Advances in Veterinary Medicine
Edited by
W. Jean Dodds HEMOPET Santa Monica, California
James E. Womack Department ...
72 downloads
860 Views
12MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Advances in Veterinary Medicine
Edited by
W. Jean Dodds HEMOPET Santa Monica, California
James E. Womack Department of Veterinary Pathobiology College of Veterinary Medicine Texas A&M University College Station, Texas
Advisory Board Kenneth C. Bovee' William S. Dernell Carolton Gyles Robert 0.Jacoby Ann B. Kier Raymond F. Nachreiner Carl A. Osborne Fred W. Quimby Alan H. Rebar Ronald D. Schultz
This Page Intentionally Left Blank
Advances in Veterinary Medicine
Volume 40
Molecular Genetics, Gene Transfer, and Therapy Edited by
W. Jean Dodds HEMOPET Santa Monica, California
James E. Womack Department of Veterinary Pathobiology College of Veterinary Medicine Texas A&M University College Station, Texas
Academic Press San Diego London Boston New York Sydney Tokyo Toronto
This book is printed on acid-free paper.
@
Copyright 0 1997 by ACADEMIC PRESS 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. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher's consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923). for copying beyond that permitted by Sections 107 or 108 of the U S . Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-1997 chapters are as shown on the title pages, if no fee code appears on the title page, the copy fee is the same as for current chapters. 1093-975)3/97 $25.00
Academic Press
u division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.apnet.com Academic Press Limited 24-28 Oval Road, London NWl 7DX, UK http://www.hbuk.co.uk/ap/ International Standard Book Number: 0- 12-03924 1-0 PRINTED IN THE U"FDSTATES OF AMERICA 97 98 9 9 0 0 01 0 2 E B 9 8 7 6
5
4
3 2 1
CONTENTS
From Molecular Genetics to Diagnosis and Gene Therapy
JENS G. HAUGE
I . Introduction to Molecular Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I1. Gene Manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I11. Studies of Metabolic and Gene Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. DiagnosisofGeneticDisease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Therapy of Genetic Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI . Use of Recombinant DNA Methods to Improve Domestic Animals . . . . . . . VII . Legal and Ethical Aspects of Patenting Biotechnology .................. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 7 18 21 33 35 39 41
Genes Involved in Oncogenesis
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I1. Detection of Genes Involved in Oncogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . I11. Mapping and Cloning Genes Involved in Oncogenesis . . . . . . . . . . . . . . . . . . . IV. Animal Model Systems Used to Study Gene Function . . . . . . . . . . . . . . . . . . . V. Identity and Function of Genes Involved in Oncogenesis . . . . . . . . . . . . . . . . VI . Applications of Oncogene Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
52 53 66 69 73 86 93 93
Diagnostic Molecular Genetics
GEORGE H . SACK.JR.
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 I1. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
vi
CONTENTS
I11. Linkage Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Single-Gene Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. FutureDirections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Gene Therapy for the Hemophilias I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I1. Advantages of Gene Therapy as a Treatment for Hemophilia . . . . . . . . . . . . I11. Advantages of Hemophilia as a Model for Gene Therapy . . . . . . . . . . . . . . . . n! Choice of Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Factor IX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI . Factor VIII . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Genes Controlling Retroviral Virulence Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Retrovirus Structure and Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Additional Retrovirus Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Feline Acquired Immunodeficiency Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . Murine AIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simian AIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Murine Leukemia Virus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endogenous Retroviruses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX. Lentiviruses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I. I1. 111. IV. V. VI . VII . VIII .
Mapping Animal Genomes I. I1. I11. IV.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Why Map Farm Animal Genomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods for Animal Genome Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Current Status of Animal Genome Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CONTENTS
V. Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii 183 184
The Canine Genome CATHRYN S. MELLERSHAND ELAINEA. OSTRANDER Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications of a Canine Genetic Map ................................. Organization of the Canine Genome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development of a Canine Genetic Map ................................ V. Targeted Cloning of Canine Genes VI. Considerations a References . . . .
191 193 196 198
INDEX.....................................................................
217
I. 11. 111. IV.
This Page Intentionally Left Blank
CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors' contributions begin. JANELLE CORTNER, National Cancer Institute/National Institutes of
Health, Division of Basic Sciences, Bethesda, Maryland 20892 (51) FREDERICK J. FULLER, Department of Microbiology, Pathology, and Parasitology, North Carolina State University, College of Veterinary Medicine, Raleigh, North Carolina 27615 (135) JENS G. HAUGE, Department of Biochemistry, Physiology, and Nutrition, Norwegian College of Veterinary Medicine, Oslo, Norway (1) KATHERINEA. HIGH, Departments of Pediatrics, Pathology, and Laboratory Medicine, University of Pennsylvania School of Medicine, and Hematology Division, The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania 19104 (119) CATHRYN S. MELLERSH, Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington 98104 (191) ELAINE A. OSTRANDER, Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington 98104 (191) GEORGE H. SACK, JR., Departments of Medicine, Pediatrics, and Biological Chemistry, The Johns Hopkins University School of Medicine, The Johns Hopkins Hospital, Baltimore, Maryland 21287 (103) SUSAN VANDE WOUDE, Laboratory Animal Resources, Colorado State University, Fort Collins, Colorado 80523 (51) GEORGE F. VANDE WOUDE, ABL-Basic Research Program, Frederick Cancer Research and Development Center, National Cancer Institute, Frederick, Maryland 21702 (51) ix
x
CONTRIBUTORS
JOHANNESWALTER,Department of Cardiothoracic Surgery, University of Vienna, Vienna, Austria (119) JAMES E. WOMACK,Department of Veterinary Pathobiology, College of Veterinary Medicine, Texas A&M University, College Station, Texas 77843 (157)
PREFACE The rapid explosion of knowledge relating recombinant DNA technology to science and medicine has laid the groundwork for its application to diagnosis of genetic disease, gene regulation, development of transgenic animals, and gene therapy. The science of molecular genetics has progressed from an understanding of the basics to manipulations of genes using specific plasmid, viral or shuttle vectors, jet injection, and particle bombardment to introduce reporter genes and transfect a variety of microbial, plant, animal, and human cells. These techniques are now being extended to map the genome of humans and other species. Early studies focused on developing the linkage map, and parallel goals include answers to questions about how genes are regulated and how they interact to control physiology and development. These questions involve studying individual genes, as well as sets of genes-the study of which is now termed "genomics." Recent experiments with transgenic mice have led to the prediction that animal agriculture would undergo an immediate revolution. One could envision "super cows, pigs, and other species," which carry transgenes that would make them particularly productive and healthy. There remains a large knowledge gap, however, in applying this technology in a scientific and ethically acceptable manner to farm animals and other species. The latest breakthrough whereby the lamb "Dolly" was successfully cloned from the cells of an adult sheep has catapulted these issues to the forefront of world debate. Recent publications of specific genetic defects of large animals have demonstrated the applicability of molecular genetics to the improvement of animal health. Examples are the studies of the leukocyte adhesion deficiency (LAD) of Holstein cattle; hyperkalemic periodic paralysis (HPP) of quarterhorses; malignant hyperthermia of swine; and canine progressive retinal atrophy, copper storage disease, pyruvate kinase deficiency, phosphofructokinase deficiency, and von Willebrand's disease (vWD). Development of genetic techniques including sequencing of specific normal and mutant genes has permitted positive identification of carrier animals for effective genetic screening. For example, the frequency
xii
PREFACE
of carriers for bovine LAD varies from 6 to 15%,whereas equine HPP has a major impact on the genotype of the quarterhorse. Similarly, breeds of dogs that carry a high freqency of vWD can be affected to the extent of 15 to 80% of current bloodlines. Not all genetic defects have undesirable consequences, for evolution of some abnormalities may have a potential beneficial effect in the carrier state. Very little is currrently known about the potential physiological or other benefits associated with certain genotypes and their observed phenotype. Identification of genetic markers permits their selection to segregate economically important animals for beneficial traits and allows for selective breeding and even positional cloning of the genes involved. The future for marker-assisted selection of the genetic loci responsible for economically valuable traits is potentially large and can involve a variety of domestic farm animals including cattle, pigs, and chickens. How these tools will be used appropriately and responsibly by the animal industry has yet to be determined. Application of these techniques to the diagnosis and control of infectious disease is also important. For example, understanding the genetic basis of various animal and human retroviruses leads to ways to control their respective diseases. Recent work on studying the equine infectious anemia virus indicated that exchange of the envelope region of a highly virulent strain of the virus with the same region of an avirulent viral strain can make the avirulent strain now become virulent. This research has opened a wide range of possibilities for viral manipulation in controlling the disease in infected horses and also for the development of other antiinflammatory and antipyretic drugs for therapy of the disease. These recent developments in biotechnology have led to an increasing plethora of related patents. The idea of patenting genes, particularly human genes, has raised a series of ethical dilemmas. While patents have become crucial to the biotechnology business, scientists and businessmen are concerned about how patents they file for portions of a gene (i.e., copies of gene fragments) could potentially prevent patenting the useful genes themselves. In other words, having a patent on a part of a gene could prevent the patenting of discoveries that affect the whole gene. Thus, the actual useful genes or gene products may now be protected by patents. A more utilitarian approach could be adopted in order to solve this problem plaguing the commercial biotechnology industry. Linked hand-in-hand with the commercial concerns are the ethical problems of holding a monopoly on patented technology which could provide lifesaving therapeutic products. The
PREFACE
...
xlll
ethical issues are particularly relevant to the application of this technology to veterinary medicine. The present volume addresses various perspectives on these exciting and challenging times. We thank the authors Jens Hauge, Janelle Cortner, Susan Vande Woude, George F. Vande Woude, George H. Sack, Johannes Walter, Katherine A. High, Frederick J. Fuller, Cathryn S. Mellersh, and Elaine A. Ostrander for their scholarly contributions. W. JEAN DODDS JAMES E. WOMACK
This Page Intentionally Left Blank
ADVANCESIN VETERINARYMEDICINE,VOL.40
From Molecular Genetics to Diagnosis and Gene Therapy JENS G. HAUGE Department of Biochemistry, Physiology, and Nutrition, Norwegian College of Veterinary Medicine, Oslo, Norway I. Introduction to Molecular Genetics A. Structure and Replication of Genes B. From Gene to Gene Product II. Gene Manipulation A. Basic Features B. Methods and Tools C. Vectors D. Studies of Gene Structure III. Studies of Metabolic and Gene Regulation A. Tissue Culture Studies B. Regulation in Transgenic Animals IV. Diagnosis of Genetic Disease A. Gene Changes in Hereditary Disease B. Detection of Gene Changes with DNA Probes C. DNA Probes and Disease Susceptibility D. Application to Domestic Animals E. Gene Changes in Cancer V. Therapy of Genetic Diseases A. Germ Line Gene Therapy B. Somatic Gene Therapy VI. Use of Recombinant DNA Methods to Improve Domestic Animals A. Markers and Maps B. Parentage Identification C. Transgenic Animals VII. Legal and Ethical Aspects of Patenting Biotechnology References
Copyright 9 1997by AcademicPress. All rights of reproduction in any form reserved. 1093-975X/97 $25.00
2
JENS G. HAUGE
I. I n t r o d u c t i o n to M o l e c u l a r G e n e t i c s
During the past 20 years, an immense expansion has taken place in our understanding of the structure, organization, and regulation of genes. This expansion has, to a large extent, been the result of the development of new methods to manipulate, multiply, and study DNA fragments. The same new methods have a number of important applications in h u m a n medicine and, gradually, also in veterinary medicine. For good overviews, see Old and Primrose (1990) and Watson et al., (1992). An agricultural perspective is given in the book edited by Hansel and Weir (1990). These surveys also discuss the large impact on microbiological diagnosis and vaccine production, aspects that are not addressed here. The development of gene technology rests on the foundations of molecular biology laid during the preceding 30 years, with the discovery of the double helical structure if DNA and the breaking of the genetic code as the major milestones. I will start this chapter, therefore, with a brief review of the molecular biological basis of gene technology. A. STRUCTURE AND REPLICATION OF GENES The notion that heredity was linked to matter, to certain cellular structures, appeared for the first time in 1903, when the chromosomal theory of heredity was formulated. Chromosomes were recognized to have properties that could accommodate the hereditary units discovered by Mendel in his experiments. A bridge was built between cytology and genetics. But the nature of the molecules that carried the hereditary information was not thereby identified. Many researchers believed that chromosomal protein was the most likely candidate. In 1944, however, Avery and co-workers showed convincingly that genes in pneumococcal bacteria are molecules of deoxyribonucleic acid (DNA). It soon became apparent that pneumococci were no exception in this respect. In animals and plants, as well as in bacteria, hereditary information is carried by DNA. Figure l a shows a section of a DNA molecule. DNA is built in the form of a chain; phosphoric acid and the sugar deoxyribose alternate as the links in this chain. Phosphoric acid always connects the 5'-OH group in one deoxyribose to the 3'-OH in the next deoxyribose. The heterocyclic bases adenine, guanine, cytosine, and thymine are bound to the 1' carbons in deoxyribose. The base-deoxyribose moiety is a nucleoside, and the repeating unit in the chain, the combination b a s e deoxyribonucleoside-phosphoric acid, a nucleotide.
FIG. 1. (a) Structure of part of a DNA chain. (From BIOCHEMISTRY, 2E by Lubert Stryer. Copyright 0 1975, 1981. Reprinted with permission of W. H. Freeman and Company.) (b) DNA double helix during replication. (Reprinted by permission from Jones and Bartlett Publishers, from Molecular Biology by David Freifelder, 1983, p. 254.)
4
JENS G. HAUGE
It is impractical to write DNA structures with all the atoms, and simplified writings have come into use. If the bases from the top on down in Fig. l a are adenine, thymine, and guanine, the most-used abbreviation is simply ATG. One uses the first letter in the name of the base to designate the corresponding nucleotide. That such chainlike DNA molecules could carry hereditary information, genes, was not difficult to conceive. Most genes determine the construction of proteins, which also are chainlike molecules, with amino acids making up the links in the chain. One needs only to postulate a one-to-one relationship between an amino acid and a group of DNA bases. The other fundamental property of the hereditary material, that it is accurately duplicated and distributed on both daughter cells when cells divide, was harder to visualize on the basis of the DNA structure, as shown in Fig. la. Watson felt that the solution to the problem perhaps lay hidden in the three-dimensional structure of DNA and, together with Crick, he set out to elucidate this structure. The result was the discovery in 1953 of the double-helix structure of DNA. The double helix is built as shown schematically in Fig. lb. A, T, G, and C here represent the bases, which point toward each other and are bound to each other through hydrogen bonds. The DNA strands of the double helix have opposite directions (a DNA chain has a 5' end and a 3' end; see Fig. la). The size of the bases and their hydrogen bonding properties make only two pairings possible, namely, A with T and G with C. One turn of the s t a n d a r d double helix has 10 base pairs (bp) and covers a length of 3.4 x 10 -6 mm. The diameter of the double helix is 2 x 10 -6 mm. It is so slim a structure that if a DNA molecule could reach from the E a r t h to the moon, it would weight only 1 mg! The model for the DNA molecule that Watson and Crick arrived at can in a simple way explain how DNA molecules are duplicated, i.e., the process of DNA replication. When we consider the double helix (Fig. lb), we see that the two strands correspond to each other. The base-pairing rules determine which base in one strand must be opposite a given base in the other strand. The strands are said to be complementary. When replication starts, the hydrogen bonds between the two strands are gradually broken (Fig. lb). New DNA chains are constructed from the four deoxyribonucleoside triphosphates with the help of the enzyme DNA polymerase, one molecule of pyrophosphate being eliminated per nucleotide laid down. Step by step, two new strands, complementary to the original strands, are synthesized. The original strands serve as templates in the process, and the result is two new double helixes, identical to each other and to the original DNA molecule.
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
5
The basic mechanism for DNA replication is simple and elegant. The practical execution of DNA replication in the cells, however, involves a large number of proteins, including some extra enzymes. The unwinding of the DNA molecules causes problems, as does the fact that DNA polymerase lays down a new strand from the 5' end toward the 3' end only. Furthermore, for this synthesis to start, a primer nucleic acid is needed for the strand to grow from. One important auxiliary enzyme is DNA ligase, which links together DNA molecules. If DNA is heated to 80-90~ the molecule will denature and the strands separate. If the temperature is lowered somewhat, the complementary strands will find each other again and re-form the double helix. This property is widely exploited in gene technology (see Section II,B,1). B. FROM
GENE
TO GENE
PRODUCT
When the information stored in DNA is used, organisms use another type of nucleic acid, ribonucleic acid (RNA). RNA differs from DNA in that it contains ribose instead of deoxyribose, and uracil instead of thymine. Uracil has the same base-pairing properties as thymine. RNA does not normally exist in double helix form, but RNA may form short stretches of double helix when the RNA strand folds back on itself locally. A DNA strand and a complementary RNA strand will form a double helix. RNA is made by transcription from one of the DNA strands in a double helix (Fig. 2). The DNA strand is used as a template in a process similar to DNA replication. The DNA double helix is opened locally, and an RNA strand complementary to the template is synthesized with help from the enzyme RNA polymerase, ribonucleoside triphosphates being used as activated building blocks. Special sequences of bases in front of the genes (promoters) are recognized by the polymerase, making it start at the right place and on the right DNA strand. At the other end of the genes, there are termination signals. In all cells, there are three main types of RNA: messenger RNA, ribosomal RNA, and transfer RNA. Messenger RNA (mRNA) carries the information specifying the amino acid sequences in the proteins. Ribosomal RNA (rRNA) molecules form part of the structure of the protein.synthesizing machinery of the cells, the ribosomes. Transfer RNA (tRNA) participates in protein synthesis (translation) by carrying amino acids to the ribosomes. The mRNA is of special importance. For each kind of protein chain the cell makes, there is a corresponding mRNA. The genetic coding
JENS G. HAUGE RNA DNA
i
!1II!I!!i NNA polymerase FIG. 2. RNA synthesis (transcription).
system that biochemists discovered between 1960 and 1965 showed t h a t organisms use groups of three nucleotides, triplets, to code for amino acids. It is, therefore, the sequence of triplets in mRNA that directly determines the sequence of amino acids in the corresponding polypeptide chain. The coding group could not have consisted of only two nucleotides. That would have given only 42 = 16 combinations, and there are 20 amino acids that need coding. But 43 = 64 gives more than enough. Nature has chosen this most economical solution and has done it in a way so t h a t 61 of these 64 triplets code for some amino acid. Almost all the amino acids, therefore, have more than one codon. When ribosomes directed by mRNA link together the amino acids, it is crucial that the ribosomes start at the correct position on the mRNA. Starting one nucleotide too early or late will give a completely different protein. A correct start is secured by having all protein synthesis start with the triplet AUG, which codes for methionine. This means that all proteins initially are synthesized with methionine as their first amino acid. In posttranslational processing, however, some proteins lose this initial methionine. As methionine also occurs internally in proteins, something more is needed for a correct start. In bacteria, the ribosomes recognize start AUGs by the presence of a short stretch of purines a short distance before the AUG. In animals and plants, the ribosomes choose the AUG that is closest to the 5' end of the mRNA. The coding triplets UAA, UAG, and UGA do not code for any amino acid, and these codons are used as termination signals. The amino acids do not by themselves have the ability to recognize their codons on mRNA. In addition, they must be activated in order to form peptide bonds with each other. Both of these problems are solved by each amino acid becoming bound to a tRNA with the help of a specific activating enzyme, using adenosine triphosphate (ATP). tRNA has in its structure an anticodon, a set of three nucleotides that form base pairs with the codon on mRNA for the corresponding amino acid. On the ribosome surface, the amino acids leave their tRNA carriers as
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
7
t h e y are linked together to form polypeptide chains. Briefly, the inform a t i o n flow is as illustrated here for the s t a r t of a hypothetical gene: 3' . .. TAC AAA AGT GAC TGT CCG CAT TAC . . . . . 5' DNA 5' . . . ATG TTT TCA CTG ACA GGC GTA ATG . . . . . 3' transcription 5 ' . . . AUG UUU UCA CUG ACA GGC GUA AUG . . . . .
3' mRNA
translation Met Phe Ser Leu Thr Gly Val Met . . . . .
protein
II. G e n e M a n i p u l a t i o n
A. BASIC FEATURES The core of the new gene m a n i p u l a t i o n techniques is t h a t one can join two pieces of DNA to form a r e c o m b i n a n t DNA molecule, which t h e n is multiplied, or cloned, in a suitable host organism to obtain sufficient a m o u n t s of the piece of DNA to study or use. One often uses as the one piece of DNA a plasmid, which is a ring-shaped, extrachromosomal DNA molecule t h a t occurs in m a n y bacteria and carries genes, for instance, for antibiotic resistance. Figure 3 illustrates molecular cloning of a piece of DNA with the help of a plasmid as vector. The plasmid vector is opened with the help of a special endonuclease, and the foreign DNA is joined in by the help of a n o t h e r enzyme. U n d e r proper conditions, some of the bacteria will t a k e in the r e c o m b i n a n t DNA molecule, which will multiply inside while the bacteria also grow and multiply. Most of the bacteria have not t a k e n up any plasmid. A selection is therefore necessary and is usually based on the plasmid c a r r y i n g an antibiotic resistance gene. One can t h e n isolate the plasmid and the foreign DNA piece from the bacteria in such a m o u n t s t h a t it is possible to s t u d y it u s i n g the electron microscope, d e t e r m i n e its base sequence, inject it in fertilized ova, or use it as a diagnostic probe, for example. W h e n one, as often is the case, does not have a homogeneous DNA p r e p a r a t i o n as s t a r t i n g m a t e r i a l for the cloning, but r a t h e r has a mixt u r e of m a n y types of DNA pieces, special selection techniques are required a f t e r w a r d to find the i n t e r e s t i n g clone or clones. P l a s m i d s are b u t one type of vector for DNA cloning. Viruses from bacteria or a n i m a l s also are used frequently.
JENS G. HAUGE
DNAto be cloned
O
Openedplasmid
1
O
RecombinantDNA
Into ~ hostcells ~romosome@ Growth ~ and selection
@o O) (o o O) (o o @ FIG. 3. Basic steps in cloning of DNA in a plasmid.
B. METHODS AND TOOLS
1. Nucleic Acid Hybridization Nucleic acid hybridization is fundamental in both basic and applied gene technology. This is the process by which sequence-specific DNA:DNA or RNA:DNA duplexes are formed from single-stranded nucleic acids. For most uses, one of the components in the hybridization is radioactively labeled, or labeled by other means, so that the hybrid can be identified. The labeled molecule is referred to as a probe. It is used to probe nucleic acid mixtures for the presence of sequences complementary to that of the probe (Fig. 4). The factors affecting the stability of the hybrid and the rate of the hybridization reaction are fairly well known (Marmur and Doty, 1961; Nygaard and Hall, 1964; Young and Anderson, 1985). The stability depends on the temperature, the ionic strength of the medium (high ionic strength reduces the repulsion between the negatively charged phosphate groups and thus stabilizes the duplex), the presence of formamide (which breaks hydrogen bonds), the G+C content of the probe (G- and C-rich probes are bound more tightly because there are three hydrogen bonds linking G to C, compared to two bonds between A and
M O L E C U L A R G E N E T I C S TO D I A G N O S I S , G E N E T H E R A P Y
Sample DNA -1I[I
Illllll
Labeled probe
Positive signal
[11 [ I T - V
Denature
Bind to support
1111111111/711/'t'/-]
7Z 1/7111 III
I IIIIf
l
FIG. 4. Hybridization.
T), and the length of the probe. The t e m p e r a t u r e at which the duplex is 50% denatured is called its T m. T m can be calculated from the formula: Tm -
81.5~
+ 16.6 log M + 0.41(%G+%C) - 500/n - 0.61(% formamide).
This formula assumes perfect complementarity between the hybridizing strands. With mismatches present, the T m is reduced by about 1~ per 1% mismatch. The factors affecting the stability also affect the rate of the hybridization. The rate is optimal about 25~ below T m. For a reaction in solution, the rate is proportional to the concentrations of the reacting species. When the DNA or RNA to be investigated is immobilized on a membrane, which is often the case, the reaction is slower by a factor of 7-10, but the general effects of the factors listed previously are the same. The reaction conditions are manipulated to give hybridizations at a desired level of stringency. High s t r i n g e n c y ~ t h a t is, high t e m p e r a t u r e or low ionic s t r e n g t h - - w i l l allow only the most stable hybrids to form, those having perfect complementarity. In certain situations, one may want to probe for a family of related DNA molecules with some sequence differences. Hybridization at a lower stringency will then be chosen. The same may be necessary when using a probe from one species to look for a homologous gene in another species. Several methods are used to label a probe with radioactivity (Arrand, 1985). In our experience, the random primer method (Feinberg and Vogelstein, 1983) is a convenient way of obtaining high specific activity, better t h a n the so-called nick translation. The probe DNA is denatured and both strands are used as a template for DNA synthesis, using short random oligonucleotides as primers, and one or more of the triphosphates being labeled with 32p.
10
JENS G. HAUGE
Nonradioactive labeling is also used because it has certain advantages for routine, diagnostic purposes. Attaching biotin to one of the bases (Leary et al., 1983) has received particular attention. The biotinylated probe is localized with streptavidin and a biotin-coupled enzyme, for which a substrate that gives a colored precipitate is used. Nonradioactive probes have the drawback of multistep detection procedures and a somewhat lower detection sensitivity t h a n do radioactive probes (Syvanen, 1986). 2. Restriction Endonucleases
Restriction endonucleases of type II, usually referred to as restriction enzymes, are a major tool in gene analysis and gene manipulation. These enzymes bind to particular sequences on DNA and cut both strands within this DNA site. The majority of the enzymes have sites of tetra-, penta-, hexa-, or heptanucleotides, which have an axis of rotational symmetry (Fig. 5a). Enzymes corresponding to over 150 different sites have been discovered so far. The names of the enzymes
4,
5' G G A T C C 3' 3' CCT~GG 5'
5' G A A T T C 3' II 3' C T T A A G 5'
A
BamHI
B
C
EcoRI
1'
4,
5' G G C C 3' 3' C C ~ G 5'
Hae III
1'
4,
5 ' G C G C 3' 3, CGttCG 5 ,
Hha I
1' $
5' C T C G A G 3' 3' GAGItC T C 5'
a
1'
Xho I
b
FIG. 5. (a) Specificities of some restriction enzymes. (From BIOCHEMISTRY, 2E by Lubert Stryer. Copyright 9 1975, 1981. Reprinted with permission ofW. H. Freeman and Company.) (b) Gel electrophoresis pattern showing SV40 DNA fragments produced with each of three restriction enzymes. (Courtesy of Dr. Jeffrey Sklar.)
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
11
have been constructed from names of the bacteria and the particular substrain from which they have been isolated. The arrows in Fig. 5a indicate where the DNA strands are cut by these 5 enzymes. Treatment of a DNA molecule with a restriction enzyme yields a reproducible set of fragments. An enzyme with a recognition site of 4 nucleotide pairs will cut relatively often, statistically once every 44 (i.e., 256) nucleotide pairs, while an enzyme with a hexanucleotide recognition site will cut less often, once every 46 (i.e., 4096) nucleotide pairs. Figure 5b shows the pattern of fragments into which three different restriction enzymes cut the SV40 virus chromosome. After the enzyme treatment, the fragments were separated by electrophoresis in an agarose gel and made visible by binding of ethidium bromide, which fluorescences in ultraviolet light. The sizes of the fragments can be found by r u n n i n g molecular weight m a r k e r s in the same gel. Short fragments move farther and long fragments a shorter distance. The fragments that the restriction enzymes produce will for most enzymes have single-stranded, complementary end sections, as the cuts are not made straight across (Fig. 5a). The fragments are said to have sticky ends. If a plasmid is opened with the same restriction enzyme that has been used to produce the DNA fragments to be cloned, this DNA will have sticky ends of a type that allows its insertion in the plasmid. Figure 6a shows what the situation would have been in Fig. 3 if the enzyme EcoRI had been used.
3. DNA Ligase To form a stable, recombinant DNA molecule from two molecules, the strands of these two molecules must be joined. This is effected by the enzyme DNA ligase. The enzyme requires for its action a 5'-phosphate and a 3'-OH group. The enzyme from phage T4 uses ATP as energy donor, while the Escherichia coli enzyme uses nicotinamideadenine dinucleotide (NAD). The ligation reaction is carried out at a low temperature, 4-15~ in order to increase the stickiness of the sticky ends.
4. Terminal Deoxynucleotidyltransferase Terminal deoxynucleotidyltransferase, or terminal transferase, is useful when the objective is to insert in a plasmid a DNA molecule that has not been produced with the help of a restriction enzyme, a frequently occurring situation. With this enzyme, tails of identical nucleotides can be attached to the 3'-end of DNA strands. The plasmid is opened with a suitable restriction enzyme and furnished with a tail of
12
JENS G. HAUGE G A-A-T-T-C C-T-T-A-A G
C-T-T-A-A G
-~G-G-G-G-G
C-C-C-C-C G-G-G-G
C-C-C-C-C
,,
FIG. 6. (a) Fitting a piece of DNA into an EcoRI site of a plasmid. (b) Fitting a piece of DNA into a plasmid after the plasmid is tailed with Gs and the insert with Cs, using terminal transferase.
about 20 G residues, while the DNA to be inserted is furnished with a C tail (Fig. 6b). 5. Adaptors Using an adaptor DNA containing a restriction site is a frequent alternative strategy when a DNA molecule is to be inserted in a plasmid opened with the corresponding restriction enzyme. A piece of DNA, 6 - 1 0 bp long, containing the desired restriction site is synthesized or purchased. The ends of the DNA to be inserted are first made blunt by filling in with DNA polymerase. All 5' ends are phosphorylated with the enzyme polynucleotide kinase, and the adaptor is attached to both ends of the insert DNA with DNA ligase. Treatment with the restriction enzyme then produces the proper sticky ends for insertion in the plasmid. This method illustrates well the role of both chemistry and enzymology in recombinant DNA technology. C. VECTORS
Through stepwise modifications of n a t u r a l E. coli plasmids, new plasmids that are safe and convenient for different types of gene manipulation have been constructed. One much used plasmid is pBR322 (Fig. 7a). This plasmid carries two antibiotic resistance genes, one for tetracycline and one for ampicillin. Within these two genes, there are several restriction sites for enzymes that have only these sites in the pBR322 molecule. If one uses the B a m H I site to insert a foreign DNA, the tetracycline resistance will be eliminated. It is then possible to select for bacteria that have taken up the plasmid by using ampicillin in the medium and then find those that have an insert in the B a m HI
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
13
Eco RI
V ' ~ PvulBam~ ~ " Pstl
Sail / |
pBR322
~Restriction
enzyme
Electrophoresis on agarosegel
dry paper towels\
l
cellulose nitrate filter
/
L containing denatured DNA
buffer
filter
paper
Hybridization with 32pcDNA ------
X
& Autoradioqraphy
FIG. 7. (a) The cloning vector pBR322 (4362 bp). Amp: gene for ampicillin resistance. Tet: gene for tetracycline resistance. (b) The Southern blot technique, x indicates the gene or sequence of interest. ("An Introduction to Recombinant DNA," Alan E. H. Emery, 1984. Reprinted by permission of John Wiley & Sons, Ltd.)
14
JENS G. HAUGE
seat by investigating which bacterial colonies have lost the tetracycline resistance. This is useful because there can be some regeneration of the original plasmid in the ligation step. Lambda virus DNA has also been modified so that it is suitable as a vector. Three of its five EcoRI sites have been removed. The DNA between the two remaining sites is not necessary for the infection process and can be exchanged with a foreign piece of DNA of 10-20 kilobase (kb) pairs. The recombinant DNA is packed in vitro using lambda coat and tail proteins. These phage particles enter E. coli bacteria very efficiently and reproduce as in a normal lytic infection. For other bacteria, for yeast, and for animal and plant cells there are other vectors. Shuttle vectors, vectors with specificity for two different hosts, have also been constructed. D. STUDIES OF GENE STRUCTURE Comprehensive solutions to a wide range of biological problems require knowledge of the structure of the gene and the regulatory sequence that govern the particular biological phenomenon.
1. cDNA Cloning For studies of genes of viruses and bacteria, cloning can start with fragments of their chromosomal DNA. With genes from animals and plants, the situation is different. Here the genome is so large and contains so many noncoding sections that one often must start on the basis on the mRNA molecules and carry out what is called a complementary, or copy, DNA (cDNA) cloning. Total RNA is isolated with chemical methods from a tissue of interest, and mRNA adsorbed on a column of poly(T)-cellulose, mRNA is bound due to the 3' tail of poly(A), which mRNA possesses in higher organisms, mRNA is eluted and used as a template for RNA-directed DNA polymerase, so-called reverse transcriptase, an enzyme occurring in retroviruses, where its role under the viral infection is to produce a double-stranded DNA corresponding to the viral RNA genome. In principle, the same reaction now takes place in vitro. A mixture of cDNAs is formed, corresponding to the mixture of mRNAs isolated. The mixture of cDNA molecules can now, after attachment of sticky ends, be inserted in a plasmid. If the bacteria to be transformed have been made competent by a special treatment, one can expect to obtain 105-107 transformed bacteria per microgram of recombinant plasmid. The clones from these transformations make up a cDNA library for that particular tissue.
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
15
In this library are cDNA clones corresponding to the different mRNAs in the tissue. Clones for the particular gene one is interested in can be found by various means. The simplest situation is when one knows something about the amino acid sequence of the protein for which the gene codes. Then one can synthesize a DNA probe of 17-19 bases, corresponding to a suitable hexapeptide. This probe is long enough that the search will be specific at high stringency. The probe preparation will have to be a mixture of DNA molecules, however, because most amino acids have more t h a n one codon. An imprint of the bacterial colonies is made on a nitrocellulose filter, the cells are opened and their DNA is denatured. If one then uses a radioactive probe, it will hybridize to a few colonies, with those colonies containing cDNA corresponding to the gene of interest. Plasmid DNA is purified and treated with the proper restriction enzyme, and the cDNA is isolated from a hybridizing band on an agarose gel after electrophoresis.
2. Gene Cloning A collection of clones t h a t includes all DNA in an organism is called a
genomic library. It is usually made in a lambda vector. DNA from the organism is subjected to incomplete digestion with a frequently cutting restriction enzyme, so t h a t fragments of about 20 kb are formed. Vector arms are ligated to both sides of these fragments, and the resulting DNA molecules packed into lambda proteins, as described in Section II,C. The lambda clones that are the result of the infection will, if one has close to a million different clones, represent all the DNA and genes of a mammal. If one now has a cDNA clone for the gene one is interested in, this cDNA can be radioactively labeled and used as a probe to find the lambda clones t h a t contain the gene or parts of it.
3. Splitting of Eukaryotic Genes One of the big surprises in molecular biology research was the discovery made in 1977 of cloned eukaryotic genes. The chicken gene for ovoalbumin was found to be 7700 bp long, while only 2000 bp would be required to code for the protein. Closer investigation of the cloned DNA with restriction enzymes and DNA sequencing revealed that the coding part of the gene was divided into eight pieces, exons, separated by larger pieces of noncoding DNA, introns.
4. Restriction Maps, Southern Blots, and DNA Sequencing Gene structure is characterized in broad outline by restriction mapping and in detail by establishing the DNA sequence. A restriction map identifies the positions of sites for a set of restriction enzymes. The
16
J E N S G. HAUGE
correct order for the fragments that are produced by a given enzyme is determined in several ways, for example, through comparison of DNA fragment lengths corresponding to different degrees of partial digestion with the enzyme. The approximate distances between the cut sites are obtained by including molecular weight m a r k e r s in the electrophoresis. Accurate distances require DNA sequencing. If one has a cDNA probe for a gene, it is possible to obtain some information on the structure of the gene without having cloned it, namely, by using the technique developed by Southern (1979) (Fig. 7b). After digestion with a restriction enzyme and separation by agarose electrophoresis, the DNA fragments are denatured in alkali and transferred to a nitrocellulose or nylon m e m b r a n e with a flow of buffer. A replica of the gel p a t t e r n of DNA is thereby produced on the membrane. When the m e m b r a n e is then treated with the radioactive probe, it will hybridize to those bands that have DNA sequences complementary to the probe. This will show up on an X-ray film. In this way, one obtains information about restriction sites in the gene region corresponding to the cDNA. DNA sequencing is carried out with a chemical method developed by Maxam and Gilbert (1977) or with an enzymatic method developed by Sanger et al., (1979). In the chemical method, one starts with a cloned DNA fragment of a few h u n d r e d bases, which has the 5' end radioactively labeled. From this, a spectrum of radioactive fragments is produced in which all sizes from the full length down to zero are represented. In principle, this is done in four reactions, one reaction each for fragments ending with a G, a C, a T, or an A. The fragments are separated by gel electrophoresis at a voltage of about 3000 V, and an autoradiograph of the gel is made (Fig. 8a). The sequence can be directly read from this autoradiograph, as indicated in Fig. 8a. The enzymatic method gives a similar end result. Here DNA polymerase synthesizes DNA in four reactions in the presence of small amounts of 2',3'dideoxynucleoside triphosphates. These cause chain termination when incorporated. DNA-sequencing work has yielded a wealth of information on gene structure and chromosome organization in virus, bacteria, and eukaryotic cells. Following are a few examples: Sequencing of the 5375 base long ~)X174 virus showed how this DNA could code for proteins of 2000 amino acids, by allowing an overlap between some of the genes. The transition between exons and introns was shown to have certain characteristic bases common to all genes. The promoter region in front of the gene, the site for a t t a c h m e n t of RNA polymerase, was also found to have a common base pattern. Light has been thrown on the struc-
17
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY A + G
G
~,~.,,.
C + T
C
~,.-,-,,,-
C G
1
A T T C
-.,---.--,"
---
G
--.,~---,-
- --
G
......
E
-
E
--
-- E
G
-
T T
_
T
m
T
-
T
E -
a
Probe
-E
b
FIG. 8. (a) Autoradiograph of a gel showing labelled fragments obtained in the Maxam and Gilbert DNA sequencing method. The reaction used for adenine affects both adenine and guanine, the reaction used for cytosine affects both cytosine and thymine. (From BIOCHEMISTRY, 2/E by Lubert Stryer. Copyright 9 1975, 1981. Reprinted with permission of W. H. Freeman and Company.) (b) An RFLP caused by presence or absence of a site for restriction enzyme E.
ture and the function of immunoglobulin genes, histocompatibility genes, and oncogenes. The sequencing technique has developed quickly. In 1982, the full sequence of lambda DNA with its 48,502 bp was published (Sanger et al., 1982). Now underway is a gigantic project, an international cooperation to sequence the total h u m a n genome, with its 3 x 109 bp, before the year 2005.
18
JENS G. HAUGE 5. The Polymerase Chain Reaction
An important use of sequence data is in enzymatic amplification of segments of DNA in the polymerase chain reaction (PCR). PCR amplification involves two oligonucleotide primers that flank the DNA segment to be amplified and repeated cycles of heat denaturation, annealing of the primers to the complementary sequences, and extension of the annealed primers with DNA polymerase. The primers hybridize to opposite strands of the target DNA and are oriented so that DNA is synthesized for the region between the primers. Because the extension products are also complementary and capable of binding the primers, each cycle doubles the amount of DNA resulting from the previous cycle. The basic idea of this method was described in 1971 by Kleppe et al., and later developed by Saiki et al. (1985). Its full potential appeared when Saiki et al. (1988) introduced the thermostable DNA polymerase from Thermus aquaticus (Taq), thus eliminating the need to add fresh polymerase during each cycle. This modification not only simplified the procedure, making it amenable to automation, but also substantially increased its specificity, sensitivity, yield, and length of targets that could be amplified. Amplification of 10 million times was demonstrated, segments up to 2000 bp were readily amplified, and longer segments were amplified with reduced yield. The PCR technique has found wide application (Erlich, 1989; Erlich et al., 1991a). Isolation of cloned segments from their vectors is simplified by PCR amplification of these, using vector-specific primers that flank the insertion site. RNA and cDNA can be amplified from the products of reverse transcription (Todd et al., 1987). DNA sequencing with the enzymatic method can be done directly on the PCR product without ligation into a sequencing vector. For the latter purpose, it is advantageous, however, to purify the sequencing template by incorporating biotin in one of the primers. After immobilization of the PCR product on streptavidin-coated magnetic beads, the unbiotinylated DNA strand is removed by alkali (Hultman et al., 1989). In diagnostic work and genetic mapping, PCR methods are used to a large extent, as will be apparent in later sections. For some applications, it is important to be aware that the Taq enzyme allows incorporation of one incorrect nucleotide for about 20,000 nucleotides incorporated. III. Studies of Metabolic and Gene Regulation Students of metabolic regulation have now new tools at their disposal through the advent of recombinant DNA technology. The area where
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
19
the impact has been the greatest is in the analysis of gene expression and its regulation by hormones. This is due to the availability of cDNA probes for a broad variety of genes of metabolic interest. These probes make possible quantitation of the level of specific messengers. A number of such studies, covering enzymes in glycolysis, gluconeogenesis, lipid metabolism and amino acid metabolism, are described by Goodridge and Hanson (1986).
A. Tissue CULTURE STUDIES Several of the corresponding genes have been isolated and their promoter-regulator regions are being identified and characterized. One example is the rat gene for cytosolic phosphoenolpyruvate carboxykinase (PEPCK), studied by Hod et al. (1986). The promoter-regulator region had been localized to within a 620-bp B a m H I / B g l I I fragment present at the 5' end of the gene. The hormonal control region was studied by linking the 621-bp fragment to the structural segment of the herpes simplex thymidine kinase (TK) gene in a plasmid vector. This construct was used to transfect cells of a TK-deficient rat hepatom cell line. The addition of dibutyrylcyclicAMP (cAMP) to the cells resulted in four- to sixfold induction of both TK activity and the level of its mRNA. To define the sequences in the PEPCK promoter that are necessary for this cAMP effect, a series of graded deletions was constructed. The transfected chimeric genes retained their responsiveness to cAMP as long as they included the sequence from - 6 1 to -108. Deletion of this 47-bp fragment completely eliminated cAMP inducibility. Further studies of this and other cAMP-regulated genes have shown that a palindromic DNA sequence, TGACGTCA, termed CRE (cAMP response element) is sufficient for the effect. Other hormones, such as glucocorticoids, androgens, mineralocorticoids, estrogen, thyroxine, and retinoic acid have specific response segments where the hormone-receptor complexes bind (Lucas and Granner, 1992). cAMP normally exerts its effect through activation of protein kinase A (PKA). This is the case also here. A CRE-binding protein (CREB) that is phosphorylated by PKA, and thereby activated, was discovered (Yamamoto et al., 1988). PEPCK is actually regulated by several hormones and the dietary state. Further transfection studies have uncovered eight protein-binding domains between - 4 6 0 and + 73 and identified proteins binding to these response elements (Park et al., 1993a). The interaction of these, partially tissue specific transcription factors, with the basic transcriptional complex is currently being studied.
20
JENS G. HAUGE B. REGULATION IN TRANSGENIC ANIMALS
While much may be learned from studies with cell cultures and how they react when recombinant DNA is introduced, this approach has its limitations. One fundamental set of regulatory problems concerns the development of an animal from the fertilized ovum. What sorts of mechanisms are responsible for tissue- and time-specific transcription of genes? Such questions are addressed through the use of transgenic organisms, particularly mice (Westphal, 1987; Grosveld and Kollias, 1992). This important technique was introduced by Gordon et al. (1980). The DNA to be investigated has mostly been injected into the male pronucleus of a fertilized ovum, and the ovum implanted. Some of the ova develop into mice with the extra DNA integrated in their chromosomes, usually in many copies arranged head to tail. The first successful experiments dealing with tissue-specific regulation were reported by Chada et al. (1985). They showed that a h u m a n ~-globin gene, in which the front end had been replaced by the corresponding mouse gene, together with 1200 bp flanking DNA, was expressed in transgenic mice exclusively in erythroid cells. They also showed that the hybrid ~-globin gene was expressed at the proper time during development (Magram et al., 1985). The 5'-flanking region must have been responsible for this specificity, along with corresponding tissue-specific regulatory proteins. Similar results have been found for a number of other genes and tissues (Palmiter and Brinster, 1986). A particularly striking example is the work on elastase I (Swift et al., 1984). Mice were made transgenic with the rat elastase I gene and its flanking regions. A marked tissue specificity was observed, rat elastase mRNA was 500,000 times more abundant in the pancreas than in the kidneys. Experiments with trimmed down 5'-flanking DNA showed that 205 bp was sufficient to uphold the pancreas-specific expression (Ornitz et al., 1985). When this regulatory region was joined to the h u m a n growth hormone gene, human growth hormone was found in the pancreas and not in other tissues. Immunofluorescence analysis, furthermore, demonstrated that the growth hormone was present in the acinar cells, not in the endocrine or connective tissue cells of the pancreas. The faithful tissue specificity observed with these kinds of constructs has been exploited in the study of oncogene expression (see Section IV, E). In one such study, H a n a h a n (1985) directed the expression of the SV40 oncogene to pancreatic ~-cells, using an insulin gene regulator sequence. This technique has also been used, for instance, to further under-
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
21
stand the complex and tissue-specific regulation of the PEPCK gene. McGrane et al. (1988) fused the promoter-regulatory sequence of the PEPCK gene to the bovine growth hormone gene as a reporter gene. The transgene was expressed only in liver and kidney and showed dietary and hormonal responsiveness. Further transgenic mice that carried mutations in specific regulatory domains were produced to establish their regulatory roles (Patel et al., 1994). Another method for producing transgenic animals uses embryonic stem (ES) cells. These are derived from an early embryo, cultivated, and infected with the retrovirus construct (Robertson et al., 1986) or transfected through the phosphate-DNA precipitation technique (Lovell-Badge et al., 1985). When they are reintroduced into the blastocyst, they contribute to the production of a chimeric animal. In the next generation, some pure transgenic animals appear. This approach has the advantage that somatic cell genetic techniques can be used to modify and to select cells with a desired potential.
IV. Diagnosis of Genetic Disease A. GENE CHANGES IN HEREDITARY DISEASE
There are many ways in which mutations can affect the production of a given gene product. Studies of hemoglobin synthesis in humans have found examples of most of these possibilities. Replacement of a single base with another base can have the effect that an amino acid is replaced by another amino acid. There are 189 such structural variants known for the h u m a n ~-globin chain (Little, 1981). Not all of these cause disease, however. The mechanism of polypeptide chain termination is the source of other disturbances. In individuals with hemoglobin Constant Spring, the stop codon TAA for ~-globin is replaced by CAA. In a type of ~~ (no production of ~-chains), the seventh codon, AAG, is replaced by TAG, terminating the chain prematurely. Single-base mutations may also affect transcription and RNA processing. In ~-thalassemia (low production of ~-chains), this is often observed. Some mutations in the promoter region reduce transcription markedly. Changes at the exon-intron junctions lead to errors in the removal of introns from the primary transcript, with unusable mRNA as the result. Mutations can also generate new, false exon-intron border sequences, again with defective mRNA as the result. Deletions and insertions are other common types of mutation. Deletions or insertions of 1, 2, or 4 bases are the cause of several
22
JENS G. HAUGE
~~ Such a mutation destroys the reading frame completely because the wrong set of triplets is read by the ribosomes. Deletions of n x 3 bases can also occur, but the reading frame is here not affected. Mutants with ~-chain shortening from 1 to 5 amino acids are known. Larger deletions do also occur; for hemoglobin genes, deletions from 0.6 kb to about 20 kb are known. Several studies have uncovered a similar mutational heterogeneity for many other genes. In addition, a new type of mutation has been discovered in myotonic dystrophy (DM) and six other diseases (Miawa, 1994). The DM gene contains the triplet repetition (GCT) n, with n varying somewhat in the population. For affected individuals, n is increased, and n appears to correlate with the severity of the disease. B. DETECTION OF GENE CHANGES WITH D N A PROBES
Recombinant DNA technology makes possible a direct approach to the studying and diagnosis of a genetic disease, instead of a study of the phenotypic expression of the disease. For animal owners and animal breeders, it will be valuable to be able to detect carriers of recessive disease genes in order to avoid propagating the disease gene or mating the animal with another carrier of the same defect. Furthermore, the homozygous state of a recessive gene, or the heterozygous state of a dominant gene, is not always expressed phenotypically. The disease may depend on an environmental factor for its expression or have a late onset. Here a DNA analysis could give an early diagnosis. Use of recombinant DNA methods in diagnosis and gene defect studies have made possible great advances in the h u m a n sector and is well underway in the veterinary sector also, as this volume demonstrates. Common to the methods is the requirement for having cloned or synthetic DNA probes for the disease gene or its neighborhood or sequence knowledge, permitting the use of PCR methods. 1. M u t a t i o n s that Create or Destroy a Restriction Site
The ideal situation exists when the mutation either creates a new restriction enzyme site or removes one that was present in the wild type. This is the case for sickle cell anemia in humans. In sickle cell anemia, the sixth amino acid in the S-chain of hemoglobin, glutamate, has been replaced by valin because an A in the corresponding codon has been replaced by T. The restriction enzyme DdeI for normal individuals cuts three places in the ~-globin gene, whenever the sequence CTNAG occurs (N can be any base). The mutation has the effect that the second of these sites is destroyed. Sickle cell patients thus will have
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
23
one longer DNA fragment hybridizing with a ~-globin probe, while normal persons will have two shorter fragments. This is revealed by the method of Southern blotting, described in Section II,D,4. A heterozygote would be revealed by all three bands being present. Each restriction digest requires 5 -lo ~g of patient DNA. This is usually obtained from the buffy coat of peripheral blood samples. The leukocytes present in 10-20 ml of human blood yield up to 200 ~g of DNA after treatment with proteinase K and extractions with phenol and chloroform. There are many other examples of diagnoses based on change in a restriction site; e.g., in the ras family of oncogenes. This is also the case for Ha-ras induced mutations in the rat (Zarbl et al., 1985). A more convenient version of this test amplifies the region around the mutation with PCR and subjects the PCR product to restriction enzyme treatment. Sufficient DNA is often present to allow direct detection of the restriction fragments after electrophoresis, using ethidium bromide and ultraviolet light. 2. Deletion a n d Insertion Mutations
When deletions or insertions of a few hundred or more base pairs are the cause of the disease, a diagnosis is often easier. These mutations will directly affect the length of the DNA fragment between two given restriction sites. Deletions occur in the human globin genes, the gene causing Duchenne type muscular dystrophy, the gene for coagulation factor VIII, and others. 3. Single-Base Replacement in a Known D N A Sequence
Base replacement mutations that do not affect any restriction site may still be diagnosed if one knows the DNA sequence around the mutation site. One can then synthesize a wild-type allele-specific oligonucleotide (ASO) of 18-20 bases, label it with radioactivity, and study how it hybridizes to DNA fragments on a membrane. Under stringent hybridization conditions, it is possible to detect a mismatch of only one base between the probe and the DNA investigated. It is prudent to also make a probe corresponding to the mutant and demonstrate its reduced hybridization to normal DNA. This method is used in diagnosis of c~i-antitrypsin deficiency (Kiddet al., 1984), Ha-ras mutations in rats (Zarbl, 1985), and several other mutations. The ASO method is more reliable and convenient when carried out on a PCR-amplified segment containing the mutation. The sensitivity thus reached makes possible the use of nonradioactive, enzyme-labeled ASO probes in dot blot hybridization (Scharf et al., 1991). The occur-
24
JENS G. HAUGE
rence of many different mutations in the same gene in the population can be analyzed using a reverse dot blot method. Instead of having numerous membranes with the PCR product applied, each to hybridize with a different ASO probe, a membrane that contains an immobilized array of ASO probes is hybridized to the PCR product (Erlich et al., 1991b).
4. Restriction Fragment Length Polymorphisms (RFLPs) Linked to the Gene of Interest Earlier, the most widely used and general diagnostic method was that of RFLPs linked to the gene that was being investigated. With this technique, one needs no knowledge of the structure of the gene. It is based on the fact that the heterozygosity at the level of DNA is very large. If the base sequences of two homologous chromosomes are compared, one finds in the human a difference for each 200 to 400 bp (Cooper and Schmidtke, 1984). The majority of these differences are neutral mutations. A few of them reside in the 3. base of coding triplets, most of them in introns and in the region between the genes. Some of these mutations will affect known restriction enzymes and will be within or close to the gene of interest, so that they can be observed as RFLPs with a probe from the same region (Fig. 8b). Such an RFLP can be used as a marker for the gene studied. The procedure is essentially the same as shown in Fig. 7b. Presence and absence of the restriction site will give two different restriction fragment lengths, moving different distances on the gel. a. Study of Linked RFLPs in Families. The use of RFLPs for diagnosis has one drawback: One does not arrive at an answer by investigating the DNA of a single individual, as was possible with the first three methods described in this section. The chromosome with the m u t a n t form of the gene will not always have the same restriction site allele, for instance, presence of the restriction site. That may have been the original constellation, but only a fraction of chromosomes with this restriction site need have the gene mutation. By recombination, the m u t a n t gene may also have landed on a chromosome with this restriction site absent. If the distance between the restriction site concerned and the gene studied is small, it is probable that the constellation will remain the same within a given family. The use of this method is illustrated in the following example from a family with ~-thalassemia. Both parents were healthy, but a child had been affected. During the three subsequent pregnancies, prenatal diagnosis was carried out. Using the restriction enzyme AvaII, Southern blotting, and a labeled ~-globin probe, the following band pattern was observed (Fig. 9).
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
+
- - - -
,,
25
,
FIG. 9. Prenatal diagnosis of ~-thalassemia. (Reproduced with permission, from the Annual Review of Medicine, Volume 37, 1986, by Annual Reviews, Inc.)
Both parents were RFLP + - , one chromosome having the restriction site and the other not. The affected child was + +. The + chromosome from both parents must have carried the disease gene. One could conclude from this that the fetuses were noncarrier, affected, and carrier, respectively. The probe cross-hybridizes with 5-globin sequences common to all. b. Usefulness of Having More than One RFLP. The example just described was a favorable situation. It would have been possible to diagnose only 50% of the offspring correctly if the thalassemia gene in one parent had been on an RFLP - chromosome. The situation might have improved if another RFLP had been close by. With two RFLPs, one would have four combinations, so-called haplotypes. This is equivalent to having four alleles for the m a r k e r locus, which gives a favorable degree of heterozygosity. If the distance between the gene and the RFLP site is large--e.g., some thousand k b ~ i t is important to have a second RFLP situated on the opposite side of the gene. A crossing over during meiosis between the first RFLP and the gene would then be detected because a double crossing over is very unlikely. Such a flanking RFLP is also useful in work attempting to localize and study the gene if the gene is not yet characterized. c. Searching for RFLPs. The search for RFLPs does not have to be a random trial and error testing of the large battery of commercially available enzymes. Cooper and Schmidtke (1984), Wijsman (1984), and Feder et al. (1985) describe rational approaches and give lists of relative efficiency for many enzymes. Enzymes containing the dinucleotide CG in their recognition sequences detect more variation of the base
26
JENS G. HAUGE
replacement type than other enzymes, probably because of mutations from methylated CGs to TGs. Larger probes are found more efficient than small probes. The screening should be limited, as suggested by Skolnick and White (1982), to a panel of about 10 unrelated individuals to avoid picking up rare polymorphisms, which have low degrees of heterozygosity and, therefore, are inferior in diagnostic work. When multiple restriction fragments are found, it must be shown that they represent alleles at the same locus. A probe might recognize nonallelic fragments due to sequence homology with a different chromosomal region. If the fragments are alleles, they will segregate in a Mendelian fashion. d. Minisatellites. Studies of gene structure have uncovered situations where a single restriction enzyme reveals a multiallele polymorphism, a polymorphism that does not affect sites for the enzyme. Between two fixed restriction sites, one finds a region consisting of tandem repetition of 15-100 bp units, and the number of these repeats varies considerably in the population. Such variable tandem repeat (VNTR) regions, also called minisatellites, have been found in humans close to the insulin gene (Bell et al., 1982), the s-related globin genes (Proudfoot et al., 1982), and the c-Ha-ras oncogene (Capon et al., 1983), but similar VNTR regions appear widely dispersed (Jeffreys et al., 1985a). The presence near a disease gene of a VNTR region is very useful, as the multiallelic nature of the corresponding RFLPs leads to high degrees of heterozygosity. A method for a systematic search for such RFLPs has been reported (Nakamura et al., 1987). e. Microsatellites. Weber and May (1989) reported the existence of a large set of highly polymorphic microsatellites, consisting of dinucleotide repeats, that could be typed using PCR. Such microsatellites have become the most important source of markers for h i g h resolution genetic maps. A h u m a n map with 814 microsatellites was presented in 1992 (Weissenbach et al., 1992). Two years later, the map included 2066 (AC) n markers, flanked by unique sequences for which PCR primers could be synthesized (Gyapay et al., 1994). The average distance between the markers was 2.9 cM. This means that there are now polymorphic microsatellite markers close enough to most disease genes for diagnostic purposes, forming at the same time starting points for work to identify the genes and their mutations. The latter task is being made easier by the parallel development of physical maps, consisting of contigs of yeast artificial chromosome (YAC) clones (Cohen et al., 1993), ordered using sequence-tagged sites (STSs) (Olson et al., 1989).
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
27
5. Disease Gene Identification Rapid progress has been made in the identification and characterization of disease genes in humans, a prerequisite for application of the direct methods of diagnosis, described in Sections IV,B,1-3. Many genes have been identified by functional cloning, using information about the protein product (sequence or antibodies). Mapping of the gene then followed cloning. Starting in 1986, another approach, positional cloning, came into use. Here, mapping to a definite location on a chromosome is the basis for the cloning effort. Some 20 disease genes have been identified in this way (Collins, 1992). As the h u m a n genome project produces high-resolution genetic maps and physical maps based on overlapping clones, positional cloning will be simplified. In other instances, a map position has been important in order to verify a gene identification, but the availability of a plausible candidate protein and its cDNA has made it possible to avoid exhaustive cloning in the specified chromosome region. This positional candidate approach will probably become increasingly common (Ballabio, 1993). In all, there are now about 400 h u m a n disease genes that can be diagnosed directly.
6. Genetic Screening Carrier screening programs have been attempted for some h u m a n genetic diseases even before DNA tests became available, notably sickle cell anemia, thalassemia, and Tay-Sach's disease. The thalassemia program in Cyprus, testing couples at marriage, has been very successful. Pilot programs testing pregnant women for the cystic fibrosis gene and the partner if the test was positive have taken place in Britain and the United States (Williamson, 1993). Although there are legitimate concerns that screening data may be used to coerce families to make reproductive decisions that meet economic or societal objectives rather than personal wishes, the availability of a screening service is generally welcomed. For h u m a n diseases, however, there is often the problem of genetic heterogeneity. The number of different cystic fibrosis mutations is now greater than 260. Although most of these are r a r e - - 11 mutations making up 91.5% of the mutations in Northwest England (Ferec et al., 1992)--a screening program for cystic fibrosis will clearly leave a small residual risk for unexpected development of the disease. For animal genetic diseases, there is less heterogeneity, making screening an effective and widespread tool (see Section IV, D).
28
JENS G. HAUGE C. D N A PROBES AND DISEASE SUSCEPTIBILITY
Recombinant DNA technology also has made it possible to begin to analyze polygenic diseases, i.e., diseases in which a combination of unfavorable genes may predispose to the development of disease. 1. H L A Genes
Genes determining susceptibility or resistance to a broad variety of diseases in humans have been localized to the HLA region, the major histocompatibility complex in humans (Vogel and Motulsky, 1986; Kostyu, 1991). Among these diseases are ankylosing spondylitis, celiac disease, dermatitis herpatiformis, and insulin-dependent diabetes mellitus (IDDM). The latter three are associated with the HLA serologic specifity DR3, IDDM also with DR4. Seventy-five percent of IDDM patients have HLA-DR4, while only 32% of controls do. This gives a DR4 individual a relative IDDM risk of 6.4. The risk is increased in DR3,4 heterozygotes (Knip et al., 1986). The association of HLA-DR genes with IDDM susceptibility is related to the fact that IDDM pathogenesis involves autoimmune phenomena (Nepom and Erlich, 1991). With recombinant DNA technology, it became possible to answer the following question: Are there DNA differences, detectable as restriction fragment length differences, between patients and controls of the same DR3 or DR4 specificity? This question was indeed answered when Owerbach et al. (1983), using a DQ~ probe, found a significantly decreased frequency of a B a m H 1 3.7-kb fragment among DR4 IDDM patients. These studies have been confirmed and extended by others, including Michelsen and Lernmark (1987). The full impact of DNA technology, however, came with PCR amplification of cDNA from the first exons of ~ and ~ chains, the exons containing most of the polymorphisms, followed by sequencing (Todd et al., 1987). This procedure has identified a large number of c~- and ~-chain alleles, denoted by a four-digit code. Typing of DNA sequence polymorphisms at HLA loci is conveniently done by PCR amplification of genomic DNA with appropriate primers, followed by probing with sequence-specific oligonucleotide probes (Erlich et al., 1991b). Such studies have shown that molecules of DQ(al*0301,~l*0302), associated with DR4, and DQ((~l*0501,~l*0201), associated with DR3 give a relative IDDM risk of 6-18 for Caucasians. In heterozygotes, two further DQ molecules may form by trans combination of the subunits (Thorsby and Rcnningen, 1993; Nepom and Erlich, 1991). Results of this nature have also been reported for malaria, where DRBI*1302-
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
29
DQBl*0501 gives some protection (Hill et al., 1991), and for papillomavirus-associated cervical carcinoma, where DRB1*1501DQBI*0602 gave increased risk (relative risk 4.8) (Apple et al., 1994). 2. I n s u l i n Gene 5 ' - V N T R Alleles
As mentioned in Section IV,B,4,c, there is a VNTR associated with the human insulin gene on chromosome 11. The variable region falls in three size classes (class I: average 570 bp, class II: average 1200 bp, and class III: average 2200 bp). A strikingly higher frequency of class I alleles was observed in Caucasians with IDDM (Bell et al., 1984). Lucassen et al. (1993) have, by sequencing the insulin gene locus from patients and controls, identified the region of association to IDDM to be 4.1 kb. Ten polymorphisms, including the VNTR, are in strong linkage disequilibrium with each other, showing a relative risk for IDDM of 3.8-4.5. Analysis of other ethnic groups may reveal exactly which polymorphism(s) confer susceptibility. It is possible that the VNTR length directly affects transcription. The insulin gene associated polymorphism and the HLA variation are not sufficient to account for the development of IDDM. Extensive mapping of the disease in the NOD (nonobese diabetic) mouse, using microsatellite markers, has revealed eight additional loci that make a contribution to this polygenic disease (Ghosh et al., 1993). One of these codes for interleukin-2, which has a different sequence in NOD mice and normal mice. Interleukin-2 has a role in autoimmunity. 3. Atherosclerosis
The most plausible link between genes and atherosclerosis is found in the structure and regulation of genes for lipoproteins, with enzymes taking part in their metabolism, and lipoprotein receptors. Genes for the apolipoproteins and the low-density lipoprotein (LDL) receptor in humans have been cloned and localized on chromosomes. One of every 500 persons is heterozygous for a mutant allele of the LDL receptor, leading to 2-3 times normal levels of cholesterol. Hobbs et al. (1992) have unraveled the mechanisms involved and characterized different types of mutations in the gene. Mapping of the defects in a population can start with a determination of mutation-associated RFLP haplotypes. This information is useful for diagnosis within families (Rcdningen et al., 1992). In order to establish direct gene tests, DNA from affected individuals was screened for mutations (Leren et al., 1993). LDL is bound to its receptor via apoB-100, and hypercholesterolemia may be caused by mutations in the apoB-100 gene as well (Innerarity
30
JENS G. HAUGE
et al., 1990). Mutations are known also in the apoA-I gene. These give low levels of high-density lipoprotein (HDL), accompanied by atherosclerosis. Type III hyperlipoproteinaemia patients have a particular allele for the apoE gene. With the current knowledge of DNA structure, oligonucleotide probes that differentiate between normal and mutant genes can be made, allowing screening of people at risk. D. APPLICATION TO DOMESTIC ANIMALS The application of recombinant DNA techniques to diagnosis in domestic animals is now quite widespread. The availability of DNA clones, or knowledge of primer sequences, from humans and small experimental animals has speeded this development. 1. Diagnosis of Disease Mutations The first disease to benefit from a DNA test was citrullinemia, which is widespread in Australian Friesian cattle. Three conditions were helpful for the establishment of a DNA test for the mutation in 1989: bovine cDNA libraries were available, sequenced cDNAs for the enzyme for humans and rats were available, and the protein is fairly small, only 412 amino acids. Normal bovine cDNA for the enzyme was isolated with a rat probe and sequenced by Dennis et al. (1989). In order to identify the mutation, total mRNA was prepared from the liver of an affected animal and cDNA synthesized. The product was amplified with PCR, using primers corresponding to sequences immediately before and after the coding region of the cDNA. These primers had been extended with B a m H I linkers, so that the amplification product could be ligated into the B a m H I site of the DNA sequencing vector. In clones from the mutant, a C to T transition in the first position of codon 86 was found. This generates a TGA termination codon, which leads to a truncated and inactive protein. The change in codon 86 leads to the disappearance of an AvaII restriction site. This was used by Dennis et al. to design a simple, PCRbased gene test on genomic DNA. A 194-bp DNA could be amplified from within the exon containing the mutation. A portion of the product was digested with AvaII, and both digested and undigested DNA analyzed by electrophoresis. The normal calf then gives two bands of 72 and 118 bp, the homozygously affected calf one band of 194 bp, and a carrier all three bands. Work with malignant hyperthermia (MH) in pigs had started earlier. We isolated clones for the glucosephosphate isomerase (GPI) gene, a
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
31
closely linked gene (Davies et al., 1987). With such DNA as probe, a multiallelic RFLP was detected and was confirmed to be tightly linked to the MH gene (Davies et al., 1988), so it could serve as a marker. The further search for the MH mutation was based on reports that calcium is more easily released from the sarcoplasmic reticulum of MH pigs. Meissner (1986) had shown that the alkaloid ryanodin interacts directly with the calcium release channel (CRC) of rabbits. Using tritiated ryanodin as a label, Lai et al. (1988) purified the channel. This work prepared the way for Takeshima et al. (1989) to clone the rabbit CRC cDNA, a 15,100-bp-long cDNA. With a small rabbit cDNA clone as a probe, we isolated a porcine cDNA clone and showed it to hybridize in situ to metaphase chromosomes in the same chromosomal band as GPI DNA hybridized (Harbitz et al., 1990). This confirmed CRC as a good candidate for the product of the MH locus. Sequencing CRC cDNA from a normal and an affected pig, Fujii et al. (1991) identified the mutation, a substitution of T for C at nucleotide 1843, changing an arginine to cysteine. The mutation was found to be the same in five Canadian pig breeds, in British Landrace (Otsu et al., 1991), and in Norwegian Landrace (Harbitz et al., 1992). The mutation destroys a H i n P site and creates a H g i A I site. Fujii et al. amplified a 74-bp piece from the exon containing the mutation and made the restriction site changes in the product the basis for their diagnosis. A more convenient and reliable method became possible after sequencing of genomic DNA flanking the exon. The last years have added three more diseases to the list: leukocyte adhesion deficiency (LAD) (Schuster et al., 1992) and uridine monophosphate synthase (DUMPS) deficiency (Schwenger et al., 1993) in cattle, and hyperkalemic periodic paralysis (HYPP) in quarter horses (Rudolph et al., 1992). In the study by Rudolph et al., the mutated region was found using single-strand conformation polymorphism (SSCP) analysis on PCR-amplified sections of the cDNA (Orita et al., 1989). This is a good general method for detecting existing two-allele polymorphic markers (Neibergs et al., 1993). 2. Diagnosis of Disease Susceptibility
Association between diseases and alleles of genes in the major histocompatibility complex (MHC) is also known for domestic animals. This is particularly well documented for chicken with Marek's disease, caused by a herpesvirus (Hanson et al., 1967; Hepkema et al., 1993). Resistance to bovine leukosis, caused by a retrovirus, also appears to have an MHC association (Lewin and Bernoco, 1986), and Mejdell et al. (1994) found an influence of the bovine MHC on resistance to mas-
32
JENS G. HAUGE
titis. Park et al. (1993b) found a highly significant association between the bovine MHC class I antigen BoLA-A8 and chronic posterior spinal p a r e s i s - - a form of ankylosing spondylitis--in Holstein bulls. The relative risk was 34.6. The special interest in developing RFLPs to establish haplotypes for the MHC genes in domestic animals must be viewed in this light. MHC RFLPs have been found with h u m a n probes in a large number of species (Andersson et al., 1986; Juul-Madsen et al., 1993). Additional studies are now being carried out based on sequencing of the MHC polymorphic exons. The data have been presented for cattle (Andersson et al., 1991), sheep (Fabb et al., 1993), horses (Szalai et al., 1993), and pigs (V~ige et al., 1994). E. GENE CHANGES IN CANCER
Cancer, like genetic diseases, is a result of changes in the hereditary material, but in cancer, the changes are mainly in somatic cells. Much new information on the phenomenon of neoplastic growth has been provided with the help of recombinant DNA techniques. The molecular structures of a long series of viral oncogenes and cellular proto-oncogenes, often coding for proteins participating in signal transduction from the cell surface to the nucleus, have been elucidated, and the mechanisms for the activation of proto-oncogenes to oncogenes to a large extent clarified (Land et al., 1983; Van de Woude et al., 1984). For the ras family of oncogenes, this activation takes place by a base replacement, and for others--e.g., myc, myb, and e t s - l - - b y translocation or amplification. Oncogene probes are available for such studies. Translocations, and sometimes mutations, result in restriction fragment size changes. Amplification leads to stronger bands of hybridizing DNA. Amplification, or stimulated transcription, leads to increased hybridization to mRNA. These developments have clinical relevance. DNA methods are becoming useful in tumor diagnosis, and they give new information on the stage of development of the tumor. Oncogenes are highly conserved in evolution, so that it is likely that human and murine oncogene probes can be used to investigate cancer development in domestic animals. Human probes for c-myc, myb, H E R - 2 / n e u , H-ras, K-ras, and N-ras thus were found by hybridize to DNA fragments from dogs (Hauge et al., 1988). While activated oncogenes act in a dominant manner, mutations in tumor suppressor genes are recessive. Their gene products inhibit cell division and thus balance the stimulatory properties of the proto-oncogene products. Loss of function of the suppressor gene on both chro-
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
33
mosomes leads to increased cell division. Because the first mutation is often t r a n s m i t t e d from a parent, while the second arises somatically, such tumors are found to be hereditary. Well-studied examples are the genes for retinoblastoma and p53, as well as several suppressor genes involved in colorectal cancer (Weinberg, 1991). Gene tests for these mutations are of obvious value for early diagnosis and treatment.
