Advances in Cancer Research, Volume 24

ADVANCES IN CANCER RESEARCH VOLUME 24 Contributors to This Volume Ann M. Ainsworth Beverly E. Griffin Claudio Basi...

Author: George Klein | Sidney Weinhouse

7 downloads 991 Views 20MB Size Report

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

DOWNLOAD PDF

ADVANCES IN CANCER RESEARCH VOLUME 24

Contributors to This Volume

Ann M. Ainsworth

Beverly E. Griffin

Claudio Basilico

J. C. Leclerc

Robert E. Bellet

J. P. levy

David Berd

Joachim Mark

Evelina A. Bernardino

Michael J. Mastrangelo

Wallace H. Clark, Jr.

Lars Ostberg

Mike Fried

Per A. Peterson Lars Rask

ADVANCES IN CANCER RESEARCH Edited by

GEORGE KLElN Department of Tumor Biology Karolinska lnstitutet Stockholm, Sweden

SIDNEY WEINHOUSE Fels Research Institute Temple University Medical School Philadelphia, Pennsylvania

VoIume 24 - 7977

cip,

ACADEMIC PRESS

New York

A Subsidiary of Harcourt Brace Jovanovich, Publishers

San Francisco

London

COPYRIGHT 0 1977, BY ACADEMIC PRESS,INC. ALL RIGHTS RESERVED. NO PART O F THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. 24/28 Oval Road, London NWI

LIBRARY O F CONGRESS CATALOG CARD NUMBER:52-13360

ISBN 0-12-006624-6 PRINTED IN THE UNITED STATES O F AMERICA

CONTENTS CONTRIBUTORS TO VOLUME 24 . . . OBITUARY-SIR ALEXANDERHADDOW .

.

.

ix xi

I. Introduction . . . . . . . . . . . . . I1. In V i m Studies of the Immunological Rejection of MSV-Induced Sarcomas . . . . . . . . . . . . . I11. Analytical Study of the Antivirus Immune Response . . . . IV . Analytical Study of the Antitumor Cell Reaction: Humoral Response . V. Analytical Study of the Antitumor Response: Cell-Mediated Immunity . . . . . . . . . . . . . (CMI) . VI . MSV Tumors and Immune Surveillance . . . . . . . VII. General Comments . . . . . . . . . . . References . . . . . . . . . . . . .

2

.

.

.

.

.

.

.

.

.

.

.

.

.

.

The Murine Sarcoma Virus-Induced Tumor: Exception or General Model in Tumor Immunology?

J . P . LEVYAND J . C . LECLERC

Organization of the Genomes

3 21 26 32 50 52 58

of Polyoma Virus and SV40

MIKE FRIED AND BEVERLY E . GRIFFIN

I. Introduction . . . . . . . . . . . . . I1. Action of Enzymes on the Viral DNAs and Construction of Physical Maps . . . . . . . . . . . . . . I11. Primary Sequence Studies . . . . . . . . . . IV. Origin and Termination of Viral DNA Replication . . . . . V. Location of Virus-Specific RNAs . . . . . . . . . VI . Virus-Induced Proteins . . . . . . . . . . . VII. Protein Binding Sites on Viral DNAs . . . . . . . . VIII . Genetic Mapping . . . . . . . . . . . . IX . Essential and Nonessential Regions of the Viral Genomes . . . . X . Defective Viral DNAs . . . . . . . . . . . XI . Comparison of the Polyoma Virus and SV40 Genomes . . . . XI1. Conclusion . . . . . . . . . . . . . References . . . . . . . . . . . . . V

67 69 76 81 83 85 88 89 94 99 102 105 107

vi

CONTENTS

P.. Microglobulin

and the Major Histocompatibility Complex

PER A . PETERSON.LARS RASK.AND LARS OSTBERG

I. Introduction . . . . . . . . I1. Genetics of the Major Histocompatibility Complex

.

.

.

.

.

.

.

.

.

.

. .

. .

. .

. 111. Traits Associated with the Major Histocompatibility Complex . IV . Isolation and Characteristics of 3/., Microglobulin . . . V. Features of the Classical Transplantation Antigens . . . VI . Biochemical Properties of the Thymus-Leukemia Antigens .

. T-Locus

VII VIII . IX . X.

Gene Products and &-Microglobulin . I-Region Defined Antigens and the Fc Receptor The S-Region and the Complement System . Conclusions and Speculations . . . . References . . . . . . . .

.

.

.

.

.

.

.

.

. . . . .

.

. .

.

. . . . .

.

. .

.

115 117 120 127 132 142 144 146 152 155 157

Chromosomal Abnormalities and Their Specificity in Human Neoplasms: An Assessment of Recent Observations by Banding Techniques

JOACHIMMARK

I. Introduction . . . . I1. Meningiomas . . . . 111. Myeloproliferative Disorders . IV . Lymphoproliferative Disorders V . Concluding Remarks . . References . . . . Addendum . . . .

. . . . . . . . . . . . .

. . . . . .

Temperature-Sensitive h..Jtations

. . . . . .

. . . . . .

. . . . . .

. .

. . . .

165 166 174 196 212 215 222

. . . . . . .

223 225 253 261 262 264 266

.

267

. . . . . . . . . . . . . . . . . . Animal Cells

I

CLAUDIOBASILICO I. Introduction . . . . . . . . . . . I1. ts Growth Mutants . . . . . . . . . 111. ts Mutations Affecting the Expression of Specialized Functions IV . Nature of the ts Mutations Described in Animal Cells . . V . Conclusions . . . . . . . . . . . References . . . . . . . . . . . Note Added in Proof . . . . . . . . .

. . . . . . .

Current Concepts of . the Biology of Human Cutaneous Malignant Melanoma

WALLACEH . CLARK,JR., MICHAELJ . MASTRANGELO. ANN M . AINSWORTH. DAVIDBERD. ROBERTE . BELLET. AND EVELINAA . BERNARDINO

I. Introduction

.

.

.

.

.

.

.

.

.

.

.

.

vii

CONTENTS

. .

I1 Inductive Circumstances . . . . . . . . . . I11 The Developmental Biology of the Primary Lesions of Malignant . . . . . . . . . . . . . Melanoma IV. The Immunobiology of Malignant Melanoma . . . . . . V . Fine-Structural Studies . . . . . . . . . . . VI Summary and Conclusions . . . . . . . . . . References . . . . . . . . . . . . .

.

SUBJECT INDEX .

.

.

.

. CONTENTSOF PREVIOUSVOLUMES.

. .

. .

. . . . . . . . . . . . . .

268 278 285 327 330 331 339

344

This Page Intentionally Left Blank

Numbers in parentheses indicate the pages on which the authors' contributions begin.

ANN M . AINSWORTH, Departments of Pathology and Dermatology, Temple University Medical School, and The Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania ( 267 ) CLAUDIOBASILICO,Department of Pathology, New York University School of Medicine, New York, New York (223) ROBERT E. BELLET,Departments of Pathology and Dermatology, Temple University Medical School, and The Institute for Cancer Research, Fox Chase Cuncer Center, Philadelphia, Pennsylvania ( 267)

DAVIDBERD,Departments of Pathology and Dermatology, Temple University Medical School, and The Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania ( 267) EVELINAA. BERNARDINO, Departments of Pathology and Dermatology, Temple University Medical School, and The Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania ( 267 ) WALLACEH . CLARK,JR., Departments of Pathology and Dermatology, Temple University Medical School, and The Institute for Cancer Research, Fox Chase Cancer Center, Philaddphia Pennsylvania ( 267) MIKE FRIED,Imperial Cancer Research Fund Laboratories, Lincoln's Inn Fields, London, England (67) BEVERLY E. GRIFFIN,Imperial Cancer Research Fund Laboratories, Lincoln's Inn Fields, London, England (67) J . C. LECLERC, Laboratoire d'Immunologk des Tumeurs, Service dHe'matologie Groupe INSERM, U 152, Pavillon Gustave Roussy, Hcpital Cochin, Paris, France (1)

J. P. LEVY, Laboratoire d'lmmunologie des Tumeurs, Service d'Hkmatologie Groupe INSERM, U 152, Pavillon Gustave Roussy, Hcpital Cochin, Paris France ( 1 ) ix

CONTRIBUTORS

X

* Cytogenetic Laboratory, Department of Pathology, Central Hospital, Skovde, Sweden (165)

JOACHIM M A R K ,

MICHAELJ . MASTRANGELO, Departments of Pathology and Dermatology, Temple University Medical School, and The Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania ( 267 ) LARSOSTBERG, Institute of Medical and Physiological Chemistry, Biomedical Center, University of Uppsala, Uppsala, Sweden (115) PERA. PETERSON, Institute of Medical and Physiological Chemistry, Biomedical Center, University of Uppsala, Uppsala, Sweden ( 115)

LARSRASK,Institute of Medical and Physiological Chemistry, Biomedical Center, University of Uppsala, Uppsala, Sweden ( 115)

* Present address: Department of Pathology, Central Hospital, 541 01 Skovde, Sweden.

