. F., Kluckman, K. D., Ledford, A,, Lee, C. A,, and Smithies, 0. (1995). Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell 83, 473-482. Morita, I., Schindler, M., Regier, M. K., Otto, J. C., Hori, T., DeWitt, D. L., and Smith, W. L. (1995). Different intracellular locations for prostaglandin endoperoxide H synthases-1 and -2. J. Biol. Chem. 270, 10902-10908. Muller, J. M., Ziegler-Heitbrock, H. W., and Baeuerle, P. A. (1993). Nuclear factor kappa B, a mediator of lipopolysaccharide effects. Zmmunobiology 187, 233-256. Muller-Decker, K., Scholz, K., Marks, F., and Furstenberger, G. (1995). Differential expression of prostaglandin H synthase isozymes diiring multistage carcinogenesis in mouse epidermis. Mul. Carcinogen. 12, 31-41. Munro, S., and Pelham, H. R. B. (1987).A C-terminal signal prevents secretion of luminal ER proteins. Cell 48, 899-907. Munroe, D. G., and Lau, C. Y. (1995). Turning down the heat: New routes to inhibition of inflammatory signaling by prostaglandin H2 synthases. Chem. B i d . 2, 343-350.

210

WILLIAM L SMITII AND DAVID L. D E W I l T

Murakami, M., Kudo, I., and Inoue, K. (1993). Molecular nature of phospholipases AL involved in prostaglandin Iz synthesis in human umbilical vein endothelial cells. Possible participation of cytosolic and extracellular type I1 phospholipases AP. J . B i d . Chein. 268, 839-844. Murakami, M., Matsumoto, R., Austen, K. F., and Arm, J. P. (1994).Prostaglandin endoperoxide synthase-1 and -2 couple to different transmeinbrane stimuli to generate prostaglandin Dz in mouse bone marrow-derived mast cells. J . B i d . Chem. 269, 22269-22275. Mutsaers, J. H., van-Halbeek, H., Kamerling, J. P., andvliegenthart, J. F. (1985). Determination of the structure of the carbohydrate chains of prostaglandin endoperoxide synthase from sheep. Eur. J . Biochem. 147, 569-574. Nalefski, E. A,, Sultzman, L. A,, Martin, D. M., Kriz, R. W., Towler, P. S., Knopf, J. L., and Clark, J. D. (1994). Delineation of two functionally distinct domains of cytosolic phospholipase A2, a regulatory CA2 t -dependent lipid-binding domain and a C a 2 f independent catalytic domain. J . B i d Chem. 269, 18239-18249. Namba, T., Sugirnoto,Y., Negishi, M., Irie, A,, Ushikubi, F., Kakizuka, A,, Ito, S., Ickikawa, A,, and Naruiniya, S. (1993). Alternative splicing of C-terminal tail of prostaglandin E receptor subtype EP3 determines G-protein specificity. Nature 365, 166-170. Nathan, C., and Xie, Q. (1994). Nitric oxide synthases: Roles, tolls, and controls. Cell 78,915-918. Nothwehr, S. F., and Stevens, T. H. (1994). Sorting of membrane proteins in the yeast secretory pathway. J. B i d . Chem. 269, 10185-10188. O’Banion, M. K., Winn, V. D., and Young, D. A. (1992). cDNA Cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase.Proc. Natl. Acad. Sci. USA 89, 4888-4892. Odenwaller, R., Maddipati, K. R., and Mamett, L. J. (1992). Detection of a higher oxidation state of manganese-prostaglandin endoperoxide synthase. J . B i d . Chein. 267, 1386313869. Ogino, N., Ohki, S., Yamamoto, S., and Hayaishi, 0. (1978). Prostaglandin endoperoxide synthetase from bovine vesicuIar gland microsomes. Inactivation and activation by heme and other metalloporphyrins. J. B i d . Chern. 253, 5061-5068. Ohki, S., Ogino, N., Yamamoto, S., and Hayaishi, 0.(1979).Prostaglandin hydroperoxidase, an integral part of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J. Biol. Chem. 254, 829-836. O’Neill, C. P., and Ford-Hutchinson, A. W. (1993).Expression of mRNA for cyclooxygenase1 and cyclooxygenase-2 in human tissues. FEBS Lett. 330, 156-160. O’Neill, G. P., Mancini, J. A,, Kargman, S., Yergey, J., Kwan, M. Y., Falgueyret, J. P., Abramovitz, M., Kennedy, B. P., Ouellet, M., Cromlish, W., et QZ. (1994). Overexpression of human prostaglandin G/H synthase-1 and -2 by recombinant vaccinia virus: inhibition by nonsteroidal anti-inflammatory drugs and biosynthesis of 15-hydroxyeicosatetraenoic acid. Mol. Pharmacol. 45, 245-254. O’Subvan, G. M., Chilton, F. H., Huggins, E. M., Jr., and McCall, C. E. (1992).Lipopolysaccharide priming of alveolar macrophages for enhanced synthesis of prostanoids involves induction of a novel prostaglandin H synthase. J. Bid. Chetn. 267, 14547-14550. O’Sullivan, M. G., Huggins, E. M., Jr., Meade, E. A,, DeWitt, D. L., and McCall, C. E. (1992). Lipopolysaccharide induces prostaglandin H synthase-2 in alveolar macrophages. Biocheni. Biophys. Res. Cornmiin. 187, 1123-1127. Otto, J. C . , and Smith, W. L. (1994). The orientation of prostaglandin endoperoxide sy”thases-1 and -2 in the endoplasmic reticulum. /. Biol. Chem. 269, 19868-19875. Otto, J. C., DeWitt, D. L., and Smith, W. L. (1993). N-glycosylation of prostaglandin endoperoxide synthases-1 and -2 and their orientations in the endoplasmic reticulum. 1.Biol. Chem. 268, 18234-18242.

CYCLOOXYGENASE ISOZYMES

211

Patrignani, P., Panarii, M. R., Greco, A., Fusco, P., Natoli, C., Iacobelli, S., Cipollone, F., Ganci, A., Creminon, C., Maclouf; J., et al. (1994). Biochemical and pharmacological characterization of the cyclooxygenaseactivity of human blood prostaglandin endoperoxide synthases. /. Pliarniacol. Exp, Ther. 271, 1705-1712. Patrono. C., Ciabattoni, G., and Davi, G. (1990).Thromboxane biosynthesis in cardiovascular diseases. Stroke 2l(Suppl. l2), IVl30-IV133. Pelhani, H. R. B. (1990). The retention signal for soluble proteins of the endoplasmic reticulum. TZBS 15, 483-485. Percival, M. D., Oullet, M., Vincent, C. J., Yergey. J. A , , Kennedy, B. P., and O’Neill, G. P. (1994). Purification and characterization of recombinant human cylooxygenase-2. Arch Biochem. Bioplzys. 315, 1 1 1- 118. Phillips, T. A,, Kujubu, D. A,, MacKay, R. I., Herschman, H. R., Russell, S. W., and Pace, J. L. ( 1993). The mouse inacrophage activation-associated marker protein, p71/73, is an inducible prostaglandin endoperoxide synthase (cyclooxygenase).J . Leukocyte Biol. 53, 411-419. Picot, D., and Garavito, R. M. (1994).Prostaglandin H synthase: Implicatioiis for membrane structure. FEBS Lett. 346, 21-25. Picot, D., Loll, P. J., and Garavito, M . (1994).The X-ray crystal structure of the membrane protein prostaglandin HLsynthase-l . Nature 367, 243-249. Pilbeani, C. C., Kawaguchi, H., Hakeda, Y., Vosnesensky, O., dander, C. B., and Raisz, L. C. ( 1993).Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic MC3T3-El cells. 1. Biol. Chem. 268, 25643-25649. Rao, C. V., and Mitra, S. B. (1982). Distribution of PGE and PGF2, receptor proteins in the intracellular organelles of bovine corpora lutea. Methods Enzymol. 86, 192-202. Raz, A,, Wyche, A,, Siegel, N., and Needleman, P. (1988). Regulation of fibroblast cyclooxygenase synthesis by interleukin-1.1. Biol. Cheni. 263, 3022-3028. Reddy, S. T., and Herschman, H. R. (1994).Ligand-induced prostaglandin synthesis requires expression of the TIS 10PCS-2 prostaglandin synthase gene in muiine fibroblasts and macrophages. 1. B i d . Chein. 269, 15470- 15480. Regier, M. K., DeWitt, D. L., Schindler, M. S., and Smith, W. L. (1993). Subcellrilar localization of prostaglandin endoperoxide synthase-2 in murine 3T3 cells. Arch. Biochem. B i ~ p h y 301, ~ . 439-444. Regier, M. K., Otto, J. C., DeWitt, D. L., and Smith, W. L. (1995).Localization ofprostaglandin endoperoxide synthase-1 to the endoplasmic reticiilum and nuclear envelope is Independent of Its C-terminal Tetrapeptide-PTEL. Arch. Biochetn. Biophys. 317, 457-463. Ren, Y., Loose-Mitchell, D. S.. and Kulmacz, R. J. (1995). Prostaglandin H synthase-1: Evaluation of C-terminus function. Arch. Biochem. Biophys. 316, 751-757. Riendeau, D., Cuay, J., Weech, P. K., Laliberte, F., and Yergey, J. (1994). Arachidonyl trifluoromethyl ketone, a potent inhibitor of85 kDa phospholipase A2, blocks production of arachidonate and 12-hydroxyeicosatetraenoic acid by calcium ionophore-challenged platelets. 1. Biol. Chern. 269, 15619-15624. Riese, I., Hoff, T., Nordhoff, A,, DeWitt, D. L., Resch, K., and Kaever, V. (1994).Transient expression of prostaglandin endoperoxide synthase-2 during mouse macrophage activation. I. Leukocyte Biol. 55, 476-482. Ristimaki, A,, Garfinkel, S., Wessendorf, J., Maciag, T., and Hla, T. (1994). Induction of cyclooxygenase-2by interleukin-la: Evidence for post-transcriptional regulation. 1. B i d . Chetn. 269, 11769-11777. Rollins, T. E., and Smith, W. L. (1980). Subcellular localization of prostaglandin-forming cyclooxygenase i n Swiss mouse 3T3 fibroblasts by electron microscopic immunocytochernistry. J. Biol. Chern. 255, 4872-4875.

212

WILLIAM L. SMITH AND DAVID L. D E W I T l

Rome, L. H., and Lands, W. E. M. (1975). Structural requirements for time-dependent inhibition of prostaglandin biosynthesis by anti-inflammatory drugs. Proc. Nutl. Acud Sci. USA 72, 4863-4865. Roshak, A., Sathe, G., and Marshall, L. A. (1994).Suppression of monocyte 85-kDaphospholipase Az by antisense and effects on endotoxin-induced prostaglandin biosynthesis. J. Biol. Chern. 269,25999-26005. Roth, G. J., Machuga, E. T., and Ozols, J. (1983). Isolation and covalent structure of the aspirin-modified, active-site region of prostaglandin synthetase. Biochemistry 22, 46724675. Ruan, K.-H., Wang, L.-H., Wu, K. K., and Kulmacz, R. J. (1993).Amino-terminal topology of thromboxane synthase in the endoplasinic reticulum.]. Biol. Chern. 268,19483-19490. Rzymkiewicz, D., Leingang, K., Baird, N., and Morrison, A. R. (1994). Regulation of prostaglandin endoperoxide synthase gene expression in rat mesangial cells by interleukinlp. Am. J. Physiol. 266(1, Pt. 2), F39-F45. Salvemini, D., Misko, T. P., Masferrer, J. L., Seibert, K., Currie, M. G., and Needleman, P. (1993). Nitric oxide activates cyclooxygenase enzymes. Proc. Nutl. Acad. Sci. USA 90, 7240-7244. Salvemini, D., Seibert, K., Masferrer, J. L., Misko, T. P., Currie, M. G., and Needleman, P. (1994).Endogenous nitric oxide enhances prostaglandin production in a model of renal inflammation.J. Clin. Inuest. 93, 1940-1947. Samet, J. M., Fasano, M. B., Fonteh, A. N., and Chilton, F. H. (1995). Selective induction of prostaglandin G/H synthase I by stem cell factor and dexamethasone in mast cells. I. Biol. Chem. 270, 8044-8049. Sano, H., Hla, T., Maier, J. A. M., Crofford, L. J., Case, J. P., Maciag, T., and Wilder, R. L. (1992).In viva cyclooxygenaseexpression in synovial tissue of patients with rheumatoid arthritis and osteoarthritis and rats with adjuvant and streptococcal cell wall arthritis. J . Clin. Inuest. 89, 97-108. Schievella, A. R., Regier, M. K., Smith, W. L., and Lin, L.-L. (1996). Calcium-mediated translocation of cytosolic phospholipase Az to the nuclear envelope. J. Biol. Chem. 270, 30749-30754. Schnabel, J. (1993). New Alzheimer’s therapy suggested. Science 260, 1719-1720. Schreiber, J., Eling, T. E., and Mason, R. P. (1986). The oxidation of arachidonic acid by the cyclooxygenaseactivity of purified prostaglandin H synthase: Spin trapping of a carboncentered free radical intermediate. Arch. Biochem. Biophys. 249, 126-136. Seibert, K., and Masferrer, J. L. (1994). Role of inducible cyclooxygenase (COX-2) in inflammation. Receptor 4, 17-23. Seibert, K., Zhang, Y., Leahy, K., Hauser, S., Masferrer, J., Perkins, W., Lee, L., and Isakson, P. (1994). Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in inflammation and pain. Proc. Natl. Acad. Sci. USA 91, 12013-12017. Shimokawa,T., and Smith, W. L. (1991). Essential histidines of prostaglandin endoperoxide synthase. J. Biol. Chern. 266, 6168-6173. Shimokawa, T., and Smith, W. L. (1992). Prostaglandin endoperoxide synthase: The aspirin acetylation region. J. Biol. Chem. 267, 12387-12392. Shimokawa, T., Kulmacz, R. J., DeWitt, D. L., and Smith, W. L. (1990). Tyrosine 385 of prostaglandin endoperoxide synthase is required for cyclooxygenase catalysis. J. Biol. Chem. 265,20073-20076. Silvia, W. J., Brockman, J. A,, Kaminski, M. A,, DeWitt, D. L., and Smith, W. L. (1994). Prostaglandin endoperoxide synthase in seminal vesicles. Mol. Androl. 6, 197-207. Simmons, D. L., Levy, D. B., Yannoni, Y., and Erikson, R. L. (1989). Identification of a phorbol ester-repressible v-src-inducible gene. Proc. Nud. Acad. Sci. USA 86,1178-1182.

CYCLOOXYGENASE ISOZYMES

213

Simmons, D. L., Xie, W., Chipman, J. G., and Evett, G. E. (1991). Multiple cyclooxygenases: Cloning of a mitogen-inducible form. In “Prostaglandins, Leukotrienes, Lipoxins, and PAF” (J. M. Bailey, Ed.), pp. 67-78. Plenum Press, New York. Sirois, J. (1994). Induction of prostaglandin endoperoxide synthase-2 by human chorionic gonadotropin in bovine preovulatory follicles in vivo. Endocrinology 135, 841-848. Sirois, J., and Richards, J. S. (1992). Purification and characterization of a novel, distinct isoform of prostaglandin endoperoxide synthase induced by human chorionic gonadotropin in granulosa cells of rat preovulatory follicles. J. Biol. Chem. 267, 6382-6388. Sirois, J., and Richards, J. S. (1993). Transcriptional regulation of the rat prostaglandin endoperoxide synthase 2 gene in granulosa cells. J. Bid. Chem. 268, 21931-21938. Sirois, J., Simmons, D. L., and Richards, J. (1992). Hormonal regulation of messenger ribonucleic acid encoding a novel isoform of prostaglandin endoperoxide H synthase in rat preovulatory follicles. J. Biol. Chern. 267, 11586-11592. Sirois, J., L e y , L. O., Simmons, D. L., and Richards, J. S. (1993). Characterization and hormonal regulation of the promotor of the rat prostaglandin endoperoxide synthase 2 gene in granulosa cells. Identification of functional and protein-binding regions. J . Bid. Chem. 268, 12199-12206. Smith, C. J., Morrow, J. D., Roberts, L. J., and Mamett, L. J. (1993). Differentiation of monocytoid THP-1 cells with phorbol ester induces expression of prostaglandin endoperoxide synthase-1 (Cox-1). Biochem. Biophys. Res. Commun. 192, 787-793. Smith, W. L. (1986).Prostaglandin biosynthesisand its compartmentation in vascular smooth muscle and endothelial cells. In “Annud Review of Physiology” (R. M. Berne, Ed.), Vol. 48, pp. 251-262. Annuals Reviews, Palo Alto, CA. Smith, W. L. (1992). Prostanoid biosynthesis and mechanisms of action. Am. /. Physiol. 263(2, Pt. 21, F181-FI91. Smith, W. L., and Bell, T. G . (1978).Immunohistochemical localization of the prostaglandinforming cyclooxygenase in renal cortex. Am. /. Physiol. 235, F451-F457. Smith, W. L., and DeWitt, D. L. (1995). Biochemistry of prostaglandin endoperoxide H synthase-1 and synthase-2 and their differential susceptibility to nonsteroidal antiinflammatory drugs. Semin. Nephrol. 15, 179-194. Smith, W. L., and Lands, W. E. M. (1972). Oxygenation of polyunsaturated fatty acids during prostaglandin biosynthesis by sheep vesicular glands. Biochemistry 11,3276-3282. Smith, W. L., and Marnett, L. J. (1994). Prostaglandin endoperoxide synthases. In “Metal Ions in Biological Systems” (H. Sigel and A. Sigel, Eds.), Vol. 30, pp. 163-199. Dekker, New York. Smith, W. L., DeWitt, D. L., and Allen, M. L. (1983).Bimodal distribution of the prostaglandin I2 synthase antigen in smooth muscle cells. J. Biol. Chern. 258, 5922-5926. Stanford, N., Roth, G. J., Shen, T. Y., and Majerus, T. Y. (1977).Lackofcovalent modification of prostaglandin synthase (cyclooxygenase)by indomethacin. Prostaglandins 13,669-675. Strieder, S.. Schaible, K., Scherer, H. J., Dietz, R., and Ruf, H. H. (1992). Prostaglandin endoperoxide synthase substituted with manganese protoporphyrin IX. Formation of a higher oxidation state and its relation to cyclooxygenase reaction. J. Bid. Chern. 267, 13870- 13878. Strittmatter, P., Machuga, E. T., and Roth, G. J. (1982). Reduced pyridine nucleotides and cytochrome b5 as electron donors for prostaglandin synthetase reconstituted in dimyristyl phosphatidylcholine vesicles. J. Biol. Chem. 257, 11883-1 1886. Stroebel, M., and Goppelt-Struebe, M. (1994). Signal transduction pathways responsible for serotonin-mediated prostaglandin G/H synthase expression in rat mesangial cells. J. Biol. Chem. 269, 22952-22957.

214

WILLIAM L. SMITH AND DAVID L. DEWI'IT

Tanaka, Y., Ward, S. L., and Smith, W. L. (1987). Immunocheinical and kinetic evidence for two different prostaglandin H-prostaglaudin E isomerases in sheep vesicular gland microsomes. I. Biot. Cherri. 262, 1374-1381. Tang, B. L., Wong, S. H., Low, S. H., and Hang, W. (1992). Retention of a Type I1 surface membrane protein in the endoplasmic reticulum by the Lys-Asp-Gh-Leu sequence. J . Biol. Chem. 267, 7072-7076. Tazawa, R., Xu, X. M., Wu, K. K., and Wang, L. H. (1994). Characterization of the genomic structure, chromosomal location and promoter of human prostaglandin H synthase-2 gene. Biochem. Biophys. Res. Commun. 203, 190-199. Thun, M. J., Namboodiri, M. M., and Heath, C. W. (1991). Aspirin use and reduced risk of fatal colon cancer. N. Engl. J. Med. 325, 1593-1596. Toh, H. (1989). Prostaglandin endoperoxide synthase contains an EGF-like domain. FEBS Lett. 258, 317-319. Topley, N., Petersen, M. M., Mackenzie, R., Neubauer,A., Stylianou, E., Kaever, V., Davies, M., Coles, G. A,, Jorres, A,, and Williams, J. D. (1994).Human peritoneal mesothelial cell prostaglandin synthesis: Induction of cyclooxygenase mRNA by peritoneal macrophagederived cytokines. Kidney Int. 46, 900-909. Tsai, A. L., Palmer, G., and Kulmacz, R. J. (1992). Prostaglandin H synthase: Kinetics of tyrosyl radical formation and of cyclooxygenasecatalysis.]. Biol. Chent. 267,17753-17759. Tsai, A,, Hsi, L. C., Kulmacz, R. J., Palmer, G., and Smith, W. L. (1994a). Characterization of the tyrosyl radicals in ovine prostaglandin H synthase-1 by isotope replacement and site-directed mutagenesis. J. Bid Chem. 269,5085-5091. Tsai, A. L., Wei, C., and Kulmacz, R. J. (1994b). Interaction between nitric oxide and prostaglandin H synthase. Arch. Biochem. Biophys. 313, 367-372. Tsai, A., Kulmacz, R. J.. and Palmer, G. (1995). Spectroscopic evidence for reaction of prostaglandin H synthase-1 tyrosyl radical with arachidonic acid. J. Bid. Chem. 270, 10503-10508. Van der Ouderaa, F. J. G., and Buytenhek, M. (1982). Purification of PGH synthase from sheep vesicular glands. Methods Enzymol. 86, 60-69. van der Ouderaa, F. J., Buytenhek, M., Nugteren, D. H., and Van-Dorp, D. A. (1977). Purification and characterisation of prostaglandin endoperoxide synthetase from sheep vesicular glands. Biochim. Biophys. Acta 487, 315-331. van der Ouderaa, F. J.. Buytenhek, M., Nugteren, D. H., and Van-Dorp, D. A. (1980). Acetylation of prostaglandin endoperoxide synthetase with acetylsalicylic acid. Eur. J. Biochem. 109, 1-8. Wdenga, R. W., Wall, S. F., Setty, B. N., and Stuart, M. J. (1986). Time-dependent inhibition of platelet cyclooxygenaseby indomethacin is slowly reversible. Prostaglandins 31,625-637. Wei, C., Kulmacz, R. J., and Tsai, A. (1995). Comparison of branched-chain and tightly coupled reaction mechanisms for prostaglandin H synthase. Biochemistry 34,8499-8512. Willard, J., Lange, R. A., and Hillis, L. D. (1992). The use of aspirin in ischemic heart disease. N. Engl. J . Med. 327, 175-181. Wong, W. Y. L., DeWitt, D. L., Smith, W. L., and Richards, J. S. (1989). Rapid induction of prostaglandin endoperoxide synthase in rat preovulatory follicles by luteinizing hormone and CAMP is blocked by inhibitors of transcription and translation. Mol. Endocrinol. 3, 1714-1723. Xie, W., and Herschman, H. R. (1995).v-src induces prostaglandin synthase 2 gene expression by activation of the c-Jun N-terminal kinase and the c-Jun transcription factor. J . Biol. Chem. 270, 27622-27628.

CYCLOOXYGEKASE ISOZYMES

215

Xie, W., Chipman, J. G., Robertson, D. L., Erikson, R. L., and Simmons, D. L. (1991). Expression of a niitogen-responsive gene encoding prostaglandin syntliase is regulated by mRNA splicing. Proc. Nntl. Acnd. Sci. USA 88, 2692-2696. Yainagata, K., Andreasson, K. I., Kaufiiiann, W. E., Barnes, C. A,, and Worley, P. F. (1993). Expression of a niitogen-inducible cyclooxygenasein brain neurons: Regulation by synaptic activity and glucocorticoids. Neuron 11, 371-386. Yokoyaina, C., and Tanabe, T. (1989). Cloning of the human gene encoding prostaglandin endoperoxide synthase and primary structure of the enzyme. Biachem. Bic@ys. Res. Corntn1Ln. 165, 888-894. Yokoyaina. C., Takai, T., and Tanabe, T. (1988). Primary structure of sheep prostaglandin endoperoxide synthase deduced from cDNA sequence. FEBS Lett. 231, 347-351. This article was accepted for publication on 28 September 1995

This Page Intentionally Left Blank

ADVANCES IN IMMUNOLOGY, VOL 62

Human Tumor Antigens Are Ready to Fly ROBERT A. HENDERSONAND OLlMRA 1. FINN h p a ~ noft Molecular Genefics and Biochemishy, Universivof Pitkbugh School of Medicine, Pitkburgh, Pennsylvania 15260

1. introduction

The immune system is fully capable of successfully eliminating harmful or infectious microbes through both humoral and cellular mechanisms. Its function may also be broadened to include detection of any sort of “danger” that threatens the integrity of an organism (Matzinger, 1994). Why is it then that tumors, which are clearly detrimental to survival, appear to go unchecked by the immune system? It is now known that the reason is not the lack of antigens recognized by either antibodies or T cells (Urban and Schreiber, 1992). Many factors influencing the immune response to tumors have begun to be appreciated, including the quality of the antigens themselves. From studies in mice, it has been known for many years that the immune system mounts a response to tumors and that immunization against certain tumors is indeed possible. Early studies in humans also demonstrated the presence of lymphocytes that were reactive against or inhibited the growth of autologous tumors (Hellstrom & Hellstrom, 1969; Vanky and Klein, 1982). Despite these observations, it was clear that most tumors had only low or undetectable immunogenicity. Their immunogenicity, however, could be enhanced by various means including, but not limited to, infection with virus (for review see Schirrmacher, 1992), the use of adjuvants, or the insertion of genes encoding either cytokines or T cell costimulatory molecules (for review see Pardoll, 1993; Hellstrom et al., 1995). The important observation derived from this line of investigationwas that tumor antigens did exist, although they were not necessarily strongly immunogenic. Molecular characterization and isolation of tumor antigens that are potentially immunogenic in humans became an important goal of tumor immunologists. The incentive was the ability to use purified antigens to create effective immunogens and thus increase the immune response to a level necessary to reject the tumor. We will review our current knowledge of tumor antigens recognized by both humoral and cellular arms of the 217 Copright 0 1996 by Academic Press, Inc.

All rights of reproduction in any form reserved.

218

ROBERT A. HENDERSON AND OLIVERA J. FINN

immune system and the promise that some of these antigens currently hold for cancer immunotherapy. II. Tumor Antigens Defined by Antibodies

The search for cell surface molecules that may immunologically distinguish tumor cells from their normal counterparts has been a constant process in tumor immunology, which greatly intensified in the late 1970s with the development of the hybridoma technology (Kohler & Milstein, 1976). To the dismay of many tumor immunologists, after almost two decades of experiments this powerful technology yielded very few molecules that showed tumor-specific expression and could be justifiably called “tumor-specific antigens” (Lloyd, 1990; Old, 1981). Moreover, the majority of these tumor-specific molecules were unique to a single tumor (for review see Schreiber et al., 1988) thus being of limited value for examining tumorspecific immune responses in general. Nevertheless, they confirmed the notion that tumor specific-antigens did exist and a further search might in fact uncover some that are shared by many tumors. When “shared tumor antigens” began to be identified, they were found to be less than tumor specific. They showed high-level expression on tumors, whereas on normal tissues they were either expressed very weakly or with a very restricted distribution; thus, these antigens were referred to as “tumor-associated’ antigens. The immunogenic potential of such antigens and the ability of the immune system to target them on tumors and not on normal tissue has not been fully evaluated to this day. Furthermore, inasmuch as these antigens on tumors were almost always detected with monoclonal antibodies generated in another species, for example, human tumor antigens detected by mouse monoclonal antibodies, the immunogenic potential in the species of tumor origin of many of these antigens has remained unknown. For a small number of them, however, these questions have been addressed and information concerning their immunogenicity is available. We will limit this review to discussing only these more fully characterized molecules. They are primarily human tumor antigens. They are also good representative examples of distinct classes of molecules into which the majority of antigens described so far can be placed. The knowledge about tumor-associated antigens detected by antibodies can be and has been used in several ways: (a) the presence of the antigen has been used for detection and diagnosis of malignancy; (b) the antibody specific for the antigen has been conjugated to toxins, drugs, or radioisotopes and used for tumor therapy; (c) the antibodies have been made bispecific or multispecific and used to bring other effector mechanisms, like T cell, natural killer (NK) cells, and macrophages, to the tumor site;

HUMAN TUMOR ANTIGENS ARE READY TO FLY

219

and (d) tumor-associated antigens identified as growth factor receptors or other types of signaling molecules have been used to induce tumor cell death or differentiation when bound by the specific antibody. These uses are all grouped under the term “passive immunotherapy” and have the drawback that repeated treatments are necessary and very high concentrations of tumor-specific antibodies are required for each treatment to see a limited effect (Roth, 1986; Kumer and Staerz, 1993; Waldman, 1991; Vitetta and Uhr, 1994). “Active immunotherapy” is perceived to be more desirable based on the hypothesis that induction of an immune response to the tumor antigen, if it could be achieved, would be long lived and would involve multiple components of the immune system. Active immunotherapy, however, is contingent on the immunogenicity of the tumor antigen in its host. Considering that the best known tumor antigens are “tumor associated’ rather than “tumor specific,”and that the host’s immune system has seen the antigen on normal tissue either during development or chronically throughout life, the magnitude of the immune response that can be generated to these molecules is hard to predict. Common immunological wisdom would in fact predict full tolerance to such molecules, but the experimental data appear to show otherwise. A CARBOHYDRATE ANTIGENS Carbohydrate antigens are present on tumor cells as either glycolipids or glycoproteins with the anti-tumor antibody specificity residing in their sugar moiety (Lloyd and Old, 1989).The best studied carbohydrate antigens are gangliosides and blood group antigens (Hakamori, 1985). 1 . Gangliosides These cell surface molecules are neuraminic acid-containing glycosphingolipids that consist of an oligosaccharide chain linked to the ceramide moiety that anchors them into the plasma membrane. They are abundantly expressed on a number of tumors, especially those of neuroectodennal origin-most notably melanomas, astrocytomas, and neuroblastomas-but are also expressed in the brain and other neural crest-derived tissues. As an example, malignant melanomas express GM3 (NeuAca2+ 3GalP1+ 4GlcCer) and GD3 (NeuAccr -+ 8NeuAccu + 3GalP1 + 4Glc-Cer) as their predominant gangliosides and, in addition, varying amounts of GM2 (GalNacPl + 4NeuAccu2 + 3GalPl + 4Glc-Cer) and GD2 (GalNacPl + 4NeuAca2 4 8NeuAccr2+ 3GalP1+ 4Glc-Cer), whereas normal melanocytes express mostly GM3 (Puke1et al., 1982; Carubia et al., 1984). These molecules are expressed on tumor cell membrane at very high density and for that reason deemed potentially good targets of an immune response. Although not tumor specific, their restricted distribution on normal tissue,

220

ROBERT A. HENDERSON AND OLIVERA J. FINN

or in an immunopriviledged site (brain), appeared to allow at least an operational window of specificity. The potential of these molecules to serve as targets of an immune response was initially investigated by looking for the presence of anti-ganglioside antibodies in sera from cancer patients and healthy individuals. Antibodies against GM2 and GD2 gangliosides were found in some healthy donors and in some melanoma patients (Lloyd, 1991), indicating that, at least at the B cell level, there was no tolerance to these molecules. The IgM isotype predominated in these responses indicating a T cell-independent response, not unexpected for carbohydratespecific antibodies. Furthermore, some patients immunized with whole melanoma cell vaccine generated anti-ganglioside antibodies (Tai et al., 1985). Based on these results, there has been an ongoing effort to evaluate ganglioside-based cancer vaccines. GM2 has received most attention (Livingston et aZ.,1987),even though GD2 and GD3 have also been considered. A number of clinical trials have been conducted to evaluate the conditions under which their immunogenicity could be enhanced (for review see Livingston, 1995). The most exciting results were derived from vaccination trials involving GMWBCG vaccine. Most patients developed IgM antibodies to GM2, and those with the highest antibody titer also showed longer disease-free interval and overall survival (Livingston et al., 1989, 1994). Despite this success, there is a perception that the predominantly IgM response induced with this vaccine may be an inferior response, and that induction of IgG antibodies may enhance the effect of GM2 vaccines. For that reason, GM2 has been conjugated to a carrier protein keyhole limpet hemocyanin (KLH) and administered with several other adjuvants (Helling et al., 1995). Initial studies indicate that conjugation to KLH induced a T helper cell response, the result of which was the antibody isotype switch to IgG. This is expected to provide additional effector mechanisms to the anti-tumor response, and a longer lasting antibody response. In turn, the expectation is that a more impressive clinical response will follow. 2. Blood Group Antigens These were primarily defined as tumor-associated antigens by antibody reactivity with epithelial tumors. Blood group antigens are expressed on hematopoietic and epithelial cells and their expression depends on glycotransferase activity, which in turn depends on the stage of cell differentiation or on malignant transformation (Lloyd, 1987; Cordon-Cardo et al., 1988; Gold and Mattes, 1988; Schuessler et al., 1991; Springer, 1984; Springer et al., 1995).The tumor-associated epitopes are created by carbohydrate chains decorating the polypeptide core primarily of the mucin family of glycoprotein molecules. The best studied have been the T (Thomsen-Friendenreich) and sialylated Tn (sTn) antigens that have been shown

HUMAN TUMOR ANTIGENS ARE READY TO FLY

22 1

to be immunogenic in cancer patients (O’Boyle et al., 1992). Based on serum antibody reactivity with these molecules and a number of murine studies, clinical trials have been initiated to explore the use of chemically synthesized forms of these antigens as cancer vaccines (MacLean et al., 1992, 1993). The ability to make synthetic forms of these carbohydrates is an advantage as long as the synthetic versions mimic faithfully the naturally occurring epitope on the tumor cell (Zhang et al., 1995). As is the case with gangliosides, these molecules also require conjugation to carrier proteins and the use of strong adjuvants to elicit antibody responses. One drawback that is attributed to carbohydrate vaccines is the general difficulty in inducing carbohydrate-specific memory responses, mostly due to the lack of T cells that can recognize carbohydrates. This is presented as a serious problem considering that cancer may recur several years after vaccination, and a secondary response capable of dealing with the tumor recurrence would be the only desirable outcome of a tumor vaccine protocol. Evidence suggests that the ganglioside GD2 is a target for cytotoxic T lymphocytes (Zhao and Cheung, 1995). It is possible that helper T lymphocytes capable of reacting with oligosaccharides will be identified eventually. This may imply that induction of memory in this cell population may indeed be possible.

B. PROTEIN ANTIGENS 1 . Carcinoemby o n i c Antigen (CEA) CEA is one of the most broadly expressed molecules on human malignancies. It is also a tumor antigen that has been most extensively studied in terms of both its molecular characteristics and its potential application for diagnosis and treatment of cancer (Shievely and Beatty, 1985; Thomson et al., 1991).It is a cell surface glycoprotein of 180-kDa molecular weight, and numerous antibodies have been generated that react with either the polypeptide core or the carbohydrate determinants on this molecule. Expression of CEA is most often associated with colon, breast, and lung adenocarcinomas, but it is likely that most tumor cells of epithelial origin express this antigen. It is also expressed on normal epithelia, most abundantly in the colon, but also on other normal epithelial and endothelial cells (Majuri et al., 1994). However, in addition to drasticdy different levels of expression, there are often differences in the processing of this molecule between normal and malignant cells (Muraro et al., 1985) or between tumors originating from different organs (Hernando et al., 1994). Taking advantage of these differences may increase tumor specificity of this antigen. Even though a lot has been done with this molecule to test its immunogenicity and the consequence of the anti-CEA immune response

222

ROBERT A. HENDERSON AND OLlVERA J. FINN

on the growth of CEA-expressing tumor cells, the majority of the work has utilized human CEA in species other than humans (Hand et al., 1993; Kantor et al., 1992; Conry et al., 1994, 1995). Despite a relatively high degree of conservation between species, the human CEA is also considerably different from that of other species and those differences could be entirely responsible for its immunogenicity. Derivation of CEA transgenic mice may provide a better system for evaluating CEA immunogenicity in an animal model (Eades-Perner and Zemmerman, 1995).Concurrently, a clinical trial under way at the University of Alabama is showing preliminary evidence that CEA is immunogenic in cancer patients (A. LoBuglio and J. Schlom, personal communication). Somewhat surprisingly, all the immunization protocols have utilized some form of a recombinant molecule, expression system, or simply the CEA DNA. In all of them, the CEA is expected to be expressed by normal cells, most likely muscle cells, and yet it is hoped that the immunity generated will preferentially target tumor cells. The reasons for choosing this unaltered form of CEA as an immunogen are difficult to appreciate. It would appear more rational to re-create as an immunogen tumor-like CEA, considering that differences in CEA expression between tumor and normal cells are not simply quantitative but also qualitative. Quantitative differences may be important in the elicitation phase of an immune response, in which perhaps low levels of CEA on normal tissues are ignored by the immune system. Once the immune system is fully activated by the CEA vaccine, it is questionable if quantitative differences will remain important. On the other hand, any difference in the processing of CEA that may yield a new tumor-specific epitope, if re-created in the vaccine, might provide a required degree of tumor specificity to the activated immune response.