V. Therapy of Genetic Diseases Gene therapy may not be of practical value in regular veterinary medicine in the foreseeable future because of the cost involved. But work with gene therapy in animals will nonetheless be important, as preparation for such therapy in humans. What needs to be shown in animals is: (1) that the gene is integrated in the target cells, (2) that it is expressed on a suitable level, and (3) that the gene does not h a r m the cells or the animal. The ideal would be to exchange the m u t a t e d gene with a wild-type gene. Techniques for doing this by homologous recombination in ES cells have been developed (Smithies et al., 1985; Thomas and Capecchi, 1986). Gene therapy can be conceived at different levels: introduction of DNA in the fertilized ovum or early embryo; in cells or tissues that are t a k e n out, modified, and reimplanted; and in the whole animal, with the help of an infective vector. A. GERM LINE GENE THERAPY
Several successful experiments in germ line gene therapy have been reported. H a m m e r et al. (1985) described now injection of the rat growth hormone (GH) gene, linked to the mouse metallothionein promoter in fertilized ova from dwarf mice with low GH levels, yielded mice with normal growth. The immune response has been restored to a tripeptide in mice from a line with a defective MHC class I I E gene by injection of DNA containing this gene. ~-Thalassemia has also been successfully corrected by injection of ~-globin DNA in a murine model of h u m a n ~-thalassemia (Costantini et al., 1986). B. SOMATIC GENE THERAPY
H u m a n gene therapy has come of age (Miller, 1992; Friedman, 1993; Tolstoshev and Anderson, 1993). Protocols for t r e a t m e n t of 14 diseases have been approved (Wivel, 1993) based on results obtained with ani-
34
JENS G. HAUGE
mals. The first disease treated was adenosine deaminase (ADA) deficiency, causing immunodeficiency. A retrovirus vector was used to introduce the normal cDNA in lymphocytes. Among the diseases recently attacked are cystic fibrosis and hypercholesterolemia. Cystic fibrosis affects about 1 in 2000 Caucasians. Mouse models for cystic fibrosis have been generated by homologous recombination in embryonic stem cells, in one case by r e p l a c e m e n t of part of the cystic fibrosis transmembrane conductance regulator (CFTR) gene with a mutated gene section (Snouwaert et al., 1992), in the other case by insertion of a mutated gene in the host gene (Dorin et al., 1992). The first type is a null mutation, while the latter has a small degree of leakiness. Hyde et al. (1993) showed that CFTR cDNA could be delivered to the lungs of the replacement mutant by direct instillation of a cDNA-liposome cocktail, resulting in correction of the ion conductance defect. Other workers have used adenovirus as a vector. Unlike retrovirus, adenovirus is capable of infecting terminally differentiated cell types and is naturally drawn to airway epithelial cells. Good expression of h u m a n CFTR cDNA, carried on replication-deficient adenovirus, was observed after intratracheal introduction in cotton rats. Adenovirus is not integrated in host DNA. Studies have therefore been carried out on the safety and efficiency of repeated CFTR cDNA transfer in cotton rats and primates (Zabner et al., 1994). Another cDNA that has been carried on adenovirus into rat lung epithelium and expressed is cDNA for ~ l-antitrypsin (Rosenfeld et al., 1991). Adenovirus can also be used for muscle diseases. Vincent et al. (1993) have observed long-term correction of dystrophic degeneration in a mouse model for Duchenne type muscular dystrophy after intramuscular injection of adenovirus carrying a human dystrophin minigene. Familial hypercholesterolemia (FH) affects individuals heterozygous for LDL receptor mutations (see Section V,B,3) and very severely, affects those who are homozygous. A strain of rabbits genetically deftcient in LDL receptors was used to demonstrate the potential efficacy of ex vivo gene therapy. Part of the liver was resected and perfused with collagenase; the hepatocytes were transduced with LDL receptor retrovirus and infused in the rabbit liver, with good effect (Chowdhury et al., 1991). The experiment was followed up using dogs and baboons. The first results in humans were reported by Grossman et al. (1994). A 29-year-old woman, homozygous for the disease, underwent the same procedure, and her LDL/HDL ratio declined from 10-13 before treatment to 5-8 following gene therapy.
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
35
VI. Use of Recombinant DNA Methods to Improve Domestic Animals
A. MARKERS AND MAPS One important use of DNA methods for the improvement of domestic animals has been discussed already: marker-assisted selection against animals that carry a disease gene that has been characterized at the molecular level. On the whole, however, specific genetic disorders are less important than traits affecting growth, reproduction, disease resistance, and the quality of the end product. These traits are usually determined by several genes by quantitative trait loci (QTLs). In order to enable marker-assisted selection for a QTL, it must be fairly precisely localized on a chromosome. This requires genetic and physical maps of high resolution. Much effort has gone into establishing such maps the last years. The work is furthest advanced for pigs. At the First Pig Gene Mapping Workshop (PGM1) (Andersson et al., 1993), the number of mapped loci was 170. Sixty of these were anonymous DNA segments, mostly microsatellites, and the rest were polymorphic blood groups and protein polymorphisms. Rohrer et al. (1994) established a genetic linkage map with 376 microsatellite and 7 RFLP loci, using a twogeneration reference population. The average distance between adjacent markers was 5.5 cM. A genetic linkage map with somewhat lower resolution (11 cM average spacing), but with 60 reference markers for comparative mapping and 47 markers physically assigned by in situ hybridization, was reported by Ellegren et al. (1994), based on a cross between the European Wild Boar and a domestic breed (Large White). This material has been analyzed with respect to some quantitative traits, and evidence for QTLs with large effects on growth, length of the small intestine, and fat deposition was found on chromosome 4 (Andersson et a/.,1994). Comparative gene mapping for pig chromosome 4 could indicate candidates for these QTLs because clusters of neighboring genes tend to be preserved between species. An example of the comparative approach was the focusing on the long arm of human chromosome 19 in the search for the malignant hyperthermia (MH) locus. It was known that the GPI gene was located here, and that GPI and MH were closely linked in pigs. Another striking example is the observation of linkage for the human secreted phosphoprotein 1 (SSP1) locus to the sheep Booroola fecundity (FecB) gene, and the subsequent finding of linkage to human epidermal growth
36
JENS G. HAUGE
factor and complement I genes, near SSP1 in 4q in humans. The latter genes define a candidate region on sheep chromosome 6 (Montgomery et al., 1993). Good progress has also been achieved in genome mapping in cattle. Barendse et al. (1994) mapped 171 loci, with an average distance between m a r k e r s of 15 cM. Fifty-six loci represent DNA information from other species. An A m e r i c a n - S w i s s collaboration produced at the same time a cattle map with 313 polymorphic markers, with an average spacing of 8 cM (Bishop et al. 1994). An extensive comparison between cattle, cat, mouse, and h u m a n s is presented by O'Brien et al. (1993). A good foundation now appears to exist for the isolation of bovine QTLs, as well as for the study of m a m m a l i a n genome evolution. B. PARENTAGE IDENTIFICATION Errors may occur, for example, in m a r k i n g semen portions used in artificial insemination. Piglets may jump from one pen to another and be falsely m a r k e d when m a r k i n g takes place. It may, at times, be a temptation for a breeder to report a false father or mother. These problems can be addressed using core sequences for minisatellites, which hybridize with VNTRs of many genes, producing a DNA fingerprint (Jeffreys et al., 1985b). Because each band represents a minisatellite locus, the probability that two individuals would have the same pattern is nearly zero, except for monozygotic twins. A multilocus microsatellite probe, such as (GTG)5, yields a similar result, with better coverage of the genome (Morsch and Leibenguth, 1994). Alternatively, one can combine the PCR results for a set of highly polymorphic microsatellites (Marklund et al., 1994). C. TRANSGENIC ANIMALS
When improving livestock by traditional breeding, one is limited to an exploitation of the alleles that are already present in the animal population. The results in 1982 of the introduction of extra DNA into pronuclei of fertilized mouse eggs suggested the possibility of a quicker and larger improvement of production animals. Experiments with injection of GH in lactating cattle had demonstrated a significant boost in milk yields. It was natural to ask: Can the same result be obtained with extra gene copies for the hormone? Will pigs with extra GH genes reach slaughter weight sooner, and with lower feed consumption? The first report on transgenic livestock animals appeared in 1985 (Hammer et al.). In the following years, these experiments were ex-
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
37
tended. Clark et al., in a 1992 review, lists 12 reports on transgenic pigs, 9 on sheep, and 5 on cattle. The average frequency of success (transgenic animals/eggs transferred) was 0.59% with pigs and 0.74% with sheep, compared to 2-5% routinely obtained with mice. This puts a severe constraint on the adoption of this technology. For sheep and cattle, however, a promising development is that viable embryos can be obtained by in vitro maturation and fertilization of oocytes removed from slaughterhouse ovaries (Lu et al., 1987). 1. Growth Regulation
Most of the pig and sheep experiments involved GH genes from various species, and mostly with the mouse metallothionein promoter. Pursel et al. (1989) found concentrations of GH in expressing founder pigs in the range 14-4000 ng/ml for human GH. The variation could reflect the influence of chromosomal position for the inserted gene. To obtain a reliable assessment of the effect of the GH transgene on growth, Pursel et al. measured the average daily weight gain and feed efficiency for two generations of transgenic and control pigs. The combined transgenic progeny showed 11% faster weight gain than the controls (p = 0.001), and there was a 16-18% increased feed efficiency. Equally dramatic was the effect on subcutaneous fat accretion. Mean backfat thickness was reduced from 21 to 7.5 mm. These favorable traits were offset by considerable deleterious side effects, many of which could be the result of the lasting, high concentrations of GH. The most common clinical signs of disease were lethargy, lameness, uncoordinated gait, exopthalmos, and thickened skin. A number of pathologic changes were noted. What appears necessary is to achieve a tightly controlled GH gene expression, so that this can be confined to 1-2 months during the period of rapid growth. A promoter construct that is turned on by a substance added in the feed could solve the problem. 2. Disease Resistance
One route to specific disease resistance is the introduction of one or more immunoglobulin genes producing antibodies against a particular pathogen. Model experiments with mice show that this could work (Rusconi and K~hler, 1985). An interesting suggestion for reducing mastitis is to fuse the lysostaphin gene with regulatory elements from the betalactoglobulin (BLG) gene before microinjection in fertilized bovine eggs. The transgenic cow would then produce lysostaphin in its udder. Lysostaphin
38
JENS G. HAUGE
hydrolyzes the cell wall of Staphylococcus aureus. The idea is under testing in mice (Clark et al., 1992). In chickens, Salter and Crittenden (1989) have described a transgenic line carrying a defective ALV retrovirus genome that expresses only the envelope proteins of the virus. This leads to resistance because the viral protein produced competes on the virus receptor binding sites on the cell surface. This is similar to a successful technique for making plants virus resistant. 3. Wool Production
In sheep, a major limitation in wool production is the availability of cysteine. Two Australian groups have cloned the necessary cysteine synthetase genes from bacteria and introduced them into sheep. At least one transgenic animal was generated (Rogers, 1990). 4. M i l k Modification
Various ways of modifying milk protein composition can be envisaged (Clark et al., 1992). Caseins could be altered in their phosphorylation sites and thereby in micelle properties. Reduction in the amount of BLG would reduce allergies and inhibit the synthesis of lactose, a disaccharide not tolerated by 90% of the world population. It would also seem useful to try to introduce in cows the gene for human lactoferrin, an antimicrobial agent, and iron transporter in order to make cow milk more suitable for human infants. 5. P h a r m a c e u t i c a l Proteins
The area of pharmaceutical proteins will probably never be a large part of commercial animal husbandry and breeding, but it deserves mention. Clark et al. (1992) have used regulatory sequences from sheep BLG to target expression of human factor IX and a~-antitrypsin (AAT) to the mammary gland. One sheep produced AAT at 30 g/liter, making up 50% of the milk proteins (Wright et al., 1991). The lack of AAT is one of the most common human genetic disorders. It leads to a life-threatening emphyzema, requiring repeated administrations of intact enzyme (200 g per patient per year). The source has hitherto been human blood plasma, with the limitations and dangers this entails. Milk is not the only possible vehicle for medically important proteins from transgenic animals. Swanson et al. (1992) have succeeded in making pigs transgenic for human globin genes, expressed in pig erythrocytes. Human hemoglobin could be separated from pig hemoglobin by ion exchange chromatography.
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
39
6. Use of Embryonic Stem Cells A major limitation in the production of transgenic animals by microinjection of DNA in fertilized ova is that there is no control over where in the genome the extra DNA finds its place. Sometimes it will disrupt a gene, introducing defects in development or function of the animal, it may land in a region where it is poorly expressed. Work with mice has shown that for ES cells in culture, one can achieve a site-specific integration of the new DNA with the help of homologous recombination. Cells that are shown to have incorporated the gene properly are then introduced in the blastocyst. The result is chimeric animals, but in the next generation animals with the new gene in their germ cells can be found. For domestic livestock, the ES technique has met with some problems. However, Notarianni et al. (1990) described the derivation of apparently pluripotent cell lines from porcine and ovine blastocysts. In cattle and sheep (but not in mice), young have been born following transfer of nuclei from cells of the inner cell mass of blastocysts to enucleated oocytes (Marx, 1988). Therefore, if gene-modified ES cells can be established from these species, it may be possible to transfer nuclei from them and avoid the chimeric stage.
VII. Legal and Ethical Aspects of Patenting Biotechnology The issue of patenting in our context refers most directly to patenting of transgenic animals (TGAs). When considering the ethics of such patents, it is necessary first to discuss the ethics of generating TGAs. If the very production of TGAs should be ethically problematic, patenting, being a stimulus for such procedures, would have to be considered even more problematic. One is faced with the question whether such manipulation is in line with the respect one should have for an animal's worth and integrity. We have here a value conflict, a conflict between the worth of animals and the worth of humans. In my view, the good consequences for humans that result, or may result, from these gene manipulations are so great that these procedures are defendable if they proceed according to the prescriptions contained in animal welfare laws. I have, in preceding sections, outlined the importance of the experiments with transgenic mice for our understanding of fundamental biological mechanisms in mammals and, thereby for our understanding of disease processes both in animals and in humans, and for the possibilities of preventing and curing disease. Improvement of domestic animal breeds must also be an important goal in a world that needs to
40
JENS G. HAUGE
produce food for a quickly increasing population. The transgenic improvement may be directed toward better animal health, a healthier composition of meat and milk, and a better feed efficiency. With regard to the third category of TGAs, it appears possible that a relatively small number of sheep, goats, or cattle will supply the world requirements for important proteins, secreted in their milk, proteins that perhaps otherwise would have to be isolated from h u m a n blood, with the limitations and contamination dangers that entails. Animal genomes are manipulated, then, based on h u m a n needs. This is not a behavior that started in 1980, with the first TGA. It started with domestication several thousand years ago. Our presentday swine do not look much like the wild swine, most of our dog races not much like the wolf. But now, as before, we are manipulating a fenced-off, limited part of nature. Wolves and wild swine continue to evolve in freedom. Of course, one may not uninhibitedly manipulate the genome of an animal. The Norwegian animal protection law states: "Animals shall be treated well and consideration taken for their instincts and n a t u r a l needs." It cannot be claimed that classical breeding has always followed this aim. Some animals have been given new burdens. Is there a danger that the use of DNA techniques to produce better production animals will entail additional burdens? Possibilities for this are present when one interferes with a balanced system. There are indeed, as mentioned, reported examples of this for transgenic pigs. During a research period, this may be acceptable. But before a TGA is put in regular production, a thorough analysis of its anatomy, physiology, and behavior must demonstrate that it does not suffer from some new stress factor. In the Norwegian gene technology law of 1993, the animal protection law was revised to cover these concerns. Work with experimental animals like mice and rats has a special legal status. The experiments must be approved and certain procedures followed to minimize pain. Does patenting of TGAs introduce additional ethical problems? The patent system exists to stimulate inventors to come up with new solutions to practical problems. If the patenting of transgenic sheep that produces ~l-antitrypsin stimulates the generation of such sheep, it would be unethical to prohibit such a patent. A number of objections have nevertheless been raised against TGA patents. One is the view that it will reduce the access of developing nations to transgenic breeds, when such become available. An answer to this is that without patentability, the transgenic production animal may never be developed. An existing patented TGA breed may, furthermore, find its way
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
41
to a developing nation if licensing fees are paid by a foreign aid organization and if farmers' privilege for use of offspring is granted. Animal rights protagonists have claimed that patenting will increase the suffering of experimental animals (Raines, 1991). But the patent system is not created to be a substitute for regulatory authority. If existing regulations are inadequate to protect animal welfare, they should be changed. Others argue that patenting of life reflects a domineering and materialistic attitude toward living creatures. Is it more materialistic than our domestication of animals? In our culture, animals are bought and sold, owned and used, slaughtered and eaten. Whether they are patented or not, they should be handled with care and compassion. REFERENCES Andersson, L., Archibald, A. L., Gellin, J., and Schook, L. B. (1993). 1st pig gene mapping workshop (PGM1). Anim. Genet. 24, 205-216. Andersson, L., Boehme, J., Peterson, P. A., and Rask, L. (1986). Genomic hybridization of bovine class II major histocompatibility genes: 2. Polymorphism of DR genes and linkage disequilibrium in the DQ-DR region. Anim. Gent. 17, 295-304. Andersson, L., Haley, C. S., Ellegren, H., Knott, S. A., Johansson, M., Andersson, K., Andersson-Eklund, L., Edfors-Lilja, I., Fredholm, M., Hansson, I., H~ikansson, J., and LundstrCm, K. (1994). Genetic mapping of quantitative trait loci for growth and fatness in pigs. Science 263, 1771-1774. Andersson, L., Sigurdardottir, S., Borsch, C., and Gustafsson, K. (1991). Evolution of MHC polymorphism: Extensive sharing of polymorphic sequence motifs between human and bovine DRB alleles. Immunogenetics 33, 188-193. Apple, R. J., Erlich, H. A., Klitz, W., Manos, M. M., Becker, T. M., and Wheeler, C. M. (1994). HLA DR-DQ associations with cervical carcinoma show papillomavirus-type specificity. Nature Genet. 6, 157-162. Arrand, J. E. (1985). In "Nucleic Acid Hybridization--A Practical Approach" (B. D. Hames and S. J. Higgins, eds.), pp. 17-45. IRL Press, Washington. Ballabio, A. (1993). The rise and fall of positional cloning? Nature Genet. 3, 277-279. Barendse, W., Armitage, S. M., Kossarek, L. M., Shalom, A., Kirkpatrick, B. W., Ryan, A. M., Clayton, D., Li, L., Neibergs, H. L., Zhang, N., Grosse, W. M., Weiss, J., Creighton, P., McCarthy, F., Ron, M., Teale, A. J., Fries, R., McGraw, R. A., Moore, S. S., Georges, M., Soller, M., Womack, J. E., and Hetzel, D. J. S. (1994). A genetic linkage map of the bovine genome. Nature Genet. 6, 227-235. Bell, G. I., Horita, S., and Karam, J. H. (1984). A polymorphic locus near the human insulin gene is associated with insulin-dependent diabetes mellitus. Diabetes 33, 176183. Bell, G. I., Selby, M. J., and Rutter, W. J. (1982). The highly polymorphic region near the human insulin gene is composed of simple tandemly repeating sequences. Nature 295, 31-35. Bishop, M. D., Kappes, S. M., Keele, J. W., Stone, R. T., Sunden, S. L. F., Hawkins, G. A., Toldo, S. S., Fries, R., Grosz, M. D., Yoo, J., and Beattie, C. W. (1994). A genetic linkage map for cattle. Genetics 136, 619-639.
42
JENS G. HAUGE
Capon, D. J., Chen, E. Y., Levinson, A. D., Seeburg, P. H., and Goeddel, D. V. (1983). Complete nucleotide sequences of the T24 human bladder carcinoma oncogene and its normal homologue. Nature 302, 33-37. Chada, K., Magram, J., Raphael, K., Radice, G., Lacy, E., and Costantini, F. (1985). Specific expression of a foreign ~-globulin gene in erythroid cells of transgenic mice. Nature 314, 377-380. Chowdhury, J. R., Grossman, M., Gupta, S., Chowdhury, N. R., Baker, J. R., Jr., and Wilson, J. M. (1991). Long term improvement of hypercholesterolemia after ex vivo gene therapy in LDLR deficient rabbits. Science 254, 1802-1805. Clark, A. J., Simons, J. P., and Wilmut, I. (1992). Germ line manipulation: Applications in agriculture and biotechnology. In '"Fransgenic Animals" (F. Grosveld and G. Kollias, eds.), pp. 249-270. Academic Press, New York. Cohen, D., Chumakov, I., and Weissenbach, J. (1993). A first-generation physical map of the human genome. Nature 366, 698-701. Collins, F. S. (1992). Positional cloning: Let's not call it reverse anymore. Nature Genet. 1, 3-6. Cooper, D. N., and Schmidtke, J. (1984). DNA restriction fragment length polymorphisms and heterozygosity in the human genome. Hum. Genet. 66, 11-16. Costantini, F., Chada, K., and Magram, J. (1986). Correction of murine ~-thalassemia by gene transfer into the germ line. Science 233, 1192-1194. Davies, W., Harbitz, I., Fries, R., Stranzinger, G., Hauge, J. G. (1988). Porcine malignant hyperthermia carrier detection and chromosomal assignment using a linked probe. Anim. Genet. 19, 203-212. Davies, W., Harbitz, I., Hauge, J. G. (1987). A partial cDNA clone for porcine glucosephosphate isomerase: Isolation, characterization and use in detection of restriction fragment length polymorphisms. Anita. Genet. 18, 233-240. Dennis, J. A., Healy, P. J., Beaudet, A. L., and O'Brien, W. E. (1989). Molecular definition of bovine argininosuccinate synthetase deficiency. Proc. Natl. Acad. Sci. USA 86, 7947-7951. Dorin, J. R., Dickinson, P., Alton, E. W. F. W., Smith, S. N., Geddes, D. M., Stevenson, B. J., Kimber, W. L., Fleming, S., Clarke, A. R., Hooper, M. L., Anderson, L., Beddington, R. S. P., and Porteous, D. J. (1992). Cystic fibrosis in the mouse by targeted insertional mutagenesis. Nature 359, 211-215. Ellegren, H., Chowdhary, B. P., Johansson, M., Marklund, L., Fredholm, M., Gustavsson, I., and Andersson, L. (1994). A primary linkage map of the porcine genome reveals a low rate of genetic recombination in the pig. Genetics, 137, 1089-1100. Emery, A. E. H. (1984). "An Introduction to Recombinant DNA." Wiley, Chichester. Erlich, H. A. (ed.) (1989). "PCR Technology. Principles and Applications for DNA Amplification." Stockton Press, New York. Erlich, H., Bugwan, R., Begovich, A. B., Scharf, S., Griffith, R., Saiki, R., Higuchi, R., and Walsh, P. S. (1991b). HLA-DR, DQ and DP typing using PCR amplification and immobilized probes. Eur. J. Immunogen. 18, 33-35. Erlich, H. A., Gelfand, D., and Sninsky, J. J. (1991a). Recent advances in the polymerase chain reaction. Science 252, 1643-1651. Fabb, S. A., Maddox, J. F., Gogolin-Ewens, K. J., Baker, L., Wu, M-J., and Brandon, M. R. (1993). Isolation, characterization and evolution of ovine major histocompatibility complex class I DRA and DQA genes. Anim. Genet. 24, 249-255. Feder, J., Yen, L., Wijsman, E., Wang, L., Wilkins, L., Schroder, J., Spurr, N., Cann, H., Blumenberg, M., and Cavalli-Sforza, L. L. (1985). Systematic approach for detecting
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
43
high-frequency restriction fragment length polymorphisms using large genomic probes. Am. J. Hum. Genet. 37, 635-649. Feinberg, A. P., and Vogelstein, B. A. (1983). A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochern. 132, 6-113. Ferec, C., Audrezet, M. P., Mercier, B., Guillermit, H., Moullier, P., Quere, I., and Verlingue, C. (1992). Detection of over 98% cystic fibrosis mutations in a Celtic population. Nature Genet. 1, 188-191. Francomano, C. A., and Kazazian, H. H. (1986). DNA analysis in genetic disorders. Annu. Rev. Med. 37, 377-398. Freifelder, D. (1983). "Molecular Biology." Jones and Bartlett, Boston. Friedman, T. (1993). Milestones and events in the early development of human gene therapy. In "Molecular Genetic Medicine" (T. Friedman, ed.), pp. 1-32. Academic Press, New York. Fujii, J., Otsu, K., Zorzato, F., De Leon, S., Khanna, V. K., Weiler, J. E., O'Brien, P. J., and MacLennan, D. H. (1991). Identification of a mutation in porcine ryanodine receptor associated with malignant hyperthermia. Science 253, 448-451. Ghosh, S., Palmer, S. M., Rodrigues, N. R., Cordell, H. J., Hearne, C. M., Cornall, R. J., Prins, J.-B., McShane, P., Lathrop, G. M., Peterson, L. B., Wicker, L. S., and Todd, J. A. (1993). Polygenic control of autoimmune diabetes in nonobese mice. Nature Genet. 4, 404-409. Goodridge, A. G., and Hanson, R. W., eds. (1986). "Metabolic Regulation: Application of Recombinant DNA Techniques." New York Acad. Sci., Vol. 478. Gordon, J. W., Scangos, G. A., Plotkin, D. J., Barbos, J. A., and Ruddle, F. H. (1980). Genetic transformation of mouse embryos by microinjection of purified DNA. Proc. Natl. Acad. Sci. USA 77, 7380-7384. Grossman, M., Raper, S. E., Kozarsky, S. E., Stein, E. A., Engelhardt, J. F., Muller, D., Lupien, P. J., and Wilson, J. M. (1994). Successful ex vivo gene therapy directed to liver in a patient with familial hypercholesterolemia. Nature Genet. 6, 335-341. Grosveld, F., and Kollias, G., eds. (1992). '~ransgenic Animals." Academic Press, New York. Gyapay, G., Morisette, J., Vignal, A., Dib, C., Fizames, C., Milasseau, P., Marc, S., Bernardi, G., Lathrop, M., and Weissenbach, J. (1994). The 1993-94 Genethon human genetic linkage map. Nature Genet. 7, 246-249. Hammer, R. E., Pursel, V. G., Rexroad, C. E., Wall, R. J., Bolt, D. J., Ebert, K. M., Palmiter, R. D., and Brinster, R. L. (1985). Production of transgenic rabbits, sheep and pigs by microinjection. Nature 315, 680-683. Hanahan, D. (1985). Heritable formation of pancreatic [~-cell tumours in transgenic mice expressing recombinant insulin/simian virus 40 oncogenes. Nature 315, 115-122. Hansel, W., and Weir, B. J., eds. (1990). "Genetic Engineering of Animals." Journals of Reproduction and Fertility. Hanson, M. P., van Zandt, J. N., and Law, G. R. (1967). Differences in susceptibility to Marek's disease in chickens carrying two different B locus blood group alleles. Poult. Sci. 46, 1268. Harbitz, I., Chowdhary, B., Thomsen, P. D., Davies, W., Kaufmann, U., Kran, S., Gustavsson, I., Christensen, K., and Hauge, J. G. (1990). Assignment of the porcine calcium release channel gene, a candidate for the malignant hyperthermia locus, to the 6pll-q21 segment of chromosome 6. Genomics 8, 243-248. Harbitz, I., Kristensen, T., Bosnes, M., Kran, S., and Davies, W. (1992). DNA sequence of the skeletal muscle calcium release channel cDNA and verification of the Arg615Cys
44
JENS G. HAUGE
mutation, associated with porcine malignant hyperthermia, in Norwegian Landrace pigs. Anim. Genet. 23, 395-402. Hauge, J. G. (1988). unpublished data. Hepkema, B. G., Blankert, J. J., Albers, G. A. A., Tilanus, M. G. J., Egberts, E., van der Zijpp, A. J., and Hensen, E. J. (1993). Mapping of susceptibility to Marek's disease within the major histocompatibility (B) complex by refined typing of White Leghorn chickens. Anim. Genet. 24, 283-287. Hill, A. V. S., Alsopp, C. E. M., Kwiatkowski, D., Anstey, N. M., Twumasi, P., Rowe, P. A., Bennett, S., Brewster, D., McMichael, A. J., and Greenwood, B. M. (1991). Common West African HLA antigens are associated with protection from severe malaria. Nature 352, 595-600. Hobbs, H. H., Brown, M. S., and Goldstein, J. L. (1992). Molecular genetics of the LDL receptor gene in familial hypercholsterolemia. Hum. Mutat. 1, 445-466. Hod, Y., Cook, J. S., Weldon, S. L., Short, J. M., Wynshaw-Boris, A., and Hanson, R. W. (1986). In "Metabolic Regulation: Application of Recombinant DNA Techniques" (A. G. Goodridge and R. W. Hanson, eds.), pp. 31-45. New York Acad. Sci. Vol. 478. Hultman, T., St~ihl, S., Hornes E., and Uhlen, M. (1989). Direct solid state sequencing of genomic and plasmid DNA using magnetic beads as solid support. Nucleic Acids Res. 17, 4937-4946. Hyde, S. C., Gill, D. R., Higgins, C. F., Trezise, Mac Vinish, L. J., Cuthbert, A. W., Ratcliff, R., Evans, M. J., and Colledge, W. H. (1993). Correction of the ion transport defect in cystic fibrosis transgenic mice by gene therapy. Nature 362, 250-255. Innerarity, T. L., Mahley, R. W., Weisgraber, K. H., Bersot, T. P., Krauss, R. M., Vega, G. I., Grundy, S. M., Friedl, W., Davignon, J., and McCarthy, B. J. (1990). Familial defective apolipoprotein B-100: A mutation of apolipoprotein B that causes hypercholesterolemia. J. Lip. Res. 8, 1337-1349. Jeffreys, A. J., Wilson, V., and Thein, S. L. (1985a). Hypervariable 'minisatellite' regions in human DNA. Nature 316, 76-79. Jeffreys, A. J., Wilson, V., and Thein, S. L. (1985b). Individual-specific 'fingerprints' of human DNA. Nature 314, 67-73. Juul-Madsen, H. R., Hedemand, J. E., Salomonsen, J., and Simonsen, M. (1993). Restriction fragment length polymorphism analysis of the chicken B-F and B-L genes and their association with serologically defined B haplotypes. Anita. Genet. 24, 243-247. Kidd, V. J., Golbus, M. S., Wallace, R. B., Itakura, K., and Woo, S. L. C. (1984). Prenatal diagnosis of (~l-antitrypsin deficiency by direct analysis of the mutation site in the gene. N. Engl. J. Med. 310, 639-642. Kleppe, K., Ohtsuka, E., Kleppe R., Molineux, I., and Khorana, H. G. (1971). Studies on polynucleotides XCVL. Repair replication of short synthetic DNAs as catalyzed by DNA polymerases. J. Mol. Biol. 56, 341-361. Knip, M., Ilonen, J., Mustonen, A., and ~kerblom, H. K. (1986). Evidence of an accelerated ~-cell destruction in HLA-Dw3/Dw4 heterozygous children with type 1 (insulindependent) diabetes. Diabetologia 29, 347-351. Kostyu, D. D. (1991). The HLA gene complex and genetic susceptibility to disease. Current Opinion Genet. Devel. 1, 40-47. Lai, F. A., Erickson, H. P., Rousseau, E., Liu, Q-Y., and Meissner, G. (1988). Purification and reconstitution of the calcium release channel from skeletal muscle. Nature 331, 315-319. Land, H., Parada, L. F., and Weinberg, R. A. (1983). Cellular oncogenes and multistep oncogenesis. Science 222, 771-778. Leary, J. J., Brigati, D. J., and Ward, D. C. (1983). Rapid and sensitive colorimetric
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
45
method for visualizing biotin-labeled DNA probes hybridized to DNA or RNA on nitrocellulose: Bioblots. Proc. Natl. Acad. Sci. USA 80, 4045-4049. Leren, T. P., Solberg, K., RCdningen, O. K., R~sby, O., Tonstad, S., Ose, L., and Berg, K. (1993). Screening for point mutations of exon 10 of the LDL receptor gene by analysis of single-strand conformation polymorphisms: Detection of a non-sense mutationm FH(469-stop). Hum. Genet. 92, 6-10. Lewin, H. A., and Bernoco, D. (1986). Evidence for BoLA-linked resistance and susceptibility to subclinical progression of bovine leukemia virus infection. Anita. Genet. 17, 197-207. Little, P. F. R. (1981). In "Genetic Engineering 1" R. Williamson, ed.), pp. 61-102. Academic Press, New York. Lovell-Badge, R. H., Bygrave, A. E., Bradley, A., Robertson, E., Evans, M. J., and Cheah, K. S. E. (1985). Transformation of embryonic stem cells with the human type-II collagen gene and its expression in chimeric mice. Cold Spring Harb. Syrup. Quant. Biol. 50, 707-711. Lu, K. H., Gordon, I., Gallagher, M., and McGovern, H. (1987). Pregnancy established in cattle by transfer of embryos derived from in vitro fertilization of oocytes matured in vitro. Vet. Rec. 121, 259-260. Lucas, P. C., and Granner, D. K. (1992). Hormone response domains in gene transcription. Annu. Rev. Biochem. 61, 1131-1173. Lucassen, A. M., Julier, C., Beressi, J. P., Boitard, C., Froguel, P., Lathrop, M., and Bell, J. I. (1993). Susceptibility to insulin dependent diabetes mellitus maps to a 4.1 kb segment of DNA spanning the insulin gene and associated VNTR. Nat Genet. 4, 305-310. Magram, J., Chada, K., and Costantini, F. (1985). Developmental regulation of a cloned adult ~-globin gene in transgenic mice. Nature 315, 338-340. Marklund, S., Ellegren, H., Eriksson, S., Sandberg, K., and Andersson, L. (1994). Parentage testing and linkage analysis in the horse using a set of highly polymorphic microsatellites. Anita. Genet. 25, 19-23. Marmur, J., and Doty, P. (1961). Thermal renaturation of deoxyribonucleic acids. J. Mol. Biol. 3, 585-594. Marx, J. L. (1988). Cloning sheep and cattle embryos. Science, 239, 463-464. Maxam, A. M., and Gilbert, W. (1977). A new method for sequencing DNA. Proc. Natl. Acad. Sci. USA 74, 560-564. McGrane, M. M., deVente, J., Yun, J., Bloom, J., Park, E., Wynshaw-Boris, A., Wagner, T., Rottman, F. M., and Hanson, R. W. (1988). Tissue-specific expression and dietary regulation of a chimeric phosphoenolpyruvate carboxykinase/bovine growth hormone gene in transgenic mice. J. Biol. Chem. 263, 11443-11451. Meissner, G. (1986). Ryanodine activation and inhibition of the Ca release channel of sarcoplasmic reticulum. J. Biol. Chem. 261, 6300-6306. Mejdell, C. M., Lie, ~., Solbu, H., Arnet, E. F., and Spooner, R. L. (1994). Association of major histocompatibility complex antigens (BoLA-A) with AI bull progeny test results for mastitis, ketosis, and fertility in Norwegian cattle. Anita. Genet. 25, 99-104. Miawa, S. (1994). Triplet repeats strike again. Nature Genet. 6, 3-4. Michelsen, B., and Lernmark,/~. (1987). Molecular cloning of a polymorphic DNA endonuclease fragment associates insulin-dependent diabetes mellitus with HLA-DQ. J. Clin. Invest. 79, 1144-1152. Miller, A. D. (1992). Human gene therapy comes of age. Science 357, 455-460. Montgomery, G. W., Crawford, A. M., Penty, J. M., Dodds, K. G., Ede, J., Henry, H. M., Person, C. A., Lord, E. A., Galloway, S. M., Schmack, A. E., Sise, J. A., Swarbrick, P. A., Hanraham, V., Buchanan, F. C., and Hill, D. F. (1993). The ovine Booroola
46
JENS G. HAUGE fecundity gene (FecB) is linked to markers from a region of human chromosome 4q.
Nature Genet. 4, 410-414.
Mcrsch, G., and Leibenguth, F. (1994). DNA fingerprinting using the digoxigenated probe (GTG)5. Anita. Genet. 25, 25-30. Nakamura, Y., Leppert, M., O'Connel, P., Wolf, R., Holm, T., Culver, M., Martin, C., Fujimoto, E., Hoff, M., Kumlin, E., and White, R. (1987). Variable number tandem repeat (VNTR) markers for human gene mapping. Science 235, 1616-1622. Neibergs, H. L., Dietz, A. B., and Womack, J. E. (1993). Single strand conformation polymorphisms (SSCPs) detected in five bovine genes. Anim. Genet. 24, 81-84. Nepom, G. T., and Erlich, H. (1991). MHC Class-II molecules and autoimmunity. Ann. Rev. Immunol. 9, 493-525. Notariani, E., Galli, C., Laurie, S., Moor, R. M., and Evans, M. J. (1990). Derivation of pluripotential, embryonic cell lines from porcine and ovine blastocysts. Proc. 4 W. Cong. Genet. Appl. Livest. Prod. XIII, 58-64. Nygaard, A. P., and Hall, B. D. (1964). Formation and properties of RNA-DNA complexes. J. Mol. Biol. 9, 125-142. O'Brien, S. J., Womack, J. E., Lyons, L. A., Moore, K. J., Jenkins, N. A., and Copeland, N. G. (1993). Anchored reference loci for comparative genome mapping in mammals. Nature Genet. 3, 103-112. Old, R. W., and Primrose, S. B. (1990). "Principles of Gene Manipulation." Blackwell, Boston. Olson, M., Hood, L., Cantor, C., and Botstein, D. (1989). A common language for physical mapping of the human genome. Science 245, 1434-1435. Orita, M., Suzuki, Y., Sekiya, T., and Hayashi, K. (1989). Rapid and sensitive detection of point mutations and DNA polymorphisms using the DNA polymerase chain reaction. Genomics 5, 874-879. Ornitz, D. M., Palmiter, R. E., Hammer, R. E., Brinster, R. L., Swift, G. E., and MacDonald, R. J. (1985). Specific expression of an elastase-human growth hormone fusion gene in pancreatic acinar cells of transgenic mice. Nature 313, 600-603. Otsu, K., Khanna, V. K., Archibald, A. L.,. and MacLennan, D. H. (1991). Cosegregation of porcine malignant hyperthermia and a probable causal mutation in the skeletal muscle ryanodine receptor gene in backcross families. Genomics 11, 744-750. Owerbach, D. A., Lernmark, A., Platz, P., Ryder, L. P., Rask, L., Peterson, P. A., and Ludvigsson, J. (1983). HLA-D region [3-chain DNA endonuclease fragments differ between HLA-DR identical healthy and insulin-dependent diabetic individuals. Nature, 303, 815-817. Palmiter, R. D., and Brinster, R. L. (1986). Germ-line transformation of mice. Annu. Rev. Genet. 20, 465-499. Park, E. A., Gurney, A. L., Nizielski, S. E., Hakimi, P., Cao, Z., Moorman, A., Hanson, R. W. (1993a). Relative roles of CCAAT/enhancer-binding protein [3 and cAMP regulatory element-binding protein in controlling transcription of the gene for phosphoenolpyruvate carboxykinase (GTP). J. Biol. Chem. 268, 613-619. Park, C. A., Hines, H. C., Monke, D. R., and Threlfall, W. T. (1993b). Association between the bovine major histocompatibility complex and chronic posterior spinal paresis--a form of ankylosing spondylitis--in Holstein bulls. Anita. Genet. 24, 53-58. Patel, Y. M., Yun, J. S., Liu, J., McGrane, M. M., and Hanson, R. W. (1994). An analysis of regulatory elements in the phosphoenol-pyruvate carboxykinase (GTP) gene which are responsible for its tissue-specific expression and metabolic control in transgenic mice. J. Biol. Chem. 269, 5619-5628.
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
47
Proudfoot, N. J., Gil, A., and Maniatis, T. (1982). The structure of the human ~-globin gene and a closely linked nearly identical pseudogene. Cell 31, 553-563. Pursel, V. G., Pinkert, C. A., Miller, K. F., Bolt, D. J., Campbell, R. G., Palmiter, R. D., Brinster, R. L., and Hammer, R. E. (1989). Genetic engineering of livestock. Science 244, 1281-1288. Raines, L. J. (1991). The mouse that roared. In '~lYansgenic Animals" (N. L. First and F. P. Haseltine, eds.), pp. 335-345. Butterworth-Heinemann, Boston. Robertson, E., Bradley, A., Kuehn, M., and Evans, M. (1986). Germline transmission of genes introduced into cultured pluripotent cells by a retroviral vector. Nature 323, 445-448. R~dningen, O. K., Leren, T. P., R~sby, O., Tonstad, D~ Ose, L., and Berg, K. (1992). Haplotype analysis at the low density lipoprotein receptor gene in Norwegian familial hypercholesterolemic subjects. Clin. Genet. 41, 288-295. Rogers, G. E. (1990). Improvement of wool production through genetic engineering. Trends in Biotechnology 8, 6-11. Rohrer, G. A., Alexander, L. J., Keele, J. W., Smith, T. P., and Beattie, C. W. (1994). A microsatellite linkage map of the porcine genome. Genetics 136, 231-245. Rosenfeld, M. A., Siegfried, W., Yoshimura, K., Yoneyama, K., Fukayama, M., Stier, L. E., Paakko, P. K., Gilardi, P., Stratford-Perricaudet, L. D., Perricaudet, M., Jallat, S., Pavrani, A., Lelcocq, J. P., and Crystal, R. G. (1991). Adenovirus-mediated transfer of a recombinant al-antitrypsin gene to the lung epithelium in vivo. Science 252, 431434. Rudolph, J. A., Spier, S. J., Byrns, G., Rojas, C. V., Bernoco, D., and Hoffman, E. P. (1992). Periodic paralysis in Quarter horses: A sodium channel mutation disseminated by selective breeding. Nature Genet. 2, 144-147. Rusconi, S., and K~hler, G. (1985). Transmission and expression of a specific pair of rearranged immunoglobulin genes in a transgenic mouse line. Nature 314, 330-334. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988). Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487-491. Saiki, R., Scharf, S., Faloona, F., Mullis, K. B., Horn, G. T., Erlich, H. A., and Arnheim, N. (1985). Enzymatic amplification of ~-globin genomic sequence and restriction site analysis for diagnosis of sickle cell anemia. Science 230, 1350-1354. Salter, D. W., and Crittenden, L. B. (1989). Transgenic chickens: Insertion of retroviral vectors into the chicken germline. Theor. Appl. Genet. 77, 457-461. Sanger, F., Coulson, A. R., Hong, G. F., Hill, D. F., and Peterson, G. B. (1982). Nucleotide sequence of bacteriophage lambda DNA. J. Mol. Biol. 1{}2, 729-773. Sanger, F., Nicklen, S., and Coulsen, A. R. (1979). DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467. Scharf, S. J., Griffith, R. L., and Erlich, H. A. (1991). Rapid typing of DNA sequence polymorphism at the HLA-DRB 1 locus using the polymerase chain reaction and nonradioactive oligonucleotide probes. Hum. Immun. 30, 190-201. Schuster, D. E., Kehrli, M. E., Jr., Ackerman, M. R., and Gilbert, R. O. (1992). Identification and prevalence of a genetic defect that causes leukocyte adhesion deficiency in Holstein cattle. Proc. Natl. Acad. Sci. USA 89, 9225-9229. Schwenger, B., Schoner, S., and Simon, D. (1993). DUMPS cattle carry a point mutation in the uridine monophosphate synthase gene. Genornics 16, 241-244. Skolnick, M. H., and White, R. (1982). Strategies for detecting and characterizing restriction fragment length polymorphisms (RFLPs). Cytogenet. Cell Genet. 32, 58-67.
48
JENS G. HAUGE
Smithies, O., Gregg, R. C., Boggs, S. S., Koralewski, M. A., and Kucherlapati, R. S. (1985). Insertion of DNA sequences into the human chromosomal ~-globin locus by homologous recombination. Nature 317, 230-234. Snouwaert, J. N., Brigman, K. K., Latour, A. M., Malouf, N. N., Boucher, R. C., Smithies, O., and Koller, B. H. (1992). An animal model for cystic fibrosis made by gene targeting. Science 257, 1083-1088. Southern, E. M. (1979). Gel electrophoresis of restriction fragments. Methods Enzymol. 68, 152-176. Stryer, L. (1981). "Biochemistry." Freeman, San Francisco. Swanson, M. E., Martin, M. J., O'Donnel, J. K., Hoover, K., Lago, W., Huntress, V., Parsons, C. T., Pinkert, C. A., Pilder, S., and Logan, J. S. (1992). Production of functional human hemoglobin in transgenic swine. Biotechniques 10, 557-559. Swift, G. H., Hammer, R. E., MacDonald, R. J., and Brinster, R. L. (1984). Tissue-specific expression of rat pancreatic elastase I gene in transgenic mice. Cell 38, 639-646. Syvanen, A. C. (1986). Nucleic acid hybridization: From research tool to routine diagnostic method. Medical Biol. 64, 313-324. Szalai, G., Bailey, E., Gerber, H., and Lazary, S. (1993). DNA sequence analysis of serologically detected ELA class II haplotypes at the equine DQ~ locus. Anim. Genet. 24, 187-190. Takeshima, H., Nishimura, S., Matsumoto, T., Ishida, H., Kanagawa, K., Minamino N., Matsua, H., Ueda, M., Hanaoka, M., Hirose, T., and Numa, S. (1989). Primary structure and expression from complementary DNA of skeletal muscle ryanodine receptor. Nature 339, 439-445. Thomas, K. R., and Capecchi, M. R. (1986). Introduction of homologous DNA sequences into mammalian cells induces mutations in the cognate gene. Nature 324, 34-38. Thorsby, E., and RCnningen, K. S. (1993). Particular HLA-DQ molecules play a dominant role in determining susceptibility or resistance to Type 1 (insulin-dependent) diabetes mellitus. Diabetologia 36, 371-377. Todd, J. A., Bell, J. I., and McDevitt, H. O. (1987). HLA-DQ~ gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Nature 329, 599604. Tolstoshev, P., and Anderson, W. F. (1993). Gene transfer techniques for use in human gene therapy. In "Genome Research in Molecular Medicine and Virology" (K. W. Adolph, ed.), pp. 35-50. Academic Press, New York. V~ige, D. I., Olsaker, I., Lingaas, F., and Lie, 0. (1994). Isolation and sequence determination of porcine class II DRB alleles amplified by PCR. Anim. Genet. 25, 73-75. Vande Woude, G. F., Levine, A. J., Topp, W. C., and Watson, J. D. (1984). "Oncogenes and Viral Genes," Cold Spring Harbor, New York. Vincent, N., Ragot, T., Gilgenkrantz, H., Couton, D., Chafey, P., Gregoire, A., Briand, H., Kaplan, J. C., Kahn, A., and Perricaudet, M. (1993). Long-term correction of mouse dystrophic degeneration by adenovirus-mediated transfer of a minidystrophin gene. Nature Genet. 5, 130-134. Vogel, F., and Motulsky, A. G. (1986). "Human Genetics," Springer, New York. Watson, J. D., Gilman, M., Wirkowsky, J., and Zoller, M. (1992). "Recombinant DNA." Freeman, New York. Weber, J. L., and May, P. E. (1989). Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 44, 388-396. Weinberg, R. A. (1991). Tumor suppressor genes. Science 254, 1138-1146. Weissenbach, J., Gyapay, G., Dib, C., Vignal, A., Morisette, J., Milasseau, P., Vaysseix,
MOLECULAR GENETICS TO DIAGNOSIS, GENE THERAPY
49
G., and Lathrop, M. (1992). A second-generation linkage map of the human genome. Nature 359, 794-801.
Westphal, H. (1987). Transgenic mice. Bio Essays 6, 73-76. Wijsman, E. M. (1984). Optimizing selection of restriction enzymes in the search for DNA variants. Nucleic Acids Res. 12, 9209-9226. Williamson, R. (1993). Universal community carrier screening for cystic fibrosis? Nature Genet. 3, 195-201. Wivel, N. A. (1993). Regulatory considerations for gene-therapy strategies and products. Trends Biotech. 11, 189-191. Wright, G., Carver, A., Cottom, D., Reeves, D., Scott, A., Simons, P., Wilmut, I., Garner, I., and Colman, A. (1991). High level expression of active human ~l-antitrypsin in the milk of transgenic sheep. Biotechniques 9, 830-834. Yamamoto, K. K., Gonzales, G. A., Biggs, W. H., and Montminy, M. R. (1988). Phosphorylation-induced binding and transcriptional efficacy of nuclear factor CREB. Nature 334, 494-498. Young, B. D., and Anderson, M. L. H. (1985). In "Nucleic Acid Hybridization--A Practical Approach" (B. D. Haines and S. J. Higgins, eds.), pp. 47-111. IRL Press, Washington. Zabner, J., Petersen, D. M., Puga, A. P., Graham, S. M., Couture, L. A., Keyes, L. D., Lukason, M. J., St. George, J. A., Gregory, R. J., Smith, A. E., and Welsh, M. J. (1994). Safety and efficacy of repetitive adenovirus-mediated transfer of CFTR cDNA to airway epithelia of primates and cotton rats. Nature Genet. 6, 75-83. Zarbl, H., Saraswati, S., Arthur, A. V., Dionisio, M. Z., and Barbacid, M. (1985). Direct mutagenesis of Ha-ras-1 oncogenes by N-nitroso-N-methylurea during initiation of mammary carcinogenesis in rats. Nature 315, 382-385.
This Page Intentionally Left Blank
ADVANCES IN VETERINARY MEDICINE, VOL. 40
Genes Involved in Oncogenesis* JANELLE CORTNER,* SUSAN VANDE WOUDE,$ AND GEORGE F. VANDE WOUDE** National Cancer Institute, National Institutes of Health, Division of Basic Sciences, Bethesda, Maryland 20892, ~Laboratory Animal Resources, Colorado State University, Ft. Collins, Colorado 80523, **ABL-Basic Research Program, Frederick Cancer Research and Development Center, National Cancer Institute, Frederick, Maryland 21702 I. Introduction II. Detection of Genes Involved in Oncogenesis A. DNA-Mediated Gene Transfer in Cell Culture B. Retroviral-Mediated Oncogenesis C. DNA Tumor Viruses D. Oncogenesis Associated with Chromosomal Rearrangements III. Mapping and Cloning Genes Involved in Oncogenesis A. Physical Mapping: Somatic Cell Hybridization, Fluorescence In Situ Hybridization, and Comparative Genome Hybridization B. Genetic Mapping and Linkage Analysis C. Cloning Disease Genes: Screening for Oncogene Homologs, Candidate Gene Cloning, Chromosome Walking, Positional Cloning, and Representational Difference Analysis IV. Animal Model Systems Used to Study Gene Function A. Transgenic Mice B. Gene Knockouts V. Identity and Function of Genes Involved in Oncogenesis A. Oncogenes and Proto-oncogenes in Signal Transduction B. Tumor-Suppressor Genes
*Research sponsored in part by the National Cancer Institute, DHHS, under contract with ABL. By acceptance of this article, the publisher or recipient acknowledges the right of the U.S. Government to retain a nonexclusive, royalty-free license in and to any copyright covering the article. 51 Copyright 9 1997 by Academic Press. All rights of reproduction in any form reserved. 1093-975X/97 $25.00
52
CORTNER, VANDE WOUDE, AND VANDE WOUDE
C. DNA Repair Genes D. Metastasis-Associated Genes VI. Applications of Oncogene Research A. Screening for Cancer Predisposition B. Human Gene Therapy C. Implications for Veterinary Medicine VII. Conclusion References
I. I n t r o d u c t i o n
Our current u n d e r s t a n d i n g of the mechanisms underlying oncogenesis has evolved from critical advances made in the diverse fields of genetics, cytogenetics, virology, cell biology, and molecular biology. Although genetics has revealed t h a t cancer can be heritable, the study of tumor viruses has shown that cancer genes, or oncogenes, can arise from normal cellular genes, or proto-oncogenes; this lesson is underscored by cytogenetic studies of chromosomal abnormalities, which have guided us to the genomic location of many important proto-oncogenes. Molecular biology has made it possible to clone, sequence, and manipulate cancer-causing genes while cellular biology has revealed their normal and oncogenic functions (Klein and Klein, 1985). Now that we have a firm grasp of many of the mechanisms of oncogenesis, we are able to begin developing therapies for some cancers at the genetic level. After considering how genes involved in oncogenesis are detected and isolated, discussing some of their normal and oncogenic functions, and presenting techniques used to elucidate these functions, we will provide an introduction to these initial approaches to gene therapy directed against cancer. We now know that oncogenesis is usually a multistep process driven by the accumulation of genetic alterations, usually over the span of years, which culminates in deregulation of normal cellular controls on differentiation, division, and proliferation. Accumulation of additional genetic changes within the transformed population of cells leads to the production of metastasis-competent tumor cells. Insults to the genome leading to cancer can be in the form of chromosomal rearrangement, deletion, point mutation, or amplification. In animals and less frequently in humans, viral infection can also be involved in oncogenesis, occasionally resulting in acute transformation but more commonly playing an indirect role. Many types of genetic lesions are repaired at high frequency, but those t h a t are not correctly repaired result in
GENES INVOLVEDIN ONCOGENESIS
53
daughter cells harboring a genetic alteration. Frequently, mutations are deleterious and novel cells containing them die. Occasionally, mutations confer a growth advantage by disrupting any of a variety of normal cellular processes and thus initiate the process of oncogenesis. In deoxyribonucleic acid (DNA) tumor virus infection, viral gene expression can stimulate growth of the infected cell. Once a population of cells possessing a growth advantage is established, additional mutations can occur in individual cells leading to the acquisition of additional characteristics necessary for the development of the cancerous, or transformed, phenotype. Such mutations include those that can cause the cell to respond to a broader array of growth factors or obviate the need for growth factors entirely and those that can release the cell from normal controls of cell division and senescence. Mutations that directly promote oncogenesis affect genes that are either oncogenes or tumor-suppressor genes. Oncogenes are dominantly acting tumorigenic versions of normal cellular genes involved in signal transduction identified largely by the study of the acutely transforming retroviruses of animal tumor model systems. In contrast, tumor-suppressor genes normally function to suppress cellular growth and thus loss or inactivation of both copies is required for transformation. Tumor-suppressor genes have been identified primarily through the study of inherited h u m a n cancer syndromes. Most known tumorsuppressor genes encode transcription factors, but some encode cytoplasmic or cytoskeletal-associated proteins involved in signal transduction. Inactivation of DNA repair enzymes is indirectly involved in oncogenesis because loss of both copies of some DNA repair genes leads to cancer by increasing the number of mutations that escape detection and repair in the proto-oncogenes and tumor-suppresser genes during DNA synthesis. Another broad class of genes, metastasis-associated genes, are also not directly tumorigenic but are involved in cancer because they render tumor cells competent to migrate to, invade, and colonize a secondary site. However, there is evidence that expression of at least some transforming oncogenes may confer all or many of the phenotypic changes associated with metastasis by virtue of aberrant expression of a normal developmental genetic program.
II. Detection of Genes Involved in Oncogenesis Oncogenes and tumor-suppressor genes, including genes encoding some of the DNA repair enzymes, have been identified by four primary
54
CORTNER, VANDEWOUDE, AND VANDEWOUDE
approaches: (1) DNA-mediated gene transfer in cell culture, (2) identification of oncogenes in the transforming retroviruses, (3) study of the transforming genes in DNA tumor viruses, and (4) identification of genes at nonrandom sites of chromosomal rearrangement associated with malignancies and heritable cancer predisposition syndromes. The transforming potential of many of the known proto-oncogenes and tumor-suppressor genes has often been revealed by more than one approach, underscoring the pivitol role these genes play in oncogenesis. A. DNA-MEDIATED GENE TRANSFER IN CELL CULTURE DNA transfection, or gene transfer, was first used to identify the transforming genes of ribonucleic acid (RNA) and DNA tumor viruses (see Section II,C) and subsequently to identify the transforming genes in a variety of tumor cell lines (Hill and Hillova, 1972). In this assay, DNA sequences suspected of harboring oncogenic properties are transfected into immortal but nontransformed cells by any of a variety of techniques. Usually, murine fibroblasts such as NIH/3T3 or Swiss3T3 cells are the recipient cell lines of choice because they can be maintained as contact-inhibited, nontumorigenic cell lines. After transfection, the recipient cells are monitored for morphological transformation resulting from loss of contact inhibition by testing for focus formation and ability to grow in soft agar (Blair et al., 1982). Once a transformed colony of cells has been established, the foreign, transforming DNA can be recovered and molecularly cloned and studied to determine the basis for its transforming effect. Many oncogenes from mouse and human tumor cell lines have been identified using this approach (Goldfarb et al., 1982). Many of the transfectable human transforming genes were shown to be members of the ras family of oncogenes. These oncogenes were first identified in acutely transforming Harvey and Kirsten sarcoma viruses and were designated c-H-ras and c-K-ras (Der et al., 1982). A ras gene family member identified in a human neuroblastoma cell line and a human promyelocytic leukemia cell line were designated N - r a s (Murray et al., 1981). Approximately 15% of human tumor cell lines and fresh tumor biopsies have activated ras oncogenes as detected by this assay, and this number may be much higher in some human tumors, such as colorectal cancers (Bos et al., 1987). R a s appears to exert its oncogenic effect by activating kinases involved in signal transduction. This process is described more fully later. A number of novel transforming genes that are not members of the
GENES INVOLVED IN ONCOGENESIS
55
ras gene family or related to the viral oncogenes of the acutely transforming retroviruses have been identified by DNA transfection assays. These include the neu, met, trk, mas, HST, and KS3 oncogenes (Perkins and Vande Woude, 1993; Benz et al., 1993; Reddy et al., 1988; see Table 1). With the exception of the mas oncogene, which appears to be a unique integral membrane protein, the other oncogenes are members of either the tyrosine kinase growth factor receptor family (neu, met, or trk) or the fibroblast growth factor (FGF) family (HST, KS3). B. RETROVIRAL-MEDIATED ONCOGENESIS
The first oncogenes were identified in studies of cancer-causing retroviruses and their continued study has led to the identification of a variety of oncogenes and proto-oncogenes. Retroviruses can activate proto-oncogenes in two primary ways: (1) by insertion into the host genome at the transcriptional level and (2) by transduction of activated oncogenes resulting from proviral recombination with the host genome in such a way that a portion of a proto-oncogene is packaged in the viral particle as part of the proviral genome (Varmus, 1988) Retroviralmediated oncogenesis by viruses carrying transduced oncogenes is quite rapid and thus these viruses are called the acutely transforming retroviruses. In human disease, retroviruses can cause transformation by insertional activation but no naturally occurring h u m a n retrovirus has been found that transduces a human oncogene. However, the oncogenes and proto-oncogenes identified by studying the acutely transforming retroviruses of animals are also involved in nonviral oncogenesis in humans. 1. Activation of Proto-oncogenes by Proviral Insertion
The stable integration of the provirus into the host chromosome is an important step in the retrovirus life cycle. One potential consequence of proviral integration is the disruption of host genes, which can result in cellular transformation if a cellular gene controlling growth or differentiation is located at the site of integration. Because integration is essentially random and animal genomes are large, the likelihood of disrupting both copies of a single, dominantly acting gene is low. Thus, although proviral disruption of wild-type tumor-suppressor genes is possible, this has not yet been observed. Because retroviruses carry their own transcriptional regulatory sequences in their so-called long terminal repeats (LTRs), random insertion of the provirus necessarily involves the random insertion of transcriptional promoters, which can alter expression of host genes near
TABLE 1 ONCOGENES Oncogene
Method of Identification
Associated Tumor
Growth Factors v-sis int-2 il2 il3 ks3 hst
SSV transduction (monkey) MMTV proviral insertion (mouse) GaLV proviral insertion (ape) IAP proviral insertion (mouse) DNA transfection DNA transfection
Glioma/fibrosarcoma Mammary carcinoma T-cell lymphoma Myelomonocyte leukemia Kaposi's sarcoma Stomach carcinoma
Integral Membrane Tyrosine Kinases SM-FeSV transduction (cat) v-erbB AEV-H transduction (chicken) c-erbB ALV proviral insertion (chicken) Amplification (human) Amplification (human) c-erbB2 HZ4-FeSV transduction (cat) v-kit V-FO8 UR2 transduction (chicken) neu DNA transfection (rat) met DNA transfection trk DNA transfection
Sarcoma Erythroleukemia/sarcoma Erythroleukemia Glioblastoma Mammary carcinoma Sarcoma Sarcoma Neuroblastoma/carcinoma Carcinomas/sarcomas Colon carcinoma
Membrane-Associated Tyrosine Kinases RSV transduction (chicken) Y73/ESV transduction (chicken) Gr-FeSV transduction (cat) FuSV transduction (chicken) STV transduction (cat) HZ2-FeSV transduction (cat) Ab-MLV transduction (mouse)
Sarcoma Sarcoma Sarcoma Sarcoma Sarcoma Sarcoma Leukemia
v-f ms
v-src v-yes v-fgr v-fps v-fes v-abl v-abl
Serine-Threonine Kinases v-mos c-mos
v-raf
Ras-Related v-H-ras
v-K-ras N-ras
Nuclear Proteins/ Transcription Factors v-myc c-myc
Mo-MSV transduction (mouse) IAP proviral insertion (mouse) MSV-3611 transduction (mouse)
Sarcoma Plasmacytoma Sarcoma
Ha-MSV (rat) MAV proviral insertion (chicken) Point mutation (human) Ki-MSV (rat) Point mutation (human) DNA transfection (human) Point mutation (human)
Sarcoma Nephroblastoma Bladder and other carcinomas Sarcoma Carcinomas and leukemias Myeloid leukemia
Carcinoma; myeloid leukemia Bursal lymphoma T-cell lymphoma
N-myc
MC29 transduction (chicken) ALV, CSV, REV proviral insertion (chicken) MLV (mouse) and FeLV (cat) proviral insertion Amplification (human) Amplification (human)
L-myc v-myb c-myb v-fos v-jun v-rel v-ets v-erbA v-ski
Amplification (human) E26 (chicken) MLV proviral insertion (mouse) FBJ-MSV (mouse) ASV transduction (chicken) REV-T (turkey) E26 (chicken) AEV-ES4 (chicken) SK-ASV (chicken)
Small-cell lung carcinoma Neuroblastoma Small-cell lung carcinoma Small-cell lung carcinoma Erythroleukemia Lymphosarcoma Osteosarcoma Leukemia Lymphatic leukemia Erythroleukemia Erythroblastosis Carcinoma
Others mas (transmembrane) int-1
DNA transfection (human) MMTV proviral insertion (mouse)
Mammary carcinoma Mammary carcinoma
58
CORTNER, VANDEWOUDE, AND VANDEWOUDE
the site of insertion. LTRs are very strong promoters of transcription; proviral integration can thus up-regulate host gene expression. Overexpression or aberrant expression of a proto-oncogene can have oncogenic consequences, frequently giving rise to leukemias or lymphomas (Hayward et al., 1981). Thus, these viruses are named leukemia or leukosis viruses. Insertional activation of proto-oncogenes occurs frequently in animal tumors, but it does not appear to play a significant role in human tumors (see Section II, B, 3). The avian leukosis virus and the mouse mammary tumor virus are well-studied examples of retroviruses that transform by insertional activation (Weiss et al., 1982). The long latent period for disease caused by these retroviruses is partially the result of the low probability that proviruses will integrate into or adjacent to a host cellular proto-oncogene. Many oncogenes have been identified on the basis of insertional activation, including ras, mos, wnt-1, erbB, fins, myc, and others (see Table 1).