OBITUARY

SIR ALEXANDERHADDOW 1907-1976 The Editors of this series sadly announce the loss of a distinguished colleague, Sir Alexander Haddow, whose death on January 21, 1976, ended the career of one of the world's outstanding leaders in cancer research. By a strange coincidence, another luminary in cancer research, War0 Nakahara, died on the same day. Alex Haddow was one of the founders of Advances in Cancer Research, and served as a coeditor with the late Jesse Greenstein from 1953 to 1958 (Volumes 1 through 5 ) . On the death of Jesse Greenstein in 1956, one of us (S.W.) took his place, and from Volumes 6 through 11, had the privilege of collaborating as coeditor with Alex Haddow. When ill health (virtual blindness) made it necessary to terminate this role in 1968, Alex continued his association with the Advances as Consulting Editor. His was a life uniquely devoted to cancer research. As he describes in a stirring autobiographical essay [Cancer Research 34, 3159-64 ( 1974) 1, his choice of a career in medicine was made when he was hardly out of the cradle. Shortly after graduating in medicine from the University of Edinburgh in 1929, he began research in chemical carcinogenesis at the same University; and continued work in this field on joining the Royal Cancer Hospital in London in 1936. At that time and place, the newly emerging field of hydrocarbon carcinogenesis was developing brilliantly under the leadership of Sir Ernest Kennaway. Succeeding Kennaway in 1946 as Director, Alex Haddow built the newly established Chester Beatty Institute into one of the world's leading cancer centers, where epoch-making progress was recorded in chemotherapy, chemical carcinogenesis, and the biology and pathophysiology of cancer. Since 1972, when the complications of diabetes necessitated his retirement, he moved to the Institute's lodge at Pollards Woods, where with the constant aid of his wife Feo, he continued his life of study and writing. Despite a busy research career and directorial responsibilities, Sir Alex was heavily involved in worldwide organizations devoted to the cancer problem, and his various posts in such external bodies are too many to list in these short paragraphs. He held several leadership positions xi

xii

OBITUARY

in the British Empire Cancer Campaign, was founder and President of the Oncology Section of the Royal College of Medicine, a Fellow of the Royal Society, Vice-president of the British Cancer Council, and from 1962 to 1966 was President of the International Union Against Cancer. Among many awards were foreign memberships in the Academy of Medical Science, USSR; Academy of Arts and Sciences, U.S.A.; the American Association for Cancer Research (Honorary Member); and the New York Academy of Sciences (Fellow). Other honors were received from France (Croix de Chevalier de Legion d'Honneur), Cuba, Belgium, and Czechoslovakia; and Honorary Doctorates from the Universities of Edinburgh, Perugia, and Helsinki. He was knighted in 1966. The Editors of this serial publication mourn a warm friend, a brilliant scientist and Ieader of scientists, and a benefactor of humanity.

ADVANCES IN CANCER RESEARCH VOLUME 24


THE MURINE SARCOMA VIRUS-INDUCED TUMOR: EXCEPTION OR GENERAL MODEL IN TUMOR IMMUNOLOGY?

. .

1. P. levy and 1 C Leclerc Loborotoire d'lmmunologie des Tumeurs. Service d'H6matolopie Groupe INSERM. U 152. Povillon Gurtove Rousry. HBpitol Cochin. Paris. France

. .

I Introduction . . . . . . . . . . . . . . . . . I1 In Viuo Studies of the Immunological Rejection of MSV-Induced Sarcomas A . Suggestions of an Immune Reaction from the Natural History of the Tumor . . . . . . . . . . . . . . . . . . B In Viuo Demonstration of a Potent Antitumor Response in MSV-TumorBearing Mice . . . . . . . . . . . . . . . . C . Tumor-Associated Transplantation Antigens (TATA) in the MSV System . . . . . . . . . . . . . . . . . . I11. Analytical Study of the Antivirus Immune Response . . . . . . A. Antigens of the Viral Particles . . . . . . . . . . . B. Anti-MSV Neutralizing Antibodies . . . . . . . . . . C In Viuo Role of Anti-MSV Neutralizing Antibodies . . . . . . IV. Analytical Study of the Antitumor Cell Reaction: Humoral Response . . A Serologically Defined Antigens of the MSV Tumor Cell Surface . . B. Antitumor Cell-Reacting Antibodies . . . . . . . . . . C. In Viuo Role of the Antitumor Cell Antibodies . . . . . . . V Analytical Studies of the Antitumor Response: Cell-Mediated Immunity (CMI) . . . . . . . . . . . . . . . . . . . A Detection of CMI by the Colony-Inhibition Test ( C I T ) and the Microcytotoxicity Assay (MA) . . . . . . . . . . . . . B. Study of the Antitumor CMI by Analytical Methods . . . . . C. Interactions of Effector Cells and Soluble Factors . . . . . . D . Antigenic Specificities Involved in CMI . . . . . . . . . E. In Vluo Relevance of the in Vitro-Detected Cell-Mediated Reactions VI MSV Tumors and Immune Surveillance . . . . . . . . . . A. MSV Tumors Escape . . . . . . . . . . . . . . B. Appearance of Leukemias after Rejection of MSV Tumors of Identical Antigenicity . . . . . . . . . . . . . . . . . VII. General Comments . . . . . . . . . . . . . . . . A Contribution of the MSV Model to the Understanding of an Efficient Antitumor Immune Response . . . . . . . . . . . . B. General Value of the MSV Model in Tumor Immunology . . . . C The MSV System: A Model of an Immune Response against a Viral Infectious Disease . . . . . . . . . . . . . . . D . Similarities of anti-MSV and anti-H-2 Immune Responses . . . . References . . . . . . . . . . . . . . . . . .

.

. .

.

.

.

.

.

2 3

3

7 14 21 21 24 25 26 26 31 32 32

32 34 41 45 49 50 50 51 52 52 54 56

57 58

2

J. P. LEVY AND J. C. LECLERC

1. Introduction

The tumors induced by murine sarcoma virus (MSV)' have been one of the most extensively studied models in tumor immunology during the past 10 years. Being autochthonous tumors with a rapid development at the site of virus inoculation, followed by a spontaneous rejection and a strong resistance to further virus challenges, they appear as an attractive model for the study of the antitumor response in the natural host of a primary tumor. Furthermore, the MSV is oncogenic for various inbred strains of mice as well as for hamsters and rats, thus providing the opportunity to compare the antitumor response in different genetic backgrounds. Numerous groups have chosen this model, and we now have a considerable amount of information about the tumor-associated antigens, the antibody response, and the cell-mediated antitumor reaction. This review will try to summarize these data and the problems that now arise about the immunology of the MSV system, since these problems are of general interest in experimental and human tumor immunology. The observations reported in this system have been especially useful, notably in cellular immunology, but we still do not know whether the highly antigenic MSV tumor must be considered as a general model, or as an exception in tumor immunology. Most of the experinients that will be reviewed, have been done with five different isolates of MSV: Harvey, or H-MSV (Harvey, 1964); Moloney, or M-MSV (Moloney, 1966); Kirsten, or K-MSV (Kirsten and Mayer, 1967); Finkel, or FBJ-MSV ( Finkel et al., 1966) and Gazdar or G,-MSV (Gazdar et al., 1972a). However, it must be emphasized that the descriptions of the antitumor response, notably when cell-mediated, concern mainly the M-MSV isolate, probably because it is characterized 'Abbreviations used in this review: MSV: murine sarcoma virus; MSV prefixed with H, K, M, Gz or FBJ: MSV pseudotypes isolated, respectively, by Harvey, Kirsten, Moloney, Gazdar, or Finkel; MuMAV: murine myeloma-associated virus; MuLV: murine leukemia virus; G, Gi, F, M, and R, respectively, Gross, Graffi, Friend, Moloney, and Rauscher strain of MuLV; WMV: woolly monkey virus; TATA: tumor-associated transplantation antigen; SCSA: sarcoma-specific antigen; GCSA: Gross specific cell surface antigen; VCSA: viral cell surface antigen; VEA: viral envelope antigen; MEV-SA: murine endogenous viral surface antigen; gs: groupspecific antigen; NP ( cells) : nonproducer ( cells) ; GVH: graft-versus-host reaction; CTL: cytolytic T lymphocytes; CMI: cell-mediated immunity; IEM: immunoelectron microscopy; IF: immunofluorescence; CIT: colony-inhibition test; MA: microcytotoxicity assay; MLTR: mixed lymphocyte tumor cells reaction; CRT: chromium release test; S. CRT: secondary CRT; MMI: macrophage migration inhibition test; PA: proline assay; ATS: antithymocyte serum; ADCC: antibody-dependent cellmediated cytotoxicity.