2. Mucins Advantages and disadvantages of CEA as a tumor antigen are very similar to those of another family of molecules, tumor-associated mucins. These molecules are expressed at very high levels on the surface of epithelial cell tumors, primarily breast, pancreas, colon, ovary, and lung adenocarcinomas (Zotter et al., 1988), and detected by numerous mouse monoclonal antibodies generated by immunization with these human tumors (TaylorPapdimitriou, 1991). They are also expressed on normal tissues, usually at lower densities, and with a different repertoire of specific epitopes (Burchell et al., 1983). Some of the epitopes appear exquisitely tumor specific (Girling et al., 1989). Mucins are very-high-molecular-weight glycoproteins that share the structural characteristic of being composed of numerous tandem repeats of varying lengths, depending on the particular mucin, and being heavily glycosylated on the repeats with O-linked carbohydrates.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

223

The best studied, and perhaps most promising as a tumor antigen, is the mucin MUC-1, also known as episialin, polymorphic epithelial mucin, or human tnilk fat globule, and encoded by the gene MUC-1 (Gendler et al., 1990). This is the only mucin molecule that is an integral membrane protein, others being secreted, and its normal tissue expression is confined to the apical surface of ductal epithelial cells. It is thus sequestered from the immune system, especially the humoral arm, throughout adult life. Malignant transformation of the ductal epithelial cells that gives rise to the previously mentioned tumors changes the normal pattern of MUC-1 expression. Tumor cells express this molecule in high amounts on the cell surface, the expression is not polarized, and furthermore, processing differences begin to uncover tumor-specific epitopes (for review see Finn et al., 1995). Some of these new epitopes are carbohydrate in nature and belong to the previously discussed blood group antigens. The MUC-1 unique epitopes preferentially expressed on tumor cells are located on the polypeptide core and exposed as a result of incomplete or aberrant glycosylation. These polypeptide epitopes are apparently immunogenic, as antibodies against them have been detected in cancer patients (Rughetti et al., 1993; Kotera et al., 1994). As we will discuss later, there is also evidence of a T cell response against the same or a similar polypeptide core epitope. This has prompted a number of attempts at induction of an efficient antitumor immune response using mucin immunogens. These attempts have been successful, although it is not clear how informative. With the exception of one study in which immune response to human MUC- 1,thought to be identical to the chimpanzee molecule,was generated in chimpanzees using autologous B cells transfected with the human MUC-1 cDNA (Pecher and Finn, submitted for publication), all other studies have been performed in mice (Hareuveni et al., 1990; Acres et al., 1993; Ding et nl., 1993; Apostolopoulos et nl., 1994).These studies have led to induction of immunity and protection from MUC-l-expressing tumors. Whether similar results can be achieved in cancer patients is unknown, and it is currently being tested in several clinical trials.

3. HER-2lneu This cell surface protein is a product of a protooncogene expressed in breast, ovarian, and several other types of epithelial tumors. It has homology to the epidermal growth factor receptor, and there is considerable evidence that this molecule can also function as a target for negative signaling (Stancovaski et nl., 1991; Drebin et al., 1988). Anti-HER-Yneu antibodies are thought to mimic the natural ligands of this molecule, several of which have been described (Peles et nl., 1992; Lupu et al., 1992). In addition, there is great interest in immunological targeting of tumors via this mole-

224

ROBERT A. HENDERSON AND OLIVERA J. FINN

cule. Similar to the previously discussed tumor antigens, HER-Yneu is characterized by only an increased rather than specific expression on tumors. It is weakly expressed on normal tissues (Press et al., 1990) and the challenge remains to use this molecule as an immunogen in a way that will induce tumor immunity and not autoimmunity. Induction of immunity in rats by immunization with purified rat HER-Z’neu does not appear to exert any untoward effect on normal rat tissues that express this antigen (Cheever et al., 1995). Although these experiments convincingly show the lack of tolerance to this autoantigen, they cannot appropriately evaluate the quality of the immune response. Rat tumor-rejection studies attempting to show that tumor rejection can be effected without damage to normal tissue have not yet been attempted. Expression of this antigen on normal human tissues also fails to induce tolerance, as both antibodies and T cells specific for this molecule can be found in cancer patients and, albeit more rarely, in healthy subjects (Disis et al., 1994). This encourages further manipulations of this antigen or its gene to be used as an anticancer vaccine, and several clinical trials based on this immunogen are pending. 4. Prostate-Spec@ Antigen (PSA) PSA was thought to be produced exclusively by the epithelial cells of the prostate, and its presence in the circulation at elevated levels has been used for diagnosis of prostate cancer (Wanget al., 1981). It is a glycoprotein of 33-kDa molecular weight and it is a member of the serine protease family with trypsin-like and chymotrypsin-like protease activity (Wat et al., 1986). It has become clear that PSA is a marker of steroid hormone action in many normal tissues and in tumors derived from these tissues. Androgenic hormones increase PSA transcription in prostate tumor cell lines (Hentu et al., 1992), and glucocorticoids, progestins, as well as androgens increase PSA in breast cancer cell lines (Yu et al., 1994). Moreover, PSA can be found in milk of lactating women and in amniotic fluid, and its production by normal breast can be stimulated by oral contraceptives (Yu and Diamandis, 1995a,b; Yu et al., 1995a). PSA expression has now been documented in various other tumors in addition to prostate-most frequently in breast, and some tumors of the skin, ovary, and salivary gland (Levesque et al., 1995).There is much interest in this molecule as apossible immunogen but no information is yet avadable on that subject. There is also no evidence yet of possible differences in the PSA molecules expressed on tumors versus the normal tissues. An interesting observation was made that the expression of PSA is a favorable prognostic factor in women with breast cancer (Yu et al., 1995b). This may be one reason to pursue the study of this molecule at the immunological level, others being that it is

H U M A N T U M O R ANTIGENS ARE READY TO FLY

22s

expressed on some very important human malignancies and its expression can be upregulated, providing the possibility of high levels of antigen.

S. Idiotypes A protein that has a unique property of being a normal molecule and yet absolutely tumor specific is the immunoglobulin receptor on B cell tumors that carries the clonally expressed idiotype, the unique antigen recognition site. This tumor antigen has been the target of passive immunotherapy with anti-idiotypic antibodies (Levy and Miller, 1990).In addition to the ability of the anti-idiotype treatment to engage the conventional effector mechanisms for tumor cell destruction, like complement and ADCC, the anti-idiotypic antibody also induced Ig signal transduction that correlated with tumor regression (Vuist et al., 1994).There are a number of other receptors, some that are members of the Ig signaling complex, like CD19 and CD20, or the growth factor receptors on epithelial carcinomas that, when targeted with monoclonal antibodies, induce tumor regression. This nonimmunological effect is considered to result in the induction of tumor dormancy through negative signaling (for review see Vitetta and Uhr, 1994).If the presence of circulating antibodies against these molecules is necessary to maintain tumor dormancy, then clearly passive immunotherapy involving repeated infusions of antibody cannot remain the treatment of choice. For this reason, the potential of these molecules to be immunogenic and to induce antibodes in vivo becomes very important. This question has been explored for the B cell lymphoma idiotypes in animal models and in patients. Immunization with purified idiotype protein under several different conditions leads to the development of anti-idiotypic immunity in vivo that is associated with tumor regression (Kaminski et al., 1987; Flamand et al., 1994; Kwak et al., 1992). The major drawback to the idiotype as a tumor antigen is the emergence of idiotype variants that lose reactivity with the antibody (Meeker et al., 1985). In that respect, active immunization with idiotype protein may have an advantage over passive therapy with infused anti-idiotypic antibody. Idiotype vaccination would likely generate a polyclonal response that will be less sensitive to single amino acid mutations. In addition to generating anti-idiotype antibodies that may induce tumor cell dormancy through negative signaling, active immunization has also been seen to generate idiotype-specific T cell responses that should provide a second antitumor effector mechanism, hopefully one with long-term memory. Idiotypic DNA vaccines have been developed and their efficacy has been tested in animal models. DNA alone has been shown to induce anti-idiotypic antibodies, but the intensity of the response is increased with the addition of plasmids encoding various cytokines (for review see Stevenson et al., 1995).

226

ROBERT A. HENDERSON AND OLIVERA J , FINN

C. MUTATED ONCOGENE PRODUCTS AS TUMOR ANTIGENS Protein products of oncogenes could have been Gods gift to tumor immunologists. Not only were they critically associated with transformation and therefore expected to be present in all tumor cells as targets, but they were also products of mutated genes and the resulting mutation in the protein could be expected to elicit the response from the immune system. Not fully appreciated on their immediate discovery because most of them were not cell surface proteins, they entered center stage when it was realized that T cells could detect their presence by recognizing the mutated peptides bound to class I MHC molecules. Unfortunately, the results of several years of investigation do not support the initial enthusiasm. We will address the reasons for this as we discuss below the three most studied oncoproteins.

1 . p53 The tumor-suppressor gene $3 is located on the short arm of chromosome 17 and serves as a negative regulator of the cell cycle. By arresting the cells in G1, it allows DNA repair (Hollstein et al., 1991; Vogelstein and Kinzler, 1992). It is thought to play an important role in suppression of malignant transformation (Finlay et al., 1989). Mutations in the p53 gene are very frequently found in human tumors, primarily with antibodies which detect high level of expression of the mutated p53 protein. The wild-type p53 protein is mostly undetectable in normal cells. The reason for this difference appears to be that subtle mutations anywhere in the p53 protein induce a profound conformational change resulting in the longer half-life, accumulation to high levels in the cytoplasm, and exposure of epitopes not accessible on the wild-type protein (Steven and Lane, 1992). All these changes are clearly noticed by the immune system as evidenced by the presence of anti-p53 antibodies in patients with breast cancer (Crawford et al., 1982; Davidoff et al., 1992), pancreatic cancer (Marxen et al., 1994), and several other cancers (Angelopoulouet al., 1994; Labrecque et al., 1993). Interestingly, antibodies are directed not only against the mutated portions of the protein (Winter et al., 1992), but also against the wild-type protein epitopes (Schlichtholtz et al., 1992). Moreover, not all patients whose tumors contain p53 mutations develop antibodies. This may be due to different outcomes of different mutations (Winter et al., 1992) or in some instances to the tumor type. Glioblastoma patients, for example, have a characteristic absence of anti-p53 antibodies in their sera despite a high percentage of glioblastoma carrying the p53 mutations (Rainovet al., 1995).There appears to be no clinical correlation between either the absence or the presence of the antibody and the course of the disease. The antibody against this tumor antigen, as in the case of

HUMAN TUMOR ANTIGENS ARE READY TO FLY

227

the previously dscussed molecules, is for now simply an indication that the immune response can be elicited. The reason to pursue immunity to p53 as a possible antitumor treatment is the hypothesis that anti-p53 responses induced in the patient through the tumor presence only may not be as effective as responses induced by vaccination. Unlike the molecules previously discussed, which are all expressed on the cell surface and could be targets for specific antibodies, p53 is an intracellular protein, and it is not immediately obvious what is the utility of the anti-p53 humoral response. It could be postulated that the presence of antibody is merely diagnostic of another ongoing immune response carried out by T cells that may be causing tumor destruction and the release of the p53 protein. As we will show later, T cells specific for p53, both wild-type and mutated forms, can be generated, but their ability to recognize tumor cells has not yet been convincingly shown. Despite this, not all enthusiasm for p53-based tumor immunotherapy has been extinguished. Immunization with p53 peptides is a part of an ongoing phase I clinical trial at the National Cancer Institute.

2. c-myc and c-myb Two other oncogene products, c-myc and c-myb, both nuclear proteins and both associated with the control of cellular growth and differentiation, have also elevated expression in a wide variety of human leukemias and in some solid tumors. Antibodies specific for these proteins have been found in cancer patients (Ben-Mahrez et al., 1988, Sirokine et al., 1991). As with p53, the significance of this humoral response is not clear, and its generation may be secondary to another antitumor response the target of which may be another tumor antigen. These responses to intracellular or nuclear proteins are reminiscent of humoral responses monitored in autoimmune diseases (Tan, 1989). There is usually a strong correlation between the presence of antibody and the disease. The antibody in most instances appears not to contribute to the pathology of the disease but rather to be a consequence of an ongoing d'isease. Because experiments in animal models appeared to indicate that antibodies alone could not elicit tumor rejection, it became important to identify antigens on tumors that could serve as targets for cellular immunity. This area of research received a great boost with the development of new methods for growing tumor-specific T cell lines and clones and a better understanding of the process of antigen recognition by T cells. 111. Tumor Antigens Defined by T Cells

The majority of tumor antigens identified to date that are targets of cellular immunity have been those recognized by CD8+ tumor-specific

228

ROBERT A. HENDERSON AND OLIVERA J. FINN

cytotoxic T lymphocytes (CTLs). Identification of tumor antigens recognized by CD4’ tumor-specific T cells has also been accomplished. A number of approaches have been employed for the identification of these antigens. Most have utilized patient-derived CTLs that specifically recognize various tumors as a tool to identify tumor-specific antigens. Tumor-specific CTLs are usually isolated by culturing in vitro irradiated tumor cells with peripheral blood lymphocytes,tumor-draining lymph node cells, or tumor-infiltrating lymphocytes with the T cell growth factor IL-2. These CTL lines recognize the tumor but display minimal reactivity with “normal” tissues such as autologous B cells or fibroblasts. In addition, these CTLs do not kill the NK cell target, the erythroleukemia cell line K562. The largest number of antigens so far derived is from melanoma tumors using melanoma-specific T cells. This is in part due to the relative ease in establishing melanoma tumor cell lines that can be used for expansion of melanoma-specific T cells in vitro. Only a limited number of T cells specific for other tumors have been developed, and consenquently fewer antigens have been identified in those systems. The antigens and their most important characteristics are listed in Table I. The first and most successful approach to the identification of a gene encoding a tumor antigen utilized melanoma-specific CTLs in combination with gene transfection. This is now known as the “genetic approach.” This approach identified the melanoma tumor cell antigen MZ2-E (van der Bruggen et al., 1991). Briefly, a cosmid library was prepared with the DNA from a melanoma tumor cell line that expressed this antigen. The cosmids were transfected into a melanoma tumor cell line that had lost the expression of the MZ2-E antigen but still expressed the restricting class I MHC molecule. The transfected cells were screened for their ability to be recognized by a CTL clone specific for this antigen. By retrieving the transfected DNA, repeated rounds of transfection using smaller fragments of originally transfected DNA were possible thus leading to the identification of the gene encoding the antigen MAGE-1. By transfecting even smaller fragments of this gene and through the use of synthetic peptides, the antigen MZ2-E was identified as a nine-amino-acid peptide with the sequence EADPTGHSY (Traversari et al., 1992). Based on previous work (Seed and Aruffo, 1987), this approach has been modified to use COS-7 cells transiently transfected with cDNA libraries cloned into expression vectors containing the SV40 origin of replication (Brichard et al., 1993). The plasmid containing the cDNA insert replicates to a very high copy number in the transfected COS-7 cells that have been stably transfected with the appropriate human class I MHC molecule. Transfectants containing the tumor antigen are identified by their ability to stimulate cytokine release (usually TNF-a) from tumor-specific T cells. Once the gene encoding the antigen has

TABLE I TUMOR ANTIGENS RECOGNIZED BY HUMAN T CELLS ~

Name (gene or protein) MAGE-1 MACE3

HLA restriction A1 Cw16 A1

A2 BAGE GAGE Tyrosinase

Cw16 CW6

A2 A2 A24 B44 DR4

MART-UMELAN-A GP100/Pme117

A2 A2 A2 A2 A2

A2 GP75 MUM-I HERWneu

A2 A31 B44

HPV-16

A2 A2 A2 A2 A2

MUC-1

Unrestricted

~~

Peptide sequence

Normal hssue

Tumors

EADPTGHSY SAYGEPRKL EVDPIGHLY FLWGPRALV AARAVFLAL YRPRPRRY MLLAVLYCL YMDGTMSQV AFLPWHRLF SEIWRDIDF ND* AAGIGILTV ILTVILGVL YLEPGPVTA LLDGTATLRL KTWGQYWQV ITDQWFSV VLYRYGSFSV MSLQRQFLR EEKLIWLF IISAWGIL KIFGSLAFL YMLDLQPE'IT LLMGTLGIV TLGIVCPI -PDTRP-

Testis Testis Testis Testis Testis Testis Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes Melanocytes NE" Epithelial cells Epithelial cells NE NE NE NE

Variable" Variable Variable Variable Variable Variable Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomas Melanomad Breast and ovarian tumors Breast and ovarian tumors Cervical carcinomas Cervical carcinomas Cervical carcinomas Epithelial cell tumors

This antigen is expressed in variable numbers of tumors of different histological origin. Not determined. This antigen is not expressed in normal tissues. This antigen is unique and is expressed in only a single individual's tumor.

230

ROBERT A. HENDERSON AND OLIVERA J FINN

been identified, the epitope recognized by the CTL can be established by transfecting gene fragments and using synthetic peptides to sensitize target cells for lysis by the CTL. This methodology has been very successful in identifying the majority of tumor antigens recognized by T lymphocytes to date. One of the advantages of this approach is that it leads not only to the identification of the peptide antigen but also to the gene that encodes it. However, there are some disadvantagesto this approach as well. Because of the high level of expression of the transfected genes in COS cells, it is possible to identify cross-reacting peptides that are recognized by CTL because of their high density of expression on MHC molecules. Furthermore, as we will discuss later, the structure of the naturally processed peptide antigen presented by the tumor may be post-translationally modified differently in COS cells than in tumors. This may be an important consideration as it has been shown that glycosylation may affect the recognition of both class I- and class II-presented peptides (Haurum et al., 1994; Michaelsson et al., 1994). An alternative approach to the isolation of tumor antigens recognized by T cells has been by direct identification and sequencing of the relevant peptide antigen by biochemical techniques. In this “peptide elution” approach, antigenic peptides are isolated from class I MHC molecules by acid denaturation and fractionated by reverse-phase high-performance liquid chromatography (RP-HPLC).The fractions are then tested for their ability to sensitize target cells expressing the appropriate class I MHC molecule for lysis by the relevant CTL. Those fractions that sensitize target cells are fractionated further by RP-HPLC and are then analyzed by tandem mass spectrometry. The advantage of tandem mass spectrometry, as opposed to Edman degradation, is that it is capable of identifying both the mass and the amino acid sequence of peptides even when they are present in complex mixtures. Thus, a number of candidate peptides that may correspond to the antigenic peptide can be identified and sequenced. These peptides are then synthesized and tested individually to confirm the exact antigenic sequence. This approach has been successfully employed by two groups to identify melanoma tumor antigens (Cox et al., 1994; Castelli et al., 1995). One of the major advantages of this approach is that it clearly identifies what the naturally processed form of the peptide tumor antigen is and can potentially identify post-translational modifications. However, the gene that encodes that antigen is not identified by this approach. In addition, this approach requires large numbers of tumor cells from which to isolate the class I-associated peptides and requires relatively complex technology to carry out the mass spectrometric analysis. One additional approach for the identification of these antigens has been to focus on peptides derived from cellular proteins that have the potential

HUMAN TUMOR ANTIGENS AHE HEADY TO FLY

23 1

of being tumor antigens. Peptides that bind to class I MHC molecules have been shown to have a length of 9-12 amino acids and a characteristic “motif” of conserved amino acids, so-called “anchor residues,” which serve to facilitate the binding of the peptide to a particular MHC molecule. The “anchor motifs” for several class I and class I1 MHC molecules have been identified (Engelhard, 1994). This information allows one to analyze the sequence of proteins and to predict which peptides from the protein may bind to a particular class I MHC molecule. The peptides can then be synthesized and tested in in vitro binding assays for their ability to bind to a particular MHC allele. Those peptides that are shown to be capable of binding, preferentially with high affinity, are then screened for their ability to induce CTL responses in vitro or to reconstitute the epitope of tumor-specific CTL. This approach has been used to screen various oncogenic proteins, tumor and tissue differentiation antigens, as well as viral proteins expressed by virus-transformed tumors. It is important to note that peptide-specific CTLs thus elicited must be further tested for their ability to recognize tumors that express both the appropriate MHC molecule and the gene encoding the antigen for this peptide to be identified as a tumor antigen. The advantage of this approach is that it allows rapid screening of a large number of potential antigens. However, this approach is limited to known proteins and to those class I MHC molecules whose motif is known. In addition, the methods for generating peptide-specific CTLs on peptide-loaded antigen-presenting cells (APCs) may lead preferentially to the generation of low-affinity T cells that may not be able to recognize the relatively low amounts of naturally processed peptide present on the tumor cell surface, even if the peptide is correctly processed and presented by the tumor. A. MAGE, BAGE, AND GAGE These genes encode tumor antigens expressed on melanoma cells and recognized by autologous CTLs. They are expressed in various other tumors in addition to melanomas but not expressed in normal tissues except for testis and placenta. Because of the reactivation of their expression in tumor cells, it has been speculated that these antigens may represent examples of oncofetal antigens, i.e., products of genes that are normally transcribed during embryonal development, silent in adult life, but reexpressed during malignant transformation. 1. MAGE

As discussed above, one of the first human tumor antigens recognized by T cells to be discovered was the melanoma tumor antigen, MZ2-E, which was found to be a nine-amino acid peptide EADPTGHSY encoded

232

ROBERT A. HENDERSON AND OLIVERA J , FINN

by the MAGE-1 gene and restricted by the class I MHC molecule HLAA 1 (Traversari et al., 1992). This gene was found to be homologous to at least two other nonidentical genes that were termed MAGE-2 and MAGE3. The exsistence of such similar genes to MAGE-1 implied the exsistence of a family of related genes. At this point in time, there are at least 12 related genes in the MAGE family (De Plaen et al., 1994) and there may be others elsewhere in the genome (Muscatell et at., 1996). MAGE-1 also encodes a second peptide antigen (SAYGEPRKL) recognized by melanoma-specific T cells restricted by HLA-Cw01601 (van der Bruggen et al., 1994). This is not a surprising result because other antigens have been shown to be capable of providing multiple peptides that can be presented by different class I MHC molecules. The novel finding is that it is restricted by an HLA-C molecule. Very few antigens to date have been found that are restricted by HLA-C molecules compared to HLA-A or HLA-B, and this finding underscores the possibility that these molecules may also play a more important role in immunosurveillance than previously suspected, despite their relatively low expression on the cell surface. As we will see below, although this was the first, this is not the only tumor peptide restricted by an HLA-C molecule. The related gene, MAGE-3, encodes several tumor antigens as well. The MAGE-3 antigens are restricted by both HLA-A1 (EVDPIGHLY) (Gaugler et al., 1994) and HLAA2 (FLWGPRALV) (van der Bruggen et al., 1994). Except for these two genes, no other member of the MAGE family has been reported to encode an antigen that can be recognized by CTLs. However, it is possible that there are other immunogenic peptides that can be derived from these two genes as peptides derived from MAGE-1 have been shown to be capable of binding to multiple class I MHC molecules (Celis et al., 1994a). In normal tissues, the expression of the MAGE genes is limited to the placenta and testis as determined by PCR analysis of a large number of normal adult tissues and a few tissues derived from >20-week-old fetuses (Van Pel et at., 1995). In tumor cells, the expression of these genes is somewhat variable between tumors of different histological origins, but it is clear that some proportion of melanomas, bladder carcinomas, mammary carcinomas, squamous cell carcinomas, non-small cell lung carcinomas, sarcomas, and prostatic carcinomas express a number of the MAGE genes. Interestingly, these genes do not appear to be expressed well in colon or rectal carcinomas and do not appear to be expressed at all in leukemias or lymphomas. It is also of interest to note that the frequency of expression of MAGE-1, -2, -3, and -4in cutaneous melanoma is higher in metastatic lesions than in primary lesions (Brasseur et al., 1996), which is important from the prospect of immunotherapy of metastatic disease.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

233

2. BAGE and GAGE Two other gene families with no sequence homology to any other known genes have been reported to encode tumor antigens recognized by melanoma-specific CTLs. The first of these, BAGE, encodes a nine-amino-acid peptide AARAVFLAL that is restricted by HLA-Cw"l601 (Boel et al., 1995),another example of an HLA-C-restricted peptide. Southern blotting of DNA extracted from normal peripheral blood lymphocytes or from the melanoma tumor from which BAGE was originally isolated with the BAGE coding sequence indicated the presence of multiple hybridizing bands. This result suggests that BAGE belongs to a multigene family similar to that of the MAGE genes. The second gene, GAGE, encodes an eightamino acid peptide YRPRPRRY that is restricted by HLA-Cw6 (Van den Eynde et al., 1995). A total of five homologous genes have been isolated that are 80-98% identical to the original GAGE cDNA. Of these, only GAGE-1 and GAGE-2 encode the antigen recognized by the CTLs. In a similar pattern as the MAGE genes, BAGE and GAGE are not expressed in normal tissues except for testis and are expressed to a variable extent in a number of tumors including melanomas, bladder carcinomas, mammary carcinomas, head and neck squamous cell carcinomas, non-small cell lung carcinomas, and sarcomas. These genes are not expressed in renal or colorectal carcinomas (Boel et al., 1995, Van den Eynde et al., 1995). Additionally, the BAGE genes are not expressed in leukemias or lymphomas (Boel et al., 1995). B. MELANOCYTIC DIFFERENTIATION ANTIGENS These antigens were also identified as tumor antigens expressed on melanoma and recognized by autologous CTLs. However, unlike the antigens discussed previously, these antigens are expressed by tumors and normal cells of the melanocytic lineage but not by other tissues or tumors of different histiologic origin. The existence of these antigens was evident from studies of antimelanoma CTLs restricted by HLA-A2 that were shown to recognize several allogeneic melanomas as well as normal melanocytes thath expressed the HLA-A2 molecule (Anichini et al., 1993). 1. Tyrosinme

The first of these differentiation antigens was found to be encoded by the tyrosinase gene. This gene was identified by cotransfection of COS-7 cells with both a cDNA library derived from a melanoma tumor and a plasmid-encoding HLA-AS. The sequence of the cDNA that encoded this antigen was virtually identical to that of a previously cloned human tyrosinase gene isolated from a melanoma tumor (Brichard et al., 1993). Tyrosinase is a part of the melanin biosynthesis pathway and catalyzes the synthe-

234

ROBERT A. HENDERSON AND OLIVERA J. FINN

sis of the melanin precursor dihydroxyphenylalanine. This gene was subsequently shown to encode two antigenic peptides restricted by HLAA2.1 that were recognized by tumor-specific CTLs (Wolfel et al., 1994). Both of these peptides were 9 amino acids in length and were derived from amino acids 1-9 (MLLAVLYCL)and amino acids 368-376 (YMNGTMSQV) of the tyrosinase protein. In addition to these peptide antigens, other peptides derived froin tyrosinase have been reported to be recognized by melanoma-specificCTLs restricted by both HLA-A24 (AFLPWHRLF) (Kang et al., 1995) and HLA-B44 (SEIWRDIDF) (Van Pel et d., 1995). Interestingly, a naturally occurring HLA-A2.l-associated pepide YMDGTMSQV was identified by mass spectrometry from peptides extracted from purified HLA-A2.1 molecules expressed by melanoma cells but not from extracts from lymphoid cells. The sequence of this peptide was identical to residues 368-376 of tyrosinase except that aspartic acid (D) was found in place of the asparagine (N) predicted by the gene sequence (Skipper et al., submitted for publication). Although both forms of this peptide bound equally well to HLA-A2.1, the naturally occurring peptide containing aspartic acid sensitized target cells for lysis by CTLs at a 100-fold lower concentration than the asparagine-containing version of the peptide. Additional studies showed that the naturally occurring peptide corresponding to the tyrosinase epitope was distinct from that deduced from the gene sequence and is the only one of these two peptides to be presented by HLA-A2.1 expressed on the tumor cell surface. It is hypothesized that naturally processed species arise as a result of a post-translational modification that converts asparagine to aspartic acid. This modification is thought to be due to the enzymatic deamidation of asparagine to aspartate through the action of peptide:N-glycanase, which may act on a glycosylated asparagine. This is an important finding because it demonstrates that changes in the post-translational modification in tumor cells may lead to the generation of new antigens that could be relevant to tumor rejection. It also indicates that in some cases it may not be possible to identify certain peptide antigens directly from a DNA sequence. Tyrosinase also contains an antigen that can be presented by class I1 MHC molecules and recognized by CD4' T cells (Topalian et al., 1994). Multiple CD4+ T cell clones restricted by DR4 were shown to recognize autologous EBV-B cells pulsed with lysates of COS-7 cells transfected with tyrosinase but not COS-7 cells transfected with control genes or untransfected COS-7 cells. The ability of tyrosinase to provide antigens restricted by both class I and class I1 MHC molecules makes it an excellent target for immunotherapy or vaccination because it would presumably be able to stimulate both cytotoxic CD8+ and helper CD4' T cells simultaneously.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

235

2. MAR T-l/MELAN-A An additional gene that encoded a melanocyte differentiation antigen recognized by melanoma-specificHLA-A2.1-restricted CTLs derived from tumor-infiltrating lymphocytes (TILs) was identified and named MART-1 (Kawakami et al., 1994). The same gene was also identified using melanoma-specific CTLs derived from peripheral blood lymphocytes and named MELAN-A (Coulie et al., 1994). This gene encoded a novel transmembrane protein whose expression was virtually identical to that of tyrosinase as it was expressed by most melanoma tumor cell lines and normal rnelanocytes but no other normal tissue except for the retina. It was not found to be expressed in any other tumors except for melanoma. The antigenic peptide from this protein was identified by testing multiple synthetic peptides from the MART-1 protein sequence that fit the HLA-A2.1 binding motif for their ability to sensitize target cells for lysis (Kawakami et al., 1994a). From this analysis, it was determined that the peptide AAGIGILTV was the optimal peptide for T cell recognition. A second antigenic peptide from MART-1 was identified from the analysis of naturally processed peptides isolated from melanoma tumor cells using mass spectrometry (Castelli et al., 1995). This peptide was found to be ILTVILGVL and to overlap with the previously discovered MART-1 peptide in the amino acids ILTV. It was also shown that two CTL clones specific for the MART-1 peptide AAGIGILTV also recognized the MART1peptide ILTVILGVL. MART-1 is a potentially good candidate for vaccination or passive immunotherapy as this antigen was shown to be recognized by 9 of 10 melanoma-specificTIL cell lines (Kawakamiet al., 1994a). However, a recent study indicates that an even better candidate may be gp100 (Kawakami et d., 1995). 3. gplOO/Prnell7 The gene encoding this protein was originally identified as a melanocyte lineage-specificantigen recognized by the antibodies NKI-beteb, HMB-50, and HMB-45, which are used as diagnostic markers for human melanoma (Adema et al., 1993). Analysis of the cDNA of this gene revealed it to be a type I transmembrane glycoprotein that was highly homologous to another melanocyte-specific protein Pme117. Like some of the proteins previously discussed, this protein is expressed in melanoma tumors but not in other tumor cell types or normal cells with the exception of melanocytes and pigmented cells in the retina. The identification of this protein as a melanoma-specific T cell antigen was accomplished when it was demonstrated that the transfection of this gene could reconstitute the epitope recognized by a CD8' HLA-A2.1-restricted TIL cell line (Bakker et al., 1994; Kawa-

236

ROBERT A. HENDERSON AND OLIVER4 J. FINN

kami et al., 1994b). An antigenic peptide from this protein was identified independently through the mass spectrometric isolation of a naturally processed peptide that was capable of reconstituting the epitope recognized by multiple melanoma-specific CTL lines (Cox et al., 1994). This peptide was restricted by HLA-A2.1 and had the sequence YLEPGPVTA. The second peptide epitope from this protein was identified by screening synthetic peptides based on the HLA-A2.1 binding motif for their ability to sensitize target cells for lysis and to stimulate IFN-.)I secretion from a melanoma-specific CTL, TIL 1200 (Kawakamiet al., 1994b). This peptide was also restricted by HLA-A2.1 and had the sequence LLDGTATLRL. An additional HLA-Ae.l-restricted peptide from gp100, KTWGQYWQV, was identified that was also capable of reconstituting the epitope recognized by TIL 1200 (Bakker et at., 1995). Because the gpl00 peptide KTWGQYWQV was capable of sensitizing target cells for lysis at a 100-fold lower concentration than was LLDGTATLRL, it was considered to be the immunodominant peptide of gpl00 recognized by TIL 1200. Other a100 peptides restricted by HLA-A2.1 have been reported (Kawakami et al., 1995). Although it is not clear which of these peptide epitopes may be immunodominant during an actual immune response, gpl00, like MART1,is potentially a good candidate for immunotherapy of melanoma because one of these peptide epitopes is recognized by five independently derived melanoma-specific CTL lines (Cox et at., 1994). In support of this, TIL 1200 and other TIL lines, which are specific for gpl00 peptides, have been shown be effective in adoptive immunotherapy (Kawakami et al., 1994b, 1995). C. MUTATEDOR ALTERNATEANTIGENS 1. gp75 The gp75 protein, a tyrosinase-related protein (TRP-l), was identified as an antigen recognized by serum IgG antibodies from one melanoma patient (Mattes et al., 1983; Vijayasyradhi et al., 1990). Similar to gpl00 and tyrosinase, gp75 is expressed in human melanocyticcells and melanoma tumors. Through a similar identification and cloning strategy as described previously, this gene was shown to encode a shared melanoma antigen recognized by HLA-A31-restricted TIL (Wang et al., 1996b). However, unlike the other melanoma differentiation antigens, the antigenic peptide recognized by the TIL was not derived from the normal gp75 protein. Instead, it was shown that the antigenic peptide, MSLQRQFLR, was derived from an alternative open reading frame that directs the translation of a small 24-amino acid protein (Wanget al., 1996b).Although the translation of overlapping reading frames has been previously described for viral

HUMAN TUMOR ANTIGENS ARE READY TO FLY

237

genes (Spiropoulou and Nichol, 19931,this may be the first known example of the same phenomenon in eukaryotic cells. Alternative translation of otherwise normal cellular genes is potentially another way by which new tumor-specific antigens may be derived. 2. MUM-1 Although it has been demonstrated in mouse tumor models that point mutations can generate new antigens recognized by syngeneic CTLs (Lurquin et al., 1989; Sibille et al., 1990; Mandelboim et al., 1994), this had not been shown to occur for human tumor antigens until the identification of the MUM-1 antigen (Coulie et al., 1995). A cDNA cloned from a melanoma tumor was shown to confer the recognition of a HLA-B44restricted melanoma-specific CTL clone. The sequence of this cDNA had no significant homology to any known gene and was expressed in normal tissues including liver, colon, muscle, and heart. Through transfection of truncated cDNA clones and the use of synthetic peptides, the antigenic peptide encoded by this cDNA was identified as a nine-amino-acid peptide with the sequence EEKLIWLF. This peptide conformed to the HLAB44 binding motif (Fleischhauer et al., 1994) and sensitized HLA-B44expressing target cells for lysis. When the sequence of this gene in tumor cells was compared to that in normal cells, it was found that there was a point mutation in the tumor gene such that a serine residue was replaced with isoleucine, EEKL(S 1)WLF. Both the mutated and the normal forms of the peptide bind equally well to HLA-B44, but the normal form is not recognized by the melanoma-specific CTL. This indicates that the mutated isoleucine residue is a critical residue for the recognition of this epitope by the T cell receptor. Furthermore, the cDNA sequence encoding the antigenic peptide is found to be a part of an intron, which suggests that this peptide is derived from the translation product of an incorrectly spliced mRNA. This may be the first example of an antigen generated by a point mutation that is recognized by human CTLs. The recognition of this antigen also has implications for the role of T cells in the surveillance of translated intronic regions and thus the integrity of the genome itself. This could be especially relevant to the immune response to viruses that integrate into the host cell genome. PROTEINS D. ONCOGENIC HER2/neu, ras, and p53 have all been investigated for their ability to serve as tumor antigens. However, despite a number of reports that have shown that peptides derived from both ras and p53 can be recognized by both CD8' and CD4+ T cells (Fossum et al., 1993, 1994; Van Elsas et al., 1995; Houbiers et al., 1993), currently there is no evidence that these

238

RORERT A. HENDERSON AND OLlVERA J. FINN

antigens are naturally processed and presented by human tumor cells. More convincing results have been achieved with the HERWneu oncoprotein. The realization that HERWneu had a role in the recognition of tumor cells came with the demonstration that the recognition of ovarian tumor cell lines by HLA-AS.l-restricted tumor-specific CTLs correlated with the level of expression of the HERWneu protein (Yoshino et al., 1994a) and from CD4' T cell responses of breast cancer patients to HERWneu peptides (Disis et al., 1994).Furthermore, it was also demonstrated that the transfection of HERWneu into HLA-A2.1t melanoma tumor cells conferred on them the ability to be recognized by ovarian-specific CTLs (Yoshinoet al., 1994a). A peptide GP2 from the HERYneu protein sequence (IISAWGIL) that had the same HLA-Ae.l-binding residues as those found in the immunogenic influenza matrix peptide 58-66 was synthesized and shown to reconstitute an epitope recognized by both breast and ovarian HLAA2.1-restricted TILs (Peoples et al., 1995). This epitope was also shown to be recognized by non-small cell lung cancer-specific CTLs (Yoshino et al., 1994b). An immunodominant HERWneu peptide restricted by HLAA2.1 (KIFGSLAFL)was also identified through a more extensive screening of HLA-A2.l-binding peptides derived from the HERWneu protein sequence (Fisk et al., 1995). Clinical trials utilizing HER2/neu peptides as an immunogen are expected to begin shortly at the University of Washington (M. Cheever, personal communication). E. EPITHELIAL ANTIGENS In the section dealing with tumor antigens detected by antibodies we discussed several epithelial antigens including CEA, PSA, and MUC- 1 mucin. Only the mucin has so far been reported to also be recognized by T cells. There is some preliminary evidence that T cells specific for CEA can also be expanded in vitro from patients vaccinated with CEA nucleotide vaccine (J. Schlom, personal communication). No evidence to date has been reported for T cell recognition of PSA. 1. M U G 1 Mucin As described previously, mucin molecules are high-molecular-weight glycoproteins expressed on epithelial cell tumors. Pancreatic, breast, and ovarian tumor cells, among others, express large amounts of one of the mucins, MUC-1, in an aberrantly glycosylatedform that uncovers an immunodominant epitope PDTRP on the polypeptide core of the protein. This epitope is immunogenic not only to B cells, as discussed previously, but also to ceil-mediated responses (for review see Finn et al., 1995). It has been demonstrated that lymph node cells of pancreatic cancer patients when stimulated with allogeneic pancreatic tumors will yield CD8+ CTL lines that will recognize and lyse mucin-expressing pancreatic tumor cell

HUMAN TUMOR ANTIGENS ARE READY TO FLY

239

lines as well as breast tumor cell lines (Barnd et al., 1989) expressing the same mucin. Similar results have been obtained with lymph node cells of breast cancer patients (Jerome et al., 1991) and in patients with multiple myeloma, a tumor that also expresses mucin (Takahashi et al., 1994). Recognition of the mucin on tumor cells is not MHC restricted and cannot be blocked by antibodies to class I MHC molecules, even though it appears to be T cell receptor mediated. It is hypothesized that the highly repetitive nature of the tandem repeat region of the much protein allows for the direct crosslinlang of T cell receptors on mucin-specific T cells and their activation. This would explain the lack of MHC restriction and suggest that the recognition of this tumor antigen is essentially an aberrant function of the mucin-specific T cell receptor and as such may be more analogous to the T cell recognition of haptens such as fluorescein (Siliciano et al., 1986).In support of this model of recognition, there have also been reports of MHC-unrestricted recognition of HSV-l-infected mononuclear cells by both d/3 and y/S T cells (Maccario et al., 1993). Because of the tumorspecific expression of this T cell epitope and the lack of a need for a particular MHC molecule, this unorthodox tumor antigen may be an excellent target for immunotherapy and an immunogen for vaccination. A nineamino acid peptide was identified in the tandem repeat region on the MUC-1 polypeptide core that binds HLA-A11 and elicits T cell responses in vitro (Domenech et al., 1995). It has not yet been determined if HLAA l l t tumor cells naturally process and present this peptide. A phase I clinical trial has just been completed at the University of Pittsburgh Cancer Center testing the feasibility of a MUC-1 peptide-based vaccine for breast, pancreas, and colon cancer. One hundred micrograms of a mucin peptide composed of five tandem repeats was administered together with BCG every 3 weeks for a total of three injections. No toxicity has been observed attributed to the peptide. Iinmunogenicity of the peptide, even though not the immediate goal of this phase I trial, is currently being evaluated on samples collected from the immunized patients.