2. Oncogene Transduction by Acutely Transforming Retroviruses In contrast to retroviruses that transform via insertional activation of cellular proto-oncogenes, acutely transforming retroviruses can produce tumors in newborn animals in less than two weeks. The difference in disease latency observed with these two virus classes is due to the fact that the acutely transforming retroviruses have acquired host genes or portions of genes that cause rapid transformation (Bishop, 1985). These host-derived genes are called viral oncogenes (v-onc), and many have been identified in acute transforming retroviruses (see Table 1). Because the maximum length of the proviral genome that can be packaged in a retroviral virion is relatively fixed, some of the viral genes are lost when host genes are transduced into the retroviral genome. As a result, the acutely transforming retroviruses usually are replication defective such that spread of the virus requires co-infection with a closely related, replication-competent, helper virus. The helper virus is required to provide the viral gene products necessary for viral replication and production of infectious viral particles containing the transforming viral genome. A notable exception is the Rous sarcoma virus (RSV) which carries the v-src oncogene. Fortunately, no acutely transforming human retroviruses have yet been identified, although proviral sequences comprise a significant percentage of the human genome. This lack is usually attributed to the paucity of actively replicating endogenous human retroviruses available to support the replication of replication-deficient, transforming retroviruses. Although the acutely transforming retroviruses have little impact on human health, the characterization of their v-onc sequences and sub-
GENES INVOLVED IN ONCOGENESIS
59
sequent identification of the proto-oncogenes from which they were derived have led to the discovery of a large number of transforming genes that are activated by non-viral mechanisms in h u m a n cancers (see Table 1). Additionally, study of the acutely transforming retroviruses has greatly enhanced our understanding of oncogenic activation and mechanisms of transformation. Retroviral replication is characterized by a high mutation rate of the viral genome. This high mutation rate, coupled with selection for increased tumorigenic potential, can result in viruses with numerous changes in the v-onc sequences. Multiple types of genetic aberrations have been identified, including multiple point mutations, deleted exons, and deleted transcriptional and posttranscriptional regulatory elements (Curran et al., 1984). Many v-onc genes are expressed as fusion products with viral genes and are transcribed at high levels under the constitutive control of the retroviral LTR (Bishop, 1985). The fusion products contribute to transforming potential by misdirecting an oncogene product to an improper cellular location or by deleting or adding a regulatory domain. Moreover, the target cell specificity of the retrovirus can result in the expression of the oncogene in an inappropriate cell type. 3. H u m a n Retroviruses in Oncogenesis
As noted previously, h u m a n cancers do not routinely arise from either proviral insertion or oncogene transduction. However, h u m a n retroviruses do exist and can contribute to oncogenesis. The transforming potential of both the h u m a n immunodeficiency viruses, type 1 and 2 (HIV-1 and -2) and the h u m a n T-cell leukemia viruses, I and II (HTLV-I and -II) is low; transformation is infrequent and occurs only after very long periods of latent infection. The mechanism of transformation for the HIVs and the HTLVs is largely indirect and probably results from the action of virally encoded regulatory proteins designed to control virus replication by regulating the expression of cellular genes and the proliferation of host cells (Kieff, 1996). Although HTLV-I and -II are thought to have similar mechanisms of infection, only HTLV-I is frequently associated with h u m a n malignancy. A very low percentage of HTLV-I infected individuals develop adult T-cell leukemia (ATL; Stewart et al., 1994). ATL usually presents as a single, clonal tumor after several decades of infection, although a vast number of T cells are infected during the long disease progression. These observations argue strongly for an indirect transforming mechanism. The HTLV-encoded Tax transcription factor alters expression of known proto-oncogenes such as c-myc, c-fos, JunB, JunD, and EGR 1, as well as the growth factor/receptor pair interleukin-2 and granulocyte-
60
CORTNER, VANDEWOUDE, AND VANDEWOUDE
macrophage colony-stimulating factor. The HTLV-encoded protein Rex also contributes to alterations in normal cellular proliferation by interfering with cellular mRNA processing. HTLV glycoproteins are expressed on the surface of infected cells and may chronically stimulate its receptor, the T-cell receptor. These alterations of normal cellular events and possibly other less well-understood virally-mediated processes are believed to predispose HTLV-I infected T cells to transformation. The HIVs are most notable as the causative agent of the human acquired immunodeficiency syndrome (AIDS). However, the HIVs are also associated with some human malignancies, the most common of these being Kaposi's sarcoma and Hodgkin's and non-Hodgkin's lymphomas (Weller, 1994). The most likely mechanisms of transformation by the HIVs probably result from the action of auxiliary viral proteins, as with HTLV-I, and from the immunosuppressive effects of HIV infection. Interestingly, some lymphomas from AIDS patients may also overexpress the fps proto-oncogene as a result of proviral insertion. If so, this represents the first instance of insertional activation by a human retrovirus (Shiramizu et al., 1994). C. DNA TUMORVIRUSES
A variety of non-retroviruses are associated with human tumors. Of these, only the Epstein-Barr virus (EBV) is acutely transforming and all require long periods of infection before the appearance of the associated tumors (Rickinson, 1994). In almost all cases, the mechanism of oncogenesis is believed to be indirect and result from secondary cellular changes occurring in hyperproliferating, infected cells or in cells proliferating as part of tissue regeneration following immune clearing of infected cells. For EBV, overexpression of the viral protein LMP1 is transforming as a result of an unknown mechanism affecting proliferation of latently infected cells. Papovaviruses, composed of the polyomaviruses and the papillomaviruses, also cause hyperproliferation of infected cells; the mechanisms papovaviruses use for overriding cellular proliferation controls involve inactivation of tumor-suppresser gene products such as p 105RB and p53 or activation of cellular proto-oncogene products by complexing with viral proteins. Polyoma, SV40, and adenovirus are all transforming polyomaviruses. The polyoma viral oncoprotein middle-T antigen binds to the src proto-oncogene product and activates its kinase activity by blocking a downregulating phosphorylation event (Courtneidge, 1985). The polyoma viral oncoprotein middle-T antigen binds to and inactivates p53. SV40-mediated transformation requires SV40-
GENES INVOLVEDIN ONCOGENESIS
61
encoded large-T antigen capable of complexing with the products of cellular p53 and RB tumor-suppresser genes. Adenovirus E IA oncoprotein also binds to and inactivates cellular pl05RB (De Caprio et al., 1988; Whyte et al., 1988). The human papillomaviruses (HPVs) cause hyperproliferation of infected epithelial cells of the mid-epidermis producing benign lesions that can occasionally develop into tumors, the most common of these being cervical carcinomas (Howley, 1993). The viral oncoproteins responsible for causing hyperproliferation and predisposition to malignant transformation are encoded by the E6 and E7 genes. E6 forms complexes with p53 and causes its degradation. E7 binds pl05RB and related proteins causing the E2F transcription factor normally sequestered by p l05RB during certain times in the cell cycle to be released and thus competent to support gene transcription throughout the cell cycle. There are over 100 HPVs with a range of transforming potentials; the most highly transforming HPVs (HPV-16 and HPV-18) bind these cellular proteins more efficiently than the equivalent proteins encoded by the less-transforming HPVs, supporting the idea that these interactions are transforming. D. ONCOGENESISASSOCIATED WITH CHROMOSOMAL REARRANGEMENTS Nonrandom chromosomal abnormalities have proven invaluable for the identification of genes involved in oncogenesis, and for providing diagnostic clues for certain tumors. Interestingly, a correlation exists between the type of chromosomal abnormality and the histopathologic type of tumor in which it is found, suggesting that certain cell types are particularly susceptible to transformation by the specific oncogene(s) created by the corresponding chromosomal rearrangement. Most of the abnormalities characterized to date occur in hematologic malignancies. This phenomenon is widely attributed to the relative ease with which these tumors can be grown in culture, but may also be caused by a higher frequency of productive chromosomal abnormalities resulting from errors occurring during normal, somatic rearrangements in these cell types. It has also been postulated that hematopoietic cells are particularly sensitive to transformation because of the delicate balance of regulatory factors involved in the process of hematopoietic differentiation (Look, 1995). Nonrandom abnormalities in chromosomes from solid tumors have also been described and have revealed the identity of many oncogenes. Table 2 lists the chromosomal rearrangements for which the resi-
TABLE 2 ONCOGENES ACTIVATED BY CHROMOSOMAL REARRANGEMENTS
Oncogene
Function
Rearrangement
Associated Tumor
bcl-2 activation bcr-abl bcr-abl Ttg-1/Rhombotin lyl-1 SCL(tal;TCL-5) SCL / SIL Hox- 11 TAN- 1 E2A-prl PRAD-1, bcl-1 cycD1 c-myc Ig locus APL-RARa PRAD-1 G1 cyclin c-myc L-myc N-myc N-myc c-erbB c-erbB2 (her2;neu) hst mdm-2 gli cdk4 ret trk
Prevents apoptosis p210 abl tyr kinase p180 abl tyr kinase Zn transcription factor HLH transcription factor HLH transcription factor Interrupter locus Homeobox transcription factor Notch homolog Homeobox transcription factor Cyclin D/cell cycle regulation Cyclin D/cell cycle regulation Transcription factor Deregulation gene expression Cyclin D/cell cycle regulation Cell cycle regulation Transcription factor Transcription factor Transcription factor Transcription factor tyr kinase GFR tyr kinase GFR
t( 14:18) (q32; q21) t(9; 22) (q34; ql 1) t(9; 22) (q34; ql 1) t(11,14) (p 15;ql 1) t(7; 19)(q34;P13) t( 1; 14)(p32; ql 1) del(1) p33 t(10;14) t(7;9) t(1; 19)(q23;p13.3) t(11; 14)(q 13; 132) Amplification t(8; 14)(q24;q32) t(15;17) inv(11) inv(11) Amplification Amplification Amplification Amplification Amplification Amplification
Follicular center cell lymphoma Chronic myelogenous leukemia (T-ALL)a T-ALL T-ALL T-ALL T-ALL T-ALL T-ALL Pre B-ALL B-cell lymphoma: breast and other carcinomas B-cell lymphoma: breast and other carcinomas Burkitt's lymphoma Acute promyelocytic leukemia Parathyroid adenoma Breast, squamous cell carcinomas Small-cell lung carcinoma Small-cell lung carcinoma Small-cell lung carcinoma Neuroblastoma Glioblastomas Mammary carcinoma
Growth factor p53 binding protein Transcription factor Cyclin-dependent kinase tyr kinase GFR tyr kinase GFR
Amplification Amplification Amplification Amplification Translocatmn Translocation
Gastric carcinoma Sarcomas Sarcomas, gliomas Sarcomas, gliomas Thyroid carcinoma Colorectal carcinomas
aT-ALL: T-cell acute lymphoblastic leukemia
GENES INVOLVED IN ONCOGENESIS
63
dent genes have been identified and cloned. These rearrangements include translocations (exchange of material between two chromosomes), inversions (rearrangement of material within a chromosome), deletions (net loss of a portion of a chromosome), and duplication or amplification (both causing net gain of chromosomal segments; Trent, 1995). Aberrations in normal chromosome recombination and replication that produce such chromosomal rearrangements occur at a much higher frequency (between 1 and 10 per 1000 cell divisions) than the rate of accumulation of uncorrected point mutations in most genes within the normal human genome (about 1 to 10 per million cell divisions; Weigel, 1995). Thus, it is not surprising that chromosomal rearrangements underlie many of the genetic alterations involved in oncogenesis. 1. Translocation
Activation of proto-oncogenes by chromosomal rearrangements can occur by creation of novel fusions between segments of genes resulting from translocations. As in the case of proto-oncogene activation by acutely transforming retroviruses, the translocations can be activating by creating a novel fusion protein or by up-regulating expression of a proto-oncogene by positioning it downstream from a transcriptional promoter. Many translocations involve either the T-cell receptor (TCR) or immunoglobulin (Ig) loci, in T- and B-cell malignancies, respectively. Sequence analysis of the joint between these loci and the foreign loci reveal the presence of canonical heptamer-nonamer sequences, which are used during normal somatic rearrangement of the TCR and Ig genes, and indicate that the translocation most likely arose by a recombination error. By juxtaposing the proto-oncogene with the TCR or Ig loci, it becomes transcriptionally deregulated, structurally altered, or both, rendering the novel gene product oncogenic. One common feature in several of the genes that are deregulated by translocation in hematopoietic malignancies is that they encode proteins with significant homology to known transcriptional regulatory proteins (see Table 2; Cleary, 1991). For example, myc, lyl-1, and tal /scl are all transcription factors containing helix-loop-helix dimerization motifs. The t(15;17) translocation characteristic of acute promyelocytic leukemia (APL) results in the deregulated production of a chimeric protein containing much of the retinoic acid receptor a, which normally functions as a transcription factor in the presence of its ligand (de Th~ et al., 1990). The deregulated expression of this chimeric factor disrupts normal cellular events by altering the expression of genes regulated normally by the proto-oncogene product.
64
CORTNER, VANDEWOUDE, AND VANDEWOUDE
About 95% of chronic myelocytic leukemias (CML) result from a translocation between chromosomes 9 and 22, producing the so-called Philadelphia chromosome. In this translocation, the c-abl proto-oncogene is translocated from chromosome 9q34 to a breakpoint cluster region (bcr) in chromosome 22ql 1. The bcr region provides a transcriptional promoter and the 5' end of a fusion gene that encodes the kinase domain of the abl proto-oncogene at its 3' end. The novel fusion protein, bcr-abl protein kinase, causes a CML-like disease when the bcr/abl gene is expressed in transgenic mice (Collins et al., 1984; Heisterkamp et al., 1990) It appears that bcr-encoded sequences in the chimeric protein activate the normal c-abl tyrosine kinase, which is known to be involved in signal transduction, by direct physical binding of the kinase regulatory domain of abl (Pendergast et al., 1991).
2. Amplification In a process known as amplification, activation of proto-oncogenes can occur by duplication of a chromosomal segment spanning a protooncogene. Chromosomal amplification of a proto-oncogene increases the dosage of the proto-oncogene proportional to the number of copies of the gene and results in overexpression of the gene product. Several cellular oncogenes are amplified in human tumors (see Table 2). The c-myc proto-oncogene locus is amplified in a promyelocytic leukemia both in the primary tumor and in the cell line HL-60, which was derived from the tumor (Dalla Favera et al., 1982). Other oncogenes, such as c-erbB (EGFR), neu (HER-2), and c-myc family members, have been shown to be amplified in specific tumor types (see Table 2), and the presence of multiple copies of these genes has been associated with poor prognosis. The presence of multiple copies of N-myc (which was first identified as an amplified gene in h u m a n neuroblastoma) correlates with advanced stages of the disease. Likewise, the amplification ofmyc family members in small-cell lung carcinoma is also associated with a more aggressive disease progression (Little et al., 1983). Thus, the myc family members appear to be associated with the progression of neuroblastomas and small-cell lung carcinomas, whereas the c-erbB, or EGFR, gene is amplified in glioblastomas and squamous carcinomas.
3. Chromosomal Deletion and Loss of Heterozygosity DeMars postulated in 1970 that cancer-prone individuals are heterozygous for the cancer-predisposing gene and that cancer develops in the individual because of a somatic mutation or loss of the remaining normal allele (DeMars, 1970). Knudson developed a mathematical model based on his observations on the incidence and age of onset of unilateral and bilateral cases of familial retinoblastoma (Knudson,
GENES INVOLVED IN ONCOGENESIS
65
TABLE 3 TUMOR SUPPRESSOR GENES Tumor Suppressor Gene and Gene Family
Syndrome
Heritable and Sporadic Tumors
Transcription Factors RB
Familial retinoblastoma
WT1 p53 BRCA1 BRCA2
Familial Wilm's tumor Li-Fraumeni syndrome Familial breast cancer Familial breast cancer
Retinoblastoma; osteosarcoma; soft tissue sarcomas; breast and small-cell lung carcinoma Wilm's tumor; hepatoblastoma Many sarcomas and carcinomas Breast and ovarian carcinomas Breast carcinoma
DNA Repair Enzymes ERCC (nucleotide excision repair) FACC (cross link repair) DNA ligase I (DNA ligase activity) (X-ray response) hMLH1 (DNA mismatch repair) hMSH2 (DNA mismatch repair)
Xeroderma pigmentosum (XP) Fanconi's anemia (FAP) 46BR
Bloom syndrome (BS) Ataxia telangiectasia (AT) Lynch syndrome (HNPCC type 1) Lynch syndrome (HNPCC type 2)
Cytoplasmic/Cytoskeletal Proteins APC Familial adenomatous polyposis (FAP) NF1 Neurofibromatosis type 1 (NF1) Merlin Neurofibromatosis type 2 (NF2) Others ret (tyr kinase GFR)
Multiple endocrine neoplasia type 2A (MEN2)
Skin cancer Leukemia Lymphoma Carcinomas Lymphoma Nonpolyposis colon cancer Nonpolyposis colon cancer
Colorectal cancer Neurofibroma Neurofibrosarcoma Acoustic neuroma; meningioma; glioma; schwannoma Thyroid carcinoma Pheochromocytoma
1971). This model predicted t h a t one additional m ut at i on was the ratelimiting step in the development of tumors in familial cases, whereas two events were needed in nonfamilial cases. This hypothesis fit with both the high incidence of bilateral tumors in familial cases and the earlier age of onset of cancer in the familial cases. The list of the heritable types of cancer to which Knudson's model of mutation applies continues to grow (Weigel, 1995; Benz 1993; see Table 3). In heritable
66
CORTNER, VANDE WOUDE, AND VANDE WOUDE
cancers, the transmitted, germ line mutation can be either a point mutation or a large or small deletion. Usually, somatic loss of the remaining wild-type allele, which results in cellular transformation, arises by chromosomal deletion or loss r a t h e r t h a n by point mutation because deletions in and loss of chromosomes occur far more frequently t h a n do uncorrected point mutations at transforming positions within genes. In cytogenetic studies, loss of heterozygosity (LOH) refers to the loss of a wild-type allele by loss or deletion of a region of a chromosome at which the gene of interest resides. If the second allele contains an inactivating mutation LOH, results in a cell lacking functional copies of both alleles. LOH can also occur in sporadic tumors, indicating that loss of tumor-suppressor genes can underly oncogenesis in nonheritable as well as heritable cancers.
III. Mapping and Cloning Genes Involved in Oncogenesis When a cancer predisposition syndrome is determined to be hereditary, the underlying gene can be mapped in the genome and cloned based on the mapping data. Proto-oncogenes can be mapped and cloned by virtue of their homology to viral oncogenes. Gene mapping techniques can be either physical or genetic. Physical mapping techniques such as in situ hybridization or somatic cell hybridization are used to roughly position genes for which at least a partial DNA sequence is available, on a particular chromosome or chromosomal region. Genetic mapping is used to determine the proximity of two DNA m a r k e r s and is based on a statistical approach called linkage analysis. The m a r k e r s used in genetic analysis can be DNA based or can be a phenotype, such as inherited cancer predisposition syndromes. Approaches to gene cloning vary depending on whether a DNA probe for the gene exists and on the n a t u r e of the mapping data available. A. PHYSICAL MAPPING: SOMATIC CELLHYBRIDIZATION, FLUORESCENCE IN SITU HYBRIDIZATION, AND COMPARATIVE GENOME HYBRIDIZATION
Physical mapping techniques can be used to roughly position a gene within the genome when a probe, such as a viral oncogene, is available. Somatic cell hybridization assigns a gene to a given chromosome by following its segregation in a panel of interspecies hybrid cell lines. For example, h u m a n - h a m s t e r cell hybridization produces several distinct cell lines with predominantly h a m s t e r chromosomes and one or two h u m a n chromosomes. H u m a n chromosomal assignments are made by
GENES INVOLVED IN ONCOGENESIS
67
determining which cell lines contain the given h u m a n gene or its marker and correlating this information with the h u m a n chromosome content of the cell line. In situ hybridization and fluorescence in situ hybridization (FISH) are more precise localization techniques used to map a DNA fragment to a particular region or band within a chromosome (Harper and Saunders, 1981) Standard in situ hybridization uses isotopically labeled probes, whereas FISH uses fluorescently labeled probes. FISH is currently the more widely used of the two techniques because it is more sensitive. Fluorescently labeled probes are applied directly to interphase chromosomes and allowed to hybridize. Regions of positive hybridization are visualized with fluorescent microscopy. The term chromosome p a i n t i n g describes the detection of large regions of homology by FISH (Breneman et al., 1995). Mapping by comparative genome hybridization (CGH) using whole chromosome probes for chromosome painting has revealed that very few major changes have occurred in the genome organization of mammals during evolution. In addition to determining the degree of synteny between the chromosomes of distantly related species, CGH can be used to identify the regions of difference, or chromosomal imbalances, between individuals of the same species. This application is quite useful as a diagnostic tool in the study of h u m a n cancers and other diseases resulting from chromosomal rearrangements (Piper et al., 1995; Thompson and Gray, 1993). B. GENETIC ~/LkPPING AND LINKAGE ANALYSIS
Genetic mapping is used to position a gene conferring a particular trait to a specific region of the genome by finding correlations between the occurrence of the trait and an identifiable DNA fragment that has been mapped. Cloning of the gene responsible for the trait under study does not require that the mapped DNA fragment be the gene but merely reside near it. Genetic markers can be any discernible genetic phenotype, such as inherited cancer predisposition syndromes, or identifiable DNA fragment, such as chromosomal rearrangements, restriction fragment length polymorphisms (RFLPs), microsatellite sequences, known genes, or anonymous, unique sequences. The degree of linkage between markers is determined by the recombination frequency or the genetic distance between them. Recombination frequency is the percentage of meioses in which a recombination is detected between two markers. Genetic distance is measured in centimorgans (cM); if recombination between two markers is observed in one of 100 meioses then they are separated by 1 cM, which in the h u m a n genome corresponds to 10 ~ base pairs (bp) of DNA. In the case of chromosomal rearrange-
68
CORTNER, VANDE WOUDE, AND VANDE WOUDE
ments with a strong genetic linkage to cancer predisposition syndromes, the gene affected by the rearrangement is frequently involved in oncogenesis and can be cloned on the basis of the correlation with the rearrangement. RFLP analysis has been used to map the genes for cystic fibrosis, neurofibromatosis type I (NFI), multiple polyposis coli, Li Fraumini syndrome, muscular dystrophy, retinoblastoma, colon cancers, and small-cell lung cancers (Rommens, 1995; Wallace et al., 1990; Dunlop et al., 1990; Kunkel et al., 1989; Lee et al., 1987; Yokota et al., 1987). C. CLONING DISEASE GENES: SCREENING FOR ONCOGENE HOMOLOGS, CANDIDATE GENE CLONING, CHROMOSOME WALKING, POSITIONAL CLONING, AND REPRESENTATIONAL DIFFERENCE ANALYSIS
If a DNA fragment is available, such as an oncogene carried by a retrovirus, the homologous proto-oncogene can be cloned from a cDNA or genomic library by hybridization screening techniques using the oncogene as a probe (Sambrook et al., 1989). Once the gene has been cloned, its regulatory regions and the structure and function of the encoded protein can be studied. No prior knowledge of the position of the gene within the genome is necessary to clone a proto-oncogene corresponding to a known oncogene. However, once the gene is cloned, it can be mapped and used as a marker in genetic linkage studies to help identify the position of other genes. A mapped proto-oncogene may become a candidate gene in the search for the cause of a particular disease if genetic mapping of that disease assigns the responsible gene to that same region of a chromosome. The strength of the candidacy is increased if the responsible gene is predicted to have a structure or function related to the newly mapped gene. Identifying genes responsible for a disease by examining all the genes known to map to the same position in the genome as the unknown gene conferring the disease is called candidate gene cloning. The sequence of the candidate gene from individuals with and without the disease is compared to determine whether mutations within this gene are responsible for the disease phenotype. About 75% of the approximately 40 cloned mouse mutations were cloned in this way (Copeland et al., 1993). Once the position of a disease gene has been genetically mapped, the gene can be cloned using chromosome walking, positional cloning, or the candidate gene method (just described). Cloning genes by chromosome walking requires that a DNA fragment that is tightly linked to the gene of interest is available as a probe. In this technique, probes
GENES INVOLVED IN ONCOGENESIS
69
derived from the end of one clone are used to rescreen the library to obtain recombinants that contain adjacent chromosomal DNA sequences (Watson et al., 1983). Large segments of the mouse and h u m a n major histocompatibility (MHC) locus have been isolated in this manner (Steinmetz et al., 1981). Clones with larger inserts, such as yeast artificial chromosomes (YACs), can also be used. For positional cloning, a fine-structure linkage map must be constructed using markers close to and flanking the gene of interest (Collins, 1992). Animals showing recombination between these markers are analyzed to refine the map position down to the approximate insert size in YAC libraries. Restriction mapping is used to confirm the physical distance between the markers. YAC libraries are then screened using the known markers as probes, and positive, overlapping YACs are ordered into contiguous units. The YACs are screened for the gene of interest by expressing the genes contained in the YAC in animals bearing the disease. If a particular YAC can reverse or "rescue" the disease phenotype, then the precise position of the exons within the YAC are identified by a technique called exon trapping (Buckler et al., 1991). The trapped exons can then be used as probes to recover the corresponding copy DNA (cDNA), which is then used to identify the entire gene. This technique has been used to clone several mutations in the mouse, including Brachyury, Snell's Waltzer, agouti, Bcg, and short ear (Herrmann et al., 1990; Avraham et al., 1995; Bultman et al., 1992; Vidal et al., 1993; Kingsley et al., 1992). Representational difference analysis (RDA) allows the regions of difference between genomes identified by CGH to be recovered and cloned (Lisitsyn et al., 1993). Subtractive hybridization is used to enrich for DNA fragments uniquely represented in one source but not another, closely related source (Sambrook et al., 1989). This technique has been used to clone proviruses from the h u m a n genome and regions of chromosomal rearrangement and deletion that can occur during normal cellular differentiation or oncogenesis. CGH has been used to reveal that a herpesvirus is linked to Kaposi's sarcoma in both HIV and nonHIV infected patients (Schalling et al., 1995).
IV. Animal Model Systems Used to Study Gene Function As discussed previously, the study of the transforming retroviruses associated with a wide variety of animal tumor systems has led to the discovery of many oncogenes and proto-oncogenes (see Table 1). Additionally, mutations arising in the fruit fly Drosophila melanogaster
70
CORTNER, VANDE WOUDE, AND VANDE WOUDE
and the worm Caenorhabidis elegans following mutagenesis that confer novel phenotypes have provided a rich source of genes involved in development. Frequently, the genes underlying these phenotypes also play important roles in m a m m a l i a n development and oncogenesis. For example, the h u m a n homologs of many transcription factors involved in drosophila development, segmentation, and differentiation have been found to play a role in leukemogenesis, presumably because hematopoietic cells require a delicate balance of transcription factor expression during hematopoietic differentiation (Look, 1995). Gene function in the control of the cell cycle and in embryogenesis can also be studied by microinjection of gene transcripts into the oocytes of the African clawed frog, Xenopus laevis (Matten and Vande Woude, 1995). Transgenic mice are frequently used to study the effects of overexpressing a wild-type or m u t a n t gene or the effects of eliminating a gene altogether in a m a m m a l i a n system. Currently, transgenic mice and knock-out (i.e., null mutation) mice provide powerful new tools for elucidating both the normal cellular function and the oncogenic mechanisms of transforming genes as well as providing methods for testing gene therapy approaches to cancer. Transgenic refers to organisms that carry genetically engineered changes in their germ lines. Currently, the mouse is the most widely used animal in transgenic studies (Hughes, 1991). The mouse genome can be modified either by introducing genes at random positions or by homologous recombination at a specific locus either by directly injecting DNA into the pronucleus of a mouse embryo or by transfecting DNA for homologous recombination into embryonic stem cells. The mouse model systems provide an excellent tool for studying the function of a wide range of genes, including oncogenes. Transgenic technology can be used to eliminate or alter the expression pattern of a gene or to express an altered version of the gene. A large number of strains of transgenic mice have been generated in the study of oncogene function using these techniques, many of which will be discussed in subsequent sections.
A. TRANSGENIC MICE
When DNA is injected directly into the pronucleus of a mouse embryo, the sites of DNA integration are random; thus, DNA from any source can be used because homologous recombination is not required (Hogan et al., 1994). The offspring resulting from the microinjected embryos are analyzed to determine whether they carry the transgene
GENES INVOLVED IN ONCOGENESIS
71
in the germ line. Such transgenic mice can be used to study the functions of both regulatory and coding regions of oncogenes. By creating a transgene, which uses the transcriptional regulatory region of a gene to drive expression of a reporter gene, the ability of the regulatory elements to direct transcription of the reporter gene in various cell types during development and in the adult can be studied. For example, although the proto-oncogene wnt-1 was known to regulate dorsal-ventral patterning of the brain, a transgene study found that the regulatory elements of the gene governing the wnt-1 expression pattern reside in the 3' untranslated region (Echelard et al., 1994). The cellular proteins that bind to and regulate expression from these regulatory sequences can now be studied. Other transgenes are typically constructed from a known transcriptional promoter driving expression of the proto-oncogene under study. Transgenic mouse strains have been developed that consistently develop certain types of tumors in response to expression of a specific oncogene. Expression of the transgene can be driven by either a constitutive or a tissue-specific promoter, depending on the range of expression desired. Constitutive promoters are useful for determining those tissues that are most easily transformed by a particular oncogene. For example, constitutive expression of c-fos results in osteosarcomas (Johnson et al., 1992). Tissue-specific promoters can be used to study the effects of a highly transforming oncogene, such as c-myc, on a limited range of cell types: Mammary adenocarcinomas result when c-myc expression is driven by a mouse mammary tumor virus promoter, whereas B-cell lymphomas arise when an immunoglobulin promoter is used (Morgenbesser and DePinho, 1994). B. GENE KNOCKOUTS
Homologous recombination in cultured, pluripotent embryonic stem (ES) cells can be used to make a cell line bearing a precise gene replacement (Capecchi, 1989). In this procedure, a specifically designed DNA construct containing two homology regions flanking a selectable drug resistance marker is transfected into ES cells. The two homology regions direct recombination and the drug-resistance marker allows selection for cells stably maintaining the construct. After drug selection and analysis of the resultant clones for successful homologous recombination, altered ES cells are injected into 16- or 32-cell stage blastocysts to produce mice chimeric for wild-type and transgenic cells. Some of the chimeric animals will have germ line cells derived from the altered ES cells. When these animals are crossed with a wild-type
72
CORTNER, VANDEWOUDE, AND VANDE WOUDE
mouse, they will produce nonchimeric animals heterozygous for one wild-type and one recombinant allele. The heterozygous animals can then be used to generate animals homozygous for the recombinant allele. Usually, the goal of gene replacement studies is either to make a null mutation (gene knockout) via targeted disruption of the gene or to modify the gene by inserting a cassette containing a mutation found in a h u m a n gene (Galli-Taliadoros et al., 1995). Many null alleles of proto-oncogenes are embryonic lethal when homozygous: c-myc knock-out mice die because of faulty yolk sac circulation, NF-1, N-myc, WT-1, and RXRa null mutants die because of heart defects, while Rb, c-myb, c-jun, met, and HGF null mutants die because of failure of liver development or hematopoiesis (Stanton et al., 1992; Jacks et al., 1994; Kreidberg et al., 1993; Sucov et al., 1994; Lee et al., 1992; Hilberg et al., 1993; Bladt et al., 1995; Schmidt et al., 1995). Unfortunately, embryonic lethal knockouts prevent the researcher from studying all roles of the gene under investigation because the animals die at the first developmental stage in which the gene product is absolutely required. Furthermore, it is difficult to determine whether the lethal event is a direct result of the null mutation on the affected organ or whether the null mutation indirectly interferes with the development of the affected organ by disrupting the microenvironment supporting its development. Additionally, because organs are composed of a variety of cell types, determination of the cell type responsible for the failure of organ development is difficult when all of the cells in the animal are homozygous for the null allele. ES cells bearing two copies of the knock-out allele can be used to gain additional information about the role of a gene during development when the null allele is embryonic lethal in the homozygous condition (Hilberg et al., 1993). Homozygous knock-out ES cells can be made using the heterozygous knock-out ES cells as the recipient cells for an additional round of homologous recombination using a similar construct bearing a different drug resistance marker to target the remaining wild-type allele. These embryos are then implanted into the uterus of a pseudopregnant mouse. Chimeric animals composed of wild-type and homozygous knock-out cells will survive only if the double knockout cells do not contribute to tissues in which the gene product is required. Thus, scoring of tissue contribution of the knock-out cells in the viable progeny will reveal the tissues in which the gene product is required. Scoring of tissues is facilitated by marking the knock-out cells with an easily detectable reporter gene. Systems for generating conditional knockouts have been or are currently being developed that will allow null and other homologous re-
GENES INVOLVED IN ONCOGENESIS
73
combinant alleles to be targeted to a specific tissue. Such conditional knockouts will facilitate the study of null alleles, which are embryonic lethal when homozygous; because the gene of interest is knocked out only in a particular tissue, other specific aspects of the gene's function besides the first lethal event can be studied. The first successful report of a conditional knockout used the so-called Cre-loxP system based on bacteriophage P1 to knock out the DNA polymerase ~ gene in T cells (Gu et al., 1994). A similar system, which uses the FLP enzyme from yeast, also is being developed (Barinaga, 1994). As of 1997, no conditional proto-oncogene knockouts have been reported.
V. Identity and Function of Genes Involved in Oncogenesis In this article, we discuss four functionally defined categories of genes involved in oncogenesis. As noted earlier, oncogenes are dominantly acting tumorigenic versions of normal cellular genes involved in signal transduction. In contrast, tumor-suppressor genes normally function to suppress cellular growth and thus loss or inactivation of both copies is required for transformation (Fearon, 1995). Tumor suppressor genes encode either transcription factors or cytoplasmic and cytoskeletal-associated proteins involved in signal transduction. Loss or inactivation of tumor-suppressor genes is largely responsible for inherited cancer predisposition syndromes, although one oncogene, ret, has recently been identified as the cause of some cases of inherited multiple endocrine neoplasia (Van Heyningen, 1994). Inactivation of DNA repair enzymes is indirectly involved in oncogenesis because loss of both copies of some DNA repair genes leads to cancer by increasing the number of mutations that escape detection and repair in the protooncogenes and tumor-suppresser genes. Another broad class of genes, referred to here as metastasis-associated genes, also are not directly tumorigenic but are involved in cancer because they render tumor cells competent to migrate to, invade, and colonize a secondary site. A. ONCOGENES AND PROTO-ONCOGENES IN SIGNAL TRANSDUCTION The dominant transforming genes are collectively called oncogenes; their normal cellular counterparts are proto-oncogenes. Oncogene-encoded proteins superimpose their activity on the cell to elicit the transformed phenotype and are thus dominant genes considered to be positive regulators of growth. To the transformed cell, oncogene expression results in a gain of function. Frequently, oncogenes encode altered
74
CORTNER, VANDEWOUDE, AND VANDEWOUDE
versions of the normal cellular proteins encoded by their cognate protooncogenes and are created by chromosomal rearrangement, by the action of transforming retroviruses, or by point mutations affecting specific domains of the protein. Sometimes, a normal cellular protein is transforming when over- or aberrantly expressed because of alterations in the gene's transcriptional promoter or by amplification of gene copy number. Although the term proto-oncogene tends to suggest that proto-oncogenes reside in the genome for the sole purpose of expressing the neoplastic phenotype, they in fact encode proteins essential to the normal biological processes by which cells respond to extracellular and intracellular signals to grow, divide, and differentiate in a regulated fashion. Proto-oncogenes have been highly conserved during evolution and it is therefore possible to clone the proto-oncogene homologues from genetically well-characterized and easily manipulated organisms such as fruit flies, worms, frogs, and mice for further study (Shilo and Weinberg, 1981). Thus, the phenotypic influence of mutated proto-oncogene homologs on cell division, differentiation, and embryological development can be tested, and genes can be identified that suppress the m u t a n t phenotype, thereby identifying other members of the biochemical pathway. The coordinated growth, division, and differentiation that are essential for viable multicellular organisms are mediated through complex pathways that propagate and amplify the signal from outside the cell to specific targets within the cell (e.g., the cytoskeleton and the nucleus). Because growth is restricted to a subset of cell types within most organs of mature multicellular organisms, these pathways are highly regulated. The chain of events allowing a cell to respond to a growth or differentiation signal is called signal transduction. Each control point along the pathway can be the target of deregulation by oncoproteins, with subsequent overexpression or ectopic expression of normal or mutated proteins. Regardless of the exact structure of the oncoprotein, loss of regulated signaling by oncogene expression can force a cell into uncontrolled cell division, or invasive growth. Hence, the classes of proteins involved in signal transduction define the classes of oncogenes (see Table 1). External growth factors initiate the signaling cascade by binding to their cognate receptors on the surface of target cells. The ligands for growth factor receptors can be either soluble or present on other cells or in the extracellular matrix. Growth factors must be supplied to untransformed cells in culture to stimulate cell proliferation. Transformed cells show partial or complete relaxation of the requirements for growth factors when oncogenes override growth factor dependency
GENES INVOLVED IN ONCOGENESIS
75
by mimicking the action of ligands, their receptors, or downstream signals (Rijsewijk et al., 1987). Some oncogene products have been determined to function as extracellular growth factors when expressed in normal cells, including the product of the v-sis oncogene (Bishop, 1991). Once expressed in nonphysiologic quantities or in aberrant tissues, uncontrolled cell proliferation results, leading to tumorigenesis. A large number of tyrosine kinase growth factor receptors has emerged from studies of oncogenes. To this sizable number can be added other tyrosine kinase receptors cloned on the basis of homology or their role in fruit fly and worm development (Hanks, 1991). These receptors contain a number of common structural features that are important to their function (Cantley et al., 1991). All have large, glycosylated, extracellular ligand-binding domains. Many function as homodimers or as heterodimers of a and b chains. The cytoplasmic portion of cell surface receptors contain tyrosine kinase domains that are required for signal transduction. Phosphorylation sites vary among the proteins and determine phosphorylation function. These sites modulate receptor activity and provide recognition sites for receptor targets. Following the binding of ligand to the extracellular domain, these receptors undergo dimerization and activation of tyrosine kinase activity. This leads to autophosphorylation as well as binding and phosphorylation of other intracellular target proteins. Both N-terminal and C-terminal rearrangements appear to activate the transforming potential of these transmembrane receptor tyrosine kinase family members. These alterations may remove downmodulating domains of the protein and result in the constitutive activation of what is normally a conditionally regulated enzyme activity. Tyrosine kinase growth factor receptor proto-oncogenes include c-fms, c-erbB, c-ros, c-kit, met, and ret (see Table 1; Hanks, 1991). Ligand binding induces a conformational change in the cytoplasmic domain of the receptor and, in the case of tyrosine kinase growth factor receptors, this results in receptor clustering and activation of receptor function by autophosphorylation (Cantley et al., 1991). The lipid bilayer of the plasma membrane allows for lateral movement, dimerization or patching of receptors, and close interaction between the cytoplasmic tails of adjacent receptors (for crossphosphorylation) as well as between the receptors and a number of proteins localized to the inner surface of the plasma membrane. The activated receptor serves as a nexus for the assembly of a multiprotein complex of signal transducers, the exact composition of which depends on the cell and signal type. Many of the proteins involved in these complexes have been implicated in oncogenesis, including the
76
CORTNER, VANDEWOUDE, AND VANDEWOUDE
membrane-associated tyrosine kinases encoded by the abl, src, fgr, yes, and fps genes, and the product of the H-ras gene, p21ras. Other proteins that complex with activated receptors at the cytoplasmic side of the cell membrane during signal transduction include phosphatidylinositol-3 (PI3) kinase, the GTPase activating protein of p21ras (rasGAP), protein kinase C (PKC), and phospholipase Cg (PLCg). Some of the signaling proteins are localized at the plasma membrane by structural domains, such as basic motifs and posttranslational modifications, including myristylation, farnesylation, and palmitoylation, while others, including ras-GAP, PI3 kinase, PKC, and PLCg, are recruited to the plasma membrane within seconds following receptor activation (Cantley et al., 1991; Aaronson, 1991). Cytoplasmic proteins, such as the serine-threonine kinases encoded by the raf and mos proto-oncogenes are also involved in the signal transduction cascade. The plasma membrane itself appears to be integrally involved in the generation of molecules that propagate the signal by serving as a pool of substrates for several key enzymes within the complex. These enzymes include PLC~, which catalyzes the breakdown of phosphatidylinositol phosphates to generate diacylglycerol (DAG); phospholipase D, which converts phosphatidylcholine to a precursor of DAG; and PI3 kinase, which causes the generation of phosphatidylinositol derivatives. These seemingly simple signaling molecules have diverse effects on the cell including activation of PKC by DAG, release of calcium from intracellular stores of inositol triphosphates, translocation of certain enzymes to the plasma membrane, and regulation of cytoskeletal proteins by polyphospho-inositides, among many other responses (Majeras et al., 1990). The endpoints of signal transduction are usually either the cytoskeleton, which is the key effector of changes in cell shape and motility, or the nucleus, where regulation of cell cycle and gene expression takes place. Although much progress has been made in elucidating components of signal transduction pathways, significant gaps exist in our understanding, particularly in bridging the events at the cell membrane with events in the nucleus. Several of the nuclear proteins encoded by proto-oncogenes and oncogenes are site-specific DNA binding proteins, which act as transcription factors to directly regulate gene expression, ultimately leading to control of cell proliferation, division, and differentiation. Nuclear oncogenes, which encode transcription factors, include myc, myb, ets, fos, jun, erbA, ets, NFkB, ski, and many others (Lewin, 1994). The balance of specific transcription factors expressed within a cell is believed to be a primary way that signal transduction specificity emerges out of a set of pathways that overlap con-
GENES INVOLVED IN ONCOGENESIS
77
siderably from one system to another. Disruption of this delicate balance by expression of oncogene-encoded transcription factors is transforming by induction of expression of inappropriate downstream genes. Not all proto-oncogenes function in signal transduction. For example, the bcl-2 proto-oncogene, identified by a chromosomal breakpoint which activates bcl-2 in some B cell lymphomas, prevents apoptosis of progenitor cells in which it is expressed by preventing oxidative damage to certain cellular components (Korsmeyer et al., 1993). Lymphoid cells of transgenics overexpressing bcl-2 fail to undergo apoptosis and develop into lymphomas (Cory et al., 1994). Conversely, bcl-2 knockout animals undergo lymphoid apoptosis of inappropriate precursor cells in the spleen and thymus, confirming the essential role of bcl-2 in apoptosis (Hockenbery, 1994).
B. TUMOR-SUPPRESSOR GENES
In contrast to the oncogenes, tumor-suppressor genes normally function to suppress cellular growth and thus act in a dominant fashion to their transforming alleles, which themselves are recessive because they are inactivated or deleted. Germ line heterozygosity for a wildtype allele and an inactivated allele of a tumor-suppressor gene is commonly the underlying cause of cancer predisposition syndromes. Tumors arise in such heterozygous individuals upon loss of the remaining wild-type allele. Frequently, inherited cancer predisposition syndromes present as autosomal dominant diseases caused by the high frequency of loss of the remaining wild-type allele in individuals with germ line heterozygosity for a m u t a n t allele (Weigel, 1995). Tumorsuppressor genes have been identified as the cellular targets of the transforming proteins of DNA tumor virus and as genes whose functions are lost in certain cancer predisposition syndromes. Known tumor-suppressor genes encode either transcription factors or proteins associated with either cytoplasmic or cytoskeletal-associated proteins (see Table 3). 1. RB1
Cytogenetic studies of individuals in families transmitting familial retinoblastoma have led to the development of the model that invokes the loss of a tumor-suppressor gene as the basis for predisposition to retinoblastoma and other heritable cancer predisposition syndromes. Retinoblastoma accounts for 1% of all cancer deaths among children and is the most common malignant eye tumor. It can be multifocal and
78
CORTNER, VANDE WOUDE, AND VANDEWOUDE
bilateral, and 40% of the cases are familial. Karyotypic analysis of tumor tissue and nontumor tissue revealed that about 5% of individuals with retinoblastoma carried deletions in the region of chromosome 13q14 in tumor cells, suggesting that tumors arose in these individuals when the second copy of the responsible gene was lost as a result of the deletion. The remaining copy was assumed to carry an undetected deletion. Unambiguous support came with the demonstration that DNA probes to the chromosome 13q14 region failed to detect this DNA in tumors, while DNA from nontumor tissue from the same patient was found to contain a single copy of DNA from the 13q14 region (Cavenee et al., 1983). These investigations revealed that there was reduction to homozygosity (or loss of heterozygosity) in the tumor DNA for genes in the vicinity of chromosome 13q14, caused by nondisjunction, mitotic recombination, or gene conversion, leaving the tumor cells with a single, presumably damaged, copy of the RB 1 gene. Cavenee and coworkers (1983) later showed that it was the allele from the unaffected parent that was lost and the allele from the affected parent that remained in somatic tissue during tumor development, consistent with the loss of a tumor-suppressor gene. Equipped with DNA probes for the region of chromosome 13 and RNA extracted from retinoblastomas, Friend and co-workers were able to isolate the gene that confers susceptibility to retinoblastoma (Friend et al., 1988). The gene (RB1) encodes a protein designated pl05-RB. That RB1 encoded a tumor-suppressor gene was demonstrated by the reversion of transformed cells cultured from tumors to a nontransformed phenotype following reintroduction of the normal RB1 gene, a finding that may ultimately have implications for cancer treatment (Bookstein et al., 1990). By RNA analysis, RB 1 gene was found to be expressed in most tissues, which was surprising because in familial retinoblastoma only two types of tumors develop: retinoblastomas and, after long latent periods, osteosarcomas (Lee et al., 1987). It is likely that other tumor-suppressor genes are expressed in other tissues in which RB1 is expressed, rendering RB1 expression r e d u n d a n t and thus protecting those tissues from RBl-mediated tumorigenesis. In addition to retinoblastomas and osteosarcomas, loss of RB function is associated at variable frequency with some sporadic tumor types, including sarcomas and carcinomas of the lung, breast, and prostate. Oncogenesis leading to these tumors is thought to involve inactivation of additional tumor-suppressor genes because these tissues are unaffected in individuals carrying familial RB predisposition genes. pl05-RB is a nuclear protein with properties of a cell cycle regulator that exists in phosphorylated and hypophosphorylated forms (De Cap-
GENES INVOLVED IN ONCOGENESIS
79
rio et al., 1988). pl05-RB is hypophosphorylated in GO and immediately after mitosis before G1 and becomes phosphorylated upon stimulation with mitogen. Phosphorylation of p l05-RB is highest at the start of S phase (Buchkovick et al., 1989, Chen et al., 1989). Hypophosphorylated p l05-RB complexes with the transcription factor E2F, blocking its ability to transcribe S phase-specific genes at times after S phase. Late in G1, phosphorylated pl05-RB cannot bind E2F, which is then free to promote transcription of S phase-specific genes involved in DNA synthesis. When complexed with hypophosphorylated pl05-RB, DNA viral oncoproteins SV40 large T antigen and HPV E7 inactivate p l05-RB and prevent it from interacting with E2F and suppressing S phase-specific transcription and thus from suppressing proliferation (Nevins, 1992). Given that RB is associated with a broad spectrum of tumors and is targeted by the transforming proteins of the DNA tumor viruses, it was believed that RB must play a critical role in controlling cellular proliferation during development and that animals homozygous for the null RB allele would die very early in embryogenesis because of a broad range of defects resulting from the loss of cellular proliferation control. Surprisingly, RB knock-out mice are viable until the fetal period when they die from defects in hematopoiesis (Lee et al., 1992). RB knock-out mice also have defective peripheral and central nervous systems; both the hematopoiesis and nervous system defects are thought to result from the failure of the cells to cease proliferation at a specified point during differentiation. Thus, as expected, the RB knock-out phenotype is embryonic lethal because of defects in cellular proliferation control. However, the relatively limited effects of the RB mutation were not expected and provide independent evidence that other tumor-suppressor genes must function in the unaffected tissues to control proliferation. This hypothesis is supported by the fact that only two types of tumors routinely occur in individuals with familial retinoblastoma. Interestingly, mice heterozygous for the null RB allele do not develop retinoblastoma, as do humans heterozygous for a null RB allele, but instead develop tumors of the pituitary when the second RB allele is lost in somatic cells. This unexpected phenotype indicates that other tumor-suppressor genes may function in the murine retina to compensate for the loss of RB and may point to a genetic treatment of retinoblastoma in humans (Matzuk and Bradley, 1994). 2. p 5 3
As with RB-1, p53 was found to form a tight complex with the transforming proteins of the DNA tumor viruses, such as SV40-encoded large T antigen, which suggested that inactivation of p53 might be
80
CORTNER, VANDEWOUDE, AND VANDEWOUDE
involved in oncogenesis. Subsequently, alterations at the p53 locus on chromosome 17p were found in a large percentage and wide variety of h u m a n tumors and is currently the most commonly altered gene associated with human cancers (Levine et al., 1991b; Hollstein et al., 1991). Indeed, about 75% of colon tumors show abnormalities at both p53 alleles: One allele is often deleted, while the other has point mutations, which are usually missense mutations that yield an altered protein product. Interestingly, inactivation of p53 in most colorectal tumors is sporadic, resulting from somatic mutations. Individuals with the cancer-predisposing Li-Fraumeni syndrome are born with mutations in one allele of the p53 gene and develop tumors that bear mutations at both alleles, as predicted by Knudson's hypothesis developed in the analysis of retinoblastoma (Malkin et al., 1990). As with RB-1, wildtype p53 can reverse the transformed phenotype of tumor cells that are transformed because of loss of both copies of p53. The point mutations found in the p53 gene are highly clustered at points that define its functional domains. Most mutations occur in one of three amino acids, 175, 248, and 273; other mutations are clustered within four short regions (Levine et al., 1991a). Certain mutations are more common in certain tumor types, which may reflect that different cell types are sensitive to the loss of specific p53 functions, p53 is a nuclear phosphoprotein containing a transcription activation domain at its amino terminus, a site-specific DNA binding motif in its central region, and a tetramerization domain at its carboxy terminus. Interestingly, some transforming mutations in the tetramerization domain act in a dominant negative fashion and can thus inactivate wild-type p53 in heterozygous cells. Similarly, p53 can be inactivated in cells overexpressing the product of the MDM2 gene, which binds to and inactivates p53. When overexpressed, MDM2 acts as an oncoprotein by virtue of its binding to p53 (Momand et al., 1992). p53 is believed to execute its role as a tumor suppressor by blocking DNA replication under certain circumstances and by regulating transcription of particular genes. Wild-type p53 positively regulates transcription of the GADD45 and WAF1 genes, which encode proteins involved in DNA repair and cell cycle control, respectively, p53 is thought to function by channeling cells with damaged DNA into either G1 arrest for repair of DNA damage or into apoptosis if damage is irreparable (Lee et al., 1994). Cells that lack wild-type p53 fail to arrest in G1 for DNA repair prior to initiating replication and thus accumulate additional mutations at an accelerated rate. Given the role of p53 in cell cycle progression, p53 knock-out mice were expected to have great genomic instability and thus a heightened frequency of tumorigenicity if they survived embryonic development. Production of these animals
GENES INVOLVED IN ONCOGENESIS
81
has shown that they are able to develop normally but have a much greater incidence of tumor development resulting from oncogene overexpression caused by genomic instability (Rong et al., 1995; Donehower et al., 1995). The heterozygous animals also develop tumors at a greater than normal rate; these animals are an excellent model for LiFraumeni tumor susceptibility syndrome, which, in humans, results from the p53 heterozygous condition (Sands et al., 1994). Interestingly, although loss of p53 in colon tumors is quite common, neither p53-deficient mice nor Li-Fraumeni patients develop colon cancer at a rate greater than that seen in the general population. This unexpected result may provide information concerning the order of acquisition of mutations and resultant phenotypes in colon cancer progression (Sands et al., 1994). In the colon, cells originating in the crypt are ultimately sloughed off in the villi; thus, although the p53-deficient colon cells have greater genetic instability compared to normal colon cells, the colon is protected from tumor formation in the p53-deficient animals by rapid cellular turnover. Colon tumors usually develop from benign polyps because cells within the polyps do not turn over as rapidly, thus allowing accumulation of cells bearing transforming mutations. 3. WT-1
Wilms' tumor is a nephroblastoma arising in young children in both sporadic and inherited modes. In both cases, tumors are caused by loss or inactivation of both alleles of WT-1 (Haber and Housman, 1992). WT-1 expression is restricted to the developing kidney, which accounts for the limited spectrum of tumors with which its m u t a n t alleles are associated. WT-1 encodes a zinc finger transcription factor, which regulates expression through the same DNA binding site as those bound by the early growth response genes EGR1 and EGR2. Normally, WT-1 can repress transcription of several growth-associated genes and upregulate expression of other genes depending on the p53 background (Rauscher, 1993). Loss of the transcriptional repression of growth-inducing genes, including several growth factor receptors and some transcription factors, is believed to cause cellular transformation. 4. B R C A 1
Over half of all cases of inherited premenopausal breast cancer, many inherited cases of ovarian cancer, and a significant fraction of sporadic breast cancers are caused by transmission of altered alleles of the BRCA1 gene (Miki et al., 1994). An additional gene linked to breast cancer but not to ovarian cancer, BRCA2, has been identified (Wooster et al., 1994). As with RB-1, p53, and WT-1, BRCA1 encodes a transcrip-
82
CORTNER, VANDE WOUDE, AND
VANDEWOUDE
tion factor. As with all the tumor-suppressor genes encoding transcription factors, predisposition for cancers with which they are associated presents as an autosomal dominant trait in individuals with germ line heterozygosity for inactivated alleles because of the frequency of loss of the remaining wild-type allele. Whereas the cumulative lifetime risk for breast cancer in individuals with no BRCA1 mutation is about 10%, the cumulative risk for individuals heterozygous for a m u t a n t BRCA1 allele is approximately 90%. Thus, detecting individuals who have inherited BRCA1 mutations would greatly enhance our ability to predict a high predisposition to breast cancer. 5. Cytoskeletal Components of Cellular Junctions As with the tumor-suppressor genes encoding transcription factors, other tumor-suppressor genes are involved in signal transduction, albeit at different points in the cascade. The genes responsible for type 1 and type 2 neurofibromatosis, NF1 and NF2, function in the cytoplasm and in the cytoskeleton, respectively. Tumors associated with the neurofibromatosis syndromes occur in individuals with germ line mutations after somatic events inactivate the remaining wild-type allele (The et al., 1993). Although NF1 is widely expressed, inactivation of both copies of NF1 results in tumors restricted to tissues of neural crest origin, presumably because of the expression of functionally redundant proteins in other tissues. NF1 encodes neurofibromin, which functions as a GTPase activating protein that can inactivate the ras gene product by converting its bound GTP to GDP (Xu et al., 1994). Thus, loss of neurofibromin function is thought to be transforming by increasing the signal transduction activity. NF2 encodes schwannomin, which is believed to link the cellular membrane with cytoskeletal components. Loss of function of schwannomin is thought to lead to tumors of the central nervous system by altering the cellular response to cell-cell and cell-substrate signals in those cells (Trofatter et al., 1993). One form of inherited predisposition to colorectal cancer, familial adenomatous polyposis (FAP), results from inactivating mutations in the adenomatous polyposis coli (APC) gene (Su et al., 1993). Additionally, somatically acquired mutations in both APC alleles appear to be involved in a large fraction of sporadic colorectal tumors. APC encodes a large protein that resides in the cytoplasm and binds to members of the catenin family, which are involved in adherins junctions of epithelial cells. Loss of function of APC is thought to be transforming by interfering with adhesion signals, leading to loss of contact inhibition of proliferation. A mouse intestinal tumor model has been identi-
GENES INVOLVED IN ONCOGENESIS
83
fled in which germ line inactivation of the mouse APC gene results in multiple intestinal neoplasias (MINs; Moser et al., 1990). The MIN mice express different phenotypes depending on their genetic background; study of the MIN mouse tumor model will thus provide additional insight into the process of oncogenesis mediated by APC. C. DNA REPAIR GENES DNA repair enzymes are also involved, albeit indirectly, in oncogenesis. Although no DNA repair genes are directly transforming, loss or inactivation of both copies of some DNA repair genes leads to cancer by increasing the rate of accumulation of mutations in the protoooncogenes and tumor-suppressor genes. Interestingly, specific DNA repair genes are associated with specific tumor types. Although their role in oncogenesis is indirect, the impact of the transforming alleles of DNA repair genes on human health is significant, especially with regard to colorectal cancer. Inherited predisposition of colorectal cancer accounts for roughly 15% of the approximately 150,000 cases of colorectal cancer reported each year (Hamilton, 1992). Two forms of inherited colorectal cancer have been identified: FAP (discussed previously) and hereditary nonpolyposis colorectal cancer (HNPCC), which accounts for the majority of cases of familial colorectal cancer. The genes predisposing individuals to HNPCC, MSH2, and MLH1, encode DNA mismatch repair enzymes, which, when inactivated, predispose the cell to "stuttering" during replication of certain dinucleotide repeats (Thibodeau et al., 1993). This stuttering effect results in genetic instability when extra copies of the dinucleotide repeats are inserted. Such insertions can cause inactivation of tumor-suppressor genes or activation of protooncogenes. HNPCC presents as an autosomal dominant disease because of the high rate of mutation of the remaining wild-type allele in individuals carrying one mutant allele. Other hereditary cancers resulting from defects in DNA repair genes are recessive, presumably because of low rate of mutation of the remaining wild-type allele in individuals carrying one mutant allele. Frequently, alterations of more than one gene can be causal; the various genes associated with each of the syndromes defines the genes involved in specific repair pathways. Defects in helicases and in the nucleotide excision repair pathway account for the skin tumors associated with xeroderma pigmentosum resulting in heightened sensitivity to ultraviolet irradiation (Friedberg, 1992). Leukemia associated with Fanconi's anemia results from loss of function of genes involved in the
84
CORTNER, VANDE WOUDE, AND VANDE WOUDE
repair of DNA cross links (Friedberg, 1992; Strathdee et al., 1992). Lymphomas in individuals with ataxia telangiectasia arise because of defects in genes of the x-ray response pathway (Gatti et al., 1991). Defects in a DNA ligase gene are believed to be responsible for tumors arising in individuals with Bloom syndrome (Heim et al., 1992). D. METASTASIS-AssoCIATED GENES
Currently, tumor metastasis is the most challenging problem faced by oncologists in both the clinic and the laboratory. As a result of genetic instability, tumors are composed of a population of heterogeneous cells that accumulate a range of genetic alterations in addition to the original transforming mutations (Fidler and Poste, 1985). Within this heterogeneous population, some cells capable of metastasis can develop. Metastasis-associated genes encode a diverse array of proteins including adhesion molecules, cytoskeletal components, proteases, and regulatory molecules, which are not themselves transforming but which equip the transformed cell to metastasize. In some cases, mutation results in loss of function, for example, of adhesion molecules, resulting in increased cellular motility. In other cases, expression of metastasis-associated genes, such as proteases, in cells that do not normally express them produces cells capable of invasion. Fortunately, metastasis is a highly selective process, which very few tumor cells are capable of successfully completing (Fidler, 1995). For tumor cells to metastasize, the tumor must be well vascularized to support growth of the tumor beyond a few millimeters in diameter and to provide a route of metastasis. Consequently, a subset of the tumor cells must produce angiogenic factors capable of inducing endothelial cells to proliferate and form capillaries within the tumor. Furthermore, metastatic cells must be capable of detaching from the tumor mass and invading either the lymphatic or circulatory systems. These steps require reduced cellular adhesion to the tumor cell mass and increased affinity for the endothelial cells of the blood vessels, as well as increased motility through extracellular matrix of the basement membranes. Once launched, the candidate metastatic cell must survive the mechanical shear forces of circulation, evade detection and clearing by the immune system, and adhere to a capillary wall at the secondary site. Adhesion to and escape through the capillary wall into the surrounding tissue again requires affinity for the endothelial cells as well as motility and the ability to invade through extracellular matrix of the basement membranes. Once the secondary site has been penetrated, the successful metastatic cell must be able to proliferate; this
GENES INVOLVEDIN ONCOGENESIS
85
step requires that the cell's proliferation not be adversely affected by factors present in the new microenvironment and that sufficient nutrients and growth factor are present to sustain the secondary tumor. The standard model for the development of the metastatic phenotype holds that each of the characteristics necessary to support metastasis are acquired as the result of independent genetic alterations with in the transformed cell. Given that a tumor cell must initiate a gene expression program that correctly balances the production of angiogenic factors, motility factors, cytoskeletal factors, proteases, and myriad regulatory molecules, as well as downregulate expression of key adhesion molecules, it is surprising that any transformed cell can do so by the accumulation of multiple independent genetic alterations. One can imagine that if such a gene expression program could be triggered with a single change, then metastasis would become a much more likely outcome. It is becoming clear that in the case of autocrine transformation by overexpression of the met proto-oncogene product and its ligand HGF/SF, activation of this single transforming gene product can indeed initiate many if not all of the ancillary cellular functions just listed that are necessary for expression of the metastatic phenotype. The c-met proto-oncogene encodes the tyrosine kinase growth factor receptor for the HGF/SF growth factor (Bottaro et al., 1991). HGF/SF is normally produced by mesenchymal cells and acts in a paracrine or endocrine fashion on met-expressing epithelial cells, although met is occasionally expressed in some populations of mesenchymal cells (Rosen et al., 1994). From gene knock-out studies it was learned that during embryonic development met-expressing myogenic precursors originating in the somites migrate to the presumptive limb buds in response to HGF/SF production at that site (Bladt et al., 1995). Although the genes expressed during this process of invasion and migration are just beginning to be identified, it is clear that many of the same gene products, when inappropriately expressed, could fulfill many of the requirements for metastasis. Besides transforming HGF/SF-expressing mesenchymal cells, met expression triggers the ability of the cells to degrade extracellular matrix and basement membrane by inducing the expression of proteases such as collagenase and the urokinase-type plasminogen activator (Rong et al., 1994; Jeffers et al., 1996a; Jeffers et al., 1996b). Additionally, m e t - H G F / S F signaling enhances cellular motility and causes a reduction in cell-cell adhesion (Rubin et al., 1993). These effects, coupled with the mitogenic effect of m e t - H G F / S F co-expression and the ability of HGF/SF to promote angiogenesis of endothelial cells, suggest that transformation by overex-
86
CORTNER, VANDE WOUDE, AND
VANDE WOUDE
pression of the c-met oncogene may result in expression of a metastatic program of gene expression (Grant et al., 1993). Co-expression of m e t and HGF/SF in mesenchymal cells is not only transforming but also produces metastases rapidly and at high frequency, supporting the idea that inappropriate expression of the m e t - m e d i a t e d embryonic migration program can be used by the transformed cells to drive metastasis (Rong et al., 1992; Jeffers et al., 1996a). Furthermore, it is likely that this mechanism plays an important role in metastasis of human sarcomas because a high percentage of sarcoma cell lines and tumor samples express m e t inappropriately (Rong et al., 1993; Rong et al., 1995; Cortner and Vande Woude, 1996).