MURINE SARCOMA VIRUS-INDUCED TUMOR

3

by ( 1 ) usual induction of sarcomas at the site of virus inoculation without other macroscopically detectable pathology, ( 2) regular spontaneous rejections, which are less frequent with the other isolates (Harvey and East, 1971). The pathology and virology of MSV will not be considered in this review. They have been extensively studied in the Harvey and East review ( 1971). In addition, a great number of subsequent reports have been published, which cannot be reviewed here. However, it is necessary to know the main characteristics of MSV, a type C RNA virus with defective replication and transforming activities in uiuo and in uitro, to comprehend the immunologic aspects of the MSV system. The constant association of a helper virus in MSV producer cells must be especially emphasized. Also, no comparison has been attempted with the immunology of other tumors, unless necessary. As far as possible, each of the four main sections of the review has been treated as a unit in its own right. The sections concern, respectively, in uiuo tumor protection ( Section II), antivirus immune response (Section III), antitumor cell antibody response ( Section IV), and cell-mediated antitumor immunity (Section V). The complexity of the antigens of the MSV system is remarkable. Therefore, we felt that to ensure a better understanding, the antigens involved in each of these four reactions must be studied separately. In addition, Table I summarizes the main antigens existing in the MSV tumor. II. In Vivo Studies of the Immunological Rejection of MSV-Induced Sarcomas

A. SUGGESTIONS OF AN IMMUNE REACTION FROM NATURAL HISTORY OF THE TUMOR

THE

The evolution of the tumors induced in mice by subcutaneous or intramuscular MSV inoculation is well known (Harvey and East, 1971). The neoplasms arise and progress rapidly at the site of inoculation, and they are frequently extensive enough to weigh up to 10%of the total body weight when newborns have been inoculated. One of the most remarkable points is that these tumors will follow a different evolution in very young and in adult animals. Whatever the inbred line, practically 100% of the adults will finally reject the local M-MSV tumor. The pathology of H-MSV and K-MSV is more complex since most of the treated animals develop at the same time a local tumor and a spleen erythroblastosis. Changes similar to those of Friend disease with erythroblast proliferation are usual after H-MSV infection, so that the mice

VEA

Type specific Surface VEA projections gs-VEA Surface projections

[

Helper virus

gp69/71

Eckner-Steevea (1972) Helper virus Gomard el ul. (1973) Aoki (1974)

Another sub-gs VEA Mulv-VEA . was described by Aoki 1974

?

Helper virus

Geering et al. GCSA (b) gsl, and (1966, 1968) gs3 are group Yoshiki et al. specific GCSA (b) could be (1973, 1974) identical to m Ferrer (1973) - l Aoki et 'al (1973j

gp69/71

GCSA (b)

?

gsl

Core shell

,,

P30

Helper virus

Core shell

,,

P30

Helper virus

gsJ SCSA

(1) Budding particles (2) Can be VCSA

SCSA SCSA (d) SCSA (b) and (c)

1 ? *

I Embryonic Embryonic antigens specificity Endogenous MEV-SA1 virus antigens

Absent ? ?

Cell surface in nonproducer cells and cell ? surface outside virus particles in producer cells Cell surface Cellular components Cell surface ?

MSV

1

Aoki et a2. (1974a)

I Host cell Endogenous type C virus

Salinas and Could be multiple Hanna (1974) Herberman et al. (1974)

For general reviews, see Levy (1974) and Bauer (1974). VEA, viral envelope antigens; VCSA, viral cell-surface antigens; SCSA, sarcoma cell-surface antigens.

+5

n -_

c


5

generally are killed by the spleen lesions without having had the time for regression of the local tumor. On the contrary, when M-MSV is used, there is no spleen erythroblastosis and the evolution of the local tumor can be studied independently. After 2 4 weeks, on the average, all the animals are tumor-free and spontaneous recurrences occur only in a very small percentage. On the other hand, 1Om of the newborn infected recipients die with a huge local tumor, sometimes with metastatic proliferation (Harvey and East, 1971). It is interesting to observe that sarcomas appear in adults as well as in newborns of the same inbred line when high virus doses are inoculated, the discrepancy between the two groups being detectable only at the stage of the tumor rejection. This suggests that the cells are sensitive to the oncogenic potency of MSV in adults as well as in newborns, but only adults are able to mount an antitumor reaction. From the beginning, it was supposed that this reaction could be immunologic, and that the newborns do not reject MSV tumors owing to their well-known immunologic immaturity. The ontogeny of the antitumor response has been studied by different groups (Fefer, 1969; McCoy et al., 1972a). The anti-M-MSV response becomes detectable in uiuo in BALB/c around the age of 3 weeks. A 50-7M rejection is observed at 4 weeks, and the maximum a little later on. However, the rejection ability may still not be total at the age of 8 weeks (Fefer, 1969). In CBA/wh the resistance to M-MSV is detectable at 2 weeks and complete at 5 weeks. A similar, or slightly more rapid, development of the antitumor response has been found in C3Hf/Gs inoculated with K-MSV, with complete protection in mice 4 5 weeks old (McCoy et al., 1972a). In our experiments (unpublished results), C57BL/6 are especially remarkable by a very rapid appearance of the ability to reject the M-MSV tumor, all being already rejected in S-weekold recipients. The level of sensitivity is different among the inbred strains of mice. For instance, C57BL/6, C57BL/10, B10-Br, DBA/2, CBA, Swiss NIH, and BALB/c are sensitive, whereas AKR and their F, hybrids with CBA, NIH, or DBA f 2 are relatively resistant to the M-MSV ( Chieco-Bianchi et al., 1974; Colombatti et al., 1975a,b). Even among sensitive lines some discrepancies can be found: it is well known, for instance, that C57BL/6 are less sensitive than BALB/c to low virus doses, and that they reject the tumor more rapidly. These variations couId be due to unequal levels of antitumor immune response, but no precise arguments have been yet given to support this hypothesis. The study of the tumor histology reinforces the idea that the tumor rejection could be an immunological phenomenon. Two different types of lesions can be found in the tumors: a clearly neoplastic proliferation

6


and an inflammatory granulomatous reaction. The neoplastic proliferation is composed of mesenchymal cells that can be fibrosarcomatous, or myoblast cells, or other mesenchymal, sometimes undifferentiated cells. In addition, hemangiosarcomas appear also relatively frequently (see notably Chesterman et al., 1966; Perk and Moloney, 1966; Perk et al., 1967; Stanton et al., 1968; Thomas et al., 1973; and for a review, Harvey and East, 1971). This problem will not be discussed here, but it can be mentioned that no differences in the antitumor response have been demonstrated according to the cell type of the neoplastic proliferation. The inflammatory reaction consists of polymorphonuclear cells, occasionally mast cells, and eosinophiles and a dominant infiltration of mononuclear cells, which are lymphocytes and possibly histiocytes. During tumor evolution in adults, this inflammatory exudate becomes more and more important, whereas the number of tumor cells decreases. Finally, the tumor cells completely disappear. On the contrary, the study of tumors induced in newborns does not reveal any mononuclear cell infiltration, but only the proliferation of tumor cells, which progress continuously until death (Perk and Moloney, 1966; Fefer et al., 1968a). The observation of tumor cell grafts confirms the correlation between mononuclear infiltration and the ability to reject the tumor (Russel and Cochrane, 1974), and the same conclusions are drawn from the study of G,-MSV tumors (Gazdar et al., 1973). Therefore, one can suppose that this infiltration represents an antitumor reaction that will provoke the tumor cell destruction. This hypothesis is strengthened by the observation that lesions with the usual morphologic characteristics of neoplasms, that is to say, with large areas almost exclusively composed of cclls of the same type, with mitotic foci and no apparent organization, are rare in adult infected mice, but occur more frequently in thymectomized or irradiated animals (Stanton et al., 1968). This kind of proliferation, with a clear neoplastic appearance, is especially frequent in tumors that develop several weeks after virus inoculation. Similarly, in addition to typical pleomorphic tumors, other neoplasms, composed of monomorphic cells with nodular or diffuse growth, reminiscent of clonal aggregates, can be observed in the resistant adult AKR inoculated with M-MSV (Ch’iecoBianchi et al., 1974). In these micc, the tumors grow slowly, but they ultimately kill the host in most cases; they are due to the spontaneous formation of a poorly immunogenic Gross ( G ) pseudotype (see Section 111,AJ ). Similarly, the naturally occurring G pscudotype of the FBJ-MSV isolated from a spontaneous osteosarcoma of CF1 mice (Finkel et al., 1966), induces progressively growing tumors with purely neoplastic morphological characteristics and very few granulomatous lesions or mononuclear cell infiltrations (Price et al., 1972). Therefore, when one


7

considers the value of MSV tumors as an in wiuo model in tumor immunology, one must remember that two different kinds of such tumors exist: 1. The sarcomas, detectable very early after the virus inoculation, usually in the first 2 weeks, are virus-producing and strongly antigenic. In most cases, adults are able to reject these tumors, which are associated with an inflammatory reaction. It is not certain whether or not the tumor cells are really autonomous; a constant production of virus with recruitment of newly infected transformed cells could be necessary to ensure tumor development, as suggested notably by the difficulty in establishing permanent transformed cell lines by in uitro infection of primary mouse embryo fibroblasts or in maintaining primary in uiuo MSV-induced tumors in a permament in uitro culture (Simons, 1970; Simons and McCully, 1970). In some way, these early sarcomas are perhaps equivalent to the “Early Foci,” dependence of virus production in the in uitro MSV-induced transformation (Aaronson et al., 1970). It is probable that really autonomous tumor cell clones would also appear inside these early sarcomas. However, in most cases, such clones would be superinfected by the viruses produced by the surrounding cells, and therefore they would be destroyed by the antitumor response, which appears to be mainly directed against viruses and/or virus products of the host cell surface (see following sections of this review). 2. By contrast, late sarcomas, which appear after several weeks could be the in uiuo equivalent of the in uitro virus-production-independent “late foci” of transformed cells ( Aaronson et al., 1970). Such sarcoma cell clones would be selected mainly in two situations: if they are nonvirus producers or if they produce a poorly immunogenic virus. In both cases, it would not be surprising if the mononuclear cell infiltration were absent or remained very weak, which could explain the slow but continuous proliferation. The rapidly growing sarcomas provide very convenient systems for study of the rejection of tumor cells in uiuo, but the slowly growing sarcomas are probably much more relevant for the natural situation. B. In Vivo DEMONSTRATION OF A POTENTANTITUMOR RESPONSE IN MSV-TUMOR-BEARING MICE