F VIRALANTIGENSAS TUMOR ANTIGENS A number of viruses have been associated with cellular transformation including human papillomaviruses (HPV) and Epstein-Barr virus. These viruses may play a role in the development of human tumors and the viral proteins that are still expressed in transfonned cells are potential tumor antigens. Most of the work to date has been done on HPV as a potential tumor antigen. 1. HPV The human papilloma viruses are DNA viruses that infect epithelial tissues. Certain types of these viruses, notably HPV-16 and HPV-18, have

240

ROBERT A. H E N D E R S O N AND OLIVERA J. F I N N

been associated with squamous cell carcinomas of the human cervix and are thought to have an important role in cervical carcinogenesis (Resnick et ul., 1990; Zur Hausen, 1991).Expression of the HPV genes E6 and E7 is constitutive in cervical tumors (Seedorfet al., 1987, Von Knebel Doeberitz et al., 1991) and required for the maintenance of the transformed state (Howley, 1991). Because of their continued expression in tumor cells, the E6 and E7 proteins are promising targets for immune intervention in cervical cancers. Because they are of viral origin, the immune response against them is expected to be exquisitely tumor specific. The immunogenicity of these proteins has been analyzed extensively (Kast et al., 1994). A set of 240 overlapping nonameric peptides derived from both E6 and E7 proteins was synthesized and tested for binding to several of the most common human HLA-A alleles. From these studies, a number of high-affinity binding peptides were determined and the immunogenicity of these peptides was tested in vivo by immunization of HJAA2.1/Kh transgenic mice and in vitro by stimulation of CTLs from normal HLA-A2.1+ human peripheral blood lymphocytes (Ressing et al., 1995). Four high-affinity binding peptides were immunogenic in the transgenic mice and three of these peptides were also immunogenic to CTLs from normal donors. Human HLA-A2.l-restricted CTL clones specific for these peptides were able to recognize and lyse peptide-pulsed targets as well as the HLA-A2.1t cervical carcinoma cell line CaSki that expresses the HPV16 E6 and E7 genes. These results strongly support these peptides as naturally processed T cell epitopes of HPV-16 and as cervical carinoma tumor antigens. A phase I and I1 immunotherapy trial using these peptides is currently being conducted at the University of Leiden Medical Center (Melief and Kast, 1995). IV. Reflections and Perspectives

When all the best known tumor antigens, discovered either by antibodies or by T cells, are put together under the magnifying glass of a review article, clear patterns emerge. The outcome of many years of searching for tumor antigens can now begin to be evaluated and certain predictions made for the future direction of this line of research. First, despite the great excitement generated by the development of cloned T cells as an additional tool with which to search for tumor antigens, the antigens discovered with this new tool, though numerous, are either exactly the same as those previously discovered by antibodies or belong to the same classes of molecules. This should not have been totally unexpected because it is just another illustration of the way the immune system works. Responses to most antigens, with the exception of some carbohydrates,

HUMAN TUMOR ANTIGENS ARE READY TO FLY

24 1

come about through antigen processing and presentation by antigenpresenting cells, including antigen-specific B cells. Amplification of the response is accomplished through cognate T-B cell interactions in which a B cell, having bound antigen through surface immunoglobulin, presents fragments of the antigen bound to MHC to T cells specific for the same antigen (Croft and Swain, 1992). This interaction encourages a response of both arms of the immune system to the same antigen only to different epitopes. Which response i s more important in tumor immunity, humoral or cellular, has been a topic of discussion that has received more attention than necessary. It would be most reasonable to assume that the effective immune response against tumors would need to include both arms of specific immunity, T cells and antibodies. They in turn would uniquely recruit the participation of numerous nonspecific effector mechanisms that would further amplify the response and add to its effectiveness. There is a wealth of data, especiaIly in the mouse models, that show the importance of specific antibodies or specific populations of T cells in tumor immunity. On the contrary, there are absolutely no data that conclusively show that any of these immune effectors can work alone. When induction of an antibody response against a tumor antigen leads to tumor regression or protection from tumor growth, the cause and effect relationship is clear. This is not to say, however, that the destruction of the tumor was antibody mediated and that it did not also include a cellular response, generated in vivo, secondary to the antibody response, either against the same tumor antigen targeted by the antibody or against other tumor antigens that are still unknown. One of the more successful clinical trials to date employed treatment of colorectal cancer patients postoperatively with 17-1A antibody directed against a tumor-associated glycoprotein (Herlyn et al., 1979). Antibody therapy extended life and prolonged remission in treated patients (Riethmuller et al., 1994). It is tempting, therefore, to laud this antibody as a good antitumor effector molecule, except for the fact that the exact effector mechanism responsible for tumor rejection is not really known. In this particular case, there is ample evidence that a group of patients that responded to the antibody treatment with tumor regression all developed T cells specific for the idiotype of the injected antibody (Fagerberg et al., 1995). These T cells also recognize the 17-1A antigen and may be the effector mechanism responsible for tumor rejection. Similarly, when adoptive transfer of a single clone of tumor-specific cytotoxic or helper T cells is credited with tumor eradication (Kahn et al., 1991), one cannot be certain that this clone worked alone. There is always a possibility that a limited effect of the T cell clone on the tumor initiated an in vivo immune

242

ROBERT A. HENDERSON AND OLIVEHA J. FINN

response that produced tumor-specific antibodies about which we know very little and that may have contributed to tumor elimination. Second, and an unexpected outcome, is the realization that the majority of shared tumor antigens are self-antigens. In retrospect, this outcome should have been expected. Tumor growth is an expansion of cells of a particular tissue bearing tissue-specific antigens beyond their anatomical barriers and to greater than physiological numbers. Immunological tolerance to tissue-specific antigens is in part regulated either by a low level of expression of these antigens or by their sequestration in immunopriviledged sites. Tumor growth changes both parameters and thus tolerance is broken. Or is it? Despite the presence in cancer patients of both antibodies and cytotoxic T cells specific for self-antigens carried on tumor cells, they appear to have no obvious effect either on the tumor that continues to grow and eventually kills the patient or on the normal tissues expressing the same antigen. One possibility to consider is that self-reactive B and especially T cells are rendered unresponsive in the periphery on encounter with low levels of antigen on normal tissues in the absence of costiinulatory molecules. The anti-tumor antibodies and T cells specific for self-antigens may be those that have encountered these antigens on or around the tumor where a certain level of activation was possible. In particular, they may have been activated by higher than normal expression of antigen. The second possibility is that these antibodies and T cells are remnants of an antitumor immune response that was defeated at the start by the large load of antigen on tumor cells that were already too numerous when the tumor broke out of its original site. Due to a large number of tumor cells, heavy antigen load, the chronic presence of antigen, or the hostile environment that the tumor creates by producing immunosuppressive cytokines, amplification of this antitumor immune response did not take place. The frequency of T cells or the titer of the antibodies specific for the antigens we have described is usually very low compared to responses generated against viruses or other foreign antigens. Furthermore, it has now been repeatedly shown that in many cancer patients, not only tumor-specific T cells, but T cells in general are abnormal in the way they process signals through the T cell receptor (Mizoguchi et al., 1992 ). Considering any or a11 of these possibilities, is there a role for the so far identified tumor antigens in tumor immunotherapy? With the disclaimer that the data available for most of the reviewed antigens are still preliminary, and basing our judgment primarily on some longer running studies, we must conclude that these antigens can definitelyparticipate in antitumor immunity, their individual contributions being relative depending on the form of immunity necessary for a particular tumor or stage of the disease.

H U M A N T U M O R ANTIGENS ARE READY TO FLY

243

Many of the tumor antigens that are cell surface proteins have been successfully targeted by infusion of antibodies specific for those antigensthe process known as passive immunotherapy. Large numbers of tumorspecific cytotoxic T cells have also been infused into patients and this too has resulted in tumor regression. Neither treatment affected in any obvious way the normal tissues. Thus, both sets of antigens, those seen by antibodies and those seen by T cells can be targets of premade effector molecules, antibodies, or T cells (“passive immunity”). Because we limited our review primarily to antigens known to be immunogenic in patients, it is obvious that they are also candidate immunogens for eliciting immunity in patients through vaccination (“active immunity”). The ultimate success of a vaccine depends on the presence of fully immunocompetent T cells and B cells specific for the tumor antigen used as immunogen. Inasmuch as most antigens are normal self-proteins, the extent of the immune response that will be possible to induce against them is still unknown. The most important role, however, that the tumor antigens so far identified may have is as catalysts for in vivo antitumor immune responses against potential new antigens through a process that could be named “provoked immunity.” T cell tolerance is an active process that depends on a certain level of presentation of the self-antigens to the immune system-enough to either induce deletion of specific T cells or anergy (Lanzavecchia, 1995). It can be postulated, and it has been experimentally shown in model antigen systems (Sercarz et al., 1993), that many self-antigens are either not processed and presented or are presented below the threshold of detection. T or B cells specific for these antigens are in the peripheral repertoire but they remain unstimulated. The same situation can be envisioned for many unknown but postulated tumor-specific antigens. Even for some known antigens, like the mutated oncoprotein ras, it appears that the presentation at subthreshold levels on tumors may be responsible for its lack of recognition by T cells. There is ample evidence, however, that cryptic epitopes do get presented under certain conditions (Sercarz et al., 1993), the most recent example involves the HIV-gpl20 molecule (Salemi et al., 1995). It appears that human T cell clones isolated from HIV patients and specific for DR-restricted CD4 epitopes recognize these epitopes only on B cells that have processed and presented engineered CD4, but not on CD4’ T cells that express this molecule and should be expected to present these epitopes. Thus, CD4 is processed differently when captured by B cells. Interestingly, the T cell clones can recognize the CD4 epitopes on T cells when it is downregulated by HIV-gpl20 ligation or anti-CD4 antibodies. This is thus an example of increased processing and presentation of cryptic epitopes on a self-molecule induced by an external ligand. Most

244

ROBERT A . HENDERSON AND OLIVER4 J. FINN

of the antigens we have discussed that have been defined by antibodies have the potential to present new epitopes to the immune system when bound by the antibody on the tumor cell itself or when bound by the antibody in their soluble form and processed by APCs. HER-Yneu, for example, is a cell surface molecule and antibodies found in cancer patients are directed to the epitopes on the extracellular domain of the molecule. However, in patients treated with anti-HER-Yneu antibody, additional antibodies develop that are specific for the previously cryptic intracellular domain epitopes (Cheever et al., 1995). “Provoked immunity” may be the comon endpoint, and thus a common denominator of a number of approaches that have been used to elicit tumor rejection. One approach has involved heat shock proteins, primarily hsp70 and gp96 (Udono and Srivastava, 1994, Srivastavaet al., 1994).These molecules have been shown to bind numerous peptides and purification of these molecules from tumor cells and their uptake and reprocessing by APCs leads to stimulation of tumor-specific CTLs. In this case as well, one can postulate presentation by APCs of unknown cryptic epitopes carried on peptides bound to the heat shock proteins that on the tumor cell were not presented above the threshold level necessary to activate the immune response. This approach does not depend on the knowledge of a specific tumor peptide bound to the heat shock proteins, but it would be of interest to use the provoked immunity, which is the result of the vaccination with tumor-derived hsp70 or gp96, to identifjr the target antigens involved in the tumor-rejection response. Another approach that leads to provoked imunity has been to use as an immunogen tumor cells modified by transfection with various cytokines (for review see Pardoll, 1993; Cavalo et al., 1994), costimulatory molecules B7-1 and B7-2 (for review see Hellstrom et aZ., 1995), or class I1 antigens (Baskaret al., 1994).All of these methods are independent of the knowledge of a particular tumor antigen and routinely induce tumor immunity in mice. When applied to human tumors, they may help identify new tumor antigens, targets of the in vivo provoked immunity responsible for tumor rejection. The importance of better understanding the provoked immunity is that it apparently leads to tumor rejection. None of the antigens we have reviewed can be called tumor-rejection antigens because they were identified by antibodies and T cells from cancer patients who succumbed to their tumors. When these antigens are used as targets for passive immunotherapy or as immunogens in active immunotherapy and when a rare tumor-rejection response is observed, it will be very important to study the immune system following the tumor rejection. We may find that the tumor-rejection response may indeed be against some of those antigens.

HUMAN TUMOR ANTIGENS ARE HEADY TO FLY

245

More likely, we will find that the majority of them only initiated a process of tumor destruction that led to antigen release, cryptic antigen presentation-in other words provoked immunity. The targets of that immunity are the real tumor-rejection antigens. They may turn out to be better at inducing immunity and may replace the antigens we know now. On the other hand, they may all turn out to be unique rather than shared, in which case we can fall back on the shared antigens described here but with a greater understanding of what it takes to achieve tumor rejection.

Drawing by S. Gross; 0 1995 The New Yorker Magazine Inc

ACKNOWLEDGMENTS This work was supported by Grants NIH R 0 1 CA56103 and NIH R 0 1 CA57820 to O.J.F. and an American Cancer Society Fellowship to R.A.H. We are grateful to d l our colleagues who sent manuscripts and shared unpublished observations. We thank Ms. Sonja Finn for help in literature collection and members of the Finn Laboratory for critically reading the manuscript. O.J.F. is a member of the Immunology Program of the Pittsburgh Cancer Institute and Faculty of the American Cancer Society.

246

R O B E R T A. H E N D E R S O N A N D OLIVEHA J. F I N N

REFERENCES Acres, R. B., Hareuveni, M., Balloul, J.-M., and Kieny, M.-P. (1993).Vaccinia virus MUCl immunization of mice: Immune response and protection against growth of murine tumors bearing the MUCl antigen. J. lmmunother. 14, 136-143. Adema, G . J., de Boer, A. J., Van? Hullenaaar, R., Denijn, M., Ruiter, D. J., Vogel, A. M., and Figdor, C. G. (1993). Melanocytic leneage-specificantigens recognized by monoclonal antibodies NKI-beteb, HMB-50 and HMB-45 are encoded by a single cDNA. Am. J . Pathol. 143, 1579-1586. Angelopoulou, K., Diamandis, E., Sutherland, D., Kellen, J. A., and Bunting P. S. (1994). Prevalence of serum antibodies against the p53 tumor suppressor gene protein in various cancers. lnt. J . Cancer 58, 480-487. Anichini, A,, Maccalli, C., Mortarhi, R., Salvi, S., Arabella, M., Suarcina, P., Herlyn, M., and Parmiani, G. (1993). Melanoma cells and normal melanocytes share antigens recognized by HLA-A2-restricted cytotoxic T cell clones from melanoma patients. J. Exp. Med. 177, 989-998. Apostolopoulos, V., Xing, P.-X., and McKenzie, I. F. C. (1994). Murine immune response to cells transfected with human MUC1: Immunization with cellular and synthetoc antigens. Cancer Res. 54,5186-5193. Bakker, A. B. H., Schreurs, M. W. J., de Boer, A. J., Kawakami, Y., Rosenberg, S. A,, Adema, G . J., and Figdor, C. (1994). Melanocyte lineage-specificantigen gpl00 is recognized by melanoma-derived tumor-infiltrating lymphocytes.J. Exp. Merl. 179, 1005-1009. Bakker, A. B. H., Schreurs, W. J., Tafazzul, G., de Boer, A. J., Kawakami, Y., Adema, G. J., and Figdor, K. (1995).Identification of a novel peptide derived from the melanocytespecific GPl00 antigen as the dominant epitope recognized by an HLA-Ae.l-restricted anti-melanoma CTL line. Znt. /. Cancer 62, 97-102. Barnd, D. L., Lan, M., Metzgar, R. S., and Finn, 0. J. (1989). Specific, MHC-unrestricted recognition of tumor-associated mucins by human cytotoxic T cells. Proc. Natt. Acacl. Sci. USA 86, 7159-7163. Baskar, S., Azarenko, V., Garcia Marschall, E., Highes, E., and Ostrand-Rosenberg, S. (1994). MHC class 11-transduced tumor cells induce long-term tumor-specific immunity in autologous mice. Cell. lmmunol. 155, 123-133. Ben-Mahrez, K., Thierry, D., Sorokine, I., Danna-Muller, A., and Kohiyama, M. (1988). Detection of circulating antibodies against c-myc protein in cancer patient sera. Br. 1, Cancer 57,529-534. Boel, P., Wildmann, C., Sensi, M., Brasseur, R., Renauld, J.-C., Coulie, P., Boon, T., and van der Bruggen, P. (1995). BAGE: A new gene encoding an antigen recognized on human melanomas by cytolyhc lymphocytes. Immunity 2, 167-175. Brasseur, F., Rimoldi, D., Lienard, D., Lethe, B., Carrel, S., Arienti, F., Suter, L., Vanwijck, R., Bourlond, A., Humblet, Y., Vacca, A,, Conese, M., Lahaye, T., Degiovanni, G., Deraemaecker, R., Beauduin, M., Sastre, X., Salmon, E., Dreno, B., Jager, E., Knuth, A., Chevreau, C., Suciu, S., Lachapelle, M., Pouillart, P., Parmiani, G., Lejeune, F., Cerottini, J.-C., Boone, T., and Marchand, M. (1996). Expression of MAGE genes in primary and metastatic cutaneous melanomas. Int. I. Cancer., in press. Brichard, V., Van Pel, A,, Wolfel, T., Wolfel, C., De Plaen, E., Lethe, B., Coulie, P., and Boon, T. (1993). The tyrosinase gene codes for an antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanoma. J. Exp. Med. 178, 489-495. Burchell, J., Durbin, H., and Taylor-Papdimitriou, J. (1983). Complexity of expression of antigenic determinants recognized by monoclonal antibodies HMFG-1 and HMFG-2, in normal and malignant mammary epithelial cells. /. lmmunol. 131, 508-511.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

247

Cambia, J. M., Yu, R. K., Macaela, L. J., Kirkwood, J. M., and Varga, J. M. (1984). Gangliosides of normal and neoplastic human melanocytes. Biochem. Biophys. Res. Comtnun. 120,500-504. Castelli, C., Storkus, W. J., Maeurer, M. J., Martin, D., Huang, E. C., Pramanik, B. N., Nagabhushan, T., Parmiani, G., and Lotze, M. T. (1995).Mass spectrometric identification of a naturally processed melanoma peptide recognized by CD8' cytotoxic T lymphocytes. J . Exp. Med. 181,363-368. Cavallo, F., Giovarelli, M., Gulino, A., Vacca, A., Scala, G., and Forni, G. (1994). Interleukininduced tumor immunogenicity. Imrnunol. Ser. 61, 183-193. Celis, E., Fikes, J., Wentworth, P., Sidney, J., Southwood, S., Maewd, A,, Del Guercio, M. F., Sette, A., and Livingston, B. (1994a). Identification of potential CTL epitopes of tumor-associated antigen MAGE-1 for five common HLA-A alleles. Mot. Immunol. 31, 1423-1430. Celis, E., Tsai, V., Crimi, C., DeMars, R., Wentworth, P. A,, Chesnut, R. W., Grey, H. M., Sette, A,, and Serra, H. M. (1994b). Induction of anti-tumor cytotoxic T lymphocytes in normal humans using primary cultures and synthetic peptide epitopes. Proc. Natl. Acad. Sci. USA 91, 2105-2109. Cheever, M. A., Disis, M. L., Bernhard, H., Gralow, J. R., Hand, S. L., Huseby, E. S., Quin, H. L., Takahashi, M., and Chen, W. (1995). Immunity to oncogenic proteins. Irnmunol. Reo. 145, 33-60. Conry, R. M., LoBuglio, A. F., Kantor, J., Schlom, J., Loechel, F., Moore, S. E., Sumerel, L. A,, Barlow, D. L., Abrams, S., and Curiel, D. T. (1994). Immune response to a carcinoembryonic antigen polynucleotide vaccine. Cancer Res. 54, 1164-1 168. Conry, R. M., LoBuglio, A. F., Loechel, F., Moore, S. E., Sumerel, L. A,, Barlow, D. L., and Curiel, D. T. (1995). A carcinoembryonic antigen polynucleotide vaccine has in vivo antitumor activity. Gene Ther. 2,59-65. Cordon-Cardo, C., Reuter, V. E., Lloyd, K. O., Sheinfeld, J., Fair, W. R., Old, L. J., Melanied, M. R. (1988). Blood-group related antigens in human urothelium: Enhanced expression of precursors, Lexand Le' determinants in urothelial carcinoma. Cancer Res. 48, 4113-4120. Coulie, P. G., Brichard, V., Van Pel, A,, Wolfel, T., Schneider, J., Traversari, C., Mattei, S., De Plaen, E., Larquin, C., Szikora, J.-P,, Renauld, J.-C., and Boone, T. (1994). A new gene coding for a differentiation antigen recognized by autologous cytolytic lymphocytes on HLA-A2 melanomas. J. Exp. Med. 180, 35-42. Coulie, P. G., Lehman, F., Lethe, B., Herman, J., Lurquin, C., Andrawiss, M., and Boon, T. (1995). A mutated iiitron sequence codes for an epitope for an antigenic peptide recognized by cytolyhc T lymphocytes on a human melanoma. Proc. Nutl. Acad. Sci. USA 92: 7976-7980. Cox, A. L., Skipper, J., Chen, Y., Henderson, R. A., Darrow, T. L., Shabanowitz,J., Engehard, V. H., Hunt, D. F., and Slingluff, C. L., Jr. (1994). Identification of a peptide recognized by five melanoma-specific human cytotoxic T cell lines. Science 264, 716-719 Crawford, L. V., Pim, D. C., and Bulbrook, R. D. (1982). Detection of antibodies against the cellular protein p53 in sera from patients with breast cancer, Int. J. Cancer. 30,403-408. Croft, M., and Swain, S. (1992). Analysis of CD4' T cells that provide contact-dependent bystander help to B cells. J. Irnmunol. 149, 3157-3165. Davidoff, A. M., Iglehart, J. D., and Marks, J. R. (1992). Immune response to p53 is dependent upon p53/HSP70 complexes in breast cancer. Proc. Natl. Acad. Sci. USA 89, 3439-3442. De Plaen, E., Arden, K., Traversari, C., Gaforio, J. J., Szikora, J.-P., De Smet, C., Brasseur, F., van der Bruggen, P., Lethe, B., Lurquin, C., Brasseur, R., Chomez, P., De Backer,

248

ROBERT A. HENDERSON AND OLIVEFLA J. FINN

O., Cavanee, W., and Boon, T. (1994). Structure, chromosomal localization and expression of twelve genes of the MAGE family. lmmunogenetics 40, 360-371. Ding, L., Lalani, E., Reddish, M., Koganty, R., Wong, T., Samuel, J., Yacyshyn, M. B., Meikle, A., Fung, P. Y. S., Taylor-Papadimitriou, J., and Longenecker, B. M. (1993). Immunogenicity of synthetic peptides related to the core peptide sequence encoded by the human MUCl mucin gene: Effect ofimmunization on the growth of murine mammary adenocarcinoma cells transfected with the human MUCl gene. Cancer Immunol. Immunother. 36, 9-17. Disis, M. L., Calenoff, E., McLaughlin, G., Murphy, A. E., Chen, W., Groner, B., Jeschke, M., Lydon, N., McGlynn, E., Livingston, R. B., Moe, R., and Cheever, M. (1994). Existent T-cell and antibody immunity to Her-Wneu protein in patients with breast cancer. Cancer Res. 54, 16-20. Domenech, N., Henderson, R. A., and Finn, 0. J. (1995). Identification of an HLA-Allrestricted epitope from the tandem repeat domain of the epithelial tumor antigen mucin. J. Immunol. 155, 4766-4774. Drebin, J. A., Link, V. C., and Greene, M. I. (1988). Monoclond antibodies specific for the neu oncogene-encoded p185 molecule exert synergistic anti-tumor effects in vivo. Oncogene 2, 273-277. Eades-Perner, A. M., and Zimmerman, W. (1995). Carcinoembryonic antigen-transgenic mice: A model for tumor immunotherapy. Tumor Biol. 16,56-61. Engelhard, V. H. (1994). Structure of peptides associated with class I and class I1 MHC molecules. Annu. Reu. Immunol. 12, 181-207. Fagerberg, J., Hjelm, A.-L., Ragnhammar, P., Frodin, J.-E., Wigzel, H., and Mellstedt, H. (1995). Tumor regression in monoclonal antibody-treated patients correlates with the presence of anti-idiotype-reactive T cells. Cancer Res. 55, 1824-1827. Finlay, C. A., Hinds, P. W., and Levine, A. J. (1989). The p53 proto-oncogene can act as a suppressor of transformation. Cell 57, 1083-1093. Finn, 0. J., Jerome, K., Henderson, R. A,, Pecher, G., Domenech, N., Magarian-Blander, J., and Barratt-Boyes, S . M. (1995). MUC-1 epithelial tumor mucin-based immunity and cancer vaccines. Immunol. Reu. 145, 61-89. Fisk, B., Blevins, T. L., Wharton, J. T., and Ioannides, C. G. (1995). Identification of an immunodominant peptide of HER-Wneu protooncogene recognized by ovarian tumorspecific cytotoxic T lymphocyte lines. ]. Exp. Med. 181, 2109-2117. Flamand, V., Sornasse, T., Thielemans, K., Demanet, C., Bakkus, M., Tielemans, F., Leo, O., Urbain, J., and Moser, M. (1994). Murine dendritic cells pulsed in vitro with tumor antigen induce tumor resistance in vivo. Eur. J. Immunol. 24, 605-610. Fleischhauer, K., A d a , D., Vilbois, F., Traversari, C., Bordignon, C., and Wallny, H.-J. (1994). Characterization of natural peptide ligands for HLA-B'4402 and B'4403: Implications for peptide involvement in allorecognition of a single amino acid change in the HLA-B44 heavy chain. Tissue Antigens 44,311-317. Fossum, B., Gedde-Dahl, T., Breivik, J., Eriksen, J. A., Spurkland, A., Thorsby, E., and Gaudernack, G. (1994).p21-ras-peptide-specificT-cell responses in a patient with colorectal cancer. CD4+ and CD8+ T cells recognize a peptide corresponding to a common mutation (13Cly + Asp). Int. J. Cancer 56, 40-45. Fossum, B., Gedde-Dahl, T., Hansen, T., Eriksen, J. A., Thorsby, E., and Gaudernack, G. (1993). Overlapping epitopes encompassing a point mutation (12 Gly + Arg) in p21 ras can be recognized by HLA-DR, -DP and -DQ restricted T cells. Eur. J. Immunol. 23, 2687-2691. Gaugler, B., Van den Eynde, B., van der Bruggen, P., Romero, P., Gaforio, J. J., De Plaen, E., Lethe, B., Brasseur, F., and Boon, T. (1994). Human gene MAGE-3 codes for an

HUMAN TUMOR ANTIGENS ARE READY TO FLY

249

antigen recognized on a melanoma by autologous cytolybc T lymphocytes. J. Exp. Med. 179.921-930. Gendler, S., Lancaster, C., Taylor-Papadimitriou, J., Duhig, T., Peat, N., Burchell, J., Pemberton, L., Lalani, E.-N., and Wilson, D. (1990). Molecular cloning and expression of human tumor-associated polymorphic epithelial mucin. J . Biol. Chem. 265, 15286. Girling, A,, Bartkova, J., Gender, S., Gillet, C., Taylor, and Papdimitriou, J. (1989). A core protein of the polymorphic epithelial mucin detected by the monoclonal antibody SM3 is selectively exposed in a range of primary carcinomas. Int. J. Cancer 43, 1072-1075. Gold, D. V., and Mattes, M. J. (1988). Monoclonal antibody B72.3 reacts with a core structure of 0-linked carbohydrates. Tumor Biol. 9, 137-144. Hakamori, S. (1985).Aberrant glycosylation in cancer-cell membranes as focused on glycolipids: Overview and perspectives. Cancer Res. 45, 2405-2414. Hand, P. H., Robbins, P. F., Salgaller, M. L., Poole, D. J., and Schlom, J. (1993). Evaluation of human carcinoembryonic antigen (CEA)-transduced and non-transduced murine tumors as potential targets for anti-CEA therapies. Cancer Immunol. Immunother. 36, 65-75. Hareuveni, M., Gautier, C., Kieny, M.-P., Wreschner, D., Chambon, P., and Lathe, R. (1990).Vaccination against tumor cells expressing breast cancer epithelial tumor antigen. Proc. Natl. Acad. Sci. USA 87, 9498-9502. Haurum, J. S., Arsequell, G., Lellouch, A. C., Wong, S. Y. C., Dwek, R. A,, McMichael, A. J., and Elliot, T. (1994). Recognition of carbohydrate by major histocompatibility complex class I-restricted glycopeptide-specific cytotoxic T lymphocytes. J. Exp. Med. 180,739-744. Helling, F., Zhang, S., Shang, A., Aduri, S., Calves, M., Koganty, R., Longenecker, M. B., Yao, T-Y., Oettgen, H. F., and Livingston, P. 0. (1995). GM2-KLH conjugate vaccine: Increased immunogenicity in melanoma patients after administration with immunological adjuvant QS-21. Cancer. Res. 55, 2783-2788. Hellstrom, K. E., and Hellstrom, I. (1969).Cellular immunity against tumor specific antigens. Adu. Cancer Res. 12, 167-223. Hellstrom, K. E., Hellstrom, I., and Chen, L. (1995). Can co-stimulated tumor immunity be therapetically efficacious? Immunol. Rev. 145, 167-178. Hentu, P., Liao, S., and Vihko, P. (1992). Androgens up-regulate the human prostatespecific antigen messanger ribonucleis acid (mRNA), but down-regulate the prostatic acid phosphatase mRNA in LNCaP cell line. Endocrinology 130, 766-772. Herlyn, M. Steplewski, Z., Herlyn, D., and Kaprowski, H. (1979). Colorectal carcinomaspecific antigen: detection by means of monoclonal antibodies. Proc. Natl. Acad. Sci. USA 76, 1438-1452. Hernando, J. J., von Kleist S . , and Grunert, F. (1994).A repertoire of monoclonal antibodies reveals extensive epitope heterogeneity in CEA purified from neoplasms originating from different organs. Int. J. Cancer. 56, 655-661. Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C. C. (1991). p53 mutations in human cancer. Science (Washington DC ) 253,49-53. Houbiers, J. G. A,, Nijman, H. W., Van der Burg, S. H., Drijfhout, J. W., Kenemans, P., Van de Velde, C. J. H., Brand, A,, Momburg, F., Kast, W. M., and Melief, C. J. M. (1993). In vitro induction of human cytotoxic T lymphocyte responses against peptides of mutant and wild type p53. Eur. 1.Immunol. 23,2072-2077. Howley, P. M. (1991). Role of human papillomaviruses in human cancer. Cancer Res. 51,5019-5022. Jerome, K. R., Barnd, D. L., Bendt, K. M., Boyer, C. M., Taylor-Papadimitriou, J., McKenzie, I. F. C., Bast, R. C., Jr,, and Finn, 0. J. (1991). Cytotoxic T lymphocytes derived from

250

ROBERT A. HENDERSON AND OLIVERA J. FINN

patients with breast adenocarcinoma recognize an epitope present on the protein core of a mucin molecule preferentially expressed by malignant cells. Cancer Rex 51,2908-2916. Kahn, M., Sugawara, H., Mccowan, P., O h n o , K., Nagoya, S., Hellstrom, K. E., Hellstrom, I., and Greenberg, P. (1991). CD4+ T cell clones specific for the human p97 melanomaassociated antigen can eradicate pulmonary metastases from a murine tumor expressing the p97 antigen. J . Immunol. 146, 3235-3241. Kaminski, M. S., Kitamura, K., Maloney, D. G., and Levy, R. (1987). Idiotype vaccination against murine B cell lymphoma: Inhibition of tumor immunity by free idiotype protein. 1.Immunol. 138, 1289-1296. Kang, X.,Kawakami, Y., El-Gamil, M., Wang, R., Sakaguchi, K., Yannelli, J. R., Appella, E., Rosenberg, S. A,, and Robbins, P. (1995). Identification of a tyrosinase epitope recognized by HLA-A24-restricted, tumor infiltrating lymphocytes. J. Immunol. 155, 1343- 1348. Kantor, J., Irvine, K., Abrams, S., Snoy, P., Olsen, R.,Greiner, J., Kaufinan, H., Eggensperger, D., and Schloni, J. (1992). Immunogenicity and safety of a recombinant vaccinia virus vaccine expressing the carcinoembryonic antigen gene in a nonhuman primate. Cancer Res. 52, 6917-6925. Kast, W.M., Brandt, R. M. P., Sidney, J., Drijfhout, J.-W., Kubo, R. T., Grey, H. M., Melief, C. J. M., and Sette, A. (1994). Role of HLA-A motifs in identification of potential CTL epitopes in human papillomavirus type 16 E6 and E7 proteins. J. Immunol. 152, 3904-3912. Kawakami, Y., Eliyahu, S., Delgado, C. H., Robbins, P. F., Rivoltini, L., Topdian, S., Miki, T., and Rosenberg, S. (1994a).Cloning of the gene coding for a shared human melanoma antigen recognized by autologous T cells infiltrating into tumor. Proc. Natl. Acad. Sci. USA 91, 3515-3519. Kawakami, Y., Eliyahu, S., Sakaguchi, K., Robbins, P. F., Rivoltini, L., Yannelli, J. R., Appella, E., and Rosenberg, S. (1994b). Identification of the immunodominant epitope of the MART-1 human melanoma antigen recognized by the majority of HLA-A2restricted tumor infiltrating lymphocytes. J . Exp. Med. 180, 347-352. Kawakami, Y., Eliyahu, S., Delgado, C. H., Robbins, P. F., Sakaguchi, K., Appella, E., Yannelli, J. R., Adema, G. J., Mild, T., and Rosenberg, S. A. (1994~).Identification of a human melanoma antigen recognized by tumor-infiltrating lymphocytes associated with in vivo tumor rejection. Proc. Natl. Acad. Sci. USA 91, 6458-6462. Kawakami, Y., Eliyahu, S., Jennings, C., Sakaguchi, K., Kang, X., Southwood, S., Robbins, P. F., Sette, A., Appella, E.,and Rosenberg, S. (1995). Recognition of multiple epitopes in the human melanoma antigen gpl00 by tumor-infiltrating T lymphocytes associated with in vivo tumor regression. J. Immunol. 154, 3961-3968. and Milstein, C. (1976).Derivation of specific antibody-producing tissue culture Kohler, G., and tumor lines by cell fusion. Eur. I. Immunol. 6, 511-519. Kotera, Y., Fontenot, J. D., Pecher, G., Metzgar, R. S., and Finn, 0. J. (1994). Humoral iinrnunity against a tandem repeat epitope of human mucin MUC-1 in sera from breast, pancreatic and colon cancer patients. Cancer Res. 54, 2856-2860. Kumer, U. and Staerz, U. (1993). Concepts of antibody-mediated cancer therapy. Cancer Inuest. 11, 174-184. Kwak, L. W., Campbell, M. J., Czerwinski, D. K., Hart, S., Miller, R., and Levy,R. (1992). Induction of immune responses in patients with B-cell lymphoma against the surfaceimmunoglobulin idiotype expressed by their tumors. N . Engl. I. Med. 327, 1209-1215. Labrecque, S., Naftaly, N., Thoinson, D., and Matlashevski, G. (1993).Analysis of the antip53 response in cancer patients. Cancer Res. 53, 3468-3471.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

25 1

Lanzavecchia, A. (1995). How can cryptic epitopes trigger autoimmunity? J. Exp. Med.