VI. Applications of Oncogene Research A. SCREENING FOR CANCER PREDISPOSITION
A primary goal of oncogene research is the development of diagnostic assays to identify individuals at heightened risk for cancer. Although widespread screening of the general population for all of the known cancer genes is currently not feasible, targeted screening of certain groups for loss of or mutations in specific genes is feasible. For example, it is estimated that 1 in 200 women carries alleles of BRCA1 predisposing to either breast or ovarian cancer. Identifying that 0.5% of of the female population would dramatically reduce the number of cases of advanced breast cancer by identifying those individuals who should be screened early and frequently by mammogram for the occurrence of nonpalpable, nonmetastatic tumors (King et al., 1993). Such early detection and treatment would save money as well as lives. By focusing screening efforts only on individuals known to be at greater risk because of higher rates of breast cancer among relatives, the majority of women predisposed to breast cancer because of BRCA1 mutations could be identified. Although screening for cancer genes has many merits, there are compelling reasons why, under current social conditions, even at-risk individuals might choose to forgo testing if it were made available. A major problem associated with making tests for cancer predisposition widely available is that they may jeopardize the ability of high-risk patients to obtain insurance (Ostrer et al., 1993). If tests are widely available, insurers or employers could conceivably require negative results as a condition for employment or coverage as a means of limiting lost productivity and excluding individuals likely to require expen-
GENES INVOLVED IN ONCOGENESIS
87
sive medical procedures. Given these problems, it is clear that we need to establish guidelines for testing such that at-risk individuals are not penalized by loss of employment or insurance coverage if we are to benefit from genetic screening for cancer predisposition. Methods for cancer gene screening vary depending on the genes being assayed and the nature of the mutations that inactivate them. If an inherited gene is inactivated by deletion of a large chromosomal region, then cytogenetic studies can be used as a screen. If submicroscopic deletions are responsible, then FISH could be used. Alternatively, biochemical testing for reduced levels of protein or protein activity of the product encoded by the tumor-suppressor gene or an adjacent gene that is also deleted could be used. If the gene is inactivated by sequence amplification or by certain mutations that alter the restriction pattern of the genomic DNA, then the cancer-predisposing alleles can be detected by DNA hybridization analysis in conjunction with RFLP mapping. If the gene is inactivated by one or more point mutations that do not produce discernible RFLPs, then sequence analysis is required to detect the mutation. Gene sequence can be determined by biochemical or hybridization-based methods. If the gene is inactivated by very specific point mutations that recur in unrelated individuals, as is the case for p53, then those regions of the gene can be analyzed specifically. If the gene can be inactivated by a wide array of mutations, then screening by sequence analysis becomes more difficult because the entire gene may need to be sequenced. B. HUMAN GENE THERAPY A wide variety of somatic cell gene therapy protocols are currently under development in the laboratory or in use in the clinic that are designed to treat single-gene diseases, a variety of cancers, or HIV infection (Kohn et al., 1989; Roth et al., 1994; Dropulic and Jeang, 1994). The goal of human somatic cell gene therapy is usually one of the following: to repair or compensate for a defective gene, to enhance the immune response directed against a tumor or pathogen, to protect vulnerable cell populations against treatments such as chemotherapy, to generate a marked population of cells for tracing the origins of recurrent tumors, or to kill tumor cells directly (Lee and Klein, 1995). Transmission of the recombinant DNA to target cells is accomplished either by ex vivo modification of the target cells followed by reintroduction into the patient or by restricted in vivo delivery (Lee and Klein, 1995). Ex vivo modification is currently more widely used than in vivo therapy. For example, cells found in the bone marrow can be tar-
88
CORTNER, VANDEWOUDE, AND VANDEWOUDE
geted ex vivo because they are readily accessible, manipulatable in vitro, and contain stem cells to perpetuate the altered genotype. When in vivo delivery is used, the transgene is targeted to the appropriate cells by localized exposure, by use of a tissue-specific promoter to drive transgene expression, or by use of a viral vector that targets specific cell types (Salmons and Gunzburg, 1993). For example, liver can be targeted by using herpesvirus-based vectors, CD4+ lymphocytes can be targeted with HIV-based vectors, muscle cells can be targeted with tissue-specific promoters or by direct DNA injection, lung cells can be targeted for therapy in cystic fibrosis and lung cancer with inhaled adenoviral vectors (Dunbar and Emmons, 1994; Efstathiou and Minson, 1995; Miller and Boyce, 1995; Trapnell and Gorziglia, 1994). Delivery vehicles can be viral or nonviral (Jolly, 1994; Ledley, 1995). Although viral vectors deliver recombinant DNA to a much higher percentage of cells than do nonviral delivery methods, nonviral methods are useful when a small dose of the therapeutic gene will suffice. These methods include in vivo liposome delivery, direct injection of naked DNA, and transfection of target cells ex vivo (Lee and Klein, 1995). If ex vivo transfection is used, the expressing cell population can be selected, thus increasing the percentage of cells expressing the therapeutic gene. Modified retroviruses are frequently used as vectors in h u m a n gene therapy because they recombine with the genome (Jolly, 1994). Unfortunately, recombination with the genome produces the risk of insertional mutagenesis. This risk is sometimes addressed by inserting a "suicide" gene into the vector in addition to the therapeutic so that cells adversely affected by random integration can be killed (Tiberghien, 1995). Because the retroviruses on which most commonly used vectors are based cannot infect nondividing cells, these vectors are unsuitable for some uses. However, this feature has been exploited to deliver genes in vivo to brain tumors, where the tumor cells are the only dividing cells (Oldfield et al., 1993). Murine retroviral vectors are used to reduce the risk of recombination with endogenous h u m a n sequences. However, modified HIVs are under development for treatment of AIDS patients (Dropulic and Jeang, 1994). To prevent the modified virus from spreading when viral vectors are used for delivery of the therapeutic gene, replication is impaired by removing essential genes from the viral genome (Levine and Friedmann, 1991a). The viral vector is thus capable of only one round of infection. The modified viral genome containing the therapeutic gene is first transfected into helper cells in the laboratory to produce infectious particles. Helper cell lines have been created by stable transfection
GENES INVOLVED IN ONCOGENESIS
89
with virus genomes lacking packaging signals but expressing viral proteins necessary to package the modified viral genome into particles. Because recombination between the helper and vector sequences can regenerate replication-competent genomes, it is now common for all viral vectors to be designed to require multiple recombination events to acquire all the necessary genes for replication, thus minimizing the likelihood of generating a replication-competent virus. Adenovirus is also used as a vector, but it does not recombine with the genome and so the therapeutic gene can be lost upon cell division (Trapnell and Gorziglia, 1994). However, because adenovirus remains episomal, insertional mutagenesis cannot occur. Adenovirus can infect nondividing cells and is thus superior to retroviral vectors in some circumstances. Furthermore, loss of the episomal gene is not a concern in nondividing cells. Inhaled adenovirus vectors are currently used to treat cystic fibrosis. Adeno-associated virus, which can recombine with the genome and infect nondividing cells also shows promise as a vector (Koitin, 1994). Herpesvirus and vaccinia virus also remain episomal and have been used in some protocols to target specific tissue, such as liver and brain (Whitman et al., 1994; Efstathiou and Minson, 1995). Treating a disease caused by a defect in a single gene requires cloning the gene and understanding the function of its product. Depending on the type of disease being treated, three strategies for gene therapy can be adopted. Dysfunctional genes can be replaced, altered to become functional, or augmented with healthy genes working at different loci within the cell or within different cells (Lee and Klein, 1995). Of these options, gene augmentation therapy is currently the most plausible. There are several protocols currently in use to treat singlegene diseases, including cystic fibrosis, Gaucher's syndrome, severe combined immunodeficiency resulting from loss of adenosine deaminase (SCID-ADA), familial hypercholesterolemia, and hemophilia (Colledge,1994; Xu et al., 1994; Blaese et al., 1993; Grossman et al., 1994; Yao et al., 1991). Protocols for treating Duchenne type muscular dystrophy (DMD) as well as other muscular dystrophies are on the horizon (Morgan, 1994). There are no protocols in use to treat such diseases as diabetes, phenylketonuria (PKU), sickle cell anemia, or ~-thalassemia even though the genes that cause these defects are known (Anderson, 1994). This is due in part to technical problems associated with designing some of the required vectors and in part to the diminishing number of individuals that are affected with these diseases. However, it is inevitable that as gene therapy becomes more routine these so-called orphan diseases will be targeted for treatment.
90
CORTNER, VANDEWOUDE, AND VANDEWOUDE
1. Gene Therapy in Oncology The largest number of gene therapy protocols currently underway is in the field of oncology (Roth et al., 1994; Culver and Blaese, 1994). Gene therapy approaches to cancer treatment seek to kill tumor cells directly, mark cells for tracing the origins of recurrent disease, enhance immunity to cancer cells, or protect vulnerable cells from chemotherapy (see Table 4). One of the most ingenious methods used to kill tumor cells directly with gene therapy uses a murine retroviral vector to target the HSVTK cells directly into brain tumor cells (Oldfield et al., 1993). Because the cells surrounding the tumor are not dividing, the virus infects the dividing tumor cells selectively. Once the suicide gene is inserted, the tumor cells are killed by administering the drug ganciclovir. Ganciclovir is a nucleotide analog that is processed only by TK and not by other enzymes involved in nucleotide biosynthesis in normal h u m a n cells. TK action on ganciclovir produces toxic metabolites that kill the cell. Interestingly, neighboring cells that are not expressing TK are also killed, presumably by the movement of toxic metabolites through apoptotic vesicles, tight junctions, and local capillary beds. As a result of this bystander effect, not every cell within the tumor must receive the TK gene to be killed. Additionally, closely associated angiogenic cells within the tumor are killed as well, thereby shutting down the tumor's blood supply and restricting the primary route of metastasis. Some gene therapy protocols are designed to achieve direct cell killing by targeting the tumor cells with tumor-suppressor genes or with antisense oncogenes. For example, expression of the tumor-suppressor gene p53 in p53-deficient non-small-cell lung carcinoma (NSLC) cells has been shown to control their growth (Cai et al., 1993). Similarly, an antisense k-ras gene has been shown to control the growth of k-ras expressing NSLC cells (Zhang et al., 1993). Both of these observations have helped pave the way for the development of gene therapy protocols designed to impose tumor cell death or growth arrest. The marking of cells to trace the source of recurrent disease exemplifies another useful application of gene therapy in oncology. For instance, patients with Acute Myeloid Leukemia (AML) or CML who have undergone irradiation to kill tumor cells require bone marrow transplants to survive (Brenner et al., 1993). Disease-free autologous bone marrow can be used for this purpose, but in the event of relapse, the donor cells are suspect. Marking these cells with a gene such as the neomycin resistance gene allows the researcher to determine whether the donor cells were the source of the recurrent tumor. Several applications of gene therapy are aimed at enhancing immu-
91
GENES INVOLVED IN ONCOGENESIS TABLE 4 GENE THERAPY
ONCOLOGY
Purpose
Tumor
PROTOCOLS
Target Cell
Gene
Marking
AML CML ALL Neuroblastoma AML CML CLL Multiple myeloma Breast cancer Hodgkin's Lymphoma Leukemia Melanoma Ovarian cancer Renal carcinoma
WBM WBM WBM WBM CD34+ CD34 + CD34+ CD34 + CD34 + CD34 + CD34 + CD34+ TIL TIL CD34 +
Neo Neo Neo Neo Neo Neo Neo Neo Neo Neo Neo Neo Neo Neo Neo
Direct Cell Killing
Ovarian cancer Brain tumors Brain tumors
Tumor cells Tumor cells Tumor cells
NSCLC
Tumor cells
HSVtk HSVtk Antisense IGF-1 Antisense
NSCLC
Tumor cells
Wild-type p53
Neuroblastoma SCLC Colorectal
Tumor cells Tumor cells Tumor cells
Solid tumors
Tumor cells
Renal cell carcinoma
Tumor cells
Renal cell carcinoma Malignant melanoma
Fibroblasts Tumor cells
Malignant melanoma Malignant melanoma
T cells Fibroblasts
IL-2 IL-2 IL-2; TNFa; HLA-B7; ~32mg HLA-B7; ~2mg IL-2; TNF~; GM-CSF IL-4 IL-2" TNF~; HLA-B7; ~2mg gIFN TNF~ IL-4
Breast cancer Ovarian cancer
CD34 + CD34 +
MDR- 1 MDR- 1
K-ras
Enhanced Immunity
Chemoprotection
n i t y d i r e c t e d a t t u m o r c e l l s . A l l of t h e s e a p p r o a c h e s a r e b a s e d o n t h e observation that increasing the immune response directed at a subset of t h e t u m o r c e l l s i n c r e a s e s t h e o v e r a l l r e s p o n s e t o t h e e n t i r e t u m o r
92
CORTNER, VANDE WOUDE, AND VANDE WOUDE
(Deisseroth et al., 1993). In one protocol, the HLA gene is inserted into HLA(-) tumor cells to enhance tumor antigenicity (Nabel et al., 1992). In another protocol, cytokine gene is inserted into the tumor cells to act as a sort of tumor vaccination (Tepper and Mule, 1994). In a variation of this approach, genes for cytokines such as tumor necrosis factor (TNF), IL-2, and INF-~ are inserted into tumor infiltrating lymphocytes (TILs) obtained from patients with advanced melanoma and transfused back into those patients (Hwu and Rosenberg, 1994). Some advanced melanomas have responded successfully to these treatments. The efficacy of the TILs bearing the various modifications is currently being assessed. Gene marking experiments with TILs indicate that modified, reintroduced TILs can persist for up to two months, that tumors are targeted by TILs, and that modified TILs are safe. The prospect of protecting vulnerable cells against chemotherapeutic agents is another exciting application of gene therapy. In some breast cancer gene therapy protocols, bone marrow is transduced ex vivo with retroviral vectors carrying copies of the multidrug resistance (MDR) gene (O'Shaughnessy et al., 1994). MDR functions to pump a wide variety of drugs out of cells; its amplification in some tumors results in the failure of chemotherapeutic approaches to treatment. Borrowing on this natural example, reintroduced bone marrow cells that have been modified to express several copies of MDR survive the high doses of chemotherapy needed to treat some aggressive breast cancers, thus increasing the chemotherapeutic dosage that the patient can withstand. C. IMPLICATIONS FOR VETERINARY MEDICINE
As alluded to in previous sections of this article, much of the current understanding of oncogene function has been augmented by the use of animal models. Many cancer-causing viruses, including retroviruses of chickens, cats, cattle, nonhuman primates, and mice (i.e., avian leukosis virus, Rous sarcoma virus, feline and bovine leukemia viruses, Mason Pfizer monkey virus, and murine mammary tumor virus, among others) and DNA viruses such as bovine papilloma virus, Marek's disease, rabbit fibroma virus, and a host of others have been studied in animals (Fenner et al., 1987; Mohanty et al., 1981). This has led to better understanding of oncogene function and pathogenesis and in some cases has aided development of preventative measures and vaccine therapy. To date, few studies have focused on oncogene or tumorsuppressor activation or deactivation in spontaneous or familial tumors arising in animals, though clearly the same mechanisms that
GENES INVOLVED IN ONCOGENESIS
93
have been defined in human disease may be responsible for oncogenesis in animals. It is hoped that future research will concentrate on these aspects of neoplasia in order to provide better diagnostic, prognostic, and therapeutic capabilities to veterinary clinicians.
VII. Conclusion
During the past decade, great insights have been made into the mechanisms surrounding the pathogenesis of cancer. Such strides have been aided by study of DNA-mediated gene transfer, oncogenic DNA and RNA viruses in animals, and evaluation of genes and chromosomes in families with apparently heritable forms of cancer. New molecular biology methodologies, including gene mapping and cloning, as well as the ability to induce mutations in mice through the use of transgenic and knock-out technology, have contributed significantly to these studies. Through this work, four categories of genes have been identified that are involved in oncogenesis: (1) oncogenes, which normally function to promote cellular growth and division in a programmed manner, but which can trigger unregulated proliferation when some aspect of their function or expression is altered; (2) tumorsuppressor genes, which normally act to suppress cellular growth and which can lead to uncontrolled proliferation when inactivated; (3) genes encoding DNA repair enzymes, which when dysfunctional lead to increased accumulation of mutations in cancer-predisposing genes; and (4) metastasis-associated genes, which when activated allow cancer cells to disseminate and establish tumors in sites peripheral to the original neoplasm. Establishment of the genetic basis of oncogenesis has provided an opportunity for new, highly accurate diagnostic methodologies to be developed and hopefully will result in improved methods of prevention and treatment for neoplastic disease. In addition, these discoveries, made possible by the use of animal models, may also ultimately be used to improve understanding and treatment of neoplasms commonly encountered in companion and large animal veterinary medicine. REFERENCES Aaronson, S. A. (1991). Growth factors and cancer. Science 254, 1146-1153. Anderson, W. F. (1994). Gene therapy for genetic diseases. Hum. Gene Ther. 5, 281-282. Avraham, K. B., Hasson, T., Steel, K. P., Kingsley, D. M., Russell, L. B., Mooseker, M. S., Copeland, N. G., and Jenkins, N. A. (1995). The mouse Snell's waltzer deafness gene
94
CORTNER, VANDE WOUDE, AND VANDE WOUDE
encodes an unconventional myosin required for structural integrity of inner ear hair cells. Nat. Genet. 1, 369-375. Barinaga, M. (1994). Knockout mice: round two. Science 265, 26-28. Benz, C. C., and Liu, E. T., eds. (1993). "Oncogenes and tumor suppressor genes in human malignancies," 2nd edition. Kluwer Academic Publishers, Boston. Bishop, J. M. (1985). Cellular oncogenes and retroviruses. Annu. Rev. Biochem. 52, 301354. Bishop, J. M. (1991). Molecular themes in oncogenesis. Cell 64, 235-248. Bladt, F., Riethmacher, D., Isenmann, S., Aguzzi, A., Birchmeier, C. (1995). Essential role for the c-met receptor in the migration of myogenic precursor cells into the limb bud. Nature 376, 768-771. Blaese, R. M. (1993). Development of gene therapy for immunodeficiency: Adenosine deaminase deficiency. Pediatr. Res. 33, $49-$55. Blair, D. G., Cooper, C. S., Oskarsson, M. K., Eader, L. A., and Vande Woude, G. F. (1982). New method for detecting cellular transforming genes. Science 218, 1122-1125. Bookstein, R., Rio, P., Madreperla, S. A., Hong, F., Allred, C., Grizzle, W. E., and Lee, W. H. (1990). Promoter deletion and loss of retinoblastoma gene expression in human prostate carcinoma. Proc. Natl. Acad. Sci. USA 87, 7762-7766. Bos, J. L., Fearon, E. R., Hamilton, S. R., Verlaan-de Vries, M., van Boom, J.H., van der Eb, A. J., and Vogelstein, B. (1987). Prevalence of ras gene mutations in human colorectal cancers. Nature 327, 293-297. Bottaro, D. P., Rubin, J. S., Faletto, D. L., Chan, A.M.-L., Kmiecik, T. E., Vande Woude, G. F., and Aaronson, S. A. (1991) Identification of the hepatocyte growth factor receptor as the c-met proto-oncogene product. Science 251, 802-804. Breneman, J. W., Swiger, R. R., Ramsey, M. J., Minkler, J. L., Eveleth, J. G., Langlois, R. A., Tucker, J. D. (1995). The development of painting probes for dual-color and multiple chromosome analysis in the mouse. Cytogenet. Cell Genet. 68, 197-202. Brenner, M. K., Rill, D. R., Moen, R. C., Krance, R. A., Mirro, J., Jr., Anderson, W. F., and Ihle, J. N. (1993). Gene-marking to trace origin of relapse after autologous bonemarrow transplantation. Lancet 341, 85-86. Buchkovich, K., Duffy, L. A., Harlow, E. (1989). The retinoblastoma protein is phosphorylated during specific phases of the cell cycle. Cell 58, 1097-1105. Buckler, A. J., Chang, D. D., Graw, S. L. Brook, J. D., Haber, D. A., Sharp, P. A., and Housman, D. E. (1991). Exon amplification: A strategy to isolate mammalian genes based on RNA splicing. Proc. Natl. Acad. Sci. USA 88, 4005-4009. Bultman, S. J., Michaud, E. J., Woychik, R. P. (1992). Molecular characterization of the mouse agouti locus. Cell 71, 1195-1204. Cai, D. W., Mukhopadhyay, T., Liu, Y., Fujiwara, T., and Roth, J. A. (1993). Stable expression of the wild-type p53 gene in human lung cancer cells after retrovirusmediated gene transfer. Hum. Gene Ther. 4, 617-624. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A., Kapeller, R., and Soltoff, S. (1991). Oncogenes and signal transduction. Cell 64, 281-302. Capecchi, M. R. (1989). Altering the genome by homologous recombination. Science 244, 1288-1292. Cavenee, W. K., Dryja, T. P., Phillips, R. A., Benedict, W. F., Godbout, R., Gallie, B. L., Murphree, A. L., Strong, L. C., and White, R. L. (1983). Expression of recessive alleles by chromosomal mechanisms in retinoblastoma. Nature 305, 779-784. Chen, P. L., Scully, P., Shew, J. Y., Wang, J. Y., Lee, W. H. (1989). Phosphorylation of the retinoblastoma gene product is modulated during the cell cycle and during cellular differentiation. Cell 58, 1193-1198.
GENES INVOLVED IN ONCOGENESIS
95
Cleary, M. L. (1991). Oncogenic conversion of transcription factors by chromosomal translocaton. Cell 66, 619-622. Colledge, W. H. (1994). Cystic fibrosis gene therapy. Current Opinion Genet. Dev. 4, 466471. Collins, F. S. (1992). Positional cloning: Let's not call it reverse anymore. Nat. Genet. 1, 3-6. Collins, S. J., Kubonishi, I., Miyoshi, I., and Groudine, M. T. (1984). Altered transcription of the c-abl oncogene in K-562 and other chronic myelogenous leukemia cells. Science 255, 72-74. Copeland, N. G., Jenkins, N. A., Gilbert, D. J. Eppig, J. T., Maltais, L. J., Miller, J. C., Dietrich, W. F., Weaver, A., Lincoln, S. E., Steen, R. G., Stein, L. D., Nadeau, J. H., and Lander, E. S. (1993). A genetic linkage map of the mouse: Current applications and future prospects. Science 262, 57-66. Cortner, J. and Vande Woude, G. F. (1996). Essentials of molecular biology. In "Cancer: Principles and Practice of Oncoogy," 5th edition (V. T. DeVita, S. Hellman, and S. Rosenberg, eds.), J.B. Lippincott Co., Philadelphia, 3-34. Cory, S., Harris, A. W., and Strasser, A. (1994). Insights from transgenic mice regarding the role of bcl-2 in normal and neoplastic lymphoid cells. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 345, 289-395. Courtneidge, S. A. (1985). Activation of the pp6Oc-src kinase by middle T antigen binding or by desphosphorylation. E M B O J. 4, 1471-1477. Culver, K. W., and Blaese, R. M. (1994). Gene therapy for cancer. Trends Genet 10, 174178. Curran, T., Miller, A. D., Zokas, L., and Verma, I. M. (1984). Viral and cellular fos proteins: A comparative analysis. Cell 36, 259-268. Dalla Favera, R., Wong-Staal, F., and Gallo, R. C. (1982). Onc gene amplification in promyelocyte leukaemia cell line HL-60 and primary leukaemic cells of the same patient. Nature 299, 61-63. Deisseroth, A. B., Kavanagh, J. J., Hanania, E. G., et al. (1993). Gene therapy: Chemoprotection, immunoenhancement and modification of tumor cells. Cancer Bull. 45, 139. DeCaprio, J. A., Ludlow, J. W., Figge, J., Shew, J. Y., Huang, C. M., Lee, W. H., Marsilio, E., Paucha, E., and Livingston, D. M. (1988). SV40 large tumor antigen forms a specific complex with the product of the retinoblastoma susceptibility gene. Cell 54, 275-283. De Caprio, J. A., Ludlow, J. W., Lynch D., et al. (1989). The product of the retinoblastoma susceptibility gene has properties of a cell cycle regulatory element. Cell 58, 1085. DeMars, R. (1970). Fundamental cancer research: 23rd annual symposium, p. 105. William & Wikins, Baltimore. Der, C. J., Krontiris, T. G., and Cooper, G. M. (1982). Transforming genes of human bladder and lung carcinoma cell lines are homologous to the ras genes of Harvey and Kirsten sarcoma viruses. Proc. Natl. Acad. Sci. USA 79, 3637-3640. de The, H., Chomienne, C., Lanotte, M., Degos, L., and Dejean, A. (1990). The t(15;17) translocation of acute promyelocytic leukaemia fuses the retinoic acid receptor a gene to a novel transcribed locus. Nature 347, 558-561. Donehower, L. A., Godley, L. A., Aldaz, C. M., Pyle, R., Shi, Y. P., Pinkel, D., Gray, J., Bradley, A., Medina, A., and Varmus, H. E. (1995). Deficiency of p53 accelerates mammary tumorigenesis in Wnt-1 transgenic mice and promotes chromosomal instability. Genes Dev. 9, 882-895. Dropulic, B., and Jeang, K. T. (1994). Gene therapy for human immunodeficiency virus
96
CORTNER, VANDE WOUDE, AND
VANDE WOUDE
infection: Genetic antiviral strategies and targets for intervention. Hum. Gene Ther. 5, 927-939. Dunbar, C. E., and Emmons, R.V. (1994). Gene transfer into hematopoietic progenitor and stem cells. Stem Cells 12, 563-576. Dunlop, M. G., Wyllie, A. H., Nakamura, Y., Steel, C. M., Evans, H. J., White, R. L., Bird, C. C. (1990). Genetic linkage map of six polymorphic DNA markers around the gene for familial adenomatous polyposis on chromosome 5. Am. J. Hum. Genet. 47, 982987. Echelard, Y., Vassileva, G., McMahon, A. P. (1994). Cis-acting regulatory sequences governing Wnt-1 expression in the developing mouse CNS. Development 120, 22132224. Efstathiou, S., and Minson A. C. (1995). Herpes virus-based vectors. Br. Med. Bull. 51, 45-55. Fearon, E. R. (1995). Oncogenes and tumor suppressor genes. In: "Clinical Oncology" (M. D. Abeloff, J. O. Armitage, A. S. Lichter, and J. E. Niederhuber, eds.), pp. 11-40. Churchill Livingstone, New York. Fenner, F. J., Gibbs, P. J., Murphy, F. A., Rott, R., Studdert, M. J., and White, D. O. (eds.) (1987). Mechanisms of viral oncogenesis In 'Veterinary Virology," 2nd edition, pp. 193-212. Academic Press, London. Fidler, I. J. (1995). Invasion and metastasis. In: "Clinical Oncology" (M. D. Abeloff, J. O. Armitage, A. S. Lichter, and J. E. Niederhuber, eds), pp. 55-76. Churchill Livingstone, New York. Fidler, I. J., Poste, G. (1985). The cellular heterogenecity of malignant neoplasms: Implications for adjuvant chemotherapy. Semin. Oncol. 12, 207-221. Friedberg, E. C. (1992). Xeroderma pigmentosum, Cockayne's syndrome, helicases, and DNA repair: What's the relationship? Cell 71, 887-889. Friend, S. H., Dryja, T. P., and Weinberg, R. A. (1988). Oncogenes and tumor supressing genes. N. Engl. J. Med. 318, 618. Galli-Taliadoros, L. A., Sedgwick, J. D., Wood, S. A., and Korner, H. (1995). Gene knockout technology: A methodological overview for the interested novice. J. Immunol. Methods 181, 1-15. Gatti, R. A., Boder, E., Vinters, H. V., Sparkes, R. S., Norman, A., and Lange K. (1991). Ataxia-telangiectasia: an interdisciplinary approach to pathogenesis. Medicine (Baltimore) 70, 99-117. Goldfarb, M., Shimizu, K., Perucho, M., and Wigler, M. (1982). Isolation and preliminary characterization of a human transforming gene from T24 bladder carcinoma cells. Nature 296, 404-409 Grant, D. S., Kleinman, H. K., Goldberg, I. D., Bhargava, M. M., Nickloff, B. J., Kinsella, J. L., Polverini, P., and Rosen, E. M. (1993). Scatter factor induces blood vessel formation in vivo. Proc. Natl. Acad. Sci. USA 90, 1937-1941. Grossman, M., Raper, S. E., Kozarsky, K., Stein, E. A., Engelhardt, J. F., Muller, D., Lupien, P. J., and Wilson, J. M. (1994). Successful ex vivo gene therapy directed to liver in a patient with familial hypercholesterolaemia. Nat. Genet. 6, 323-324,. Gu, H., Martin, J. D., Orban, P. C., Mossmann, H., and Rajewsky, K. (1994). Deletion of a DNA polymerase ~ gene segment in T cells using cell type-specific gene targeting. Science 265, 103-106. Haber, D. A., and Housman, D. E. (1992). The genetics of Wilms' tumor. Adv. Cancer Res. 59, 41-68. Hamilton, S. R. (1992). Molecular genetics of colorectal carcinoma. Cancer 70, 12161221.
GENES INVOLVED IN ONCOGENESIS
97
Hanks, S. K. (1991). Eukaryotic protein kinases. Current Opinion Struc. Biol. 1, 369-384. Harper, M. E., Saunders, G. F. (1981). Localization of single copy DNA sequences of G-banded human chromosomes by in situ hybridization. Chromosoma 83, 431-439. Hayward, W. S., Neel, B. G., and Astrin, S. M. (1981). Activation of a cellular ONC gene by promoter insertion in ALB-induced lymphoid leukosis. Nature 290, 475-480. Heim, R. A., Lench, N. J., and Swift, M. (1992). Heterozygous manifestations in four autosomal recessive human cancer-prone syndromes: Ataxia telangiectasia xeroderma pigmentosum, Faconi anemia, and Bloom syndrome. Mutat. Res. 284, 25-36. Heisterkamp, N., Jenster, G., ten Hoeve, J., Zovich, D., Pattengale, P. K., and Groffen, J. (1990). Acute leukaemia in bcr/abl transgenic mice. Nature 344, 251-253. Herrmann, B. G., Labeit, S., Poustka, A., King, T. R., and Lehrach, H. (1990). Cloning of the T gene required in mesoderm formation in the mouse. Nature 343, 617-622. 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. Hill, M., and Hillova, J. (1972). Virus recovery in chicken cells tested with Rous sarcoma cell DNA. Nature New Biol. 237, 35-39. Hockenberry, D. M. (1994). bcl-2 in cancer, development and apoptosis. J. Cell Sci. Suppl. 18, 51-55. Hogan, B., Beddington, R., Constantini, F., and Lacey, E. (1994). "Manipulating the Mouse Embryo: A Laboratory Manual," Cold Spring Harbor Laboratory Press, Plainview, NY. Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C.C. (1991). p53 mutations in human cancers. Science 253, 49-53. Howley, P. (1993). Principles of carcinogenesis. In: Cancer: Principles & Practice of Oncology," 4th edition, (V. T. DeVita, S. Hellman, and S. Rosenberg, eds.), pp. 182199. J.B. Lippincott, Philadelphia. Hughes, S. H. (1991). Will retroviral vectors be used to make transgenic livestock? Agric. Biotechnol. News Info. 3, 25-28. Hwu, P., and Rosenberg, S. A. (1994). The genetic modification of T cells for cancer therapy: An overview of laboratory and clinical trials. Cancer Detect. Prev. 18, 43-50. Jacks, T., Shih, T. S., Schmitt, E. M., Bronson, R. T., Bernards, A., and Weinberg, R.A. (1994). Tumour predisposition in mice heterozygous for a targeted mutation in Nfl. Nat. Genet. 7, 353-361 Jeffers, M., Rong, S., and Vande Woude, G. F. (1996a). Enhanced tumorigenicity and invasion/metastasis by HGF/SF-Met signalling in human cells concomitant with induction of the urokinase proteolysis network. Mol. Cell Biol. 16, 1115-1125. Jeffers, M., Rong, S., and Vande Woude, G. F. (1996b). Hepatocyte growth factor/scatter factor-Met signalling in tumorigenicity and invasion/metastasis. J. Mol. Med., in press. Johnson, R. S., Spiegelman, B. M., and Papaioannou, V. (1992). Pleiotropic effects of a null mutation in the c-fos proto-oncogene. Cell 71, 577-586. Jolly, D. (1994). Viral vector systems for gene therapy. Cancer Gene Ther. 1, 51-64. Kieff, E. (1995). Tumor viruses. In: "Clinical Oncology" (M. D. Abeloff, J. O. Armitage, A. S. Lichter, and J. E. Niederhuber, eds.), pp. 135-150. Churchill Livingstone, New York. King, M. C., Rowell, S., and Love, S. M. (1993). Inherited breast and ovarian cancer. What are the risks? What are the choices? J A M A 269, 1975-1980. Kingsley, D. M., Bland, A. E., Grubber, J. M., Marker, P. C., Russell, L. B., Copeland, N. G., and Jenkins, N. A. (1992). The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF beta superfamily. Cell 71, 399-410.
98
CORTNER, VANDE WOUDE, AND VANDE WOUDE
Klein, G., and Klein, E. (1985). Evolution of tumours and the impact of molecular oncology. Nature 315, 190-195. Knudson, A. G., Jr. (1971). Mutation and cancer: Statistical study of retinoblastoma. Proc. Natl. Acad. Sci. USA 68, 820-823. Kohn, D. B., Anderson, W. F., and Blaese, R. M. (1989). Gene therapy for genetic diseases. Cancer Invest. 7, 179-192. Koitin, R. M. (1994). Prospects for the use of adeno-associated virus as a vector for human gene therapy. Gene Ther. 5, 793-801. Korsmeyer, S. J., Shutter, J. R., Veis, D. J., Merry, D. E., and Oltvai, Z. N. (1993). Bcl-2/Bax: A rheostat that regulates an anti-oxidant pathway and cell death. Semin. Cancer Biol. 4, 327-332. Kreidberg, J. A., Sariola, H., Loring, J. M., Maeda, M., Pelletier, J., Housman, D., and Jaenisch, R. (1993). WT-1 is required for early kidney development. Cell 74, 679-691. Kunkel, L. M., Beggs, A. H., Hoffman, E. P. (1989). Molecular genetics of Duchenne and Becker muscular dystrophy: Emphasis on improved diagnosis. Clin. Chem. 35, B2124. Ledley, F. D. (1995). Nonviral gene therapy: The promise of genes as pharmaceutical products. Hum. Gene Ther. 6, 1129-1144. Lee, E. Y., Chang, C. Y., Hu, N., Wang, Y. C., Lai, C. C., Herrup, K., Lee, W. H., and Bradley, A. (1992). Mice deficient for Rb are nonviable and show defects in neurogenesis and haematopoiesis. Nature 359, 288-294. Lee, J. H., and Klein, H. G. (1995). Cellular gene therapy. Hematol. Oncol. Clin. North Am. 9, 91-113. Lee, J. M., Abrahamson, J. L., and Bernstein, A. (1994). DNA damage, oncogenesis and the p53 tumour-suppressor gene. Mutat. Res. 307, 573-581. Lee, W. H., Bookstein, R., Hong, F., Young, L. J., Shew, J. Y., and Lee, E. Y. (1987). Human retinoblastoma, susceptibility gene: Cloning, identification and sequence. Science 235, 1394-1399. Levine, A. J., Momand, J., and Finlay, C. A. (1991b). The p53 tumour suppressor gene. Nature 351, 453-456. Levine, F., Yee, J. K., and Friedmann, T. (1991a). Efficient gene expression in mammalian cells from a dicistronic transcriptional unit in an improved retroviral vector. Gene 108, 167-174. Lewin, B. (1994). Genes V, 5th ed. Oxford Press, Cary, NC. Lisitsyn, N., Lisitsyn, N., and Wigler, M. (1993). Cloning the differences between two complex genomes. Science 259, 946-951. Little, C. D., Nau, M. M., Carney, D. N., Gazdar, A. F., and Minna, J. D. (1983). Amplification and expression of the c-myc oncogene in human lung cell lines. Nature 306, 194-196. Look, T. A. (1995). Oncogenic role of "master" transcription factors in human leukemias and sarcomas: A developmental model. In "Advances in Cancer Research," Vol. 67 (G. Vande Woude and G. Klein, eds.), pp. 25-76. Academic Press, San Diego. Majeras, P. W., Ross, T. S., Cunningham, T. W., Caldwell, K. K., Jefferson, A. B., and Basal, V. S. (1990). Recent insights in phosphatidylinositol signaling. Cell 63, 459465. Malkin, D., Li, F. P., Strong, L. C., Fraumeni, J. F., Jr., Nelson, C. E., Kim, D. H., Kassel, J., Gryka, M. A., Bischoff, F. Z., Tainsky, M. A., et al. (1990). Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250, 1233-1238. Matten, W., and Vande Woude, G. F. (1995). Microinjection into Xenopus oocytes.
GENES INVOLVED IN ONCOGENESIS
99
In "Methods in Enzymology: Oncogene Techniques," Vol. 254 (P. K. Vogt and I. M.
Verma, eds.), pp. 458-466. Academic Press, New York. Matzuk, M. M., and Bradley, A. (1994). Identification and analysis of tumor suppressor genes using transgenic mouse models. Semin. Cancer Biol. 5, 37-45. Miki, Y., Swensen, J., Shattuck-Eidens, D., Futreal, P. A., Harshman, K., Tavtigian S., Liu, Q., Cochran, C., Bennett, L. M., Ding, W., et al. (1994). A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 266, 66-71. Miller, J. B., and Boyce, F. M. (1995). Gene therapy by and for muscle cells. Trends Genet. 11, 163-165. Mohanty, S. B., and Dutta, S. K., eds. (1981). Tumor (oncogenic) viruses. In '"Veterinary Virology," pp 310-339. Lea and Febiger, Philadelphia. Momand, J., Zambetti G. P., Olson D. C., George, D., and Levine A. J. (1992). The m d m 2 oncogene product forms a complex with the p53 protein and inhibits p53 mediated transactivation. Cell 69, 1237-1245. Morgan, J. E. (1994). Cell and gene therapy in Duchenne muscular dystrophy. Hum. Gene Ther. 5, 165-173. Morgenbesser, S. D., and DePinho, R. A. (1994). Use of transgenic mice to study myc family gene function in normal mammalian development and in cancer. Semin. Cancer Biol. 5, 21-36. Moser, A. R., Pitot, H. C., and Dove, W. F. (1990). A dominant mutation that predisposes to multiple intestinal neoplasms in the mouse. Science 247, 322-324. Murray, M. J., Shilo, B.-Z., Shih, C., Cowing, D., Hsu, H. W., and Weinberg, R. A. /1981). Three different human tumor cell lines contain different oncogenes. Cell 25, 355361. Nabel, G. J., Chang, A., Nabel, E. G., and Plautz, G. (1992). Immunotherapy of malignancies by in vivo genes transfer into tumors. Hum. Gene Ther. 3, 399-410. Nevins, J. R. (1992). E2F: A link between the Rb tumor suppressor protein and viral oncoproteins. Science 258, 424-429. Oldfield, E. H., Ram, Z., Culver, K. W., Blaese, R. M., DeVroom, H. L., and Anderson, W. F. (1993). Gene therapy for the treatment of brain tumors using intra-tumoral transduction with the thymidine kinase gene and intravenous ganciclovir. Hum. Gene Ther. 4, 39-69. Ostrer, H., Allen, W., Crandall, L. A., Moseley, R. E., Dewar, M. A., Nye, D., and McCrary, S. V. (1993). Insurance and genetic testing: Where are we now? Am. J. Hum. Genet. 52, 565-577. O'Shaughnessy, J. A., Cowan, K. H., Nienhuis, A. W., McDonagh, K. T., Sorrentino, B. P., Dunbar, C. E., Chiang, Y., Wilson, W., Goldspiel, B., Kohler, D., et al. (1994). Retroviral mediated transfer of the human multidrug resistance gene (MDR-1) into hematopoietic stem cells during autologous transplantation after intensive chemotherapy for metastatic breast cancer. Hum. Gene Ther. 5, 891-911. Pendergast, A. M., Muller, A. J., Havlik, M. H., Maru, Y., and Witte, O. N. (1991). BCR sequences essential for transformation by the BCR / A B L oncogene bind to ABL SH2 regulatory domain in a nonphosphotyrosine-dependent manner. Cell 66, 161-171. Perkins, A. S., and Vande Woude, G. F. (1993). Principles of molecular cellular biology of cancer: Oncogenes. In "Cancer: Principles and Practices of Oncology," 4th Edition (V. De Vita, S. Hellman, and S. A. Rosenberg, eds.), pp. 35-59. J.B. Lippincott Co., Philadelphia. Piper, J., Rutovitz, D., Sudar, D., Kallioniemi, A., Kallioniemi, O. P., Waldman, F. M., Gray, J. W., and Pinkel, D. (1995). Computer image analysis of comparative genomic hybridization. Cytometry 19, 10-26.
100
CORTNER, VANDE WOUDE, AND VANDE WOUDE
Rauscher, F. J. (1993). The WT1 Wilms' tumor gene product: A developmentally regulated transcription factor in the kidney that functions as a tumor suppressor. FASEB 7, 896. Reddy, E. P., Skalka, A. M., and Curran, T., eds. (1988). "The Oncogene Handbook." Elsevier Science Pub. Co., New York. Rickinson, A. B. (1994). EBV infection and EBV-associated tumours. In: '~Viruses and Cancer," Vol. 51 (A. Minson, J. Nell, and M. McCrae, eds.), pp. 81-100. University Press, Cambridge. Rijsewijk, F., Schuermann, M., Wagenaar, E., Parren, P., Weigel, D., and Nusse, R. (1987). The Drosophila homolog of the mouse mammary oncogene int-1 is identical to the segment polarity gene wingless. Cell 50, 649-657. Rommens, J. M., Iannuzzi, M. C., Kerem, B., Drumm, M. L., Melmer, G., Dean, M., Rozmahel, R., Cole, J. L., Kennedy, D., Hidaka, N., et al. (1989). Identification of the cystic fibrosis gene: Chromosome walking and jumping. Science 245, 1059-1065. Rong, S., Bodescot, M., Blair, D., Dunn, J., Nakamura, T., Mizuno, K., Park, M., Chan, A., Aaronson, S., and Vande Woude, G. F. (1992). Tumorigenicity of the met protooncogene and the gene for hepatocyte growth factor. Mol. Cell. Biol. 12, 5152-5158. Rong, S., Donehower, L. A., Hansen, M. F., Strong, L., Tainsky, M., Jeffers, M., Resau, J. H., Hudson, E., Tsarfaty, I., and Vande Woude, G. F. (1995). Met proto-oncogene product is overexpressed in tumors of p53-deficient mice and tumors of Li-Fraumeni patients. Cancer Res. 55, 1963-1970. Rong, S., Jeffers, M., Resau, J. H., Tsarfaty, I., Oskarsson, M. and Vande Woude, G. F. (1993). Met expression and sarcoma tumorigenicity. Cancer Res. 53, 5355-5360. Rong, S., Segal, S., Anver, M., Resau, J. H., and Vande Woude, G. F. (1994). Invasiveness and metastasis of NIH/3T3 cells induced by Met-HGF/SF autocrine stimulation. Proc. Natl. Acad. Sci. USA 91, 4731-4735. Rosen, E. M., Nigam, S. K., and Goldberg, I. D. (1994). Scatter factor and the c-met receptor: A paradigm for mesenchymal/epithelial interaction. J. Cell Biol. 127, 17831787. Roth, J. A., Mukhopadhyay, T., Zhang, W. W., Fujiwara, T., and Georges, R. (1994). Gene replacement strategies for the prevention and therapy of cancer. Eur. J. Cancer 30A, 2032-3037. Rubin, J. S., Bottaro, D. P., and Aaronson, S. A. (1993). Hepatocyte growth factor/scatter factor and its receptor, the c-met proto-oncogene product. Biochim. Biophys. Acta 1155, 357-371. Salmons, B., and Gunzburg, W. H. (1993). Targeting of retroviral vectors for gene therapy. Hum. Gene Ther. 4, 129-141. Sambrook, J., Fritch, E. F., Maniatis, T. (1989). "Molecular Cloning: A Laboratory Manual," ed. 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Sands, A., Donehower, L. A., and Bradley, A. (1994). Gene-targeting and the p53 tumorsuppressor gene. Mutat. Res. 307, 557-572. Schalling, M., Ekman, M., Kaaya, E. E., Linde, A., and Biberfeld, P. (1995). A role for a new herpes virus (KSHV) in different forms of Kaposi's sarcoma. Nat. Med. 1, 707708 Schmidt, C., Bladt, F., Goedecke, S., Brinkmann, V., Zschiesche, W., Sharpe, M., Gherardi, E., and Birchmeier, C. (1995). Scatter factor/hepatocyte growth factor is essential for liver development. Nature 373, 699-702. Shilo, B.-Z., and Weinberg, R. A. (1981). DNA sequences homologous to vertebrate oncogenes are conserved in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 78, 6789-6792.
GENES INVOLVED IN ONCOGENESIS
101
Shiramizu, B., Herndier, B. G., and McGrath, M. S. (1994). Identification of a common clonal human immunodeficiency virus integration site in human immunodeficiency virus-associated lymphomas. Cancer Res. 54, 1069-1072. Shuyer, M., Ab, G., Brinkmann, A. O., and Blankenstein, R. A. (1993). "Zinc-finger Proteins in Oncogenesis: DNA-binding and Gene Regulation." New York Academy of Sciences, New York. Stanton, B. R., Perkins, A. S., Tessarolo, L., Sassoon, D. A., and Parada, L. F. (1992). Loss of N-myc function results in embryonic lethality and failure of the epithelial component of the embryo to develop. Genes Dev. 6, 2235-2247 Steinmetz, M., Moore, K. W., Frelinger, J. G., Sher, B. T., Shen, F. W., Boyse, E. A., and Hood, L. (1981). A pseudogene homologous to mouse transplantation antigens: Transplantation antigens are encoded by eight exons that correlate with protein domains. Cell 25, 683-692. Stewart, S. A., Poon, B., and Chen, I.S.Y. (1994). Mechanism of HTLV leukemogenesis. In: "Viruses and Cancer" (A. Minson, J. Neil, and M. McCrae, eds.), pp. 189-212. University Press, Cambridge. Strathdee, C. A., Gavish, H., Shannon, W. R., and Buchwald, M. (1992). Cloning of cDNAs for Franconi's anaemia by functional complementation. Nature 356, 763-767. Su, L. K., Vogelstein, B., and Kinzler, K. W. (1993). Association of the APC tumor suppressor protein with catenins. Science 262, 1734-1737. Sucov, H. M., Dyson, E., Gumeringer, C. L., Price, J., Chien, K. R., and Evans, R. M. (1994). RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007-1018. Tepper, R. I., and Mule, J. J. (1994). Experimental and clinical studies of cytokine genemodified tumor cells. Hum. Gene Ther. 5, 153-164. The, I., Murthy, A. E., Hannigan, G. E., et al. (1993). Neurofibromatosis type1 gene mutations in neuroblastoma. Nat. Genet. 3, 62-66. Thibodeau, S. N., Bren, G., and Schaid, D. (1993). Microsatellite instability in cancer of the proximal colon. Science 260, 816-819. Thompson, C. T., and Gray, J. W. (1993). Cytogenetic profiling using fluorescence in situ hybridization (FISH) and comparative genomic hybridization (CGH). J. Cell Biochem. Suppl. 17G, 139-143. Tiberghien, P. (1995). Use of suicide genes in gene therapy. J. Leukocyte Biol. 56, 203209. Trapnell, B. C., and Gorziglia, M. (1994). Gene therapy using adenoviral vectors. Current Opinion Biotechnol. 5, 617-625. Trent, J. M. (1995). Cytogenetics. In: "Clinical Oncology" (M. D. Abeloff, J. O. Armitage, A. S. Lichter, and J. E. Niederhuber, eds.), pp. 41-54. Churchill Livingstone, New York. Trofatter, J., MacCollin, M., Rutter, J., et al. (1993). A novel myosin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 suppressor gene. Cell 72, 791. van Heyningen, V. (1994). One gene--four syndromes. Nature 367, 319-320. Varmus, H. (1988). Retroviruses. Science 240, 1427-1435. Vidal, S. M., Malo, D., Vogan, K., Skamene, E., Gros, P. (1993). Natural resistance to infection with intracellular parasites: Isolation of a candidate for Bcg. Cell 73, 469485. Wallace, M. R., Marchuk, D. A., Andersen, L. B., Letcher, R., Odeh, H. M., Saulino, A. M., Fountain, J. W., Brereton, A., Nicholson, J., Mitchell, A. L., et al. (1990). Type 1 neurofibromatosis gene: Identification of a large transcript disrupted in three NF1 patients. Science 249, 181-186.
102
CORTNER, VANDE WOUDE, AND VANDE WOUDE
Watson, J. D., Tooze, J., and Kurtz, D. T. (1983). "Recombinant DNA: A short Course." Scientific American Books, New York. Weigel, R. J. (1995). Inherited cancer. In: "Clinical Oncology" (M. D. Abeloff, J. O. Armitage, A. S. Lichter, and J. E. Niederhuber, eds.), pp. 167-186. Churchill Livingstone, New York. Weiss, R., Teich, N., Varmus, H., et al. (1982). "RNA Tumor Viruses." Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Whyte, P., Buchkovick, K. J., Horowitz, J. M., et al. (1988). Association between an oncogene and an antioncogene: The adenovirus EIA protein binds to the retinoblastoma gene product. Nature 334, 124. Weller, I.V.D. (1994). HIV and predisposition to cancer. In: '~Viruses and Cancer," Vol. 51 (A. Minson, J. Neil, and M. McCrae, eds.), pp. 293-306. University Press, Cambridge. Whitman, E. D., Tsung, K., Paxson, J., and Norton, J. A. (1994). In vitro and in vivo kinetics of recombinant vaccinia virus cancer-gene therapy. Surgery 11{}, 183-188. Whyte P., Williamson, N. M., Harlow, E. (1989). Cellular targets for transformation by the adenovirus EIA proteins. Cell 56, 67-75. Wooster, R., Neuhausen, S. L., Mangion, J., Quirk, Y., Ford, D., Collins, N., Nguyen, K., Seal, S., Tran, T., Averill, D., et al. (1994). Localization of a breast cancer susceptibility gene, BRCA2, to chromosome 13q12-13. Science 265, 2088-2090. Xu, G. F., O'Connell, P., Viskochil, D., Cawthen, R., Robertson, M., Calvert, M., Dunn, D.,Stevens, J., Gesteland, R., White, R., et al. (1990). The neurofibromatosis type 1 gene encodes a protein related to GAP. Cell 62, 599-608. Xu, L., Stahl, S. K., Dave, H. P., Schiffmann, R., Correll, P. H., Kessler, S., and Karlsson, S. (1994). Correction of the enzyme deficiency in hematopoietic cells of Gaucher patients using a clinically acceptable retroviral supernatant transduction protocol. Exp. Hematol. 22, 223-230. Yao, S. N., Wilson, J. M., Nabel, E. G., Kurachi, S., Hachiya, H. L., and Kurachi, K. (1991). Expression of human factor IX in rat capillary endothelial cells toward somatic gene therapy for hemophilia B. Proc. Natl. Acad. Sci. USA 88, 8101-8105. Yokota, J., Wada, M., Shimosato, Y., Terada, M., Sugimura, T. (1987). Loss of heterozygosity on chromosomes 3,13 and 17 in small-cell carcinoma and on chromosomes 3 in adenocarcinoma of the lung. Proc. Natl. Acad. Sci. USA 84, 9252-9256. Zhang, Y., Mukhopadhyay, T., Donehower, L. A., Georges, R. N., and Roth, J. A~ (1993). Retroviral vector-mediated transduction of K-ras antisense RNA into human lung cancer cells inhibits expression of the malignant phenotype. Hum. Gene Ther. 4, 145-460.
ADVANCESIN VETERINARYMEDICINE,VOL.40
Diagnostic Molecular Genetics GEORGE H. SACK, JR. Departments of Medicine, Pediatrics, and Biological Chemistry, The Johns .Hopkins University School of Medicine, The Johns Hopkins Hospital, Baltimore, Maryland 21287 I. Introduction II. Methods A. Restriction Fragment Length Polymorphisms B. Polymerase Chain Reaction C. Mapping III. Linkage Analysis A. Background B. Positional Cloning C. Candidate Genes D. Quantitative Trait Loci IV. Single-Gene Disorders A. Canine Mutations B. Canine Genetic Homogeneity C. Technical Considerations V. Future Directions Acknowledgments References
I. I n t r o d u c t i o n D e v e l o p m e n t s in m o l e c u l a r g e n e t i c s a r e e n l a r g i n g t h e scope of app r o a c h e s to q u e s t i o n s i n v o l v i n g i n h e r i t e d t r a i t s or d i s e a s e s in dogs. C u r r e n t efforts a r e l i m i t e d only by t h e a m o u n t of basic g e n e t i c knowledge a b o u t t h e c a n i n e g e n o m e , w h i c h is g r o w i n g s t e a d i l y (see "The C a n i n e Genome"), a n d specific m o l e c u l a r d a t a a b o u t i n d i v i d u a l M e n d e l i a n t r a i t s in dogs. T h e s e a r e only r e l a t i v e l i m i t a t i o n s , however, bec a u s e of t h e i n c r e a s i n g i n t e r e s t in a n d possibilities for e x p l o i t i n g p a r a l 103 Copyright 9 1997by AcademicPress. All rightsof reproductionin any formreserved. 1093-975X/97 $25.00
104
GEORGE H. SACK, JR.
lel developments in human and murine genetics. We will consider several examples that show the value of using these genetic data from other mammals for comparative studies in dogs. Molecular genetics is based on detecting locus-specific differences between individuals. The loci studied may be single genes of known identity, which necessarily occupy specific chromosome locations, or unique but anonymous chromosomal position markers identified only by their deoxyribonucleic acid (DNA) sequences. The latter are more powerful analytically because they frequently show greater interindividual variations; they are the basis for the best current h u m a n (Gyapay et al., 1994) and murine (Dietrich et al., 1994) linkage maps. Unfortunately, these anonymous markers cannot generally be related directly to specific biologic function(s). By contrast, the former are directly related t o - - a n d frequently derived from--specific genes and can be used to identify the precise mutation(s) responsible for specific traits. Ideally, and in the forseeable future, both sorts of loci will be interrelated in the canine genetic map as has been the case for maps of humans and mice. Currently, however, most Mendelian conditions in dogs are delineated on the basis of detailed study of their individual responsible genes. II. Methods A. RESTRICTION FRAGMENT LENGTH POLYMORPHISMS DNA-based gene analysis can proceed by several established methods. The first method relies on producing site-specific cleavage of chromosomal DNA with restriction enzymes and separating the thousands of resulting fragments by gel electrophoresis. The array of fragments can then be transferred to a supporting filter membrane, becoming a so-called Southern blot. Such a blot then can be used to determine the size (length) of immobilized DNA fragment(s) that contain(s) sequences identical (or sufficiently similar, depending on test conditions) to a specific DNA test sequence or probe. Exposing the blot to the probe permits alignment of the complementary nucleotide sequences and the size(s) (length) of the blotted fragment(s) can be determined by following the label (radioactivity, enzyme activity, antigenicity, etc.) on the probe. The critical variable in such a study is the size of the hybridizing fragment(s) on the Southern blot. Variation(s) in fragment size (length) between individuals is (are) the basis for identifying a restriction frag-
DIAGNOSTIC MOLECULAR GENETICS
105
ment length polymorphism (RFLP). RFLPs have served as the basis for much gene discovery and mutation detection. Each RFLP, however, relies on an imprecise measurement (the specific length of a DNA fragment, which may comprise thousands of base pairs) and may fail to detect many different DNA base changes (if the change does not disrupt the restriction enzyme recognition sites, which determine the end of the DNA fragment recognized by the probe, it will be invisible by this technique). Furthermore, an important DNA change, such as a point mutation, may not grossly alter the length of the fragment. An additional technical difficulty is that extensive RFLP analysis requires a substantial amount of DNA. B. POLYMERASE CHAIN REACTION
The second approach to DNA-based gene analysis uses variations of the basic technique of copying any given DNA fragment by the polymerase chain reaction (PCR). PCR can use a very small amount (in theory, a single molecule) of DNA (target DNA) and generate large numbers of copies by geometric amplification. PCR products can be tested directly for length by conventional gel electrophoresis. They also can be subjected to electrophoresis in denaturing enviroI.ments, which cause the two DNA strands to separate and thus reveal nucleotide sequence changes that have not affected the length of the fragment but which may alter the mobility of the single strands. PCR products may be combined with automated DNA sequencing technology to determine the actual nucleotide sequence of a specific region of an individual's genome and thus the presence of small changes can be detected. Variations on the basic PCR technique can provide specific, rapid testing for mutations that have been identified previously. PCR can be used to study both anonymous DNA sequences and those for which the function has been defined. Examples of the former are simple DNA sequence repeats (e.g., (CA)n where n can vary widely [e.g., Sack and Talbot, 1991]). Such dinucleotide repeats are spread throughout mammalian genomes. Other repeats may have tri- or tetranucleotide units; the latter have been most useful (because they show considerable variability) in canine genome studies. These repeated sequences are without known function but have distinct chromosomal positions defined by the unique stretches of DNA that flank them. Despite their anonymity, such repeat sequences are very valuable for linkage studies. Several groups are developing and characterizing these markers for the canine genome (Ostrander et al., 1993; Ostrander et al., 1995; Holmes et al., 1993; Mellersh et al., 1994;
106
GEORGE H. SACK, JR.
Fredholm and Winterr 1995). Undoubtedly, these will form the basis of the first canine molecular genetic linkage map. PCR also can be used to study single, defined regions of individual genes. Regional DNA changes can then be related directly to specific positions in their respective genes. This can permit detailed piece-bypiece analysis of large genes because PCR is most useful for relatively short lengths of DNA. PCR testing will remain the most flexible and far-reaching approach for genomic research for an extended time. C. MAPPING
To be maximally useful, individual DNA markers must be related to each other as well as to specific canine chromosomes. Dogs have 38 pairs of autosomes and a pair of sex chromosomes. These have been difficult to study because many are quite small. However, a consensus karyotype is now available and each can be distinguished (Selden et al., 1975).
1. Chromosome Assignment The first task in mapping is to assign individual markers to their respective chromosomes. This can be done by several methods (discussed in more detail in "Mapping Animal Genomes" and "The Canine Genome"). Cell lines can be established carrying small numbers of individual canine chromosomes on a background of, for example, the entire mouse or hamster genome. Such heterokaryons collectively contain all canine chromosomes. Many such cell lines are then prepared and the specific canine chromosomes that each line contains are determined. DNA from each cell line can be isolated and PCR testing can show which line(s) contains the specific canine gene or DNA marker. Referring the test results to the chromosome complement of each cell line can then show which chromosome contains which marker. Because of the large number of canine chromosomes, 75-100 hybrid cell lines will likely be needed for these studies. Although much work is involved in preparing and characterizing the hybrid cell lines, their DNA can serve as a permanent and standardized reference for all subsequent studies. 2. In Situ Hybridization Another approach to assigning genes and anonymous markers to specific chromosomes uses in situ hybridization. Here, the probe corresponding to the locus of interest and labeled with an identifying dye is hybridized to a metaphase array of chromosomes from either a hetero-
DIAGNOSTIC MOLECULAR GENETICS
107
karyon or a regular canine cell. The single site where the probe hybridizes can then be determined by light microscopy. In situ hybridization can identify the chromosome to which a probe corresponds (the chromosome must be distinguishable by other criteria as well). Also, it is frequently possible to determine the relative position of the probe on the chromosome--e.g., telomeric, centromere-proximal or -distal. In ideal circumstances, the relative positions of different probes can be determined simultaneously when each is labeled with a dye of a different color. A third approach to assigning markers to specific chromosomes (and relative positions on those chromosomes) uses linkage methods as described in the following section. These approaches are complementary and each builds on information supplied by the others. Canine genome mapping will begin by identifying a few genes or polymorphic markers on each chromosome; others will be added subsequently as positional relationships are determined. The initial goal (to be useful for subsequent efforts) is to construct a map with markers at intervals of 10 centimorgan (cM). This likely will require about 300 markers, allowing for redundancy and clustering, and will be the basis for the linkage studies proposed in the next section.
III. Linkage Analysis
A. BACKGROUND Genomic linkage studies are based on analyzing the passage of specific genes (or anonymous DNA markers) through families (or kindreds). Such studies ask whether any two identifiable traits (e.g., anonymous DNA markers, variations in known genes, or clinical features) are passed together in a kindred more frequently than would be expected on the basis of chance alone. The more frequently these traits are transmitted together through a kindred on the basis of simple pedigree analysis the more closely their responsible DNA locations will be on the actual genetic map. Efficient use of linkage analysis presupposes a set of highly variable markers throughout the genome which can be used for testing; DNA-based markers have been discussed previously. In theory, any trait or marker that has Mendelian segregation can be placed spatially on the growing linkage map for the organism. The limiting requirements are a sufficiently extensive kindred for analysis of the trait or marker and an adequate set of variable markers for comparison with the trait or marker of interest. With respect to the
108
GEORGE H. SACK, JR.
kindred, a m i n i m u m of three generations is usually needed, although this depends on the structure of the kindred. In terms of the genetic markers, it is important to note that many individual canine breeds already represent genetic isolates because of purposeful inbreeding, which may have reduced intrabreed genetic variation substantially (see Section IV,B). This may make it more difficult to find large numbers of highly variable genetic markers for certain breeds. The growing number of polymorphic markers now being developed should reduce, but not necessarily eliminate, this problem. B. POSITIONAL CLONING
Linkage analysis has been used effectively for human diseases that have Mendelian segregation patterns but for which the underlying causes are unknown. In human medicine, a particularly vivid example is provided by Huntington's disease in which the early finding linking the trait to an anonymous, variable DNA marker preceded discovery of the actual gene by a decade (Gusella et al., 1983; The Huntington's Disease Collaborative Research Group, 1993). Isolating the responsible gene beginning with a chromosomal location has been called positional cloning. An important use of linkage analysis is to establish the chromosomal location for a given genetic trait. Often, this can serve as the first step in identifying and isolating the responsible gene. Examples of using this approach in humans include Huntington's disease, as previously noted, as well as neurofibromatosis and cystic fibrosis (The Huntington's Disease Collaborative Research Group, 1993; Wallace et al., 1990; Riordan et al., 1989). Once a chromosomal location has been identified, subsequent efforts are directed to examining genes known to be in that chromosomal region as possible candidate genes to explain the disease. If no obvious candidate gene exists, recombinant DNA techniques (e.g., positional cloning) can be used to search for genes in the appropriate region and examine them for mutation(s) that might explain the disease. In linkage studies, it is important that an identifiable marker in or near the candidate gene must move consistently (segregate) through the kindred with the trait of interest. Finding divergence between kindred segregation patterns of the DNA and gene marker change(s) and the trait of interest excludes that change from explaining the condition. Conversely, finding consistent cosegregation between the DNA marker and the trait is supportive--but not necessarily diagnost i c - o f a direct relationship between them. These important aspects of
DIAGNOSTIC MOLECULAR GENETICS
109
using linkage analysis emphasize the value of large, well-defined kindreds to permit gathering statistically significant linkage segregation data and to support the identification of unequivocal gene positions before extensive molecular searches are undertaken. C. CANDIDATE GENES
A different but related application of linkage analysis is its use in candidate gene studies. Here, one can begin with a reasonable hypothesis about the pathophysiologic basis of the trait of interest; for example, an apparent enzyme deficiency. Next, a group of DNA clones is isolated representing potentially responsible genes. These clones are then studied for changes or mutations in affected animals. If such changes are found, the segregation patterns of the trait and potentially responsible gene change(s) can be analyzed for segregation in the kindred. Failure to find cosegregation excludes the candidate gene(s). Finding cosegregation strengthens the hypothesis. Here it is essential that a kindred be available for confirmatory studies. Another value of studying a large kindred in this context is that it provides reasonable assurance that only a single trait is present. The presence of a clinically similar but genetically distinguishable trait (phenocopy) complicates the analysis and can lead to rejecting a legitimate mutation if it is not present in all of the overtly similar animals. This approach to discovering genes responsible for specific traits depends on the availability of DNA clones of canine genes or their close counterparts from other mammals. Our recent study of the cross-reactivity of a group of single-copy h u m a n gene clones with canine DNA indicated that at least a third failed to give specific hybridization patterns, presumably because of significant divergence between h u m a n and canine DNA sequences (Sack et al., 1996). Thus, at least in some cases, clones for the actual canine genes may be needed to perform effective candidate gene studies in dogs. By contrast, some genes isolated from humans are very similar to their canine counterparts and cross-react readily. This cross-reactivity between the gene DNA sequences was fundamental to studies of golden retriever muscular dystrophy (Sharp et al., 1992) as well as to studies of phosphofructokinase deficiency in English springer spaniels (Giger et al., 1994), proteolipid protein in Welsh springer spaniels (Duncan and Nadon, 1994), clotting factor IX in Labrador retrievers in which the gene was deleted (Brooks, 1994) as well as others where a point mutation was present (Evans et al., 1989), and ~5(IV) collagen in Samoyed X-linked hereditary nephritis (Zheng et al., 1994).
110
GEORGE H. SACK, JR. D. QUANTITATIVE TRAIT LOCI
Another application of linkage analysis relates to many of the more common diseases of dogs. Canine hip dysplasia is a prominent example of a condition of widespread interest that has both a recognized spect r u m of clinical severity (e.g., age of onset, progression) and also breedspecific features (Hedhammer et al., 1979; Willis, 1989; Smith et al., 1990). An extensive DNA-based marker map of the canine genome can be used with extended kindreds to identify the chromosomal locations of genes responsible for at least some of the variation encountered clinically. Mathematical approaches are being developed to quantitate the contributions of individual genes to the final clinical picture. Genes or DNA regions contributing to such common traits are referred to as quantitative trait loci (QTL) and may (at least initially) be anonym o u s - i . e . , identified by chromosomal position alone. Here, in addition to a reasonably dense gene map of informative markers, it is essential to have extensive kindreds and a consistent set of criteria for measuring and defining the severity of the admittedly variable condition. QTL encompass some of the most common traits and disorders in dogs, including temperament, stature, and tumor susceptibility. Identifying critical genes will be very important in counseling breeders and owners. At least in principle, there is no reason why using this approach in dogs will not be as informative as, for example, QTL studies of hypertension in rats (Jacob et al., 1991). Despite the importance of this approach, however, it will be a labor-intensive undertaking, critically dependent on the prerequisites noted in this section.
IV. Single-Gene Disorders Molecular biologic approaches are most directly applicable to this category of traits. This also is the area in which progress is likely to be most rapid. The speed of developments in this area will reflect several factors: (1) There already are many well-defined Mendelian traits in dogs and at least some implicate malfunction of specific genes. (2) Breeding records can serve as starting points for kindred analysis and many study kindreds have been assembled. (3) Many DNA clones of single genes are being isolated in other organisms (particularly humans and mice). At least some of these should provide excellent starting points for studying their canine counterparts. (4) The technology of gene analysis (outlined in Section II) is directly applicable to studies in dogs.
DIAGNOSTIC MOLECULAR GENETICS
111
A. CANINE MUTATIONS
Table I lists nine canine Mendelian disorders that already have been studied by molecular techniques and for which actual nucleotide change(s) have been identified. In most of these, the responsible gene was already known or strongly implicated on clinical or metabolic grounds (for instance, the dystrophin gene in golden retriever muscular dystrophy [Sharp et al., 1992], phosphofructokinase deficiency in English springer spaniels [Giger et al., 1994], pyruvate kinase deficiency in Basenjis [Whitney et al., 1994], and ~-L-iduronidase deficiency in Plott hound mucopolysaccharidosis I [Menon et al., 1992]). These results are immediately applicable to clinical studies, such as those for presymptomatic testing of animals at risk and for carrier screening in breeding populations. Many conditions originally described in humans as single-gene disorders actually represent different mutations within the responsible gene (for example, ~-glucocerebrosidase in Gaucher disease [Horowits et al., 1993], neurofibromin in neurofibromatosis [Wallace et al., 1990], and the cystic fibrosis transmembrane regulator in cystic fibrosis [The cystic fibrosis genotype-phenotype consortium, 1993]). This situation, which may be more the rule than the exception in humans, makes carrier screening and presymptomatic detection in outbred human populations considerably more difficult because one cannot tell a priori which mutation to anticipate. B. CANINE GENETIC HOMOGENEITY
In studies of canine genes, by contrast, the long history of selective breeding has turned many breeds into "genetic isolates" and substantially homogenized their gene pool. This was shown by our recent studies of RFLPs in a large colony of Brittany spaniels in which we examined the autoradiographic signals developed from hybridizing 17 single-copy human DNA sequences against conventional Southern blots of canine DNA (Sack et al., 1996). Ten of these 17 probe DNAs represented so-called anchor-loci, which have been proposed for comparative mammalian gene mapping and which can reasonably be expected to be conserved between mammals (O'Brien et al., 1993). Despite considerable variation in hybridization conditions, only 8 of the 17 human DNA clones gave clear cross-hybridication signals and only 5 of these 8 showed diallelic RFLPs in the kindred. No multiallelic patterns were encountered using these probes. Unfortunately, diallelic RFLPs are of only limited value in kindred studies because they show relatively few variants.