1. Development of a Specific anti-MSV Tumor Resistance in Regressor Mice Regressor mice are strongly immunized against a booster MSV injection ( Fefer et al., 1968a) or against the graft of live sarcoma cells (Fefer et al., 1967a; Burstein, 1970). The same is true in regressor rats (Jones

8


et al., 1974). The immune protection is even stronger after repeated immunizations, and it concerns not only MSV transformed cells, but also lymphoma cells, which share some membrane transplantation antigens with M-MSV (see Section I1,C). The immune protection against syngeneic cell grafts is already detectable 8 days after virus inoculation at a time when the autochthonous tumors are still growing, and it is complete on day 14, when the clinical regression of the autochthonous tumor begins (Fefer et al., 1968a,b). Thereafter, the immune protection persists for a long time, at least for 46 weeks in C57BL/6. A very small number of BALB/c have been tested in uiuo after 46 weeks. Fefer et al. (1968a) examined only two mice, which did not resist the tumor grafts. This could explain why some of these mice develop leukemias of common antigenicity with MSV tumors several months after the initial tumor rejection. The reason for such a decrease of the specific immunity is not clear. One explanation could be that a nonspecific defect of the immune response associated with aging occurs, since it has been shown that old mice become more sensitive to MSV oncogenesis, either by virus infection (Pazmifio and Yuhas, 1973) or by cell transplants ( Strausser and Bober, 1972). Whatever the late evolution of the antitumor response, in the weeks that follow tumor regression practically 100%of the mice are strongly resistant to MSV-tumor development, showing that unusually potent tumor-associated transplantation antigens ( TATA ) are present on the MSV-tumor cells (see Section I1,C). The in uiuo immune resistance is confirmed by the development of a high level of antibodies (see Sections I11 and IV) and of a cell-mediated immunity ( see Section V). 2. Enhancement of MSV Tumor Growth in Zmmunosuppressed Animals a. X-Rays and Chemical Agents. A 350-rad X-ray irradiation 1 day prior to MSV inoculation does not modify the susceptibility to M-MSV oncogenesis, the tumor incidence being similar in irradiated and control mice. However, whereas the tumors regress in 100%of controls, they progress and kill a large number of irradiated animals. The same results have been observed as well with K-MSV and M-MSV and in primary or transplanted tumors (Fefer et al., 1967b,c; McCoy et al., 1972a). A similar suppression of the capacity to cure MSV-induced tumors has also been observed in about 601%of mice that have received 2.2 mg of cortisone acetate 24 hours prior to virus inoculation (Shachat et al., 1968). In these experiments, an increase in tumor incidence was also reported that was not found after irradiation (Fefer et al., 1967b). An enhancement of virus production in cortisone-treated animals could be responsible, but an immunologic phenomenon cannot be excluded, since the same 100%


9

tumor incidence has been detected as well after thymectomy or antithymocyte treatment ( see section below), Tumor regression can be inhibited also in adult mice if they are treated with various antimitotic agents, such as cyclophosphamide ( Fefer, 1969, 1970) or daunorubidomycin (Casazza et al., 1971; Giuliani et al., 1973). These drugs can be administered 6 days after the virus or 1 3 days before the virus. The second schedule has the advantage of greatly weakening their antitumor effect, so that only their immunosuppressive action is tested. In these experimental conditions, the tumors follow a normal evolution for about 2 weeks, including the beginning of regression. However, later they enlarge again, so that finally most of the immunosuppressed animals die with huge tumors. b. Enhancement of Tumor Growth in Animals with Altered Thymic Functions. Thymectomy performed in newborn mice and rats or in lethally irradiated 2-month-old mice, reconstituted with bone marrow cells, increases tumor incidence and inhibits tumor rejection (Ting, 1967; East and Harvey, 1968; Law et al., 1968a; Collavo et al., 1974). On the contrary, such thymectomized mice are protected against the appearance of H-MSV-induced late-developing lymphatic leukemia ( East and Harvey, 1968). The ability to resist MSV-induced tumors can be reconstituted by a thymus graft under the kidney capsule of CBA mice, provided the graft was performed early enough to have restored the thymic function, at the time of tumor development (Collavo et al., 1974). The graft restores resistance when done 30 days before MSV inoculation; when performed 1 5 days after MSV inoculation, it does not inhibit tumor appearance, but still allows rejection; thymectomy in adults does not modify tumor evolution (unpublished data). Similarly, an antithymocyte serum ( ATS ) treatment increases both tumor incidence and mortality in MSV-infected mice (Law et al., 1968b; Hook et al., 1969; Varet et al., 1968, 1971). The effects are detectable in sensitive as well as in resistant ( AKR) lines ( Chieco-Bianchi et al., 1974). When ATS has been given before the MSV inoculation, the immunosuppression is complete and 100%of the recipients die. A weak effect is still detectable 1 day after the virus inoculation, the tumors being more frequent but being always rejected. After 2 days, the immunosuppression becomes inefficient, even with repeated ATS injections: the tumor incidence and mortality are identical in ATS-treated animals and in controls, and the same treatment followed up to 60 days does not induce any recurrence (Varet et al., 1971). However, this treatment which is not able to stop the rejection of the first tumor, can depress the secondary antitumor response, with clear enhancement of an MSV tumor due to a secondary virus inoculation ( Varet et al., 1971 ) .

10


On the whole, thymectomy and ATS experiments are very informative: (1) They show that both tumor incidence and tumor rejection are under the control of immune responses. ( 2 ) They demonstrate that the antiMSV reaction is a T-cell-dependent reaction and involves a T-celldependent memory phenomenon. (3) They show that the immune reaction is very strong and cannot be suppressed more than 1 day after the virus inoculation; this suggests that at that time effector cells or their progenitors are already present in the tumor and do not circulate, or that they are in a differentiation state that makes them resistant to ATS. The experimental data obtained in thymectomized or ATS-treated animals, are confirmed by the observation of nude mice that possess a defective thymus (Pantelouris, 1968) and are unable to reject the MSVinduced tumors (Allison et aZ., 1974; Davis, 1975; Stutman, 1975). In some experiments (Stutman, 1975) a paradoxical increased latency of tumors has also been reported, which could be due to the absence of immunosuppressive cells (see Sections VI,A) or to the natural occurrence of nonspecific cytotoxic antibodies in nude mice (J. W. Martin and Martin, 1974; S . E. Martin and Martin, 1975). However, this long incubation period of the tumor is not mentioned in all reports. The role of the thymus in tumor rejection is confirmed by the observation that thymosin given in very young CBA/ wh accelerates the development of resistance to tumor growth (Zisblatt et al., 1970). c. Enhancement of MSV Tumors by Graft-uersus-Host Reaction ( G V H ) .The inoculation of 10' living parental spleen cells in (BALB/c X C57BL/8) F, 10 days after the MSV inoculation strongly increases the tumor incidence, and 25%of the mice die with MSV tumors (Varet et al., 1973). This observation shows that the enhancing effect of GVH is not restricted to lymphoid malignancies allowing us to conclude that the chronic stimulation of lymphoid cells during GVH is not sufficient to explain the facilitation of spontaneous lymphomas ( Schwartz and Beldotti, 1965) or Graffi-virus-induced leukemias (Varet et al., 1973). However, the mechanism of MSV-tumor enhancement by GVH is not clear. It has been shown that it increases leukemia virus production (Hirsch et al., 1972) : this improvement of helper virus replication could be responsible for a better MSV production, Another explanation, the well known immunosuppressive effect of the GVH, cannot be ruled out (Lapp and Moller, 1969; Moller, 1971) d. Enhancement of MSV Tumors by Coinfection with Other Viruses. Preinfection with Rauscher, Friend, or Moloney leukemia viruses reverses the local tumor regression usually observed in M-MSV inoculated adult mice ( Chirigos et al., 1968). Similarly, the leukemogenic component associated with H-MSV increases the sarcoma incidence in H-MSVinfected animals (Harvey and East, 1989, 1970). An increase of the MSV

.