181, 1945-1948. Levesque, M., Yu, H., D'Costa, M., and Didmantis, E. P. (1995).Prostate specific antigen expression by various tumors. J. Clin. Lab. Anal. 9, 123-128. L e y , R., and Miller, R. (1990).Therapy of lymphoma directed at idiotypes.J. Natl. Cancer

Inst. 10, 61-68. Livingston, P. 0. (1995). Approaches to augmenting the immunogenicity of melanoma gangliosides: From whole melanoma cells to ganglioside-KLH conjugate vaccine. Immnzrnol. Reu. 145, 147-166. Livingston, P. O., Natoli, E. J., Jr.. Jones Calves. M., Stockert, E., Oettgen, H. F., and Old, L. J. (1987). Vaccines containing purified GM2 ganglioside elicit GM2 antibodies in melanoma patients by immunization. Proc. Natl. Acad. Sci. USA 84, 2911-2915. Livingston, P. O., Ritter, G., Srivastava, P., Calves, M. J., Oettgen, H. F., and Old, L. J. (1989). Characterization of IgG and IgM antibodies induced in melanoma patients by immunization with purified GM2 ganglioside. Cancer Res. 49, 7045-7050. Livingston, P. O., Wong, C . Y., Adluri, S.,et al. (1994). A randomized trial of adjuvant vaccination with BCG versus BCG plus the melanoma ganglioside GM2 in AJCC stage I11 melanoma patients. 1.Clin. Oncol. 12, 1036-1044. Lloyd, K. 0. (1987). Blood group antigens as markers for normal cell differentiation and malignant change in human tissues. Am, J. Clin. Pathol. 87, 129-139. Lloyd, K. 0. (1990). Molecular characteristics of tumor antigens. Zrntnunol. Allergy Clin. North Am. 10, 765-779. Lloyd, K. 0.(1991).Humoral immune responses to tumor-associated carbohydrate antigens. Cancer Biol. 2, 421-431. Lloyd. K. O., and Old, L. J. (1989). Human monoclonal antibodies to glycolipids and other carbohydrate antigens: Dissection of' the humoral immune response in cancer patients. Cancer Res. 49,3445-3451. Lupu, R., Colomer, R., Kannan, B.. and Lippman, M. (1992).Characterization of a growth Factor that binds exclusively to the erbB-2 receptor and induces cellular responses. Proc. Natl. Acarl. Sci. USA 89, 2287-2291. Lurquin, C., Van Pel, A,, Mariame, B., De Plaen. E., Szikora, J.-P., Janssens, C., Reddehase, M., Lejeune, J., and Boon, T. (1989). Structure of the gene of tum- transplantation antigen P91A: The mutated exon encodes a peptide recognized with Ld by cytolytic T cells. Cell 58, 293-303. Maccario, R., Revello, M. G., Comoly, P., Montagna, D., Locatelli, F., and Gerna, G. (1993). HLA-unrestricted lalling of HSV-1-infected mononuclear cells. Involvement of either ylS or cU/p human cytotoxic T lymphocytes. J . lmmunol. 150, 1437-1445. MacLean, G. D., Bowen-Yacyshyn, M. B., Samuel, J,, Meikle, A., Stuart, G., Nation, J., Poppenla, S., Jerry, M., Koganty, R., Wong, T., and Longenecker, B. M . (1992). Active immunization of human ovarian cancer patients against a common carcinoma (ThomsenFriendenreich) determinant using a synthetic carbohydrate antigen. J. hmnunother.

11, 292-305. MacLean, G. D., Reddish, M., Koganty, R. R., Wong, T., Gandhi, S., Smolenski,M., Samuel, J., Nabholtz, J. M., and Longenecker, B. M. (1993). Immunization of breast cancer patients using a synthetic sialyl-Tn glycocongugate plus Detox adjuvant. Cancer InLmunol. I t n t nunother. 36, 2 15-222. Majuri, M. L., Hakkarainen, M., Paavonen, T., and Renkonen, R. (1994).Carcinoembryonic antigen is expressed on enctothelial cells. A putative mediator of tumor cell extravasation and metastasis. APMIS 102, 432-438.

252

ROBERT A. HENDERSON AND OLIVERA

1. FINN

Mandelboim, O., Berke, G., Fridkin, M., Feldman, M., Eisenstain, M., and Eisenbach L. (1994). CTL induction by a tumor-associated antigen octapeptide derived from a murine lung carcinoma. Nature 369, 67-71. Mamsen, J., Schmiegel, W., Roder, C., Harder, R., Juhl, H., Henne-Bruns, D., Kremer, B., and Kalthoff, H. (1994). Detection of the anti-p53 antibody response in malignant and benign pancreatic disease. Br. ]. Cancer 70, 1031-1034. Matzinger, P. (1994). Tolerance, danger, and the extended family. Annu. Rev. lmmunol. 12,991-1045. Mattes, M. J., Thomson, T. M., Old, L. J., and Lloyd, K. 0. (1983). A pigmantationassociated, differentiation antigen of human melanoma defined by a precipitating antibody in human serum. lnt. ]. Cancer 32, 717-721. Meeker, T., Lowder, J., Cleary, M. L., Stewart, S., Wamke, R., Sklar, J., and Levy, R. (1985). Emergence of idiotype variants during treatment of B-cell lymphoma with antiidiotype antibodies. N . Engl. ]. Med. 312, 1658-1665. Melief, C. J. M., and Kast, W. M. (1995). T-cell immunotherapy of tumors by adoptive transfer of cytotoxic T lymphocytes and by vaccination with minimal essential epitopes. lmmunol. Rev. 146, 167-177. Michaelsson, E., Malmstrom, V., Reis, S., Engstrom, A., Burkhardt, H., and Holmdahl, R. (1994).T cell recognition of carbohydrates on type I1 collagen.]. Exp. Med. 180,745-749. Mizoguchi, H., OShea, J. J., Longo, D. L., LoefAer, C. M., McVicar, D., and Ochoa, A. C. (1992). Alterations in signal transduction molecules in T lymphocytes from tumorbearing mice. Science 258, 1795-1797. Muraro, R., Wundulich, D., Thor, A., Lundy, J., Noguchi, P., Cunningham, R., and Schlom, J. (1985).Definition by monoclonal antibodies of a repertoire of epitopes on CEA different i d y expressed in human colon carcinomas versus normal tissues. Cancer Res. 45,57695780. Muscatelli, F., Walker, A. P., De Plaen, E., Stafford, A. N., and Monaco, A. P. (1996). Isolation and characterization of a new MAGE hene family in Xp21.3 region. Proc. NatZ. A c ~ Sci. . USA 92,4987-4991. OBoyle, K. P., Zamore, R., Adluri, S., Cohen, A., Kemeny, N., Welt, S., Lloyd, K. O., Oettgen, H. F., Old, L. J., and Livingston, P. 0. (1992). Immunization of colorectal cancer patients with modified ovine submaxillary gland mucin and adjuvants induces IgM and IgG antibodies to sialylated Tn. Cancer Res. 52,5663-5667. Old, L. J. (1981). Cancer immunology: The search for specificity. Cancer Res. 41,361-375. Pardoll, D. M. (1993). New strategies for enhancing the immunogenicity of tumors. Cum. Opinion Immunol. 5, 719-725. Peles, E., Bacus, S., Koski, R., Lu, H., Wen, D., Ogden, S., Levy, R., and Yarden, Y. (1992). Isolation of the neu/Her-2 stimulatory ligand: A 44 kd glycoprotein that induces differentiation of mammary tumor cells. Cell 69, 205-216. Peoples, G. E., Goedegebuure, P. S., Smith, R., Linehan, D. C., Yoshino, I., and Eberlein, T. J. (1995). Breast and ovarian cancer-specific cytotoxic T lymphocytes recognize the same HERYneu derived peptide. Proc. Natl. Acad. Sci. USA 92, 432-436. Press, M., Cordon-Cardo, C., and Slamon, D. (1990). Expression of HER-Yneu protooncogene in normal human adult and fetal tissues. Oncogene 5, 953-962. Pukel, C. S., Lloyd, K. O., Travassos, L. R., Dippold, W. G . , Oettgen, H. F., and Old, L. J. (1982). GD3, a prominent gangliosideof human melanoma. Detection and characterization by a mouse monoclonal antibody. ], Exp. Med. 155, 1133-1147. Rainov, N. G., Dobberstein, K.-U., Fittkau, M., Bahn, H., Holzhausen, H.-J., Gantchev, L., and Burkert, W. (1995). Absence of p53 autoantibodies in sera from glioma patients. Clin. Cancer Res. 1, 775-781.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

253

Resnick, R. M., Cornelissen, M. T. E., Wright, D. K., Eichinger, G. H., Fox, H. S., ter Schegget, J., and Manos, M. M. (1990). Detection and typing of human papillomavirus in archival cervical specimens by DNA amplification w’th consensus primers. J. Natl. Cancer Inst. 82, 1477-1484. Ressing, M. E., Sette, A., Brandt, R. M. P., Ruppert, J., Wentworth, P. A., Hartman, M., Oseroff, C., Grey, H. M., Melief, C. J. M., and Kast, W. M. (1995). Human CTL epitopes encoded by human papillomavirus type E6 and E7 identified through in vivo and in vitro immunogenicity studies of HLA-A0201-bindingpeptides. J. Immunol. 154,5934-5943. Riethmuller, G., Schneider-Cadicke, E., Schlimok, G., Schmiegel, W., Raab, R., Hofken, K., Gruber, R., Pichelmaier, H., Hirche, H., Pichlmyr, R., Buggisch, P., Witte, J., and the German Cancer Aid 17-1A Study Group (1994). Randomised trial of monoclonal antibody for adjuvant therapy of resected Dukes’ C colorectal carcinoma. Lancet 343, 1177- 1183. Roth, J. A. (Ed.) (1986). “Monoclonal Antibodies in Cancer. Advances in Diagnosis and Treatment.” Futura, Mount Kisco, NY. Rughetti, A., Turchi, V., Ghetti, C. A,, Scambia, G., Panici, P. B., Ronucci, G., Mancuso, S., Frati, L., and Nutti, M. (1993). Human B-cell immune response to the polymorphic epithelial mucin. Cancer Rex 53, 2457-2459. Salemi, S., Caporossi, A. P., Boffia, L., Longobardi, M. G., and Barnaba, V. (1995). HIVgp120 activates autorective CD4-specific T cell responses by unveiling of hdden CD4 peptides during processing. J. Exp. Med. 181, 2253-2257. Sato, T.,Maguire, H. C., Mastrangelo, M. J., and Berd, D. (1995). Human immune response to DNP-modified autologous cells after treatment with a DNP-conjugated melanoma vaccine. Clin. Immunol. Immunopath. 74, 35-43. Schiirmacher, V. (1992). Immunity and metastases: In situ activation of protective T cells by virus modified cancer vaccines, Cancer Survey 13, 129-154. Schlichtholtz, B., Legros, Y., Gillet, D., Gailard, C., M a g , M., Lane, D., Calvo, F., and Soussi, T. (1992). The immune response to p53 in breast cancer patients is directed against immunodominant epitopes unrelated to the mutational hot spot. Cancer Res. 52, 6380-6384. Schreiber, H., Ward, P. L., Rowley, D. A,, and Stauss, H. J. (1988). Unique tumor-specific antigens. Annu. Rev. Immunol. 6,465-483. Schuessler, M. A., Pintado, S., Welt, S., Real, F. X., Xu, M., Melamed, M. R., Lloyd, K. O., and Oettgen, H. F. (1991). Blood-group and blood-group-related antigens in normal pancreas and pancreas cancer: Enhanced expression of precursor type 1,Tn and SialylTn in pancreas cancer. Int. J. Cancer. 47, 180-187. Seed, B., and Aruffo, A. (1987). Molecular cloning of the CD2 antigen, the T-cell erythrocyte receptor, by a rapid immunoselection procedure. Proc. Natl. Acad. Sci. USA 84,33653369. Seedorf, K., Oltersdorf, T., Krammer, G., and Rowekamp, W. (1987). Identification of early proteins of the human papillomavirus type 16 (HPV16) and type 18 (HPV18) in cervical carcinoma cells. EMBO J. 6, 139-144. Sercm, E. E., Lehmann, P. V., Ametani, A,, Benichou, G., Miller, A., and Moudgil, K. (1993). Dominance and crypticity of T cell antigenic determinants. Annu. Reu. ImmunoZ. 11, 729-766. Shievely, J., and Beatty, J. (1985). CEA related antigens: Molecular, biological and clinical significance. CRC Crit. Rev. Oncol. Hemutol. 2, 355-399. Sibille, C., Chomez, P., Wildmann, C., Van Pel, A., De Plaen, E., Maryanski, J., de Bergeyck, V., and Boone, T. (1990). Structure of the gene of tum- transplantation antigen P198: A point mutation generates a new antigenic peptide. J. Exp. Med. 172, 35-45.

254

ROBERT A. HENDERSON AND OLIVERA J. FINN

Siliciano, R., Hemesath, T., Pratt, J,, Dintzis, R., Dintzis, H., Acuto, O., Shin, H., and Reinherz, E. (1986). Direct evidence for the existence of nominal antigen binding sites of T cell surface Ti a-p heterodimers of MHC-restricted T cell clones. Cell 47, 161-171. Sirokine, I., Ben-Mahrez, K., Bracone, A,, Thierry, D., Ishii, S., Imamoto, F., and Kohiyama, M. (1991). Presence of circulating anti c-myb oncogene product antibodies in human sera. Znt. I. Cancer 47, 665-669. Spiropoulou, C. F., and Nichol, S. T. (1993). A small highly basic protein is encoded in overlapping frame within the P gene of vesicular stoinatatis virus.]. Virol. 67,3103-3110. Springer, G. F. (1984). T and Tn, general carcinoma autoantigens. Scierice 224,1198-1206. Springer, C. F., Desai, P. R., Spencer, B. D., Tegtmeyer, H., Carlstedt, S. C., and Scanlon, E. F. (1995). T/Tn antigen vaccine is effective and safe in preventing recurrence of advanced breast carcinoma. Cancer Det. Preu. 19, 374-380. Srivastava, P. K., Udono, H., Blachere, N. E., and Li, Z. H. (1994). Heat-shock proteins transfer peptides during antigen processing and CTL priming. Zmmunogenetics 39,93-98. Stancovaski, I., Huwitz, E., Leitner, O., Ullrich, A,, Yarden, Y., and Sela, M. (1991). Mechanistic aspects of the opposing effects of monoclonal antibodies to the erbB-2 receptor on tumor growth. Proc. Natl. Acad. Sci. USA 88, 8691-8695. Steven, C. W., and Lane, D. P. (1992). Mutant confirmation of p53-precise epitope mapping using a filamentous phage epitope library. J. Mol. Bid. 225, 577-583. Stevenson, F. K., Zhu, D., King, C. A,, Ashworth, L. J., Kumar S., and Hawkins, R. E. (1995). Idiotypic DNA vaccines against B-cell lymphoma. Immunol. Reu. 145, 211-228. Tai, T., Cahan, L. D., Tsuchida, T., Saxton, R. E., Irie, R. F., and Morton, D.L (1985). Immunogenicity of melanoma-associated gangliosides in cancer patients. Znt. J. Cancer 35,607-612. Takahashi, T., Makiguchi, Y., Hinoda, Y, Kakiuchi, H., Nakagawa, N., Imai, K., and Yachi, A. (1994). Expression of MUCl on myeloma cells and induction of HLA-unrestricted CTL against MUC-1 from multiple myeloma patient. /. Zmmunol. 153, 2102-2109. Tan, E. M. (1989). Antinuclear antibodies: Diagnostic markers for autoimmune disease. Adu. Zmmunol. 44,93-151. Taylor-Papadimitriou,J. (1991). Report on the First International Workshop on CarcinomaAssociated Mucins. Int. J. Cancer 49, 1-11. Thomson, J., Grunert, F., and Ziemerman, W. (1991). CEA gene family: Molecular biology and clinical perspective, J. Chn. Lab. Anal. 5, 344-366. Topalian, S. L., Rivoltini, L., Mancini, M., Markns, N . R., Robbins, P. F., Kawakami,Y., and Rosenberg, S. A. (1994). Human CD4+ T cells specifically recognize a shared melanomaassociated antigen encoded by the tyrosinase gene. Proc. Natl. Acad. Sci. U S A 91,94619465. Traversari, C., van der Bruggen, P., Luescher, I. F., Lurquin, C., Chomez, P., Van Pel, A,, De Plaen, E., Amar-Costesec, A., and Boone, T. (1992). A nonapeptide encoded by human gene MAGE-1 is recognized on HLA-A1 by cytolytic T lymphocytes directed against tumor antigen MZ2-E. J. Exp. Med. 176, 1453-1457. Udono, H., and Srivastava, P. K. (1994). Comparison of tumor-specific immunogenicities of stress-induced proteins Gp96, Hsp9O and Hsp70. J. Immunol. 152, 5398-5403. Urban, J. L., and Schreiber, H. (1992).Tumor antigens. Annu. Reu. Zmmunol. 10,617-644. Van den Eynde, B., Peeters, O., De Backer, O., Gaugler, B., Lucas, S., and Boone, T. (1995). A new family of genes coding for an antigen recognized by autologous cytolyhc T lymphocytes on a human melanoma. 1.Exp. Med. 182, 689-698. van der Bruggen, P., Szikora, J-P., Boel, P., Wildmann, C., Somville, M., Sensi, M., and Boon, T. (1994).Autologous cytolytic lymphocytes recognize a MAGE-1 nonapeptide on melanomas expressing HLA-Cwl601. Eur. 1. Zmmunol. 24, 2134-2140.

HUMAN TUMOR ANTIGENS ARE READY TO FLY

255

van der Bruggen, P., Traversari, C., Chomez, P., Durquin, C., De Plaen, E., Van den Eynde, B., Knuth, A,, and Boon, T. (1991). A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma. Science 254, 1643-1647. Van Elsas, A,, Nijman, H. W., Van Der Minne, C. E., Mourer, J. S., Kast, W. M., Melief, C. 1. M., and Schrier, P. I. (1995). Induction and characterization of cytotoxic T-lymphocytks recognizing a mutated p21RAS peptide presented by HLA-AO2Ol. Znt. 1.’Cancer 61. 389-396. V a n 6 F., and Klein, E. (1982). Specificity of auto-tumor cytotoxicity exerted by fresh, activated and propagated human T lymphocytes. Int. 1.Cancer 29, 547-558. Van Pel, A,, van der Bruggen, P., Coulie, P. G., Brichard, V. G., Leth6, B., van den Eynde, B., Uyttenhove, C., Renauld, J.-C., and Boone, T. (1995).Genes codingfor tumor antigens recognized by cytolpc T lymphocytes. Immunol. Rev. 145, 229-250. Vijayasyradhi, S., Bouchard, B. B., & Houghton, A. N. (1990). The melanoma antigen gp75 is the human homologue to mouse b (BROWN) locus gene.]. Exp. Med. 171,1375-1380. Vitetta, E., and Uhr, J. (1994). Monoclonal antibodies as agonists: An expanded role for their use in cancer therapy. Cancer Res. 54, 5301-5309. Vogelstein, B., and Kinzler, K. W. (1992). p53 function and dysfunction. Cell 70,523-526. Von Knebel Doeberitz, M., Baukneckt, T., Bartsch, D., and Zur Hausen, H. (1991). Influence of chromosomal integration on glucocorticoid-regulated transcription of growthstimulating papillomavirus genes E6 and E7 in cervical carcinoma cells. Proc. Nad. Acad. Sci. USA 88, 1411-1415. Vuist, M. J., Levy, R., and Maloney, D. G. (1994). Lymphoma regression induced by monoclonal anti-idiotypic antibodies correlates with their ability to induce Ig signal transduction and is not prevented by tumor expression of high levels of Bcl-2 protein. Blood 83,899-906. Waldman, T. (1991). Monoclonal antibodies in diagnosis and therapy. Science (Washington DC) 252, 1657-1662. Wang, M. C., Papsider, L. D., Kuriyama, M., Valenzuela, L. A., Murphy, G. P., and Chu, T. M. (1981). Prostate antigen: a new potential marker for prostatic cancer. Prostate 2, 89-96. Wang, R.-F., Robbins, P. F., Kawakami, Y., Kang, X.-Q., and Rosenberg, S. A. (1995a). Identification of a gene encoding a melanoma tumor antigen recognized by HLA-A31restricted tumor-infiltrating lymphocytes./. Erp. Med. 181, 799-804. Wang, R. F., Parlehurst, M. R., Kawakami, Y., Robins, P. F., and Rosenberg, S. A. (1996). Utilization of an alternative open reading frame of a normal gene in generating a novel human cancer antigen. 1. Exp. Med., in press. Wat, W. W. K., Lee, P.-J., M’Timkulu, T., Chan, W.-P., and Loor, R. (1986). Human prostate-specific antigen: Structural and functional similarity with serine proteases. Proc. Natl. Acad. Sci. USA 83, 3166-3170. Winter, S. F., Minna, J. D., Johnson, B. E., Takahashi, T., Gazdar, A., and Carbome, D. (1992). Development of antibodies against p53 in lung cancer patients appears to be dependent on the type of p53 mutation. Cancer Res. 52,4168-4174. Wolfel, T., Van pel, A., Brichard, V., Schneider, J., Seliger, B., Meyer zum Buschenfelde, and Boon, T. (1994). Two tyrosinase nonapeptides recognized on HLA-A2 melanomas by autologous cytolpc T lymphocytes. Eur. 1.Zmmunol. 24, 759-764. Yoshino, I., Peoples, G. E., Goedegebuure, P. S., Maziarz, R., and Eberlein, T. J. (1994a). Association of HER’Wneu expression with sensitivity to tumor-specific CTL in human ovarian cancer. 1.Immunol. 152, 2393-2400. Yoshino, I., Goedegebuure, P. S., Peoples, G. E., Parikh, A. S., DiMaio, M. J., Lyerly, H. K., Gazdar, A. F., and Eberlein, T. J. (1994b). HERWneu-derived peptides are shared

256

ROBERT A. HENDERSON AND OLIVERA J. FINN

antigens among human non-small cell lung cancer and ovarian cancer. Cancer Res. 54,3387-3390. Yu, H., and Diamandis, E. P. (1995a). Prostate specific antigen in milk of lactating women. Clin. Chem. 41, 54-58. Yu, H., and Diamandis, E. P. (1995b). Prostate specific antigen immunoreactivityin amniotic fluid. Clin. Chem. 41, 54-58. Yu, H., Diamandis, E. P., Zarghami, N., and Grass, L. (1994). Induction of prostate specific antigen production by steroids and tamoxifen in breast cancer cell lines. Breast Cancer Res. Treat. 32, 291-300. Yu, H., Diamandis, E. P., Monne, M., and Croce, C. M. (1995a). Oral contraceptive-induced expression ofprostate specific antigen in the female breast.]. Biol. Chem. 270,6615-6618. Yu, H., Giai, M., Diamandis, E. P., Katsaros, D., Sutherland, D. J. A., Levesque, M. A., Roagna, R., Ponzone, R., and Sismondi, P. (1995b). Prostate-specific antigen is a new favorable prognostic indicator for women with breast cancer. Cancer Res. 55,2104-2110. Zhang, S., Walberg, L. A,, Ogata, S., Itzkowitz, S. H., Rao Koganty, R., Reddish, M., Gandhi, S. S . , Longenecker, S. H., Lloyd, K., and Livingston, P. 0. (1995). Immune sera and monoclonal antibodies define two configurations for the sialyl Tn antigen. Cancer Res. 55, 3364-3368. Zhao, X.-J., and Cheung, N.-K. (1995). GD2 oligosaccharide:Target for cytotoxic T lymphocytes. 1.Exp. Med. 182, 67-74. Zotter, S., Hageman, P. C., Lossnitzer, A., Mooi, W. J., and Hilgers, J. (1988). Tissue and tumor distribution of human polymorphic epithelial mucin. Cancer Reu. 11, 55-101. Zur Hausen, H. (1994). Molecular pathogenesis of the cervix and its causation by specific human papillomavirus types. Cum Top. Microbiol. Immunol. 186, 131-156. This article was accepted for publication on 28 September 1995.

ADVANCES IN IMMUNOMGY. VOL. 62

Inflammatory Mediators, Cytokines, and Adhesion Molecules in Pulmonary Inflammation and Injury NICHOLAS W. LUKACS AND PETER A. WARD Depadment of fahobgy, University of Michigan M i c a \ School, Ann Arbor, Michigan A8 109

1. Introduction

The disruption of lung structure and pulmonary function can be a devastating event leading to lifelong illness and deteriorating health. Regulation of lung function is maintained by intricate mechanisms that promote a balance between host defenses against injurious (infectious and noninfectious) and reactive inflammatory responses. An excessive inflammatory response can result in damage to the lung, leading to organ dysfunction. The mechanisms attributed to lung damage resulting during pulmonary inflammation are the focus of this review. In addition, we will focus on specific models of pulmonary inflammation that have contributed to our knowledge of inflammatory mechanisms responsible for lung damage. Induction of inflammatory responses is mediated through a multistep process initiated by the release of the early response cytokines, TNF-a and IL-1, as well as other components, such as activation products of complement, oxidants, and proteinases (1).The release of early response medators leads to the upregulation of selectins (E and P) and other adhesion molecules (ICAM-I, VCAM-1, etc.) on the surface of vascular endothelial cells within and around the site of inflammation (2-4). Selectin molecules (L, P, and E) interacting with their “counterreceptors” have been shown to play a role in early adhesion events (5).E- and P-selectins are rapidly inducible and initiate rolling of leukocytes on activated endothelium through Ca2+-dependentrecognition of cell surface carbohydrates of the sialyl Lewis X family and related ohgosaccharides that may be expressed on glycoproteins or glycolipids. L-selectin on lymphocytes, originally referred to as the “lymphocyte homing receptor,” has been associated with adhesion of lymphocytes to high endothelial venules (HEV) of lymph nodes (6). The counterreceptor on HEVs for lymphocytes seem to vary with the location of the lymph node or lymphocyte structure and may include glyCAM-1 as well as “mucosal addressin” cell adhesion molecule (MadCAM-1). These molecules have been the subject of several reviews (7, 8). Selectin molecules function to slow the flow of leukocytes in the vascular compartment and allow transient adherence to counterreceptors 257 Copyight 0 1996 by Academic Press, Inc. All nghts of reproduction in any form reserved.

258

NICHOLAS W. LUKACS AND PETER A. WARD

expressed on activated vascular endothelium (9).Normally, adhesion molecules or their counterreceptors are tightly regulated. Inappropriate expression may contribute to inflammatory disorders. Although P-selectin (CD62) is expressed on activated platelets and endothelial surfaces by translocation of storage granules to the cell surface, E-selectin is expressed on activated endothelium and L-selectin is normally expressed on all blood leukocytes (5). The currently held concept is that the rapid expression of selectin molecules and/or their counterreceptors allows the reversible binding resulting in “rolling” of leukocytes along the vascular endothelial surface (10, 11).This rolling allows the leukocytes to be slowed from circulatory flow, permitting development of firm adhesion to endothelial cell-expressed adhesion molecules (VCAM-1 and ICAM-1, P- and E-selectins) at the site of inflammation (12, 13). These interactions have been outlined in Fig. 1. The expression of endothelial adhesion molecules leads to a sequence of adherence promoting events whereby leukocytes ultimately bind firmly to the vascular endothelium via P-integrin receptors on leukocyte surfaces. Although the mechanism is not entirely clear, it appears, at least in the case of Pz integrins, a change in affinity occurs during cell activation events and allows an increased avidity of adhesion-promoting molecules (14, 15). The Bla4 integrins (VLA-4),expressed primarily on mononuclear cells and eosinophils, have been shown to bind to vascular cell adhesion molecule-1 (VCAM-1) (16), whereas P2 integrins (CDll/CD18), expressed on all leukocytes, bind varyingly to intracellular adhesion molecules-1, -2, and

- Slectin (E-selectin)

u-

Adherion Molecule (ICAM-1) Rolling

Firm adhesion

MigraUon

Endothrllum

IL-1,TNFa

C5a, IL-8, ENA-78 (Chrmorltnct~nts)

FIG.1. Extravasationof leukocytesacrossvascular endothelium is dependent on cytokineinduced selectin and adhesion molecule interactions, followed by the directed migration into tissue toward chemotactic factors.

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

259

-3 (ICAM-1, -2, and -3) (17,18),the first of which is expressed on endothelial cells. These two families of adhesion molecules are able to facilitate leukocyte binding to the activated endothelium. Once firmly adhered, leukocytes may then enter the tissue following chemotactic gradients through a series of detachmentheadherence events typified by the polar expression of integrins specific for the adhesion molecules on the surface of mesenchymal-derived cells. Cell-to-cell communication during inflammatory events is mediated by a class of soluble proteins known as cytokines. This diverse group of molecules mediates, maintains, and regulates the inflammatory response and dictates the intensity of the reaction. The early response cytokines IL-1 and TNF-a appear to play a pivotal role in the induction of inflammatory responses through the initiation of cytokine cascades (19).The exuberant production of IL-1 and TNF-a may lead to multisystem injury and systemic complications, as exemplified in septic shock syndromes (19,2O). IL-1 and TNF-a initially upregulate the selectin (E-selectin) and adhesion molecules (ICAM-1 and VCAM-1) needed for the first step of leukocyte extravasation into tissue. In addition, IL-1 and TNF-a upregulate other inflammatory cytokines involved in the chemotactic responses of leukocytes into inflamed tissue. The production of chemotactic factors appears to be paramount for the movement of leukocytes from the vascular compartment to the interstitium of the lung. The chemokines, a family of chemotactic cytohnes, are specificallyupregulated by IL-1 and TNF-a in most immune and nonimmune cells during inflammation and exhibit specificity for particular subpopulations of leukocytes during disease development (22). An important concept that has been realized has been that chemokines can be produced by nearly every cell type, including platelets. These observations have important implications in the overall understanding of the role of these molecules during development of inflammatory responses. Because of the near omnipresence of chemokines in association with inflammation and the specificityof chemokmes for particular leukocyte subpopulations, chemokines have become the center of interest. The members of the chemokine family are more thoroughly reviewed in a later section. Regulation of inflammatory responses in the lung has been a critically important focus of research. In severe cases of lung injury and inflammation, glucocorticoids have been the therapy of choice. However, this broad spectrum inhibitor is associated with significant systemic side effects. Studies into the regulation of inflammatory proteins have demonstrated several cytokine molecules that specifically modulate the production of inflammatory cytokines. One of the most potent cytokine synthesis regulators appears to be IL-10 (22,23). IL-10 was first described by its ability to downregulate a number of inflammatory cytokines, which now include interluken-1

260

NICHOLAS W. LUKACS AND PETER A . WARD

( IL-l), TNF-a, interferon-? (IFN-y), IL-8, macrophage inflammatory protein-1, (MIP-la),as well as a number of other cytokines. IL-4 and IL13 have demonstrated some of the same properties as IL-10. However, unlike IL-10, IL-4 and IL-13 appear to have inflammatory properties that may exacerbate certain types of chronic lung inflammation that may be controlled by Th2 lymphocytes. The inflammatory functions of IL-4 and IL-13 include induction of B cell growth, antibody class switching (to IgE or IgGl), maturation of Th2-type lymphocytes, upregulation of VCAM-1 expression, hematopoietic functions, and a significant role in promoting allergic inflammatory responses (25,26). These three cytokines collectively have been termed Th2-type cytokines. In general, they downregulate Thltype ( IFN-y-mediated) reactions and acute inflammatory responses (27). The investigation into the mechanisms that are operative with regulatory cytokines may lead to development of therapeutics with specific local effects for controlling pulmonary inflammatory responses. 11. Pathophysiologic Mechanisms of Inflammatory Injury

Injury to the lung may have long-term physiological effects including decreased efficacy of gas exchange, increased vascular permeability, and irreversible lung fibrosis. Many pulmonary diseases, including emphysema and acute respiratory distress syndrome (ARDS), may have several similar mechanisms that lead to the injury of the lung. The injury may be the result of multiple mediators, including complement products, radicals (derived from oxygen or L-arginine), and proteinases, or the events leading to injury may be due to direct and indirect effects of cytokine cascades. The exacerbated production or expression of these multiple factors has an effect on the local environment that results in cell injury and lung dysfunction (Table I). A. COMPLEMENT The initiation of the complement cascade can be accomplished via multiple mechanisms including antibody-antigen complexes, bacterial products, and toxins. Of the complement cascade products, C3 and C5 have the most profound effects on the inflammatory response (28, 29). The split products of C3, C3a and C3b, generated by C3 convertase [as well as further products (iCSb, C3d, C3g)], have important activating roles in the inflammatory pathway. C3a is an anaphylatoxin that induces the activation of mast cellshasophils and appears to have direct and indirect effects on vascular permeability. C3b acts as a potent opsinizing component binding to bacteria and allowing accelerated phagocytosis and clearance of pathogens via its receptor on neutrophils and macrophages, Mac-l (CDllb/

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

261

TABLE I SOLUBLE FACTORS LEADING TO LUNGPATHOMGY Component

Source

Direct Effects

Leukocyte recruitment, Complement (C3a, C5a) Cleaved serum proteins, activation, and degranulation epithelial cell protein Oxygen radicals Neutrophils, macrophages, Cellular activation, structural cell damage, disruption of normal monocytes physiology Proteases Recruited neutrophils Cleavage of tissue matrix, direct toxic effects to structural cell populations Upregulation of leukocyte Inflammatory cytokines Leukocytes adhesion, induction of (TNF-a and IL-1) cytokine cascades Recruitment and activation of Chemokines Most cell populations specific leukocyte subsets

CD18). The split products of C5, C5a and C5b, can subsequently be induced through the aid of C3b and C5 convertase. Similar to C3a, but much more potent, C5a is an anaphylatoxin,which activates mast cell and basophil degranulation and immediate changes in vascular permeability. In addition, C5a stimulates smooth muscle contraction and has neutrophil chemotactic and activating characteristics that promote directed migration of these leukocytes toward a concentration gradient. Furthermore, C5a stimulates neutrophil oxidative metabolism, granule discharge, and adhesiveness to vascular endothelium. In addition, C5a can directly stimulate endothelial cells in a G protein receptor-dependent fashion to cause signal transduction events resulting in increased intracellular Ca2+,induction of superoxide (02), and expression of P-selectin (30).C3a lacks these activities. Accordingly, C5a has the ability to stimulate both leukocyte and endothelial cells. Altogether, these functions of C3 and C5 spilt products suggest that they are potent inflammatory mediators. Elevated complement component levels have been described with several pulmonary diseases including sarcoidosis,idiopathic pulmonary fibrosis (IPF), ARDS, and chronic obstructive pulmonary disease (COPD) (3133). Bronchoalveolar lavage (BAL) fluid samples from sarcoid and IPF patients demonstrate a significant increase in complement component C2/ CH50 ratio when compared to BAL fluid analysis from normal individuals, suggesting complement pathway activation within the lung. In addition, BAL fluid from IPF patients indicated a correlation between the Ba fragment of the alternative pathway of complement and clinical disease (31). In trauma patients at risk for ARDS, an increased ratio of BAL fluid C3a to

262

NICHOLAS W. LUKACS AND PETER A. WARD

plasma C3a has been associated with development of ARDS and indicated substantial complement activation locally within the lung (32). In COPD patients, significantly lower levels of serum C3 and C4 have been detected in BAL fluids and correlated with chronic cough and expectoration (33), suggesting the probability of intrapulmonary consumption of the complement proteins. Interestingly, a major source of complement component C3 within the lung has been identified from type I1 epithelial cells (34) with synthesis being induced by several mechanisms including IL-4 activation (35).Although pulmonary diseases have been shown to be accompanied by increased levels of complement components that likely amplify the inflammation and injury within the lung, it is not clear whether these components have causative or potentiating effects on endstage lung disease. Mechanisms of complement activation and lung injury have more clearly been defined in animal models of pulmonary disease and will be discussed in later sections.

B. OXIDANT FORMATION DURING LUNGINFLAMMATION The activation of pulmonary macrophages and infiltrating leukocytes (i.e., neutrophils) induces the release of oxidative intermediates that have additional activating and damaging effects to structural cells in the lung. Within the lung, the induction of oxidant stress, which is accomplished either by raising ambient oxygen concentrations or by activating local phagocytic populations to release toxic oxygen radicals, appears to induce functional lung damage (36). The scheme of oxygen radical formation by NADPH oxidase of phagocytic cells (neutrophils, macrophages, and monocytes) is shown in Fig. 2. NADPH oxidase, which is a memberassociated enzyme assembled on the surface of phagocytic cells through activation events as well as transposition of cytoplasmic subunits to the cell surface, causes a sequence of electron (e-) additions to fully reduced molecular oxygen (dioxygen).NADPH oxidase adds single electrons rather than duplexes of electrons, thus producing intermediate oxygen-centered radicals. The first product in the sequence is 02* (superoxide anion), which has a limited but not exquisitely brief half-life. The addition of an electron to 02results in the generation of hydrogen peroxide, which can be acted on by myeloperoxidase in the presence of halide to form products such as hypocholrous acid (HOC1). HOCl is a powerful oxidant but appears to be relatively nontoxic to tissues. However, HOCl is bactericidal and also has the ability to activate the precursor arm of matalloproteinases, such as collagenase and gelatinase, to their active forms. The addition of an electron of Hz02results in the formation of the most highly reactive of all oxygen radicals, HO- (hydroxyl radical). This conversion and reduction of HzOz requires the presence of a transition metal such as reduced iron

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

263

02

r I

2'

t H202

(Superoxide)

MPO+Hallde

I

Fe2$

moct e-

r-

0' (hydroxyl radical)

H20

FIG.2. NADPH oxidase pathway of oxygen radicd formation.