TABLE 1 MUTATIONS IN SINGLE-GENE CANINE DISORDERS
Disorder
Breed
Mutation
X-linked severe combined immunodeficiency Muscular dystrophy Shaking pup syndrome Phosphofructokinase deficiency Rod-cone dysplasia I Hemophilia B Pyruvate kinase deficiency Mucopolysaccharidosis I
Basset hound Cardigan Welsh corgi Golden retriever Welsh springer spaniel English springer spaniel Irish setter Geulph cairn terrier Labrador retriever Basenji Plott hound
X-linked hereditary nephritis
Samoyed
4 bp" deletion in IL-2R~ 1 bp insertion A-~G in dystrophin Point mutation in proteolipid protein G--~A at 2228 in phosphofructokinase G ~ A at 2420 in cGMP b phosphodiesterase B G ~ A at 1477 in Factor IX Factor IX deletion Deletion of C 4:~:~in pyruvate kinase G ~ A in donor splice side of intron 1 of c~a-iduronidase G ~ T in ~5(IV) collagen chain
abp: base pair bcGMP:
Reference Henthorn et al., 1994 Sharp et al., 1992 Duncan and Nadon, 1994 Giger et al., 1994 Ray et al., 1994 Evans et al., 1989 Brooks, 1994 Whitney et al., 1994 Menon et al., 1992 Zheng et al., 1994
DIAGNOSTIC MOLECULAR GENETICS
113
These findings imply that we may expect considerable genetic homogeneity within individual canine kindreds. Thus, at least within breeds, it is likely that individuals affected with simple Mendelian disorders will have the same mutation because there will be a concentration of heterozygous carriers. Another consequence of this high frequency of heterozygotes is the likely appearance of individuals homozygous for recessive conditions, which would be rare in outbred populations. For example, the large kindred of Brittany spaniels assembled for studies of hereditary canine spinal muscular atrophy (HCSMA [Cork et al., 1979]) also has been found to segregate a deftciency of the third component of complement (Winkelstein et al., 1981) as well as a recessive form of cleft palate (Richtsmeier et al., 1994). These three traits are unlinked in Brittany spaniels. Recognizing this inbreeding should simplify screening and counseling because fewer mutations will need to be assessed once the breed-specific mutation(s) has (have) been identified. Mutations causing the same clinical picture (phenocopies) may differ between different breeds, however. Recall that hemophilia B (clotting Factor IX deficiency) shows a point mutation in Guelph cairn terriers and a deletion in Labrador retrievers (Evans et al., 1989; Brooks, 1994). C. TECHNICAL CONSIDERATIONS
1. Analytic Methods As noted previously, there are no unique technical barriers to genebased studies in dogs, especially when the studies are based on PCR. PCR permits the use of very small amounts of DNA, which can come from blood, skin, tail, hair bulb, and so on. Approaches such as conventional RFLP analysis require large amounts of blood or biopsy material and there is yet no reliable method for propagating canine lymphocytes in culture without using noncanine viruses. (By contrast, preparing lymphoblasts by transforming peripheral lymphocytes is a widely used method for studying h u m a n DNA beginning with only a small blood sample.) The analytical barriers are those of identifying mutations using the methods described earlier. Specific mutations are likely to be discovered by individual laboratories concentrating on intensive study of single illnesses or traits. Such laboratories generally will have collected sufficiently informative kindreds and affected individuals to permit detailed studies and also may have obtained pathophysiologic data, which can aid their search by implicating candidate genes. This has been a frequent pattern for similar studies of h u m a n and murine mutations.
114
GEORGE H. SACK, JR. 2. D a t a M a n a g e m e n t
As observations from various laboratories accumulate, it will be essential to organize the data and make them readily accessible to all workers. Of necessity, this will be computer based, preferably on-line, and should be able to exploit the data-handling methods of the h u m a n and murine genome projects. Having open access to all data will speed this work considerably. It also will minimize duplication of efforts, which is inefficient and wasteful of limited resources. The DOGMAP internet site is a prototype for this approach (Dolf et al., 1994). A particularly valuable method for h u m a n gene studies has been the publication of sequence-tagged sites (STSs), which are pairs of unique PCR primer DNA sequences flanking a defined genetic region (gene, anonymous marker, etc. [Olson et al., 1989]). Making all canine STS data available to all workers will permit quick dissemination of essential genomic information, which can be used by any laboratory interested in a particular gene region. Hudson et al. (1995) published a h u m a n gene map based on 15,086 STSs. DOGMAP includes sequence identification for unique, polymorphic canine markers (Dolf et al., 1994). V. Future Directions
Canine molecular genetics is in its infancy. The field will grow rapidly based on considerations noted in this article with special contributions from parallel efforts such as the H u m a n Genome Project. Important developments can be expected in the following: (1) delineating the molecular bases for single-gene diseases in dogs (this clearly will move most rapidly), (2) establishing a readily accessible public molecular genetic marker map for the canine genome (likely to be based on simple-sequence repeat polymorphisms spread across all canine chromosomes [Ostrander et al., 1993, 1995; Holmes et al., 1993; Mellersh et al., 1994]), and (3) dissecting complex canine traits based on contributions from individual Mendelian loci discovered by quantitative linkage analysis. The canine genome project will be particularly valuable for dissecting neurologic and behavior traits. Dogs permit detailed neurologic examinations, exposing fine nuances of clinical, behavioral, and neurophysiologic changes. HCSMA in Brittany spaniels provides an excellent example of the details that can be established (Cork et al., 1979) by thorough study. Many specific traits have been identified in other breeds and all are candidates for genetic and molecular dissection. Such studies will identify genes important for neuronal function. Still other, more
DIAGNOSTIC MOLECULAR GENETICS
115
broadly defined features (docility, herding behavior, aggressiveness), while undoubtedly polygenic, show sufficiently breed-specific features to be approached as quantitative trait loci (see Section II,D) so at least some individual genetic contributions to them will be identified. A particularly important outgrowth of this work will be a central data collection and coordination service. This will prevent duplicated efforts (because resources are limited) and also can serve as a referral center for specific diagnostic tests performed by individual laboratories. The rarity of some single-gene canine traits may prevent molecular-diagnostic studies from being financially self-supporting. Thus, proprietary, investigator-based diagnostic services may remain valuable for some time. Moreover, having additional numbers of referrals for the diagnosis of single-gene canine disorders in specialized laboratories will likely improve the breadth of experience and reliability in each laboratory and may identify additional mutations as has been the case in diagnostic laboratories for human genetic diseases (e.g., Wallace et al., 1990; Riordan et al., 1989; Horowits et al., 1993; The cystic fibrosis genotype-phenotype consortium, 1993; Lei et al., 1995). Combining the wide variation of canine breed characteristics and numerous specific genetic traits already recognized with molecular analytic tools will undoubtedly broaden the understanding of canine biology. ACKNOWLEDGMENTS Studies in the author's laboratory have been supported by the ALS Foundation and generous gifts from Mr. and Mrs. William Griffin and Mr. Daniel M. Kelly. I am grateful to Dr. Linda C. Cork for inspiration as well as to Dr. Elaine Ostrander for valuable discussions, critical reading, and suggestions; to Dr. Gustavo Aquirre for helpful references; and to Mrs. Mary A. Mix for secretarial assistance. REFERENCES Brooks, M. (1994). Canine hemophilia B: Affected dogs show phenotypic and molecular genetic heterogeneity. American Kennel Club Conference, Morristown, NJ, abs. 21. Cork, L. C., Griffin, J. W., Munnell, J. F., Lorenz, M. D., Adams, R. J., and Price, D. L. (1979). Hereditary canine spinal muscular atrophy. J. Neuropathol. Exp. Neurol. 38, 209-221. The cystic fibrosis genotype-phenotype consortium. (1993). Correlation between genotype and phenotype in patients with cystic fibrosis. N. Eng. J. Med. 329, 1308-1313. Dietrich, W. F., Miller, J. C., Steen, R. G., Merchang, M., et al. (1994). A genetic map of the mouse with 4006 simple sequence length polymorphisms. Nat. Genet. 7, 220-225. Dolf, G., Gaillard, C., Fredholm, M., Winterr A. K., Juneja, K., Lingaas, F., Sampson, J., Mellersh, C., Molyneux, K., Batt, R., Schelling, C., Switonski, M., Stranzinger, G., Fries, R., Ward, O. G., Giger, U., Goldspink, G., Mohan-Ram, V., Morris, B., Van de Weghe, A., Kelly, E. P., Rothuizen, J., Lubas, G., Niini, T., Gerlach, J., Holmes, N.,
116
GEORGE H. SACK, JR.
Binns, M., Troyer, D., Kent, M. G., Graphodatsky, A. S., Sargan, D. R., Pihkanen, S., Schook, L., Zhang, P., Panthier, J. J., Mariat, D., and M~irki, U. (1994). DOGMAP-An international collaboration towards a canine low resolution marker map. American Kennel Club Conference, Morristown, NJ, abs. 1. Duncan, I. D., and Nadon, N. L. (1994). Pathology and molecular genetics of the shaking pup syndrome. American Kennel Club Conference, Morristown, NJ, abs. Evans, J. P., Brinkhous, K. M., Brayer, G. D., Reisner, H. M., and High, K. A. (1989). Canine hemophila B Resulting from a poiint mutation with unusual consequences. Proc. Natl. Acad. Sci. USA 86, 10095-10099. Fredholm, M., and WinterO, A. K. (1995). Variation of short tandem repeats within and between species belonging to the Canidae family. Mammal. Genome 6, 11-18. Giger, U., Smith, B. F., Stedman, H., Rajpurohit, Y., Henthorn, P. S., Wolfe, J. H., and Patterson, D. F. (1994). Muscle-type phosphofructokinase (PFK) deficiency: A common cause of hemolysis and myopathy in English springer spaniels. American Kennel Club Conference, Morristown, NJ, abs. Gusella, J. F., Wexler, N. S., Conneally, P. M., Naylor, S. L., Anderson, M. A., Tanzi, R. E., Watkins, P. C., Ottina, K., Wallace, M. R., Sakaguchi, A. Y., Young, A. B., Shoulson, I., Bonilla, E., and Martin, J. B. (1983). A polymorphic DNA marker genetically linked to Huntington's disease. Nature 306, 234-238. Gyapay, G., Morissette, J., Vignal, A., Dib, C., Fizames, C., Millasseau, P., Marc, S., Bernardi, G., Lathrop, M., and Weissenbach, J. (1994). The 1993-1994 G~n6thon human genetic linkage map. Nat. Genet. 7, 246-339. Hedhammer,/~., Olsson, S.-E., Andersson, S.-/~., Persson, L., Pettersson, L., Olausson, A., and Sundgren, P.-E. (1979). Canine hip dysplasia: Study of heritability in 401 litters of German shepherd dogs. J. Am. Vet. Med. Assn. 174, 1012-1016. Henthorn, P. S., Somberg, R. L., Fimiani, V. M., Puck, J. M., Patterson, D. F., and Felsburg, P. J. (1994). IL-2R~ gene microdeletion demonstrates that canine X-linked severe combined immunodeficiency is a homologue of the human disease. Genomics 23, 6 9 - 74.
Holmes, N. G., Mellersh, C., Humphreys, S. J., Binns, M. M., Holliman, A., Curtis, R., and Sampson, J. (1993). Isolation and characterization of microsatellites from the canine genome. Anim. Genet. 24, 289-292. Horowits, M., Tzuri, G., Eyal, N., Berebi, A., Kolodny, E. H., Brady, R. O., Barton, N. W., Abrahamov, A., and Zimran, A. (1993). Prevalence of nine mutations among Jewish and non-Jewish Gaucher disease patients. Am. J. Hum. Genet. 53, 921-930. Hudson, T. J., Stein, L. D., Gerety, S. S., Ma, J., et. al. (1995). An STS-based map of the human genome. Science 270, 1945-1954. The Huntington's Disease Collaborative Research Group. (1993). A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes. Cell 72, 971-983. Jacob, H. J., Lindpainter, K., Lincoln, S. E., Kusumi, K., Bunker, R. K., Mao, Y.-P., Ganten, D., Dzau, V. J., and Lander, E. S. (19917. Genetic mapping of a gene causing hypertension in the stroke-prone spontaneously hypertensive rat. Cell 67, 213-224. Lei, K.-J., Chen, Y.-T., Chen, H., Wong, L.-J. C., Liu, J.-L., McConkie-Rosell, A., Van Hove, J. L. K., Ou, H. C.-Y., Yeh, N. J., Pan, L. Y., and Chou, J. Y. (1995). Genetic basis of glycogen storage disease type l a: Prevalent mutations at the glucose-6-phosphatase locus. Am. J. Hum. Genet. 57, 766-771. Mellersh, C., Holmes, N., Binns, M., and Sampson, J. (1994). Dinucleotide repeat polymorphisms at four canine loci (LEI 003, LEI 007, LEI 008 and LEI 015). Anita. Genet. 25, 125-126.
DIAGNOSTIC MOLECULAR GENETICS
117
Menon, K. P., Tien, P. T., and Neufeld, E. F. (1992). Architecture of the canine IDUA gene and mutation underlying canine mucopolysaccharidosis I. Genomics 14, 763-768. O'Brien, S. J., Womack, J. E., Lyons, L. A., Moore, K. J., Jenkins, N. A., and Copeland, N. G. (1993). Anchored reference loci for comparative genome mapping in mammals. Nat. Genet. 3, 103-112. Olson, M., Hood, L., Cantor, C., and Botstein, D. (1989). A common language for physical mapping of the human genome. Science 245, 1434-1435. Ostrander, E. A., Mapa, P., Yee, M., and Rine, J. (1995). One hundred and one new simple sequence repeat-based markers for the canine genome. Mammal. Genome 6, 192-195. Ostrander, E. A., Sprague, G. F., and Rine, J. (1993). Identification and characterization of simple sequence repeat markers for genetic mapping in dogs. Genomics 16, 207-213. Ray, K., Baldwin, V. J., Acland, G. M., Blanton, S. H., and Aguirre, G. D. (1994). Cosegregation of Codon 807 mutation of the canine rod cGMP phosphodiesterase [3 gene and rcd 1. Invest. Ophthalmol. Vis. Sci. 35, 4291-4299. Richtsmeier, J. T., Sack, G. H., Jr., Grausz, H. M., and Cork, L. C. (1994). Cleft palate with autosomal recessive transmission in Brittany spaniels. Cleft Palate-Craniofacial J. 31, 364-371. Riordan, J. R., Rommens, J. M., Kerem, B., Alon, N., Rozmahel, R., Grzelczak, Z., Zielenski, J., Lok, S., Plavsic, N., Chou, J.-L., Drumm, M. L., Iannuzzi, M. C., Collins, F. S., and Tsui, L.-C. (1989). Identification of the cystic fibrosis gene: Cloning and characterization of complementary DNA. Science 245, 1066-1073. Sack, G. H., Jr., and Talbot, C. C., Jr. (1991). Highly polymorphic domains of the human serum amyloid A (SAA) gene GSAA1. Scand. J. Immunol. 33, 485-488. Sack, G. H., Jr., Taylor, W. E., Meyers, D. A., Dragwa, C. R., and Cork, L. C. (1996). Canine genetic linkage study using heterologous DNA probes. J. Hered. 87, 15-20. Selden, J. R., Moorhead, P. S., Oehlert, M. L., and Patterson, D. F. (1975). The giemsa banding pattern of the canine karyotype. Cytogenet. Cell Genet. 15, 380-387. Sharp, N. J. H., Kornegay, J. N., Van Camp, S. D., Constantinou, C. D., Herbstrieth, M., Secore, S., Hung, W. Y., Dykstra, M. J., Roses, A. D., and Bartlett, R. J. (1992). Exon skipping during dystrophin mRNA processing in a canine model of Duchenne muscular dystrophy. Genomics 13, 115-121. Smith, G. K., Biery, D. N., and Gregor, T. P. (1990). New concepts of coxofemoral joint instability and the development of a clinical stress-radiographic method for quantitating hip joint laxity in the dog. J. Am. Vet. Med. Assn. 196, 59-70. Wallace, M. R., Marchuk, D. A., Andersen, L. B., Letcher, R., Odeh, H. M., Saulino, A. M., Fountain, J. W., Brereton, A., Nicholson, J., Mitchell, A. L., Brownstein, B. H., and Collins, F. S. (1990). Type 1 neurofibromatosis gene: Identification of a large transcript disrupted in three NF1 patients. Science 249, 181-186. Whitney, K. M., Goodman, S. A., Bailey, E. M., and Lothrop, C. D., Jr. (1994). The molecular basis of canine pyruvate kinase deficiency. Exp. Hematol. 22, 866-874. Willis, M. B. (1989). In Genetics of the Dog. London: H. F. and G. Witherby. Winkelstein, J. A., Cork, L. C., Griffin, D. E., Griffin, J. W., Adams, R. J., and Price, D. L. (1981). Genetically determined deficiency of the third component of complement in the dog. Science 212, 1169-1170. Zheng, K., Thorner, P. S., Marrano, P., Baumal, R., and McInnes, R. R. (1994). Canine X chromosome-linked hereditary nephritis: A genetic model for human X-linked hereditary nephritis resulting from a single base mutation in the gene encoding the ~5 chain of collagen type IV. Proc. Natl. Acad. Sci. USA 91, 3989-3993.
This Page Intentionally Left Blank
ADVANCESIN VETERINARYMEDICINE,VOL. 40
Gene Therapy for the Hemophilias JOHANNES WALTER* AND KATHERINE A. HIGHt *Department of Cardiothoracic Surgery, University of Vienna, Vienna, Austria, and tDepartments of Pediatrics, Pathology, and Laboratory Medicine, University of Pennsylvania School of Medicine, and Hematology Division, The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania 19104 I. Introduction A. Clinical Manifestations B. Structure, Biology, and Genetics II. Advantages of Gene T h e r a p y as a T r e a t m e n t for Hemophilia III. Advantages of Hemophilia as a Model for Gene Therapy IV. Choice of Vector A. Retroviral Vectors B. Adenoviral Vectors C. Adeno-associated Viral Vectors V. Factor IX VI. Factor VIII VII. S u m m a r y References
I. Introduction A.
CLINICAL MANIFESTATIONS
Hemophilias A and B are X-linked recessive bleeding disorders caused by molecular defects in the genes for factors VIII and IX, respectively. The incidence of the disease is about 25 in 100,000 males, with hemophilia A accounting for about 80% of the cases. Although the two forms of hemophilia result from mutations in two different genes, they are indistinguishable clinically. The severity varies as a function of the activity level of the missing or defective factor. Patients with F.VIII or F.IX levels less t h a n or equal to 1% of normal are severely 119 Copyright 9 1997 by Academic Press. All rights of reproduction in any form reserved. 1093-975X/97 $25.00
120
JOHANNES WALTER AND KATHERINE A. HIGH
affected, generally requiring clotting factor infusion as often as two to four times a month as treatment for bleeding episodes. Spontaneous bleeding into the joint space is the most common manifestation of the disease and over a period of years can result in significant arthropathy. The most serious complications arise from bleeding into closed spaces, e.g., intracranial, retroperitoneal, and epidural or retropharyngeal bleeds; these can be fatal if not treated quickly. Patients with F.VIII or F.IX levels from 1% to 5% have more moderate disease with spontaneous bleeds occurring less frequently. Those with levels from 5% to 15% have a mild phenotype and generally experience bleeds only in the setting of t r a u m a or surgery. Mild forms of the disease may not be diagnosed until adult life. It is the bleeding into the joints that causes the major morbidity of the disease and often leads to crippling chronic arthritis and arthropathy. Because hemophilia patients have normal platelet function, they do not manifest excessive hemorrhage from minor cuts and abrasions. B. STRUCTURE, BIOLOGY, AND GENETICS The genes for both F.VIII and F.IX are located on the X-chromosome at positions q28 and q27, respectively. The gene for F.VIII, one of the largest genes known, is composed of 26 exons and spans 186 kilobases (kb) (Gitschier et al., 1984). The F.VIII messenger ribonucleic acid (mRNA) is about 9 kb. The complete sequence of the gene has been determined (Gitschier et al., 1984; Vehar et al., 1984) and numerous mutations, including gross and partial gene deletions, as well as point mutations, have been identified in hemophilia A patients (Tuddenham et al., 1991). The gene for F.IX is about 34-kb long and consists of eight exons and seven introns (Yoshitake et al., 1985). Before secretion of the mature protein, F.IX must undergo cleavage of its signal and propeptide sequences, as well as ~/-carboxylation of 11 glutamic acid residues at the N-terminus of the protein. The mature protein is secreted by hepatocytes as a single-chain polypeptide consisting of 415 amino acids. Expression of functional F.IX from target cells in gene therapy requires that these cells perform necessary posttranslational processing functions efficiently.
II. Advantages of Gene Therapy as a Treatment for Hemophilia Current treatment for hemophilia consists of intravenous infusion of clotting-factor concentrates, either plasma derived or recombinant. Because of the expense of clotting-factor concentrates (adults with severe
GENE THERAPYFOR THE HEMOPHILIAS
121
disease typically spend $50,000-$100,000 per year on concentrates), infusions are generally given only in response to bleeds, not prophylactically. Thus, during the inevitable delay between the onset of a bleed and the infusion of factor, tissue damage results. Gene therapy potentially could provide sustained synthesis of the clotting factor from genetically modified cells, so a constant level of circulating factor would result and bleeds could be prevented rather than treated. Gene therapy also has the potential to provide a less expensive alternative to infusion of clotting-factor concentrates, as it should obviate the need for frequent treatments. A major problem with plasma-derived concentrates has been the transmission of blood-borne pathogens, especially hepatitis viruses (Troisi et al., 1993) and the human immunodeficiency virus (HIV) (Goedert et al., 1989). Current concentrates, however, are essentially safe because efficient methods for heat-inactivation, antibody-purification, and virucidal treatments have been developed (Schimpf et al., 1987, 1989). III. Advantages of Hemophilia as a Model for Gene Therapy Compared to other genetic and acquired disorders for which gene therapy has been proposed or contemplated, hemophilia has several features that make it an appealing model in which to attempt gene therapy. For example, the hemophilias do not require precise regulation of the levels of expression of factors VIII and IX. Levels as low as 5% of normal would have a substantial therapeutic effect, and levels as high as 200% of normal have not been associated with ill effects in patients infused with clotting-factor concentrates. This is at least partly because the factors circulate in the plasma as inactive precursors that become activated only when the coagulation cascade is initiated. Another advantage of hemophilia as a model for gene therapy is that there is no requirement for tissue-specific expression. Although the clotting factors are normally synthesized in the liver, work by a number of investigators has established that biologically active F.IX can be synthesized in a variety of other tissues, including fibroblasts, endothelial cells, and myoblasts (Dai et al., 1992; Palmer et al., 1989; Yao and Kurachi, 1992). Similar findings have been reported for F.VIII (Eaton et al., 1987). Finally, large- and small-animal models exist that faithfully mirror the human disease. For both hemophilia A and B, there are naturally occurring dog models of the disease. Canine hemophilia colonies are maintained at several centers in North America; the canine disease is X-linked and is characterized by frequent spontaneous bleeding episodes similar to those seen in hu-
122
JOHANNES WALTERAND KATHERINE A. HIGH
mans. The canine F.IX copy deoxyribonucleic acid (cDNA) has been cloned (Evans et al., 1989a), but the canine F.VIII cDNA is not yet available. As is the case for humans, the molecular defects leading to canine hemophilia appear to be heterogeneous. Evans et al. (1989b) have delineated the defect causing severe hemophilia B (F.IX activity <1% normal) in the Chapel Hill colony; these animals exhibit a Gly Glu substitution at residue 379, which appears to result in an unstable protein that is destroyed intracellularly (i.e., is not secreted). Other variants are caused by gene deletions or other point mutations (Marjorie Brooks and James Catalfamo, New York State College of Veterinary Medicine, Cornell University, Ithaca, New York, unpublished observations). Mutations causing canine hemophilia A have not yet been reported. A mouse model of hemophilia A was created by knock-out technology (Bi et al., 1995), and similar models are in progress for hemophilia B (Darrel Stafford, UNC-Chapel Hill, personal communication). The hemophilia A mice have <1% F.VIII levels by an activity assay. They do not appear to bleed spontaneously but exhibit severe (lethal) bleeding following routine tail vein sampling. The availability of both large- and small-animal models constitutes an unusual (though not unique) resource for the study of a genetic disease; the animal models have been used extensively for trials of gene therapy (vide infra).
IV. Choice of Vector Vectors used in gene therapy for hemophilia must transfer the gene of interest efficiently and safely and must either result in long-term expression or allow for repetitive administration. It should also be noted that there is a distinction between gene repair, in which the defective gene is replaced by a normal copy of the gene, and gene insertion, in which the defective copy remains in the target cell, but a normal copy is inserted elsewhere within the host genome or the cell nucleus. With current techniques, any method of gene repair therapy remains speculative because the efficiencies in vitro and in vivo are much too low to achieve any substantial level of gene correction. Thus, all current approaches to gene therapy are gene insertion techniques. Vectors currently under study for gene therapy can be divided into two large groups: viral and nonviral vectors. Among viral vectors, most work has been focused on retroviral, adenoviral, and adeno-associated virus (AAV) vectors. Most nonviral techniques are based on liposomes that can encapsulate large DNA sequences, and techniques in which
B.
[ '
'] . . . .
] humanF.IXcDNA
] SV
]
neo
H
LTR
FIG. 1. Structure of a retroviral vector expressing human factor IX. (A) Segments of the retroviral genes, gag, pol, and env, have been excised and (B) replaced by the human F.IX cDNA and a selectable marker neo under the control of the SV40 promoter. Transcription of the gene of interest, in this case F.IX, is controlled by a promoter in the long terminal repeat (LTR).
the DNA is bound to polylysine together with adenoviral particles and cell receptor ligands. A. RETROVIRAL VECTORS
Retroviral vectors have generated a great deal of interest because they are capable of transferring genetic material stably into a cell's genome and subsequently expressing it in a m a n n e r that is generally not detrimental to the host cell. Most of the currently used retroviral vectors are derived from the Moloney murine leukemia virus (MoMLV). Because parts of their coding regions are replaced by the gene to be transferred (Fig. 1), they are replication defective and have to be grown in special cell lines (packaging cells) that supply the structural and replicative proteins in trans. However, co-infection with helper virus can lead to viral replication, which may cause serious side effects such as in vivo tumorigenesis as has been observed in some animal models (Donahue et al., 1992). Thus, for any h u m a n application of the virus, extensive testing for helper virus contamination must be carried out. There is no m i n i m u m requirement for the size of the insert in a retroviral vector; the maximum size insert that can be accommodated is approximately 7.5 kb (Morgenstern and Land, 1991). The majority of published gene transfer studies use recombinant retroviruses. These vectors are capable of stably transducing up to 100% of the target cells in tissue-culture experiments. However, replication of the target cell is necessary for efficient transduction and pro-
A number of recombinant adenoviral vectors expressing different transgenes have been developed and shown to deliver genes with high efficiency to a broad range of target cells (reviewed in Kozarsky and Wilson (1993)). These vectors are of special interest because they are easily prepared in very high titers, are capable of transducing nondividing cells in vitro and in vivo, and direct high levels of expression of the transgene. Integration into the host cell genome does not appear to be an integral part of the adenoviral life cycle. Rather, the viral DNA enters the nucleus, where it remains episomal. Most current recombinant adenoviral vectors are derived from human adenovirus type 2 or 5, which cause respiratory disease in humans but which have not been associated with any human malignancies (Strauss, 1984). They contain deletions of parts of their genome (e.g., of the E l a - E l b and/or E3 genes) that render them replication deficient (Fig. 2). These replication-deficient adenoviruses are grown to high titers in 293 cells, a human embryonic kidney cell line, which is transformed with the adenoviral E1 gene; this supplies E1 gene products in trans. The usefulness of recombinant adenoviral vectors for in vivo applica-
d e l e t i o n E3
CMV F.IX 5'UT poly A
1 -
8.4 - 86
I
I
0
20
I
40
I
I
I
60
100
80
E3
E1
v
A
E2
A
,,
E4
A
v
Late Transcription
125
GENE THERAPY FOR THE HEMOPHILIAS
tions has been limited by the immune response of the host organism to the virus and virally transduced cells; the immune response results in both limited duration of expression and a failure of expression following readministration of the vector (Yang et al., 1994). Strategies proposed to circumvent this obstacle have included immunosuppression of the host (Dai et al., 1995), engineering of the adenoviral vector to reduce expression of adenoviral proteins (Engelhardt et al., 1994), and the induction of tolerance in the host animal (Walter et al., 1996). C. ADENO-ASSOCIATED VIRAL VECTORS
Although there are not yet any data on the use of AAV vectors for gene transfer of F.VIII or F.IX, these vectors are noteworthy for several features that make them potentially useful as gene delivery vehicles. AAV is a single-stranded DNA virus that encodes two major genes, rep and cap, and includes inverted terminal repeat (ITR) sequences (Fig. 3). The ITR sequences are required in cis to function as the origin of replication, and they also contain the sequences required for packaging, integration into the host cell genome, and rescue from recombinant plasmids (Hermonat et al., 1984; Samulski et al., 1987). The rep and cap genes are only required in trans for viral replication and encapsidation, respectively. In a recombinant vector, the promoter and transgene cassette is placed between the ITRs and rep and cap are supplied in trans (on a different plasmid) during the generation of the recombinant virus. A helper virus, such as adenovirus, is needed for efficient AAV replication in vitro. Once generated, these vectors are replication defective because they lack rep and cap. Thus, recombinant AAV vectors have the advantage that coding sequences for viral gene products are not transferred to the host. After gaining entry into a host
lr
B9 ~
l ,,
l
re.
promoter
'I
[
cap
'I
}
t
transgene
I,
,:-ff-]
FIG. 3. AAV vector. (A) The genome of the wild-type AAV is -4.7 kb. The ITRs are 145 bp long. (B) In the vector, the promoter and transgene cassette replaces the rep and cap genes, which are supplied in t r a n s by a second plasmid during generation of the recombinant vector.
126
JOHANNES WALTERAND KATHERINE A. HIGH
cell, AAV can stably integrate its DNA via the ITRs and has been shown to transduce efficiently a wide range of human target cells (Cukor et al., 1984). In contrast to retroviruses and adenoviruses, AAV is not associated with any human disease (Berns et al., 1982). Wild-type AAV possesses the unique feature of integrating preferentially at a specific location on chromosome 19 (Kotin et al., 1990). Such a feature would be an advantage in a vector because it would decrease the likelihood of integration occurring at a potentially harmful site. However, conservation of this property would most likely require inclusion of additional AAV gene sequences in the recombinant vector. Thus far, low titers have been a problem. AAV vectors have not yet been used in any human trials but have been used successfully for gene transfer in rodents (Kaplitt et al., 1994) and rabbits (Flotte et al., 1993). A clearer understanding of AAV biology, and technical advances in high-titer vector preparation, will be required for more extensive use of this vector.
V. Factor IX
Because the F.IX cDNA is only about 1.4-kb long (about 2.7 kb when the long 3' untranslated region is included), it fits the size constraints of all currently useful viral vectors. Much of the experience with in vivo gene transfer of F.IX (summarized in Table 1) has been with retroviral vectors. These vectors have been shown to express F.IX at high levels in cell culture (up to 3000 ng/ml) in various cell types. When transduced with vectors carrying either the human or canine F.IX cDNA, fully active F.IX was produced in primary culture human (Palmer et al., 1989) and mouse (St. Louis and Verma, 1988) fibroblasts as well as in primary skin fibroblasts of hemophilic dogs (Axelrod et al., 1990; Lozier et al., 1994), primary rabbit hepatocytes (Armentano et al., 1990), rat capillary endothelial cells (Yao et al., 1991), and mouse myoblasts (Dai et al., 1992; Yao and Kurachi, 1992). This demonstrates that the necessary posttranslational modifications can be performed by a wide variety of target cells. However, successful gene therapy would require that cells transduced ex vivo also survive and express the transgene at therapeutic levels when transferred to a host animal. When Palmer et al. (1989) transduced human fibroblasts ex vivo and implanted these subcutaneously into nude mice, they achieved moderate levels (50-200 ng/ml) of active human F.IX, but levels declined to baseline after 4 weeks. Similar results were obtained by St. Louis and Verma (1988). Dai et al. (1992) reported long-term expression of biologically
TABLE 1 IN Y l v o ANIMAL TRIALS OF GENE THERAPY FOR HEMOPHILIA B
Vector
Target Cells
Animal Model
Maximal Levels of Active F.IX (%)
Duration of Expression
Retrovirus
Hapatocytes
Hemophilic Dogs
<0.2%
>9 mos.
Retrovirus
P r i m a r y Myoblasts
Nude Mice
0.2%
>6 mos.
Adenovirus
Hepatocytes
-6%
9 weeks
Adenovirus
Hepatocytes
Immunocompetent Mice Hemophilic Dogs
300%
4•
Adenovirus
Hepatocytes
-500%
> 6 mos. with 2 injections
Immunocompetent Mice
weeks
Reference Kay et al. (#35) Science '93 Dai et al. (# 10) P N A S '92 Smith et al. (#46) N a t u r e Genetics '93 Kay et al. (#36) P N A S '94 Walter et al. (#22) P N A S '96
128
JOHANNES WALTER AND KATHERINE A. HIGH
active F.IX from transduced primary myoblasts that were injected into hind legs of recipient mice. Although the expression was stable for over 6 months, levels ofF.IX expression were subtherapeutic at 10 ng/ml. In a different in vivo approach, Kay et al. (1993) infused a recombinant retrovirus expressing canine F.IX into the portal vein of hemophilic dogs that had undergone partial hepatectomy (to induce cell division in the remaining hepatocytes). The authors reported a modest shortening of the whole blood clotting time, but levels of F.IX expression (~10 ng/ml) were still well below the therapeutic range. These results indicate that in vivo gene therapy with retroviral vectors is feasible, but also that levels of expression will have to be increased by at least an order of magnitude for possible h u m a n application. Adenoviral vectors have been used in dogs with hemophilia B to achieve high-level expression of canine F.IX with complete phenotypic correction, but expression is only transient and, on readministration of the vector, there is no expression, presumably because of the immune response to the vector. Kay et al. (1994) injected an adenoviral vector into the portal vein and achieved supraphysiologic levels (~33 ~g/ml) of canine F.IX in the plasma. Levels fell steadily over a period of 1-2 months and were well below the therapeutic range (< 10 ng/ml) at that point. Readministration of vector was not specifically addressed in this report. In a subsequent paper, Dai et al. (1995) showed that intramuscular administration of an adenoviral vector expressing canine F.IX to nude mice resulted in long-term (> 300 days), high-level expression of the transgene, whereas injection into normal mice resulted in only 7-10 days' expression. Normal mice could not be reinjected because of neutralizing antibodies to adenovirus. Co-administration of the immunosuppressive drug cyclosporin A resulted in longer-term expression. Several strategies to circumvent the obstacles to reinjection have been proposed and investigated. Walter et al. (1996) have taken advantage of the immaturity of the newborn immune system to achieve successful repeat administration of an adenoviral vector expressing h u m a n F.IX. Normal adult mice injected intravenously with the vector showed supraphysiologic levels of expression, which declined over a period of 12-16 weeks to zero; these mice could not successfully be reinjected. However, newborn mice injected on day one of life also showed high levels of expression, which declined over a period of 12-16 weeks, but when these mice were reinjected with the same vector, high-level expression was again achieved. Whether third or subsequent injections will be possible is not yet clear, but the data suggest that early exposure may alter the immune response. Other investigators have used immunomodulatory agents to achieve expression following repeat administration (Yang et al., 1995) or to achieve
GENE THERAPY FOR THE HEMOPHILIAS
129
longer-term expression from a single administration of an adenoviral vector (Kay et al., 1995). Neither of these strategies has yet been tried with vectors expressing coagulation factors.
VI. Factor VIII
Experience with F.VIII gene transfer is more limited than that with F.IX. This is because of several disadvantages of F.VIII compared to F.IX. First, the full-length F.VIII cDNA is too large (9.0 kb) to be accommodated in many currently available vectors. However, this disadvantage can be overcome by deleting the nonessential B domain, which renders the F.VIII cDNA small enough for most vectors. Second, F.VIII is a considerably larger molecule than F.IX and the half-life of human F.VIII in h u m a n plasma is only 10-12 h (only about 2.5 h without von Willebrand factor (vWF) (Tuddenham et al., 1982). In addition, it has been shown by several groups (Dwarki et al., 1995; Hoeben et al., 1992) that the half-life of human F.VIII in murine plasma is even shorter (less than 1 h) making it even harder to achieve detectable levels of h u m a n F.VIII in murine model systems. (The recent availability of a knock-out mouse model of hemophilia A may solve this experimental problem) (Bi et al., 1995). Finally, vWF is required to stabilize F.VIII in plasma because it acts as a carrier protein that protects F.VIII from proteolysis in vivo. Thus, the expressed F.VIII must gain easy and rapid access to the circulation to bind to vWF. Several groups have successfully used recombinant retroviruses for gene transfer of F.VIII. In in vitro experiments, Hoeben et al. (1990) have achieved moderate levels (35-125 mU/106 cells for 24 hours) of functional human F.VIII in murine fibroblast cell lines as well as in primary h u m a n skin fibroblasts. These vectors contained a partially deleted F.VIII cDNA (B domain deleted) driven by the retroviral long terminal report (LTR). When the transduced cells were transplanted subcutaneously into nude mice, no human F.VIII antigen was detected, although the cells seemed to survive (Hoeben et al., 1992). It was unclear whether the failure to detect F.VIII antigen after implantation was caused by the short half-life of human F.VIII in murine plasma, insufficient vascularization of the implant, inadequate transport of F.VIII into the circulation, or specific inactivation of the virally derived transcripts (Challita and Kohn, 1994). Dwarki et al. (1995), using the retroviral vector to express a B domain-deleted F.VIII construct achieved therapeutic levels of active F.VIII in nude mice. When this group transduced primary human fibroblasts in vitro and implanted them into the peritoneal cavity of
130
JOHANNES WALTER AND KATHERINE A. HIGH
nude mice, they detected levels of F.VIII as high as 100 ng/ml for about 10 days after which the levels fell to zero. Both cell death and extinction of the LTR-driven mRNA expression may have played a role in the limited duration of expression. Interestingly, when transduced C2C12 cells expressing F.VIII were injected intramuscularly into nude mice, no F.VIII was detectable. This is in contrast to the same experiments done with F.IX, where F.IX can be detected in the plasma of nude mice for prolonged periods (>100 days) at levels of 10 ng/ml (Dai et al., 1992). More recently, Kaleko and coworkers have achieved sustained levels of Factor VIII in the plasma of C57/B16 mice using an adenoviral vector (Connelly et al., 1996). These investigators have shown that inclusion of genomic elements in the adenoviral construct results in higher level expression, allowing for administration of lower doses of vector. Duration of expression is longer with lower doses, presumably due to a less vigorous immune response. In a different, nonviral approach, Zatloukal et al. (1994) used adenoviral-polylysine-transferrin-DNA complexes to express human F.VIII in mice. In this gene delivery method, pioneered by Birnstiel and colleagues (Cotten et al., 1992), polylysine serves to condense the DNA, and transferrin serves as a ligand that allows the complex to bind to receptors on the surface of the target cell, leading to endocytosis of the DNA-containing complex. Inclusion of (ultraviolet-inactivated) replication-defective adenovirus in the complex results in 100to 1000-fold enhancement of gene expression, presumably because of disruption of the endosome and resulting improved delivery of DNA to the target cell nucleus. In the report by Zatloukal et al., murine fibroblasts and myoblasts were "transferrinfected" in vitro and then injected into the spleens of mice. F.VIII levels detected in the plasma were around 8-17% of normal at 24 hours after injection, but declined to undetectable levels at 48 hours. The reasons for this rapid decline in expression are not known. When the same cells were injected into skeletal muscle, however, no expression was seen, which is in accord with the report by Dwarki et al. (1995).
VII. Summary There are many lines of evidence that suggest the eventual success of gene therapy as a treatment strategy for hemophilia. Because current treatment protocols using plasma-derived or recombinant proteins are far from ideal, the safe and efficient substitution of the defective gene by a normal copy of the gene, or at least its addition, would be
GENE THERAPY FOR THE HEMOPHILIAS
131
of great benefit to the p a t i e n t and m a y even be a potential cure. However, the construction of efficient gene t h e r a p y vehicles has proven quite difficult in the past and, so far, t h e r e is no system t h a t promises to have all the desired features without any serious disadvantages. In general, either the levels of t r a n s g e n e expression are too low (because of the low titers achieved d u r i n g the generation of the virus) or shortlived (e.g., because of the specific shut-off of the t r a n s f e r r e d promoter) as is often seen with retroviruses, or in the case of adenoviral vectors, expression is limited because of a strong i m m u n e response of the host. Clearly, much work r e m a i n s to be done to optimize these promising t h o u g h still imperfect vector systems. In the case of adenovirus, the development of less immunogenic vectors or in v i v o modulation of the host i m m u n e s y s t e m m a y hold promise for improvements. Reports by Yang et al. (1995) and Kay et al. (1995) are promising steps in the direction of i m m u n o m o d u l a t i o n . Both a t t e n u a t e the i m m u n e reaction to the adenoviral vector by s i m u l t a n e o u s application of either an interleukin or an immunoglobulin, respectively. W h e n IL-2 was administered, the a m o u n t s of IgA were reduced and successful a d m i n i s t r a t i o n of a second dose of virus was possible. W h e n CTLA4-Ig, an immunoglobulin t h a t blocks the second signal d u r i n g antigen presentation, was administered, a m a r k e d l y prolonged expression of the t r a n s g e n e resulted. I n vivo trials with AAV vectors have been carried out for some diseases (Flotte et al., 1993; Kaplitt et al., 1994) but not for hemophilia. Advances in high-titer AAV vector p r e p a r a t i o n will m a k e this approach more feasible. The pace continues to quicken in the developm e n t of nonviral modes of gene delivery (Perales et al., 1994). Although these results are encouraging for the future of gene t h e r a p y as a treatm e n t for genetic diseases, much work r e m a i n s to be done to m a k e this potential a l t e r n a t i v e a reality for t r e a t m e n t of hemophilia. REFERENCES Armentano, D., Thompson, A. R., Darlington, G., and Woo, S. L. C. (1990). Expression of human factor IX in rabbit hepatocytes by retrovirus-mediated gene transfer: Potential for gene therapy of hemophilia B. Proc. Natl. Acad. Sci. USA 87, 6141-6145. Axelrod, J. H., Read, M. S., Brinkhous, K. M., and Verma, I. M. (1990). Phenotypic correction of factor IX deficiency in skin fibroblasts of hemophilic dogs. Proc. Natl. Acad. Sci. USA 87, 5173-5177. Berns, K. I., Cheung, A., Ontrove, J., and Lewis, M. (1982). Adeno-associated virus latent infection. In "Virus Persistence" (B. W. J. Mahy, A. C. Minson, and G. K. Darby, eds.), p. 249. Cambridge University Press, Cambridge. Bi, L., Lawler, A. M., Antonarakis, S. E., High, K. A., Gearhart, J. D., and Kazazian, H. H. (1995). Targeted disruption of the mouse factor VIII gene produces a model of haemophilia A. Nat. Genet. 10, 119-121.
132
JOHANNES WALTER AND KATHERINE A. HIGH
Challita, P., and Kohn, D. B. (1994). Lack of expression from a retroviral vector after transduction of murine hematopoietic stem cells is associated with methylation in vivo. Proc. Natl. Acad. Sci. U.S.A. 91, 2567-2571. Connelly, S., Gardner, J. M., Lyons, R. M., McClelland, A., Kaleko, M. (1996). Sustained expression of therapeutic levels of human factor VIII in mice. Blood 87, 4671-4677. Cotten, M., Wagner, E., Zatloukal, K., Phillips, S., and Birnstiel, M. L. (1992). Highefficiency receptor mediated delivery of small and large (48 kilobase) gene constructs using the endosome-disruption activity of defective or chemically inactivated adenovirus particles. Proc. Natl. Acad. Sci. USA 89, 6094-6098. Cukor, G., Blacklow, N. R., Hoggan, D., and Berns, K. I. (1984). Biology of adeno-associated virus. In 'The Parvoviruses" (K. I. Berns, ed.), pp. 33-66. Plenum Press, New York. Dai, Y., Roman, M., Naviaux, R. K., and Verma, I. M. (1992). Gene therapy via primary myoblasts: Long-term expression of factor IX protein following transplantation in vivo. Proc. Natl. Acad. Sci. USA 89, 10892-10895. Dai, Y., Schwarz, E. M., Gu, D., Zhang, W. W., Sarvetnick, N., and Verma, I. M. (1995). Cellular and humoral immune responses to adenoviral vectors containing factor IX gene: tolerization of factor IX and vector antigens allows for long-term expression. Proc. Natl. Acad. Sci. USA 92, 1401-1405. Donahue, R. E., Kessler, S. W., Bodine, D., McDonagh, K., Dunbar, C., Goodman, S., Agricola, B., Byrne, E., Raffeld, M., Moen, R., et al., (1992). Helper virus induced T cell lymphoma in nonhuman primates after retroviral mediated gene transfer. J. Exp. Med. 17{}(4), 1125-1135. Dwarki, V., Belloni, P., Nijjar, T., Smith, J., Couto, L., Rabier, M., Clift, S., Berns, A., and Cohen, L. K. (1995). Gene therapy for hemophilia A: Production of therapeutic levels of human factor VIII in vitro in mice. Proc. Natl. Acad. Sci. USA 92, 1023-1027. Eaton, D. L., Hass, P. E., Riddle, L., Mather, J., Wiebe, M., Gregory, T., Vehar, G. (1987). Characterization of recombinant human F. VIII. J B C 263, 1115-1118. Engelhardt, J. F., Ye, X., Doranz, B., and Wilson, J. M. (1994). Ablation of E2A in recombinant adenoviruses improves transgene persistence and decreases inflammatory response in mouse liver. Proc. Natl. Acad. Sci. USA 91, 6196-6200. Evans, J. P., Brinkhous, K. M., Brayer, G. D., Reisner, H. M., and High, K. A. (1989a). Canine hemophilia B resulting from a point mutation with unusual consequences. Proc. Natl. Acad. Sci. USA 86, 10095-10099. Evans, J. P., Watzke, H. H., Ware, J. L., Stafford, D. W., and High, K. A. (1989b). Molecular cloning of a cDNA encoding canine factor IX. Blood 74(1), 207-212. Flotte, T. R., Afione, S. A., Conrad, C., McGrath, S. A., Solow, R., Oka, H., Zeitlin, P. L., Guggino, W. B., and Carter, B. J. (1993). Stable in vivo expression of the cystic fibrosis transmembrane conductance regulator with an adeno-associated virus vector. Proc. Natl. Acad. Sci. USA 90, 10613-10617. Gitschier, J., Wood, W. I., Goralka, T. M., Wion, K. L., Chen, E. Y., Eaton, D. H., Vehar, G. A., Capon, D. J., and Lawn, R. M. (1984). Characterization of the human factor VIII gene. Nature 312(5992), 326-330. Goedert, J. J., Kessler, C. M., and Aledort, L. M. (1989). A prospective study of human immunodeficiency virus type 1 infection and the development of AIDS in subjects with hemophilia. N. Engl. J. Med. 321, 1141-1152. Hermonat, P. L., Labow, M. A., Wright, R., Berns, K. I., and Muzyzcka, N. (1984). Genetics of adeno-associated virus: Isolation and preliminary characterization of mutants of adeno-associated virus type 2. Virol. 51, 329-339. Hoeben, R. C., Einerhand, M. P., Briet, E., van Ormondt, H., Valerio, D., and van der Eb, A. J. (1992). Toward gene therapy in heamophilia A: Retrovirus-mediated transfer of a
GENE THERAPY FOR THE HEMOPHILIAS
133
factor VIII gene into murine haematopoietic progenitor cells. Thrombosis and Haemostasis {;7(3), 341-345.
Hoeben, R. C., van der Jagt, R. C., Schoute, F., van Tilburg, N. H., Verbeet, M., Briet, E., van Ormondt, H., and van der Eb, A. (1990). Expression of functional factor VIII in primary human skin fibroblasts after retrovirus-mediated gene transfer. J. Biol. Chem. 265(13), 7318-7323. Kaplitt, M. G., Leone, P., Samulski, R. J., Xiao, X., Pfaff, D. W., O'Malley, K. L., and During, M. J. (1994). Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet. 8, 148-154. Kay, M. A., Holterman, A. X., Meuse, L., Gown, A., Ochs, H. D., Linsley, P. S., and Wilson, C. B. (1995). Long-term hepatic adenovirus-mediated gene expression in mice following CTLA4Ig administration. Nat. Genet. 11(2), 191-197. Kay, M. A., Landen, C. N., Rothenberg, S. R., Taylor, L. A., Leland, F., Wiehle, S., Fang, B., Bellinger, D., Finegold, M., Thompson, A. R., et al., (1994). In vivo hepatic gene therapy: Complete albeit transient correction of factor IX deficiency in hemophilia B dogs. Proc. Natl. Acad. Sci. USA 91, 2353-2357. Kay, M. A., Rothenberg, S., Landen, C. N., Finegold, M., Thompson, A. R., Read, M. S., Brinkhous, K. M., and Woo, S. L. C. (1993). In vivo gene therapy of hemophilia B: Sustained partial correction in factor IX-deficient dogs. Science 262, 117-119. Kotin, R. M., Siniscalco, R., Samulski, R. J., Zhu, X., Hunter, L., Laughlin, C. A., McLaughlin, S., Muzyczka, N., Rocchi, M., and Berns, K. I. (1990). Site-specific integration by adeno-associated virus. Proc. Natl. Acad. Sci. USA 87, 2211-2215. Kozarsky, K. F., and Wilson, J. M. (1993). Gene therapy: Adenovirus vectors. Current Opinion Genet. Dev. 3, 499-503. Lozier, J. N., Thompson, A. R., Hu, P.-C., Read, Brinkhous, K. M., High, K., and Curiel, D. T. (1994). Efficient transfection of primary cells in a canine hemophilia B model using adenovirus-polylysine-DNA complexes. Hum. Gene Ther. 5, 313-322. Morgenstern, J. P., and Land, H. (1991). Choice and manipulation of retroviral vectors. In 'WIethods in Molecular Biology: Gene Transfer and Expression Protocols" (E. J. Murray, ed.), pp. 181-206. Humana Press, Clifton, New Jersey. Palmer, T. D., Thompson, A. R., and Miller, A. D. (1989). Production of human factor IX in animals by genetically modified skin fibroblasts: Potential therapy for hemophilia B. Blood 73(2), 438-445. Perales, J. C., Ferkol, T., Molas, M., and Hanson, R. W. (1994). An evaluation of receptormediated gene transfer using synthetic DNA-ligand complexes. Eur. J. Biochem. 226(2), 255-266. Samulski, R. J., Chang, L. S., and Shenk, T. E. (1987). A recombinant plasmid from which an infectious adeno-associated virus genome can be excised in vitro and its use to study viral replication. J. Virol. 61, 3096-3101. Schimpf, K., Brackman, H. H., and Kreuz, W. (1989). Absence of anti-human immunodeficiency virus types 1 and 2 seroconversion after the treatment of hemophilia A or von Willebrand's disease with pasteurized factor VIII concentrate. N. Engl. J. Med. 321, 1148-1152. Schimpf, K., Mannucci, P. M., and Kreuz, W. (1987). Absence of hepatitis after treatment with a pasteurized factor VIII concentrate in patients with hemophilia and no previous transfusion. N. Engl. J. Med. 316, 918-922. St. Louis, D., and Verma, I. (1988). An alternative approach to somatic cell gene therapy. Proc. Natl. Acad. Sci. USA 85, 3150-3154. Strauss, S. E. (1984). Adenovirus infection in humans. In '~Fhe Adenoviruses" (H. S. Ginsberg, ed.), pp. 451-496. Plenum Press, New York.
134
JOHANNES WALTER AND KATHERINE A. HIGH
Troisi, C. L., Hollinger, F. B., and Hoots, W. K. (1993). A multicenter study of viral hepatitis in a United States hemophilic population. Blood 81, 412-418. Tuddenham, E. G. D., Cooper, D. N., Gitschier, J., et al., (1991). Haemophilia A: Database of nucleotide substitutions, deletions, insertions and rearrangements of the factor VIII gene. Nucleic Acids Res. 19, 4821-4833. Tuddenham, E. G. D., Lane, R. S., Rotblat, F., Johnson, A. J., Snape, T. J., Middleton, S., Kernoff, P. B. (1982). Response to infusions of polyelectrolyte fractionated human factor VIII concentrate in human hemophilia A and yon Willebrand's disease. Br. J. Hematology 52, 259-267. Vehar, G., Keyt, B., Easton, D., et al., (1984). Structure of human factor VIII. Nature 313, 337-342. Walter, J., You, Q., Hagstrom, N., Sands, M., and High, K. A. (1996). Successful expression of human factor IX following repeat administration of an adenoviral vector in mice. Proc. Natl. Acad. Sci. USA, 93, 3056-3061. Yang, Y., Nunes, F. A., Berencsi, K., Furth, E. E., and Gonczol, E. (1994). Cellular immunity to viral antigens limits El-deleted adenoviruses for gene therapy. Proc. Natl. Acad. Sci. USA 91, 4407-4411. Yang, Y., Trinchieri, G., and Wilson, J. (1995). Recombinant IL-12 prevents formation of blocking IgA antibodies to recombinant adenovirus and allows repeated gene therapy to mouse lung. Nat. Med. 9(1), 890-893. Yao, S., and Kurachi, K. (1992). Expression of human factor IX in mice after injection of genetically modified myoblasts. Proc. Natl. Acad. Sci. USA 89, 3357-3361. Yao, Y., Wilson, J. M., Nabel, E. G., Kurachi, S., and Hachiya, H. L. (1991). Expression of human factor IX in rat capillary endothelial cells: Toward somatic gene therapy for hemophilia B. Proc. Natl. Acad. Sci. USA 88, 8101-8105. Yoshitake, S., Schach, B. G., Foster, D. C., Davie, E. W., and Kurachi, K. (1985). Nucleotide sequence of the gene for human factor IX (antihemophilic factor B). Biochemistry 24, 3736-3750. Zatloukal, K., Cotten, M., Berger, M., Schmidt, W., Wagner, E., and Birnstiel, M. L. (1994). In vivo production of human factor VIII in mice after intrasplenic implantation of primary fibroblasts transfected by receptor-mediated adenovirus-augmented gene delivery. Proc. Natl. Acad. Sci. USA 91(11), 5148-5152.
ADVANCES IN VETERINARY MEDICINE, VOL. 40
Genes Controlling Retroviral Virulence FREDERICK J. FULLER Department of Microbiology, Pathology, and Parasitology, North Carolina State University, College of Veterinary Medicine, Raleigh, North Carolina 27606 I. II. III. IV. V. VI. VII.
VIII. IX.
Introduction Retrovirus Structure and Replication Additional Retrovirus Genes Feline Acquired Immunodeficiency Syndrome A. FeLV Subgroups Murine AIDS Simian AIDS Murine Leukemia Virus A. Spleen Focus-Forming Virus B. MuLV Neurologic Disease C. Osteopetrosis D. HTLV-1-Associated Myelopathy E. Reticuloendotheliosis Virus Endogenous Retroviruses A. Mouse Mammary Tumor Virus Lentiviruses A. Human Immunodeficiency Virus B. Equine Infectious Anemia Virus C. Feline Immunodeficiency Virus D. Visna, Maedi, Progressive Pneumonia Virus E. Caprine Arthritis Encephalitis Virus References
I. I n t r o d u c t i o n
Retroviruses under normal circumstances infect cells and yet inflict relatively little damage to the infected cell. The cell, however, remains infected for the lifespan of that cell or any daughter cells derived from 135 Copyright 9 1997 by Academic Press. All rights of reproduction in any form reserved. 1093-975X/97 $25.00
136
FREDERICK J. FULLER
t h a t infected cell. Retroviruses that belong to the lentivirus subgroup are an exception because these viruses typically can directly kill the infected cell. The types of diseases induced by lentiviruses are also somewhat different presumably, at least in part, because of additional genes encoded in these complex retroviruses. Some retrovirus and retrovirus-like sequences are present in germline cells and are referred to as endogenous retroviruses. These sequences usually are nonexpressed or are expressed at very low levels and are discussed in this article. A wide range of biologic effects are discerned from a variety of retrovirus infections from malignancies, benign proliferative states that can lead to malignancies, degenerative diseases, viremias, and insertional activation or inactivations of cellular genes and transactivations of cellular genes. These activities can result from normal replication of these retroviruses or the less frequent generation of replication-defective eeiruses that can arise during persistent infections. These replicationdefective viruses are probably not normally t r a n s m i t t e d from host to host but have profound implications in the host in which they arise. "Genes Involved in Oncogenesis" reviews the role that retroviral genes play in the development of malignancies and the reader is referred there for an excellent coverage of this topic. This article discusses those retroviral diseases and what is known concerning the genes and controlling regions that have been shown to influence the development of the broad category of nonmalignant diseases.
II. Retrovirus Structure and Replication The reader is referred to several comprehensive reviews of the details of both retrovirus structure and replication (Coffin, 1996). This review summarizes the available information concerning retrovirus gene structure and replication of exogenous retroviruses. The retrovirus family was previously subdivided into three main groups: the oncoviruses, the lentiviruses, and the spumaviruses. However, the International Committee on the Taxonomy of Viruses currently recognizes seven genera of retroviruses (Coffin, 1992) with the likely possibility of additional genera being added as a result of recently described retroviruses of fish and insects. The seven genera include the avian leukosis-sarcoma virus group, the m a m m a l i a n C-type virus group, the B-type virus group, the D-type virus group, the h u m a n T-cell leukemia virus (HTLV)-bovine leukemia virus (BLV) group, the lentivirus group, and the spumavirus group. Table 1 lists the retrovirus g e n e r a - -
GENES CONTROLLINGRETROVIRALVIRULENCE
137
recognized as of 1996mwith examples of viruses as well as some of the currently unclassified retroviruses. Table 1 also includes some biologic features of some of these retroviruses along with the designation of exogenous (transmitted horizontally) or endogenous (transmitted from parent to offspring as a provirus integrated into the germ line). The genomes of all retroviruses have some common features. Figure 1 demonstrates the basic features of the approximately 8 to 12 kilobase (kb) retrovirus ribonucleic acid (RNA). Two copies of this RNA are packaged per infectious virus particle. Features of this RNA include a 5' cap structure, a 3' polyadenylate tail, a short repeat (R) sequence at both ends, a transfer RNA (tRNA) primer binding (PB) site, a splice donor (D) site and at least one (but can be more) splice acceptor (A) site(s), an encapsidation site required for packaging of viral RNA into virus particles, and a U5 region, which contains unique information and is the first region copied into DNA during reverse transcription. The U5 region then becomes the 3' end of the long terminal repeat (LTR). The U3 becomes the 5' end of the LTR and contains a number of cis-acting signals necessary for virus replication. The U3 region contains signals that can be recognized by the cellular transcriptional machinery and is an important determinant of viral gene expression and, therefore, is known to play an important role in many retrovirusinduced disease processes. The four coding regions of a generic retrovirus include the gag, pro, pol, and env genes. The gag gene region encodes the group-specific antigens. The gag gene is translated from the full-length viral RNA to yield a polyprotein that is processed to yield three to five capsid proteins depending on the retrovirus (minimally, the matrix, capsid, and nucleic acid binding protein). The pro gene encodes a protease that processes the gag and pol polyproteins. The pol gene encodes the reverse transcriptase (RT) and an integrase (IN) protein. In addition, some retroviruses (equine infectious anemia virus, feline immunodeficiency virus, Mason-Pfizer monkey virus) have
Splice donor Primer binding site site %~1~ - -
,
, .
.
.
Splice acceptor Polypurine site~, tract
.
pol
env
H ! AAAAAA. 99
A "~ ev,~., Encapsidation U3 R U5 site FIG. 1. Schematicdiagram of the gene coding regions and important cis-acting signals of retroviruses. These features are present on all retroviruses.
TABLE 1 LIST OF RETROVIRUS GROUPS Genus
Virus
Properties
Avian leukosis-sarcoma virus group
Rous sarcoma virus Avian myeloblastosis virus Avian erythroblastosis virus Avian myelocytomatosis virus Rous-associated virus Rous-associated virus-0
Exogenous, 8rc oncogene Exogenous, myb oncogene Exogenous, erb-A and -B oncogene Exogenous, myc oncogene Exogenous, B-lymphomas, osteopetrosis Endogenous, no known disease
Mammalian C-type virus group
Moloney murine leukemia virus Harvey murine sarcoma virus Abelson murine leukemia virus Feline leukemia virus Simian sarcoma virus Reticuloendotheliosis virus Spleen necrosis virus
Exogenous, T-cell lymphomas Exogenous, H-ras oncogene Exogenous, abl oncogene Exogenous, T-cell lymphoma, immunodeficiency Exogenous, sis oncogene Exogenous viruses of birds
B-type viruses
Mouse mammary tumor virus
Endogenous and exogenous, milk-borne, mammary carcinoma, T lymphoma
D-type viruses
Mason-Pfizer monkey virus Simian AIDS viruses
Exogenous Immunodeficiencies in monkeys
HTLV-BLV group
Human T-cell leukemia virus 1 and 2 Bovine leukemia virus
T-cell lymphoma, neurologic disease B-cell lymphoma
Lentivirus group
Human immunodeficiency virus 1 and 2 Simian immunodeficiency virus
Acquired immunodeficiency syndrome Usually no disease in natural primate host and AIDS-like disease in other primates Immune disorders, ocular abnormalities, gingivitis Exogenous, possibly benign Neurologic or lung disease in sheep (occasionally goats) Encephalitis in young, arthritis in adults Periodic episodes of fever, anemia, thrombocytopenia
Feline immunodeficiency virus Bovine immunodeficiency virus Visna/maedi-ovine progressive pneumonia virus Caprine arthritis-encephalitis virus Equine infectious anemia virus Spumavirus group
Simian foamy virus Human foamy virus Feline syncytium-forming virus
Exogenous, possibly benign Exogenous, possibly benign Exogenous, possibly benign
Unclassified retroviruses
Walleye dermal sarcoma virus Damselfish neurofibromatosis virus
Exogenous, benign dermal sarcomas of fish Exogenous, Schwann-cell neoplasms in damselfish Probably endogenous, probably benign Associated with neoplasms Associated with sarcomas in clams
Russell's viper retrovirus Corn snake retrovirus Retrovirus of molluscs
140
FREDERICK J. FULLER
a deoxyuridine triphosphatase (dUTPase) encoded in the pol gene region. The pro and pol genes are not strictly considered separate genes because in the case of the pro region it is either encoded as a portion of the gag or pol region depending on the retrovirus and the pol region is encoded on the same messenger RNA (mRNA) as the gag gene. The env gene encodes two envelope glycoproteins that are cleaved from a precursor protein. The surface (SU) protein binds to cell-surface receptors and mediates attachment of virus particles to cells and the transmembrane (TM) protein anchors the SU protein to the lipid envelope of the virus particle. The env gene is translated from a singly spliced mRNA, unlike the gag, pro, and pol gene products. III. Additional Retrovirus Genes
Retroviruses belonging to the murine leukemia-related and avian leukemia-related virus groups require only the gag, pro, pol, and env genes for all aspects of their replication. However, the other retrovirus groups encode a variety of additional gene products (Fig. 2) that are either necessary for replication or important for modifying the pattern of viral (or perhaps cellular) gene expression. The m a m m a l i a n B-type viruses, like mouse m a m m a r y tumor virus (MMTV), encode a gene designated the sag gene for superantigen. MMTV superantigen is expressed on the surface of infected cells and activates a subpopulation of CD4 + T cells. The sag gene may play an important role in ensuring efficient transmission of MMTV via milk to neonates possibly by recruitment of susceptible T cells to m a m m a r y tissue (Golovkina et al., 1992; Held et al., 1993). HTLV and BLV encode two additional genes at their 3' ends from doubly spliced mRNAs. The 3' coding region of these viruses were originally designated the X-gene region. The tax gene is named for the transactivator from the X-gene region. The tax gene product, a 40,000 dalton protein is required for efficient promoter activity from the viral LTRs and is therefore essential for replication. The rex gene is desig-
FIG. 2. Diagram displaying the coding regions of each of the seven representative retrovirus groups. The open boxes indicate the coding regions of the gag, pro, pol, or env gene regions of each respective virus. The solid boxes indicate either regulatory or accessory genes that are encoded by these selected retroviruses. Note that the complexity (number of regulatory and accessory genes) increases from top to bottom in this listing.
1-
gag
Pro
env
U 1
-
1
gag
pro
U
env
1-
pol
I
POI
Pro
env
I
1 0 gag 0
0
POI
1-1
env
Pro
CIIl
0I env gag
O
1-
POI
pro
I--&;---
H-
Human Immunodeficiency mrus
, 0 m vlf
gag
i
pro
Pol
vpr
r
rex
I
7
env
~---ia!---
I
1
nef
Irn
- - -rev ----#
VPU
m
gag
pro
0
be12
I env be13 Pol
bell
n.-
bet
142
F R E D E R I C K J. F U L L E R
nated as a regulator of RNA splicing from the X region. The rex gene product is a 27,000 dalton protein that is required for the synthesis of full-length and e n v mRNAs. When the rex is not expressed, only tax and rex mRNAs are produced and therefore no virus replication can occur.