MUFUNE SARCOMA VIRUS-INDUCED TUMOR

11

production (Turner and Chirigos, 1969) or the strong immunosuppression of antibody production, notably observed with the Rauscher virus ( Hook et al., 1969), could be responsible, but no definite demonstrations have been offered. A potentiation of MSV oncogenicity has also been reported in lactic-dehydrogenase-elevatingvirus ( LDV) infected mice (Turner et al., 1971). A possible inhibition of cell-mediated immune response is suggested, but has not been proved. 3. Inhibition of MSV Tumors by Nonspecific Stimulation of the Zmmune Functions The MSV tumor is probably not a good system for this kind of experiment, since in most cases adults are spontaneously able to reject the tumor. However, some experiments have been done with highly activated viruses that could be lethal even for adult mice. Good protection by Bacillus Calmette-GuBrin ( BCG ) has been reported in animals which received this immunostimulator at the moment of virus inoculation, especially if they had been preimmunized 28 days earlier by a first BCG injection (Schwartz et al., 1971). The mechanism by which BCG could interact has not been studied. A nonspecific protection of newborns has been also observed with poly (2’,O-methyladenilic acid) [poly(Am) 1, a single-strand polynucleotide. This agent is efficient only against relatively weak doses of infectious virus. It could play this role by a nonspecific stimulation of B cells improving the antibody synthesis in young animals (Tennant et al., 1974). However, another effect of poly(Am), notably on the virus, has not been excluded. The effect of splenectomy, which decreases the maximum volume of the virus-induced tumor and accelerates tumor rejection, especially in females, could also be immunologic, but its exact mechanism remains unknown (Pollack, 1971), In other experiments with G,-MSV splenectomy did not modify the tumor evolution ( Gazdar et al., 1973). 4 . Mechanism of the Anti-MSV Immune Protection in Vivo

a. In Vivo Transfer of Anti-MSV Sera. The serum of regressors sampled 1 month after MSV-tumor rejection, protects sensitive recipients very efficiently against an MSV inoculation performed 24 hours laters (Law et al., 1968a; Pierce, 1971). An even better protection can be obtained with hyperimmune sera from mice that have received repeated injections of MSV or irradiated sarcoma cells (Bubenik et al., 1969; Fefer, 1969, 1970). Various schedules of serum transfer have been used, including notably single or multiple serum injections. Relatively small volumes

12


(0.1. ml) of serum donor, given 1 day prior to the virus infection (Pierce, 1971), are sufficient to protect the recipients. In fact, the timing of serum and virus inoculations appears to be the most important factor, and, in all experiments, good protection was obtained when the serum was given either before or a few days after the virus. By contrast, practically no protection was found if the treatment began on day 6, or later, after virus infection, even if the tumors were still not palpable at that time (Bubenik et al., 1969; Giuliani et al., 1973). In immunosuppressed animals, a secondary recurrence of the tumor is frequent, probably owing to the short life-span of passively transferred antibodies (Fefer, 1989, 1970; Giuliani et al., 1973). A similar passive immunity against MSV oncogenesis exists in offspring of immunized mothers which are inoculated at the age of 7 to 14 days ( Chieco-Bianchi et al., 1973). In other experiments, the challenge was done with a tumor cell graft instead of a virus inoculation. Both transplantable sarcomas and antigenically related lymphomas induced by the Moloney leukemia virus have been used. The results are less clear than after a virus challenge. Very good protection was obtained with regressor or hyperimmune sera given 3 days before the tumor graft (Pearson et al., 1973). However, other groups reported only weak in uiuo protection in these conditions with regressor sera (Bubenik et al., 1989; Weissman, 1973). It seems that the protection is highly variable according to the serum used, even after hyperimmunization, and it is sure that not only protection but also complete absence of effect or an enhancement can be found (Bubenik and Turano, 1988a; Bubenik et al., 1989), the enhancing factor being in the 7 S fraction (Bubenik and Turano, 1968a). An enhancing effect was also reported with progressor sera (Chirigos et al., 1968; Pearson et al., 1973; Jones et al., 1974) or sera of early regressors harvested 5 days after the beginning of the tumor rejection (Pierce, 1971). The possible role of blocking antibodies, circulating antigens, or antigen-antibody complexes in these phenomena will be discussed in Section V,C,1. On the whole, these experiments show that the regressor serum is able to inhibit the in duo proliferation of MSV or lymphoma cells. However, this effect is mainly demonstrated when the tumor cells have been preincubated in vitro with the serum, before their in duo inoculation (Fefer et al., 1967c; Varet et al., 1968; Bubenik et al., 1969). Such a procedure does not necessarily reflect the in uiuo situation, and it does not allow to determine up to what point the humoral factors are important in the spontaneous MSV-tumor rejection. Similarly, the use of hyperimmune sera is hardly quantitatively relevant to the in uioo situation. The respective roles of antivirus neutralizing (see Section 111) and antitumor cell antibodies (see Section IV) also remains an open subject.


13

It was observed that sera containing the highest titers of neutralizing antibodies were also the most effective in preventing MSV carcinogenesis (Bubenik et al., 1969), and we have seen above that in viuo protection is easier to obtain against a booster virus inoculation than against a tumor cell graft. This suggests that neutralizing antibodies are probably responsible, in large part, for the in duo protection obtained by the passive transfer of immune sera. However, the fact that the transfer of regressor sera inhibits the in viuo growth of sarcoma cell grafts and of lymphoma transplants (Varet et al., 1968; Pearson et al., 1973), which are actually independent of virus replication, suggests that the protective effect of regressor serum is not unequivocal. In conclusion, in viuo neutralization of the virus, and less certainly, the destruction of tumor cells by an antibody-dependent mechanism seem to be important in anti-MSV tumor protection. Nevertheless, it must be emphasized that the rapidly growing tumors due to M or to H-MSV infections possibly would not constitute a perfectly valid model for comparing these two kinds of humoral immunities, since the role of autonomous tumor cells is not established in this system. Unfortunately, there is no information about the in viuo role of antibodies in latedeveloping autonomous MSV tumors ( see Section I1,A). b. Zn Vioo Transfer of Zmmune Lymphoid Cells. The possible transfer of anti-MSV immunity by regressor spleen or lymph node cells is well demonstrated. Protection is found when in vitro mixtures of immune lymphocytes and tumor cells are inoculated in duo (Fefer et al., 1967c; Varet et al., 1968; Jones, et al., 1974). More significant is the protection obtained when tumor cells and lymphocytes are injected in different ways (Fefer, 1969) or when the recipients have been inoculated with MSV before or after the lymphoid cell transfer (Fefer, 1969, 1970; Giuliani et al., 1973; Gorczynski, 1974a; Pollack, 1971; Leclerc et al., 1976). Protection by immune lymphocytes has been found as well in X-irradiated animals (Pollack, 1971; Gorczynski, 1974a), in mice treated with various drugs (Fefer, 1969, 1970; Giuliani et al., 1973), and in newborn (Varet et al., 1968; Leclerc et al., 1976) or very young mice ( Fefer, 1969). As for antibody transfer, the timing between lymphocyte and virus or tumor cell inoculations is the main determining factor. Protection is obtained when the immunocompetent cells are given 4 days before (Varet et al., 1968) or at the time of virus inoculation (Gorczynski, 1974a) or even 4 days later (Leclerc et al., 1976). However, on day 6 after inoculation, the transfer of lymphoid cells becomes inefficient. Pretreatment of immune spleen cells by anti-Thy.1.2 and complement completely abolishes their ability to transfer any anti-MSV immune protection (Gorczynski, 1974a; Leclerc et d., 1976), which demonstrates that T cells play a major role in the phenomenon. The exact nature of

14


the involved T cells is not yet established. An attractive hypothesis would be that the cytolytic T lymphocytes (CTL) which are present in the spleens of immune mice could be directly responsible for tumor cell destruction (see Section V,B,2,c). However, arguments have been proposed that are not in agreement with this assumption, notably: (1) Protection by allogeneic immune cells would be dependent on their persistence for at least 1 week in the host (Fefer, 1969, 1970). (2) The transferred CTL are efficient, whatever the level of their in uitro activity ( Leclerc et al., 1976; see Section V,E ) , These arguments are not definitely conclusive, but they suggest that the donor CTL are possibly not directly involved. Two main hypotheses could explain the in uioo protection by T cells: ( a ) T cells are necessary in T-B cell cooperation to produce antitumor cells or virus-neutralizing antibodies. ( b ) CTL or their progenitors must be restimulated in vivo to reach an efficient level of antitumor activity. The recent demonstration that CTL rapidly lose their activity when they are separated from tumor cells, but can be strongly restimulated in a secondary reaction, supports this hypothesis (Plata et al., 1975; Senik et al., 1975a). A definite but lesser immune protection has been described in recipients reconstituted with normal, nonimmune, spleen or thymus cells. Anti Thy.l.2-treated spleen cells or untreated bone marrow cells were devoid of activity (Gorczynski, 1974a). This result is in agreement with the possible reconstitution of an anti-MSV response in thymectomized mice, by thymus graft (see above Section II,A,2,b). However, the results are less clear in the transfer of nonimmune cells than in the transfer of anti-MSV lymphocytes; and negative results have been reported by several groups, despite inoculations of very high doses of spleen cells or thymocytes (Varet et al., 1968; Fefer, 1970; Hirsch et al., 1972; Pollack, 1971; Leclerc et al., 1976). The time schedules could be a determining factor in these experiments, since the reconstitution of the thymus functions requires some delay. In conclusion, the reconstitution of immunodeficient recipients by immune lymphoid cells of MSV regressor mice allows one to obtain a strong anti-MSV protection and even the regression of already palpable tumors. In all comparative experiments reported, this kind of immunotherapy was more efficient than the inoculation of immune sera. C. TUMOR-ASSOCIATED TRANSPLANTATION ANTIGENS (TATA) IN THE MSV SYSTEM The cell surface of MSV sarcomas is a complex mosaic of antigens, like those of any other RNA-oncogenic-virus-producer tumors ( Levy, 1974). Some of these antigens can induce an immune reaction resulting

MURIh'E SARCOMA VIRUS-INDUCED TUMOR

15

in tumor rejection by syngeneic hosts. This is the definition of TATAs that will be studied in this section. The other antigenic specificities, which are defined only by in uitro experiments, will be studied in the following sections of this review, Most of the studies on MSV-tumor antigens have been done with tumor cells that produce both MSV and the helper-murine leukemia virus. In such conditions it is difficult to distinguish antigenic changes associated with virus replication from those due to virus-coded but nonvirion cell-surface antigens. The use of MSV-transformed cells that lack any evidence of virus production (nonproducer, or NP, cells), but from which the MSV genome can be rescued by Mulv superinfection, provides the opportunity to study whether the virus genome is able to induce detectable virus-specific antigens in the absence of virus production. Therefore, it is logical to distinguish the study of the TATAs present on virus-producer and on nonproducer MSV tumor cells.