(Fez'). The Fez' ion does not normally exist in tissues. Its source is thought to be from stores of intracellular ferritin in which all iron is present in the oxidized state ( Fe3+).It has been demonstrated that 02* can cause reduction and release of Fe3' from the ferritin molecule and conversion of HzOzto the hydroxyl radical can readily occur. The final addition of an electron to the hydroxyl radical results in the formation of fully reduced forms of oxygen, HzO.The resulting oxygen radicals that form, Oz*and HzOzleading to HO. formation, may cause direct damage of cells within the lung and airway that have functional roles in gas exchange. In vitro studies have suggested that oxygen radical production may be central to airway epithelial and endothelial cell damage. In vivo damage to airway epithelial cells by infiltrated and activated neutrophils appears to correlate with several disease states within the lung. In ARDS, a disease associated with the shock/ sepsis syndrome, neutrophil infiltration and oxygen radical formation are closely associated with disease pathology (37, 38). In this disease, damage to the epithelial and endothelial cells appears to be induced by activated and degranulated neutrophils that release reactive oxygen species (ROS) and that, together with proteases, promote local damage and sloughing of the airway epithelial cells and vascular endothelial cell injury. An important event in ischemia-reperfusion injury and in hyperoxia-induced lung injury is the release of arachidonic acid and formation of eicosanoids such as prostaglandins, prostacylin, thromboxane, and leukotrienes. ROS-induced

264

NICHOLAS W. LUKACS AND PETER A. WARD

release of arachidonic acid metabolites may lead to airway hyperreactivity as well as resulting in vaso- and bronchoconstriction and edema formation within the lung (41).Subsequently, these same substances display chemotactic functions and can further damage the lung. These events may play a signficant role in several respiratory diseases including ARDS, hyperoxic syndromes, and asthma. In addition to the direct damage generated by oxygen radicals, these reactive intermediates can also augment local lung damage by inducing chemokines responsible for specific leukocyte infiltration. Observations have demonstrated that free radical production during anoxiahyperoxia conditions in vitro can induce the production of a neutrophil-specific chemokine, IL-8, in isolated monocyte populations (42).In addition, under the same conditions IL-1 and TNF gene expression may also be induced, suggesting that the generation of oxygen radicals can augment the inflammatory response via the induction of inflammatory and chemotactic cytokines (43). These results may apply directly to events that occur during anoxiahyperoxia, in vivo, and culminate in lung inflammation and damage. Nitrogen oxide (.NO) and related products have been shown to be important bioactive species of metabolites within the lung. .NO is formed in epithelial cells, macrophages, neutrophils, mast cells, autonomic neurons, smooth muscle cells, fibroblasts,and endothelial cells utilizing the inducible constitutive enzyme, nitric oxide synthase (NOS), via L-Arg-dependent pathways (44).As demonstrated in Fig. 3, in the nitric oxide pathway, whether involving constitutive or inducible NOS, the substrate, L-arginine, is converted to L-citrulline along with generation of .NO. Nitric oxide is relatively nonreactive radical that, on interaction with superoxide generated from some other source, can form the highly reactive peroxynitrite anion (ONOO.), which is reactive with thiol groups as well as a large number of other chemical targets. Protonation of the peroxynitrite anion produces the hydroperoxy nitrite compound, which undergoes homolytic cleavage to form the hydroxyl radical (HO.). It should be noted that in this situation hydroxyl radical generation occurs in the absence of a heavy metal catalyst. The other product, NO2, undergoes conversion to nitrite (NO;) and nitrate (NO3),both of which can be measured and represent quantitative surrogates of the original amount of .NO generated. Accordingly, in this pathway peroxynitrite anion and hydroxyl radical are the species that are most reactive with potential targets. *NOitself has been demonstrated to react with iron in the iron sulfur-centered enzymes, such as those present in mitochondrial, resulting in impaired oxidative metabolism. Hemoglobin within red cells has also been shown to be a scavenger for -NO. One of the primary functions of *NOappears to be induction of airway and vascular smooth muscle cell relaxation. For this reason, .NO may

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

265

L-arginine

c

nitric oxide synthase (NOS)

'NO (nitric oxide) + L-citrilline

P

ON 0 (peroxynitrite anion)

H+

H d + 'NO,

ik

NO,- +

NO,'

~ multiple CL populaFIG.3. The nitric oxide pathway. The generation of nitric o x i in tions is mediated via the enzyme nitric oxide synthase and L-arginine. Other reactive species can also be formed when superoxide is available. Peroxynitrite anion is a highly reactive species that subsequently can be protonated to form the highly reactive hydroxyl radical that can damage cells.

be therapeutically beneficial in various disease states such as pulmonary hypertension, ARDS, and asthma. *NO plays an important role in the killing of organisms such as Mycobacterium tuberculosis, Toroplasm gondii, T y p a n o s o m cruzi, Leishmania donovani, etc. in macrophages. This production of * N Oduring inflammatory responses appears to have a generalized beneficial effect in clearance of infectious agents and vasodilatation, which enhances blood flow (and delivery of leukocytes and proteins). However, there are two lines of evidence that .NO generated in vivo may be harmful. In patients with ARDS and shockhepsis syndromes, it is reported that the hypotensive state that is nonresponsive to vasopressive agents is corrected by infusion of antagonists of L-arginine (44-46), suggesting that overproduction of *NO may cause systemic vasorelaxation and shock. However, these data need to be verified by additional clinical trials. Excessive production of .NO may also lead to compromise of hemodynamic tone and to tissue injury through direct chemical modulation of proteins, ligands, and carbohydrates.

266

NICHOLAS W. LUKACS AND PETER A. WARD

C. PROTEINASES INVOLVED IN LUNG INJURY The release of proteolytic enzymes during inflammation in the lung appears to be a significant factor in damage to lung structures, resulting in mechanical dysfunction, The most important sources of damaging proteases are neutrophils and macrophages. During neutrophil activation by bacteria, proteases released from neutrophils function to slow, disrupt, and destroy bacterial growth and proliferation (41). However, a major problem associated with these responses is the damage and disruption of tissue integrity. The most potent of the neutrophil proteases is elastase (48).Neutrophil elastase functions at neutral pH and is capable of cleaving not only elastin but nearly all connective tissue proteins in the alveolar wall. Therefore, elastase by itself may be sufficient to destroy the integrity of the alveolar wall. A primary disease in which elastase appears to have a destructive role is emphysema. Airway instillation of pancreatic or neutrophi1 elastase will produce emphysema in a manner related to breakdown of elastin. Emphysema has been associated with circumstances linked to the inability to regulate the neutrophil elastase response (49).Of particular interest is al-antitrypsin deficiency, a hereditary disorder in which patients with specific haplotypes have 35% lower q-antitrypsin levels in serum. Interestingly, al-antitrypsin provides the lower respiratory tract with a defense against the damaging effects of neutrophil-derived elastase. The decreased presence of a1-antitrypsin in this disease does not allow sufficient protection against the adverse effects of elastase and is associated with progressive destruction of lung elastin. A typical individual with genetically linked al-antitrypsin deficiency will begin to experience symptoms of emphysema between the ages of 30 and 45 years (50). Neutrophils also release other proteases that can contribute to the alveolar damage and lung dysfunction. Neutrophil-derived collagenase can break down type I collagen, an important component of the alveolar wall. However, collagenase cannot by itself induce emphysema. In addition, collagenase can be produced by structural cells such as fibroblasts within the alveolar wall (47). Cathepsin G from neutrophils has many similarities to neutrophil elastase because it can also break down matrix components and elastin within the alveolar wall. The activity of cathepsin G, however, does not seem to be as potent as that of neutrophil elastase (51). Overall, it appears that the release of neutrophil-derived proteases can have a damaging effect on the integrity and function of the lung itself and likely contributes to the pathogenesis of multiple diseases that damage normal lung function. Matrix metalloproteases are proteolyhc enzymes specifically directed against extracellular matrix components. Similar to the above proteases,

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

267

metalloproteases are secreted by inflammatory cells and may significantly contribute to damage during acute pulmonary injury. These metalloproteases have been detected during endotoxin-induced inflammation in guinea pigs and appeared to be produced by both resident macrophages and recruited neutrophils (52). Furthermore, metalloproteases have been detected during hyperoxic exposure in murine lungs that corresponded to subsequent increased expression in mRNA for tissue inhibitor of metalloproteases (TIMP-1) (53).The tissue inhibitors of metalloprotease (TIMP1 and TIMP-2) may serve as a therapeutic option for controlling tissue damage mediated by the metalloproteases.

D. ADHESIONMOLECULES A N D PULMONARY INFLAMMATION The upregulation of selectins and other adhesion molecules (ICAM-1 and VCAM-1) on the vascular endothelium and on various tissues within the lung appears to be a primary initiating factor responsible for localization of leukocytes to a site of inflammation. Increased expression of these molecules has been described in idiopathic pulmonary fibrosis (54),sarcoidosis (55), adult respiratory distress syndrome (56), as well as in extrinsic allergic alveolitis (54).Interestingly, the presence of soluble adhesion molecules in the sera of patients has also been correlated with the disease severity. In particular, the levels of circulating soluble ICAM-1 were examined in patients with IPF, sarcoidosis, hypersensitive pneumonitis, or multiple pneumonias-only IPF patients displayed increased circulating levels of soluble ICAM-1 (57). In addition, levels of soluble E-selectin were elevated in the sera of patients with septic shock (58). Shedding of adhesion molecules into the plasma during inflammation in the lung and other organs may be indicative of the inflammatory process and may therapeutically serve to modulate the response due to decreased surface receptors/ ligands at a site of inflammation. Moreover, these soluble molecules may bind to their appropriate ligandheceptor and further downregulate the inflammatory response by competing with cell surface expressed adhesion molecules.

E. CYTOKINES Cytokines produced during an inflammatory response allow the cellto-cell communication that orchestrates recruitment, extravasation, and activation of immune and nonimmune cells. Studies over recent years have outlined a wide array of cell types that contribute to initiation, maintenance, and modulation of the inflammatory response. The initial expression of early response cytokines, IL-1 and TNF-a, mediates the upregulation of vascular adhesion molecules and subsequently activates inflammatory cytokine cascades. TNF-a and IL-1 production has been associated with

268

NICHOLAS W. LUKACS AND PETER A. WARD

both acute and chronic inflammatory disorders including septic shock, pulmonary and hepatic injury, rheumatoid arthritis, and parasitic and viral infections. Specifically in pulmonary inflammation, production of both TNF-a and IL-1 has been associated with asthma, IPF, sarcoidosis, ARDS, and cystic fibrosis. TNF-a was originally described by its ability to inhibit and/or directly kill tumor cells (57). However, the direct toxic effects of TNF-a appear to represent only a minor role compared to its homeostatic and inflammatory functions. As discussed previously, TNF-a has a major role in induction of adhesion molecule expression and inflammatory cytokine production. Other cytokines have also been described during pulmonary inflammation and likely mediate important functions during the response. An important concept that has emerged from various studies has been that resident stromal cells, by their expression of cytokines, can act as effector cells during an inflammatory response. These nonimmune cells were once viewed as bystander cells, primarly serving structural or architectural roles for the organs and tissues. However, this view has been revised because they have been identified as important cytokine-secreting cells. A cytokine network can be initiated through the production of early response mediators, such as TNF-a or IL-1, which affect surrounding stromal cell populations and lead to the production of chemotactic cytokines that then mediate recruitment of leukocytes to the site of inflammation (60). The continuous activation and production of the leukocyte-derived early response mediators, as well as immune-associated cytokines, can maintain the activation of the surrounding stromal cell populations and therefore leukocyte recruitment continues. This process can be maintained for as long as the initiating stimuli remains. A variety of cytokines are likely involved in initiating inflammation and have been extensively studied. Platelet-derived growth factor (PDGF) is found in abundance in a granules of platelets and functions as a mitogen, with its primary regulatory role directed at cell cycle activation. PDGF is also chemotactic for leukocytes and may be an early inducer of leukocyte migration into a site of inflammation (61).Although platelets appear to be the primary and most abundant source of PDGF, most cell types are capable of synthesizing and secreting this protein. Cells within the lung with PDGF receptors include fibroblasts, vascular smooth muscle cells, and epithelial and endothelial cells. Although the function of PDGF in the lung inflammatory response is not entirely clear, much of the evidence suggests an initiating role in pulmonary fibrotic diseases. This contention is supported by the original observation that PDGF induced 3T3 fibroblasts to proliferate and produce collagen (62).Because of this original observation, several other groups have implicated PDGF in fibrosis using human specimens and animal models.

MEDIATORS. CYTOKINES, AND ADHESION MOLECULES

269

Transforming growth factor$ (TGF-6) is a potent cytokine that has potent bidirectional modulatory mechanisms. TGF-P can inhibit cell activation (i.e., leukocytes), upregulate cellular proliferation, and induce matrix gene activation (i.e., collagen in fibroblasts) (63).These properties provide TGF-P with unique attributes that contribute directly to lung inflammation. The ability of TGF-P to inhibit endothelial cell proliferation and migration into connective tissue matrix suggests a blockade of endothelium regeneration (64, 65). TFG-P acts as a multistep activator of matrix genes, upregdating expression transcriptionally, translationally, as well as at posttranslational points. The production of TGF-/3, overall, has been linked directly to the progression of lung fibrosis in animal models and in human pulmonary diseases. The production of IL-6 during pulmonary inflammation has been well documented. IL-6 appears to have a broad range of effects on the immune system including plasma cell growth, B cell differentiation (66-68), induction of acute-phase proteins (67),induction of hematopoesis, enhancement of T cell activation (69), and facilitation of development of cytotoxic T cells (69). IL-6 is also associated with the memory T cell phenotype (70). IL-6 infused into rabbits causes fever and elevations in acute-phase proteins (71). In the lung, elevated IL-6 levels correlate clinically with allogeneic lung rejection (72), idiopathic pulmonary fibrosis, hypersensitivity pneumonitis, and asthmatic attacks. Within the lung, the primary cell populations that produce IL-6 in response to direct stimulation with IL-1 and TNF-a are the fibroblast and alveolar macrophage. These observations suggest that IL-6 may have direct activatingeffects and promote deleterious events within inflamed lung tissue. Currently, some of the most interesting cytokines are the chemokines. Induction of chemotactic factors seems to be pivotal for the extravasation of leukocytes from the vascular compartment into inflamed tissue. LTB4, platelet-activating factor, and C5a are well-studied chemoattractants with relatively little specificity for leukocyte populations. However, several chemotactic peptides selective for neutrophils, monocytes, T cells, or eosinophils have been described including chemotactic cytokines (chemokines). On the basis of chemical structure, these chemoattractants, which are listed in Table 11, have been divided into two distinct families, the C-X-C (a)and the C-C (p) families, designated by the position of the first two cysteine residues (22). The C-X-C family of chemokines is primarily chemotactic for neutrophils and contains a number of related molecules, as typified by interleukin-8. The C-X-C family members appear to be more closely related to acute inflammatory reactions and generally are characterized by their ability to initiate neutrophil influx into tissues. The C-X-C family includes IL-8 ENA-78, GROa,@,y,CTAP 111,NAP-2, MIP-

270

NICHOLAS W. LUKACS AND PETER A. WARD

TABLE I1 THECHEMOKINE FAMILY

c-x-c

c-c

IL-8 ENA-78 GROa,P,y CTAP I11 NAP-2 MIP-2 PF-4" IP-10" MIC"

MIP-la,P MCP-1,2,3 RANTES 1-309 c10 Eotaxin MARCSYK

~

Non-ELR motifs.

2, PF-4, IP-10, and MIG (Table 11).This family can be divided according to the presence of an E-L-R motif, which appears to confer the ability to attract neutrophils and provide an angiogenic signal (i-e.,endothelial cell growth and migration). The C-C chemokines appear to be primarily chemotactic for monocytes, lymphocytes, basophils, and eosinophils. The family contains several related proteins: MCP-1, MCP-2, MCP-3, MIPla, MIP-1P, C10, 1-309, Eotaxin, MARCFYK, and RANTES (73). The C-C family members are of particular interest in chronic airway responses because a correlation exists with the expression of C-C chemokines and chronic inflammatory diseases. The production of chemokines can be observed in nearly every cell after stimulation by classical cytokines (i.e., IL1and TNF-a) (22), during platelet aggregation (74),during hyperoxia (42), and during cell-to-cell interactions (75,76).The interest in these two groups of chemokines has become intense due to the possibilities of modulating inflammatory responses through the regulation of these molecules. As discussed previously, these proteins have a role in promoting the recruitment of specific leukocyte populations during inflammatory events. IL-8, a neutrophil-specific chemokine, has been shown to be associated with many acute and chronic inflammatory events in the lung. Accumulation of neutrophils in patients with cystic fibrosis has been directly correlated with IL-8 levels in the BAL fluid. Its presence is associated with clinical exacerbations in these patients (77). During ARDS, the primary inducer of lung damage may be the infiltrating neutrophil whose number correlate with IL-8 levels in the BAL fluid of these patients. In addition, increased levels of IL-8 have been observed in other disorders such as IPF and asthma (78,79).MIP-la has also been shown to increase during pulmonary inflammation, specifically in alveolar macrophages and isolated fibroblasts

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

271

from IPF patients (80). Other members of the C-C chemokine family have also been identified. 111. Animal Models of Lung Injury

A. SEPSIS

The clinical manifestations of endotoxemia and/or sepsis are usually the consequence of intense cellular interactions that often result in tissue injury and organ dysfunction (81). A primary organ targeted for injury is the lung. Septic conditions often lead to pulmonary complications, including ARDS, impaired lung function, and death. Despite significant advances in antibiotic treatment, intensive care unit technology, and mechanical ventilatory support, mortality due to sepsis-mediated multiorgan dysfunction has not significantly changed over the past three decades. The limited therapeutic options that are currently available for effective management of these patients directly reflect our poor understanding of the mechanisms underlying these pathophysiologic disorders. Thus, the key to improving morbidity and mortality requires an in-depth understanding of endogenous mediators that either initiate and maintain or regulate the responses. Endotoxin models of acute inflammation have been used in an attempt to understand the pathophysiologic pathways operative during septic conditions. The development of applicable animal models of sepsis may prove to be useful for determining the mechanistic and activational pathways that control the septic responses. One of the earliest pathways activated during sepsis is the complement cascade. The activation of complement in trauma patients with bacterially related septic shock syndromes has been correlated with the early stages of the response (82). In a rabbit model of septicemia, the levels of C5a have been correlated with the number of neutrophils migrating into the lung and with the degree of lung damage (83).The activation of complement leading to inflammatory split products, such as C3a and C5a, induces potent effects on neutrophil degranulation (84, 85). Moreover, inhibition of the complement system (either by complement depletion or infusion of sCR1, the soluble complement receptor) significantly reduces neutrophil recruitment, adherence, and degranulation, thus blocking release of destructive proteases and oxygen radicals (86). Using a Fischer rat model of endotoxemia, inhibition of complement (sCR1) completely blocked the upregulation of Mac-1 (CDllbICD18) on the surface of neutrophils (87), demonstrating a direct relationship between complement and neutrophil activation. Together, these studies indicate that activation of the complement cascade plays a significant role in the development of lung injury during sepsis.

272

NICHOLAS W. LUKACS A N D PETER A. WARD

A complex interaction of multiple mediators during sepsis has made it difficult to identify the relevant mechanisms that are operative during acute inflammatory responses. In critically iil patients with the septic syndrome, a corresponding respiratory burst in neutrophils compared to those obtained from nonseptic patients was observed, suggesting a relationship between reactive oxygen metabolites and development of septicemia (88).The main sources of oxygen radical formation within the lung appear to be the recruited neutrophil and activated residental macrophages. Formation of toxic oxygen products from neutrophils was attenuated using cyclooxygenase inhibitors in a swine model of LPS-induced lung injury (89). In a rat model of endotoxemia, in addition to the infiltrating neutrophils, alveolar and interstitial macrophages were both found to produce signficant levels of reactive oxygen intermediates (go),indicating that the resident macrophage populations also contribute to the production of oxygen radicals that cause cell and tissue damage. As discussed earlier, the release of reactive oxygen intermediates not only damages the local cell populations and connective tissue matrix, but these likely have an activating effect due to increased generation of inflammatory and chemotatic cytokines. The ability to recruit neutrophils into lung requires expression of adhesion molecules on the vascular endothelium, but the recruitment also appears to rely on increased avidity of P2-integrin molecules for ICAM1. In animal models of sepsis, the adhesion molecules appear to have a signficant role in the localization and activation of leukocytes (neutrophils in particular). In Cynologous monkeys, upregulation of E-selectin expression after LPS infusion has been found (91).E-selectin in the vascular endothelium was induced rapidly by 2 hr and peaked at 4 hr after LPS infusion. This rapid increase in E-selectin may mediate the initial neutrophil adherence or rolling response following endotoxin infusion. The second step to leukocyte adherence is the firm adhesion to vascular endothelium via binding to ICAM-1 by neutrophil p-integrins (CDllICD18). In the same model, administration of a blocking anti-CD18 antibody significantly attenuated neutrophil accumulation within the lung and alveolar capillary membrane injury and reduced the capillary leak. Anti-CD18 antibody treatment did not, however, reverse systemic hypotension or decrease cardiac index, pulmonary hypertension, or relative hypoxemia. These data suggest multiple mechanisms are operative during sepsis and begin to distinguish between neutrophil and non-neutrophil-dependent events (92). Using the same anti-CD18 antibody in a model of abdominal sepsis, although neutrophi1 accumulation was attenuated in the gut wall, this treatment did not protect against the weight loss, infectious complications, or mortality rates (93). Therefore, blocking of neutrophil accumulation alone may not, by itself, be sufficient for reversing the adverse effects of the septic response.

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

273

The production of inflammatory and chemotactic cytokines has been shown to correlate directly with the severity of the septic response in some patient populations. In particular, TNF-a, IL-1, and IL-8 levels in plasma correspond with those of the clinical outcome of patients with the septic syndrome. Animal models of endotoxemia have provided a number of mechanisms that are likely to be operative during human sepsis. In particular, neutralization of TNF-a at the onset of the septic response in mice has demonstrated significant decreases in subsequent mortality. TNF-a production during endotoxemia induces the appearance of other cytokines (94, 95) and chemokines (96-98) within the lung. Neutralization of TNFa in a porcine model of sepsis blocked the upregulation of CD18 on the surface of neutrophils and attenuated acute lung injury (99). TNF-a receptors have been observed to be released into circulation during septic responses (100) and appear to contribute to the regulation of TNF-ainduced activation through competitive binding of TNF-a. Synthetic soluble human p80 TNF-a cell surface receptor constructs have demonstrated effective treatment in vivo in endotoxin models of sepsis (101). These studies outline the importance of TNF-a in endotoxin-induced responses. In contrast, in a cecal ligation and puncture (CLP) model of sepsis, antiTNF-a antibody treatment failed to alter lethality (102), indicating differences between results in the endotoxin and active bacterial CLP models. Furthermore, in clinical trials the delayed administration of anti-TNF-a after the onset of sepsis has not been clinically useful. These latter data indicate that TNF-a may not be a therapeutic target for septic patients already exhibiting complications of multiple organ failure. This lack of protection in septic patients by anti-TNF-a treatment may be related to the temporal production of TNF-a that likely has activated the system and returned to background levels prior to anti-TNF-a treatment (103, 104). Another early response cytokine that appears to have similar functions as those of TNF-a during sepsis is IL-1. The production of IL-1 in animal models of endotoxemia occurs after the production of TNF-a (103) and IL-1 has also been considered a likely target for therapeutic control of septic shock/endotoxemia. The injection of IL-1 into the lungs of rabbits induced a shock-like syndrome similar to that observed during endotoxemia (105),suggesting a contributing role for IL-1 during the septic response. In contrast, when recombinant IL-1 and TNF-a were injected into canine lungs, only TNF-a induced lethal injury (106), suggesting a minor role for IL-1 in this model. However, in a study of endotoxin-induced mortality, an IL-1 receptor antagonist ( IL-ha) significantly reduced mortality in rabbits (107), suggesting that IL-lra may be an appropriate therapy during sepsis. Although the use of IL-lra held great promise for control of septic

274

NICHOLAS W. LUKACS A N D PETER A. WARD

responses, clinical trials have suggested that IL- Ira is not useful therapy for septic shock patients. Another target for inflammation intervention is the family of chemotactic cytokines such as IL-8 (produced later in the response and involved in induction of recruitment of leukocytes into the lung). The recruited and activated leukocytes damage the surrounding cells in the lung and damage tissue integrity, resulting in organ dysfunction. Macrophage inflammatory protein-2 is a member of the C-X-C chemokine family, is chemotactic for neutrophils, and may be the murine homologue for IL-8. Previous studies have demonstrated increased MIP-2 production during LPSinduced endotoxemia in animal models. The production of MIP-2 during a septic response in mice may be a pivotal stimulus for the recruitment of neutrophils (108). Inhibition of MIP-2 by antibody has demonstrated significant attenuation of mortality in a mouse model of endotoxemia. However, serum MIP-2 levels peaked within 30 min after LPS injection, similar to changes in TNF-a levels, possibly suggesting difficulties in a clinical setting similar to those with TNF-a in which early detection is difficult. In additional studies examiningthe role of chemokines in endotoxemia or airway LPS challenge, MIP-la (109) and RANTES (96, 97), two C-C family chemokines, were shown to be expressed during endotoxemia. The inhibition of these chemokines in vivo attenuated leukocyte infiltration into the lungs and blocked the lethality associated with endotoxemia. In general, chemokines appear to play a significant role in septidendotoxin responses. These cytokines are produced later in the response, in a temporal pattern that would allow therapeutic application to block leukocyte recruitment and thus lung damage. The ability to regulate cytokine expression during sepsis may be one of the most promising strategies for treating septic responses. During septic responses in patient populations (110),a suppressive cytokine, IL-10, has been shown to be produced in high levels by monocyte/macrophage cell populations (111). IL-10 is a potent suppressive agent for monocyte/macrophages and downregulates multiple inflammatory cytokines, including TNF-a, IL-1, IL-6, and chemokines. In turn, adhesion molecule expression is blunted. When injected in combination with LPS into mice, IL-10 protected the mice and attenuated the lethality normally observed (112). In a recent study, peak IL-10 levels were observed in serum and in lungs of mice 4-6 hr after endotoxin infusion, corresponding to the decrease in appearance of TNF-a and MIP-2 in lungs. Administration of anit-IL-10 antibody signficantly increased lethality, TNF-a and MIP-2 levels, and myeloperoxidase (MPO) content in lungs, suggesting a role for endogenous IL-10 in regulation of the endotoxin-induced responses (108).Use of gene therapy to induce IL-10 expression may represent a future approach for

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

275

the treatment of the sepsis syndrome. Using a liposome transfer system 48 hr prior to an endotoxin challenge, IL-10 intratracheal gene transfer reduced pulmonary TNF-a levels by 62% and neutrophil infiltration by 55%and blocked the lethality of the response (113).Together, these studies indicate that IL-10 may be a potent therapeutic option for use in the clinical setting of sepsis.

B. ISCHEMINREPERFUSION LUNGI N J U R Y Ischemia/ reperfusion injury of lungs most often occurs during cardiopulmonary bypass and during lung transplantation in which blood flow is interrupted (114). Removal of a thrombotic obstruction in a pulmonary vessel followed by reprefusion of blood (115),expansion of an endematous lung, during pneumothorax (116), or ARDS after resolution of vascular thrombosis (117) can all result in ischemidreperfusion (IIR) injury to develop. In addition, it appears that VR injury to the intestinal tract or limbs can affect lungs in a manner similar to that induced in local pulmonary injury. Regardless of the nature of the clinical circumstance, lung vascular injury development causes increased microvascular permeability and intravascular leukocyte aggregation (118).Mechanisms that result in vascular damage in the lung from induction of IR have been evaluated by the use of animal models of I/R. One of the initial events occurring during VR is the activation of the complement system. The release of C3a and C5a leads to leukocyte activation within the vascular compartment. The most convincing evidence that complement plays a role in organ injury after L/R can be found in studies in which complement receptors were blocked following infusion of sCRl (119,120). Soluble CR1 blocks activation of the complement by interfering with assembly of the C3 and C5 convertases. In these studies, the VR response was intiated in intestine or hindlimbs. The addition of sCRl blocked pulmonary vascular permeability and edema. The mechanism of complement-mediated damage was likely indirect through the recruitment and activation of neutrophils. As previously discussed, neutrophils release oxygen radicals and proteases, which induce tissue damage. The pulmonary dysfunction that results from local or distal VR correlates with the neutrophil accumulation in tissues (121). In a model of I/R injury in rats in which blood flow in the hindlimbs is interrupted for up to 4 hr followed by reperfusion, injury occurs both locally (in the limbs) and distally (within the lung). Complement activation occurs, as reflected by falling serum CHm levels. Concomitantly, blood neutrophils show evidence of activation as reflected by increased surface P2-integrin (CDllbICD18; Mac-1) expression. Tissue injury is clearly neutrophil dependent. On the basis of the use of oxygen radical blocking reagents, the injury is also

276

NICHOLAS W. LUKACS AND PETER A. WARD

oxygen radical dependent. Studies have indicated that neutrophils are not necessary in some cases of VR lung injury (122, 123) and that oxygenderived radicals can be derived from nonneutrophil cell populations. However, the neutralization of IL-8 in a rabbit lung model of VR demonstrated totally abrogated neutrophil influx and lung reperfusion injury, linking neutrophils to lung injury in this particular model (124). Whatever the source, oxygen metabolites appear to play a major role in VR lung injury. Increased levels of HzOl can be detected in BAL fluids from lungs of animals after VR, and the damage can be signficantlyattenuated by treating the animals with catalase or other oxygen radical scavengers (125-127). In addition, xanthine oxidase, a member of the flavin-linked enzymes responsible for HzOzproduction, is also elevated during I/R and inhibition of this enzyme signficantly attenuates the injurious response (128, 129). Arachadonic acid metabolites, such as thromboxane A2 (130), may have a role in I/R responses in the lung. Inhibition of cyclooxygenase metabolite production can attenuate I/R-mediated damage ( 131).Finally, it appears that nitric oxide (endothelium-derived relaxing factor) can also decrease the effects of I/R in a rat model (132). This latter effect may be due to the ability of nitric oxide to inhibit neutrophil adherence to vascular endothelium. Localization of leukocytes to a site of inflammation relies on the initial adherence events between leukocytes and the vascular endothelium. The role of selectins in lung and limb I/R injury suggested differential roles for these molecules. Neutrophil recruitment into the affected tissues depends on CDlldCD18 and CDllb/CDlS as well as ICAM-1, along with L- and E-selectins (133).In the neutrophil recruitment process, P-selectin also appears to be involved but only transiently. The early response cytokines, TNF-(I! and IL-1, can be detected in plasma approximately 2 hr after the reperfusion process, and they appear to be vitally involved in the outcome of tissue injury because antibody-induced blockade substantially reduces neutrophil buildup in the affected tissues, resulting in protection. It appears highly likely that a major role for these cytokines is upregulation of vascular E-selectin and ICAM-1. In a sheep model of lung VR, administration of an antibody that blocks both E- and L-selectin improved survival (134). Blockage of ICAM-1 (CD54) interactions during the development of VR lung injury induced a decrease in lung tissue MPO, whereas lung weight increase was not affected (135).Likewise, blocking CD18 inhibited injury mediated by both lung and lower torso during UR-induced inflammation (135, 136). In another study of lung inflammation after intestinal VR, anti-CD18 prevented pulmonary injury but did not affect systemic hypotension (137), suggesting separate mechanisms for lung injury and impairment of cardiovascular function. The expression of adhesion mole-

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

277

cules may be secondary to the production of TNF-a and IL-1 (138, 139), which are potent inducers of adhesion molecule expression. Overall, these studies suggest that blockade of adhesion molecules (E-selectin and ICAM1)or leukocyte adhesion-promoting molecules (L-selectin, CDlldCD18, and CDllb/CD18) may be effective in treatment of I/R-mediated lung injury. C. IMMUNE COMPLEX-MEDIATED LUNGINFLAMMATION Acute lung injury as well as chronic diseases, such as Wegner’s granulomatous, sarcoidosis, and several other pulmonary and nonpulmonary diseases (SLE), may in part be linked to the presence of immune complexes (140). The injury that is induced by immune complex deposition in lung includes a vascular leak syndrome, significant recruitment of leukocytes, leukocyte activation, and damage of alveolar walls. Intrapulmonary deposition of IgG or IgA immune complexes in rats leads to acute lung injury to vascular and alveolar epithelium (141, 142). In the former, within 3 or 4 hr a severe hemorhagic alveolitis ensues with intense infiltration of leukocytes (neutrophils) in the interstitial or intraalveolar space. Subsequently, this inflammation results in the loss of lung integrity in the alveolar capillary wall. A better understanding of the events that regulate the development of the inflammation and injury may result in novel therapeutic applications to combat immune complex-mediated diseases. Neutrophils that are recruited into the site of immune complex deposition have been shown to be dependent on the class of antibody that was present. The deposition of IgG facilitates the local accumulation of neutrophils (Fig. 4A), whereas deposition of IgA immune complexes appears to activate residental macrophages (Fig. 4B) (141, 142). Regardless of the type of leukocyte associated with the injury, the damage induced during the inflammatory response appears to be mediated by toxic oxygen radical release from phagocytic cells (143-145). The activation pathways of the IgG and IgA immune complex models differ substantially. In both the IgG and the IgA immune complex models of lung injury, complement activation occurs. When rats with IgG or IgA immune complex-induced inflammation in lungs were pretreated with soluble complement receptor-1, a marked attenuation of the inflammatory damage was observed (146). The activation of residental macrophages also appears to have a differential role between the two models. In the IgG immune complex model, a substantial upregulation of TNF-a and IL-1 appears to be associated with the upregulation of selectin (E-selectin) and adhesion molecules (ICAM-1) resulting in recruitment, activation, and degranulation of neutrophils into the alveolar compartment. The activation and degranulation of migrated neutrophils and activation of residental macrophages releases multiple

278

NICHOLAS W. LUKACS AND PETER A. WARD

FIG.4. Transmission electron micrographs of lungs from rats 4 hr after intrapulmonary deposition of IgG (A) or IgA (B) immune complexes. In the former, large numbers of neutrophils (arrows) are present in the interstitial and intraalveolar spaces. Within the alveolar space are also red blood cells (closed arrowheads) as well as protein and fibrin deposits (open arrowheads). In the case of deposited IgA immune complexes (B), few if any neutrophils are present, along with intraalveolar hemorrhage, cell debris, and protein. (Osmuim tetroxide stained, X3500).

oxidants (O,., HO., *NO,ONOO., etc.) and proteases (serine proteases and metalloproteases), leading to structural damage of cells and connective tissue matrix. In the IgG immune complex model, the associated injury could be signficantly attenuated by the administration of catalase (143). Together, these studies suggest that activation of neutrophils and macrophages results in release of oxygen radicals (primarily HzOzand hydroxyl radicals) that are crucial events leading to lung damage. In related studies the release of nitric oxide, a free radical generated form L-arginine, was described to participate in damage in both IgG and IgA immune complexinduced lung injury (145).The expression of both TNF-a and IL-1 production was required for the most advanced injury in the IgG immune complex model (147). In these same studies, the use of antibodies directed against IL-1 and TNF-a or administration of IL-lra diminished pulmonary MPO levels as well as the related injury. In contrast, use of antibodies to TNF-

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

279

LY does not appear to alter the inflammation induced by IgA-mediated immune complexes (147). In the IgA immune complex model, the activation of resident pulmonary macrophage populations leads to the production of MCP-1, a member of the C-C chemokine family, followed by the generation of oxidants and release of metalloproteases which in turn damage cells and connective tissue matrix in a similar manner as in the IgG immune complex model. The role forP2-integrins (CDllICDllb) in these two models ofimmune complex-mediated inflammation has also demonstrated interesting differences. The participation of the P2-integrins in the damage appears to be twofold. In the IgG model, CDlldCD18 (LFA-1) appears to be responsible for the interactions of neutrophils with the endothelial expressed ICAM-1, leading the recruitment of neutrophils into the alveolar space. In addition, CDllbICD18 (Mac-1) present on alveolar macrophages (which may facilitate macrophage-alveolar epithelial cell interaction) appears to facilitate the production of TNF- and IL-1 in the IgG immune complex model. Airway blockade of either CDllbICD18 or ICAM-1 greatly reduced the production of TNF-a and IL-1 by pulmonary macrophages resulting in the reduced recruitment of neutrophils. Furthermore, in both the IgA and the IgG immune complex-induced inflammation, antiCD 18antibodies significantlyattenuated the inflammatoryresponses ( 148). These latter studies demonstrate that although the mechanisms and leukocyte populations involved in the development of the lung injury differ between the two models (IgA vs IgG), a common role of p2-integrins was involved in the inflammatory response. It seems probable that the requirement for CDllbICDl8 (Mac-1) and ICAM-1 in the airway compartment of the lung following intrapulmonary deposition of immune complexes is related to adherence interactions between macrophage Mac-1 and ICAM-1, which are constitutively expressed on alveolar epithelial cells. This tethering may serve to enhance cytokine production in activated alveolar macrophages. Evaluation of the regulatory pathways involved in the immune complexinduced inflammation may shed light onto particularly relevant therapies in conditions of acute inflammation involving a variety of organs (lung, joint, renal glomeruli, skin, etc.). In a study of the IgG immune complex model of lung injury, the use of IL-10 (The-type cytokine) was highly suppressive in a manner associated with reduced levels of BAL fluid TNFa as well as diminished influx of neutrophils (149). In addition, the administration of IL-4 suppressed the IgG immune complex-induced inflammation, further demonstrating differences between these two models. Future studies should elucidate the mechanisms by which these latter TH2-type cytokines function to regulate the inflammatory response.