The lentiviruses have a complex pattern of additional genes that, in the case of the h u m a n immunodeficiency virus (HIV), number six that have been identified as of 1996. Some of the lentiviruses have predicted open reading frames that have yet to be determined if they are expressed or their functions determined. It is also clear that not all lentiviruses have all of the additional genes contained in the HIV genome. For example, equine infectious anemia virus has only three additional open reading frames besides the g a g , p o l , and e n v genes designated tat, rev, and $2. However, by alternate splicing, at least one additional protein, ttm (truncated transmembrane), is encoded by combining a portion of the t a t and e n v genes. The t a t gene has a transactivating function that can significantly stimulate transcription of the HIV LTR by increasing transcriptional initiation or elongation (Peterlin et al., 1993). In HIV, t a t requires a cisacting RNA element, designated the TAR element. The TAR element forms a stem-loop structure located at the 5' end of viral RNA transcripts. Genes with a t a t or t a t - l i k e function have been identified for most lentiviruses. Although t a t is essential for efficient replication of lentiviruses, the role of t a t in the pathogenesis of lentivirus disease is not clearly defined. A number of studies have demonstrated that HIV-1 t a t has the capacity to activate promoters of some cellular genes and heterologous viruses. Kaposi's sarcoma-like lesions have been observed in some mice made transgenic for t a t gene expression (Vogel et al., 1988) suggesting a possible link between this gene and the h u m a n disease. The rev gene performs a function analogous to the rex gene of HTLV and BLV. The rev gene regulates splicing and transport of viral RNA from the nucleus to the cytoplasm (Parslow, 1993). When the rev gene product is present in infected cells, the production of full-genomelength viral RNAs is greatly favored over subgenomic spliced viral mRNAs. The v i f (viral infectivity factor) gene is present in HIV-1 and in most of the other lentiviruses, including HIV-2, simian immunodeficiency virus (SIV), and feline immunodeficiency and visna viruses. The v i f gene product has been shown to significantly enhance the infectivity of HIV-1 virus particles. This gene has been shown to be essential for establishing a productive HIV-1 infection in peripheral blood T lym-
GENES CONTROLLINGRETROVIRALVIRULENCE
143
phocytes as well as monocytes and macrophages. Although it is not currently known how the vif gene product performs its functions, it is known that the protein is a cytoplasmic protein that tightly associates with cell membranes (Goncalves et al., 1995). It has been speculated that v i f m a y counteract a cell-specific inhibitory factor(s) that prevents HIV-1 particle assembly (Michaels et al., 1993). The vpr gene product of HIV-1, HIV-2, and SIV is a virion-associated protein that is believed to play a role in facilitating entry of the viral core into the nucleus of nondividing cells. In addition, vpr prevents the establishment of chronically HIV producer cells by arresting the infected cells in the G2/M phase of the cell cycle. This cell-cycle arrest in G2 is brought about by a vpr-induced inhibition of the activation of the p34cdc2/cyclin B complex (He et al., 1995). It is speculated that the cell-cycle arrest might delay or prevent apoptosis of infected cells thereby allowing the infected cells to produce more virus (Hoch et al., 1995). A vpr or vpx m u t a n t of SIV still induces AIDS in Rhesus monkeys; however, an SIV double v p r / v p x deletion m u t a n t was highly attenuated on an SIVmac239 virus (Gibbs et al., 1995). The vpx gene is a homolog of vpr that is present in various SIV strains and in HIV-2. The authors of these studies suggested that, because vpr and vpx are related genes that may have overlapping or duplicative functions, deletion of both of these virion-associated gene products may completely eliminate their overlapping functions and result in less virus growth in vivo and result in virus attenuation. The vpu gene of HIV-1 is associated with two known functions. The first function is an induction of CD4 protein (HIV-cell receptor) degradation in the endoplasmic reticulum. The second function that has been recently described is a monovalent cation channel activity when expressed in Xenopus oocytes (Ewart et al., 1996). The vpu protein is an oligomeric, integral membrane protein that contains a hydrophobic t r a n s m e m b r a n e domain and a polar phosphorylated cytoplasmic tail. The nef gene of HIV-1, HIV-2, and SIV encodes a 27-34 kDa myristylated, inner plasma membrane protein that is not required for virus replication. Experimental infections of rhesus macaques with nef gene mutants of SIV have shown that nef is required for the maintenance of high viral loads and progression to AIDS. The nef gene has also been shown to downregulate the expression of CD4 molecules on the surface of cells by mediating the internalization and subsequent degradation of the receptor in the lysosomes (Michael et al., 1995; Kestler et al., 1991). The nef gene also has been shown to enhance the infectivity of HIV-1 virions. This enhancement of infectivity appears to be mediated at the level of altering the processing of the inter-
144
FREDERICK J. FULLER
nalized virus core, making it more competent for viral deoxyribonucleic acid (DNA) synthesis (Aiken and Trono, 1995). There is also evidence that nef is capable of both enhancement and inhibition of T-cell activation. This may be a consequence of the release of the p561ck tyrosine kinase from the CD4 cytoplasmic domain (Aiken et al., 1994; Rhee and Marsh, 1994).
IV. Feline Acquired Immunodeficiency Syndrome Feline leukemia virus (FeLV) infection can result in either neoplastic or a variety of nonneoplastic diseases. The neoplastic diseases are classified based on the cell type that has undergone malignant transformation and the location of the primary lesion. The wide variety of nonneoplastic diseases has made diagnosis of many of these diseases very difficult. These nonneoplastic diseases can be confused with other feline diseases. Diseases such as a nonregenerative anemia, hemolytic anemia, thymic atrophy, a panleukopenia-like syndrome, glomerulonephritis, reproductive disorders, and immunosuppression (feline acquired immunodeficiency syndrome, FAIDS) (reviewed in Hardy, 1993). A. FELV SUBGROUPS There are three subgroups of FeLV, which are designated A, B, and C. These subgroups are based on the property of superinfection interference in which two A-type viruses will interfere with each other when infecting a population of cells previously infected with an A-type virus. For example, no interference is exhibited between an A- and B- or Aand C-type virus. The A subgroup FeLV is the most common subgroup isolated from domestic cats. This virus replicates to high titers in infected cats and is always present when the B and C subgroup viruses are isolated. The subgroup type determines the type of disease that can be induced. For example, the Rickard strain of FeLV is a combination of A and B virus types, which induces a high incidence of persistent viremia and severe immunosuppression by 20-30 weeks postinfection. A subgroup A infection alone (Glasgow strain of Rickard) results in a low incidence of persistent viremia, hemorrhagic enteritis, and neutropenia. The subgroup C virus is found in plasma after a delay and may well be replication defective. The Kawakami-Theilen strain of FeLV is a mixture of A, B, and C viruses and causes severe erythrosuppresion and death rapidly (within 9 weeks) postinoculation of newborn kittens. Surprisingly, this virus is nonpathogenic in kittens older than 6 months
GENES CONTROLLING RETROVIRAL VIRULENCE
145
of age and adults. The Sarma strain is a subgroup C virus that induces a viremia and erythroid aplasia in infected kittens. In FeLV-infected cats, immunosuppression coincides with the onset of bone marrow cell infection and precedes detectable neoplastic transformation. During this time, infected cats are susceptible to a variety of opportunistic infections. Diseases such as enteritis, gingivitis, bacteremias, pneumonia, feline infectious peritonitis, HemobartoneUa felis infections, or a number of parasitic infections. Kristal et al. (1993) demonstrated that the FeLV surface protein gp70 determines the subgroup interference type (A, B, or C), the T-cell killing properties of the virus, the host range specificities, as well as the growth kinetic properties. This group's FeLV-FMDS-EECC chimeric clone causes T-cell killing presumably because of efficient replication associated with massive superinfection. The superinfection and the cell killing can be inhibited by adding virus neutralizing antiserum to block superinfection.
V. Murine AIDS Some inbred strains of mice when infected with the LP-BM5 strain of murine leukemia virus (MuLV) develop a severe immunodeficiency that shares some similarities with human AIDS. Clinical features including lymphadenopathy, profound immunodeficiency of both the humoral and cellular immunity, splenomegaly, and hypergammaglobulinemia are displayed. B-cell lymphomas can develop in the late stage of this disease process. The LP-BM5 strain of MuLV is actually a mixture of ecotropic, polytropic, and defective MuLV viruses. The defective MuLV encodes an unusual 60-kDa protein related to gag that is probably responsible for the immunosuppression. More recent data indicates that the defective gag-related protein may have superantigen properties that result in a selective expansion of certain T cells. These properties may misdirect the immune response leading to an inability to mount an effective immune response against a potential pathogen (Mosier et al., 1985; Klinken et al., 1988).
VI. Simian AIDS The Simian type-D retroviruses (SRV) are a group of exogenous viruses that are indigenous to Asian macaques. There are also related type-D endogenous retroviruses found in simians. Only the exogenous viruses have been shown to be pathogenic. The exogenous type-D vi-
146
FREDERICK J. FULLER
ruses are nononcogenic, horizontally transmitted, and have the capacity to cause a fatal immunodeficiency syndrome in macaques. SRVs are not closely related to HIV or to SIV. SRV infections in primate centers have been shown to be important causes of fatal immunosuppressive diseases in Asian macaques. For example, in 1983-84 at the California Primate Center, nearly all of the adult macaques in a breeding colony were infected with either SRV-1 or SRV-2 and the mortality was close to 50% in macaques less t h a n 2 years of age. The prototype virus of this group, Mason-Pfizer monkey virus (MPMV) was isolated in 1970 from a rhesus monkey m a m m a r y tumor. Experimental infection of infant rhesus monkeys with MPMV results in a wasting syndrome with anemia, thymic atrophy, neutropenia, and opportunistic infections (reviewed in Gardner et al., 1994). VII. Murine L e u k e m i a Virus A. SPLEEN FOCUS-FORMING VIRUS There are some strains of MuLV, designated the Friend strains, that have the capacity to induce an abnormal proliferation of erythroblasts in infected mice. Although the nondefective MuLV possesses the capacity to induce this condition alone, it has been frequently observed that a replication-incompetent defective virus known as spleen focus-forming virus is a potent inducer of the erythroblastosis. Infected mice displaying the erythroblastosis condition are predisposed to developing a malignant erythroleukemia condition. It has been shown that a modified MuLV envelope protein, p55, is responsible for the erythroblastosis condition. Three modifications of the MuLV env gene are required for the conversion of this retroviral protein into a protein that can apparently function as a potent erythroid growth factor. The MuLV env gene recombines with an endogenous murine retrovirus env gene, resulting in an altered sequence, a portion of the surface domain of the envelope gene; the t r a n s m e m b r a n e region is deleted; and the membrane-spanning domain of the t r a n s m e m b r a n e region is extended (Ruscetti et al., 1990). B. MuLV NEUROLOGIC DISEASE Wild mice captured from the Lake Casitas region in California have a high incidence of hind-limb paralysis associated with elevated levels of MuLV expression (Gardner et al., 1973). An isolate of this MuLVo designated Cas-Br-E causes the same hind-limb paralysis when infect-
GENES CONTROLLING RETROVIRAL VIRULENCE
147
ing inbred strains of mice. The neurologic disease seems to be a consequence of high levels of virus replication in the central nervous system. Mice older than 10 days of age that are inoculated with this virus do not develop the disease and do not become viremic. Mapping of this phenotype with nonneurovirulent strains of MuLV indicate that the env gene is responsible for this property, although the LTR and some gag-pol sequences influence the rate of disease progression. The development of neurologic disease can take up to 12 months and the precise mechanism responsible for disease has not been defined. It has also been shown that a temperature-sensitive mutant of Moloney MuLV (MoMLV) and a rat-passaged variant of Friend MuLV have the capacity to induce neurologic disease as well. The neurovirulence of the temperature-sensitive mutant of MoMLV has been mapped to the env gene like the Cas-Br-E strain of virus. The envelope proteins accumulate in the infected cell because of defective processing (Wong, 1983). C. OSTEOPETROSIS
Certain strains of avian leukemia virus (ALV) can induce osteopetrosis in chickens. A proliferation of osteoblasts in the long bones of the leg leads to a very apparent thickening of the legs. The osteoblasts contain large amounts of unintegrated viral DNA (Robinson and Miles, 1985) and high amounts of virus. Although the mechanism of this condition is not clear, it is likely that the unusually high amount of virus replication in this tissue site results in a disruption of the delicate balance between bone growth and resorption that normally occurs. Genetic mapping studies comparing viral chimeras between osteopetrosis-inducing and nonosteopetrosis-inducing ALV strains of virus indicate that sequences in the g a g gene region near the 5' LTR determine the ability to induce osteopetrosis. D. HTLV-1-AssocL~TED MYELOPATHY
In 1985, an association between infection of a group of West Indian humans with a neurologic disease and infection with HTLV-1 was noted. The neurologic disease was identified as tropical spastic paraparesis (TSP). A Japanese group also noted a group of humans with similar neurologic problems also with associated HTLV-1 infections (Osame et al., 1986; Akizuki et al., 1987). This disease was designated HTLV-l-associated myelopathy (HAM). The pathogenesis of this neurologic disease is poorly understood and is under study (Furukawa et al., 1992).
148
FREDERICK J. FULLER
E. RETICULOENDOTHELIOSIS VIRUS
Some strains of reticuloendotheliosis virus (REV) can cause a variety of nonneoplastic diseases in chickens and ducks. These diseases include anemia (Kawamura et al., 1976), abnormal feather development (Koyama et al., 1980), atrophy of the bursa and thymus (Mussman and Twiehaus, 1971), enteritis (McDougall et al., 1978), proventriculitis (Jackson et al., 1977), necrosis of the liver and spleen (Purchase and Witter, 1975), a runting disease syndrome (Mussman and Twiehaus, 1971), and immunosuppression (Witter et al., 1979). In addition, REV can induce neoplastic conditions. The immunosuppressive properties of REV are thought to be responsible for secondary bacterial and viral infections. Cell-mediated immune responses are impaired because of infection with this virus. The genetic basis of these diseases is not currently known.
VIII. Endogenous Retroviruses A. MOUSE
MAMMARY
TUMOR
VIRUS
The typical mouse contains approximately 1000 copies of endogenous retroviral sequences. There are at least three known cases in which genetic-based diseases have been caused by insertion of endogenous murine retrovirus sequences in inbred strains of mice. The hairless mutation (Stoye et al., 1988), a coat-color mutation (Jenkins et al., 1981), and a lethal collagen defect (Harbers et al., 1984). In addition, a m a m m a l i a n B-type endogenous virus, MMTV, is a less common endogenous murine virus that requires virally encoded superantigen (sag) gene to facilitate efficient transmission to neonates via the milk. Superantigens bind to MHC proteins and T-cell receptor protein variable regions and, as a consequence, they activate all T lymphocytes of a particular T-cell receptor class. The proper receptor bearing T cells will proliferate for a period of time but eventually will be clonally deleted from the T-cell population. MMTV superantigen from C3H mice causes T cells of the V-beta-14 class to proliferate. Transgenic mice from a background that do not normally express the C3H MMTV sag gene will delete the V-beta-14 class of T lymphocytes when made to express this gene. The elimination of T cells responsive to MMTV superantigen renders the mouse resistant to infection by that virus. Therefore, the sag gene is likely required in vivo for maintenance and transmission of this virus.
GENES CONTROLLING RETROVIRAL VIRULENCE
149
IX. Lentiviruses A. HUMAN IMMUNODEFICIENCY VIRUS Although it is not possible to directly test the influence of various HIV-1 and -2 genes on pathogenesis in h u m a n s (as compared to SIV), there have been some interesting studies done on h u m a n s infected with HIV-1 who have survived for 10 years or more and remain healthy with no signs of clinical progression to acquired immunodeficiency syndrome (AIDS) and normal CD4 lymphocyte counts. In come cases, the virus from these long-term nonprogressors have been analyzed and found to contain some interesting and suggestive mutations. Deacon et al. (1995) analyzed the virus from a cohort (unrelated individuals) of transfusion recipients from a B loodbank in Sydney, Australia. The HIV-1 infected blood donor and seven recipients of blood from the infected individual had been infected for 10.755 to 14 years without any cohort member developing any AIDS- or HIV-related symptoms or receiving any antiviral therapy. Proviral DNA was analyzed from seven individuals and it was discovered that all seven virus populations contain deletions of various sizes in the nef gene and the U3 LTR domain because there is some overlap between the nef gene and the U3 region. Similar results have been reported for another individual long-term nonprogressor infected with HIV-1 in which deletions in the same nef-LTR regions were observed (Kirchhoff et al., 1995). These studies are consistent with the SIV studies that indicated that the SIV nef gene played an essential role in the induction of disease (as discussed previously). B. EQUINE INFECTIOUS ANEMIA VIRUS Equine infectious anemia virus (EIAV) causes a persistent, life-long infection of horses characterized by periodic episodes of fever, thrombocytopenia, and anemia, which if severe can result in death within the first two weeks of infection. A comparison of sequence variation between a highly virulent strain of EIAV and an a t t e n u a t e d strain of this virus reveals a large number of sequence differences that are predominantly in the U3 region of the LTRs and the surface protein sequence (gp90). It has been possible only since 1996 to obtain infectious molecular clones of this virus t h a t promptly induce disease; it now should be feasible to determine the role of various viral genes and cis-acting sequences on the influence of disease expression.
150
FREDERICK J. FULLER C. FELINE IMMUNODEFICIENCYVIRUS
The disease induced by feline immunodeficiency virus (FIV) appears in many ways to mimic the effects of the HIV-1- and -2-induced immunosuppressive disease seen in humans. Cats develop an acute infection syndrome with a mild fever and transient generalized lymphadenopathy. The acute phase of infection is followed by an asymptomatic stage when the CD4:CD8 cell ratio declines because of a progressive decrease in CD4 cells. The asymptomatic phase is followed by a phase in which a variety of clinical disorders occur that closely resemble those in humans with AIDS. Clinical signs in cats include a variety of opportunistic infections, chronic gingivitis and stomatitis, chronic upper respiratory infections, as well as superficial and systemic fungal infections (Pedersen et al., 1989). While the virus by itself can induce inversions of CD4 to CD8 lymphocytes in infected cats, this infection appears to most importantly reduce the efficiency of a Thl (T-cell-based) immune response. For example, cats that were experimentally infected with the NCSU strain of FIV and were subsequently challenged with a normally nonlethal dose of the intracellular parasite Toxoplasma gondii succumbed to this dual infection. An analysis of the immune response during the T gondii infection, as shown in Fig. 3, indicated that the FIV-infected cats failed to generate an IL12-dependent Type-1 immune response to T gondii infection but rather responded with an ineffective IL4-induced Type-2 response (Levy et al., personal communication). Because resistance to T gondii depends primarily on a Thl-immune response in cats, FIV-infected cats were unable to respond to the parasite infection with a good Thl response. As a result, the FIV-infected cats developed a severe generalized toxoplasmosis resulting in acute and often fatal interstitial pneumonia. In contrast, cats infected only with T gondii developed a transient, mild clinical disease characterized by anorexia, lethargy, and multifocal chorioretinitis. D. VISNA, MAEDI, PROGRESSIVE PENUMONIA VIRUS
The most common disease manifestation of ovine lentivirus infected sheep is dyspnea and a wasting disease characterized by severe loss of body weight (Narayan and Cork, 1985). The disease was first described in epidemic form in Iceland by Sigurdsson after the introduction of infected sheep from Europe during the mid-1930s. The primary clinical manifestations included dyspnea and wasting in infected sheep. A neurologic complication (visna) of the virus infection was observed in the Icelandic breeds of sheep during the height of the epidemic in the
151
GENES CONTROLLING RETROVIRAL VIRULENCE
|
IL ll3/TNFcx IFNs~ IL 12 IL I~ II::NT/TNFa
TNFa
|
@
L
j
.|
~ IL 10 9
t
)| IFN'y/IL12
IFN? IgN~
d E )
) Positive Signal 9 l ~ Negative Signal (greatly enhanced in FIV infected cats and H1V infected humans) FIG. 3. Autocrine and paracrine cytokine regulation of the innate and T-cell-mediated immune response to T gondii. The initial interaction of the macrophage (MP) with the pathogen (TOX) causes synthesis of IFN~ and TNF~ and their respective receptors (step 1). The macrophage-derived ILl2 in collaboration with TNF~, or ILI~ (step 2) induces synthesis of IFN~ by the NK cell, which becomes the major IFN~ producer (step 3). NKcell-derived IFN~ combined with TNF~ causes increased microbiocidal activiation of the intracellular macrophage functions (step 3) as well as an upregulation of ILl2 necessary to stimulate the T h l immune response (step 4). Antigen-presenting cells (APCs) stimulate naive T cells (Th0) to differentiate into Thl (inflammatory-T) (step 5) or Th2 (helperT) cells and, although the specific signals are not clear, the cytokines present at the time of antigen presentation play an important role. Cytokines that promote a T h l response include ILl2, IFN% IFN~, and TNF~. Once activated, Thl cells leave the lymph nodes, enter the bloodstream, and specifically migrate to regions of inflammation as directed by adhesion molecules where they perform their primary effector function of activating macrophages through the production of IFN~ (step 6). ILl0 is an important regulator of this process because a macrophage stimulated to produce ILl0 will inhibit the synthesis of ILl2 and therefore inhibit the Thl response (step 7) in FIV-infected cats and HIVinfected humans, ILl0 produced from either macrophages, CD8 § T cells or Th2 cells downregulate ILl2 production from macrophages. The decreased ILl2 production greatly inhibits the development of the T h l inflammatory T-cell response. Courtesy of Wayne B. Tompkins.
1950s. The disease is characterized primarily by inflammation of specific organ or tissue. Affected tissues are characterized by a marked infiltration of mononuclear cells to such an extent that normal tissue
152
FREDERICK J. FULLER
architecture is disrupted and tissue and organ function is impaired. Affected tissues can include the lung, spleen, joints, lymph nodes, brain, and spinal cord where these infiltrates can be so extensive as to disrupt normal function. Further study is required to determine the role of viral gene expression in the development of this lentivirus infection. E. CAPRINE ARTHRITIS ENCEPHALITIS VIRUS The most common clinical manifestation associated with caprine arthritis encephalitis virus (CAEV) infection in goats is synovitis in adult dairy goats. The synovitis has a gradual onset and slowly progresses to a crippling arthritis (Narayan and Cork, 1985). Newborn goats in herds of CAEV-infected goats can develop a rapidly progressing neurologic disease leading to paralysis and death. As in ovine lentivirus infection, CAEV-infected goats will display severe inflammation in affected tissues. Additional study is required to determine the role of viral gene expression in the development of disease for this lentivirus. REFERENCES Aiken, C., Konner, J., Landau, N. R. Lenberg, M. E., and Trono, D. (1994). N e f induces CD4 endocytosis: Requirement for a critical dileucine motif in the membrane-proximal CD4 cytoplasmic domain. Cell 76, 853-864. Aiken, C., and Trono, D. (1995). N e f stimulates human immunodeficiency virus type 1 proviral DNA synthesis. J. Virol. 69, 5048-5056. Akizuki, S., Nakazato, O., Higuchi, Y., et al. (1987). Necropsy findings in HTLV-1 associated meylopathy. Lancet 1, 156-157. Aziz, D. D., Hanna, Z., and Jolicoeur, P. (1989). Severe immunodeficiency disease induced by a defective murine leukaemia virus. Nature 338, 505-508. Ball, J. K., Diggelmann, H., Dekaban, G. A., et al. (1988). Alterations in the U3 region of the long terminal repeat of an infectious thymotropic type B retrovirus. J. Virol. 62, 2985-2993. Barklis, E., Mulligan, R. C., and Jaenisch, R. (1986). Chromosomal position or virus mutation permits retrovirus expression in embryonal carcinoma cells. Cell 47, 391399. Berger, S. A., Sanderson, N., Bernstein, A., and Hankins, W. D. (1985). Induction of the early stages of Friend erythroleukemia with helper-free Friend spleen focus-forming virus. Proc. Natl. Acad. Sci. USA 82, 6913-6917. Boone, L. R., Glover, P. L., Innes, C. L., Niver, L. A., Bondurant, M. C., and Yang, W. K. (1988). Fv-1 N- and B-tropism-specific sequences in murine leukemia virus and related endogenous proviral genomes. J. Virol. 62, 2644-2650. Boral, A. L., Okenquist, S. A., and Lenz, J. (1988). Identification of the S13-3 virus enhancer core as a T-lymphoma cell-specific element. J. Virol. 63, 76-84. Bosze, Z., Thiesen, H., and Charnay, P. (1986). A transcriptional enhancer with specificity for erythroid cells is located in the long terminal repeat of the Friend murine leukemia virus. E M B O J. 5, 1615-1624.
GENES CONTROLLING RETROVIRAL VIRULENCE
153
Bova, C. A., Manfredi, J. P., and Swanstrom, R. (1986). env genes of avian retroviruses: Nucleotide sequence and molecular recombinants define host range determinants. Virology 152, 343-354. Bova, C. A., Olsen, J. C., and Swanstrom, R. (1988). The avian retrovirus env gene family: Molecular analysis of host range and antigenic variants. J. Virol. 62, 75-83. Brown, D. W., Blais, B. P., and Robinson, H. L. (1988). Long terminal repeat (LTR) sequences, env, and a region near the 5' LTR influence the pathogenic potential of recombinants between Rous-associated virus types 0 and 1. J. Virol. 62, 3431-3437. Coffin, J. M. (1996). Retroviridae: The viruses and their replication. In "Fields Virology" (B. N. Fields, D. M. Knipe, Howley, P. M. et al., eds.), pp. 1767-1848. LippincottRaven Publishers, Philadelphia. Coffin, J. M. Structure and classification of retroviruses. In "The Retroviridae" (J. A. Levy, ed.), pp. 19-50. Plenum Press, New York. Deacon, N. J., Tsykin, A., Solomon, A., Smith, K., Ludford-Menting, M., Hooker, D. J., McPhee, D. A., Greenway, A. L., Ellett, A., Chatfield, C., Lawson, V. A., Crowe, S., Maerz, A., Sonza, S., Learmont, J., Sullivan, J. S., Cunningham, A., Dwyer, D., Dowton, D., and Mills, J. (1995). Genomic structure of an attenuated quasi species of HIV-1 from a blood transfusion donor and recipients. Science 270, 988-991. Ewart, G., Sutherland, T., Gage, P., and Cox, G. (1996). The Vpu protein of human immunodeficiency virus type 1 forms cation-selective ion channels. J. Virol. 70, 71087115. Furukawa, Y., Fujisawa, J., Osame, M., et al. (1992). Frequent clonal proliferation of HTLV-l-infected T cells in HTLV-l-associated myelopathy (HAM-TSP). Blood 80, 1012-1016. Gardner, M. B., Endres, M., and Barry, P. (1994). The simian retroviruses SIV and SRV. In "The Retroviridae," Vol. 3 (J. A. Levy, ed.), pp. 133-276. Plenum Press, New York. Gardner, M. B., Henderson, B. E., Officer, J. E., Rongey, R. W., Parker, J. C., Oliver, C., Ester, J. D., and Huebner, R. J. (1973). A spontaneous lower motor neuron disease apparently caused by indigenous type-C RNA virus in wild mice. J. Natl. Cancer Inst. 51, 1243. Gessain, A., Vernant, J. C., Maurs, L., et al. (1985). Antibodies to human T-lymphotropic virus type-1 in patients with tropical spastic paraparesis. Lancet 2, 407-410. Gibbs, J. S., Lackner, A. A., Lang, S. M., Simon, M. A., Sehgal, P. K., Daniel, M. D., and Desrosiers, R. C. (1995). Progression to disease in the absence of genes for vpr or vpx. J. Virol. 69, 2378-2383. Golovkina, T. V., Chervonsky, A., Dudley, J. P., and Ross, S. R. (1992). Transgenic mouse mammary tumor virus superantigen expression prevents viral infection. Cell 69, 637645. Golovkina, T. V., Prescott, J. A., and Ross, S. R. (1993). Mouse mammary tumor virusinduced tumorigenesis in sag transgenic mice: A Laboratory model of natural selection. J. Virol. 67, 7690-7694. Goncalves, J., Shi, B., Yang, X., and Gabuzda, D. (1995). Biological activity of human immunodeficiency virus type 1 v i f requires membrane targeting by C-terminal basic domains. J. Virol. 69, 7196-7204. Harbers, K., Soriano, P., Muller, U., and Jaenisch, R. (1984). Insertion of retrovirus into the first intron of alpha l(I) collagen gene leads to embryonic lethal mutation in mice. Proc. Natl. Acad. Sci. U S A 81, 1504. Hardy, W. D. (1993). Feline oncoretroviruses. In '~Fhe Retroviridae," Vol. 2 (J. A. Levy, ed.), pp. 109-180. Plenum Press, New York. He, J. Choe, S., Walker, R., Di Marzio, P., Morgan, D. O., and Landau, N. R. (1995).
154
FREDERICK J. FULLER
Human immunodeficiency virus type 1 viral protein R (vpr) arrests cells in the G2 Phase of the cell cycle by inhibiting p34 cdc2 activity. J. Virol. {}9, 6705-6711. Held, W., Waanders, G. A., Shakhov, A. N., Scarpellino, L., Acha-Orbea, H., and MacDonald, H. R. (1993). Superantigen-induced immune stimulation amplifies mouse mammory tumor virus infection and allows virus transmission. Cell 74, 529-540. Hoch, J., Lang, S. M., Weeger, M., Stahl-Hennig, C., Coulibaly, C., Dittmer, U., Hunsmann, G., Fuchs, D., Muller, J., Sopper, S., Fleckenstein, B., and Uberla, K. T. (1995). vpr deletion mutant of simian immunodeficiency virus induces AIDS in rhesus monkeys. J. Virol. 69, 4807-4813. Jackson, C. A. W., Dunn, S. E., Smith, D. I., Gilchrist, P. T., and MacQueen, P. A. (1977). Proventriculitis, "Nakanuke" and reticuloendotheliosis in chickens following vaccination with herpesvirus of turkeys (HVT). Aust. Vet. J. 53, 457. Jenkins, N. A., Copeland, N. G., Taylor, B. A., and Lee, B. K. (1981). Dilute (d) coat colour mutation of DBA/2J mice is associated with the site of integration of an ecotropic MuLV genome. Nature 293, 370. Kawamura, H., Wakabayashi, T., Yamaguchi, S., Taniguchi, T., Takayanagi, N., Sato, S., Sekiya, S., and Horiuchi, T. (1976). Inoculation experiment of Marek's disease vaccine contaminated with reticuloendotheliosis Virus. Natl. Inst. Anim. Health Q. 16, 135. Kestler, H. W., Ringler, D. J., Mori, K., Panicali, D. L., Sehgal, P. K., Daniel, M. D., and Desrosiers, R. C. (1991). Importance of the nef gene for maintenance of high virus loads and for development of AIDS. Cell 65, 651-662. Kirchhoff, F., Greenough, T. C., Brettler, D. B., Sullivan, J. L., and Desrosiers, R. C. (1995). Brief report: Absence of intact nef sequences in a long-term survivor with nonprogressive HIV-infection. N. Engl. J. Med. 332, 4,228-232. Klinken, S. P., Frederickson, T. N., Hartley, J. W., Yetter, R. A., and Morse, H. C., III. (1988). Evolution of B cell lineage lymphomas in mice with a retrovirus-induced immunodeficiency syndrome, MAIDS. J. Immunol. 140, 1123. Koyama, H., Sasaki, T., Ohwada, Y., and Saito, Y. (1980). The relationship between feathering abnormalities ("Nakanuke") and tumour production in chickens inoculated with reticuloendotheliosis virus. Avian Pathol. 9, 331. Kristal, B. S., Reinhart, T. A., Hoover, E. A., and Mullins, J. I. (1993). Interference with superinfection and with cell killing and determination of host range and growth kinetics mediated by feline leukemia Virus surface glycoproteins. J. Virol. 67, 41424153. Lackner, A. A. (1994). Pathology of simian immunodeficiency virus induced disease. In "Simian Immunodeficiency virus" (N. L. Letvin and R. C. Desrosiers, eds.), pp. 35-64. Springer-Verlag, Berlin. Marthas, M. L., Ramos, R. A., Lohman, B. L., Van Rompay, K. K., Unger, R. E., Miller, C. J., Banapour, B., Pedersen, N. C., and Luciw, P. A. (1993). Viral determinants of simian immunodeficiency Virus (SIV) virulence in rhesus macaques assessed by using attenuated and pathogenic molecular clones of SIVmac. J. Virol. 67, 6047-6055. McDougall, J. S., Biggs, P. M., and Shilleto, R. W. (1978). A leukosis in turkeys associated with infection with reticuloendotheliosis virus. Avian Pathol. 7, 557. Michael, N. L., Chang, G., D'Arcy, L. A., Tseng, C. J., Birx, D. L., and Sheppard, H. W. (1995). Functional characterization of human immunodeficiency virus type 1 nef genes in patients with divergent rates of disease progression. J. Virol. 69, 6758-6769. Michaels, F. H., Hattori, N., Gallo, R. C., and Franchini, G. (1993). The human immunodeficiency virus type 1 (HIV-1) vif protein is located in the cytoplasm of infected cells and its effect on viral replication is equivalent in HIV-2. AIDS Res. Hum. Retroviruses 9, 1025-1030.
GENES CONTROLLING RETROVIRAL VIRULENCE
155
Mosier, D. E., Yetter, R. A., and Morse, H. C., III. (1985). Retroviral induction of acute lymphoproliferative disease and profound immunosuppression in adult C57BL/6 mice. J. Exp. Med. 161, 766. Mussman, H. C., and Twiehaus, M. J. (1971). Pathogenesis of reticuloendothelial virus disease in chicks--An acute runting syndrome. Avian Dis. 15, 483. Narayan, O., and Cork, L. C. (1985). Lentiviral diseases of sheep and goats: Chronic pneumonia leukoencephalomyelitis and arthritis. Rev. Infect. Dis. 7, 89-98. Niimi, N., Wajjwalku, W., Ando, Y., Nakamura, N., Ueda, M., and Yoshikai, Y. (1995). A novel V~2-specific endogenous mouse mammary tumor virus which is capable of producing a milk-borne exogenous virus. J. Virol. 69, 7269-7273. Osame, M., Usuku, K., Izumo, S., et al. (1986). HTLV-1 associated meylopathy: A new clinical entity. Lancet 1, 1031-1032. Parslow, T. G. (1993). Post-transcriptional regulation of human retroviral gene expression. In "Human Retroviruses (B. R. Cullen, ed.), pp. 101-136. IRL Press, Oxford. Pedersen, N. C., Yamamoto, J. K., Ishida, T., and Hansen, H. (1989). Feline immunodeficiency virus infection. Vet. Immunol. Immunopathol. 21, 111-129. Peterlin, B. M., Adams, M., Alonso, A., Baur, A., Ghost, S., Lu, X., and Luo, L. (1993). Tat trans-activator. In "Human Retroviruses" (B. R. Cullen, ed.), pp. 75-100. IRL Press, Oxford. Purchase, H. G., and Witter, R. L. (1975). The reticuloendotheliosis viruses. Current Top. Microbiol. Immunol. 71, 103. Rhee, S. S., and Marsh, J. W. (1994). Human immunodeficiency virus type 1 nef-induced down-modulation of CD4 is due to rapid internalization and degradation of surface CD4. J. Virol. 68, 5156-5163. Robinson, H. L., and Miles, B. D. (1985). Avian leukosis virus induced osteopetrosis is associated with the persistent synthesis of viral DNA. Virology 141, 130-143. Ruscetti, S. K., Janesch, N. J., Chakraborti, A., Sawyer, S. T., and Hankins, W. D. (1990). Friend spleen focus-forming virus induces factor independence in an erythropoietindependent erythroleukemia cell line. J. Virol. 63, 1057. Stoye, J. P., Fennder, S., Greenoak, G. E., Moran, C., and Coffin, J. M. (1988). Role of endogenous proviruses as insertional mutagens: The hairless mutation of mice. Cell 54, 383. Vogel, J., Hinrichs, S. H., Reynolds, R. K., Luciw, P. A., and Jay, G. (1988). The HIV tat gene induces dermal lesions resembling Kaposi's sarcoma in transgenic mice. Nature 335, 606-611. Witter, R. L., Lee, L. F., Bacon, L. D., and Smith, E. J. (1979). Depression of vaccinal immunity to Marek's disease by infection with reticuloendotheliosis virus. Infect. Immun. 26, 90. Wong, P. K. Y., Soong, M. M., MacLeod, R., Gallick, G. E., and Yuen, P. H. (1983). A group of temperature-sensitive mutants of Moloney leukemia virus which is defective in cleavage of env precursor polypeptide in infected cells also induces hind-limb paralysis in newborn CFW/D mice. Virology 125, 513.
This Page Intentionally Left Blank
ADVANCES IN VETERINARY MEDICINE, VOL. 40
Mapping Animal Genomes JAMES E. WOMACK Department of Veterinary Pathobiology, College of Veterinary Medicine, Texas A&M University, College Station, Texas 77843 I. Introduction II. Why Map Farm Animal Genomes A. To Study Chromosomal Evolution B. To Permit Marker-Assisted Selection C. To Facilitate Positional Cloning III. Methods for Animal Genome Mapping A. Synteny Mapping B. In Situ Hybridization C. Linkage Mapping D. Comparative Mapping IV. Current Status of Animal Genome Maps A. Physical Mapping B. Linkage Mapping C. Comparative Mapping D. Economic Traits V. Future Directions References
I. I n t r o d u c t i o n A d v a n c e s in m o l e c u l a r genetic technology a n d a worldwide a s s a u l t on m a p p i n g a n d s e q u e n c i n g t h e h u m a n g e n o m e h a v e m a d e g e n o m e r e s e a r c h h i g h l y visible to t h e public a n d one of t h e exciting n e w frontiers in t h e life sciences. M a p p i n g genes to discrete r e l a t i v e positions on a n i m a l c h r o m o s o m e s . i s not a n e w concept a m o n g biologists, however. S t u r t e v a n t (1913) successfully followed alleles t h r o u g h meiotic r e c o m b i n a t i o n in drosophila to order genes on t h e X-chromosome, t h u s i n t r o d u c i n g our c u r r e n t g e n e r a t i o n of a n i m a l biologists to t h e u n d e r l y 157 Copyright 9 1997 by Academic Press. All rights of reproduction in any form reserved. 1093-975X/97 $25.00
158
JAMES E. WOMACK
ing principles of genomic research. Unfortunately, the 50 years immediately following Sturtevant's work produced only a scattering of genes mapped in a variety of animal species with the notable exception of the laboratory mouse. The availability of large numbers of inbred strains and a collection of mutant stocks produced a linkage map in Mus musculus, which became the prototypic mammalian genetic map. It has subsequently been enhanced by the availability of biochemical and molecular markers and the widespread use of interspecific hybrid backcrosses. The genomic revolution is most apparent, however, in the human genome project. It should not be surprising that the agricultural sciences, long concerned with germ plasm improvement, are beginning to invest in building maps and localizing important traits to chromosomal regions. Perhaps it is surprising that it has taken so long for animal geneticists, when compared to plant breeders and medical geneticists, to enthusiastically support the construction of genomic maps in economically important animals. Nonetheless, the genomic revolution has finally extended its borders to the study of animals used in agriculture and this article presents an overview of the progress and problems of genome analysis in those domestic animals. II. Why Map Farm Animal Genomes A. To STUDY
CHROMOSOMAL
EVOLUTION
The organization of extant genomes and their relationship to each other provides valuable insight into the organization of ancestral genomes and how chromosome structure may have evolved. Comparative gene mapping, which is nothing more than mapping homologous genes in multiple species, provides important information about chromosomal evolution between distant species that we cannot retrieve even with today's most advanced cytogenetic technologies. Consequently, a driving force in animal genome mapping is to be able to answer questions regarding chromosomal rearrangements that accompany evolution and thus to advance the fundamental biological knowledge of the species of interest and animals in general. Comparative mapping has become one of the major tools for the study of chromosomal evolution in animals. The well-funded and well-organized international human genome initiative is well on the way to providing the standard mammalian genomic map to which all others will undoubtedly be compared. The map of the laboratory mouse continues to advance at a rapid pace
MAPPING ANIMAL GENOMES
159
and will undoubtedly continue to be the first yardstick for evaluating chromosomal conservation and genomic evolution. Other mammals, including the livestock species, have extremely important roles to play, however, in helping us understand the pathways of chromosomal rearrangement that have accompanied m a m m a l i a n evolution. It is important to note in this context that all mammals have essentially the same amount of deoxyribonucleic acid (DNA) in their genomes and a similar complement of structural genes. Chromosome numbers, however, differ markedly even within some m a m m a l i a n orders, and banding-pattern divergence suggests that this homologous m a m m a l i a n genome has undergone various combinations of reciprocal chromosomal exchanges and internal rearrangements as taxa have diverged. It has already become clear through comparative mapping that some maremals, such as cattle and cats, have genomes that are more highly conserved relative to the h u m a n genome than is the most often compared mouse genome. These more highly conserved genomes may most nearly reflect the chromosomal arrangement of the ancestral mammal. At the minimum, they demonstrate that m a m m a l i a n chromosome evolution cannot be fully represented by differences in the genomes of humans and mice. Continued comparative mapping including the genomes of domestic animals will continue to make a valuable contribution to understanding m a m m a l i a n chromosomal evolution in a universal context.
B. To PERMIT MARKER-ASSISTED SELECTION
The practical goal of economic enhancement through marker-assisted selection (MAS) of desirable and marketable traits drives gene mapping in many laboratories and provides the motivation for several national programs that fund animal genomic research. Genetic markers of advantageous alleles for economic trait loci (ETL), including quantitative trait loci (QTL), have the potential to enhance the rate and efficiency of genetic gain through selective breeding, a concept advanced well before the technical tools become available for its implementation (Soller and Beckmann, 1982; Weller et al., 1990; Smith and Simpson, 1986). The ideal marker for use in a selective breeding program would obviously be the gene actually responsible for the trait. While such markers are still rare, a few have been identified by relentless searches for variation in genes thought to be involved in the physiological pathways leading to the phenotype of interest. This so-called
160
JAMES E. WOMACK
candidate gene approach to marker identification requires a sound fundamental knowledge of the physiological processes underlying the trait and then extensive and sometimes expensive screening for variation in candidate genes related to these processes. Ideally, once sequence variation related to the trait has been found, it can be incorporated directly into the marker assay. As an excellent example of a successful candidate gene search, Shuster et al. (1992) identified the genetic defect responsible for leukocyte adhesion deficiency (LAD) in Holstein cattle as a missense mutation coding amino acid 128 in CD18. The mutant and normal alleles were then distinguished by the polymerase chain reaction (PCR) providing the ultimate genetic marker of this ETL. The physiological bases for most economic traits are unknown and consequently candidate genes are not obvious. Alternatively, the complexity of the trait often presents a long and unduly cumbersome list of candidate genes. ETL, even QTL, can be genetically mapped, however (Lander and Botstein, 1989; Paterson et al., 1989; Georges et al., 1993a; Georges et al., 1993b; Andersson et al., 1994; Georges et al., 1995). Markers mapped in close proximity to ETL might then be used to assist in selection for the ETL if the recombination frequency (map distance) is sufficiently small and the chromosomal phase of marker and ETL alleles is known. Efficiency of MAS can be increased by identifying markers flanking the ETL because recombination in the region spanned by two markers can be detected. A major goal of animal gene mapping has therefore been to produce maps of highly polymorphic markers spaced at intervals of 20 centimorgans (cM) (lcM = 1% recombination) or less across every chromosome of the targeted species. These markers can then be applied to families segregating QTL, hopefully resulting in the linkage associations of the QTL with one or more markers. Under appropriate breeding protocols, these markers may then be used for MAS. Linked markers can be used in selection only if the phase of the marker allele and the ETL allele is known. P h a s e refers to the association of trait alleles and marker alleles on a common parental chromosome. In some gametes, for example, allele a of the marker locus might be physically associated with the positive (+) trait allele, whereas in another it might be associated with the negative ( - ) allele. This does not present a problem in defined pedigrees where the phase is initially known and map distances are small. Linked markers cannot be used in the absence of pedigree information, however, unless the marker and trait are in linkage disequilibrium.
MAPPING ANIMAL GENOMES
161
C. To FACILITATE POSITIONAL CLONING
A third, and perhaps most important, goal of gene mapping is to identify and clone genes responsible for ETL. It is obvious from the previous discussion that MAS can be done much more efficiently with variation in the gene actually responsible for the ETL. Moreover, a complete understanding of the potential interaction of the trait with other physiological processes is possible only when we know the genes involved. The term reverse genetics has been replaced in common usage by positional cloning or map-based cloning to describe the process by which the application of map information is used to clone a gene responsible for a specific trait in the absence of information about the biochemical or molecular basis of the trait. Success of positional cloning has been experienced a number of times in searches for human disease genes, perhaps best exemplified by the cloning of the gene for cystic fibrosis (Rommens et al., 1989). While the task of positionally cloning genes in any species is a formidable one, cloning genes for ETL in livestock is almost prohibitive. Animal maps will most certainly never be as dense as those of the human. Large insert libraries for livestock species are only beginning to be developed and used. Chromosomal deletions of economically important genes, important tools in many of the human and mouse successes, have not been identified and propagated in livestock. The task is further complicated by the quantitative nature of most of the traits of economic interest in farm animals and the paucity of research support worldwide for animal agriculture relative to the human genome initiative. Alternative strategies to conventional positional cloning must be planned and developed. One proposed approach is comparative candidate positional cloning, which takes advantage of knowledge of the evolutionary history of chromosomes and rapid advances in the human and mouse maps. This concept will be further developed in Section V. III. Methods for Animal Genome Mapping A. SYNTENY M_~kPPING
Synteny simply means on the same strand or, in genetic terminology, on the same chromosome. A synteny map is nothing more than a list of genes known to reside on the same chromosome in a particular species. An unfortunate misuse of the word synteny abounds in recent genetic literature. Such statements as "a greater degree of synteny between
162
JAMES E. WOMACK
cattle and human than between cattle and mouse," or "a limited degree of synteny has been demonstrated between human, mouse and swine" appear repeatedly in genetic literature and are meaningless in genetics. "Conserved synteny" was used by Nadeau (1989) to describe the location of two or more homologous genes on the same chromosome in different species. "Synteny" should not be substituted for "conserved synteny" in our comparison of maps between species. Synteny mapping is probably associated with comparative mapping because the only maps available for comparison between most animal species have heretofore been synteny maps. The current revolution in human gene mapping began with the building of synteny maps, the success of which can be traced to the hybridization of cells from different species (Weiss and Ephrussi, 1966) and to the subsequent elimination of human chromosomes in the proliferation of hybrid cells (Ruddle, 1972). The assignment of genes to syntenic groups was accomplished by observing the concordant loss or retention of gene products. If two genes were lost or retained concordantly in the random segregation of human chromosomes, they were assumed to be located on a common chromosome and said to be syntenic. In the early days of somatic cell genetics, electrophoretic differences in homologous gene products of the progenitor species allowed the identification of human specific gene products and subsequently the mapping of the human genes encoding them. Cytogenetic analysis of hybrid clones then permitted the correlation of a particular gene product with a specific human chromosome or chromosome fragment and thus the assignment of that gene to a chromosome or chromosomal region. These methods with numerous refinements have been thoroughly reviewed (Ruddle and Creagan, 1975; Shows, 1977; Shows et al., 1982). Somatic cell genetics is a powerful mapping tool because it circumvents both the requirements of large numbers of offspring from sexual matings and genetic variation within a species. Because of the evolutionary divergence of such species as humans and mice, the probability of finding differences in their homologous genes or gene products is much greater than the probability of finding variation within a species. Whereas cell hybridization facilitated by Sendai virus or polyethylene glycol has been the principal method for transferring donor chromosomes into recipient cells, other methods were developed for specific experimental strategies (Ruddle, 1981). The isolation of donor chromosomes into microcells (Fournier and Ruddle, 1977) provides experimental control over chromosome segregation and allows the transfer of one or a few intact donor chromosomes. Recipient cells can also
MAPPING ANIMAL GENOMES
163
be transformed with cell-free preparations of donor chromosomes, which are degraded into chromosomal fragments after endocytosis (Klobutcher land Ruddle, 1981). Eukaryotic cells can also be transformed by purified DNA (Wigler et al., 1977), either by endocytotic uptake of calcium precipitated DNA or by injection with micropipets. The size of the donor DNA molecule can be specifically controlled in these experiments and is often 50 kilobases (kb) or smaller. Each of these methods provides a means of transferring progressively smaller segments of donor chromosome into recipient cells and consequently progressively extending the limit of refinement of gene mapping by parasexual methods. Though still available as a useful method of gene identification in somatic cell gene mapping, the electrophoretic separation of gene products is no longer necessary for parasexual genetics and is in fact seldom used. The limitation of mapping only those genes whose products are expressed in cultured cells has been overcome by the use of molecular probes and PCR primers that distinguish host and recipient DNA. One useful mapping tool provided by recombinant DNA technology is the use of Southern blotting for restriction fragment mapping. Successfully applied to the mapping of mouse and h u m a n genes (Ruddle, 1981; Shows et al., 1982) as well as to many cattle genes as will be described in detail later, this method depends on differences in the location of restriction enzyme sites in the host and donor genomes. Heterologous probes have proved extremely useful in mapping livestock species in which only a small number of genes have been cloned and sequenced. Use of these probes provides a convenient entry into comparative gene mapping as well. Both enzyme electrophoresis and Southern blotting with probe hybridization have given way to the PCR as the analytical tool of choice in synteny mapping. Primers from conserved regions of genes often amplify across species, providing the same advantages of heterologous probes. As with other methods, conditions and primers must be such that the presence or absence of the DNA of one species can be determined against the background genome of another. Somatic cell genetics is still the most common method for building synteny maps. Hybrid somatic cells can be constructed such that the chromosomes of practically any progenitor species are preferentially lost. Each hybrid clone will retain a partial genome of that species along with the complete genome of the other, which is usually a transformed rodent cell line. Because chromosome loss is more or less random, each clone will retain a different subset of chromosomes from the species being mapped. As in h u m a n gene mapping, analysis of pairs of
164
JAMES E. WOMACK
genes in a panel of hybrid cell lines will reveal concordance or discordance of their retention. Concordance of retention is evidence for the location of two genes on the same chromosome. Conversely, discordance of retention is evidence for asynteny, their location on different chromosomes. Gene products or DNA sequences may be mapped by synteny analysis in any species so long as the presence or absence of the gene or gene product of the targeted species can be ascertained against the fully retained rodent genomic background. Enzyme electrophoresis, Southern blotting with unique sequence probes, and PCR amplification with species discriminating primers have all been effective analytical tools for synteny mapping. Somatic cell genetics does not typically assign markers to specific chromosomal sites or even to chromosomal regions. Consequently, genes on a synteny map are usually unordered. Somatic cell methods using rearranged chromosomes are an exception to this generalization and have been used very effectively to order genes in the human map. Radiation hybrid mapping (Cox et al., 1990; Walter et al., 1994) has become an important tool for constructing high resolution maps of h u m a n chromosomes. The techniques used are variations of basic somatic cell genetics in which the donor cells have been irradiated to achieve chromosome fragmentation. Statistical analysis is based on the principles of linkage analysis, i.e. the closer two loci are to each other the less likely they are to be separated by random chromosomal rearrangement. First used by Goss and Harris (1975), the technique can be used with single chromosome hybrids as the irradiated donor (Cox et al., 1990) or with total genome irradiation in a diploid donor cell (Walter et al., 1994). Whether used in mapping a single chromosome or a whole genome, the technology is effective for constructing contiguous maps of mammalian chromosomes at a 500-kb level of resolution. This method may prove to be the ideal approach to comparative gene mapping because it provides an ordered map without the requirement of segregating polymorphisms in breeding populations. B. In S i t u HYBRIDIZATION
Unique sequences, repetitive elements, and whole genomes have all been successfully localized to chromosomal sites by in situ hybridization. This technique uses the attachment of a microscopically detectable marker to a DNA probe followed by hybridization of the probe to denatured DNA of an otherwise intact chromosome. The specificity of hybridization is determined by the uniqueness of the probe. While radioactive probes dominated the early application of this technology,
MAPPING ANIMAL GENOMES
165
fluorescent probes are now generally used. In her review of fluorescence in situ hybridization (FISH), Trask (1991) notes the following advantages of FISH over isotopic labeling. It provides superior spatial resolution and generally requires the visualization of fewer labeled chromosomes. It is faster and the probe is more stable. The sensitivities are similar, each requiring a few kb pairs of uninterrupted sequence as the hybridizing probe. Schemes have been developed that allow multiple probes with different color signals to be used on the same chromosomes. The latter is particularly important for gene mapping because it provides the potential for ordering loci within the limits of approximately 100-kb resolution. Cosmids, cloning vectors that accommodate large DNA sequences (up to 45 kb), have been effectively used as probes for FISH. Because these large inserts often contain repetitive DNA, the target DNA must first be prehybridized with unlabeled total genomic DNA. As described later, this method has been effectively used to anchor linkage maps to chromosomes by hybridizing cosmids that contain highly polymorphic markers used in linkage mapping. The technologies just discussed result in physical maps that describe the physical relationships of loci to the chromosomes on which they reside. Higher resolution physical maps that define markers in contiguous clones (contig maps) are forthcoming in livestock but will most likely span small genomic regions of special interest rather than whole chromosomes as is targeted in the h u m a n genome initiative and is prerequisite to total genome sequencing. C. LINKAGE MAPPING
Linkage maps are defined in biological rather than physical terms, a map unit representing 1% recombination in meiosis. Because linkage is measurable only in gametic products, linkage mapping requires detection of maternal and paternal alleles in gametes of heterozygous individuals. A linkage map is then made from the percentage of recombinants in the parental arrangements of alleles of two loci on a chromosome. The segregation of three or more loci permits ordering of genes on the map because double recombinants are rare relative to single recombinants. Markers for mapping purposes have been catagorized as Type I or II by O'Brien (1992). Type I markers are expressed sequences (genes) and are usually conserved from one m a m m a l i a n species to another. Thus, they are the material for comparative gene mapping. Unfortunately, they are usually not highly polymorphic and therefore are difficult to
166
JAMES E. WOMACK
incorporate into linkage maps. Type II markers are highly polymorphic and are more widely used for linkage mapping. The value of these markers lies in their high level of polymorphism. They are not necessarily expressed genes but often anonymous stretches of DNA. The polymorphic DNA markers that dominate the linkage maps of most plant and animal species are one of two general classes. Restriction fragment length polymorphisms (RFLPs) identify alleles on the basis of polymorphism in the presence or absence of specific restriction enzyme cleavage sites. Genomic DNA is digested with a particular restriction enzyme, separated by electrophoresis, and blotted onto a solid membrane (Southern, 1975). Hybridization to a labeled probe often reveals polymorphism in the size of the DNA fragments produced. One major advantage of RFLPs as linkage markers is that they can identify polymorphisms in or around Type I markers, assuming the probe used for detection is a cloned gene or complementary DNA (cDNA). They are, therefore, useful for comparative mapping and also provide a potential source of variation in candidate genes for economic traits. Blotting and probe development are laborious and time consuming, however, and blotting exhausts DNA resources much faster than do PCRbased methods. Microsatellites (Weber and May, 1989; Fries et al., 1990) have become the Type II marker of choice in linkage mapping in most animal species. These islands of di-, tri-, or tetranucleotide repeats are proving to be ubiquitous throughout animal genomes. The number of repeats is usually the polymorphic entity, with a half dozen or more alleles not uncommon in a breeding population. Microsatellites are defined by unique flanking sequences that serve as primers for PCR amplification and also as tags of unique sites in the respective genome. Other types of markers on linkage maps include enzyme polymorphisms, blood group antigens, and phenotypic traits such as coat color or horn development. These all add interest and biological significance to the map. Moreover, they provide potential for enhanced comparative mapping across species. Linkage mapping requires observable meiotic products, usually in the form of offspring from individuals segregating the markers just discussed. Large full sib-ships or half sib-ships are ideal for linkage analysis in farm animals. Because a comprehensive map requires large numbers of markers scored on a common set of meiotic products, sharing of family material among laboratories is the most effective arrangement for generating a linkage map. Sets of shared reference families are available for most species, including cattle (Barendse et al., 1994). Offspring are not essential for linkage analysis, however, be-
MAPPING ANIMAL GENOMES
167
cause multiple DNA loci can now be analyzed for allelic variation directly on meiotic products. Van Eijk et al. (1993) successfully ordered three loci on bovine chromosome 23 by PCR after primer extension preamplification (PEP) of individual sperm. D. COMPARATIVE MAPPING
Comparative gene mapping depends entirely on the accurate assessment of homology between genes of different species. It is not always easy to make this assessment. The issue of gene homology has been addressed repeatedly in reports of the Committee on Comparative Mapping, heretofore an integral component of the H u m a n Gene Mapping (HGM) workshop. Although the technologies for detecting genes have evolved at a rapid rate, the principles as outlined in HGM 10 can still be regarded as standards for the assessment of interspecific gene homology (Lalley et al., 1989). These principles are as follows: 1. Molecular structure a. Similar nucleotide or amino acid structure b. Similar immunological cross-reaction c. Similar subunit structure d. Formation of functional heteropolymeric molecules in interspecific cell hybrids in cases of multimeric proteins e. Cross-hybridization to the same molecular probe 2. Biological or biochemical function a. Similar tissue distribution b. Similar developmental time of appearance c. Similar pleiotropic effects d. Identical subcellular locations e. Similar substrate specificity f. Similar response to specific inhibitors The increased use of heterologous probes has had a major impact on comparative gene mapping. Although there is probably not a better single criterion for gene homology t h a n cross-hybridization with a molecular probe, the Committee on Comparative Mapping has recognized an inherent danger in using probes characterized in other laboratories for comparative mapping purposes. Obviously, a mislabeled probe used for mapping can lead to an erroneous comparative chromosomal assignment. The committee strongly recommends that investigators receiving probes from other laboratories perform the tests necessary to verify that the probe being used is the same as the one used in previously published mapping experiments. These criteria include the following:
168
JAMES E. WOMACK
1. Verification of the restriction sites in the plasmid or phage clone 2. Verification of several restriction fragments after Southern analysis of h u m a n or genomic DNA 3. Partial sequence analysis of clones 4. PCR homology with an included oligonucleotide 5. Other traditionally accepted methods of probe verification If this recommendation is not generally followed, some misassignments are certain to occur and to result in significant problems. The same strict guidelines must be applied to genes analyzed by the PCR. Primer sequences may be selected from conserved regions of a gene in one species and used for amplification in another. Sufficient exon sequence should be included in cross-species primer design to allow for sequencing and homology assessment. The previous discussion assumes the use of biochemical or molecular gene markers for comparative mapping. However, just as albino and pink-eye dilute were identified as homologous mutations in different species (Haldane et al., 1915; Castle and Wachter, 1924; Feldman, 1924; Clark, 1936), mutations with similar phenotypic expression are still useful comparative markers. Unfortunately, there does not appear to be an abundance of obvious homologous mutations in different species. The mapping of two similar mutations to a known conserved region in two different species, however, is good preliminary evidence for homology of the mutations. IV. Current Status of Animal Genome Maps A. PHYSICAL ~IIAPPING More than 400 Type I loci have been mapped in cattle (Fries et al., 1993; O'Brien et al., 1993) primarily through somatic cell genetics. Of the 100 or so chromosomal assignments in the sheep map (Echard et al., 1994), most were made by synteny mapping. More than 80 of these are Type I markers. While these synteny-mapped markers indicate the boundaries of chromosomal conservation relative to the map-rich genomes of mice and humans and show extensive sheep-cattle conservation, they provide an incomplete comparative map. Conservation of gene order is not addressed by the synteny map. In situ hybridization, especially FISH, has been used effectively to address the order of Type I loci, to assign syntenic groups to specific chromosomes, and to anchor the rapidly growing linkage map to chromosomes (Fries et al., 1993; Solinas-Toldo et al., 1993; Iannuzzi et al.,
MAPPING ANIMAL GENOMES
169
1993; Gallagher et al., 1993b). There are currently more than 50 in situ localizations of unique sequences on cattle chromosomes. All bovine syntenic groups are now assigned to specific chromosomes and the bovine linkage map is anchored at 35 sites on 26 chromosomes. Thirtyfour in situ assignments in sheep were reviewed by Echard et al. (1994). At least 80 chromosomal assignments have been made in pigs, most by in situ hybridization (Andersson et al., 1993). Sixty of these are Type I markers. A promising new technique uses in situ PCR and permits the visualization of amplified sequences directly on porcine chromosomes (Troyer et al. 1994; Xie et al., 1995). B. LINKAGE MAPPING
The published bovine linkage map of 200 markers (Barendse et al., 1994) has grown to almost 800 markers since its initial publication (Barendse, personal communication), which was made possible in large measure by international cooperation and the use of a common set of reference families. Combined with the independent development of other maps (Bishop et al., 1994), there are probably more than 1200 markers currently assigned to cattle linkage groups. The goal of 20 cM resolution has clearly been achieved over at least 95% of the total genome. The international map includes 65 type I markers. A pig linkage map of 128 markers, 60 of which are informative for comparative mapping, was published by Ellegren et al. (1994). Rohrer et al. (1994) published a 383 marker map of the pig, almost all of which were microsatellites. Two reference families of chickens have each produced linkage maps in excess of 200 markers (Burt et al., 1995). C. COMPARATIVE MAPPING
We began comparative mapping in cattle in our laboratory by constructing a panel of hybrid somatic cells and assessing gene segregation by enzyme electrophoresis (Womack and Moll, 1986). Initially, these c a t t l e - h a m s t e r hybrid somatic cells were analyzed by celluloseacetate electrophoresis for 28 enzyme gene products, including such markers as GAPD, ITPA, ADA, ACO1, GDH, GUK, CAT, and GLO1. The 28 loci were organized into 21 independent syntenic groups bringing the composite bovine gene map at that time to 35 loci on 24 syntenic groups. A total of 32 homologous genes were then mapped in humans, mice, and cattle. Conservation of cattle and human linkage groups was evidenced by only three syntenic discordancies among 32 loci as contrasted to nine discordancies among the same loci in the
170
JAMES E. WOMACK
h u m a n and mouse maps. This was the initial suggestion that cattle and h u m a n chromosomal conservation exceeded that of either to the laboratory mouse. Another different but related somatic cell approach facilitated the mapping of the PRGS and PAIS genes in cattle (McAvin et al., 1988). Individual c o w - h a m s t e r hybrid cell lines established by fusion of mutant CHO cells, ade-C and ade-G, with cattle leukocytes required complementing bovine genes for PRGS and PAIS, respectively, when propagated on selective media. Homogenates of 12 PRGS + hybrid clones and 12 PAIS + hybrid clones retained the bovine electromorph of SOD 1 while extensively segregating 14 biochemical markers of other cattle syntenic groups. Secondary c a t t l e - h a m s t e r hybrid subclones, which segregated bovine PRGS and PAIS in late passages on nonselective media, concordantly segregated bovine SOD1. These data supported a syntenic relationship among PRGS, PAIS, and SOD 1 on cattle syntenic group U10. An interferon receptor (IFREC) locus was previously known to be syntenic with SOD 1. This synteny demonstrated an extensive conservation of bovine U10 and the Down syndrome region of h u m a n chromosome 21. Moreover, it expanded the boundaries of somatic cell genetics in domestic animals to use of auxotrophic recipient cell lines to preferentially retain a specific donor chromosome. Verification of these results and expansion of the map of the SOD gene family were provided by Gallagher et al. (1992a). cDNA probes of h u m a n extracellular superoxide dismutase (EC-SOD) and bovine superoxide dismutase 1 (SOD1) genes were hybridized to Southern blots containing genomic DNAs from the same cow-rodent somatic cell lines segregating bovine chromosomes. The SOD1 probe identified two loci: the coding locus (SOD1), which mapped to bovine U10; and a related locus (SOD1L), which mapped to U l l . EC-SOD mapped to bovine U15. The mapping of EC-SOD to U15 further defined a region of extensive syntenic conservation between humans and domestic cows. Mapping families of genes is an efficient and biologically interesting approach to the comparative mapping of animal genomes. DNA from the bovine-hamster hybrid cell panel was analyzed by blot hybridization with alpha and beta interferon probes (Adkison et al., 1988a). Retention and loss of the bovine interferon genes was compared to segregation of bovine isozyme loci representing the previously described syntenic groups. Families of bovine alpha (IFNA) and beta (IFNB) interferon genes were segregated in concordance with each other and with aconitase-1 (ACO1) on bovine syntenic group U18. This syntenic relationship is conserved on h u m a n chromosome 9p and on the portion of mouse chromosome 4 proximal to the centromere. In
MAPPING ANIMAL GENOMES
171
addition, cattle RFLPs were identified with both IFNA and IFNB probes. These polymorphisms provided valuable tools for subsequent searches for relationships of host resistance to disease and variation in candidate genes. These studies were expanded to another chromosomal region by Adkison et al. (1988b). DNA from c o w - h a m s t e r and cow-mouse somatic hybrid cells segregating bovine chromosomes were analyzed by Southern blotting and hybridization with h u m a n fibronectin and g a m m a crystallin probes. Concordance of retention of these bovine genes was compared to cattle isozyme loci representing all known syntenic groups. Bovine fibronectin (FN1) and g a m m a crystallin (CRYG) fragments were concordant with each other and with isocitrate dehydrogenase 1 (IDH1), representing bovine syntenic group U17. The syntenic relationship of these genes is conserved on h u m a n chromosome 2q and also on mouse chromosome 1. In addition, bovine RFLPs were identified with both fibronectin and g a m m a crystallin probes. These polymorphisms proved valuable for the study of recombination between the syntenic loci in pedigreed herds and to mark a segment of the bovine genome that is homologous to the L s h region of mouse chromosome 1, which confers resistance in mice to several intracellular parasites. Hallerman et al. (1988) mapped the genes encoding bovine prolactin and rhodopsin to syntenic groups on the basis of hybridization of DNA from hybrid cells with cloned prolactin and rhodopsin gene probes. Prolactin was found to be syntenic with previously mapped glyoxalase, BoLA, and 21-hydroxylase genes, establishing a syntenic conservation with h u m a n chromosome 6. The presence of bovine rhodopsin sequences among the various hybrid cell lines was not concordant with any gene previously assigned to one of the 23 defined autosomal syntenic groups. Thus, rhodopsin marked a new bovine syntenic group, leaving only five cattle autosomes unmarked at that time by at least one biochemical or molecular marker. Parathyroid hormone and the beta hemoglobin gene cluster, which are closely linked on h u m a n chromosome 11p15, were localized to bovine syntenic group (U7) with the gene for catalase by Foreman and Womack (1989). RFLPs were followed through informative pedigrees to determine a linkage map distance of approximately 15 cM between the parathyroid hormone and hemoglobin genes. This work began the integration of the rapidly growing synteny map to the fledgling linkage map of the bovine genome. The immunoglobulin genes are an important gene family both from the evolutionary and host-resistance perspective. Their mapping was
172
JAMES E. WOMACK
the target of a study by Tobin-Janzen and Womack (1992). Comparative gene mapping in mammals suggested that the bovine immunoglobulin heavy chain genes, IGHG4 and IGHM might be syntenic with the FOS oncogene. Interestingly, however, when these genes were assigned to bovine syntenic groups in our panel of hybrid somatic cells, IGH genes were shown to be syntenic with the FES oncogene rather than FOS. IGH and FES were assigned to Bos t a u r u s chromosome 21 while FOS was assigned to chromosome 10. In addition, bovine-specific immunoglobulin-like sequences were observed in the hybrid somatic cells, and one, IGHML1, was mapped to bovine syntenic group U16. The probes used for somatic cell mapping were also used to screen a small number of cattle of several different breeds for RFLPs. IGHG4 and IGHM were shown to be highly polymorphic, while FOS and FES were not. Two bovine DNA probes (LC a and LC b) complementary to the clathrin light chain genes were hybridized to DNA from the bovinehamster hybrid somatic cell lines by Gallagher et al. (1991). Concordancy of retention of the clathrin genes was compared to existing syntenic data for the domestic cow, which, by this time, consisted of a substantial data base. LC b identified a single locus, CLTB, concordant with the genes encoding bovine anti-Mfillerian hormone (AMH) and bovine osteonectin from bovine syntenic group U22. LC a recognized two loci: CLTAL1, from a previously unidentified bovine syntenic group; and CLTAL2, which is concordant with GGTB2, a gene marker for bovine syntenic group U18. A significant advance in the bovine genome map was precipitated by a combination of somatic cell genetics and in situ hybridization (Fries et al., 1991). The chromosomal locations of bovine class I and class II cytokeratin sequences were determined in the Fries laboratory using in situ hybridization and in our laboratory using Southern blot hybridization to DNA from the hybrid somatic cells to establish syntenic relationships. The main signals were found over chromosome region 19q16 -~ qter after in situ hybridization with two probes for the class I cytokeratin gene subfamily (KRT10 and KRT19) and over region 5q14 q23 after hybridization with probes for the class II gene subfamily (KRT1, KRT5, and KRT8). These regions are thought to contain the loci of functional cytokeratin genes, with KRT10 and KRT19 mapping to 19q21 and KRT1, KRT5, and KRT8 to 5q21. The in situ hybridization data were then corroborated by analysis of the somatic hybrid cell panel. The genes for the class I keratins segregated concordantly with each other and with syntenic group U21 in the somatic cell panel but were discordant with the class II keratin genes. The class II keratin
MAPPING ANIMAL GENOMES
173
genes segregated concordantly with each other and with syntenic group U3. Two class II gene probes gave an additional minor signal above chromosome region 5q25 ~ q33 after in situ hybridization, while another class II probe yielded a minor signal above chromosome region 10q31 ~ qter. When the latter probe and an additional linked probe were hybridized to DNA from the hybrid panel, two independently segregating loci were recognized, one of which cosegregated with the class II subfamily in syntenic group U3 and the other with syntenic group U5. These data confirmed the chromosomal assignment of two syntenic groups and allowed the assignment of a formerly unassigned syntenic group to a specific chromosome, a significant step at that time in the growth of the gene map. Bovine genes encoding T-cell receptor, CD3, and CD8 molecules were mapped to syntenic groups in hybrid somatic cells by Li et al. (1992). T-cell receptor ~ and 8 chains were assigned to bovine syntenic group U5, and the 13 and ~/genes were syntenic with each other and with markers on U13. CD3E and CD3D genes were syntenic with each other and located to bovine syntenic group U19. CD8 was most concordant with markers of syntenic group U16. These data expanded the comparative gene maps of h u m a n chromosome 7, bovine syntenic group U13, and mouse chromosomes 6 and 13 and suggested extensive evolutionary conservation among the m a m m a l i a n taxa. To establish syntenic relationships of phototransduction genes, Gallagher et al. (1992b) mapped the genes encoding the ~-, 13-, and ~/-subunits of rod cGMP phosphodiesterase (PDE) (PDEA, PDEB, PDEG), the ~'-subunit of cone PDE (PDEA2), and the rod cGMP-gated channel (CNCG) to bovine syntenic groups. The rod cGMP PDE ~-, ~-, and ~/-subunit genes mapped to bovine syntenic groups U22, U15 (chromosome 6), and U21 (chromosome 19), respectively. The rod cGMPgated channel gene also mapped to syntenic group U15, and the bovine cone a'-subunit gene mapped to U26 (chromosome 26). With the exception of the cone PDE ~'-subunit gene, which had not been mapped in other mammals, all of these genes were assigned to conserved chromosomal regions shared among bovine, human, and mouse. A compilation of known syntenic assignments and hypotheses regarding future assignments of phototransduction genes in human, mouse, and cattle was used to demonstrate the value of the bovine map for comparative genomic predictions of gene location in humans and mice. A phage library of bovine genomic DNA was screened for hybridization with a h u m a n HSP70 cDNA probe by Grosz et al. (1992) and 21 positive plaques were identified and isolated. Restriction mapping and blot hybridization analysis of DNA from the recombinant plaques dem-
174
JAMES E. WOMACK
onstrated that the cloned DNAs were derived from three different regions of the bovine genome. One region contained two tandemly arrayed HSP70 sequences, designated HSP70-1 and HSP70-2, separated by approximately 8 kb of DNA. Single HSP70 sequences, designated HSP70-3 and HSP70-4, were found in two other genomic regions. Locus-specific probes of unique flanking sequences from representative HSP70 clones were hybridized to restriction endonuclease-digested DNA from the somatic cell hybrid panel to determine the chromosomal location of the HSP70 sequences. The probe for the tandemly arrayed HSP70-1 and HSP70-2 sequences mapped to bovine chromosome 23, syntenic with glyoxalase 1,21 steroid hydroxylase, and major histocompatibility class I loci. HSP70-3 sequences mapped to bovine chromosome 10, syntenic with nucleoside phosphorylase and murine osteosarcoma viral oncogene (v-fos), and HSP70-4 mapped to bovine syntenic group U6, syntenic with amylase 1 and phosphoglucomutase 1. Based on these data, bovine HSP70-1,2 was proposed to be homologous to human HSPA1 and HSPA1L on chromosome 6p21.3, bovine HSP70-3 was proposed as the homolog of an unnamed human HSP70 gene on chromosome 14q22-q24, and bovine HSP70-4 was proposed homologous to one of the human HSPA-6, -7 genes on chromosome 1. Thus, comparative mapping was used as a tentative assessment of gene homology in a complex gene family. Ryan et al. (1992) confirmed the previous assignment of bovine a(IFNA) and ~- (IFNB) interferon gene families to syntenic group U18 and assigned this syntenic group to chromosome 8. FISH localized these genes to bovine chromosome 8, band 15, and demonstrated that with biotinylated plasmids, as few as five tandemly arrayed sequences can be detected by conventional fluorescent microscopy. Amplification of an ancestral lysozyme gene in artiodactyls appears to be associated with the evolution of foregut fermentation in the ruminant lineage and has resulted in about 10 lysozyme genes in true ruminants. Hybridization of a cow stomach lysozyme 2 cDNA clone to restricted DNA from our panel of cow-hamster hybrid cell lines revealed that all but one of the multiple bovine-specific bands segregated concordantly with the marker for bovine syntenic group U3 (Gallagher et al., 1993a). The anomalous band was subsequently mapped to bovine syntenic group U22 (chromosome 7) with a second panel of hybrids representing all 31 bovine syntenic groups. By two-dimensional pulsed-field gel electrophoresis, the lysozyme genes on cattle U3 (chromosome 5) were shown to be clustered on a 2- to 3-Mb DNA fragment, while the lactalbumin gene and pseudogenes that are paralogous and syntenic with the lysozymes were outside the lysozyme gene cluster.