1. TATAs of MSV Producer Cells a. Subgroup and Type-Specific TATAs of MSV Tumors Due to the Helper Virus. The strong antigenicity of MSV-producer cells is easily demonstrated by the high level of resistance to a second challenge of MSV-sarcoma cells or antigenically related lymphomas after a primary tumor rejection (see Section II,B,l). Moreover, one can immunize previously untreated recipients with irradiated tumor cells. Such results have been reported repeatedly as well with tumors induced by H-MSV (Bubenik and Turano, 1968a; Chuat et al., 1969; Koldovsky et al., 1969), M-MSV (Fefer et al., 1967a, 1968a; Law et al., 1968a; Chuat et al., 1969; Jones et al., 1974), or K-MSV (McCoy et al., 1972a). In addition, complete cross-reaction has been demonstrated in transplantation assays between H-MSV and M-MSV or between them and Moloney lymphomas (Fefer et al., 1967a; Law et al., 1968a; Varet et al., 1968; Chuat et al., 1969; Law and Ting, 1970; Pearson et al., 1973). Conversely, mice preimmunized with the Moloney leukemia virus are protected against MSV (Fefer et d.,1967a; Burstein, 1970; Law and Ting, 1970). Furthermore, AKR made tolerant to Moloney virus by neonatal inoculation, are sensitive to MSV-oncogenesis whereas normal AKR are resistant ( ChiecoBianchi et al., 1974). Few experiments have been done to compare the TATAs of MSVtumors and those of lymphomas induced by murine leukemia viruses other than the Moloney agent. It has been mentioned that immunization against Rauscher leukemia virus (RLV) or RLV pseudotypes of MSV does not protect against M-MSV (Burstein, 1970; Law and Ting, 1970), whereas in reverse experiments, M-MSV could protect mice against the

16


graft of Friend, Rauscher (Fefer et al., 1967a), or Graffi (Levy et al., 1968) virus-induced lymphomas. On the whole, it seem that the M-MSVinduced immunity is regularly stronger for Moloney than for Rauscher, Friend, or Graffi lymphomas, and that the highly malignant and fast growing sarcomas induced by M-MSV inoculation could only be prevented by anti-Moloney vaccination. These observations suggest that several different antigens are involved as TATAs in these experiments, one being common to MSV Friend, Moloney, Rauscher, and Graffi tumors, and the other mainly type-specific. Similar remarks have been made in comparative studies on the induction of transplantation resistance against these various lymphomas (Bianco et al., 1966; Levy et al., 1968; Ting et aZ., 1974). The subgroupspecific TATA shared by MSV, Moloney, Friend, Graffi, and Rauscher tumors evidences the same pattern of specificities that has been also serologically defined for the FMR or FMRGi antigen (Old et al., 1964; Levy et al., 1968; see also Section IV,A,l), The type-specific TATA could be related to the viral envelope antigens (VEA) of type C budding particles as suggested by the observation that UV-irradiated Rauscher virus can be used to vaccinate BALB/c mice against a further challenge of MSV and Rauscher virus-producer cells (Stephenson and Aaronson, 1972). Recent experiments in which in d u o protection induced by Graffi and Moloney MSV pseudotypes have been compared, confirm that both subgroup-( possibly FMRGi ) and type-specific antigens ( possibly VEA) are involved as TATAs in MSV tumors (Leclerc et aZ., 1976). Any cross-reaction between Gross virus-induced lymphomas and H- or M-MSV sarcomas, has never been reported in mice. The demonstration that the FBJ virus (Finkel et al., 1966), initially extracted from a spontaneous osteosarcoma of CF1 mice, is a natural MSV pseudotype (see Section III,A,l) of which the helper agent is an endogenous “Gross type” virus (Kelloff et aZ., 1969; J. A. Levy et al., 1973), has provided a new model. It is remarkable that the sarcomas induced by the FBJ-MSV are slowly growing tumors, seldom rejected, this contrasts with the usual characteristics of H-MSV or M-MSV malignancies. The same properties are also found in tumors that are due to pseudotypes naturally appearing in AKR, after these mice have resisted a M-MSV inoculation (ChiecoBianchi et al., 1974; see also Section 11,A). All the FBJ tumors share a common TATA, which can be demonstrated by preimmunization of the recipient by irradiated syngeneic tumor cells, or surgical excision of developing subcutaneous tumor grafts (Jones and Moore, 1973), but they elicit only a low level of resistance, never exceeding more than lo4 or lo5 living tumor cells, which is much less than in the M-MSV system. The possibility that a G-type MSV tumor could grow in AKR that have


17

resisted the M-MSV challenge further illustrates the subgroup specificity of the in uivo transplantation immunity in the MSV system. b. Group-Specific TATAs of MSV Tumors Due to the Helper Virus. It was recently demonstrated that the gsl (species-specific) and gs3 (interspecies specific) antigens of Mulv, which are associated with the major internal polypeptide of the virion (p30) are also represented at the host cell surface (Nowinski et al., 1972; Yoshiki et al., 1973, 1974; Ferrer, 1973). Although the experiments were generally conducted with lymphomas, it has been shown that the antigens exist also on MSV producer cells (Ferrer, 1973; Epstein and Knight, 1975). Rats are able to recognize gsl and gs3 and to produce a high level of specific antibodies (Geering et al., 1966, 1968). It has also been suggested that the gs specificities could play a major role in cell-mediated antilymphoma reactions in rats (see Section V,D). Therefore, it would not be surprising if gsl, gs3 antigens could be TATAs in tumors induced by murine leukemia or sarcoma viruses, in rats. More recently, it has been shown that murine lymphocytes sensitized in vitro by p30 extracted from Moloney virus, protect immunosuppressed recipients against a M-MSV challenge (Gorczynski and Knight, 1975b). This result is somewhat surprising for at least two reasons: ( 1) In contrast to rats, mice react poorly against the p30-associated antigens gsl-gs3 as revealed by antibody production (Geering et al., 1966). ( 2 ) As seen above, in transplantation there are no cross-reactions between the tumors induced by Gross and Moloney pseudotypes of the MSV, whereas both viruses possess the same p30 group antigen (Schafer et al., 1969; Strand and August, 1974). Therefore, the conclusion that gsl-gs3 antigens are TATAs in mice must be considered carefully, since it cannot be completely excluded that contamination by a polypeptide other than p30 might have occurred in these experiments. Moreover, Gorczynski and Knight (197513) have observed that the immune lymphocytes from In viuo regressors do not react strongly in uitro with p30, which could suggest that the protection obtained after in uitro sensitization would not be necessarily relevant to the in uivo situation. In conclusion, it is possible that p30-associated gsl-gs3 play the role of weak TATAs in mice, but usually they can be only minor components of the in uiuo reaction. By contrast, they could perfectly well be TATAs in rats, but these animals are not the natural hosts of murine viruses, and they recognize many antigens that are not, or only poorly, immunogenic for mice (Geering et al., 1966; Igel et al., 1967; Levy et al., 1968; Ferrer, 1973). The rat system must be considered as an especial, artificial model of the naturally induced tumors. c. Sarcoma-Specific Antigens as TATA. Antigens specific to the MSV-

18


transforming genome can be defined only in transformed but not in leukemia-virus-producer cells ( see Section II,C,2), However, it must be mentioned that the “sarcoma-specific antigen” ( SCSA ) described in nonproducer cells is also expressed in producer cells (Aoki et aZ., 1973). It seems well established that this antigen does not act as TATA in nonproducer cells (see Section II,C,2), but it could play this role when associated with other specificities in producer cells. d . Antigens of Endogenous Viruses as TATAs. It has been proposed that antigens due to the expression of etiologically unrelated type C endogenous viruses could be responsible for the cell-mediated anti-MSV reactions that are detected in uitro (see Section V,D,S,d). However, it is unlikely that the same antigens could be major TATAs in uivo since they exist on different tumors which evidence no cross immunity in transplantation assays. On the whole, it can be concluded that (1) TATAs exist in all MSVinduced producer tumors that have been tested so far. ( 2 ) Their exact number is not known, but probably both subgroup- and type-specific TATAs are involved. In addition, group-specific antigens could play a minor role as TATA in mice, and a more important role in rats. (3) The TATAs associated with naturally occurring “Gross-type” MSV are always less potent than the TATAs induced by H or M-MSV. 2. Search for TATAs on MSV-Transformed but Nonproducer ( N P ) Cells a. Zmmunosensitiuity. The existence of TATA in MSV-NP cells remains debatable. A weak transplantation immunity was first reported in rats challenged with the MSB1-NP tumor after preimmunization by Moloney or MSV infection (Ting, 1967). Similarly, a small number of the “NP” H-MSV-induced hamster tumor cells PD4T.1. can be rejected in hosts that have been preimmunized by MSV producer cells or H-MSV infection (McCoy et al., 1972b). Similar results were reported for the XM1 M-MSV-induced murine hemangiosarcoma ( Law and Ting, 1970). However, these experiments are not definitely demonstrative since the recipients have been previously infected with viruses which can replicate in uiuo, at least in rats and mice, so that finally the immunizing virus could have superinfected and antigenically converted the challenged “NP” cells. It has been shown in another system that an in uiuo superinfection of NP MSV tumors by Rauscher virus provokes a tumor rejection, even in animals that already bear palpable tumors (Greenberger and Aaronson, 1973). Furthermore, the interpretation of the above experiments is especially difficult, since XM1 occasionally produce MSV and/or Mulv (Law et d., 1974), and that MSB.1 produce both a focus-forming