280

NICHOLAS W. LUKACS AND PETER A. WARD

D. PULMONARY GRANULOMA FORMATION Granulomatous inflammation is characterized by an accumulation of leukocytes around an infectious or noninfectious agents (150, 151). The cellular constituents of granulomas include immune as well as nonimmune cell populations including macrophages, lymphocytes, mast cells, epithelioid cells, and fibroblasts. In addition, multinucleated giant cells are common in delayed-type hypersensitivity granulomas; eosinophils are often found in chronic parasite-induced granulomas. Each of these cell populations likely contributes to the overall pathology and tissue damage associated with granulomatous reactions. The residual fibrosis that accompanies the resolution phase of the granuloma may result in irreversible tissue damage and organ dysfunction. Noninfectious foreign body-type granulomas can be induced by environmental or industrial agents, such as talc, silica, or beryllium, whereas infectious granulomas are induced by bacteria (tubercle bacilli), fungi, viruses, or parasites (leishmania, schistosomal). In addition, other classes of granulomatous inflammation that have been associated with sarcoid and Wegener’s granulomatous have been classified as “nonspecific” immune-type reactions. These latter responses are of unknown etiology, but they have classical immune-associated characteristics (high rate of cellular turnover and cytokine-producing lymphoid cells). The treatment strategies for infectious lesions include the elimination of the infectious agent, whereas in most cases of noninfectious granuloma formation, treatment with glucocorticoids effectively attenuates the reaction presumably by downregulating the inflammatory response. The limited options for treatment largely reflect the paucity of information concerning the mechanisms that are operative during the formation of granulomas. The initiation of granuloma formation involves multiple mechanisms and inflammatory mediators. Reactive oxygen metabolites appear to play an important role in the development of granulomas. In experimental pulmonary granuloma formation, synchronously developing immune lesions spontaneously release significant quantities of superoxide anion (compared to nonimmune foreign body-type granulomas). In addition, daily administration of the oxygen scavenger, d-tocopherol, suppresses lesion development by >60%. Moreover, addition of specific inactivators of oxygen and HzOz (superoxide dismutase and catalase, respectively) results in a 30-40% reduction in granuloma size (152). These studies clearly suggest a role for reactive oxygen species during pulmonary granuloma formation and may reflect the ability of these molecules to initiate important cascades that result in production of inflammatory mediators during delayed-type hypersensitivity responses. Some of the most potent classes of inflammatory cytokines in hypersensitivity granuloma formation have been identified as early response cytokines,

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

281

IL-1 and TNF-a. Studies of TNF-a production during schistosome egg and tuberculoid granuloma formation indicate that the production of inflammatory cytokines correlates well with granuloma development ( 153, 154). Most convincingly, in vivn TNF-a, blockade inhibits development of mycobacterial granulomas (155),whereas exogenous administration of TNF-a into nonresponsive SCID mice reconstitutes the ability of these mice to mount a circumovum schistosome egg granulomatous response (156). This latter response in SCID mice appears to be similar to a foreign body-type reaction, consisting primarily of mononuclear cells. TNF-a has also been shown to induce the expression of ICAM-1 in the lung, which is required to mediate the granulomatous response and accumulation of leukocytes around the egg nidus (157). The role of IL-1 in granulomatous lesions has also been investigated. Increased mRNA expression and production of IL-1 protein has been associated with the development and severity of several types of granulomatous diseases including those associated with sarcoidosis (158),schistosomiasis (159), and leishmaniasis (160), as well as in foreign body reactions. The regulation of IL-1 during granulomatous lung diseases may be crucial for limiting the size of the lesion. Lung biopsies from sarcoid patients display intense immunostaining of IL-lra within granulomatous lesions ( 161).The neutralization of endogenous IL-1 receptor antagonist accelerated lesion development and severity ( 162). Together, these data suggest that the regulation of IL-I is crucial for controlling the size of granulomatous lesions. Other cytokines also appear to regulate the development and resolution of granulomas. Immune-associated cytokinesAymphokines, such as JFNy , IL-4, IL-10, and IL-12, have a striking ability to regulate the granuloma formation, leukocyte infiltration, chemokine production, and possibly endstage fibrosis of the developing lesions. As indicated previously, these lymphokines can be classified as Thl- (type 1) or Th2 (type 2)-associated cytokines (27). The type 1 cytokines, IL-12 and IFN-y, appear to be associated primarily with intracellular infectious agents, such as mycobacteria, which depend on intracellular killing by phagoc9c cells to clear infectious agent. In intracellular infections, the ability to produce type 1 cytokines dictates the success of clearance of the agent and granuloma resolution (163).The Th2-related cytokines, IL-4 and IL-10, are associated most closely with parasite-elicited responses such as those observed in schistosomiasis (164) and leishmaniasis (165).The production of Th2-type cytokines during intracellular infections may promote prolonged granuloma formation and increased lesion size leading to exacerbated fibrotic responses. Utilizing a functional analysis of granuloma formation, laboratories have begun to classify granulomatous diseases by the pattern of cytokines

282

NICHOLAS W. LUKACS AND PETER A. WARD

produced during the response. These specific responses may dictate the pattern of chemotactic factors produced and, therefore, the leukocytes recruited. In a glucan-induced foreign body granulomatous inflammation in rats, a signficant mononuclear cell infiltrate (present in an angiocentric location) has been associated with the expression and production of a C-C chemokine, MCP-1 (166, 167). Neutralization of MCP-1 in these animals significantly decreased the size of the lesions and the intensity of the mononuclear cell infiltrate (166).The regulation of MCP-1 and mononuclear cell recruitment appears to be via IL-lP and TNF-a production (168). A significant attenuation of glucan-induced lesions was obtained by passive immunization of animals with anti-IL-1 or anti-TNF-a. The expression and production of MCP-1 were significantly attenuated when IL-1 and TNF-a! were neutralized in uiuo, suggesting the presence of cytokine cascades leading to the accumulation of mononuclear cells. Chemokine production in the schistosome egg-mediated granuloma has been associated with two chemokines that demonstrate differential participation at distinct stages of lesion development. MIP-la, shown to be chemotactic for monocytes, T lymphocytes, and eosinophils,is found within the primary and/or secondary granulomas. Initial observations demonstrated significant induction of MIP-la mRNA in the primary and secondary lesions that corresponded to the growth patterns of granuloma formation (169). In viuo neutralization of MIP-la during primary and secondary granuloma formation produced different outcomes. During primary granuloma formation, the size of the lesion was reduced >60%, whereas during secondary granuloma formation the size was reduced by only 15-20%. In contrast, MCP-1 displayed a different profile during granuloma formation. Isolated egg granulomas demonstrated very high constitutive levels of MCP-1 in secondary, but not primary lesions,which directly correlated with the size of the Th2 cytokine-mediated secondary lesion (170). Interestingly, when MCP-1 was neutralized in uiuo,there was no evidence of diminished granuloma formation of primary lesions, but a significant decrease was observed in secondary lesions. When lung tissue sections were immunostained for MIP-la and MCP-1, the predominant expression of MIP-la was in mononuclear cell populations, whereas MCP-1 was immunolocalized to vascular smooth muscle cells within the secondary lesions (170). Results from neutralization of MCP-1 and MIP-la suggested that specific chemokines were operative at certain stages of the immune response, possibly related to the type of immune response involved (Thl vs Th2 type). In addition, the immunolocalizationof chemokines in leukocytes or mesenchyma1 cells suggests that certain cell types contribute at different phases of granuloma formation. The latter results suggest a prominent role for stro-

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

283

mal cell-derived chemokines (171) within chronic lesions, resulting in sustained leukocyte recruitment. These observations may extend to other diseases with distinct phases or cytokine profile associations and suggest a differential role of chemokines during granuloma formation and other inflammatory reactions. E. ALLERGICAIRWAYINFLAMMATION Allergic pulmonary diseases affect significant numbers of humans worldwide and include diseases such as asthma, pulmonary eosinophilia, and bronchopulmonary mucomycosis ( 172). The inflammation induced during allergic airway inflammation is mediated by the coordination of several immune-specific events. The initial induction of IgE-mediated mast cell degranulation, upregulation of adhesion molecules, and production of inflammatory cytokines resulting in infiltration of specific leukocytes are orchestrated in a sequential manner. The long-term pathological effects of asthma have been attributed, in part, to infiltrating leukocytes that surround the bronchus and infiltrate into the bronchiaVbronchiolar wall (173, 174). Immune responses associated with asthma produce histopathological features of a chronic, cell-mediated immune reaction characterized by the infiltration of the bronchial mucosa with neutrophils, basophils, eosinophils, macrophages, and lymphocytes (173). Eosinophils are considered to be the cells chiefly responsible for the production of bronchial mucosal injury and bronchial dysfunction and are thought to induce the bronchial constriction responses associated with the asthmatics (172, 173, 176, 177). In atopic asthmatic inflammation, a corresponding production of Th2 cell-type cytokines (IL-4 and IL-5) has been observed that correlates with disease intensity and eosinophil infiltration ( 176-183). The expression of IgE and the accumulation of eosinophils are both characteristic of allergic responses (184, 185) and correspond directly to the production of the The-type lymphokines. Depletion of CD4+ T cells in a mouse model of allergic airway inflammation results in an abrogation of eosinophilia and a reduction in airway hyperreactivity (186), verifylng the importance of T cells in allergic inflammation. Examination of the mechanisms of cellular recruitment in human asthmatic airway inflammation has been difficult. This is primarily due to the difficulty in performing longitudinal studies in patient populations. Animal models of airway inflammation have been used to examine the allergic response in the lung. Several groups have employed various antigens in dogs, subhuman primates, and guinea pigs (187-194), whereas others have utilized pharmacological reagents to induce airway reactivity (195, 196). These models have provided data that would not be attainable in humans including the effects of antiasthmatic drugs, longitudinal studies

284

SICHOLAS \Y. LUKACS A S D PETER A. \YARD

of airway reactivity after challenge, the nature of cellular infiltrates, and examination of the late asthmatic inflammatory response. A primate model of ascaris antigen-induced airway inflammation appears to be a useful model, with responses similar to those of human disease. In this model, ICAM-1 was identified as an important adhesion molecule leading to eosinophil recruitment into the airway ( 197). Subsequently, other animal models have been used to examine the role of adhesion molecules in airway inflammation. In a Brown Norway rat model of ovalbumin-induced airway inflammation, anti-ICAM-1 treatment demonstrated significantly reduced airway responsiveness without a decrease in eosinophil or lymphocyte recruitment (198). A mouse model of ovalbumin-induced airway inflammation has demonstrated that VLA-WCAM-1 interactions are paramount to induction of eosinophil accumulation in the airway wall (199). Interestingly, treatment of the mice with anti-ICAM-1 and anti-LFA-1 did not inhibit eosinophil recruitment in the lung interstium, possibly suggesting species differences and/or the specificity of the antigen used to induce the response. Reasons for these discrepent results are not clear. Perhaps the most widely utilized models for examining airway responses and hyperreactivity have involved rats and guinea pigs. Within these models, animals are either sensitized with alum-precipitated ovalbumin and challenged with nebulized aerosols containing relatively high concentrations of dissolved ovalbumin or challenged directly into the airway with a known airway irritant. The resulting responses include immediate and delayed airway hyperreactivity, peribronchial inflammation, and associated late-phase eosinophil recruitment. The induction of airway responsiveness could be initiated using a number of early response mediators including leukotrienes (200), products of xanthine/xanthine oxidase (201), bradykinins (202), IL-lP (203,204), as well as multiple chemicaVenvironmenta1 mediators of airway responsiveness. These studies demonstrate that multiple mediators contribute to the overall airway responsiveness and hyperreactivity during inflammatory responses within lung and airway. In addition, the rat and guinea pig models have also been utilized for examining immune-specific cytokine reactivity. In particular, IL-5 has been shown to be involved in the recruitment of eosinophils in and around the airway (205). Although relatively few reagents have been developed for guinea pig products, significant advances have been made related to induction and prevention of airway inflammation and hyperreactivity. The use of mouse models of allergic airway inflammation has the advantage of defining mechanisms of inflammation, in part because of the availability of a wide array of reagents. The mouse models of allergic inflammation have followed procedures established in guinea pig and rat models utilizing alum-precipitated ovalbumin for sensitizing the animals. Studies

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

285

have demonstrated a role for IL-4 and IL-5 in eosinophil recruitment and airway hyperreactivity (206-208). In a recent study, the role of mast cells in allergic airway responses was examined. The sensitization of mast celldeficient mice demonstrated a decreased eosinophil as well as a diminished airway hyperreactivity response to allergen, suggesting an important mechanism in allergic responses mediated via mast cell activation. In a model similar to the primate ascaris model, a Schistosomu mansoni soluble egg antigen (SEA)was utilized to induce allergic responses in lungs. This model has demonstrated the requirement for IL-4 (209) and TNF-a (210) as early response cytokines for the inflammatory reaction in the lung and airway responsible for eosinophil recruitment. In addition, the SEAinduced model has identified two C-C chemokines, MIP-la (211) and RANTES (N. Lukacs, unpublished data), which together mediated eosinophi1 recruitment and accumulation around and within the airway (Table 111). In contrast, neutralization of MCP-1 did not alter the recruitment of eosinophils. Interestingly, MIP-la and RANTES appear to have a specific effect on eosinophils because mononuclear cell accumulation within the lung is not decreased when these two chemokines are neutralized. In contrast, neutralization of MCP-1 significantly attenuated the lymphocyte recruitment into the lung and around the airway. This latter observation possibly suggests a specific role of chemokines during particular types of inflammation affecting only one cell population. The use of the mouse models of airway inff ammation appears to have many advantages over that of the other models of inflammation; however, as with all animal models, their relationship to asthmatic responses in patient populations has been questioned. IV. Therapeutic lnkrventions

The treatment of pulmonary inflammatory conditions has been constrained by a lack of understanding of the mechanisms that are operative TABLE I11 CELINLAR SOURCESOF EOSINOPHIL-RELATED IN ALLERGIC AIRWAY INFLAMMATION CYTOKINES Cytoldne

Cellular Source of Cytokine

IL-4 IL-5 TNF-a MIP-la RANTES

Mast cells, T lymphocytes Mast cells, T lymphocytes Mast cells, alveolar macrophages Airway epithelial cells, macrophages Epithelial cells

286

NICHOLAS W. LUKACS AND PETER A. WARD

during disease onset or progression. The use of animal models offers a unique opportunity to evaluate specific reagents for their effectiveness as well as determine the mechanism of action during pulmonary inflammation. As outlined in various animal models, multiple mediators have been identified that appear to be appropriate targets for therapeutic intervention in many of the diseases commonly found in patient populations. Interestingly, most of these mediators, including complement cascade products, reactive oxygen intermediates, multiple inflammatory and chemotactic cytokines, as well as particular adhesion molecules, all have an intricate role in the development of the inflammation that damages lung structure and function in most of the diseases. Interruption of any one of these mediators may be sufficient to alleviate inflammation-induced damage within the lung. However, determining the proper reagent for a particular disease may be crucial for the appropriate intervention (Table IV) as each of the mechanisms may be related to a particular type of injury. One of the primary events occurring during pulmonary inflammatory responses is activation of the complement cascade. Because the complement breakdown products, especially C3a and C5a, have potent activating and chemotactic effects, the inhibition of these events may sufficiently block the subsequent inflammatory response. Pulmonary diseases, such as ARDS, IPF, and COPD, all have been reported to be associated with complement activation products. The activation of complement by itself constitutes an important mechanism for lung inflammation. This has been elucidated using animal models of systemic complement activation such TABLE IV POTENTIAL THERAPEUTIC INTERVENTIONS IN PULMONARY DISEASES Therapeutic

Disease Process

Target( s)

sCRl

C3a, C5a receptor

SLPI TIMP-1, TIMP-2 Antiselectin

Serine proteases Metalloproteases E-, L-, and P-selectins

Antiadhesion molecule

ICAM-1, VCAM-1

Anti-TNF

TNF-D

IL-lra IL-10

IL-1 receptor Macrophages

Immune complexes, hyperoxia, sepsis (ARDS), reperfusion injury ARDS, emphysema ARDS, emphysema ARDS, immune complexes, reperfusion injury, hyperoxic inflammation ARDS, granulomatous (DTH), allergic (Asthma) Possibly all diseases (timing is important)

? Sepsis (ARDS), immune complexes, granulomatous (DTH)

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

287

as with cobra venom factor (CVF) administration (212). The CVF model of inflammation has been effective at demonstrating the role of complement cascade activation in initiating an inflammatory response (Fig. 5 ) . As discussed in previous sections, the most effective manner to inhibit the inflammatory events associated with Complement products is by complement depletion or complement blockade [with soluble complement receptor-1 (SCRl)].SCRl competitively binds the reactive complement convertases inhibiting subsequent activation events in leukocytes. The specific inhibition of oxygen radicals with the use of catalase or superoxide dismutase may offer another mode of therapy to reduce the inflammatory lung injury. Because both complement activation and oxygen radical release are early events in inflammation, the timing of treatment protocols may be critical. The damage to airway epithelial cells and vascular endothelial cells during pulmonary inflammation may be a result of neutrophil-released oxidants as well as proteases including elastase and cystein proteases. These proteases may be a candidate target for therapeutic blockade using specific inhibitors for inactivation of the damage. Secretory leukoprotease inhibitor

FIG.5. Transmission electron micrograph of rat lung 30 min after intravenous infusion of cobra venom factor. Neutrophils (arrows) in close contact in the capillary walls are present. There is extensive evidence of endothelial cell destruction, intravascular fibrin deposits (open arrowhead), intraalveolar hemorrhage, and fibrin deposits (closed arrowheads). (Osmium tetroxide stained, X5600).

288

NICHOLAS W. LUKACS AND PETER A. WARD

(SLPI) blocks neutrophil elastase activity and is produced endogenously by distal bronchial epithelial cells (213). Aerosolization of recombinant SLPI inhibits pulmonary epithelial cell damage (214). Using a hamster model of LPS-induced emphysema, the intratracheal instillation of SLPI blocked the airway damage and the appearance of emphysema-like findings (215). Metalloproteases, such as collagenase, stromelysin, and gelatinase, have also been implicated in pulmonary disease and lung damage. The activity of these proteases is regulated in vivo by endogenously and constitutively expressed tissue inhibitors of metalloproteases, TIMP-1 and TIMP2. These two inhibitors have been found to be associated with resolution of inflammatory responses and reduction of tumor mass (216). TIMP-1 and TIMP-2 demonstrate therapeutic utility in diminishing the damage induced by IgG immune complexes (217). Adhesion molecule expression is the initial step in localization and extravasation of leukocytes into tissue. These molecules may represent appropriate targets for therapeutic intervention. Selectin molecules have been shown to be essential in vivo for the intial localization on endothelial surfaces of leukocytes, whereas adhesion molecules, such as ICAM-1, appear to mediate a firm adhesion event on the vascular endothelium leading to subsequent tissue transmigration. The therapeutic effectiveness in binding of endothelial selectins or ICAM-1 in the experimental models of lung inflammation was described. To date, no clinical trials in humans involving adhesion molecule blockade have been completed in lung inflammation. However, a study of blockade of ICAM-1 in patients with rheumatoid arthritis has demonstrated efficacy (218).The fact that selectin and ICAM-1 molecules are released into circulation during inflammatory responses may be indicative of the upregulation of these molecules. The low levels in plasma of these molecules do not likely cause competitive inhibition of inflammatory responses. The use of exogenous soluble receptor proteins to competitively block the crucial interactions between leukocytes and endothelial cells may provide a substantial reduction in lung inflammatory responses. These same interactions can also be blocked by use of antibodies specific for various molecules. The development of cytokine networks is probably an important aspect for the initiation and maintenance of inflammation. These cascade networks allow several points of regulation and potential intervention. During inflammation, release of cytokine receptors into tissue and circulation has been observed. The importance of these soluble receptors in blocking inflammatory responses is emphasized by the observation that various pox viruses encode polypeptides with functional and structural homology to the extracellular sections of the TNF-a, IL-1, and IFN-7 receptors, suggesting that these viral-derived receptors aid the virus in evading the host's

MEDIATORS, ClTOKINES, AND ADHESION MOLECULES

289

immune response. The production of these “captured’ genes produced during viral infection likely aids the virus in limiting the normal immune response (219,220).Soluble receptor proteins specific for TNF-a and IL-1 have also demonstrated significant attenuation of inflammatory responses including those observed in models of rheumatoid arthritis, pulmonary granulomatous inflammation, allergic airway responses, as well as in other models of disease. Other endogenously produced regulators of the immune response have also been identified. In particular, IL-1 receptor antagonist protein (ILIra) has demonstrated modulatory aspects of the inflammatory response. Within the lung, high levels of IL-lra are produced constutitively by alveolar and interstial macrophages. In contrast, monocytes isolated from peripheral blood produce no IL-lra constitutively; however, it can be activated by LPS, TNF, or IL-1 to produce high levels. IL-lra functions as a true effector antagonist without any agonist activities. The constitutive production within cells in the lung and airway may provide a mechanism for the regulation of acute responses for cell populations in constant contact with external stimuli. In diseased lungs from patients with IPF, extremely high levels of IL-lra were identified both by ELISA and by immunohistochemical localization (221).Exogenous administration of IL-Ira in animal models of pulmonary inflammation has not provided the protection expected. However, when endogenous IL-lra was removed using neutralizing IL-Ira-specific antibodies, a tremendous increase in granulomatous inflammation was observed (222). These studies suggest that endogenous IL-lra levels may already be sufficiently high and only on removal of the protein coult its anti-inflammatory affects be observed. Furthermore, these studies may indicate that exogenous administration of IL-lra may not be an appropriate therapeutic. Perhaps the most attractive approach to modulation of inflammation in the lung and other organs is the use of suppressive cytokines such as IL10. IL-10, a cytokine synthesis inhibitory factor, has been assessed in many types of inflammation such as those outlined previously. Interestingly, IL10 has a high degree of homology (-80%) with an Epstein-Barr virusencoded protein, and its expression during vital infection likely modulates the host response (223).IL-10 is predominantly produced by macrophage populations, Th2-type lymphocytes, and B cells. Most notably, administration of IL-10 to mice protected them from a lethal endotoxin challenge (112). These early observations defined a significant role for IL-10 in the management of acute inflammatory responses. In addition, IL-10 has the ability to downregulate multiple inflammatoy cytokines, including TNFa,IL-1, IL-6, IFN-7, as well as adhesion molecule expression and the production of nitric oxide. The downregulation of these mediators by IL-

290

NICHOLAS W. LUKACS AND PETER A. WARD

10 therapy may demonstrate an extremely potent anti-inflammatory agent to be used for intervention of inflammation-induced injury. Interleukin-4, a product of T lymphocytes and mast cells, has also demonstrated regulatory effects on cytokine production primarily with monocytel macrophage function. LPS-stimulated monocytes cocultured with IL-4 exhibited diminished IL-1 and TNF-a gene expression (224) as well as decreased adhesion molecule expression (225). IL-4 also decreased the tumoricidal activity of alveolar macrophages, suggesting a potential role of IL-4 in downregulation of antitumor responses (226). However, the effects of IL-4 as a regulatory cytokine may activate properties that complicate therapy. For example, IL-4 has demonstrated an intimate role in progression of pulmonary granulomatous responses (164),in allergic airway eosinophilia (206, 208), and in proliferation and collagen gene expression in pulmonary fibroblast populations (227). Together, the use of IL-4 may have activating side effects that would preclude the use in pulmonary inflammation. V. Conclusions

Many mediator pathways appear to be involved in the generation of an inflammatory response leading to lung injury and appear to be commonly shared in a variety of pulmonary diseases. The activation of complement, production of inflammatory and chemotactic cytokines, upregulation of adhesion molecule expression, localization and activation of inflammatory leukocyte populations, and release of damaging proteases together contribute to injury induced during various pulmonary diseases. This complex cascade of events offers multiple points of therapeutic targets, which could be approached for the modulation of tissue-threatening responses. The treatment to be applied likely will depend on the type of inflammatory response, the time frame of disease progression, and the severity of the disease. Although common pathways of activation are utilized, a single therapeutic approach will probably not be sufficient for every type of pulmonary inflammation. In fact, the inhibition of a single factor may not be effective during any disease and a “cocktail” approach may be needed to effectively control vigorous inflammatory responses. The elucidation of mechanisms, through use of clinical and animal model research, during pulmonary inflammation is imperative for a better understanhng of disease processes and development of appropriate and effective therapeutic approaches. REFERENCES

1. Ward, P. A., Warren, J. S., and Johnson, K. (1988). Oxygen radicals, inflammation, and tissue injury. Free Rad. B i d Med. 5(5-6), 403-408.

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

291

2. Luscinskas, F. W., Cybuisky, M. I., Kiely, J. M., Peckins, C. S., Davis, V. M., and Gimbrone, M. A,, Jr. (1991).Cytohne-activated human endothelial monolayers support enhanced neutrophil transmigration via a mechanism involving both endothelialleukocyte adhesion molecule-1 and intracellular adhesion molecule 1.J. Zmmunol. 146, 1617. 3. Bevilacqua, M. P., Pober, J. S., Wheeler, M. E., Cotran, R. S., and Gimbrone, M. A,, Jr. (1985). Interleukin 1 acts on cultured human vascular endothelium to increase the adhesion ofpolymorphonuclear leukocytes, monocytes, and related leukocyte cell lines. J. Clin. Intiest. 76, 2003. 4. Dustin, M. L., Rothlein, R., Bhan, A. K., Dinarello, C. A,, and Springer, T. A. (1986). Induction by IL-1 and interferon-gamma: Tissue distribution, biochemistry, and function of a natural adherence molecule (ICAM-1).J. Immunol. 137, 245. 5. Jutila, M. A. (1994). Selectins in leukocyte extravasation:Function of a common epitope on L- and E-selectin. Adti. Phannacol. 25, 235-262. 6. Zimmerman, C . A,, Prescott, S. M., and McIntyre, T. M. (1992). Endothelial cell interactions with granulocytes: Tethering and signaling molecules. Zminunol. Today 13, 93. 7. Imhof, B. A,, and Dunon, D. (1995).Leukocyte migration and adhesion. Adti. Immunol. 58, 345-416. 8. H o g , J. C., and Doerschuk, C. M. (1995). Leukocyte traffic in the lung. Annu. Reti. Physiol. 57, 97-114. 9. Mulligan, M. S., Watson, S. R., Fennie, C., and Ward, P. A. (1993).Protective effects of selectin chimeras in neutrophil-mediated lung injury.J. Immunol. 151(ll),6410-6417. 10. Mulligan, M. S., Padson, J. C., De Frees, S., Zheng, Z. L., b w e , J. B., and Ward, P. A. (1993). Protective effects of oligosaccharidesin P-selectin-dependent lung injury. Nature 364, 149-151. 11. Hansel, T. T., Braunstein, J. B., Walker, C., Blaser, K., Bruijnzeel, P. L., Virchow, J. C., Jr., and Virchow, C., Sr. (1991). Spuhim eosinophils from asthmatics express ICAM-1 and HLA-DR. C h . Exp. Zmmunol. 86, 271-277. 12. Hakkert, B. C., Kuijpers, T. W., Leeuwenberg, J. F., van Mourik, J. A., and Roos, D. (1991). Neutrophil and monocyte adherence to and migration across monolayers of cytokine-activated endothelial cell: The contribution of CD18, ELAM-1, and VLA-4. Blood 78, 2721-2726. 13. Carlos, T., Kovach, N., Schwartz, B., Rosa, M., Newman, B., Wayner, E., Benjamin, C., Osborn, L., Lobb, R., and Harlan, J. (1991). Human monocytes bind to two cytokineinduced adhesive ligands on cultured human endothelial cells: Endothelid-leukocyte adhesion molecule-1 and vascular cell adhesion molecule-I. Blood 77, 2266-2271. 14. Horejsi, V. (1991). Surface antigens of human leukocytes. Ado. Zmmunol. 49,75-147. 15. H o g , N. (1992). Leukocyte integrin activation. C . R. Acad. Sci., Ser. 3 314(9 Suppl.), 39-43. 16. Nakajima, H., Sano, H., Nishimura, T., Yoshida, S., and Iwamoto, I. (1994). Role of vascular cell adhesion molecule l/very late activation antigen 4 and intracellular adhesion moleculeAymphocytefunction-associated antigen 1interaction in antigen-induced eosinophil and T cell recruitment into tissue. J. Exp. Med. 179, 1145-1154. 17. Springer, T. A,, Dustin, M. L., Kishimoto, T. K., and Martin, S. D. (1987). The lymphocyte function-associated LFA-1, CD2, and LFA-3 molecules: Cell adhesion receptors of the immune system. Annu. Reti. Zmnzunol. 5, 223. 18. Springer, T. A. (1990). Adhesion receptors of the immune system. Nature 346, 425-434.

292

NICHOLAS W. LUKACS AND PETER A. WARD

19. Lukacs, N. W., Strieter, R. M., and Kunkel, S. L. (1993). Cytokines in acute inflammation. Curt-. Opin. Hemutol. 1993, 26-31. 20. Strieter, R. M., Kunkel, S. L., and Bone, R. C. (1993). Role of tumor necrosis factoralpha in disease states and inflammation. Crit. Care Med. 21(10 Suppl.), S4474463. 21. Kunkel, S. L., Remick, D. G., Strieter, R. M., and Larrick, J. W. (1989). Mechanisms that regulate the production and effects of tumor necrosis factor-alpha. Crit. Reu. Immunol. 9(2), 93-117. 22. Oppenheim, J. J., Zachariae, C. 0. C., Mukaida, N., and Matsushima, K. (1991). Properties of the novel proinflammatory “intercrine” cytoldne family. Annu. Rev. Immunol. 9, 617. 23. Mosmann, T. R. (1994). Properties and functions of interleukin-10. Adu. Immunol. 56, 1-26. 24. Mosman, T. R., and Moore, K. W. (1989). The role of IL-10 in crossregulation of Thl and Th2 response. Immunol. Today 12, A49-A58. 25. Keegan, A. D., Nelms, K., Wang, L. M., Pierce, J. H., and Paul, W. E. (1994). Interleukin 4 receptor: Signaling mechanisms. Immunol. Today 15(9), 423-432. 26. Zurawsld, G., and de Vries, J. E. (1994). Interleukin 13, an interleukin 4-like cytokine that acts on monocytesand B cells, but not on T cells. Immunol. Today 15(l), 19-26. 27. Mosmann, T. R., Chenvinski, H., Bond, M. W., Giedlin, M. A,, and Coffman, R. L. (1986). Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins. J. Immunol. 136, 2348-2359. 28. Johnston, R. B., Jr. (1993).The complement system in host defense and inflammation: The cutting edges of a double edged sword. Pediatr. Infect. Dis. J. 12(11),933-941. 29. Ward, P. A. (1972). Biological activities of the complement system. Ann. Allergy 30(6), 307-314. 30. Foreman, K. E., Vaporciyan, A. A,, Bonish, B. K., Jones, M. L., Johnson, K. J., Glovsky, M. M., Eddy, S. M., and Ward, P. A. (1994). C5a-induced expression of P-selectin in endothelial cells. J. Clin. Inoest. 94, 1147-1155. 31. Meliconi, R., Senaldi, G., Sturani, C., Gdavotti, V., Facchini, A,, Gasbarrini, G., and Vergani, D. (1990). Complement activation products in idiopathic pulmonary fibrosis: Relevance of fragment Ba to disease severity. Clin Immunol. Immunopathol. 57,64-73. 32. Zilow, G., Joka, T., Obertacke, U., Rother, U., and Kirschfink, M. (1992). Generation of anaphylatoxin C3a in plasma and bronchoalveolar lavage fluid in trauma patients at risk for the adult respiratory distress syndrome. Crit. Care Med. 20, 468-473. 33. Miller, R. D., Kueppers, F., and Offerd, K. P. (1980). Serum concentrations of C3 and C4 of the complement system in patients with chronic obstructive pulmonary disease. J. Lab. Clin. Med. 95, 266-271. 34. Strunk, R. C., Eidlen, D. M., and Mason, R. J. (1988).Pulmonary alveolar type I1 cells synthesize and secrete proteins of the classical and alternative complement pathways. J. Clin. lnuest. 81, 1419-1426. 35. Christian-Ritter, K. K., Hill, L. D., Hoie, E. B., and Zach, T. L. (1994). Effect of interleukin-4 on the synthesis of the third component of complement by pulmonary epithelial cells. Am. J. Puthol. 144, 171-176. 36. Barnes, P. J . (1990). Reactive oxygen species and airway inflammation. Free Rad. Biol. Med. 9, 235-243. 37. Sibille, Y., and Reynolds, H. Y. (1990). Macrophages and polymorphonuclear cells in lung defense and injury. Am. Reu. Respir. Dis. 141, 471-501. 38. Repine, J. E. (1992). Scientific perspectives on adult respiratory distress syndrome. Lancet 339,466-472.

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

293

39. Klausner, J. M., Paterson, I. S., Lester, K., Valeri, C. R., Shepro, D., and Hechtman, H. B. (1989). Surgery 105, 192-199. 40. Gurtner, G. H., Farrukh, I. S., Adkinson, N. F., Sciuto, A. M., Jacobson, J. M., and Michael, J. R. (1987). The role of arachidonate mediators in perioxide-induced lung injury. Am. Rev. Respir. Dis. 136, 480-483. 41. Olafsdottir, K., Ryrfeldt, A., Atzori, L., Berggren, M., and Moldeus, P. (1991). Hydrogenperoxide-induced broncho- and vasoconstriction in the isolated rat lung. Exp. Lung Res. 17, 615-627. 42. Metinko, A. P., Kunkel, S. L., Standiford, T. J.. and Strieter, R. M. (1992). Anoxiahyperoxia induces monocyte-derived interleukin-8. J . Clin. Invest. 90,791-798. 43. O’Brien-Ladner, A. R., Nelson, M. E., Cowley, B. D., Jr., Bailey, K., and Wesselius, L. J. (1995). Hyperoxia amplifies TNF-alpha production in LPS-stimulated human alveolar macrophages. Am. J. Respir. Cell. Mol. Biol. 12(3), 275-279. 44. Gaston, B., Drazen, J. M., Loscalzo, J., and Stamler, J. S. (1994). The biology of nitrogen oxides in the airways. Am. j . Respir. Crit. Care Med. 149, 538-551. 45. Palmer, R. M. J., Femge, A. G., and Moncada, S. (1987). Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 327,524-526. 46. Hogman, M., Frostell, C. G., Hedenstrom, H., and Hedenstiema, G. (1993). Inhalation of nitric oxide modulates adult human bronchial tone. Am. Rev. Respir. Dis. 148,14741478. 47. Hubbard, R. C., and Crystal, R. G. (1986). Antiproteases and antioxidants: Strategies for the pharmacologic prevention of lung destruction. Respiration 50, 56-73. 48. Gadek, J. E., Fells, G. A.. Wright, D. G., and Crystal, R. G. (1980). Human neutrophil elastase functions as atype I11 collagen “collagenase.”Biochem. Biuphys. Res. Commun. 95, 1815-1822. 49. Janoff, A. (1985). Elastases and emphysema: Current assessment of the protease hypothesis. Am. Rev. Respir. Dis. 132, 417-433. 50. Kueppers, F., and Black, L. F. (1974).Alpha-1-antitrypsinand its deficiency.Am. Rev. Respir. Dis. 110, 176-194. 51. Reilly, C. F., and Travis,J. (1980).The degradation of human lung elastin by neutrophil proteinases. Biochem. Biophys. A d a 621, 147-157. 52. Hoogsteden, H. C., van Hal, P. T., Wiljkhulis, J. M., Hop, W., and Hilvering, C. (1992). Expression of the CDlUCD18 cell surface adhesion glycoprotein family and MHC class I1 antigen on blood monocytes and alveolar macrophages in interstitial lung diseases. Lung 170, 221-224. 53. Melis, M., Gjomarkaj, M., Pace, E., Malizia, G., and Spatafora, M. (1991). Increased expression of leukocyte function associated antigen-1 (LFA-1) and intercellular adhesion molecule-I (ICAM-1) by alveolar macrophagesofpatients with pulmonarysarcoidosis. Chest 100, 910-916. 54. Donnelly, S. C., Haslett, C., Dransfield, I., Robertson, C. E., Carter, D. C., Ross, J. A,, Grant, I. S., and Tedder, T. F. (1994). Role of selectins in development of adult respiratory distress syndrome. Lancet 344, 215-219. 55. Shijubo, N., Imai, K., Aoki, S., Hirasawa, M., Sugawara, H., Koba, H., Tsujisaki, M., Sugiyama, T., Hinoda, Y., Yachi, A., et al. (1992). Circulating intercellular adhesion molecule-1 (ICAM-1) antigen in sera of patients with idiopathic pulmonary fibrosis. Clin. Exp. Immunol. 89, 58-62. 56. Newman, W., Beall, L. D., Carson, C. W., Hunder, G. G., Graben, N., Randhawa, Z. I., Gopal, T. V., Wiener-Kronish, J., and Matthay, M. A. (1993). Soluble E-selectin is found in supernatants of activated endothelid cells and is elevated in the serum of patients with septic shock. J. Zmmunol 150, 644-654.