MAPPING ANIMAL GENOMES
175
FISH of a cocktail of lysozyme genomic clones localized the lysozyme gene cluster to cattle chromosome 5, band 23, corroborating the somatic cell assignment. The asymmetric unit membrane (AUM) of the apical surface of mammalian urinary bladder epithelium contains several major integral membrane proteins, including uroplakins IA and IB (both 27 kDa), II (15 kDa), and III (47 kDa). These proteins are synthesized only in terminally differentiated bladder epithelial cells. They are encoded by separate genes and, except for uroplakins IA and IB, appear to be unrelated in their amino acid sequences. The genes encoding these uroplakins were mapped by Ryan et al. (1993) to chromosomes of cattle through their segregation in this same panel of bovine-rodent somatic cell hybrids. Genes for uroplakins IA, IB, and II were mapped to bovine chromosomes 18 (UPK1A), 1 (UPK1B), and 15 (UPK2), respectively. Two bovine genomic DNA sequences reactive with a uroplakin III cDNA probe were identified and mapped to BTA 6 (UPK3A) and 5 (UPK3B). They also mapped genes for uroplakins IA and II in mice, to the proximal regions of mouse chromosomes 7 ( U p k l a ) and 9 (Upk2), respectively, by analyzing the inheritance of restriction fragment length variants in recombinant inbred mouse strains. These assignments are consistent with linkage relationships known to be conserved between cattle and mice. The mouse genes for uroplakins IB and III were not mapped because the mouse genomic DNA fragments that hybridized with these probes were invariant among the inbred strains tested. Although the stoichiometry of AUM proteins is nearly constant, the fact that the uroplakin genes are unlinked indicates that their expression might be independently regulated. These results also predict likely positions for two h u m a n uroplakin genes. These studies demonstrate the success of the gene family approach to developing the map of the cattle genome, an approach equally applicable to any species for which hybrid cell lines are available. Another approach has been to directly evaluate regions of possible chromosome conservation by mapping in cattle genes that span the entirity of a specific h u m a n (or mouse) chromosome. Using the same panel of hybrid somatic cells, sequences homologous to genes spanning h u m a n chromosome arm 8q were syntenically assigned in cattle by Threadgill et al. (1990b). Thyroglobulin (TG), carbonic anhydrase II (CA2), and the proto-oncogenes myc and mos were assigned to a newly identified bovine syntenic group, U23. Additionally, in situ hybridization of the TG probe to bovine metaphase chromosomes revealed this syntenic group to be on bovine chromosome 14 and the bovine TG gene to reside at 14q12 ~ q15.
176
JAMES E. WOMACK
Four genes having homologous loci on the short arm of human chromosome 8 were mapped to two different bovine syntenic groups (Threadgill and Womack, 1991b). The gene coding for the tissue-type plasminogen activator mapped with GSR, a h u m a n chromosome 8 marker of bovine syntenic group U14, while lipoprotein lipase and the medium and light neurofilament polypeptide genes were shown to be syntenic with the h u m a n chromosome 9 marker GGTB2, which is on bovine syntenic group U18. The somatic cell panel was also used to investigate the syntenic relationship of nine loci in the bovine that have homologous loci on h u m a n chromosome 9 (Threadgill and Womack, 1990). Six loci-ALDH1, ALDOB, C5, GGTB2, GSN, and ITIL--were assigned to the previously identified bovine syntenic group U18 represented by ACO 1, whereas the other three loci--ABL, ASS, and A R P 7 8 m m a p p e d to a new, previously unidentified autosomal syntenic group. Additionally, a secondary locus, ABLL, which cross-hybridized with the ABL probe, was mapped to bovine syntenic group U1 with the HSA 1 loci PGD and ENO1. The results predicted that ACO1 would map proximal to ALDH1; GRP78 distal to ITIL and C5; GSN proximal to AK1, ABL, and ASS on HSA 9; GRP78 to MMU 2; and ITIL and GSN to MMU 4. Hybrid somatic cells were used again to investigate the syntenic relationship of nine loci in the bovine that have homologous loci on h u m a n chromosome 12 (Threadgill et al., 1990a). Eight loci, including A2M, GLI, HOX3, IFNG, INT1, KRAS2, NKNB, and PAH, were assigned to the previously identified bovine syntenic group U3 represented by GAPD. However, a single locus from the q-terminus of HSA 12, ALDH2, mapped to a new, previously unidentified autosomal syntenic group. These results indicated the existence of a very large ancestral syntenic group spanning from the p-terminus to q24 of HSA 12 and containing over 4% of the m a m m a l i a n genome. Additionally, the results predicted that ALDH2 is distal to PAH and IFNG on HSA 12, the type II keratin gene complex resides between q l l and q21 of HSA 12, A2M will map to MMU 6, and LALBA and GLI are on MMU 15. To determine the extent of conservation between bovine syntenic group U10, h u m a n chromosome 21 (HSA 21), and mouse chromosome 16 (MMU 16), 11 genes were physically mapped by segregation analysis in somatic cells by Threadgill et al. (1991). The genes chosen for study spanned MMU 16 and represented virtually the entire q arm of HSA 21. Because the somatostatin (SST) gene was previously shown to be in U10, the transferrin (TF) gene, an HSA 3/MMU 9 marker, was also mapped to determine whether U10 contained any HSA 3 genes not represented on MMU 16. With the exception of the protamine gene
MAPPING ANIMAL GENOMES
177
PRM1 (HSA 16/MMU 16), all of the genes studied were syntenic on bovine U10. Thus, all homologous loci from HSA 21 that have been studied in the cow are on a single chromosome. The bovine homolog of HSA 21 also carries several HSA 3 genes, two of which have homologous loci on MMU 16. The syntenic association of genes from the q arm of HSA 3 with HSA 21 genes in two mammalian species, the mouse and the cow, indicated that HSA 21 may have evolved from a larger ancestral mammalian chromosome that contained genes now residing on HSA 3. Additionally, the syntenic association of TF with SST in the cow permits the prediction that the rhodopsin (RHO) gene is proximal to TF on HSA 3q. Loci homologous to those on human chromosome 10 (HSA10) map to five mouse chromosomes: MMU2, MMU7, MMU10, MMU14, and MMU19. In cattle, one unassigned syntenic group (U26) was previously defined by the HSA10/MMU19 isoenzyme marker glutamic-oxaloacetic transaminase 1 (GOT1). To evaluate the syntenic arrangement of other HSA10 loci in cattle, seven genes were physically mapped by Threadgill and Womack (199 la). The genes mapped include vimentin (VIM) on HSA10 and MMU 2; interleukin 2 receptor (IL2R) on HSA10 and MMU?; ornithine aminotransferase (OAT) on HSA10 and MMU7; hexokinase 1 (HK1) on HSA10 and MMU10; retinol-binding protein 3 (RBP3) on HSA10 and MMU14; plasminogen activator, urokinase type (PLAU) on HSA10 and MMU14; and a-2-adrenergic receptor (ADRA2) on HSA10 and MMU 19. VIM and IL2R mapped to bovine U l l ; ADRA2 and OAT mapped to bovine U26; and RBP3, PLAU, and KH1 mapped to bovine U28. Homologs to genes residing on human chromosome 3 (HSA 3) map to four mouse chromosomes (MMU) 3, 6, 9, and 16. In the bovine, two syntenic groups that contain HSA 3 homologs, unassigned syntenic groups 10 (U10) and 12 (U12), were known. U10 also contains HSA 21 genes, which is similar to the situation seen on MMU16; whereas, U12 apparently contains only HSA 3 homologs. The syntenic arrangement of other HSA 3 homologs in the bovine was investigated by Threadgill and Womack (1991c). The genes mapped include Friend-murine leukemia virus integration site 3 homolog (FIM3; HSA 3/MMU 3), sucraseisomaltase (SI) and glutathione peroxidase 1 (GPX1) (HSA 3/MMU ?), murine leukemia viral (v-raf-1) oncogene homolog 1 (RAF1; HSA 3/MMU 6), and ceruloplasmin (CP; HSA 3/MMU 9). FIM3, SI, and CP mapped to bovine syntenic group U10, while RAF1 and GPX1 mapped to U12. Genes homologous to those located on human chromosome 4 (HSA4) were mapped in the bovine by Zhang et al. (1992) to determine regions
178
JAMES E. WOMACK
of syntenic conservation among humans, mice, and cattle. Previous studies had shown that two homologs of genes on HSA4, PGM2, and PEPS were located in bovine syntenic group U15 (chromosome 6). The homologous mouse genes, Pgm-1 and Pep-7, are on MMU5. Using the panel of bovine-hamster hybrid somatic cells, they assigned homologs of 11 additional HSA4 loci to their respective bovine syntenic groups. D4S43, D4S10, QDPR, IGJ, ADH2, KIT, and IF were assigned to syntenic group U15. This syntenic arrangement is not conserved in the mouse, where D4s43, D4sl0, Qdpr, and Igj are on MMU5 while A d h - 2 is on MMU3. IL-2, FGB, FGG, and F l l , which also reside on MMU3, were assigned to bovine syntenic group U23. These data suggested that breaks or fusions of ancestral chromosomes carrying these genes occurred at different places during the evolution of humans, cattle, and mice. In an effort to generate a more complete bovine syntenic map of Type I comparative anchor loci, seven homologs to genes found on HSA5 were mapped by Zhang and Womack (1992). Five HSA5 genes--CSF2, RPS14, PDGFRB, FGFA, and C S F i R ~ w e r e assigned to bovine syntenic group U22 (chromosome 7), while two o t h e r s ~ C 9 and H G M C R ~ mapped to U10 and U5, respectively. Previous studies had assigned the HSA5 marker SPARC to bovine syntenic group U22. The mapping of genes spanning the length of HSA5 in cattle and also in mouse permits syntenic comparisons between prototypic genomes of three m a m m a l i a n orders, providing insight into the evolutionary history of this region of the ancestral mammalian genome. The studies just discussed demonstrated a high level of conservation between h u m a n chromosomes and bovine syntenic groups. One such comparison was between h u m a n chromosome 12 and bovine chromosome 5, where at least 16 loci have been shown to be conserved in a single homologous segment. However, the degree of conservation of order of the loci on bovine chromosome 5 was unknown and, in general, the conservation of order in comparisons between humans and cattle can only be speculated. We began studying the conservation of gene order by estimating the recombination fractions between five of the loci that were previously published as mapping to bovine chromosome 5 by a combination of in situ hybridization and analysis of bovine-rodent somatic cell hybrid lines to determine whether order has been conserved in the homologous segment of bovine chromosome 5 and h u m a n chromosome 12 (Barendse et al., 1992). Recombination fractions were estimated in reference pedigrees of cattle for the loci A2M, GSNL, HOX3, INT1, KRAS2, and PAH. RFLPs for all loci were defined by screening a panel of eight restriction endonucleases. A multipoint
MAPPING ANIMALGENOMES
179
analysis was performed to estimate support for the most likely order. This map shows an inverted order of some of the loci. Moreover, the distance spanned in cattle is less than a quarter the distance spanned in humans. Together, these data indicate that several chromosomal evolutionary events have occurred in the homologous segment shared by humans and cattle. This study emphasized the demand for extending comparative gene mapping to a study of comparative order of genes within conserved segments of synteny. These studies relied almost exclusively on hybridization to molecular probes. The advent of the PCR, however, opened new doors for rapid expansion of the genome maps of domestic animals. The PCR was combined with hybrid somatic cell technology to extend the bovine physical map by Dietz et al. (1992). Eight bovine loci--glycoprotein hormone alpha (CGA), coagulation factor X (F10), chromogranin A (CHGA), low-density lipoprotein receptor (LDLR), human prochymosin pseudogen (CYM), oxytocin (OXT), arginine-vasopressin (ARVP), and cytochrome oxidase c subunit IV pseudogene (COXP)m were assigned to bovine syntenic groups with this approach. CGA was assigned to bovine syntenic group U2, F10 to U27, CHGA to U4 [bovine chromosome 21], LDLR to U22, CYM to U6, OXT and ARVP to U l l , and COXP to U3 (bovine chromosome 5). Seven of these genes, CGA, F10, CHGA, LDLR, OXT, ARVP, and CYM, further delineated regions of chromosomal conservation on human chromosomes 6, 13, 14, 19, 20, 20, and 1, respectively. CHGA, OXT, and ARVP were yet unmapped in the mouse. Comparative mapping predicted that mouse CHGA would map to chromosome 12, and mouse OXT and ARVP would map to mouse chromosome 2. Furthermore, human CYM was predicted to be sublocalized to lp32-q21. The primers developed for these eight loci were the beginning of a collection of bovine expressed sequence tags for characterizing the hybrid cell panel. Neibergs et al. (1993) used the PCR primers designed to amplify bovine specific sequences of the ARVP, CGA, COXP, CYM, F10, inhibin A (INHBA), LDLR, and OXT genes in hybrid cells in a search for single-strand conformation polymorphisms in cattle populations. DNA from 75 animals comprising cross-bred and 7 purebred breeds were analyzed. ARVP, COXP, CYM, LDLR, and OXT were found to be polymorphic while CGA, F10, and INHBA were not. Polymorphic regions were identified within 206 bp of exon 1 of ARVP, 582 bp of the pseudogene COXP, 253 bp of exon 9 of CYM, 519 bp ofLDLR cDNA, and 160 bp of the upstream regulatory region of OXT. This was the first report of bovine polymorphisms for these genes and an important step in incorporating type I comparative anchor loci into the bovine linkage
180
JAMES E. WOMACK
map. Polymorphic loci were subsequently analyzed in pedigreed fullsib families and shown to be inherited in a Mendelian fashion. Despite the success with RFLPs and single-strand conformation polymorphisms (SSCPs), it was apparent by 1993 that microsatellites would become the marker of choice for linkage mapping in domestic animals. Several studies were initiated to isolate and map bovine microsatellites, including one by Steffen et al. (1993). A partial plasmid library with bovine genomic inserts of about 500 bp was screened with a (dC-dA)n(dG-dT) ~ oligonucleotide probe for the repeated nucleotide motif (CA)~. Eleven positive clones (0.3% of all colonies screened) were discovered and were subsequently isolated and sequenced. Eight microsatellite loci were analyzed, one with eight alleles, one with seven alleles, three with six alleles, one with three alleles, and two with two alleles. Six of these microsatellites were mapped by PCR analysis of a panel of somatic hybrid lines. This initial synteny mapping was later followed by linkage mapping when the appropriate reference families were available and the best available microsatellites had been selected. Some markers have both Type I and Type II characteristics. Such is the bovine gene for the p21 ~a~ protein activator (RASA), which includes in its 5' untranslated region a ( T G ) n repeat. Analysis of this ( T G ) n repeat by PCR amplification of genomic DNA revealed a four-allele polymorphism. A cDNA probe was used to assign RASA to the region 2.4-qter of bovine chromosome 7 by in situ hybridization. PCR analysis of our panel of somatic hybrid lines allowed the assignment of RASA to the unassigned syntenic group 22 (U22) and thus localizes U22 on chromosome 7 (Eggen et al., 1992). With the advent of FISH, the bovine synteny groups were quickly assigned to chromosomes. A 260-bp genomic pstI fragment, which encodes a portion of the carbohydrate recognition domain, for example, was used along with hybrid somatic cells to map the conglutinin gene (CGN1) to domestic cow (Bos taurus) syntenic group U29 by Gallagher et al. (1993). In turn, a cosmid containing the entire bovine CGN 1 was used with FISH to sublocalize this gene to cattle (BTA) 28 band 18. Because BTA 28 and several of the other small acrocentric autosomes of cattle are difficult to discriminate, they also chromosomally sublocalized CGN1 to the p arm of the lone biarmed autosome of the gaur (Bos gaurus). The use of the gaur 2/28 Robertsonian as a marker chromosome and the assignment of CGN 1 to BTA 28 helped resolve some of the early nomenclatural questions involving this cattle chromosome. The bovine linkage map developed quickly after the discovery of
MAPPING ANIMAL GENOMES
181
microsatellites but only after the assembly of reference families for linkage analysis. Barendse et al. (1993a) reported the first genetic map of highly polymorphic index DNA loci in livestock for bovine chromosome 21. This map consisted of six loci with an average heterozygosity of 82%, each with a minimum of five alleles, spaced at an average genetic distance of 9.7 cM, and covering most of the expected length of the acrocentric bovine chromosome 21. The order of markers along the chromosome was determined to be cen-ETH 131-UWCA 4-TGLA 337TGLA 122-CSSM 18-GMBT 16-tel. There was heterogeneity reported among the recombination fractions between the sexes. To identify physical and genetic anchor loci on bovine chromosomes, 13 cosmids, obtained after the screening of partial bovine cosmid libraries with the ( C A ) n microsatellite motif, were mapped by FISH by Solinas-Toldo et al. (1993). Eleven cosmid probes yielded a specific signal on one of the bovine chromosomes and identified the following loci: D5S2, D5S3, D6S3, D8S1, D l l S 5 , D13S1, D16S5, D17S2, D19S2, D19S3, D21S8. The microsatellite-containing regions were subcloned and sequenced. Primers were designed for eight of the nonsatellite cosmids, and seven of these microsatellites were polymorphic with between three and eight alleles on a set of outbred reference families. The polymorphic and chromosomally mapped loci were subsequently used to physically anchor other bovine polymorphic markers by linkage analysis. The microsatellite primers were also applied to DNA samples of the now fully characterized panel of somatic hybrid cell lines, allowing the assignment of seven microsatellite loci to defined syntenic groups. These assignments confirmed earlier mapping results and placed two formerly unassigned syntenic groups on specific chromosomes. Approximately 400 loci have been mapped in both cattle and humans. Most of these have also been mapped in mice. While extensive conservation of synteny has been observed between cattle and humans (Womack and Moll, 1986; Threadgill et al., 1991), conservation of linkage (conservation of gene order) may not be so prevalent. Barendse et al. (1994) incorporated a sufficient number of Type I loci into the linkage map to demonstrate the presence of several rearrangements of gene order within conserved syntenies. The sheep and cattle maps, as expected, are highly conserved with only three apparent differences (Echard et al., 1994). Developing these maps in parallel should enhance and economize the efforts targeted to each map. To make the process of comparative gene mapping in mammals more
182
JAMES E. WOMACK
efficient, O'Brien et al. (1993) proposed a list of comparative mapping anchor loci (CMAL). These 321 Type I markers were selected to span the h u m a n genome at approximately 10 cM intervals and to include the anchor loci for h u m a n and mouse maps. Concerted efforts to map domestic animal homologs of these loci by syntenic, cytogenetic, and linkage approaches will greatly facilitate comparative mapping and the use of the growing h u m a n and mouse data base for animal improvement. A breakthrough in comparative gene mapping is heterologous chromosomal painting, or ZOO-FISH painting. Rettenberger et al. (1995) painted the pig genome with 22 human autosome-specific libraries. Solinas-Toldo et al. (1995) have done the same experiments including the X chromosome in cattle. These studies delineate the boundaries of chromosomal conservation at the cytogenetic level and are thus far highly consistent with the results of comparative synteny mapping. Like synteny mapping, they do not address conservation of gene order. D. ECONOMIC TRAITS A number of traits of economic significance are beginning to appear on animal genome maps. In addition to bovine LAD (Shuster et al., 1992; Threadgill et al., 1991), UMPS is also mapped to a specific site on chromosome 1 (Schwenger et al., 1993; Ryan et al., 1994; Barendse et al., 1993). BoLA is associated with susceptibility to leukemia virus infection (Lewin et al., 1988). Georges et al. (1993a) have linked the polled locus to microsatellites on chromosome 1. The weaver disease maps to markers on chromosome 4 (Georges et al., 1993b) and also appears to be associated with a quantitative trait for improved milk production. Variation around the prolatin gene on chromosome 23 (Cowan et al., 1990) is related to milk production in certain Holstein sire families, and Georges et al. (1995) have used mapped microsatellites to locate an additional five QTL for milk production. The Booroola fecundity gene (Montgomery et al., 1993) and a hyper muscling gene called callipyge (Cockett et al., 1994) have been mapped using the genomic approach to ETL mapping in sheep. Growth and fatness QTL were mapped in pigs in a cross involving the European wild boar (Andersson et al., 1994) and previously the candidate gene approach identified the genetic defect for porcine malignant hyperthermia (Otsu et al., 1991). The number of ETL on animal genome maps is rapidly expanding.
MAPPING ANIMAL GENOMES
183
V. Future Directions
Sufficient numbers of mapped markers now exist in cattle, sheep, pigs, and chickens for extensive genome coverage of families segregating ETL. Unfortunately, different ETL require different segregating families. These resource families are usually expensive to develop and maintain. Nonetheless, the families are an integral and necessary step in the ultimate application of the gene map to economic improvement. The mapping of ETL to 10-20 cM regions of a chromosome may be followed by high resolution mapping to ultimately identify and clone the responsible genes. Chromosome-specific libraries will aid this process. Despite the difficulties and expense of obtaining adequate resource family material, ETL will continue to appear on cattle linkage maps over the next few years. Animal breeders will not be satisfied with ETL-marker distance of 5-10 cM, however, and the next major step, identifying and cloning ETL, is a formidable one. The high-resolution linkage maps, chromosomal deletions, and large insert contigs that have contributed greatly to positional cloning of h u m a n disease loci are simply not available for animal ETL cloning. Positional cloning of human genes is rapidly shifting toward the positional candidate (Collins, 1995) approach, however, which relies on the availability of a pool of expressed genes mapped to the same chromosomal regions as the disease gene. In livestock as in humans, the three-step process for positional candidate cloning will be (1) to localize the ETL to a chromosomal subregion, (2) to search data bases for reasonable candidate genes, and (3) to test the candidate gene for variation correlated with phenotype. Obviously, step 2 is unrealistic in cattle where currently only 400 of the 70,000 or so genes are even assigned to chromosomes. This step was almost as unrealistic in h u m a n genetics until very recently when several international initiatives targeted the large-scale mapping of expressed sequence tags (ESTs). The success of these efforts suggest that more than half of the h u m a n transcripts could be mapped when this article is published (Collins, 1995). Thus, the key to animal ETL cloning may be through comparative genome data bases, which translate a 10-cM bovine segment into its h u m a n counterpart, then search for h u m a n ESTs with attractive features relative to the bovine phenotype. Assuming 20 or so potential candidate genes per centimorgan, the work will then have only begun. Nonetheless, there is hope for ETL cloning through such a comparative positional candidate cloning strategy that is not available outside comparative mapping and extensive mapping of h u m a n transcripts.
184
JAMES E. WOMACK
Exploitation of human EST data bases for recovery of animal genes will require much more precise comparative maps than are currently available. Identifying the boundaries of conserved synteny will not be sufficient. We must identify breakpoints of internal rearrangements as well as the reciprocal rearrangements that have accompanied mammalian chromosomal evolution. REFERENCES Adkison, L. R., Leung, D. W., and Womack, J. E. (1988a). Somatic cell mapping and restriction fragment analysis of bovine alpha and beta interferon gene families. Cytogenet. Cell Genet. 47, 62-65. Adkison, L. R., Skow, L. C., Thomas, T. L., Petrash, M., and Womack, J. E. (1988b). Somatic cell mapping and restriction fragment analysis of bovine genes for fibronectin and gamma crystallin. Cytogenet. Cell Genet. 47, 155-159. Andersson, L., Archibald, A. L., Gellin, J., Schook, L. B. (1993). 1st pig gene mapping workshop (PGM1). Anim. Genet. 24, 205-216. Andersson, L., Haley, C. S., Ellegren, H., Knott, S. A., Johansson, M., Andersson, K., Andersson-Eklund, L., Edfors-Lilja, I., Fredholm, M., Hansson, I., H~kansson, J., and LundstrSm, K. (1994). Genetic mapping of quantitative trait loci for growth and fatness in pigs. Science 236, 1771-1774. Barendse, W., Armitage, S. M., Womack, J. E., and Hetzel, J. (1992). Linkage relations between A2M, HOX3, INT1, KRAS2, and PAH on bovine chromosome 5. Genomics 14, 38-42. Barendse, W., Armitage, S. M., Kirkpatrick, B. W., Moore, S. S., Georges, M., Womack, J. E., and Hetzel, J. (1993a). A genetic map of index DNA loci on bovine chromosome 21. Genomics 18, 598-601. Barendse, W., Armitage, S. M., Ryan, A. M., Moore, S. S., Clayton, D., Georges, M., and Womack, J. E. (1993b). A genetic map of DNA loci in bovine chromosome 1. Genomics 18, 602-608. Barendse, W., Armitage, S. M., Kossarek, L. M., Shalom, A., Kirkpatrick, B. W., Ryan, A. M., Clayton, D., Li, L., Neibergs, H. L., Zhang, N., Grosse, W. M., Weiss, J., Creighton, P., McCarthy, F., Ron, M., Teale, A. J., Fries, R., McGraw, R. A., Moore, S. S., Georges, M., Soller, M., Womack, J. E., and Hetzel, D. J. S. (1994). A genetic linkage map of the bovine genome. Nat. Genet. 6, 227-235. Beckmann, J. S., and Soller, M. (1988). Detection of linkage between marker loci and loci affecting quantitative traits in crosses between segregating populations. Theor. Appl. Genet. 76, 228-236. Bishop, M. D., Kappes, S. M., Keele, J. W., Stone, R. T., Sunden, S. L. F., Hawkins, G. A., Solinas-Toldo, S., Fries, R., Grosz, M. D., Yoo, J., and Beattie, C. W. (1994). A genetic linkage map of cattle. Genetics 136, 619-639. Burt, D. W., Bumstead, N., Bitgood, J. J., Ponce de Leon, F. A., and Crittenden, L. B. (1995). Chicken genome mapping: A new era in avian genetics. Trends Genet. 11, 190194. Castle, W. E., and Wachter, W. L. (1924). Variation of linkage in rats and mice. Genetics 9, 1-2. Clark, F. H. (1936). Linkage of pink-eye and albinism in the deer mouse. J. Hered. 27, 256-260. Cockett, N. E., Jackson, S. P., Shay, T. L., Neilson, D., Moore, S. S., Steele, M. R.,
MAPPING ANIMAL GENOMES
185
Barendse, W., Green, R. D., and Georges, M. (1994). Chromosomal location of the callipyge gene in sheep (Ovis aries) using bovine DNA markers. Proc. Natl. Acad. Sci. USA 91, 3019-3022. Collins, F. S. (1995). Positional cloning moves from perditional to traditional. Nat. Genet. 9, 347-350. Cowan, C. M., Dentine, M. R., Ax, R. L., and Schuler, L. A. (1990). Structural variation around prolactin gene linked to quantitative traits in an elite Holstein sire family. Theor. Appl. Genet. 79, 577-582. Cox, D. R., Burmeister, M., Price, E. R., Kim, S., and Myers, R. M. (1990). Radiation hybrid mapping: A somatic cell genetic method for constructing high-resolution maps of mammalian chromosomes. Science 250, 245-250. Dietz, A. B., Neibergs, H. L., and Womack, J. E. (1992). Assignment of eight loci to bovine syntenic groups by use of PCR: Extension of a comparative gene map. Mammal. Genome 3, 106-111. Echard, G., Broad, T. E., Hill, D., and Pearce, P. (1994). Present status of the ovine gene map (Ovis aries); comparison with the bovine map (Bos taurus). Mammal. Genome 5, 324-332. Eggen, A., Solinas-Toldo, S., Dietz, A. B., and Womack, J. E. (1992). RASA contains a polymorphic microsatellite and maps to bovine syntenic group U22 on chromosome 7q2.4-qter. Mammal. Genome 3, 559-563. Ellegren, H., Chowdhary, B. P., Johansson, M., Marklund, L., Fredholm, M., Gustavsson, I., and Andersson, L. (1994). A primary linkage map of the porcine genome reveals a low rate of genetic recombination. Genetics 137, 1089-1100. Feldman, H. W. (1924). Linkage of albino allelomorphs in rats and mice. Genetics 9, 487492. Foreman, M. E., and Womack, J. E. (1989). Genetic and synteny mapping of parathyroid hormone and beta hemoglobin in cattle. Biochem. Genet. 27, 541-549. Fournier, R. E. K., and Ruddle, F. H. (1977). Microcell-mediated transfer of murine chromosomes into mouse, Chinese hamster, and human somatic cells. Proc. Natl. Acad. Sci. USA 74, 319-323. Fries, R., Eggen, A., and Stranzinger, G. (1990). The bovine genome contains polymorphic microsatellites. Genomics 8, 403-406. Fries, R., Eggen, A., and Womack, J. E. (1993). The bovine genome map. Mammal. Genome 4, 405-428. Fries, R., Threadgill, D. W., Hediger, R., Gunawardana, A., Blessing, M., Jorcano, J. L., Stranzinger, G., and Womack, J. E. (1991). Mapping of bovine cytokeratin sequences to four different sites on three chromosomes. Cytogenet. Cell Genet. 57, 135-141. Gallagher, D. S., Threadgill, D. W., Jackson, A. P., Parham, P., and Womack, J. E. (1991). Somatic cell mapping of bovine clathrin light chain genes: Identification of a new bovine syntenic group. Cytogenet. Cell Genet. 56, 154-156. Gallagher, D. S., Jr., Gibbs, L. S., Shaffer, J. B., and Womack, J. E. (1992a). Somatic cell mapping of bovine EC-SOD and SOD1L loci. Genomics 12, 610-612. Gallagher, D. S., Womack, J. E., Wolfgang, B., and Pittler, S. J. (1992b). Syntenic assignments of Visual transduction genes in cattle. Genomics 14, 699-706. Gallagher, D. S., Threadgill, D. W., Ryan, A. M., Womack, J. E., and Irwin, D. M. (1993a). Physical mapping of the lysozyme gene family in cattle. Mammal. Genome 4, 368373. Gallagher, D. S., Jr., Ryan, A. M., Liou, L. S., Sastry, K. N., and Womack, J. E. (1993b). Somatic cell mapping of conglutinin (CGN 1) to cattle syntenic group U29 and fluorescence in situ localization to chromosome 28. Mammal. Genome 4, 716-719.
186
JAMES E. WOMACK
Georges, M., Drinkwater, R., King, T., Mishra, A., Moore, S. S., Nielsen, D., Sargeant, L. S., Sorensen, A., Steele, M. R., Zhao, X., Womack, J. E., and Hetzel, J. (1993a). Microsatellite mapping of a gene affecting horn development in Bos taurus. Nat. Genet. 4, 206-210. Georges, M., Lanthrop, M., Dietz, A. B., Mishra, A., Nielsen, D., Sargeant, L., Sorensen, A., Steele, M. R., Zhao, X., Leipold, H., and Womack, J. E. (1993b). Microsatellite mapping of the gene causing weaver disease in cattle will allow the study of an associated quantitative trait locus. Proc. Natl. Acad. Sci. USA 90, 1058-1062. Georges, M., Nielsen, D., Mackinnon, M., Mishra, A., Okimoto, R., Pasquino, A. T., Sargeant, L. S., Sorensen, A., Steele, M. R., Zhao, X., Womack, J. E., and Hoeschelel, I. (1995). Mapping quantitative trait loci controlling milk production in dairy cattle by exploiting progeny testing. Genetics 139, 907-920. Goss, S. J., and Harris, H. (1975). New method for mapping genes in human chromosomes. Nature 255, 680-684. Grosz, M. D., Womack, J. E., and Skow, L. C. (1992). Syntenic conservation of HSP70 genes in cattle and humans. Genomics 14, 863-868. Haldane, J. B. S., Sprunt, A. D., Haldane, N. M. (1915). Reduplication in mice. J. Genet. 5, 133-135. Hallerman, E. M., Theilmann, J. L., Beckmann, J. S., Soller, M., and Womack, J. E., Mapping of bovine prolactin and rhodopsin genes in hybrid somatic cells. (1988). Anim. Genet. 19, 123-131. Iannuzzi, L., Di Meo, G. P., Gallagher, D. S., Ryan, A. M., Ferrara, L., and Womack, J. E. (1993). Chromosomal localization of omega and trophoblast interferon genes in goat and sheep by fluorescent in situ hybridization. J. Hered. 84, 301-304. Klobutcher, L. A., and Ruddle, F. H. (1981). Chromosome mediated gene transfer. Annu. Rev. Biochem. 50, 533-554. Lalley, P. A., Davisson, M. T., Graves, J. A. M., O'Brien, S. J., Womack, J. E., Roderick, T. H., Creau-Goldberg, N., Hillyard, A. L., Doolittle, D. P., Rogers, J. A. (1989). Report of the committee on comparative mapping (HGM10). Cytogenet. Cell Genet. 51, 503532. Lander, E. S., and Botstein, D. (1989). Mapping mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121, 185-199. Lewin, H. A., Wu, M.-C., Stewart, J. A., and Nolan, T. J. (1988). Association between BoLA and subclinical bovine leukemia virus infection in a herd of Holstein-Friesian cows. Immunogenet. 27, 338-344. Li, L., Teale, A., Bensaid, A., Dunlap, S., Dietz, A. B., and Womack, J. E. (1992). Somatic cell mapping of T-cell receptor CD3 complex and CD8 genes in cattle. Immunogenet. 36, 224-229. McAvin, J. C., Patterson, D., and Womack, J. E. (1988). Mapping of bovine PRGS and PAIS genes in hybrid somatic cells: Syntenic conservation with human chromosome 21. Biochem. Genet. 26, 9-18. Montgomery, G. W., Crawford, A. M., Penty, J. M., Dodds, K. G., Ede, A. J., Henry, H. M., Pierson, C. A., Lord, E. A., Galloway, S. M., Schmack, A. E., Sise, J. A., Swarbrick, P. A., Hanrahan, V., Buchanan, F. C., and Hill, D. F. (1993). The ovine Booroola fecundity gene (FecB) is linked to markers from a region of human chromosome 4q. Nat. Genet. 4, 410-414. Nadeau, J. H. (1989). Maps of linkage and synteny homologies between mouse and man. Trends Genet. 5, 82-86. Neibergs, H. L., Dietz, A. B., and Womack, J. E. (1993). Single-stranded conformation polymorphisms (SSCPs) detected in five bovine genes. Anim. Genet. 24, 81-84.
MAPPING ANIMAL GENOMES
187
O'Brien, S. J. (1992). Mammalian genome mapping; lessons and prospects. Current Opinion Genet. Dev. 1, 105-111. O'Brien, S. J., Womack, J. E., Lyons, L. A., Moore, K. J., Jenkins, N. A., and Copeland, N. G. (1993). Anchored reference loci for comparative genome mapping in mammals. Nat. Genet. 3, 103-112. Otsu, K., Khanna, V. K., Archibald, A. L., and MacLennan, D. H. (1991). Cosegregation of porcine malignant hyperthermia and a probable causal mutation in the skeletal muscle ryanodine receptor gene in backcross families. Genomics 11, 744-750. Paterson, A. H., Lander, E. S., Hewitt, J. D., Peterson, S., Lincoln, S. E., and Tanksley, S. D. (1989). Resolution of quantitative traits into Mendelian factors by using complete linkage map of restriction fragment length polymorphisms. Nature 335, 721726. Rettenberger, G., Klett, C., Zechner, U., Kunz, J., Vogel, W., and Hameister, H. (1995). Visualization of the conservation of synteny between humans and pigs by heterologous chromosomal painting. Genomics 26, 372-378. Rohrer, G. A., Alexander, L. J., Keele, J. W., Smith, T. P., and Beattie, C. W. (1994). A microsatellite linkage map of the porcine genome. Genetics 136, 231-245. Rommens, J. M., Iannuzzi, M. C., Kerem, B.-S., Drumm, M. L., Melmer, G., Dean, M., Rozmahel, R., Cole, J. L., Kennedy, D., Hidaka, N., Zsiga, M., Buchwald, M., Riordan, J. R., Tsui, L.-C., and Collins, F. S. (1989). Identification of the cystic fibrosis gene: Chromosome walking and jumping. Science 245, 1059-1065. Ruddle, F. H. (1972). Linkage analysis using somatic cell hybrids. Adv. Hum. Genet. 3, 173-235. Ruddle, F. H. (1981). A new era in mammalian gene mapping: Somatic cell genetics and recombinant DNA methodologies. Nature 294, 115-120. Ruddle, F. H., and Creagan, R. P. (1975). Parasexual approaches to the genetics of man. Annu Rev. Genet. 9, 407-486. Ryan, A. M., Gallagher, D. S., Jr., SchSber, S., Schwenger, B., and Womack, J. E. (1994). Somatic cell mapping and in situ localization of the bovine uridine monophosphate synthase gene (UMPS). Mammal. Genome 5, 46-47. Ryan, A. M., Gallagher, D. S., and Womack, J. E. (1992). Syntenic mapping and chromosomal localization of bovine a and ~ interferon genes. Mammal. Genome 3, 575-578. Ryan, A. M., Womack, J. E., Yu, J., Lin, J.-H., Wu, X.-R., Sun, T.-T., Clarke, V., and D'Eustachio, P. (1993). Chromosomal localization of uroplakin genes of cattle and mice. Mammal. Genome 4, 656-661. Schwenger, B., SchSber, S., and Simon, D. (1993). DUMPS cattle carry a point mutation in the uridine monophosphate synthase gene. Genomics 16, 241-244. Shows, T. B. (1977). Genetic and structural dissection of human enzymes and enzyme defects using somatic cell hybrids. In "Isozymes: Current Topics in Biological and Medical Research," Vol. 2. (M. C. Rattazi, J. G. Scandalios, and G. S. Whitt, eds.), pp. 107-158. Alan R. Liss, New York. Shows, T. B., Sakaguchi, A. Y., Naylor, S. L. (1982). Mapping the human genome, cloned genes, DNA polymorphisms, and inherited disease. Adv. Hum. Genet. 12, 341-452. Shuster, D. E., Kehrli, M. E., Jr., Ackermann, M. R., and Gilbert, R. O. (1992). Identification and prevalence of a genetic defect that causes leukocyte adhesion deficiency in Holstein cattle. Proc. Natl. Acad. Sci. USA, 89, 9225-9929. Smith, C., and Simpson, S. P. (1986). The use of genetic polymorphism in livestock improvement. J. Anim. Breed. Genet. 103, 205-217. Solinas-Toldo, S., Fries, R., Steffen, P., Neibergs, H. L., Barendse, W., Womack, J. E., Hetzel, D. J. S., and Stranzinger, G. (1993). Physically mapped, cosmid-derived
188
JAMES E. WOMACK
micro-satellite markers as anchor loci on bovine chromosomes. Mammal. Genome 4, 720-727. Solinas-Toldo, S., Lengauer, C., and Fries, R. (1995). Comparative genomic map of human and cattle. Genomics 27, 489-496. Soller, M., and Beckmann, J. S. (1982). Restriction fragment length polymorphisms and genetic improvement. Proc. 2nd World Cong. Genet. Appl. Livestock Production 6, 396-404. Southern, E. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503-517. Steffen, P., Eggen, A., Dietz, A. B., Womack, J. E., Stranzinger, G., and Fries, R. (1993). Isolation and mapping of polymorphic microsatellites in cattle. Anim. Genet. 24, 121124. Sturtevant, A. H. (1913). The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. J. Exptl. Zool. 14, 43-59. Threadgill, D. W., Adkison, L. R., and Womack, J. E. (1990a). Syntenic conservation between humans and cattle. II. Human chromosome 12. Genomics 8, 29-34. Threadgill, D. W., Fries, R., Faber, L. K., Vassart, G., Gunawardana, A., Stranzinger, G., and Womack, J. E. (1990b). The thyroglobulin gene is syntenic with the myc and mos protooncogenes and carbonic anhydrase II and maps to chromosome 14 in cattle. Cytogenet. Cell Genet. 53, 32-36. Threadgill, D. S., Krause, J. P., Krawetz, S. A., and Womack, J. E. (1991). Evidence for the evolutionary origin of human chromosome 21 from comparative gene mapping in the cow and mouse. Proc. Natl. Acad. Sci. USA 88, 154-158. Threadgill, D. W., and Womack, J. E. (1990). Syntenic conservation between humans and cattle. I. Human chromosome 9. Genomics 8, 22-28. Threadgill, D. S., and Womack, J. E. (1991a). Mapping HSA10 homologous loci in cattle. Cytogenet. Cell Genet. 57, 123-129. Threadgill, D. W., and Womack, J. E. (1991b). Synteny mapping of human chromosome 8 loci in cattle. Anim. Genet. 22, 117-122. Threadgill, D. S., and Womack, J. E. (1991c). Mapping HSA 3 loci in cattle: Additional support for the ancestral synteny of HSA 3 and 21. Genomics 11, 1143-1148. Tobin-Janzen, T. C., and Womack, J. E. (1992). Comparative mapping of IGHG1, IGHM, FES, and FOS in domestic cattle. Immuno. 36, 157-165. Trask, B. J. (1991). Fluorescence in situ hybridization. Trends Genet. 7, 150-154. Troyer, D. L., Xie, H., Goad, D. W., and Skinner, D. Z. (1994). Use of a new technique to map the porcine ~ interferon gene to chromosome 1. Mammal. Genome 5, 112-114. Vaimen, D., Bahri-Darwich, I., Mercier, D., Yerle, M., Eggen, A., Leveziel, H., Guerin, G., Gellim, J., and Cribu, E. P. (1993). Mapping of new bovine microsatellites on cattle chromosome 15 with somatic cell hybrids, linkage analysis, fluorescence in situ hybridization. Mammal. Genome 4, 676-679. Van Eijk, M. J. T., Russ, I., and Lewin, H. A. (1993). Order of bovine DRB3, DYA, and PRL determined by sperm typing. Mammal. Genome 4, 113-118. Walter, M. A., Spillett, D. J., Thomas, P., Weissenbach, J., and Goodfellow, P. N. (1994). A method for constructing radiation hybrid maps of whole genomes. Nat. Genet. 7, 2228. Weber, J. L., and May, P. E. (1989). Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 44, 388-396. Weller, J. L., Kashi, Y., and Soller, M. (1990). Power of daughter and granddaughter designs for determining linkage between marker loci and quantitative trait loci in diary cattle. J. Dairy Sci. 73, 2525-2537.
MAPPING ANIMAL GENOMES
189
Weiss, M. C., and Ephrussi, B. (1966). Studies of interspecific (rat x mouse) somatic hybrids. II. Lactic dehydrogenase and f3-glucuronidase. Genetics 54, 1111-1112. Wigler, M., Silverstein, S., Lee, L.-S., Pellicer, A., Cheng, Y.-C., and Axel, R. (1977). Transfer of purified herpes virus thymidine kinase gene to cultured mouse cells. Cell 11, 223-232. Womack, J. E., and Moll, Y. D. (1986). A gene map of the cow: Conservation of linkage with mouse and man. J. Hered. 77, 2-7. Xie, H., Alexander, L. J., Rohrer, G. A., Beattie, C. W., and Troyer, D. L. (1995). Use of DISC-PCR and FISH to assign a linkage group to pig chromosome 10. Mammal. Genome 6, 139-141. Zhang, N., Threadgill, D. W., and Womack, J. E. (1992). Synteny mapping in the bovine: Genes from human chromosome 4. Genomics 14, 131-136. Zhang, N., and Womack, J. E. (1992). Synteny mapping in the bovine: Genes from human chromosome 5. Genomics 14, 126-130.
This Page Intentionally Left Blank
ADVANCESIN VETERINARYMEDICINE,VOL. 40
The Canine Genome CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington 981 04 I. Introduction II. Applications of a Canine Genetic Map A. Canine Models for H u m a n Genetic Diseases B. Diseases of Interest to Specific Breeds III. Organization of the Canine Genome A. Chromosomes B. Repeated Elements in the Canine Genome IV. Development of a Canine Genetic Map A. Overview and Approach to Map Construction B. Microsatellite Repeats C. Assignment of Genetic M a r k e r s to Chromosomes D. Ordering of Markers on Chromosomes V. Targeted Cloning of Canine Genes A. Genes of Interest B. The Canine I m m u n e and Hematopoietic Systems VI. Considerations and Conclusions References
I. Introduction Selected breeding to exploit physical attributes and behavioral skills has made the modern dog the most diverse species on earth. All dogs are members of the same species, and dogs of any breed can be crossed to produce fertile offspring. Yet, among the over 320 different breeds recognized by various kennel clubs, there exists a range of morphological variation that is unsurpassed among other species of animals. In contrast, because of the institution of breed standards, dogs within a 191 Copyright 9 1997 by Academic Press. All rights of reproduction in any form reserved. 1093-975X/97 $25.00
192
CATHRYN S. MELLERSHAND ELAINE A. OSTRANDER
breed are relatively homogeneous. From a genetic viewpoint, this combination of uniformity and diversity is a tremendous asset. The uniformity ensures that nearly all dogs of a given breed will share significant homogeneity at many loci in addition to those that are important to the breed standard. The diversity, however, ensures a wealth of opportunities to genetically map and characterize genes responsible for differences. One unfortunate consequence of inbreeding is the maintenance of genetic diseases within many breeds. Autosomal recessive traits present the greatest problems. Asymptomatic carrier dogs produce apparently healthy puppies until they are crossed with another dog carrying the same disease-associated allele. Thus, carrier status may not be evident until several litters have been produced. Alternatively, undesirable traits may be maintained within a breed because symptoms do not develop until after reproductive age. Some diseases that fall into this category include various heart defects, epilepsy, and susceptibility to autoimmune disorders. Two other factors that complicate genetic mapping are the occurrence of diseases that have variable penetrance and diseases caused by multiple genes. In the case of variable penetrance, affected dogs may display the trait at differing levels, with some dogs even appearing unaffected. As a result, apparently healthy dogs may be naively crossed, producing dogs that are later recognized as diseased. Traits prone to variable penetrance are often affected by environmental factors such as nutrition and infectious diseases. Complex traits, by comparison, are those that are controlled by several genes, each of which contributes partially to a final phenotype. The genetics of complex traits is sophisticated and requires pedigrees with large numbers of offspring and well-documented disease status. Even among the most diligent breeders, this information is often incomplete. Mapping both simple and complex diseases in dogs will become possible through the development of a canine genetic map. A genetic map is composed of polymorphic genetic markers. When analyzed together with the disease status of members of a family, the genetic map functions as a series of molecular signposts that highlight regions of the genome likely to contain genes of interest. In the following sections, we briefly review our motivation to construct a canine genetic map, the organization of the canine genome, and the development and use of genetic markers in constructing the map. Finally, we conclude with a brief discussion of genetic registries and the use of genetic markers as diagnostic tools.
THE CANINE GENOME
193
II. Applications of a Canine Genetic Map Purebred dogs are plagued with many inherited diseases. One major application of a canine genetic map will be to facilitate study of such diseases. This article is not an attempt to detail all canine diseases; other excellent resources exist for this purpose. These resources include (1) Online Mendelian Inheritance in Animals (OMIA), (2) the Canine Genetic Disease Information System, and (3) Genetics of the Dog by Malcolm Willis (Willis, 1989). OMIA is a catalog of inherited disorders listing clinical and diagnostic features, providing summaries of current knowledge, and presenting a discussion of the presumed mode of inheritance for each disorder. It was developed by several investigators including Steve Brown and Paul LeTissier. The Canine Genetic Disease Information System, developed by Dr. Donald Patterson and colleagues in the section of Medical Genetics at the Veterinary School of the University of Pennsylvania provides descriptive and clinical information about a wealth of diseases. Genetics of the Dog lists inherited disorders of the dog by system and includes comments on breed susceptibility. Other excellent resources exist, or are in preparation, but the three listed here are sources of detailed and comprehensive information for anyone interested in canine inherited diseases. To date, the most extensively studied diseases are either homologs of h u m a n conditions or diseases that are widespread and threatening within particular breeds. Following are a few relevant examples. A. CANINE MODELS FOR HUMAN GENETIC DISEASES
Many canine pedigrees with inherited diseases are studied because they are expected to provide useful data about similar diseases in humans, such as inherited neurological disorders. Among the most well studied is hereditary canine spinal muscular atrophy (HCSMA), which is similar to human spinal muscular atrophy, an autosomal recessive degenerative motor neuron disease. HCSMA was originally identified in Brittany Spaniels (Cork et al., 1979). The disease produces progressive weakness and atrophy of muscle groups because of lower motor neuron loss. The course of illness depends on the particular variant of the disease, but the illness is often fatal, usually as a result of the involvement of the muscles of respiration and swallowing. In dogs, the disease appears to be autosomal dominant, with variable penetrance. Dogs that are presumed homozygotes for the disease locus have a profound phenotype, characterized by accelerated
194
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
disease, with weakness beginning between 6 and 8 weeks of age and progressing rapidly to complete tetraparesis by 13-16 weeks (Sack et al., 1984). Extensive pathologic studies on HCSMA showed that the disease recapitulates many aspects of human motor neuron diseases. Human spinal muscular atrophy (SMA) is complex, because there appear to be three forms with different ages of onset, each of which maps to a wellcharacterized region of chromosome 5. The causative gene (or genes) is likely one of several located in a region of generalized instability on the chromosome. The unstable nature of this region has made the genetics of human SMA difficult to study. Hence, studies have been initiated on the canine disorder. One additional example is provided by the work of Mignot and colleagues, who have studied the genetics of canine narcolepsy, an autosomal recessive trait with full penetrance in Doberman pinschers and Labrador retrievers (Baker et al., 1982; Mignot et al., 1991). Unlike h u m a n narcolepsy, which appears to be polygenic, recent work from this group suggests that a single autosomal recessive gene, canarc-1, is responsible for the disorder. Canarc-1 is an immunoglobulin muswitchlike gene. The identification of a non-Major Histocompatibility Complex (MHC) gene for this disorder is noteworthy because h u m a n narcolepsy has been linked thus far only to MHC gene class II alleles (Marcadet et al., 1985). Because the disorder in humans involves multiple genes, the findings from Mignot and colleagues may suggest additional strategies for researchers studying human narcolepsy. An understanding of the genetic basis of some diseases has raced ahead of that of other disorders, simply because of the simple patterns of inheritance. A well-studied example is an X-linked severe combined immunodeficiency syndrome called X-SCID. In dogs, this disease is similar to one form of human SCID and is characterized by profound defects in cellular and humoral immunity, resulting in a normal number of circulating B cells and IgM, but low levels of serum IgG and IgA (Jezyk et al., 1989). In humans, the disease is associated with mutations in the gene for the gamma chain of the interleukin-2 receptor (Noguchi et al., 1993). Henthorn and colleagues have shown that mutations in the same gene are responsible for the canine disease in a colony of dogs established from a single X-linked SCID carrier female (Henthorn et al., 1994). More recently, a different mutation in the same gene has been found in a Welsh Corgi puppy presenting with classic symptoms of X-SCID, suggesting that different mutations in the same gene may account for the same disease in different dogs or
THE CANINE GENOME
195
different breeds (Somberg et al., 1995). Maintenance of this colony and cloning of the relevant genes offered a unique opportunity to characterize this devastating disorder. B. DISEASES OF INTEREST TO SPECIFIC BREEDS The predisposition of various dog breeds to specific inherited disorders is well recorded. Some of the inherited diseases described in the literature are thought to be breed specific. For instance, collie eye anomaly (CEA) is believed to affect the rough and smooth collie, the Shetland sheepdog, and, rarely, the border collie. The disease has variable expression and confusion remains as to its true mode of inheritance. Affected dogs share a series of ocular defects that typically include tortuous retinal arteries and veins, a pale area or specifically placed streaks on the fundus of the eye, and coloboma in the area near the optic nerve head (Roberts et al., 1966). The disorder ranges from slight to total blindness. The incidence of the different forms varies, but the disease is clearly problematic for the threatened breeds, and genetic studies are needed to facilitate testing. Other disorders, because of their high frequency, are nearly synonymous with the breed in which they were first reported. An example is Scottie or Scotch cramp. Although most common in Scottish terriers, it has also been reported in wire-haired fox terriers, cocker spaniels, and dalmatians (Meyers et al., 1970; Smythe, 1945; Woods, 1977). The disease appears to be autosomal and recessive; affected dogs undergo periods of muscular hypertension, which are relieved after resting. The dogs appear to neither be in pain nor lose consciousness. Similar diseases are reported in other breeds. Although the disease is not generally life threatening, its unique characteristics and clear genetics have made it a focus of significant study. Perhaps the most devastating diseases are those for which the dog appears normal at birth and early life, and then displays a syndrome during mid-life. Epidemiological data exist, for instance, to suggest that particular forms of cancer are much more common in some breeds than others. Historically, the boxer breed is associated with a variety of types of tumors, including thyroid, cutaneous, nervous system, and mammary tumors (MacVean et al., 1978). Anecdotal evidence suggests other breeds also have genetic predisposition to particular tumors including melanomas in Scottish terriers (Cotchin, 1954; Cotchin, 1955), mammary tumors in cocker spaniels, and mastocytoma for the Boston terriers.
196
CATHRYN S. MELLERSHAND ELAINE A. OSTRANDER
The genetic basis of these traits has been difficult to study because there are probably multiple genes involved. In general, the strategies necessary to study multi-gene traits are much more complex than those used to map the genes responsible for single-gene disorders. Mapping complex traits typically requires families that display the trait at its extreme levels, i.e. profoundly affected dogs as well as mildly affected dogs. Rigorous quantitative data regarding strength of phenotype must be available for multiple generations of the family. Finally, a certain amount of luck is required. If a particular trait is controlled by a large number of genes, each of which contributes only minimally to the final complex phenotype, it will be difficult to dissect out the contributions of any one gene. If, however, the phenotype is controlled largely by a small number of genes, i.e. two or three, the genetics will be much easier to resolve. The highest level of resolution is achieved when a high resolution map is used to analyze rigorous quantitative data from several generations of a family. Complex diseases such as cancer, epilepsy, and hip dysplasia are most productively studied using a genetic map. A genetic map allows an investigator to navigate around the genome and, when properly developed and implemented, is a key resource for identifying regions of the genome involved in disease. The development of a canine map will enormously influence canine genetic research in the future. Because of the central role it will play, the organization of the canine genome and the construction of a canine genetic map are discussed in detail in the following section.
III. Organization of the Canine Genome A. CHROMOSOMES The dog has 38 pairs of autosomes, most of which are small, acrocentric chromosomes, plus an X and Y. Historically, however, it has been difficult to take advantage of the organization of the dog genome into so many discrete linkage groups because cytogenetic analysis of dog chromosomes is extremely difficult. A number of reports on banding of dog chromosomes were published in the mid-1970s, but they failed to agree on the banding patterns of many of the chromosomes, and numbering assignments were not consistent. As cytogenetic technologies have improved, interpretation of Giemsa-banded dog chromosomes has become much easier. Stone et al. (1991) have used cell synchronization techniques to define higher resolution banding patterns
THE CANINE GENOME
197
at the 327 band level for canine chromosomes. The only breed difference reported among the dogs examined was the degree of chromosome condensation. A higher resolution karyotype was reported by Graphodatsky et al. (1995), who used GTG banding to define dog and silver fox chromosomes to the 400 band level. Langford et al. (1995) have used molecular genetic techniques to develop a set of chromosome "paints" specific for the canine genome, in collaboration with investigators from the Animal Health Trust in Newmarket and the Sanger Centre in Cambridge. To generate these paints, investigators used flow cytometry to separate the 38 canine chromosomes into 32 reproducible and recognizable peaks, with most of the peaks representing one chromosome and a small number representing two. The deoxyribonucleic acid (DNA) from each peak was amplified and fluorescently labeled using polymerase chain reaction (PCR) and then hybridized to metaphase spreads of dog chromosomes. The labeled products anneal specifically to the particular chromosome from which they were derived (hence the name paints) and are becoming an important resource with which to reproducibly distinguish one chromosome from another. As a result, chromosome numbers can now unambiguously be assigned to the 21 largest chromosomes (Langford et al., 1995), and it is expected that further development of additional reagents will allow all of the chromosomes to be definitively numbered. B. REPEATED ELEMENTS IN THE CANINE GENOME
The canine genome, like other m a m m a l i a n genomes, is characterized by several classes of repetitive elements, one of which is the Jeffreys minisatellite probe. These probes are DNA sequences isolated from the h u m a n genome that have been shown to cross-hybridize to dog DNA to give individual specific fingerprints in dogs (Jeffreys and Morton, 1987, and references therein). The probes consist of a reiterated short core sequence that hybridizes to many families of tandemly repeated DNA sequences that show multiallelic length variations. The typical assay hybridizes radiolabeled core elements to DNA cleaved with restriction enzymes to produce a characteristic Southern blot fingerprint. With the exception of monozygotic twins, DNA fingerprints are absolutely individual-specific. These probes are useful to study migration of small isolated populations (Gilbert et al., 1990) and to resolve cases of questionable percentage. In one clear illustration of this phenomenon, Jeffreys and Morton (1987) demonstrated that DNA fingerprints of any two purebred whippets were no more similar than fingerprints derived from dogs of completely unrelated breeds.