19

virus MSV ( 0 ) (Ting, 1968), and a rat tropic helper agent (Aaronson, 1971). Similarly, PD4T.1 possess the group-specific antfgen of hamster leukemia virus, although the role of this agent was not demonstrated in the immunosensitivity experiments (McCoy et al., 197213). b. Zmmunogenicity. More convincing arguments could be drawn from the demonstration that NP cells could induce an immune response in noninfected recipients. X-Irradiated PD4T.1. cells protect against a further challenge by viable cells (McCoy et al., 1972b), and the surgical excision of XM-1 tumors similarly immunizes the recipients against XM-1 cells (Law and Ting, 1970). In both cases, the protection is very weak. The TATAs are probably different on XM-1 and PD4T-1 since their radiosensitivities are completely different. A further argument in favor of the existence of TATA in PD4T-1 cells was obtained from the observation that X-ray treatment of the recipients enhanced the growth of these NP cells ( McCoy et al., 1972b). In contradiction with these results, other groups completely fail to reveal any immunogenicity of MSV-NP cells. An X-ray pretreatment of the recipients does not modify the growth of K.MSV NP cells (K.234) but enhances the growth of K234-superinfected producer cells ( Stephenson and Aaronson, 1972). Despite some contradictory results (McCoy et al., 1974), it appears that no TATA can be evidenced in K-MSV transformed BALB/c or rat NP cells (Stephenson and Aaronson, 1972). This is especially remarkable, since SCSA have been demonstrated in these cells by immunoelectron microscopy (IEM) (Aoki et al., 1973, 1974a) with the same immunization procedure that fails to reveal any TATA (Greenberger et al., 1974). Similarly, with the exception of one test in which a very weak and not certainly significant protection was found, all the attempts of Chuat et al. (1969) to protect mice against syngeneic MSV tumor cells by preimmunization with hamster NP cells were negative. On the whole, the only possible conclusion at the present time is that no TATA have been really demonstrated in NP MSV-transformed cells although they possess other non-TATA surface antigens (see Section V,A,2). Special attention must be given to the fact that weak protection after in uioo NP-cell immunizations is not necessarily significant when it has been passed in uitro, since such weak protection can also be found by virus-free 3T3 normal cells (Stephenson and Aaronson, 1972; Greenberger et al., 1974). It can be due to antigenic changes in tissue culture cells or to the expression of embryonic antigens. However, the second hypothesis is less probable, since BALB/c 3T3 normal cells do not immunize against hematopoietic fetal cells, whereas K-MSV-transformed cells do ( Salinas and Hanna, 1974).

20


3. Embryonic Antigens Acting as TATA on M SV-Transf orrned Cells

In recent years, it has been shown that fetal or embryonic antigens are very frequently, perhaps constantly, present on experimental tumor cells, including RNA virus-induced lymphomas and sarcomas (Hanna et al., 1971; Ishimoto and Ito, 1972; Ting et al., 1972; Ting and Herberman, 1974). Only one experiment with MSV tumor has been reported showing that embryonic specificities are effectively present on K-MSV cultured cells transformed in vitro (Salinas and Hanna, 1974). These antigens can function as TATAs and immunize mice, so that after irradiation and reconstitution by fetal liver cells, the recipients will develop a reduced number of hematopoietic spleen colonies. The role of the same antigen in tumor immunology is not known (see Section II,C,2). 4. Conclusion

From the data presently available, one can draw the following conclusions. 1. MSV producer cells are antigenic in viuo, the reactions being directed against mainly virus products present on the host cell surface, as discussed above (end of Section II,C,l ). Several different viral antigens can be involved with different immunogenicities, according to the helperassociated virus. 2. Fetal or embryonic antigens are very likely associated with the viral antigens, but all data coincide to affirm that they are only very weak TATA in this system and are possibly inefficient. 3. Sarcoma-specific antigens probably exist ( see Section II,C,l,c), but they seem not to be functional in transplantation immunity in NP cells. 4. Other antigens can be detected when immunized with in vitrocultured cells. These reactions are not related to the MSV genome, nor to any specific antitumor reactions, and they must be considered as probable artifacts due to experimental conditions. 5. The absence, or at least the very weak activity, of cell-surface antigens in MSV-transformed NP cells, contrasts with the regular presence of highly immunogenic antigens, in cells transformed by DNA viruses (Habel and Eddy, 1963; Koch and Sabin, 1963; Sjogren et al., 1961; Trentin and Bryan, 1966). This suggests that MSV-transformed cells when NP, could be very bad targets for immunologic rejection, which does not exclude the possibility that they could be a good model for the approach of naturally occurring tumors.

MURINE SARCOMA VBUS-INDUCED TUMOR

21

111. Analytical Study of the Antivirus Immune Response

A. ANTIGENSOF

THE

VIRAL PARTICLES 1. MSV Pseudotypes

The MSV virions can be strongly antigenic for their host as shown by the rapid appearance of neutralizing antibodies in adult mice infected by M-MSV or H-MSV (Fefer et al., 1967d, 1968a; Bubenik and Turano, 1968b; Law et al., 1968a; McCoy et al., 1972a; Pearson et al., 1973). However, the MSV being a defective agent, the reactions studied are mainly directed against antigens of the helper virus, which seems to provide notably the antigenic components of the viral envelope. Recently, some isolates of supposed competent MSV have been described that would allow one to determine whether some antigenic component of the virion could be due to the MSV, not to the helper genome. Available information about “competent MSV” is still too scarce to allow of answering the question ( Ball et al., 1973; Lo and Ball, 1974). The MSV defectiveness (Hartley and Rowe, 1966) allows one to obtain MSV-transformed NP cells in hamsters, rats, and mice. These cells possess the MSV genome, but do not produce any type C virus. It is easy to rescue the MSV genome by superinfecting with a MuLV, which produces an MSV pseudotype, with the envelope specificities of the helper (Huebner et al., 1966). This property has been used to produce in vitro,a large number of pseudotypes, the envelope of which was one of various MuLV including Moloney, Friend, Rauscher, Gross (Huebner et aZ., 1966), Kaplan Rad. leukemia virus (Igel et al., 1967), Graffi (Oppenheim et al., 1968), Tennant, Rich, LLV (Gomard et al., 1973). The oncogenic ability of these pseudotypes is due to the transforming MSV genome, but their infectivity is directed by the helper genome, and they are neutralized by anti-helper virus antisera (Hartley et al., 1970). Complete cross-reactivity exists between MSV pseudotypes and their helper, without any additional specificity that could be related to the MSV genome, Therefore, MSV pseudotypes have been frequently used to study the antigens of MuLV, taking advantage of the transforming ability of the MSV genome which provides in vitro as well as in V ~ U Oa rapidly detectable marker of the biological activity of the virus (Igel et al., 1967; Levy et al., 1969; Ferrer, 1973; Gomard et al., 1973). Similarly, other groups have studied the Mulv envelope antigens by using pseudotypes of the spleen focus-forming agent ( Eckner and Steeves, 1971,1972; Steeves and Axelrad, 1967).

22


The rescue of the MSV genome by a competent MuLV is possible not only in vitro but also in uiuo. AKR mice resistant to the M-MSV produce after M-MSV infection an oncogenic pseudotype that exhibits the antigenic properties of Gross or AKR viruses (Chieco-Bianchi et al., 1974). The same antigenic properties were found for the naturally occurring FBJ-MSV: it is neutralized by the serum of tumor-bearing mice and by rat anti-AKR, but not by rat anti-M-MSV (Kelloff et al., 1969; J. A. Levy et al., 1973). Complement fixation with mouse antisera also confirms this antigenicity (Jones and Moore, 1974a), By contrast, H-MSV and M-MSV appear as natural pseudotypes of the Moloney virus, being completely cross-reactive with each other and with this virus (Fefer et al., 1967a)b;Chuat et al., 1969; Strouk et al., 1972; Gomard et al., 1973; Pearson et al., 1973). Another pseudotype with the same antigenicity has also been isolated from a spontaneous tumor of ( NZW X NZB ) F, mice (Gazdar et al., 1972a,b). However, contamination by another virus cannot be excluded. It would have been interesting to know the original antigenicity of this pseudotype as well as that of a recently described natural MSV isolate from old BALB/c (Peters et al., 1974). This would allow one to determine whether only Gross type MSV or both Gross and other, more immunogenic, pseudotypes can be responsible for spontaneous sarcomas in mice. The in uivo production of MSV pseudotypes has been observed not only in mice, but also in other species of which an endogenous C-type virus may perfectly well rescue the MSV genome. Such, heterologous pseudotypes are notably known in rats and hamsters (see Harvey and East, 1971). 2. Diversity of the Virion Antigens