294

NICHOLAS W. LUKACS AND PETER A. WARD

57. Beutler, B. (1990). The tumor necrosis factors: Cachectin and lymphotodn. Hosp. Pract. 2, 45-56. 58. Standiford, T. J., Kunkel, S. L., Basha, M. A,, Chensue, S. W., Lynch, J. P., Toews, G. B., and Stieter, R. M. (1990).Interleukin-8gene expression byapulmonary epithelial cell line: A model for cytokine networks in the lung. J. C h Invest. 86, 1945-1953. 59. Fabisiak, J. P., and Kelley, J. (1993). Platelet-derived growth factor. In “Cytokines of the Lung” ( J . Kelley, Ed.), pp. 3-26. Dekker, New York. 60. Pledger, W. J., Stiles, C. D., Andoiades, H. N., and Scher, D. (1977). An ordered sequence of events is required before BALB/C-3T3 cells become committed to DNA synthesis. Proc. Natl. Acad. Sci. USA 74,4481-4490. 61. Roberts, A. B., and Sporn, M. B. (1990).The transforminggrowth factor-/3s. In “Peptide Growth Factors and Their Receptors. Handbook of Experimental Pharmacology” (M. B. Sporn and A. B. Roberts, Ed.), pp. 419-473. Vol. 95, Springer-Verlag, Heidelberg. 62. Baird, A,, and Durkin, T. (1986). Inhibition of endothelid cell proliferation by type ptransforming growth factor: Interaction with acidic and basic fibroblast growth factors. Biochem. Biophys. Res. Commun. 138, 476-482. 63. Muller, G . , Behrens, J., Nussbaumer, U., Bohlen, P., and Bichmeier, W. (1987). Inhibitory action of transforming growth factor b on endothelial cell Proc. Natl. Acad. Sci. USA 84,5600-5604. 64. Van Snick, J. (1990). Interleukin-6: An overview. Annu. Rev. Zmmunol. 8, 253-278. 65. Van Damme, J., Cayphas, S., Opdennakker, G., Billiau, A., and Van Snick, J. (1987). Interleukin-1 and poly(rl):poly(rC) induce production of a hybridoma growth factor by human fibroblasts. Eur. J. Zmmunol. 17, 1-7. 66. Van Damme, J . , Cayphas, S., Van Snick, J., Coning, R., Put, W., Lenaerts, J. P., Simpson, R. J., and Billiar, A. (1987). Purification and characterization of human fibroblast-derivedhybridoma growth factor identical to T-cell derived B-cell stimulatory factor-2 (interleukin-6). Eur. J. Biochem. 168, 543-550. 67. Gauldie, J., Richards, C., Hamish, D., Landsdorp, P., and Baumann, H. (1987). Interferon-beta2/BSF-2 shared identity with monocyte-derived hepatocyte stimulating factor (HSF) and regulates the major acute phase protein responses in liver cells. Proc. Natl. Acad. Sci. USA 84, 7251-7255. 68. Uyttenhove, C., Coulie, P. G., andvan Snick, J. (1988).Tcell growth and differentiation induced by interleukin HPlAL-6, the murine hybridomdplasmacytoma growth factor by human fibroblasts.J. Exp. Med. 167, 1417-1427. 69. Okada, M., Kitahara, M., Kishimoto, S., Matsuda, T., Hirano, T., and Kishimoto, T. (1988). IL-6IBSF-2 functions as a killer helper factor in the in vitro production of cytotoxic T lymphocytes. J. Immunol. 141, 1543-1549. 70. Kasahara,Y . , Miyawaki,T., Kato, K., Kanegane, H., Yachie, A., Yodoi, T., and Taniguchi, N. (1990). Role of interleukin 6 for differential responsiveness of naive and memory CD4+ T cells in CD2 mediated activation.1.Exp. Med. 172, 1419-1424. 71. Helle, M., Brakenhoff,P. J., De Groot, R., and Aarden, L. A. (1989). Interleukin-6 is involved in interleukin 1-induced activities. Ezw. J. Zmmunol. 18, 957-959. 72. Jordan, S . C., Kondo, T., Prehn, J., Marchevsloj, A., and Waters, P. (1991). Cytokine gene activation in rat lung allografts: Analysis by Northern blotting. Transplant. Proc. 23, 604-606. 73. Schall, T. J . (1991).Biology of the RANTES/SIS cytokine family. Cytokine 3,165-183. 74. Kameyoshi, Y., Dorschner, A., Mallet, A. I., Christophers, E., and Schroder, J. M. (1993).Cytokine RANTES released by thrombin-stimulated platelets is a potent attractant for human eosinophils. J. Exp. Med. 176, 587-592.

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

295

75. Lukacs, N. W., Strieter, R. M., Elner, V. M., Evanoff, H. L., Burdick, M. D., and Kunkel, S. L. (1994). Intrace1lular adhesion molecule-1 mediates the expression of monocyte-derived MIP-la during rnonocyte-endothelial cell interactions, Blood 83, 1174. 76. Lukacs, N. W., Strieter, R. M., Elner, V., Evanoff, H. L., and Kunkel, S. L. (1995). Monocyte-endothelial cell interactions induce chemokine production, IL-8 and MCP1. Blood. 86, 2767-2773. 77. McElvaney, N. G., Nakamura, H., Birrer, P., Hebert, C. A,, Wong, W. L., Aphonso, M., Baker, J. B., Catalano, M. A,, and Crystal, R. G. (1992). Modulation of airway inflammation in cystic fibrosis. In vivo suppression of interleukin 8 levels on the respiratory epithelial surface by aerosolization of recombinant secretory leukoprotease inhibitor. J Clin. Invest. 90, 1296-1301. 78. Car, B. C., Meloni, F., Luisetti, M., Semenzato, G., Gialdroni-Grassi, G., and Walz, A. (1994). Elevated IL-8 and MCP-l in the bronchoalveolar lavage fluid of patients with idiopathic pulmonary fibrosis and pulmonary sarcoidosis. Elevated IL-8 and MCP1 in the bronchoalveolar lavage fluid of patients with idiopathic pulmonary fibrosis and pulmonary sarcoidosis. Am. /. Respir. Crit. Care Med. 149, 655-659. 79. Ischweizer, R. C., Welmers, B. A. C., Raaijmakers, J. A. M., Zanen, P., Lammers, J. W. J., and Koenderman, L. (1994). RANTES- and interleukin-8-induced responses in normal human eosinophils:Effects ofpriming with interleukin-5. Blood 83,3697-3704. 80. Standiford, T. J., Rolfe, M. E., Kunkel, S. L., Lynch, J. P., Burdick, M. D., Gilbert, A. R., Orringer, M. B., Whyte, R. I., and Strieter, R. M. (1993). Macrophage inflammatory protein-1 alpha expression in interstitial lung disease.]. Immunol. 151,2852-2863. 81. Ghosh, S., Latimer, R. D., Gray, B. M., Hanvood, R. J., and Oduro, A. (1993). Endotoxin-induced organ injury. Crit. Care Med. 21, S19-S28. 82. Leon, C., Rodrigo, M. J., Tomasa, A,, Gallart, M. T., Latorre, F. J., Rius, J., and Brugues, J. (1982). Complement activation in septic shock due to gram-negative and gram-positive bacteria. Crit. Care Med. 10, 308-310. 83. Bergh, K., Olsen, P. O., Hdgunset, J . , and Iversen, 0.J. (1991).Complement activation and pulmonary dysfunction in experimental Escherichia coli septisemia. Acta Anaestheiol. Scand. 35, 267-274. 84. Grisham, M. B., Everse, J., and Janssen, H. F. (1988). Endotoxemia and neutrophil activation in vivo. Am. J. Physial. 254, H1017-1022. 85. Taggart, D. P., Sundaram, S., McCartney, C., Bowman, A,, McIntyre, H., Courtney, J. M., and Wheatley, D. J. (1994). Endotoxemia, complement, and white blood cell activation in cardiac surgery: A randomized trial of laxatives and pulsatile perfusion. Ann Thorac. Surg. 57, 376-382. 86. Bone, R. C. (1992). Inhibitors of complement and neutrophils: A critical evaluation of their role in the treatment of sepsis. Crit. Care Med. 20, 891-898. 87. Witthaut, R., Farhood, A,, Smith, C. W., and Jaeschke, H. (1994). Complement and tumor necrosis factor-alpha contribute to Mac-1 (CD1lbKD18) upregulation and systemic neutrophil activation during endotoxemai in vivo. 1. Leukocyte Biol. 55, 105-111. 88. Trautinger, F., Hammerle, A. F., Poschi, G., and Micksche, M. (1991). Respiratory burst capability of polymorphonuclear neutrophils and TNF-alpha serum levels in relationship to the development of septic syndrome in critically ill patients.]. Leukocyte Biol. 49, 449-454. 89. Carey, P. D., Jenkens, J. K., Byme, K., Walsh, C. J., Fowler, A. A., and Sugerman, H. J. (1992). The neutrophil respiratory burst and tissue injury in septic acute lung injury: The effect of cyclooxygenase inhibition in swine. Surgery 112, 45-55.

296

NICHOLAS W. LUKACS AND PETER A. WARD

90. Wizemann, T. M., and Laskin, D. L. (1994). Enhanced phagocytosis, chemotaxis, and production of oxygen intermediates by interstitial lung macrophages following acute endotoxemia. Am. /. Respir. Cell. Mol. B i d . 11, 358-365. 91. Engelberts, I., Samyo, S. K., Leeuwenberg, J. F., van der Linden, C. J., and Buurman, W. A. (1992). A role for ELAM-1 in the pathogenesis of MOF during septic shock. J . Surg. Res. 53, 136-144. 92. Walsh, C. J., Carey, P. D., Cook, D. J., Bechard, D. E., Fowler, A. A., and Sugerman, H. J. (1991). Anti-CD18 antibody attenuates neutropenia and alveolar capillarymembrane injury during gram-negative sepsis. Surgenj 110, 205-21 1. 93. Mileski,W. J., Winn, R. K., Harlan, J. M., and Rice, C . L. (1991).Transient inhibition of neutrophil adherence with the anti-CD18 monoclonal antibody 60.3 does not increase mortality rates in abdominal sepsis. Surgery 109, 497-501. 94. Ulich, T. R., Guo, K., Yin, S., del Catillo, J., Yi, E. S., Thompson, R. C., adn Eisenberg, S. P. (1992). Endotoxin-induced cytokine gene expression in vivo. IV. Expression of interleukin-1 alphaheta and interleukin 1receptor antagonist mRNA during endotoxemia and during endotoxin-initiated local acute inflammation.Am. J. Puthol. 14,61-68. 95. Remick, D. G., Strieter, R. M., Eskandari, M. K., Nguyen, D. T., Genord, M. A., Raiford, C . L., and Kunkel, S. L. (1990). Role of tumor necrosis factor-a in lipopolysaccharide-induced pathologic alterations. Am. J. Puthol. 136, 49-55. 96. Standiford, T. J., Kunkel, S. L., Lukacs, N. W., Greenberger, M. J., Danforth, J. M., Kunkel, R. G., and Strieter, R. M. (1995). Macrophage inflammatory protein-la mediates lung leukocyte recruitment, lung capillary leak, and early mortality in murine endotoxemia. J. Zmmunol. 155, 1515-1524. 97. Xing, Z., Jordana, M., Kirpalani, H., Driscoll, K. E., Schall, T. J., and cauldie, J. (1994).Cytokineexpression by neutrophils and macrophages in vivo: Endotoxin induces tumor necrosis factor-a, macrophage inflammatoryprotein-2, interleukin-lb, and interleukin-6 but not RANTES or transforming growth factor-bl mRNA expression in acute lung inflammation. Am. J. Respir. Cell. Mol. B i d . 10, 148-156. 98. Vanotteren, G. M., Strieter, R. M., Kunkel, S. L., Paine, R., Creenberger, M. J., Danforth, J. M., Burdick, M. D., and Standiford, T. J. (1995). Compartmentalized expression of RANTES in a murine model of endotoxemia. J. lmmunol. 154, 19001908. 99. Windsor, A. C., Walsh, C. J., Mullen, P. G., Cook, D. J., Fisher, B. J., Biocher, C. R., Leeper-Woodford, S. K., Sugerman, H. J., Fowler, A. A,, 111, and Powler, A. A. (1993). Tumor necrosis factor-alpha blockade prevents neutrophil CD18 receptor upregulation and attenuates acute lung injury in porcine sepsis without inhibition of neutrophil oxygen radical generation. J. Clin. Invest. 91, 1459-1468. 100. Van Zee, K. J., Kohno, T., Fischer, E., Rock, C. S., Moldawer, L. L., and Lowry, S. F. (1992). Tumor necrosis factor soluble receptors circulate during experimental and clinical inflammation and can protect against excessive tumor necrosis factor alpha in vitro and in vivo. Proc. Natl. Acad. Sci. USA 89, 4845-4849. 101. Mohler, K. M., Torrance, D. S., Smith, C. A,, Coodwin, R. G., Stremler, K. E., Fung, V. P., Madani, H., and Widmer, M. B. (1993). Soluble tumor necrosis factor (TNF) receptors are effective therapeutic agents in lethal endotoxemia and function simultaneously as both TNF carriers and TNF antagonists. J. ZmmunoZ. 151, 1548-1561. 102. Eskandari, M. K., Bolgos, G., Miller, C., Nguyen, D. T., DeForge, L. E., and Remick, D. C. (1992). Anti-tumor necrosis factor antibody therapy fails to prevent lethality after cecal ligation and puncture or endotoxemia. 1. Zmmunol. 148,2724-2730. 103. Chensue, S. W., Terebuh, P. D., Remick, D. C., Scales, W. E., and Kunkel, S. L. (1991). In vivo biologic and immunohistochemical analysis of interleukin-1 alpha, beta

MEDIATORS, CYTOKINES, AND ADHESION MOLECULES

297

and tumor necrosis factor during experimental endotoxemia. Kinetics, Kupffer cell expression, and glucocorticoid effects. Am. J. Pathol. 138(2), 395-402. 104. Ulich, T. R., Watson, L. R., Yin, S., Guo, K., Wang, P., Thang, H., and del Catillo, J. (1991).The intratracheal administration of endotoxin and cytokines I. Characterization of LPS-induced IL-1 and TNF mRNA expression and the LPS-, IL-1, and TNFinduced inflammatory infiltrate. Am. J. Puthol. 138, 521-524. 105. Okusawa, S., Gelfand, J. A., Ikejuma, T., Connolly, R. J., and Dinarello, C. A. (1988). Interleukin 1 induces a shock-like state in rabbits. Synergism with tumor necrosis factor and the effect of cyclooxygenase inhibition. J. CZin Inuest. 81, 1162-1172. 106. Eichacker, P. Q., Hoffman, W. D., Farese, A,, Banks, S. M., Kuo, G. C., MacVittie, T. J., and Natanson, C. (1991).TNF hut not IL-1 in dogs causes lethal lung injury and multiple organ dysfunction similar to human sepsis. J. AppZ. PhysioE. 71, 1979-1989. 107. Ohlsson, K., Bjork, P., Bergenfeldt, M., Hageman, R., and Thompson, R. C. (1990). Interleukin-1 receptor antagonist reduces mortality from endotoxin shock. Nature 348,550-552. 108. Standiford, T. J., Strieter, R. M., Lukacs, N. W., and Kunkel, S. L. (1995).Neutralization of IL-10 increases lethality in endotoxemia: Cooperative effects of macrophage inflammatory protein-2 and tumor necrosis factor. J. Immunol. 155, 2222-2229. 109. Shanley, T. P., Schmal, H., Friedl, H. P., Jones, M. L., and Ward, P. A. (1991). Role of macrophage inflammatory protein-1 alpha (MIP-1 alpha) in acute lung injury in rats. J . Immunol. 154(9),4793-4802. 110. Marchant, A., Deviere, J., Byl, B., De Groote, D., Vincent, J. L., and Goldman, M. (1994). Interleukin-10 production during septiceaemia. Lancet 343, 707-708. 111. Gomez-Jimenez, J., Martin, M. C., Sauri, R., Segura, R. M., Esteban, F., Ruiz, J. C., Nuvials, X., Boveda, J. L., Peracaula, R., and Salgado, A. (1995). Interleukin-10 and the monocyte/macrophage-induced inflammatory response in septic shock. 1.Infect. Dis. 171, 472-475. 112. Howard, M., Muchamuel, T., Andrade, S., and Menon, S. (1993). Interleukin 10 protects mice from lethal endotoxemia. 1.Exp. Med. 177, 1205-1208. 113. Rogy, M. A,, Auffenberg, T., Espat, N. J., Philip, R., Remick, D., Wollenberg, G. K., Copeland, E. M., and Moldawer, L. L. (1995). Human tumor necrosis factor receptor (p55)and interleukin 10 gen transfer in the mouse reduces mortality to lethal endotoxemia and also attenuates local inflammatory responses. J . Exp. Med. 181, 2289-2293. 114. Bruade, S., Nolop, K. B., Flemming, J. S., Krausz, T., Taylor, K. M., and Royston, D. (1986).Increased pulmonary transvascular protein flux after caninecardiopulmonary bypass. Am. Reu. Respir. Dis. 135, 463-481. 115. Martin, R. J., Sandblom, R. L., and Johnson, R. J. (1983). Adult respiratory distress syndrome following thromholytic therapy for pulmonary embolism. Chest 1, 151-153. 116. Sprung, C. L., Leowenherz, J. W., Baier, H., and Hauser, M. J. (1981). Evidence of increased permeability in reexpansion pulmonary edema. Am. /. Med. 71, 497-500. 117. Tomashefski, J. F., Jr., Davier, P., Boggin, C., Green, R., Zapol, W. M., and Reid, L. M. (1983).The pulmonary vascular lesion of the adult respiratory distress syndrome. Am. J. Pathol. 112, 112-116. 118. Packer, C. S., and Rhoades, R. A. (1992).Impaired pulmonaryvascular smooth muscle function in lung injury. In “Lung Vascular Injury: Molecular and Cellular Response” (A. Johnson and T. J. Ferro, Eds.), pp. 249-252. Dekker, New York. 119. Lindsay, T. F., Hill, J., Oritz, F., Rudolph, A., Valeri, C. R., Hechtman, H. B., and Moore, F. D., Jr. (1992).Blockade of complement activation prevents local and pulmonary albumin leak after lower torso ischemia-reperfusion. Ann. Surg. 216, 677-683.

298

NICHOLAS W. LUKACS AND PETER A. WARD

120. Hill, J., Lindsay, T. F., Oritz, F., Yeh, C. G., Hechtman, H. B., and Moore, F. D., Jr. (1992). Soluble complement receptor type 1 ameliorates the local and remote organ injury after intestinal ischemia-reperfusion in the rat. 1. Immunol. 149, 1723-1728. 121. Seekamp, A,, Mulligan, M. S., Till, G. O., and Ward, P. A. (1993). Requirements for nuetrophil products and L-arginine in ischemia-reperfusion injury. Am. /. Pathol. 142, 1217-1226. 122. Deeb, G. M., Grum, C. M., Lynch, M. J.. Guynn, T. P., Gallagher, K. P., Lijungman, A. G., Bolling, S. F., and Morganroth, M. L. (1990). Neutrophils are not necessary for induction of ischemia-reperfusion injury. J. Appl. Physiol. 68, 374-381. 123. Steimle, C. N., Guynn, T. P., Morganroth, M. L., Bolling, S. F., Carr, K., and Deeb, G. M. (1992). Neutrophils are not necessary for ischemia-reperfusion lung injury. Ann Thorac. Surg. 53, 64-72. 124. Sekido, N., Mukaida, N., Harada, A,, Nadanishi, I., and Watnabe, Y. (1993). Prevention of lung reperfusion injury in rabbits by a monoclonal antibody against interleukin-8. Nature 365, 654-657. 125. Detterbeck, F. C., Keagy, B. A,, Paul, D. E., and Wilcox, B. R. (1990). Oxygen free radical scavengers decrease reperfusion injury in lung transplantation. Ann. Thorac. Surg. 50, 204-210. 126. Jackson, R. M., Russell,W. J., andVeal, C. F. (1992).Endogenous catalase in reoxygenation lung injury. J . Appl. Physiol. 72, 858-864. 127. Fisher, P. W., Huang, Y. C., Kennedy, T. P.. and Piantadosi, C. A. (1993). P02dependent hydroxyl radical production during ischemia-reperfusion lung injury. Am. 1.Physiut. 265, 279-285. 128. Adkins, W. K., and Taylor, A. E. (1990). Role of xanthine oxidase and neutrophils in ischemia-reperfusion injury in rabbit lung. J . Appl. Physiol. 69, 2012-2018. 129. Terada, L. S., Dormish, J. J., Shanley, P. F., Leff, J. A,, Anderson, B. O., and Repine, J. E. (1992).Circulating xanthine oxidase mediates lung neutrophil sequestration after intestinal ischemia-reperfusion. Am. J . Physiol. 263, L394-UOl. 130. Zamora. C. A.. Baron, D. A.. and Heffner. I. E. (1993). Thromboxane contributes to pulmonary hpertension in ischemia-reperfusion lung injury. 1. Appl. Physiol. 74, 224-229. 131. Liungman, A. G., Grum, C. M., Deeb, G. M., Bolling, S. F., and Morganroth, M. L. (1991). Inhibition of cyclooxygenase metabolite production attenuates ischemiareperfusion lung injury. Am. Reu. Respir. Dis. 143, 610-617. 132. Abdih, H., Kelly, C. J., Bouchier-Hayes, D., William, R., Watson, G.. Redmond, H. P., Burke, P., and Bouchier-Hayes, D. J. (1994). Nitric oxide (endothelium-derived relaxing factor) attenuates revascularization-induced lung injury. J. Surg. Res. 57, 39-43. 133. Seekamp, A., Till, G. O., Mulligan, M. S., Paulson, J. C., Anderson, D. C., Miyasaka, M., and Ward, P. A. (1994).Role of selectins in local and remote tissue injury following ischemia-reperfusion. Am. J. Pathol. 144, 592-598. 134. Steinberg, J. B., Mao, H. Z., Niles, S. D., Jutila, M. A., and Kapelanski, D. P. (1994). Survival in lung reperfusion injury is improved by an antibody that binds and inhibits L- and E-selectin. J. Heart Lung Transplant. 13, 306-318. 135. Horgan, M. J., Wright, S. D., and Malik, A. B. (1990). Antibody against leukocyte integrin (CD18) prevents reperfusion-induced lung vascular injury. Am. J. Physiol. 259, L315-L319. 136. Welboum, C. R., Goldman, G., Paterson, I. S., Valeri, C. R., Shepro, D., and Hechtman, H. B. (1991). Neutrophil elastase and oxygen radicals: Synergism in lung injury after hindlimb ischemia. Am. 1.Physiol. 260, H1852-H1856.

MEDIATORS, CITOKINES, AND ADHESION MOLECULES

299

137. Hill, J., Lindsay. T.. Valeri, C. R.. Shepro, D., and Hechtman, H. B. (1993). A CD18 antibody prevents lung injury but not hypotension after intestinal ischemiareperfusion. ]. Appl. Physiol. 74, 659-664. 1361. Caty, M. G., Guice, K. S., Oldham, K. T., Remick, D. G., and Kunkel, S. L. (1990). Evidence for tumor necrosis factor-induced pulmonary microvascular injury after intestinal ischeniia-reperfusion injury. Ann. Surg. 212, 694-700. 139. Seekamp, A., Warren, J. S., Remick, D. G., Till, G. O., and Ward, P. A. (1993). Requirements for tumor necrosis factor-alpha and interleukin-1 in limb ischemiareperfusion injury associated lung injury. Am. J. Puthol. 143, 453-463. 140. Shasby, D. M., Schwarz, M. I., Forstot, J. Z., Theofilopoulos, A. N., and Kassan, S. S. (1982). Pulmonary immune complex deposition in Wegner’s granulomatous. Chest 81, 338-340. 141. Johnson, K. J.. Chapman, W. E., and Ward, P. A. (1979). Inimunopathology of the lung: A review. Ani. 1.Pathol. 95, 795-844. 142. Johnson, K. J., Ward, P. A,, Kunkel, R. G., and Wilson, B. S. (1986). Mediation of IgA induced lung injury in the rat. Role of macrophages and reactive oxygen products. Lab. Invest. 54, 499-506. 143. Johnson, K. J., and Ward, P. A. (1981).Role of oxygen metabolites in immune complex injury of the lung. I. Immunol. 126, 2365-2369. 144. Fligiel, S . E., Ward, P. A., Johnson, K. J., and Till, G. 0. (1984). Evidence for a role of hydroxyl radical in immune-complex-induced vaculitis.Am. 1.Puthol. 115,375-382. 145. Mulligan, M. S., Hevel, J. M., Marletta, M. A,, and Ward, P. A. (1991). Tissue injury caused by deposition of immune complexes is L-arginine dependent. Proc. Nutl. Acad. Sci. USA 88, 6338-6342. 146. Mulligan, M. S . , Yeh, C. G., Rudolph, A. R., and Ward, P. A. (1992).Protective effects of soluble CR1 in complement- and neutrophil-mediated tissue injury, 1. Immunol. 148, 1479-1485. 147. Mulligan, M . S., and Ward, P. A. (1992). Immune coniplex-induced lung and dermal vascualr injury. Differing requirements for tumor necrosis factor-alpha and IL-1. J. Zmmunol. 149,331-339. 148. Mulligan, M. S., Warren, J. S., Smith, C. W., Anderson, D. C., Yeh, C. G., Rudolph, A. R., and Ward, P. A. (1992).Lung injury after deposition of IgA immune complexes. Requirements for CD18 and L-arginine. J . Zmmunol. 148, 3086-3092. 149. Mulligan, M. S., Jones, M. L., Vaporciyan, A. A,, Howard, M . C., and Ward, P. A. (1993). Protective effects of IL-4 and IL-10 against immune complex-induced lung injury. I. Immunol. 15, 5666-5674. 150. Boros, D. L. (1989). Immunopathology of Schistosom mnsoni infection. Clin. Microbiol. Rev. 2, 250-269. 151. Kunkel, S. L., Strieter, R. M., Lukacs, N . W., and Chensue, S. W. (1994). Molecular aspects of granuioinatous inflammation.1. Immunol. Immunophumcol. 14, 71-77. 152. Chensue, S. W., Quinlan, L., Higashi, G. I., and Kunkel, S. L. (1984). Role of oxygen reactive species in Schistosoma mansoni egg-induced granulomatous inflammation. Biochem. Biophys. Res. Commun. 122, 184-190. 153. Kindler, V., Sappino, A. P., Grau, G. E., P i p e t , P. F., and Vassalli, P. (1989). The inducing role of tumor necrosis factor in the development of bactericidal granulomas during BCG infection. Cell 56, 731-740. 154. Joseph, A. L., and Boros, D. L. (1993).TNF plays a role in Schistosoma mansoni egginduced granuloniatous inflammation.J. Zmmunol. 151, 5461-5468. 155. Tumang, M. C., Keogh, C., Moldawer, L. L., Helfgott, D. C., Teitelbaum, R., Hariprashad, J., and Murray, H. W. (1994). Role and effect of TNF-alpha in experimental visceral leishnianiasis.J. Immunol. 153, 768-775.

300

NICHOLAS W. LUKACS AND PETER A. WARD

156. Amiri, P., Locksley, R. M., Parslow, T. G., Sadick, M., Rector, E., Ritter, D., and McKerrow, J. H. (1992). Tumor necrosis factor-a restores granulomas and induces parasite egg laying in schistosome-infected SCID mice. Nature 356, 604-610. 157. Lukacs, N . W., Chensue, S. W., Strieter, R. M., Warmington, K., and Kunkel, S. L. (1994).Inflammatorygranuloma formation is mediated by TNF-a-inducible intracellular adhesion molecule-1. I. Zmmunol. 152, 5883-5889. 158. Rolfe, M. W., Standiford, T. J., Kunkel, S. L., Burdick, M. D., Gilbert, A. R., Lynch, J. P., and Strieter, R. M. (1993).Interleukin-1 receptor antagonist expression in sarcoidosis. Am. Rev. Respir. Dis. 148, 1378-1384. 159. Chensue, S. W., Otterness, I. G., Higashi, G. I., Forsh, C. S., and Kunkel, S. L. (1992). Monokine production by hypersensitivity (Schistosoma mansoni egg) and foreign body (Sephadex bead)-tpe granuloma macrophages: Evidence for sequential production of IL-1 and tumor necrosis factor. J. Zmmunol. 142, 1281-1287. 160. Curry, A. J., and Kaye, P. M. (1992). Recombinant interleukin 1 alpha augments granuloma formation and cytokine production but not parasite clearence in mice infected with Leishmania donavani. Infect Zmmun. 60,4422-4426. 161. Chensue, S. W., Warmington, K. S., Berger, A. E., and Tracey, D. E. (1992). Immunohistochemical demonstration of interleukin-1 receptor antagonist protein and interleukin-1 in human lymphoid tissue and granulomas. Am. /. Pathol. 140(2), 269-275. 162. Chensue, S. W., Bienkowski, M., Eessalu, T. E., Warmington, K. S., Hersey, S. D., Lukacs, N. W., and Kunkel, S. L. Endogenous interleukin 1receptor antagonist protein ( I M P ) regulates schistosome egg granuloma formation and the regional lymphoid response. 1.Zmmunol. 151, 3654-3662. 163. Huygen, K., Abramowicz, D., Vanderbussche, P., Jacobs, F., De Bruyn, J., Kentos, A., Drowart, A,, Van Vooren, J. P., and Goldman. M. (1992). Spleen cell cytokine secretion in Mycobacterium bovis SCG-infecter mice. Infect. Zmmun. 60,2880-2886. 164. Grzych, J. M., Pearce, E., Cheever, A., Caulada, Z. A., Caspar, P., Henry, S., Lewis, F., and Sher, A. (1991). Egg deposition is the majo stimulus for the production of Th2 cytokines in murine schistosomiasis mansoni. 1.lmmzmol. 146, 1322-1329. 165. Chakkalath, H. R., and Titus, R. G. (1994). Leishmania major-parasitized macrophages augment The-type T cell activation. I. Zmmunol. 153, 4378-4387. 166. Jones, M. L., and Warren, J. S. (1992). Monocyte chemoattractant protein 1 in a rat model of pulmonary granulomatosis. Lab. Invest. 66, 498-503. 167. Flory, C. M., Jones, M. L., and Warren, J. S. (1993). Pulmonary granuloma formation in the rat is partially dependent on monocyte chemoattractant protein 1. Lab. Znoest. 69,396-404. 168. Flory, C. M., Jones, M. L., Miller, B. F., and Warren, J. S. (1995). Regulatory roles of tumor necrosis factor-alpha and interleukin-1 beta in monocyte chemoattractant protein-1-mediated pulmonary granuloma formation in the rat. Am. 1. Puthol. 146, 450-462. 169. Lukacs, N. W., Kunkel, S. L., Strieter, R. M., Warmington, K., and Chensue, S. W. (1993).The role of macrophage inflammatory protein-1 alpha in Schistosomu munsoni egg-induced granulomatous inflammation. J. Exp. Med. 177, 1551-1559. 170. Chensue, S. W., Warmington, K. S., Lukacs, N. W., Lincoln, P. M., Burdick, M. D., Strieter, R. M., and Kunkel, S. L. (1995). Monocyte chemotactic protein expression during schistosome egg granuloma formation: Sequence of production, localization, contribution, and regulation. Am. 1.Puthol. 146, 130-138. 171. Lukacs, N. W., Kunkel, S. L., Allen, R., Evanoff, H. L., Shaklee, C. L., Sherman, J. S., Burdick, M. D., and Strieter, R. M. (1995). Stimulus and cell-specificexpression

MEDIATORS, ClTOKINES. AND ADHESION MOLECULES

301

of C-X-C and C-C chemokines by pulmonary stromal cell populations. Am. 1.Physiol. 268, L856-L861. 172. Kay, A. B. (1988). The eosinophil. In “Pulmonary Diseases and Disorders” (A. P. Fishman, Ed.), 2nd ed, pp. 599-606. McGraw-Hill, New York. 173. Kay, A. B., and Comgan, C. J. (1992). Asthma, eosinophils, and neutrophils. Br. Med. Bull. 48, 51-64. 174. Gleich, G. J. (1986).The eosinophilicleukocyte: Structure and function. Adv. lnamunof. 39, 177-253. 175. Djukanovic, R., Poche, W. R., Wilson, J. W., Beasley, C. R. W., Twentyman, 0. P., ~ in asthma. Am. Reu. Howarth, P. H., and Holgate, S. T. (1990). M L I C O Sinflammation Respir. Dis. 142, 434-457. 176. Corrigan, C. J.. and Kay, A. B. (1992).Asthma. Role of T-lymphocytesand lymphokines. Br. Med. Bull. 48, 72-84. 177. Corrigan, C. J., and Kay, A. B. (1992). T cells and eosinophils in the pathogenesis of asthma. Zmmunol. Today 13, 501-505. 178. Def Prete, G. (1992). Human Thl and Th2 lymphocytes:Their role in the pathophysiology of atopy. Allergy 47, 450-455. 179. Gratziou, C., Carroll, M., Walls, A., Howarth, P. H., and Holgate, S. T. (1992). Early changes in T lymphocytes recovered by bronchoalveolar lavage after local allergen challenge of asthmatic ainvays. Am. Reu. Respir. Dis. 145, 1259-1264. 180. Kay, A. B. (1991). T lymphocytes and their products inatopic allergy and asthma. Znt. Arch. Allergy Appl. Imniunol. 94, 189-193. 181. Robinson, D. S., Hamid, Q., Ying, S., Tsicopoulos, A,, Barkans, J., Bentley, A. M., Corrigan, C., Durham, S. R., and Kay, A. B. (1992). Predominant TH2-like bronchoalveolar T-lympocyte population in atopic asthma. N . Engl. 1. Med. 326, 298-304. 182. Romagnani, S., Ma@, E., Parronchi, P., Macchia, D., Piccinni, M. P., and Ricci, M. (1991). Increased numbers of Th2-like CD4+ T cells in target organs and in the allergen-specificrepertoire of allergic patients. Possible role of IL-4 produced by nonT cells. lnt. Arch. Allergy Appl. Immunol. 94, 133-136. 183. Walker, C., Virchow, J. C., Jr., Bruijnzeel, P. L., and Blaser, K. (1991). T cells and asthma. 11. Regulation of eosinophilia of asthma by T cell cytohnes. Znt. Arch. Allergy Appl. Immunol. 94, 248-250. 184. Zhang, X., Polla, B., Hauser, C., and Zubler, R. H. (1992).T cells from atopic individuals produce IgE-inducing activity incompletely blocked by anti-interleukin-4 antibody. Eur. 1. Immunol. 22, 829-833. 185. Spry, C. J., Kay, A. B., and Gleich, G. J. (1992). Eosinophils. lmmunol. Today 13, 384-387. 186. Gavett, S. H., Chen, X., Finkleman, F., and Wills-Karp, M. (1994). Depletion of murine CD4+ T lymphocytes prevents antigen-induced airway hyperreactivity and pulmonary eosinophilia. Am. J. Respir. Cell. Mol. B i d . 10, 587-593. 187. Iijima, H., Ishii, M., Yamauchi, K., Chao, C. L., Kimura, K., Shimura, S., Shindoh, Y., Inoue. H., Mue, S., andTakishima, T. (1987).Bronchoalveolarlavage and histologic characterization of late asthmatic response in guinea pigs. Am. Reu. Respir. Dis. 136,922-929. 188. Kasama, T., Strieter, R. M., Lukacs, N. W., Burdick, M. D., and Kunkef, S. L. (1994). Regulation of neutrophil-derived chemokine expression by IL-10.1. Immunol. 152, 3559-3569. 189. Kay, A. B. (1988). The eosinophil. In “Pulmonary Diseases and Disorders” (A. P. Fishman, Ed.), 2nd ed., pp. 599-606. McGraw-Hill, New York.

302

NICHOLAS W. LUKACS AND PETER A. WARD

190. Pritchard, D. I., Eady, R. P., Harper, S. T., Jackson, D. M., Orr, T. S., Richards, I. M., Trig, S., andwells, E. (1983).Laboratory infection ofprimateswith Ascaris suum to provide a model of allergic bronchoconstriction. Clin. Exp. Immunol. 54,469-476. 191. Sasaki, H., Yanai, M., Shimura, S., Okayama, H., Aikawa, T., Sasaki, T., and Takishima, T. (1988). Late asthmatic response to Ascaris antigen challenge in dogs treated with metyrapone. Am. Rev. Respir. Dis. 136, 1459-1465. 192. Gundel, R. H., Wegner, C. D., and Letts, L. G. (1992). Antigen-induced acute and late-phase responses in primates. Am. Rev. Respir. Dis. 146, 369-373. 193. Vertes, C., Gonczy, S., Lendvay, N., and Dbreczeni, L. A. (1987).A model for experimental asthma: Provocation in guinea-pigs immunized with BordeteZla pertussis. Bull. Eur. Physiopathol. Respir. 10, 111s-113s. 194. Yamada, N., Kadowaki, S., and Umezu, K. (1992). Development of an animal model of late asthmatic response in guinea pigs and effects of anti-asthmatic drugs. Prostaglandins 43, 507-521. 195. Hayes, J. P., Lotvall, J. O., Baraniuk, J., Daniel, R., Barnes, P. J., Taylor, A. J., and Chung, K. F. (1992). Bronchoconstriction and airway microvascular leakage in guinea pigs sensitized with trimellitic anhydride. Am. Rev. Respir. Dis. 146, 1306-1310. 196. Mapp, C., Hartiala, J., Frick, 0. L., Shields, R. L., and Gold, W. M. (1985). Airway responsiveness to inhaled antigen, histamine, and methacholine in inbred, ragweed sensitive dogs. Am. Reu. Respir. Bs.132, 292-298. 197. Gundel, R. H., Wegner, C. D., Torcellini, C. A., and Letts, L. G. (1992). The role of intercellular adhesion molecule-1 in chronic airway inflammation. CZin. Exp. Allergy 22(5), 569-575. 198. Sun, J., Elwood, W., Haczku, A,, Barnes, P. J., Hellewell, P. G., and Chung, K. F. (1994). Contribution of intercellular-adhesion molecule-1 in allergen-induced airway hyperresponsiveness and inflammation in sensitised brown-Norway rats. Int. Arch. Allergy Immunol. 104(3),291-295. 199. Nakajima, H., Sano, H., Nishimura, T., Yoshida, S., and Iwamoto, I. (1994). Role of vascular cell adhesion molecule llvery late activation antigen 4 and intracellular adhesion moleculeAymphocytefunction-associatedantigen 1interaction in antigen-induced eosinophil and T cell recruitment into tissue. J. Exp. Med. 179, 1145-1154. 200. Richards, I. M., Sun, F. F., Taylor, B. M., Shields, S. K., Griffin, R. L. Morris, J., Wishka, D. G., Smith, H. W., Johnson, R. A., and Dunn, C. J. (1991). Contribution of leukotriene B4 to airway inflammation and the effect of antagonists. Ann. N . Y. Acad. Sci. 629,274-287. 201. Misawa, M., and Arai, H. (1993). Airway inflammation induced by xanthine oxidase in guinea pigs. Agents Actions 38, 19-26. 202. Nakajima, N., Ichinose, M.,Takahashi, T.,Yamauchi, H., Igarashi, A., Miura, M., Inoue, H., Takishima, T., and Shirato, K. (1994). Bradykinin-induced airway inflammation. Contribution of sensory neuropeptides differs according to airway site. Am. J Respir. Grit. Care Med. 149, 694-698. 203. Watson, M. L., Smith, D., Bourne, A. D., Thompson, R. C., and Westwick, J, (1994). Cytokines contribute to airway dysfunction in antigen-challenged guinea pigs: Inhibition of airway hyperreactivity, pulmonary eosinophil accumulation, and tumor necrosis factor generation by pretreatment with an interleukin-1 receptor antagonist. Am. J. Respir. Cell Mol. Biol. 8, 365-369. 204. Hernandez, A., Omini, C., and Daffonchio, L. (1991). Interleukin-1 beta: A possible mediator of lung inflammation and airway hyperreactivity Pharmucol. Res. 24, 385-393.