198
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
IV. Development of a Canine Genetic Map A. OVERVIEW
AND
APPROACH
TO MAP
CONSTRUCTION
A genetic map consists of a series of genes or genetic markers for which the specific order and distances between them are known. A framework map is composed entirely of markers for which the order and spacing is very well supported and, although it may be of low density, is very reliable. This is in contrast to a comprehensive map that aims to include all syntenic markers. Although comprehensive maps are usually very dense, they may be locally unreliable with respect to both order and spacing. Maps are constructed by examining the inheritance of polymorphic alleles in the offspring of a specific mating. If two markers are on different chromosomes, their alleles will be inherited independently; if they are on the same chromosome, they may segregate nonrandomly and are said to be linked. Genes that are close together are usually inherited together, while genes that are far apart on the same chromosome segregate independently because genetic recombination between chromosomes at meiosis disrupts their association. For a given area of the genome, the probability of a genetic recombination event occurring between a pair of m a r k e r s is proportional to the distance between them. The probability of a genetic recombination event occurring is expressed as a recombination fraction or, after minor mathematical adjustment for the possibility of double recombinants, as centiMorgans (cM). One percent recombination is equal to approximately lcM, which roughly corresponds to a million base pairs in h u m a n s and just over two million in the mouse. To be useful, a genetic map should have m a r k e r s placed at least every 10 cM. If the size of the dog genome is assumed to be similar to that of humans, such a map will consist of about 350 well-spaced markers. The more densely the map is covered, however, the more useful it will be. The h u m a n genetic map, for instance, currently has over 6000 markers, and is still considered incomplete (Murray et al., 1994; Weissenbach et al., 1992). Increasing map density improves the precision with which a disease region can be localized. A 10 cM map, for instance, will allow localization of a disease gene to within only 5 million base pairs, whereas a map of I cM density, composed of 3000 markers, will localize a gene to within half a million base pairs. The construction of a genetic map takes place in several sequential steps. First, large numbers of polymorphic genetic markers are isolated from the genome of interest. Second, the m a r k e r s are placed into
THE CANINE GENOME
199
linkage groups and, finally, the order of markers in a given linkage group is determined. Each of these steps is discussed in the next section. B. MICROSATELLITE REPEATS
A genetic marker, by definition, must have two or more alleles; if the frequency of the most common allele is less than 95% the marker is described as polymorphic. One measure of the degree of polymorphism displayed by a given marker is the polymorphism information content (PIC) value (Botstein et al., 1980). The PIC value defines the probability that for a particular marker, the genotype of a given offspring will allow deduction of which the parents' two marker alleles it received. The PIC has a value of 0.0-1.0 and uses information about the number of alleles for a given marker in the population and the frequency with which each allele appears. The ideal map is made up of large numbers of highly informative markers, typically with PIC values greater than 0.7. The most useful type of genetic markers are microsatellites. Microsatellites are small stretches of DNA composed of small mono-, di-, trior tetranucleotide motifs repeated in tandem, such as (CA)n, where n is typically 5-30 (Hamada and Kakunaga, 1982; Miesfeld et al., 1981; Stallings et al., 1991; Tautz and Renz, 1984). Microsatellites are polymorphic for repeat number within populations, yet with an estimated mutation rate to new alleles of only 5 x 10 -4 to 10 -5 per allele per meiosis in the h u m a n genome (Kwiatkowski et al., 1992), are sufficiently stable to use as genetic markers of Mendelian inheritance (Litt and Luty, 1989; Weber and May, 1989). Microsatellites are found in large numbers in all eukaryote genomes and, equally importantly for map building, the distribution of these simple sequence repeats around the genome is approximately random. In both humans and dogs, for instance, a dinucleotide (CA)~ repeat is found on average every 30-60 kilobases (kb) (Stallings et al., 1991; Ostrander et al., 1993). Individual microsatellite markers are distinguished from one another by PCR primers designed to match the unique DNA sequences that bracket each repeat (Fig. 1). The inheritance of individual alleles is easily measured by the PCR, which detects the length of polymorphisms that characterize different alleles (Litt and Luty, 1989; Weber and May, 1989). Microsatellite-based markers have been used to construct genetic maps of many organisms including h u m a n (Murray et al., 1994; Weissenbach et al., 1992), mouse (Dietrich et al., 1992), and cattle and pigs (Bishop et al., 1994;
Heterozygote
(CA)n lllllllllllllilllllllllillllli/llllll/
Allele I t Allele 2
Homozygote
I
Illllllllliill!
(CA)n-x
PCRprimers ,'-I~ ~ i I
(CA)n -~
1111111111111111111111111111111111111 ~ l
J
illilillllllllllllllllllllillllllllll
(CA)n
~PCR
Allele 1 J Allele 1
~-
~PCR Electrophoresis
m
2
----
m
FIG. 1. Oligonucleotide primers are designed to match the unique DNA sequences that bracket microsatellites. The inheritance of individual alleles is monitored by amplifying the DNA by the polymerase chain reaction (PCR), followed by fractionation on polyacrylamide gels.
THE CANINE GENOME
201
Ellegren et al., 1993; Fries et al., 1993; Toldo et al., 1993). Efforts have been underway since 1992 to isolate large numbers of microsatellitebased markers for the canine genome (Ostrander et al., 1992; Ostrander et al., 1993; Holmes et al., 1993; Deschenes et al., 1994; Mellersh et al., 1994; Francisco et al., 1996). (CA)n repeat-based microsatellites are typically isolated from shortinsert genomic libraries screened with radioactively labeled ( C A ) n o r (TG) n probes. DNA is isolated from positive clones, it is sequenced, and marker-specific PCR primers are designed to flank each repeat. The development of a strategy to construct genomic libraries highly enriched for microsatellite repeats has been useful for the isolation of large numbers of microsatellite bearing clones (Ostrander et al., 1992). Although less frequent than ( C A ) n repeats, efforts in the human genome project have focused on the development of markers based on tetranucleotide repeats (Murray et al., 1994). Microsatellites based on tetranucleotide repeat motifs have been shown to give PCR products that lack the lower molecular weight "stutter" bands characteristic of some (CA)n repeats (Edwards et al., 1991; Edwards et al., 1992). In dogs, as in humans, a correlation has been demonstrated between the length of a microsatellite's repeat array and its PIC value, with longer repeats being more polymorphic than shorter ones (Weber, 1990; Ostrander et al., 1993). In canines, ( G n A t ) n repeats are usually long, with many having repeat units greater than (GAAA)5o. This particular class of tetranucleotide repeats has average PIC values of 0.75 (Francisco et al., 1996). A ( G A A A ) n repeat occurs about once every 100-300 kb in the canine genome (Francisco et al., 1996), only about threefold less than the frequency of ( C A ) n repeats, suggesting that these are sufficiently frequent and polymorphic to be a good resource for map building.
C. ASSIGNMENT OF GENETIC MARKERS TO CHROMOSOMES
In the 1960s, techniques were developed to create r o d e n t - h u m a n cell hybrids. These hybrids retain only a portion (typically one or two chromosomes) of the h u m a n genome together with a full complement of rodent chromosomes. It is possible to assign any h u m a n gene or DNA segment that can be distinguished from the rodent version, to the h u m a n chromosome contained within the hybrid cell line. Hybrid cell lines are used to assign new markers or genes to chromosomes by testing the marker against a comprehensive panel of cell lines in which each chromosome is represented at least once. For the canine genome,
202
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
a panel of canine-rodent hybrid cell lines would be of use for exactly this purpose. Hybrids also provide a mechanism with which to investigate the genomic organization of individual species and can be used to probe the relationships between genomes. As the numbers of genes mapped in organisms other than humans increases, the opportunity to study the syntenic relationships between chromosomes of different animals increases correspondingly. A set of anchored reference loci suitable for comparative mapping in mammals and other vertebrate classes has been proposed (O'Brien et al., 1993). These loci are cloned mouse and h u m a n functional genes, spaced an average of 5-10 cM apart throughout their respective genomes. Importantly, the genes are evolutionarily conserved and many are represented in the gene maps of other mammalian orders, particularly cattle and the domestic cat. Somatic cell hybrids can be used to assign anchored reference loci to syntenic groups in a particular animal, and then the linkage relationships of homologous genes can be compared between animals. Information regarding the conservation of synteny of homologous genes between different organisms, as well as considerations of gene order within homologous gene segments, will then contribute to our understanding of mammalian genome evolution. In addition, improved knowledge of conserved synteny will allow researchers investigating any disease to take full advantage of candidate gene approaches and explore the role of genes already characterized in other animals. D. ORDERING OF MARKERS ON CHROMOSOMES
1. Overview If two markers segregate consistently with a particular chromosome, they must be on the same chromosome and are said to be syntenic. Syntenic loci are not necessarily genetically linked; they may be so far apart on a given chromosome that genetic recombination between them approaches 50% and they segregate independently. Thus, while somatic cell hybrids can assign genes or DNA segments to particular chromosomes, they provide no information regarding the position or order of genes on chromosomes. The generation of a set of linearly arranged loci for which the genetic distance between adjacent markers has been determined produces a genetic map. The genetic distance between two markers is a measure of the expected number of crossovers occurring on a single chromosome strand between the two markers at meiosis. In meiosis (the cell division leading to gamete formation), homologous chromosomes pair up.
203
THE CANINE GENOME
At this stage, each chromosome consists of two identical s t r a n d s (chromatids), so each chromosome pair consists of four strands. D u r i n g meiosis, homologous chromosomes s e p a r a t e from each other except at one or two zones of contact, which are called chiasmata. C h i a s m a t a involve one c h r o m a t i d from each pair of homologous chromosomes and are the points at which t h e r e is formation of crossovers between chromatids. This process leads to genetic recombination and can involve either of the two chromatids. The alleles received by an individual from one p a r e n t are called a haplotype; if the same alleles are passed from g r a n d p a r e n t to F 2 offspring the haplotype is described as p a r e n t a l or nonrecombinant. In Fig. 2, the f a t h e r is heterozygous for both genes, h a v i n g inherited the haplotype AB from his f a t h e r and ab from his mother. If these two genes were i n h e r i t e d i n d e p e n d e n t l y (as they would be if t h e y were on different chromosomes), four types of g a m e t e would be expected (AB, Ab, aB, and ab) in a p p r o x i m a t e l y equal proportions. If, in contrast, the genes were relatively close together on the s a m e chromosome, this AaBb individual would be expected to produce an excess of the two
~
AA B B
aa
bb
)
i H
Parental haplotype
A a
Bb
~
Non-parental haplotype
i ,1 !
I
[:, I
l l
I
aa bb
l:
)
FIo. 2. The alleles inherited by an indi~dua] from one parent are known as a haplot-
ype; if the same alleles are passed from grandparent to F 2 offspring, the haplotype is parental. Linked genes give rise to an excess of parental gametes, whereas unlinked genes produce equal numbers of parental and nonparental gametes. Five of the six F 2 offspring inherit the parental haplotypes AB or ab, whereas one offspring receives the nonparental, or recombinant Ab haplotype.
204
CATHRYN S. M E L L E R S H AND E L A I N E A. O S T R A N D E R
parental haplotypes (AB and ab) over a smaller number of the nonparental or recombinant haplotypes (Ab and aB). Recombinant haplotypes are generated when a crossover occurs between two linked markers (Fig. 3). Genetic recombination is essentially random, and no two sperm or eggs ever carry an identical combination of genetic information. Therefore, no two offspring ever inherit exactly the same combination of alleles, regardless of how many offspring a set of parents produce. Hence, it will always be possible to tell the offspring one from another, with the exception of monozygotic twins.
2. Recombination Frequency and Map Distance The likelihood a crossover will occur between two markers on the same chromosome is proportional to the distance between them. The probability that a gamete or offspring will be recombinant is measured by the recombination frequency (0). Genes that are unlinked segregate independently, with a recombination frequency of 0.5. In contrast, genes for which there is a consistent deviation from the 1:1 ratio of recombinant to nonrecombinant offspring are said to be genetically linked, with 0 < 0.5. Crossovers cannot be seen; they are counted by
12
34
II
12
a
i iA
1
34
a
2
a
A
b
B
I
3
4
b
c
c
1 and 4 =
c
Non
-
recombinant haplotypes
2 and 3 = Revombinant h a p l o t y p e s FIG. 3. Genetic crossovers involve one chromatid from each of a pair of homologous chromosomes. R e c o m b i n a n t haplotypes are g e n e r a t e d when a crossover occurs b et w e e n linked m a r k e r s .
THE CANINE GENOME
205
observing the recombinant haplotypes of resulting offspring. In small intervals, when the probability of multiple crossovers is negligible, recombinations may be counted directly as crossovers. Then, the relationship between the recombination fraction and the distance between two genes (x) is simply x = 0, (Morgan, 1928) and is generally true when 0 < 0.10. When many closely linked m a r k e r s are available on a chromosome and their order is known, the simplest method of determining map distances between these m a r k e r s is to estimate the recombination fractions in each interval of adjacent loci. The map distance between more distant m a r k e r s is the sum of the map distances in the intervals between these loci (Sturtevant, 1913). This method can only be applied when a large n u m b e r of m a r k e r s are available, enabling a given interval effectively to be subdivided into small intervals. Recombination fractions are not additive over large distances because of the occurrence of multiple crossovers that decrease the number of recombinant haplotypes observed and, in turn, the genetic distance. Various mapping functions have been derived to convert a recombination fraction (0) into a map distance (x). For example, the widely used Kosambi function (Kosambi, 1944) 1+20 x = 1/2 tan-i(20) = 1/4 In 1 - 20 yields a map distance of 0.236 Morgans, or 23.6 cM, for a value of 0 = 0.22.
3. Genetic Linkage and the LOD Score The process of counting numbers of recombinant and nonrecombinant offspring of a given mating followed by the estimation of map distance and determination of the likelihood of linkage versus nonlinkage is called genetic linkage analysis; this process is fundamental to the construction of genetic linkage maps. Recombination events can be detected only by observing haplotypes passed from parents to offspring so genetic linkage analysis and map construction cannot be carried out on unrelated individuals but requires observation of families. In linkage analysis between two loci, the likelihood for linkage (0 < 0.5) versus the likelihood for free recombination (0 = 0.5) is calculated based on the n u m b e r of observed recombinant and nonrecombinant offspring produced by a given mating. Conventionally, the logarithm of the likelihood ratio, the LOD score, Z(O) = log~o [L(O)/L(0.5)],
206
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
is used as the measure of support for linkage. For example, if n observations consist of k recombinants and n - k nonrecombinants, the corresponding LOD score is given by Z(0) = n log(2) + klog (q) + (n - k) log (1 - q) if 0 > 0 n log(2) if 0 = 0.
A linkage group is a set of markers in which each m a r k e r is linked to at least one other marker with a specified LOD score. Generally, linkage is accepted if a recombination fraction of 0 < 0.5 is supported by a LOD score of at least 3.0, reflecting an odds ratio of 1000:1. In the direct method of linkage analysis, recombinant and nonrecombinant offspring of a particular mating are observed directly. However, for recombinants to be distinguishable from nonrecombinants the parental and nonparental haplotypes must be known. In principal, a doubly heterozygous individual (AaBa) could have received the A allele in coupling with either the B or the b allele from one parent. These two possibilities are known as the two phases of a double heterozygote. In a two-generation family, the phase of the parents is unknown; therefore, to be able to count recombinants directly it is essential to have genotype information from three generations. The genotype of each individual is recorded and the recombination frequency of every pairwise combination of markers calculated. A parent must be heterozygous for a given pair of markers for it to be informative for linkage for those markers. When, for example, a parent with the genotype Ab / ab produces the haplotype Ab it is impossible to know if a recombination has occurred because both recombinant and nonrecombinant haplotypes can have this genotype. A mating is, therefore, only informative for linkage between a pair of given markers if at least one of the parents is heterozygous for both markers. 4. Canine Families for Mapping The resolution and integrity of a genetic map depends directly on the number of informative meioses available to be typed with each marker. Consequently, genetic maps are constructed with genotyping data from large, three-generation families. The h u m a n genetic map has been constructed with a panel of over 40 three-generation families known as the CEPH (Centre d'Etude du Polymorphisme Humain) reference families (Dausset et al., 1990). These families were chosen because both parents and most of the grandparents were alive. In addition, each of the CEPH families has a large number of offspring. Cells from all members of each family have since been immortalized. The construction of genetic maps of other animals has been possible
THE CANINE GENOME
207
because of the commitment to generate very large purpose-built pedigrees by crossing genetically diverse strains or breeds. For example, investigators at the Whitehead Institute/MIT Center for Genome Research have constructed a mouse genetic map by genotyping over four thousand polymorphic microsatellites in 46 F2 intercross progeny from a cross between two strains of mice known as the OB and CAST strains. The map provides an average spacing between markers of 0.35 cM, which corresponds to about 750 kb in the mouse (Dietrich et al., 1994). The pig genetic map is being constructed using a three-generation pedigree generated by crossing two European wild boars with eight sows of the domesticated large white breed. The subsequent intercross of 26 F~ animals has generated 200 F 2 individuals. This cross was chosen because the founder lines exhibit considerable phenotypic as well as genotypic variation (Ellegren et al., 1993), increasing the chances the pedigree would be informative for any given marker. The very high levels of inbreeding associated with most breeds of dog presents the dog mapping project with a challenge. Any single dog family is only informative for a subset of markers. A panel of several unrelated families, representing different breeds, is therefore necessary. The informativeness of this panel is further maximized by avoiding families with inbreeding loops. The distance over which linkage between a pair of markers can be confirmed, with a LOD score of at least 3.0, depends on the number of meioses available that are informative for both markers. Whereas only 10 meioses are needed to confirm that two markers are linked if there are no recombinants, data from 100 meioses are necessary to support linkage over a distance of 31 cM, illustrating the importance of large families for linkage analysis. Ultimately, the resolution and integrity of the dog genetic map, and its eventual value to breeders and the scientific community, will depend on appropriate pedigrees being made available to scientists constructing the map. Placing the markers into linkage groups and then determining the order and spacing of the markers requires genotyping each family member with every marker for which that family is informative. Data from the genotyping of each new marker must be analyzed together with that from all previous markers in the region. Although there are a variety of specifically designed computer programs that can calculate recombination frequencies and determine the likelihood of linkage, such as the LINKAGE program package (Lathrop et al., 1984), MAPMAKER (Lander and Green, 1987), and Multimap (Matise et al., 1994), the process of building a genetic map remains a large and ex-
208
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
pensive task t h a t takes the coordinated efforts of several laboratories. Ongoing efforts to share and coordinate family collection and genotyping data will greatly reduce the overall effort and cost necessary to build the map.
5. Analysis of Markers for Linkage to Disease Genes In the mapping of disease genes, one uses the same strategies just described but looks for linkage between a m a r k e r and a disease state, r a t h e r t h a n between two markers. Efforts to map canine disease genes are best illustrated by the work of Yuzbasiyan-Gurkan and collaborators at Michigan State University and their studies of canine copper toxicosis. Canine copper toxicosis is an autosomal recessive disorder that is prevelant among a subset of breeds, including Bedlington terriers, in which the gene frequency is estimated to be about 50%. Previous work by the same laboratory had established that the canine disease is similar to a h u m a n disorder called Wilson's disease, in which affected individuals suffer from progressive hepatic disease because of cooper accumulation in the liver (Yuzbasiyan-Gurkan et al., 1993). In humans, the Wilson's disease gene has been mapped to chromosome 13q and is closely linked to the esterase D and retinoblastoma genes. In dogs, these two loci are not closely linked and appear unlinked to copper toxicosis (Yuzbasiyan-Gurkan et al., 1993). In a more recent study, Yuzbasiyan-Gurkan et al. (1996) describe a microsatellite marker that yields a LOD score of 5.96 at a recombination fraction of 0, which indicates t h a t the m a r k e r is close to the copper toxicosis gene. It is important to make the distinction between cases in which genes of interest have actually been cloned, such as those described in Section V, and cases in which the actual gene remains unidentified, but a closely positioned m a r k e r has been identified. In the case of canine copper toxicosis, the gene itself remains unidentified, but the m a r k e r is sufficiently close to the gene that it is suitable for diagnostic use.
V. T a r g e t e d Cloning of Canine Genes A. GENES OF INTEREST
Simultaneous with the analysis of canine pedigrees for m a r k e r s linked to disease, efforts are ongoing to directly clone canine genes that may be important in disease. In most cases, these genes are identified by taking advantage of the extensive similarity in DNA sequence that exists in coding regions between mammals. Genes likely to be impor-
THE CANINE GENOME
209
tant in canine cancers have been of particular interest. For instance, the evolutionarily conserved part of the canine homolog of the p53 gene has been cloned and sequenced. Studies reveal that somatic mutations in p53 are important in canine thyroid cancers (Devilee et al., 1994). The canine transforming growth factor-beta-1 gene, tumor necrosis factor (TNF) alpha, and c-yes proto-oncogene have also been cloned (Manning et al., 1995; Zhao et al., 1995; Zucker et al., 1994). As with other genes, the level of homology is generally high. The canine TNFalpha gene, for instance, displays 90% sequence homology to the corresponding h u m a n TNF-alpha cDNA (Zucker et al., 1994). Finally, the canine BRCA1 homolog has been cloned and sequenced (Szabo et al., 1996). The gene is highly homologous to the h u m a n gene, which has been shown to be a major cause of familial breast cancer in humans (Hall et al., 1990; Miki et al., 1994). Its relative importance in canine breast cancer, however, remains to be tested. Several genes have been cloned in an effort to learn more about specific canine diseases. For instance, the gene for canine factor IX was cloned by Evans et al. (1989). This gene possesses 86% identity at the amino acid level with the h u m a n counterpart and is likely to be important in canine hemophilia B. In other studies, the canine homolog rodopsin gene, which is associated with several forms of h u m a n retinitis pigmentosa was cloned (Petersen-Jones et al., 1994). Because many dog breeds suffer from apparently hereditary progressive retinopathies that are similar to retinitis pigmentosum, this gene is a likely candidate for inherited canine eye diseases, although this remains to be shown. The description of a male-specific progressive retinal atrophy (PRA) syndrome in Siberian huskies by Acland et al. (1994), suggests that multiple genes are important in canine PRA, at least one of which is on the X chromosome. Analysis of several other genes has been useful for studies comparing gene functions between animals. For instance, cloning and sequencing of the canine insulin gene predicts that the canine preproinsulin molecule contains an additional C-peptide fragment of 31 amino acids that is not observed in humans. The role of this protein remains unclear, but ongoing studies are expected to provide further insight into the mechanism by which the protein works (Kwok et al., 1983). B. THE CANINE IMMUNE AND HEMATOPOIETIC SYSTEMS
Another class of genes that have been the target of several studies are those involved in the canine immune system. Differences in the canine immune system are likely to account for the differential suscep-
210
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
tibility to disease and autoimmune disorders that is noted between breeds of dog and can serve as models for some h u m a n diseases. The dog major histocompatibility locus (DLA) is among the most well-studied canine gene families. The DLA is divided into three classes of molecules: Ia, Ib, and II. Class I genes are expressed on most tissues and cells and encode m e m b r a n e glycoproteins. B u r n e t t and Geraghty (1995) have studied the structure and expression of a canine class I gene, DLA-79. The limited polymorphism, low messenger ribonucleic acid (mRNA) expression, and divergent structure of this gene suggest that it is an analog of the major histocompatibility complex (MHC) Ib genes in h u m a n s and rodents. The canine class II genes encode heterodimeric glycoproteins consisting of an alpha and a beta chain that are involved in the control of the i m m u n e response and antigen presentation. Sarmiento et al. (1990) first reported the isolation and characterization of a canine class II gene that appeared to belong to the canine DRB family. Since then, Wagner et al. (1995, 1996a, 1996b, 1996c) have reported on the polymorphism of canine DRB genes as well as isolation of DRA and DQA genes. As in other m a m m a l i a n species, these genes are highly homologous to the h u m a n and murine counterparts. In addition to work on the DLA system just described, several other canine genes have been cloned and studied in an effort to better understand the canine hematopoietic and immune system. These include hematopoietic growth factors and cytokines such as the stem cell factor, IL-8, interferon, and GM-CSF genes. The canine CD34 gene, which encodes a protein by which progenitor stem cells are recognized, has also been cloned and sequenced (McSweeney et al., 1996). Although the primary interest in these genes stems from a desire to better use the dog as a model for bone marrow transplantation, the cloning and sequencing of these genes are a useful measure of homology among the human, canine, and murine genomes.
VI. Considerations and Conclusions The identification of pedigrees suitable for genetic mapping and the generation of a genetic map is not an end in itself. The map is simply a tool to accelerate the rapid positional assignment of inherited traits as an initial step toward gene identification and characterization. AIthough a real possibility, the t r e a t m e n t of dogs affected with inherited disease is many years in the future; the ability to recognize phe-
THE CANINE GENOME
211
notypically healthy dogs carrying genetic diseases, however, is already a reality for a small number of diseases. As the map is assembled, genetic markers linked to inherited diseases will be identified and the list of diseases for which carrier detection is possible will grow very rapidly. The ability to identify dogs that carry the gene for an inherited disease is exciting. It provides the means with which to control that disease in the population by appropriate breeding programs. However, there are important issues associated with the ability to recognize carriers that must be considered by breeders, kennel clubs, and veterinarians in this new era in canine genetics. For instance, the development of a registry of dogs tested, and their carrier status, should be considered. Such a resource would play a central role in implementing breeding programs to control or eliminate diseases for which a genetic test is available, and it would provide a source of information for consumers concerned with buying a healthy dog. A related issue is that of confidentiality. For instance, should it be a requirement of kennel club registration for puppies to be tested for a disease known to be a problem for their particular breed? In addition, once a genetic test becomes available, should it be compulsory for test results to be entered in the registry? As each new genetic test becomes available, a major consideration will be deciding which breeds should be tested. Disease alleles will exist in different breeds at varying frequencies and genetic testing is expensive. Therefore, mechanisms must be established to guide breed clubs as they decide how common a disease need be before genetic testing is required. The biggest problem currently facing breeders is that dogs carrying recessive inherited disorders are usually indistinguishable from noncarriers; only when they have produced affected offspring is their carrier status recognized. The ability to identify dogs as carriers as soon as they are born means choices can be made as to whether carriers are permitted to breed. Crossing a carrier with an unaffected dog will produce approximately 50% carrier offspring but will never produce affected dogs. Breeders and veterinarians will need to seek advice from geneticists regarding the best way to eliminate each particular disease. Breeders need to decide whether merely preventing the birth of affected dogs is satisfactory; the elimination of disease alleles from a population would require that only noncarriers be crossed. Stipulations of this nature would obviously reduce the number of dogs available for breeding and may mean that dogs carrying other very desir-
212
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
able traits will not be bred. Breeders and breed clubs need to come to terms with any compromises in the breed standard that may result from this practice. Finally, it is important for these issues to be addressed now. Breeders, veterinarians, and kennel clubs must prepare for the dilemmas that genetic testing will present. Dog breeders, long the advocates for their dogs, must be willing to educate themselves so they may contribute in a positive way as these decisions are debated in the coming years. Clearly, the potential to eliminate the inherited diseases that plague so many breeds cannot be ignored, but it will be hard to turn away from short-term losses and keep focused on long-term goals. The desire to have happier, healthier, more long-lived animals must remain the long-term goal. Toward that goal, each breeder should be prepared to take full advantage of the opportunities provided by genetic testing and the development of a canine genetic map. REFERENCES Acland, G. M., Blanton, S. H., Hershfield, B., and Aguirre, G. D. (1994). XLPRA: A canine retinal degeneration inherited as an X-linked trait. Am. J. Med. Genet. 52, 27-33. Baker, T. L., Foutz, A. S., McNerney, V., Mitler, M. M., and Dement, W. C. (1982). Canine model of narcolepsy: Genetic and developmental determinants. Exp. Neurol. 75, 729742. Bishop, M. D., Kappes, S. M., Keele, J. W., Stone, R. T., Sunden, S. L. F., Hawkins, G. A., Toldo, S. S., Fries, R., Grosz, M. D., Yoo, J., and Beattie, C. W. (1994). A genetic linkage map for cattle. Genetics 136, 619-639. Botstein, D., White, R. L., Skolnick, M., and Davis, R. W. (1980). Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am. J. Hum. Genet. 32, 314-331. Burnett, R. C., and Geraghty, D. E. (1995). Structure and expression of a divergent canine class I gene. J. Immunol. 155, 4278-4285. Cork, L. C., Griffin, J. W., Munnell, J. F., Lorenz, M. D., Adams, R. J., and Price, D. I. (1979). Hereditary canine spinal muscular atrophy. J. Neuropath. Exp. Neural. 38, 209-221. Cotchin, E. (1954). Neoplasia in the dog. Vet. Rec. 66, 879-888. Cotchin, E. (1955). Malanotic tumours in the dog. J. Comp. Path. 65, 115-129. Dausset, J., Cann, H., Cohen, D., Lathrop, M., Lalouel, J. M., and White, R. (1990). Centre d'etude du polymorphisme humain (CEPH): Collaborative genetic mapping of the human genome. Genomics 6, 575-577. Deschenes, S. M., Puck, J. M., Dutra, A. S., Somberg, R. L., Felsburg, P. J., and Henthorn, P. S. (1994). Comparative mapping of canine and human proximal Xq and genetic analysis of canine X-linked severe combined immunodeficiency. Genomics 23, 62-68. Devilee, P., van Leeuwen, I. S., Voesten, A., Rutteman, G. R., Vos, J. H., and Cornelisse, C. J. (1994). The canine p53 gene is subject to somatic mutations in thyroid carcinoma. Anticancer Res. 14, 2039-2046. Dietrich, W., Katz, H., Lincoln, S., Shin, H., Friedman, J., Dracopoli, N., and Lander, E.
THE CANINE GENOME
213
(1992). A genetic map of the mouse suitable for typing intraspecific crosses. Genetics 131, 423-447. Dietrich, W. F., Miller, J. C., Steen, R. G., Merchant, M., Damron, D., Nahf, R., Gross, A., Joyce, D. C., Wessel, M., Dredge, R. D., Marquis, A., Stein, L. D., Goodman, N., Page, D. C., and Lander, E. S. (1994). A genetic map of the mouse with 4,006 simple sequence length polymorphisms. Nat. Genet. 7, 220-245. Edwards, A., Civitello, A., Hammond, H. A., and Caskey, C. T. (1991). DNA typing and genetic mapping with trimeric and tetrameric tandem repeats. Am. J. Hum. Genet. 49, 746-756. Edwards, A., Hammond, H. A., Jin, L., Caskey, C. T., and Chakraborty, R. (1992). Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups. Genomics 12, 241-253. Ellegren, H., Johansson, M., Chowdray, B., Marklund, S., Ruyter, D., Marklund, L., Brauner-Nielson, P., Edfors-Lilja, I., Gustavsson, I., Juneja, R., and Andersson, L. (1993). Assignment of 20 microsatellite markers to the porcine linkage map. Genomics 16, 431-439. Evans, J. P., Brinkhous, K. M., Brayer, G. D., Reisner, H. M., and High, K. A. (1989). Canine hemophilia B resulting from a point mutation with unusual consequences. Proc. Natl. Acad. Sci. USA 86, 10095-10099. Francisco, L. V., Langston, A. A., Mellersh, C. S., Neal, C. L., and Ostrander, E. A. (1996). A class of highly polymorphic tetranucleotide repeats for canine genetic mapping. Mammal. Genome 7, 359-362. Fries, R., Eggen, A., and Womack, J. E. (1993). The bovine genome map. Mammal. Genome 4, 405-428. Gilbert, D. A., Lehman, N., O'Brien, S. J., and Wayne, R. K. (1990). Genetic fingerprinting reflects population differentiation in the California Channel Island fox. Nature 344, 764- 767. Graphodatsky, A. S., Beklemisheva, V. R., and Dolf, G. (1995). High-resolution GTGbanding patterns of dog and silver fox: Description and comparative analysis. Cytogenet. Cell Genet. 69, 226-231. Hall, J. M., Lee, M. K., Newman, B., Morrow, J. E., Anderson, L. A., Huey, B., and King, M. C. (1990). Linkage of early-onset familial breast cancer to chromosome 17q21. Science 250, 1684-1689. Hamada, H., and Kakunaga, T. (1982). Potential Z-DNA forming sequences are highly dispersed in the human genome. Nature 298, 396-398. Henthorn, P. S., Somberg, R. L., Fimiani, V. M., Puck, J. M., Patterson, D. F., and Felsburg, P. J. (1994). IL-2R gamma gene microdeletion demonstrates that canine X-linked severe combined immunodeficiency is a homologue of the human disease. Genomics 23, 69-74. Holmes, N. G., Mellersh, C. S., Humphreys, S. J., Binns, M. M., Holliman, A., Curtis, R., and Sampson, J. (1993). Isolation and characterization of microsatellites from the canine genome. Anita. Genet. 24, 289-292. Jeffreys, A. J., and Morton, D. B. (1987). DNA fingerprints of dogs and cats. Anim. Genet. 18, 1-15. Jezyk, P. F., Felsburg, P. J., Haskins, M. E., and Patterson, D. F. (1989). X-linked severe combined immunodeficiency in the dog. Clin. Immunol. Immunopathol. 52, 173-189. Kosambi, D. D. (1944). The estimation of map distances from recombination values. Ann. Eugen. 12, 172-175. Kwiatkowski, D. J., Henske, E. P., Weimer, K., Ozelius, L., Gusella, J. F., and Haines, J. (1992). Construction of a GT polymorphism map of human 9q. Genomics 12, 229-240.
214
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
Kwok, S. C., Chan, S. J., and Steiner, D. F. (1983). Cloning and nucleotide sequence analysis of the dog insulin gene. Coded amino acid sequence of canine preproinsulin predicts an additional C-peptide fragment. J. Biol. Chem. 258, 2357-2363. Lander, E., Abrahamson, J., Barlow, A., Day, M. J., Lincoln, S. E., Newberg, L., and Green, P. (1987). MAPMAKER: An interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174-181. Langford, C. F., Fischer, P. E., Binns, M. M., Holmes, N. G., and Carter, N. P. (1996). Chromosome-specific paints from a high-resolution flow karyotype of the dog. Chrom. Res. 4, 115-123. Lathrop, G. M., Lalouel, J. M., Julier, C., and Ott, J. (1984). Strategies for multilocus linkage analysis in humans. Proc. Natl. Acad. Sci. USA 81, 3443-3446. Litt, M., and Luty, J. A. (1989). A hypervariable microsatellite revealed by in vitro amplification of a dinucleotide repeat within the cardiac muscle actin gene. Am. J. Hum. Genet. 44, 397-401. MacVean, D. W., Monlux, A. W., Anderson, P. S., Silberg, S. L., and Roszel, J. F. (1978). Frequency of canine and feline tumors in a defined population. Vet. Pathol. 15, 700715. Manning, A. M., Auchampach, J. A., Drong, R. F., and Slightom, J. L. (1995). Cloning of a canine cDNA homologous to the human transforming growth factor-beta 1-encoding gene. Gene 155, 307-308. Marcadet, A., Gebuhrer, L., Betuel, H., Seignalet, J., Freidel, A. C., Confavreux, C., Billiard, M., Dausset, J., and Cohen, D. (1985). DNA polymorphism related to HLADR2 Dw2 in patients with narcolepsy. Immunogenetics 22, 679-683. Matise, T. C., Perlin, M., and Chakravarti, A. (1994). Automated construction of genetic linkage maps using an expert system (Multimap): A human genome linkage map. Nat. Genet. 6, 384-390. McSweeney, P. A., Rouleau, K. A., Storb, R., Bolles, L., Wallace, P. M., Beauchamp, M., Krizanac-Bengez, L., Moore, P., Sale, G., Sandmaier, B., Revel, T. D., Appelbaum, F. R., and Nash, R. A. (1996). Canine CD34: Cloning of the cDNA and evaluation of an antiserum to recombinant protein. Blood 88, 1992-2003. Mellersh, C., Holmes, N., Binns, M., and Sampson, J. (1994). Dinucleotide repeat polymorphisms at four canine loci (LEI 003, LEI 007, LEI 008 and LEI 015). Anim. Genet. 25, 125-126. Meyers, K. M., Padgett, G. A., and Dickson, W. M. (1970). The genetic basis of a kinetic disorder of Scottish terrier dogs. J. Hered. 61, 189-192. Miesfeld, R., Krystal, M., and Arnheim, N. (1981). A member of a new repeated sequence family which is conserved throughout eucaryotic evolution is found between the human delta and beta globin genes. Nucleic. Acids. Res. 9, 5931-5947. Mignot, E., Wang, C., Rattazzi, C., Gaiser, C., Lovett, M., Guilleminault, C., Dement, W. C., and Grumet, F. C. (1991). Genetic linkage of autosomal recessive canine narcolepsy with a mu immunoglobulin heavy-chain switch-like segment. Proc. Natl. Acad. Sci. USA 88, 3475-3478. Miki, Y., Swensen, J., Shattuck-Eidens, D., Futreal, P. A., Harshman, K., Tavtigian, S., Liu, Q., Cochran, C., Bennett, L. M., and Ding, W. (1994). A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 266, 66-71. Morgan, T. H. (1928). "The Theory of Genes." Yale University Press, New Haven, CT. Murray, J., Buetow, K. H., Weber, J. L., Ludwigsen, S., Scherpbier-Heddema, T., Manion, F., Quiellen, J., Sheffield, V. C., Sunden, S., and Duyk, G. M. (1994). A comprehensive human linkage map with centiMorgan density. Science 265, 2049-2054.
THE CANINE GENOME
215
Noguchi, M., Yi, H., Rosenblatt, H. M., Filipovich, A. H., Adelstein, S., Modi, W. S., McBride, O. W., and Leonard, W. J. (1993). Interleukin-2 receptor gamma chain mutation results in X-linked severe combined immunodeficiency in humans. Cell 73, 147157. O'Brien, S. J., Womack, J. E., Lyons, L. A., Moore, K. J., Jenkins, N. A., and Copeland, N. G. (1993). Anchored reference loci for comparative genome mapping in mammals. Nat. Genet. 3, 103-112. Ostrander, E. A., Sprague, G. F., Jr., and Rine, J. (1993). Identification and characterization of dinucleotide repeat (CA)n markers for genetic mapping in dog. Genomics 16, 207-213. Ostrander, E. A., Jong, P. M., Rine, J., and Duyk, G. (1992). Construction of small-insert genomic DNA libraries highly enriched for microsatellite repeat sequences. Proc. Natl. Acad. Sci. USA 89, 3419-3423. Petersen-Jones, S. M., Sohal, A. K., and Sargan, D. R. (1994). Nucleotide sequence of the canine rod-opsin-encoding gene. Gene 143, 281-284. Roberts, S. R., Dellaporta, A., and Winter, F. C. (1966). The collie ectasia syndrome. Pathology of eyes of young and adult dogs. Am. J. Ophthalmol. 62, 728-752. Sack, G. H., Cork, L. C., Morris, J. M., Griffin, J. W., and Price, D. L. (1984). Autosomal dominant inheritance of hereditary canine spinal muscular atrophy. Ann. Neurol. 15, 369-373. Sarmiento, U. M., Sarmiento, J. I., and Storb, R. (1990). Allelic variation in the DR subregion of the canine major histocompatibility complex. Immunogenetics 32, 13-19. Smythe, R. H. (1945). Recurrent tetany in the dog. Vet. Rec. 57, 380. Somberg, R. L., Pullen, R. P., Casal, M. L., Patterson, D. F., Felsburg, P. J., and Henthorn, P. S. (1995). A single nucleotide insertion in the canine interleukin-2 receptor gamma chain results in X-linked severe combined immunodeficiency disease. Veterinary Immunology and Immunopathology 47, 203-213. Stallings, R. L., Ford, A. F., Nelson, D., Torney, D. C., Hildebrand, C. E., and Moyzis, R. K. (1991). Evolution and distribution of (GT)n repetitive sequences in mammalian genomes. Genomics 10, 807-815. Stone, D. M., Jacky, P. B., and Prieur, D. J. (1991). The Giemsa banding pattern of canine chromosomes, using a cell synchronization technique. Genome 34, 407-412. Sturtevant, A. H. (1913). The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. J. Exp. Zool. 14, 43-59. Tautz, D., and Renz, M. (1984). Simple sequences are ubiquitous repetitive components of eukaryotic genomes. Nucleic Acids Res. 12, 4127-4137. Toldo, S. S., Fries, R., Steffen, P., Neibergs, H. L., Barendse, W., Womack, J. E., Hetzel, D. J., and Stranzinger, G. (1993). Physically mapped, cosmid-derived microsatellite markers as anchor loci on bovine chromosomes. Mammal. Genome 4, 720-727. Wagner, J. L., DeRose, S. A., Burnett, R. C., and Storb, R. (1995). Nucleotide sequence and polymorphism analysis of canine DRA cDNA clones. Tissue Antigens 45, 284287. Wagner, J. L., Burnett, R. C., DeRose, S. A., and Storb, R. (1996a). Molecular analysis and polymorphism of the DLA-DQA gene. Tissue Antigens 48, 199-204. Wagner, J. L., Burnett, R. C., and Storb, R. (1996b). Molecular analysis of the DLA DR region. Tissue Antigens 48, 549-553. Wagner, J. L., Burnett, R. C., Works, J. D., and Storb, R. (1996c). Molecular analysis of DLA-DRBB1 polymorphism. Tissue Antigens 48, 554-561. Weber, J. L. (1990). Informativeness of human (dC-dA)n (dG-dT)n polymorphisms. Genomics 7, 388-396.
216
CATHRYN S. MELLERSH AND ELAINE A. OSTRANDER
Weber, J. L., and May, P. E. (1989). Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 44, 388-396. Weissenbach, J., Gyapay, G., Dib, C., Vignal, A., Morisette, J., Millasseau, P., Vaysseix, G., and Lathrop, M. (1992). A second-generation linkage map of the human genome. Nature 359, 794-801. Willis, M. B. (1989). "Genetics of the Dog." Macmillan Publishing Company, New York, New York. Woods, C. B. (1977). Hyperkinetic episodes in two Dalmatian dogs. J. Am. Anim. Hosp. Ass. 13, 255-257. Yuzbasiyan-Gurkan, V., Blanton, S. H., Cao, Y., Ferguson, P., Li, J., Venta, P. J., and Brewer, G. J. (1996). Linkage of a microsatellite marker to the canine copper toxicosis locus in the Bedlington terrier. Am. J. Vet. Res. 58, 23-27. Yuzbasiyan-Gurkan, V., Wagnitz, S., Blanton, S. H., and Brewer, G. J. (1993). Linkage studies of the esterase D and retinoblastoma genes to canine cooper toxicosis: A model for Wilson disease. Genomics 15, 86-90. Zhao, D. M., Tateyama, S., Miyoshi, N., Uchida, K., Yamaguchi, R., Yamagami, T., and Hayashi, T. (1995). Sequence of the canine c-yes proto-oncogene. 59, 230-233. Zucker, K., Lu, P., Fuller, L., Asthana, D., Esquenazi, V., and Miller, J. (1994). Cloning and expression of the cDNA for canine tumor necrosis factor-alpha in E. coli. Lymphokine and Cytokine Research 13, 191-196.
Index
Acquired immunodeficiency syndrome human immunodeficiency virus, s e e Human immunodeficiency virus retroviral virulence influence, 144-146 feline, 144-145, 150 murine, 140, 145, 147 simian, 145-146 Adeno-associated virus, hemophilia gene therapy, 125-126, 131 Adenomatous polyposis, tumor-suppressor gene function, 82 Adenosine deaminase deficiency, gene therapy, 34 Adenovirus, gene therapy hemophilia, 124-125, 131 human cancers, 89 Amplification, oncogenesis detection, 64 Arthritis encephalitis virus, nonmalignant disease association, 152 Atherosclerosis, susceptibility, 29-30
Biotechnology, ethics, 39-41 Blot hybridization, comparative genome mapping, 170 Bovine comparative genome mapping, 169175, 177-181 economic traits, 182 leukemia virus, 136, 140-141, 147 BRCA1 gene cancer predisposition screening, 86 functional identity, 81-82
Cancer domestic animals canine genes, 209 genetic changes, 32-33 human gene therapy, 87-92 oncogenesis, s e e Oncogenesis predisposition screening, 86-87 Canine genome map, 191-212 applications, 193-196 breed diseases, 193-196 human genetic disease models, 193-195 future research directions, 210-212 map development, 198-208 canine families, 206-208 construction approach, 198-199 genetic linkage, 205-206, 208 logarithm of likelihood score, 205206 map distance, 204-205 marker assignment, 198, 201-202 marker ordering, 202-208 microsatellite repeats, 199-201 recombination frequency, 204-205 organization chromosomes, 196-197 repeated elements, 197 overview, 191-192 targeted cloning, 208-210 canine immune system, 209-210 genes of interest, 208-209 hematopoietic system, 209-210 molecular genetic disease diagnosis future research directions, 114-115 217
218
INDEX
Canine ( c o n t . ) linkage analysis copper toxicosis, 208 gene candidates, 109 map development, 205-206, 208 positional cloning, 108-109 quantitative trait loci, 110 methods, 104-107 chromosome assignment, 106 i n s i t u hybridization, 106-107 polymerase chain reaction, 105106 restriction fragment length polymorphism, 104-105, 113 restriction mapping, 106-107 overview, 103-104 single-gene disorders, 110-114 genetic homogeneity, 111-113 mutations, 111 technical considerations analytic methods, 113 data management, 114 Caprine arthritis encephalitis virus, virulence influence, 152 Cattle, see Bovine cDNA cloning disease mutation diagnosis, 30-31 historical perspective, 14-15 DNA sequencing, 15-17 probes, see DNA probes somatic gene therapy, 33-35 Chromosomes canine disease diagnosis, 106 genome organization, 196-197 evolution study, 158-159 mapping, see Gene mapping oncogenesis detection, 61-66 amplification, 64 chromosomal deletion, 64-66 heterozygosity loss, 64-66 translocation, 63-64 Citrullinemia, diagnosis, 30 Cloning canine genome disease linkage analysis, 108-109 targeted cloning, 208-210 canine immune system, 209-210
genes of interest, 208-209 hematopoietic system, 209-210 disease gene detection canine linkage analysis, 108-109 historical perspective, 27, 30-31 gene manipulation, 7, 14-15 genome mapping rational, 161 oncogenesis detection, 67-69 positional cloning, see Positional cloning c - m e t oncogene, metastasis association, 85-86 Comparative genome hybridization blot hybridization, 170 domestic animal genome mapping, 167-182 bovine, 169-175, 177-181 current genome map status, 169-182 hamster, 170-171 mouse, 169-171, 173, 175 sheep, 181 electrophoresis, 169 human genome map, 167-170, 173181 methods, 167-168 microsatellites, 181 oncogenesis detection, 66-67 restriction fragment length polymorphism, 171 Southern blot, 170-171 Copper toxicosis, canine linkage analysis, 208 Current trends, see Historical perspectives Cystic fibrosis, transmembrane conductance regulator gene therapy, 34
Deletion mutations historical perspective, 23 oncogenesis detection, 64-66 Deoxynucleotidyltransferase, historical perspective, 11-12 Disease canines genome models, 193-195 inbreeding role, 192-193, 211
INDEX molecular genetic diagnosis analytic methods, 113 chromosome assignment, 106 data management, 114 future research directions, 114115 gene candidates, 109 genetic homogeneity, 111-113 in s i t u hybridization, 106-107 linkage analysis, 107-110, 208 methods, 104-107 mutations, 111 overview, 103-104 polymerase chain reaction, 105106 positional cloning, 108-109 quantitative trait loci, 110 restriction fragment length polymorphism, 104-105, 113 restriction mapping, 106-107 single-gene disorders, 110-114 diagnosis, historical perspective, 22-30 cloning, 27, 30-31 deletion mutations, 23 disease susceptibility, 28-30 gene identification, 27 genetic screening, 27 insertion mutations, 23 restriction fragment length polymorphism, 24-26 restriction site mutations, 22-23 single-base replacement, 23-24 nonmalignant forms, retroviral virulence influence, 135-152 acquired immunodeficiency syndrome, 144-146 caprine arthritis encephalitis virus, 152 endogenous retroviruses, 148-152 equine infectious anemia virus, 149150 feline immunodeficiency virus, 144145, 150 HTLV- 1-associated myelopathy, 147 human immunodeficiency virus, 140-142, 149 mouse mammary tumor virus, 140, 145, 148 murine leukemia virus, 146-147
219
neurologic disease, 145-146 osteopetrosis, 147 overview, 135-136 progressive pneumonia virus, 150-152 reticuloendotheliosis virus, 147-148 retroviral groups, 140-142 spleen focus-forming virus, 146 virus biology, 136-140 resistance, domestic animal genetic improvement methods, 37-38 DNA canine chromosome paints, 197 cDNA cloning disease mutation diagnosis, 30-31 gene manipulation, 14-15 DNA sequencing, 15-17 probes, see DNA probes somatic gene therapy, 33-35 historical perspective adaptors, 12 genetic disease diagnosis, see DNA probes recombinant animal improvement methods gene maps, 35-36 genetic markers, 35-36 parentage identification, 36 transgenic animals, 36-39 replication overview, 2-5 polymerase chain reaction, 18 sequencing, 15-17 single-base replacement, 23-24 oncogenesis detection DNA-mediated gene transfer, 54-55 genetic mapping, 67-68 DNA ligase, gene manipulation, 11 DNA probes, genetic disease diagnosis, 22-30 deletion mutations, 23 disease susceptibility, 28-30 gene identification, 27 genetic screening, 27 insertion mutations, 23 restriction fragment length polymorphism, 24-26 restriction site mutations, 22-23 single-base replacement, 23-24
220
INDEX
DNA repair genes, functional identity, 83-84 DNA tumor viruses, oncogenesis detection, 60-61 Dog, see Canine Domestic animals, see also specific types genetic improvement methods, historical perspective, 35-39 gene maps, 35-36 genetic markers, 35-36 parentage identification, 36 transgenic animals, 36-39 disease resistance, 37-38 embryonic stem cell use, 39 growth regulation, 20-21, 37 milk modification, 38 pharmaceutical proteins, 38 wool production, 38 genome mapping current genome map status, 168-183 comparative mapping, 169-182 economic traits, 182 linkage mapping, 169 physical mapping, 168-169 future research directions, 183-184 historical perspective, 157-158 methods, 161-168 comparative mapping, 167-168 in situ hybridization, 164-165 linkage mapping, 165-167 synteny mapping, 161-164 rational, 158-161 chromosomal evolution study, 158-159 marker-assisted selection, 159160, 182 positional cloning, 161 Duchenne muscular dystrophy, gene therapy, 34
Embryonic stem cells domestic animal genetic improvement methods, 39 oncogene function study, gene knockouts, 71-73 Encephalitis virus, virulence influence, 152 Endonucleases, gene manipulation, 1011 Epstein-Barr virus, oncogenesis detection, 60-61 Equine infectious anemia virus, virulence influence, 149-150 Ethics, biotechnical manipulation, 39-41 Exons, 15
Factor IX gene canine disease, 209 hemophilia gene therapy, 119-120, 126-129 Factor VIII gene, hemophilia gene therapy, 119-120, 129-130 Familial adenomatous polyposis, tumorsuppressor gene function, 82 Familial hypercholesterolemia, gene therapy, 34 Feline immunodeficiency virus, virulence influence, 144-145, 150 Fluorescence in situ hybridization domestic animal genome mapping, 165, 168, 180-182 oncogenesis detection, 66-67 G Gene cloning, see Cloning Gene manipulation, see also specific m a nipulations
Economic trait loci, domestic animal improvement future research directions, 183-184 marker-assisted selection, 159-161, 182 Electrophoresis, comparative genome mapping, 169
historical perspective, 7-18 methods, 8-12 adaptors, 12 DNA ligase, 11 nucleic acid hybridization, 8-10 restriction endonucleases, 10-11 terminal deoxynucleotidyltransferase, 11-12
221
INDEX overview, 7 structure studies, 14-18 cDNA cloning, 14-15 DNA sequencing, 15-17 eukaryotic gene splitting, 15 gene cloning, 15 polymerase chain reaction, 18 restriction maps, 15-17 Southern blot, 15-17 vectors, 7, 12-14 Gene mapping canine genome applications breed diseases, 193-196 human genetic disease models, 193-195 disease diagnosis, 106-107 future research directions, 210-212 map development, 198-208 canine families, 206-208 construction approach, 198-199 genetic linkage, 205-206, 208 logarithm of likelihood score, 205206 map distance, 204-205 marker assignment, 198, 201-202 marker ordering, 202-208 microsatellite repeats, 199-201 recombination frequency, 204-205 organization chromosomes, 196-197 repeated elements, 197 overview, 191-192 targeted cloning, 208-210 canine immune system, 209-210 genes of interest, 208-209 hematopoietic system, 209-210 current genome map status, 168-183 comparative mapping, 169-182 economic traits, 182 linkage mapping, 169 physical mapping, 168-169 future research directions, 183-184 historical perspective, 157-158 domestic animal genetic improvement methods, 35-36 gene manipulation, 15-17 methods, 161-168 comparative mapping, 167-168 in situ hybridization, 164-165
linkage mapping, 165-167 synteny mapping, 161-164 oncogenesis detection genetic mapping, 67-68 linkage analysis, 67-68 physical mapping, 66-67 rational, 158-161 chromosomal evolution study, 158159 marker-assisted selection, 159-160 positional cloning, 161 Gene markers, see Genetic markers Gene regulation cystic fibrosis transmembrane conductance regulator gene, 34 historical perspective tissue culture studies, 19 transgenic animals, 20-21, 37 Gene replication historical perspective, 2-5 polymerase chain reaction, see Polymerase chain reaction retrovirus RNA, 137-142 Gene splitting, historical perspective, 15 Gene therapy hemophilia, 119-131 clinical manifestations, 119-120 IX factor, 119-120, 126-129 VIII factor, 119-120, 129-130 gene biology, 120 model advantages, 121-122 treatment advantages, 120-121 vector choice adeno-associated viral vectors, 125-126, 131 adenoviral vectors, 124-125, 131 retroviral vectors, 123, 131 historical perspective, 33-35 germ line, 33 somatic genes, 33-35 human cancer research, 87-92 Genetic disease diagnosis, see also specific dis e as e s
historical perspective, 21-33 disease susceptibility atherosclerosis, 29-30 diagnosis, 31-32 HLA genes, 28-29 insulin gene 5'-VNTR alleles, 29
222
INDEX
Genetic disease diagnosis ( c o n t . ) DNA probes, 22-30 deletion mutations, 23 disease susceptibility, 28-30 gene identification, 27 genetic screening, 27 insertion mutations, 23 restriction fragment length polymorphism, 24-26 restriction site mutations, 22-23 single-base replacement, 23-24 domestic animal applications, 30-32 cancer, 32-33 disease mutation diagnosis, 30-31 disease susceptibility diagnosis, 31-32 gene changes, 21-22 Genetic markers canine genome disease diagnosis, 106 map development assignment, 198, 201-202 microsatellite repeats, 199-201 ordering, 202-208 economic trait loci selection, 159-161, 182 genome mapping rational, 159-160 historical perspective, 35-36 linkage mapping, see Linkage analysis quantitative trait loci selection, 35-36, 159-160 Genetic screening cancer predisposition, 86-87 historical perspective, 27 Genome mapping, see Gene mapping Growth hormone canine disease, 209 germ line gene therapy, 33 oncogene function identity, 74-75 regulation, historical perspective, 37
H
Hamster, comparative genome mapping, 170-171 Hematopoietic system, canine targeted cloning, 209-210 Hemophilia, gene therapy, 119-131 clinical manifestations, 119-120
IX factor, 119-120, 126-129 VIII factor, 119-120 gene biology, 120 model advantages, 121-122 treatment advantages, 120-121 vector choice adeno-associated viral vectors, 125126, 131 adenoviral vectors, 124-125, 131 retroviral vectors, 123, 131 Herpes simplex thymidine kinase gene, characterization, 19 Historical perspectives biotechnology ethics, 39-41 domestic animal genetic improvement methods, 35-39 gene maps, 35-36 genetic markers, 35-36 parentage identification, 36 transgenic animals, 36-39 disease resistance, 37-38 embryonic stem cell use, 39 growth regulation, 37 milk modification, 38 pharmaceutical proteins, 38 wool production, 38 gene manipulation, 7-18 methods, 8-12 adaptors, 12 DNA ligase, 11 nucleic acid hybridization, 8-10 restriction endonucleases, 10-11 terminal deoxynucleotidyltransferase, 11-12 overview, 7 structure studies, 14-18 cDNA cloning, 14-15 DNA sequencing, 15-17 eukaryotic gene splitting, 15 gene cloning, 15 polymerase chain reaction, 18 restriction maps, 15-17 Southern blot, 15-17 vectors, 7, 12-14 gene regulation tissue culture studies, 19 transgenic animals, 20-21 gene therapy, 33-35 germ line, 33 somatic genes, 33-35
223
INDEX genetic disease diagnosis, 21-33 disease susceptibility atherosclerosis, 29-30 diagnosis, 31-32 HLA genes, 28-29 insulin gene 5'-VNTR alleles, 29 DNA probes, 22-30 deletion mutations, 23 disease susceptibility, 28-30 gene identification, 27 genetic screening, 27 insertion mutations, 23 restriction fragment length polymorphism, 24-26 restriction site mutations, 22-23 single-base replacement, 23-24 domestic animal applications, 30-32 cancer, 32-33 disease mutation diagnosis, 30-31 disease susceptibility diagnosis, 31-32 hereditary disease, 21-22 genome mapping, domestic animals, 157-158 metabolic regulation, 18-21 molecular genetics, 2-7 gene replication, 2-5 gene structure, 2-5 transcription, 5-7 patent law, 39-41 HLA genes disease susceptibility, 28-29 gene therapy, 91-92 Human genome map cancer therapy, 87-92 canine models, 193-195 comparative mapping, 167-170, 173181 synteny mapping, 162-164 Human immunodeficiency virus, see also Acquired immunodeficiency syndrome hemophilia treatment, 121 retroviral virulence influence, 140, 142, 149 Human narcolepsy, canine models, 194 Human papillomaviruses, oncogenesis detection, 61 Human spinal muscular atrophy, canine models, 193
Human T-cell leukemia virus oncogenesis detection, 59-60 virulence influence, 136, 140, 142, 146 Hypercholesterolemia, gene therapy, 34 Hyperthermia, diagnosis, 30-31, 35
Immune system, canine targeted cloning, 209-210 Immunodeficiency syndromes, see specific types
Immunoglobulin A, hemophilia treatment, 131 Insertion mutations historical perspective, 23 oncogenesis detection, 55-58 In situ hybridization canine disease diagnosis, 106-107 domestic animal genome mapping, 164-165, 168, 172-178 oncogenesis detection, 67 Insulin-dependent diabetes mellitus HLA gene, 28-29 5 ' - V N T R gene, 28-29 Insulin gene 5'-VNTR, disease susceptibility, 29 Interleukin 2, hemophilia treatment, 131 Intestinal neoplasias, tumor-suppressor gene function, 83 Introns, 15
Knockout genes, oncogene function study, 71-73
Law, patents, 39-41 Leukemia virus oncogenesis detection, 59-60 virulence influence, 136, 140, 142, 145-147 Lineage identification canine family gene map development, 206-208 historical perspective, 24-25, 36
224
INDEX
Linkage analysis domestic animal genome mapping, 165-167, 169 genetic disease diagnosis canine gene candidates, 109 genome map development, 205206, 208 overview, 107-108 positional cloning, 108-109 quantitative trait loci, 110 historical perspective, 24-25 oncogenesis detection, 67-68 Logarithm of likelihood score, canine genome, 205-206 Lymphocytes, tumor infiltrating lymphocytes, gene therapy, 92
M
Maedi, retroviral virulence influence, 150-152 Major histocompatibility complex canine disease, 210 disease susceptibility role, 31-32 Malignant hyperthermia, diagnosis, 3031, 35 Mammary tumor virus, virulence influence, 148 Mapping, s e e Gene mapping Markers, s e e Genetic markers Metabolism, regulation, 18-21 Metastasis-associated genes, functional identity, 84-86 Microsatellites canine gene repeats, 199-201 comparative genome mapping, 181 historical perspective, 26 linkage mapping, 166, 199-201 Milk production economic traits, 182 genetic improvement methods, 38 Minisatellites canine genome repeated elements, 197 genetic disease diagnosis, restriction fragment length polymorphism, 26 Molecular genetics canine disease diagnosis, 103-115 future research directions, 114-115 linkage analysis, 107-110
gene candidates, 109 positional cloning, 108-109 quantitative trait loci, 110 methods, 104-107 chromosome assignment, 106 i n s i t u hybridization, 106-107 polymerase chain reaction, 105106 restriction fragment length polymorphism, 104-105, 113 restriction mapping, 106-107 overview, 103-104 single-gene disorders, 110-114 genetic homogeneity, 111-113 mutations, 111 technical considerations analytic methods, 113 data management, 114 historical perspective, 2-7 gene replication, 2-5 gene structure, 2-5 transcription, 5-7 Monkey, acquired immunodeficiency syndrome, 145-146 Mouse comparative genome mapping, 169171, 173, 175 mammary tumor virus, 140, 145, 147 Multiple intestinal neoplasias, tumorsuppressor gene function, 83 Murine leukemia virus, virulence influence, 145-147 Muscular atrophy, canine models, 193 Muscular dystrophy, gene therapy, 34 Mutations canine single-gene disorders, 111 deletion mutations historical perspective, 23 oncogenesis detection, 64-66 domestic animal disease diagnosis, 23, 30-31 insertion mutations, oncogenesis detection, 55-58 restriction site mutations, 22-23 Myelopathy, retroviral virulence influence, 147
Narcolepsy, canine models, 194
INDEX Neoplasia, tumor-suppressor gene function, 83 Neurofibromatosis, tumor-suppressor gene function, 82 Neurologic disease, retroviral virulence influence, 146-147 Nonmalignant disease, see Disease, nonmalignant forms Nucleic acid hybridization, 8-10
Oncogenesis, 51-93 animal model systems, 69-73 gene knockouts, 71-73 transgenic mouse, 70-71 cloning, 66-69 gene detection, 53-66 chromosomal rearrangement, 61-66 amplification, 64 chromosomal deletion, 64-66 heterozygosity loss, 64-66 translocation, 63-64 DNA-mediated gene transfer, 54-55 DNA tumor viruses, 60-61 retroviral-mediated oncogenesis, 5560 acutely transforming retroviruses, 58-59 human retroviruses, 59-60 oncogene transduction, 58-59 proto-oncogene activation, 55-58 proviral insertion, 55-58 gene functional identities, 73-86 DNA repair genes, 83-84 metastasis-associated genes, 84-86 oncogenes, 73-77 proto-oncogenes, 73-77 tumor-suppressor genes, 77-83 BRCA1 gene, 81-82, 86 cellular junction cytoskeletal components, 82-83 p53 gene, 79-81 RB1 gene, 77-79 WT-1 gene, 81 mapping techniques, 66-69 genetic mapping, 67-68 linkage analysis, 67-68 physical mapping, 66-67 overview, 52-53
225
research applications, 86-93 cancer predisposition screening, 8687 human gene therapy, 87-92 veterinary medicine, 92-93 Osteopetrosis, retroviral virulence influence, 146
Papillomaviruses, oncogenesis detection, 61 Papovaviruses, oncogenesis detection, 60-61 Parentage identification canine family gene map development, 206-2O8 historical perspective, 24-25, 36 Patent law, historical perspective, 39-41 p53 gene, functional identity, 79-81 Pharmaceutical proteins, genetic improvement methods, 38 Phosphoenolpyruvate carboxykinase gene characterization, 19 regulation, 21 Pig, genetic map development, 207 Pneumonia virus, virulence influence, 150-152 Polymerase chain reaction canine chromosome paints, 197 disease diagnosis, 105-106, 113114 historical perspective, 18 synteny mapping, 163 Polyoma virus, oncogenesis detection, 60-61 Polyposis, tumor-suppressor gene function, 82 Porcine, genetic map development, 207 Positional cloning canine linkage analysis, 108-109 domestic animal gene mapping rational, 161 historical perspective, 27 oncogenesis detection, 68-69 Progressive pneumonia virus, virulence influence, 150-152 Protein kinase C, oncogene signal transduction, 76
226
INDEX
Proto-oncogenes, see a l s o Oncogenes function identity, 73-77 retroviral-mediated oncogenesis, 55-58
Q Quantitative trait loci canine disease linkage analysis, 110 domestic animal improvement, markerassisted selection, 35-36, 159-160
R
gene, functional identity, 77-79 Representational difference analysis, oncogenesis detection, 69 Restriction endonucleases, gene manipulation, 10-11 Restriction fragment length polymorphism canine disease diagnosis, 104-105, 113 comparative genome mapping, 171 genetic disease diagnosis, historical perspective, 24-26 family linkage studies, 24-25 microsatellites, 26 minisatellites, 26 multiple polymorphisms, 25 search methods, 25-26 linkage mapping, 166 oncogenesis detection cancer predisposition screening, 87 genetic mapping, 67-69 Restriction mapping, see Gene mapping Restriction site mutations, genetic disease diagnosis, 22-23 Reticuloendotheliosis virus, virulence influence, 147-148 Retrovirus gene therapy hemophilia, 123, 131 human cancers, 88-89, 91 nonmalignant disease virulence influence, 135-152 acquired immunodeficiency syndrome feline, 144-145, 150 murine, 140, 145, 147 simian, 145-146
RB1
biology, 136-140 endogenous retroviruses, 148-152 caprine arthritis encephalitis virus, 152 equine infectious anemia virus, 149-150 human immunodeficiency virus, 149 mouse mammary tumor virus, 140, 145, 148 progressive pneumonia virus, 150152 murine leukemia virus, 146-148 HTLV-l-associated myelopathy, 147 neurologic disease, 146-147 osteopetrosis, 147 reticuloendotheliosis virus, 147-148 spleen focus-forming virus, 146 overview, 135-136 retroviral groups, 140-142 oncogenesis detection, 55-60 acutely transforming retroviruses, 58-59 human retroviruses, 59-60 oncogene transduction, 58-59 proto-oncogene activation, 55-58 proviral insertion, 55-58 RNA historical perspective cDNA cloning, 14-15 transcription, 5-7 retrovirus replication, 137-142
Severe combined immunodeficiency syndrome, canine models, 194 Sheep comparative genome mapping, 181 genetic improvement methods economic traits, 182 wool production, 38 Signal transduction, oncogene detection, 58-59, 73-77 Single-base replacement, disease diagnosis genetic homogeneity, 111-113 historical perspective, 23-24 mutations, 111
227
INDEX Somatic cell hybridization gene therapy, 33-35 oncogenesis detection, 66-67 Southern blot comparative genome mapping, 170171 historical perspective, 15-17 Spinal muscular atrophy, canine models, 193-194 Spleen focus-forming virus, virulence influence, 146 SV40 virus, oncogenesis detection, 60-61 Synteny mapping, see Gene mapping
Tumor infiltrating lymphocytes, gene therapy, 92 Tumor necrosis factor-a, canine disease, 209 Tumors, see Cancer; Oncogenesis Tumor-suppressor genes, functional identities, 77-83 B R C A 1 gene, 81-82, 86 cellular junction cytoskeletal components, 82-83 p 5 3 gene, 79-81 RB1 gene, 77-79 WT-1 gene, 81 Tyrosine kinase growth factor receptors, oncogene function identity, 75-76
Tag enzyme, polymerase chain reaction,
V
18 T-cell leukemia virus oncogenesis detection, 59-60 virulence influence, 136, 140-141, 147 Terminal deoxynucleotidyltransferase, historical perspective, 11-12 Thymidine kinase gene, characterization, 19 Transcription, historical perspective, 5-7 Transduction, oncogene detection, 58-59, 73-77 Transgenic animals historical perspective domestic animal genetic improvement methods, 36-39 disease resistance, 37-38 embryonic stem cell use, 39 growth regulation, 37 milk modification, 38 pharmaceutical proteins, 38 wool production, 38 gene regulation, 20-21 oncogene function study, 70-71 Translocation, oncogenesis detection, 6364 Trends, see Historical perspectives
Vectors, see also specific types hemophilia therapy, 123-126, 131 historical perspective, 7, 12-14 Visna, retroviral virulence influence, 150-152
W
Wilson's disease, canine linkage analysis, 208 WT-1 gene, functional identity, 81
X-linked severe combined immunodeficiency syndrome, canine models, 194
Yeast artificial chromosomes, oncogenesis detection, cloning, 69
This Page Intentionally Left Blank