JIEDIATORS. CITOKISES. .ASD .\DHESIOS JIOLECCLES

303

205. \’an Ooosterhout, A. J., Ladenius, A. R.. Savelkroul, H. F., \‘an Ark, I., Delsman, K. C., and Sijkamp, F. P. (1993).Effect ofanti-IL-5 and IL-3 on ainvay h!perreacti\it\and eosinophils in guinea pigs. A i n Rec. Respir. Dis. 147, 348-552. 206. Bnisselle, G., Kips, J., Joos, G., Bluethmann, H., and Pauwels, R. (1995).Allergeninduced ainvay inflammation and bronchial responsiveness in uild-t\pe and interleukin4-deficient mice. A m J. Respir. Cell Mol. B i d . 12, 254-259. 207. Brusselle, G. G., Kips, J. C., Tavernier, J. H., van der Heyden, J. G., Cuvelier, C . A , , Pauwels, R. A., and Bluethmann, H. (1994).Attenuation of allergic ainva\.inflainination in IL-4 deficient mice. Clin. Erp. Allergy 24, 73-80. 208. Nagai, H., Yamaguchi, S., Inagaki, S . , Tsuruoka, N., Hitoshi, Y., and Takatsu, K. (1993). Effect of anti-IL-5 monoclonal antibody on allergic eosinophilia and ainvay h,yperresponsiveness in mice. Lije Sci. 53, PL243-PL247. 209. Lukacs, N. W., Strieter, R. M., Chensue, S. I\’., and Kunkel, S. L. (1994). Interleukin4-dependent pulmonary eosinophil infiltration in a murine model of asthma. Aiii. J . Respir. Cell. Mol. Biol. 10, 526-531. 210. Lukacs, N. W., Strieter, R. M., Chensue, S. W., Widmer, M., and Kunkel, S. L. (1995). TNFa mediates recruitment of neutrophils and eosinophils during ainvay inflammation. J . Zrniizunol. 154, 5411-5417. 211. Lukacs, N. W., Strieter, R. M., Shaklee, C. L., Chensue, S. W., and Kunkel. S. L. (1995). Macrophage inflammatory protein-1 a influences eosinophil recruitment in antigen-specific airway inflammation. Eur. J. Itninunol. 25, 245-251. 212. Till, G. O., Johnson, K. J., Kunkel, R., et al. (1982). Intravascular activation of complement and acute lung injuy: Dependency on neutrophils and toxic oxygen metabolites. J. Clin. 1nt;est. 69, 1126-1135. 213. Vogelmeier, C., Hubbard, R. C . , Fells, G. A,, Schnebli, H. P., Thompson, R. C., Fritz, H., and Crystal, R. G. (1991). Anti-neutrophil elastase defense of the normal human respiratory epithelial surface provided by the secretory leukoprotease inhibitor. J. Clin. Incest. 87, 482-488. 214. Vogelmeier, C., Buhl, R., Hoyt, R. F.,Wilson, E., Fells, G. A., Hubbard, R. C., Schnebli, H. P., Thompson, R . C., and Crystal, R. G. (1990).Aerosolization of recombinant SLPI to augment antineutrophil elastase protection of pulmonary epithelium. J. Appl. Physiol. 69, 1843-1848. 215. Rudolphus, A,, Stolk, J., Dijkntan, J. H., and Kranips, J. A. (1993). Inhibition of lipopolysaccharide-induced pulmonary emphysema by intratracheally instilled recombinant secretory leukocyte proteinase inhibitor. Ain. Rec. Respir. Dis. 147, 442-447. 216. Urbanski, S. J., Edwards, D. R., Maitland, A,, Leco, K. J., Watson, A,, and Nossakowska, A. E. (1992). Expression of metalloproteases and their inhibitors in primary pulmonary carcinomas. Br. J. Cancer 66, 1188-1194. 21 7. Smith, C. A,, Davis, T., Wignall, J. M., Din, W. S., Farrah, T., Upton, C., McFadden, G., and Goodwin, R. G. (1991). T2 open readmg fram forin the shope fibroma ~ i r u s encodes a soluble form of the TNF receptor. Biochetn. Biophys Res. Coininun. 176, 33.5-340. 218. Sprigs, C. A,, Hruby, D. E., Maliszewski, C. R., Pickup, D. J., Sims, J. E., B u l k , R. M. L., and Van Slyke, J, (1992).Vaccina and cowpox viruses encode a novel secreted IL-1 binding protein. Cell 71, 145-153. 219. Mulligan, M. S., Desrochers, P. E., Chinnaiyan, A. M., Gibbs, D. F.,Varani, J., Johnson, K. J,, and Weiss, S. J. (1993).In vivo suppression of immune complex-induced dveohtis by secretory Ieukoproteinase inhibitor and tissue inhibitor of metalloproteinases. Proc. Natl. Acad. Sci. USA 90, 11523-11527.

304

NICHOLAS W. ISJKACS AND PETER A. WARD

220. David, L. S., Kavanaugh, A. F., Nichols, L. A,, and Lipsky, P. E. (1995). Induction of persistent T cell hyporesponsiveness in vivo by monoclonal antibody to ICAM-1 in patients with rheumatoid arthritis. 1.Immunol. 154, 3525-3537. 221. Smith, D. R., Kunkel, S. L., Standiford, T. J., Rolfe, M. W., Lynch, J. P., Arenberg, D. A., Wilke, C. A,, Burdick, M. D., Martinez, F. J., Hampton, J. N., Whyte, R. I., Orringer, M. B., and Strieter, R. M. (1995).Increased interleukin-1 receptor antagonist in idiopathic pulmonary fibrosis: A compartmental analysis. Am. Rev. Respir. Dis. 151, 1965-1973. 222. Chensue, S. W., Bienkowski, M., Eessalu, T. E., Warmington, K. S., Hershey, S. D., Lukacs, N. W., and Kunkel, S. L. (1994). Endogenous IL-1 receptor antagonist protein ( I M P ) regulates schistosome egg granuloma formation and the regional lymph node response. 1. Immunol. 151, 3654-3662. 223. Fiorentino, D. F., Zlotnik, A., Mosmann, T. R., Howard, M., and O’Garra, A. (1991). IL-10 inhibits cytokine production by activated macrophages.]. Immunol. 147,38153822. 224. Sone, S., Yanagawa, H., Nishioka, Y., Orino, E., Bhaskaran, G., Nil, A,, Mizuno, K., Heike, Y., ogushi, F., and Ogura, T. (1992). Interleukin-4 as a potent down-regulator for human alveolar macrophages capable of producing tumor necrosis factor-alpha and interleukin-1. Eur, Respir. J. 5, 174-181. 225. Moore, K. W., O’Garra, A., de Wall Malefyt, R., Vieira, P., and Mosmann, T. R. (1993). Interleukin-10. Annu. Rev. Immunol. 11, 165-190. 226. Nishioka, Y., Sone, S., Orino, E., Nili, A., and Ogura, T. (1991). Downregulation by interleukin 4 of activation of human alveolar macrophages to the tumoricidal state. Cancer Res. 51,5526-5531. 227. Sempowski, G. D., Beckmann, M. P., Derdak, S., and Phipps, R. P. (1994). Subsets of murine lung fibroblasts express membrane-bound and soluble IL-4 receptors. Role of IL-4 in enhancing fibroblast proliferation and collagen synthesis. J. Immunol. 152,3606-3614.

This article was accepted for publication on 28 September 1995.

INDEX

A Active immunotherapy, 219 Acute respiratory distress syndrome (ARDS), 260-265, 267, 270, 271, 286 Acyl salicylates, PGHS, 187 Addressin, 257 Adhesion molecules ICAM-1, 77-78 pulmonary inflammation, 267, 288 Adipocytes, IFN-y and, 91 &way inflammation,see Pulmoualy inflammation Allergic pulmonary inflammation, 283-28.5 Allograft rejection CD28-B7 costimulation, 142 IFN-y and, 101-103 Alveolitis, extrinsic allergic, 267 Anaphylaxis, interferon-? and, 103 Anchor residues, 231 Antibodies IFN-y, 63, 87-89 PGHS-1 or PGHS-2, 178 Antigen-presenting cells (APCs), 231 Antigens IFN-y and, 80 tumor antigens, 217-228 BAGE, 229, 233 blood group antigens, 220-221 carbohydrate antigens, 219-221 carcinoembryonic antigen, 221-222 c-myc and c-nyb, 227 defined by antibodies, 218-229 defined by T cells, 227-231

epithelial antigens, 238-239 GAGE, 229, 233 gangliosides, 219-220 g~100, 229, 235-236 HER-Ynm, 223-224, 237-238 HPV, 229, 239-240 idiotypes, 225 MAGE, 228, 229, 231-232 MART-1, 229, 235 MELAN-A, 229, 235 melanocyhc differentiation antigens, 233-236 MHC class 11, 80 MUC-1 mucin, 229, 238-239 mucins, 222-223, 238-239 MUM-1, 229, 237 mutated antigens, 236-237 mutated oncogene products, 226-227 oncogenic proteins, 237-238 ~53,226-227,237-238 Pmell7, 229, 235 prostate-specific antigen, 224-225 protein antigens, 221-225 tyrosinase, 233-234 viral antigens, 239-240 Antiproliferative effect, IFN-y, 76-77 cq-Antitqpin, 266 Arachidonate, cyclooxygenase reaction, 178-183, 198 Argl20, 17.5 Arg277, 174 Arthritis, autoimmune, 99-100, 107 hsn410, 178 Asn580, 174 305

306

INDEX

Aspirin antiplatelet cardiovascular therapy, 200 colon cancer, 190 PGHS, 187 Asthma, 283-285 Astrocytes, 84 Autoantibodies germline origin, 19 VH segments, 19 Autoimmune disease CD28 costimulation, 143 IFN-y and, 78, 93, 98-101, 105, 107 VHpolymorphisms, 9

B B7, IGN-7 and, 78 B7-1, 132, 135-138, 156-157 B7-1/B7-2 ligand family, 135-138, 156 B7-2, 132, 135-138, 156-157 B7-CD28 costimulation, Td activation of B cells, 132, 142-144, 148-157 BAGE, 229, 233 B cells IFN-y, 87 T cell-dependent responses, 131 B7-1/B7-2 ligand family, 135-138, 140-142 CD28-B7 costimulation, 142-144, 148-157 CD28lCTLA-4 receptor family, 132-135, 138-140 cellular events, 144-148 costimulation, 138-144, 148-154 model, 131-132 Bcl-2, 43, 53 Bcl-x, 43, 53 Blastocysts, RAG-2-deficient blastocyst complimentation, 31-41 Blood group antigens, 220-221 Bronchopulmonary mucomycosis, 283 Bruton’s tyrosine kinase (Btk), 42, 47

C C1 inhibitor protein (ClINH), 80 C3a, 261, 262, 275, 286 C5a, 261, 269, 275, 286

Cachectin, 96 Cachexia, 96 Cdcineurin, 42, 46-47 CAMP response elements (CRE), 194 Cancer interferon-y and, 94-96 PGSH-2, 190, 201 Cancer antigens, 240 Carbohydrate antigens, 219-221 Carcinoembryonic antigen (CEA), 221-222 Cathepsin G, 266 CD11, 258, 272, 276, 279 CD18, 258, 272, 276, 279 CDI9, 225 CD20, 225 CD21,42,43-44 CD28, 132-135, 156 CD28-B7 costimulation, Td activation of B cells, 132, 142-144, 148-157 CD40, 41-43, 146, 153-154 CD80, 132 CD86, 132 Central nervous system autoimmunity in, 98-99 IFN-y and, 91-92 Cervical cancer, antigens, 240 Chemokines, 80 dfergic airway inflammation, 285 granulomatous lesions, 282 pulmonary inflammation, 259, 268-270 Chimeric mouse, HAG-2-deficient blastocyst complementation. 36 CH locus, 1 Chromosome 14, V,, locus mapping, 4-7, 23, 24 Chromosome 15,orphon VH and DHloci, 10-12, 19, 20, 23 Chromosome 16, orphon VH and DHloci, 10-12, 19, 20, 23 Chronic obstructive pulmonary disease (COPD), 261, 286 ck-regulatory elements, 37, 42-43, 49-52 C-jun, 42, 48 Chss I1 inactivator (CIITA), 106 c-myb, 227 c-myc, 227 Cobra venom factor, 287-288 Collagenase, 266, 288 Colon cancer, 190, 201

307

INDEX

Complement cascade, puhnonary inflammation. 260-262, 271, 286288 Connective tissue. IFN-y, 90-91 Conserved pseudogenes, 16, 24 Constitutive NO synthase, 72-73 Costimulation, 132 07-1 and 07-2, 140-142 CD28-B7, 132 in vivo immune response. 142-144 Td B cell response, 148-157 CD28 and CTLA-4, 138-139 Csk, 42, 45 CTLA-4, 132, 132-135, 156 CTLA4Ig, 138, 141, 143, 149, 150 CX5a, 261 C-X-C family. 269-270 Cyclooxygenase active site, PGHS-2, 174-176 Cyclooxygenase, i s o p e s catalysis by PGHS. 178-188 celldar and physiologic action of PGHS, 196-200 future work, 201-202 pathophysiologies, 200-201 regulation of gene expression, 188- 196 structure-function relationships, 169-178 Cyclophospharnide, diabetes induced by, 100-101 Cytokine release syndromes, IFN-y and, 103-104 Cytokines, 195 granulomatous lesions, 281 pulmonary inflammation, 267-271, 280. 288 Cytokine synthesis inhibitory factor, 85 Cytotoxic T lymphocytes (CTLs), 87, 228, 23 1

Diabetes, interferon-y and, 100-101 Diclofenac, 184, 186 Diverged pseudogenes, 16 DuPG97. 186

E a , 69 E,, 43, 51-52 Ey, 69 Early response cytokines, 267, 276, 280 Elastin, 266 Emphysema, 188, 266 Encephalomyelitis. experimental autoimmune (EAE), 98-99, 105, 107 Endocrine cells, IFN-y and, 93-94 Endothelial cells, IFN-y, 81 Endothelial venules (HEV), L-selectin association with lymphocyte adhesion, 257 Eosinophils, asthma, 283, 285 Epidermal growth factor domain, PGHS, 169, 171 Episalin, 223 Epithelial antigens, 238-239 Epithelial barriers, IFN-y, 90 Epithelial tumors, antigens, 220-221, 223-224 Elythropoiesis, IFN-y and, 93 ES cells, mutant, 32, 36-37, 40 E-selectin, 259, 267, 277 Etodolac, 186 Ets-1, 42, 47-48 EvohJtion, VH loci, 19-26 Experimental autoimmune encephalomyelitis (EAE), 98-99, 105, 107 Experimental autoimmune uveitis, 98-99 Extrinsic allergic alveolitis, 267

D Delayed-type hypersensitivity (DTH), IFN-y and, 96-98 Dengue virus, IFN-y and infection, 77 Dexamethasone, 192, 195 DI-IA, 186 Dr, segments, 1, 23 mouse and human, 17-19 organization, 9-10

F Fc receptors, 77 Fibroblast interferons, 61 Flurbiprofen, 184, 186 N-Formyl-kynurenine, 71 FS cells, IFN-y effects, 94

308

INDEX

Hyper-IgE recurrent infection (HIE), 89 Hypophysis, IFN-y effect on, 93 GAGE, 229,233 Gamma-activated factor (GAF), 68 Gangliosides, 219-220 GEP-1, 76 GCs, 146-148, 150 GD2, 219 GD3, 219 Gelatinase, 288 Gene conversion, 21-26 Gene function, lymphocytes, 31-53 Genes neo, 36-38 PGHS, 192 Rb, 42, 48-49 tk, 36 Germinal center cells (GCs), 146-148, 150 Germline mutational approach, 31, 34-36 Germline VHsegments, 1-26 6111203, 176-177 Glucocorticoids, 259 N-Glycosylation sites, PGHS, 173-174 GM2, 219 GM3, 219 GMSBCG vaccine, 220 ~ 7 5229, , 236-237 gp96, 244 gp100, 229, 235-236 Graft vs. host reactions, IFN-y, 101-103, 105 Granuloma, in pulmonary inflammation, 280-283 Graves’ disease, 19 Guanylyl cyclase, 74

H Heat shock proteins, 78, 244 hsp70,244 Heavy-chain locus, immunoglobulin, 1-26 Hematopoiesis, IFN-y and, 92-93 HER-Uneu, 223-224,237-238 HEV, 257 His207, 176-177 HLA-A2.1, 234 Human papilloma virus (HPV), 229, 239-240 HPV -16, 239-240 HPV -18, 239-240

i IE, 43, 49-50 IyZb, 43, 49-50 Ibuprofen, 184, 186 ICAM-1 adhesion molecule, 77-78 central nervous system cells, 91 interferon-y, 92 pulmonary inflammation, 267, 272, 276, 277, 281, 288 Idiopathic pulmonary fibrosis (IPF), 261, 267, 270, 286 IDO, 71 Immune complexes, lung inflammation, 277-283 Immune interferon, 61 Immune response Td B cell response, 131-157 tumor antigens, 217-245 Immune suppression, IFN-y as mediator, 89-90, 105-107 Immunoglobulin, heavy-chain locus, 1-26 Immunoglobulin E, allergic airway inflammation, 283 Immunoglobulin K enhancers, 43, 50-51 Immunoglobulin K 3’ enhancer, 43, 49 Immunopathology, IFN-y and, 96-104, 106-107 Immunotherapy pulmonary inflammation, 259,285-290 tumors, 218-219 Indoleamine-2,3-dioxygenase(IDO), 71 Indomethacin, 184, 186, 201 Infection, interferon-y and, 94-96 Inflammation IFN-y and, 103-104, 105, 281, 288-289 PGHS-2, 201 pulmonary, see Pulmonary inflammation Insulitis, autoimmune, 100-101 Plcur Integrins, 258 &-Integrins, 279 Interferon, immune, 61 Interferon-a (IFN-a), 61, 84-85 Interferon-p (IFN-p), 61, 84-85

INDEX

Interferon-y (IFN-y), 61-62, 105-107 antagonists, 82-83, 105 B7 expression, 136 biochemistry cellular, 71-76 signal transduction, 67-71 biological effects adipocytes and, 91 antibody formation, 87-89 antigen presentation, 80 antiproliferative effect. 76-77 central nervous system cells and, 91-92 connective tissue and, 90-91 endocrine ceIls and, 93-94 endothelial cells, 81 epithelial barriers and, 90 in hematopoiesis, 92-93 immune suppression mediator, 89-90, 105 lymphocytes and, 86-87 mem brane protein expression, 77-78 mononuclear phagocytes, 78-80 natural IFN-y antagonists, 82-86 synergy between IFN-y and TNF, 81-82 CD28 costimulation, 139 immunopathology allograft rejection, 101-103 autoimmune disease, 98-101 cytokine release syndrome, 103-104 delayed-type hypersensitivity, 96-98 granulomatous lesions, 281 nonspecific inflammation, 103- 104 infection and cancer, 94-96 monocytes, 136 production, 63-66, 105 pulmonary inflammation, 281, 288-289 recent research, 106-107 structure-function relationship, 62-63 Interferon-y receptor biochemistry, 66-67, 105 IFN-yRa, 66 IFN-yRP, 66 Interferon-regulatory factors, 69-70 Interferon-sensitive response elements ( ISREs), 69-70 Intergenic probes, 19-20 Interleukin-1 (IL-1) adhesion molecules, 81 autoimmune arthritis, 99-100 CD28 costimulation, 139

309

granulomatous lesions, 281 interferon-y, 80 puhnonaly jnflamrnatios, 259, 267, 273, 276, 277, 279, 280, 288-289 Interleukin-2 (IL-2), 88 allograft rejection, 102, 103 B7 expression, 136 CD28 costimulation. 139 interferon-y, 80 Th cells, 146 Interleukin-4 (IL-4), 83-84, 85 allograft rejection. 102 antibody response, 88 asthma, 283, 285 B7 expression, 136 CD28 costimulation, 139 granulomatous lesions, 281, 290 interferon-y, 105 pu~monaryinflammation, 260, 290 Interleukin-5 (IL-5) allergic airway inflammation, 283-285 Th cells, 146 Interleukin-6 (IL-6). 86 antibody resonse, 88 CD28 costimulation, 139 pulmonary inflammation, 289 Interleukin-8 (IL-8) interferon-y, 80 pulmonary inflammation, 260, 270, 273, 274 Interleukin-10 (IL-lo), 85 allograft rejection, 102 antibody response, 88 granulomatous lesions, 281 interferon-?, 105 PGHS-2, 192, 195 pulmonary inflammation, 259-260, 274-275, 279, 289 Th cells, 146 Interleukin-12 (IL-12) granulomatous lesions, 281 interferon-?, 105, 107 Th cells, 146 Interleukin-13 (IL-l3), 85-86, 260 Inter~eukin-1S(IL-15). 106-107 Interleukin-16 (IL-16), 136, 146 Interleukin receptor-antagonist protein (IRAP), 289 IP-10, 76, 80 IPF, 261, 267, 270,286

310

INDEX

IRF-1, 69-70 IRF-2,69-70 IRG-47, 76 Ischemidreperfusion, pulmonary inflammation, 275-277 Islets of Langerhans, interferon-y and, 100- 101 ISREs, 69-70

J JAK-1, 68-69, 106 JAK-2, 68, 69, 106 Jak family, 68-69, 106 Jak-STAT mechanism, IFN-y binding, 67-71 Janus kinase family, 68 JH segments, 1 deletion, 32-33, 42 in mouse and human, 17-19 organization, 9-10

KC, 80 Keratinocytes, IFN-y effect on, 76 Ketorolac tromethamine, 186 Keyhole limpet hernocyanin (KLH), 149, 220 Kynurenine, 71

L Leukocyte interferons, 61 LFA-1, IFN-y and, 78, 81 L-selectin, 257-258 LTBI, 269 Lung inflammation, see Pulmonary inflammation Lymphocyte homing receptor, 257 Lymphocyte receptors, 41-44 Lymphocytes gene function Bcl-2 and Bcl-x, 43, 53 cis-regulatory elements, 42-43, 49-52 perforin, 43, 52-53 signal transduction, 42, 44-47 TdT, 43, 52 transcription, 42, 47-49

IFN-y and, 86-87 RAG-2-deficient blastocyst coplementation, 31-41 T cell activation. 131-157

Macrophage-activating factor (MAF), 61-62, 96 Macrophage inflammatory protein-1 (MIP-1), 260, 270, 280, 285 Macrophage inflammatory protein-2 (MIP-2), 274 Macrophages activation, 78-80, 86, 277 PGHS-2, 191-192 MadCam-1, 101, 257 MAGE, 228, 229, 231-232 MAGE-3, 232 Malignant transformation, p53, 226-227 Maiigano-heme PGHS-1, 188 Mapping, Vt, locus, 4-7 MART-1, 229, 235 MCP-1, 80, 282 Meclofenamate, 184, 186 MELAN-A, 229, 235 Melanocytic differentiation antigens, 233-236 Melanoma, antigens, 219-220, 228, 230, 231-236 Membrane-binding domain, PGHS, 172-173 Membrane proteins, IFN-y, 66, 77-78 Metalloproteases, 266-267, 288 Mg21, 76 MHC class XI molecules, 80 6-MNA, 186 Monocytes interferon-y and, 80, 136 PGHS-2, 191-192 N'-Monomethyl-L-arginine (L-NMMA),72, 7599 Mononuclear phagocytes, IGN-y and, 78-80 MUC-1, 223, 229, 238-239 Mucins, 222-223, 238-239 Mucosal addressin, 257 Multiple sclerosis, IFN-y and, 75, 91 MUM-1, 229, 237 Myelopoiesis,92-93 MZ2-E, 228, 231-232

INDEX

Naproxen, 184, 186 Natural killer cells, interferon-y production, 63-66, 105, 106-107 Natural killer-cell stimulating factor, 105 neo gene, 36-38 Neutrophil elastase, 266 NF-IL-GICIEBP response elements, 194 Nitric-oxide synthase, 72-73 Nitrogen monoxide, in biological systems, 72-76, 99, 105, 264-265 Nonspecific inflammation, IFN-y and, 103-104 Nonsteroidal antiinflammatory drugs (NSAIDs), PGHS and, 167, 184-187 class I NSAIDs, 184, 185-186, 201 class I1 NSAIDs, 184, 185, 186 class 111 NSAIDs. 185 NS398, 186 NSAIDs, see Nonsteroidal antiinflammatory drugs

Oncogenic proteins, 237-238 Orphons, 10-12, 20, 22-24 Oxidant formation, 262-267 Oxygen radicals, 72, 262-264

P p53, 226-227,237-238 Pancreas, interferon-y and, 100-101 Pancreatitis, autoimmune, 98 Pentanoylsalicylate, 187 Perforin, 43, 52-53 Peroxidase active site, PHGS-2, 176-177, 187-188 Phagocytes, mononuclear, IGN-y and, 78-80 Phenylbutazone, 186 Phosphoinositide-3 kinase, 133-135 Physical mapping, VH locus, 4-7 Piroxiiicam, 184, 186, 201 Platelet aggregation, PGHS-1, 200-201 Platelet-derived growth factor (PDGF), 268 Pmell7, 229, 235 Polymorphism, VH locus, 8-9

311

Promoter regions, PGHS-1 and PGHS-2, 192-196 Propionylsalicylate, 187 Prostaglandin En, 198 Prostaglandin endoperoxide H synthase (PGHS) cellular and physiologic action of PCHS, 196-200 future work, 201-202 NSAIDs, 184-187 pathophysiologies, 200-201 regulation of gene expression, 188-196 structure-function relationships, 169-178 Prostaglandin endoperoxide H synthase-l (PCHS-1) catalysis by, 178-188 cellular and physiologic action, 196-200 future work, 201-202 gene structure, 192 mangano-heme PGHS-1, 188 NSAIDs, 184-187 pathophysiology, 200-201 platelet aggregation, 200-201 promoters, 192-196 prostanoid production, 196-199 regulation of gene expression, 188-189 structure-function relationship, 169-178 Prostaglandin endoperoxide H synthase-2 (PCHS-2) catalysis by, 178-188 cellular and physiologic action, 196-200 cyclooxygenase active site, 174-176 future work, 201-202 gene structure, 192 inflammation, 201 NSAIDs, 184-187 pathophysiology, 200, 201 peroxidase active site, 176-177, 187-188 promoters, 192-196 prostanoid production, 199-200 regulation of gene expression, 188-196 structure-function relationship, 169-178 trypsin cleavage sites, 174 Prostaglandin E synthase, 198 Prostaglandin G, 174 Prostaglandin Hn, 198 Prostaglandin In synthase, 200 Prostanoids, 167, 168, 196-200 Prostate-specific antigen (PSA), 224-225 Protein antigens, 221-225

312

INDEX

Proteinases, pulmonary inflammation, 266-267 Provoked imunity, 244 P-selectin, 276 Pseudogenes, 16-17, 24-26 P/STEL sequences, PHGS-2, 177 pTa, 42, 44 Pteridin metabolism, IFN-y, 72 Pdmonary inflammation, 257-260, 290 adhesion molecules, 267, 288 allergic, 283-285 animal models, 271-285 complement cascade, 260-262, 271, 286-288 cytokines, 267-271, 280, 288 granuloma formation, 280-283, 290 immune-complex mediated, 277-279 ischemidreperfusion, 275-277 oxidant formation, 262-267 proteinases, 266-267 sepsis, 271-275 therapeutic interventions, 259, 285-290

Secretory leukoprotease inhibitor (SLPI ), 287-288 Selectins, 257-258, 259, 267, 276, 277 Sepsis, pulmonary inflammation, 271-275 Septic shock complement cascade, 271 interferon-y and, 103, 104 Ser530, 175 Serum sickness, interferon-y and, 103 Signal transduction B7-1 and B7-2, 138 CD28, 134 IFN-y binding, 67-71, 106 in lymphocytes, 42, 44-47 Sjogren’s syndrome, 76-77 STAT-1, 69 STAT-2, 69 STAT48,69 STAT91, 68 STAT proteins, 68, 69 Stem cell factor, I89 Streptozotocin, diabetes induced by, 100 Stromelysin, 288 Sulindac, 184, 186, 201

R RAG-2-deficient blastocyst complementation, 31-41 RANTES, 285 Ras, 42, 46 Rb gene, 42, 48-49 Reactive oxygen IFN-y, 72 pulmonary inflammation, 262, 263 Receptors interferon-y action, 66-67, 105 in lymphocytes, 41-44 Recombination, V-D-J, 1 Reperfusion, pulmonary inflammation, 275-277 Reproduction, PGHS-1, 191 Retinoblastoma gene, 42, 48-49

Sdicylates, PGHS, 187 Salicylic acid, 184-187 Sarcoidosis, 267, 281 SC58125, 186

T T-cell activation, 131 B cells, 144-148 B7-1/B7-2 ligand family, 135-138, 140-142 CD28-B7 costimulation, 142-144, 148-157 CD28/CTLA-4 receptor family, 132-135, 138-140 costimulation, 138-144, 148-154 model, 131-132 T cell receptor p enhancer, 43, 51-52 T cells interferon-y production, 63-66, 105 tumor antigens defined by, 227-231 Terminal deoxylnucleotidyltransferase,43, 52 Tetrahydrobiopterin, IFN-y, 72 T helper 2 cells, 146 Thromboxane A2, 198 Thromboxane-A synthase, 198, 200 Thyroid, IFN-y effect on, 93 Tissue inhibitor of metalloprotease-1, 267, 288

INDEX

Tissue inhibitor of metalloprotease-2, 267, 288 tk gene, 36 Toxic shock syndrome, interferon-y and, 103 Transcription, in lymphocytes, 42, 47-49 Transformation, malignant, p53. 226-227 Transforming growth factor-p (TGF-P), 84-85, 90, 105, 269 Translocation, V,, and DHsegments, 11, 19 Trypsin cleavage sites, PGHS-2. 174 Tryptophan metabolism, IFN-y, 71-72 Tumor necrosis factor, synergy between IFNy and, 81-82 Tumor necrosis factor a,81, 86, 139 allergic ainvay inflammation, 285 granulomatous lesions, 281, 282 pulmonary inflammation, 259, 260, 267, 273, 288-289 in hypersensitivity granuloma formation, 280-281 regulation in IC-mediated inflammation, 279 upregulation of vascular E-selectin and ICAM-1, 276, 277 Tumor antigens, 217-218, 240-245 Tumors defined by antibodies, 218-227 defined by T cells, 227-240 interferon-? and, 96, 101-103 Two-signal model, T cell activation, 131-132 TYK-2, 68 Type I interferon, see Interferon-a; Interferon+ Type II interferon, see Interferon-y Tyr385, 175, 178, 180 Tyrosinase, 233-234

U937 cells, PGHS-2 expression regulation, 194-195 Uveitis, experimental autoimmune, 98-99

V V1-2 segment, 18 V1-3 segment, 18 V1-8 segment, 18

313

V1-12P pseudogene, 23 V1-18 segment, 24 V1-24P segment, 16 V1-45 segment, 17 V1-58P segment, 16, 17 V1-69 segment. 18, 19 V2-5 segment, 18 V3 region, 24 V3-7 segment, 18, 19 V3-9 segment, 18, 20 V3-11 segment, 18 V3-13 segment, 20 V3-15 segment, 18, 19 V3-16P segment, 16 V3-20 segment, 17, 20 V3-21 segment, 20 V3-23 segment, 19 V3-30 segment, 18-19, 20 V3-33 segment, 20 V3-35 segment, 17 V3-38P segment, 16 V3-43 segment, 20, 25 V3-47P segment, 20 V3-48 segment, 19, 20 V3-53 segment, 18 V3-54P segment, 16 V3-60P segment, 25 V3-62P segment, 25 V3-64 segment, 17 V3-72 segment, 17 V4-4 segment, 18, 25 V4-28 segment, 17, 20, 25 V4-31, 20 V4-39, 18, 19 V4-59, 18, 19 V4-61, 19 V5-51, 19 V6-1 segment, 18 V7-81 segment, 17 VISC segment, 24 V54 segment, 24 Vaccination, GMS/BCG vaccine, 220 Valerylsalicylate, 187 Vav, 42, 45-46 VCAM-1, 258, 267 V-D-J recombination, 1 Venules, endothelial, L-selectin association with lymphocyte adhesion, 257 VFl-12P pseudogene, 23 VH2family, 13

314 VH3 family, 18 V H family, ~ 13, 25 Vt15 family, 13, 17 VH6family, 13, 17 v117 family, 14, 15 v,, locus evolution, 19-26 physical mapping, 4-7 polymorphisms, 8-9 repetitive sequences, 20, 22 VH pseudogenes, 16-17

INDEX

VH segments, 1-2, 7-8, 12-19 Viral antigens, 239-240 Viruses, IFN-y and infection, 77, 95 VLA-4, 258 V region, 1

Y Yeast artificial chromosome (YAC) system, 3

CONTENTS OF RECENT VOLUMES

Volume 58

Positive Selection of Thymocytes PAMELA J. FINK AND MICHAELJ. BEVAN

NF-kB and Re1 Proteins in Innate Immunity EL~ZABETH B. KOPP A N D SANKAH GHOSH

Molecular and Cellular Aspects of X-Linked Agaminaglobulinemia PASCHALIS SIDERAS AND C. I. EDWARD SMITH

V(D)J Recombination and Double-Strand Break Repair DAVID T. WEAVER

The Common &Chain for Multiple Cytokine Receptors ASAO, K ~ z u oSUGAMURA, HIRONOBU MOTONAHIKONDO, NOBUYUKI TANAKA, NAOTOISHII,MASATAKA NAKAMURA, AND TOSHIKASU NAKAMUHA

Development and Selection of T Cells: Facts and Puzzles PAWEL KISIELOW A N D HARALD VON BOCHMER The Pharmacology of T Cell Apoptosis GUIDOKHOEMER Intraepithelial Lymphocytes and the Immune System GEK-KEE SIM Leukocyte Migration and Adhesion BEATA. IMHOF AND DOMINIQUE DUNON

Self-Tolerance Checkpoints in B Lymphocyte Development CHRISTOPHER C. GOODNOW, JASON 6. CYSTEH, SUZANNE B. HAHTLEY, SAHAII E. BELL,MICHAEL P. COOKE, JAMES L. HEALY, SRINIVAS AKKARAJU, JEFFREY c. RATHMELL, SARAH L. POCUE, AND KEVIN

P. S H O U T

Gene Transfer as Cancer Therapy GLENN DRANOFF A N D RICHARD C. Mu LLIGAE;

The Regulation of Pulmonary Immunity MARYF. LIPSCOMB, DAVID E. RICE,C. RICHARDLYONS, MARK R . SCHUYLEH, AND DAVID WILKES

INDEX

INDEX

Volume 59

Volume 60

The CD1 Family: A Third Lineage of Antigen-Presenting Molecules STEVEN A. PORCELLI

The Janus Protein Tyrosine Kinase Family and Its Role in Cytokine Signaling JAMES N. IHLE

315

316

CONTENTS OF RECENT VOLUMES

X-Linked Agammaglobulinemiaand Immunoglobulin Deficiency with Normal or Elevated IgM: Immunodeficiences of B Cell Development and Differentiation RAMSAY FULEIHAN, N AHAYANASWAMY RAMESH, AND RAIFS. GEHA

The Role of Nitric Oxide in Inflammation C. RICKLYONS INDEX

Volume 61 Defective Glycosyl Phosphatidylinositol Anchor Synthesis and Paroxysmal Nocturnal Hemoglobinuria KINOSHITA,NOHIMITSU INOUE, TAROH AND JUNJI TAKEDA

CD40-CD40 Ligand: A Multifunctional Receptor-Ligand Pair CEESKOOTENAND JACQUES BANCHEREAU Antibody Class Switching JANET STAVNEZEH

The Use of Multiple Antigen Peptides in the Interleukin-2 Receptor Signaling Mechanisms Analysis and Induction of Protective Immune LARHY M. KARNITZ A N D ROBERT T. Responses against Infectious Diseases ABRAHAM G. A. OLIVEIHA, J. M. E. H. NARDIN, CALVO-CALLE. AND R. S. NUSSENZWEIG Control of the Complement System M KATHRYNLISZEWSKI. TIMOTHY C. Eosinophils: Biology and Role in Disease FARHIES. DOUGLAS M. LUBLIN. ISABELLE A. ROONEY, A. J. WAHDLAW, R. MOQBEL, AND A. B. AND JOHN P. ATKINSON KAY

In Situ Studies of the Germinal Center Reaction GARNETTKELSOE

Cytotoxic T Lymphocytes: The New Identified FAS (CD95)-Mediated Killing Mechanism and a Novel Aspect of Their Biological Functions HAJIME TAKAYAMA, HIDEFUMI KoJIMA, AND NOBUKATA SHINOHAHA

V(D)J Recombination Pathology KLAUSSCHWARZ AND CLAUS R. BARTHAM Major Histocompatibility Complex Class I1 Deficiency: A Disease of Gene Regulation VIKTOHSTEIMLE. WALTER REITH,AND BERNARD MACH TH1-TH2 Cells in Allergic Responses: At the Limits of a Concept IWAN AEBISCHERAND BEDAM. STADLER INDEX

This Page Intentionally left blank

ISBN 0-12-022462-3 90051