It is well established that multiple antigens are simultaneously present on the virus particles (viral envelope antigens, VEAs) and inside the virion. They are associated with different proteins or glycoproteins, each of them bearing several antigenic activities with type, group (species), or interspecies specificities (see, for review, August et al., 1974; Bolognesi et al., 1974; Lilly and Steeves, 1974). The virion antigens have been extensively described in MuLV, and, as discussed above, they have no distinctive features in the MSV virions, since they are directed by the helper virus genome (see Table I ) , The VEAs, which are associated with the major glycoprotein gp69/71 of the virion surface, are the most important specificities for the natural antivirus immune reaction, since they react with neutralizing antibodies. Group specific antigens associated with p15 are also present at the cell surface but apparently not concerned in virus neutralization (Lee and Ihle, 1975); the other antigens are


23

less important for the natural protection of the host, being internal components that are not represented on the surface of the virion. Several reviews have been recently published which summarize the main properties of the VEAs (Eckner and Steeves, 1972; Gomard et al., 1973; Aoki, 1974; Bauer, 1974; Levy, 1974). The essential characteristics of the VEA are the following: ( a ) They are associated with the virion surface projections, the main component of which is gp69171 (Witter et al., 1973; Hunsmann et al., 1974). ( b ) They can be revealed by virus neutralization (Eckner and Steeves, 1972; Gomard et al., 1973), or by IEM ( Aoki et al., 1974b), or by immunoprecipitation of the whole virus (Ihle et al., 1974). ( c ) These antigens are probably important for virus infectivity, and they are involved in virus-induced hemagglutination ( Witter et al., 1973; Hunsmann et al., 1974). ( d ) Several different specificities are associated on the same virions. A complete classification of the VEAs is still not available in murine C-type viruses. However, there is general agreement that both type and group-specific antigens exist on the virus envelope. Type-specific VEAs allow the recognition of individual strains in MuLV by mouse-typing serum (Eckner and Steeves, 1972; Gomard et al., 1973; Aoki, 1974). They are specific to Gross, AKR, Moloney, Rauscher, Friend, and Graffi viruses, etc., and therefore they may be named G-VEA, AKR-VEA, M-VEA, R-VEA, F-VEA, Gi-VEA, etc. However, some relationship among different MuLV has frequently been observed so that even with mouse-typing serum, cross neutralizations are frequent ( Eckner and Steeves, 1972; Gomard et al., 1973). These could be due either to the association of several virion populations, since nonclonal viruses have been used, or to the existence of shared subtype specificities ( Aoki, 1974). Nevertheless, G-VEA and AKR-VEA appear completely different from R-VEA, F-VEA, M-VEA, or Gi-VEA in all published experiments. On the other hand, clear separations of Gi-VEA from M-VEA (Levy et al., 1969) or of M-VEA from F-VEA (Bassin et al., 1973; Aoki et al., 1974b) have also been reported. By contrast, the group-specific VEAs (gs VEA) are common to all strains of Mulv, and they are mainly recognized by rat antisera, or other heterologous typing sera (Ferfer et al., 1967d; Levy et al., 1969; Ferrer, 1973; Gomard et al., 1973; Aoki, 1974). Their existence has been confirmed by the observation that GP69/71 possess a speciesspecific and an interspecies-specific determinant that are common to the different types of MuLV (Hunsmann et al., 1974; Strand and August, 1974). For this reason, rat-typing sera must be used carefully in the identification of MuLV, all the MSV pseudotypes being neutralized up to some degree by any anti-MuLV or anti-MSV antisera (Fefer et al., 1967d; Levy et al., 1969, Ferrer, 1973; Gomard et al., 1973). Aoki et al.

24

J. P. LEVY AND J.

C. LECLERC

(1974b) have described by IEM two different group-specific VEAs, named, respectively gs-VEA, which is common to all C-type murine viruses and identified by the serum of normal NZB mice, and MuLVVEA, identified by rat anti-MuLV typing serum, The last one could be present on all MuLV, but not on the MuMAV (murine myelomaassociated virus), which are produced notably by some murine plasmocytomas. The relationship between these two antigens and the previously described gs.VEA is not known, nor is their relationship with the species and interspecies determinants of gp69/71. On the whole, one can conclude that MSV, like any other RNA type-C virus of mice, possess different kinds of VEA, which could play different parts in the natural antivirus response of the host. Type-specific reactions are very predominant in mice, but both type- and group-specific reactions exist as well in other species infected with any MSV pseudotype. B. ANTI-MSV NEUTRALIZING ANTIBODIES Efficient neutralizing antibodies are obtained in MSV hyperimmunized mice (Gomard et al., 1973; Pearson et al., 1973) or rats (Bubenik et al., 1969; Gomard et al., 1973; Igel et al., 1967, Levy et al., 1969). They are helpful for typing. However, it is probably more interesting for comprehension of the autochthonous antivirus response to study the neutralizing antibodies in progressor or regressor mice which received only one virus inoculation, or those of naturally infected animals. Neutralizing antibodies are easily detected in the serum of M-MSV regressors (Fefer et al., 1967a,b, 1968a; Law et al., 1968a; Pearson et al., 1973)) or in mice inoculated with H-MSV (Bubenik and Turano, 1968b) or K-MSV (McCoy et al., 1972a). The activity is already found in BALB/c between 8 and 15 days after infection (Fefer et al., 1968a), and sometimes earlier even with low dose inocula. After an initial rise of antibody titer, the maximum is reached around day 30, and high levels of neutralizing antibodies persist several months thereafter (Fefer et al., 1968a; Lamon et al., 1 9 7 3 ~ )During . this evolution, a transitory fall was observed at the maximum tumor size ( Lamon et al., 1 9 7 3 ~ ) . Newborns with progressively growing tumors have no detectable neutralizing antibodies (Fefer et al., 1968a). However, these antibodies are found in both the rare newborn regressors (Fefer et al., 1967b), and rare adult progressors (Fefer et al., 1968a). The neutralizing activity exists both in 7 S and in 19 S fractions, but it is higher in 19 S on day 30, and in 7 S from 3 to 6 months after the MSV infection (Lamon et al., 1 9 7 3 ~ )The . production of neutralizing antibodies is not equivalent in all inbred lines. Notably, C57BL/6 or their F, hybrids with BALB/c are


25

much better responders than BALB/c (Fefer et al., 196%). Similarly C3H and (NZB X NZW)Fl infected by G,-MSV produce a high level of neutralizing antibodies whereas the antibody level remains low in BALB/c ( Gazdar et al., 1973). C. In Viuo ROLEOF ANTI-MSVNEUTRALIZING ANTIBODIES

1. Antivirus Activity The rapid appearance of neutralizing antibodies explains the early disappearance of the virus from tumors of MSV-infected adults, and the very low levels of infectious virus found in their blood and spleen, whereas high levels are detectable in both organs of inoculated newborns (Fefer et al., 1968a). In G,-MSV infected mice the disappearance of the virus in tumors and blood plasma correlates the evolution of neutralizing antibodies (Gazdar et al., 1973). The virus neutralization could have two different kinds of protective effects in uiuo: ( a ) They would play a major role in nonautonomous tumors dependent on continuous cell transformation (see Section II,A), since by limiting the spreading of virus they would allow natural cure of the disease. ( b ) They could have a minor role if autonomous malignant cells exist, The virus neutralization would limit the number of initially transformed cells without acting against the multiplication of transformed cells.

2. Antitumor Activity A direct role of neutralizing antibodies in the destruction of tumor cells is less probable. In theory, one could imagine that anti-VEA antibodies would be able to react with the cell-surface budding particles to destroy the host cells. However, cytotoxic complement-fixing antibodies are different from the neutralizing antibodies in the MSV system (Levy et al., 1969; Lamon et al., 1973c) as well as in other tumor systems (Pasternak, 1967; Steeves, 19sS). No direct information is available that could support the role of antivirus-neutralizing antibodies in the immune destruction of MSV tumor cells. It has been reported that the best neutralizing antisera were also the most effective in the in uiuo protection against virus-induced oncogenesis, but no correlation can be established between these activities and the effect of the sera on transplanted tumor growth ( Bubenik et al., 1969). The comparison between tumor evolution and the kinetics of neutralizing antibodies synthesis does not allow a solution to the problem. A detectable level of neutralizing antibodies appears in BALB/c at a monment when the capacity to reject transplanted syngeneic sarcoma cells happens for the first time before the

26


beginning of the primary tumor regression (Fefer et al., 1968a)b). However, I F and cytotoxic antibodies become detectable at the same time (Fefer et al., 1968a), and it would be impossible to conclude from these observations whether any are involved in tumor cell elimination. In fact, “when individual mice are studied, no relationship can be found between antibody production, presence or absence of tumor, tumor size, regression, and time of tumor regression, not only in normal but in X-ray treated animals” (Fefer et al., 1967b). Similarly, no clear correlation can be established between the enhancement of MSV tumor growth in immunosuppressed animals and the level of neutralizing antibodies in the same mice. After X-ray treatment, the antibody response is delayed and could be completely abolished in BALB/c (Fefer et aZ., 1967b); this correlates with the high level of progressors found in irradiated BALB/c. However, despite 1001%MSV progressors, thymectomized CBA evince only a very weak decrease of neutralizing antibodies, in comparison with intact mice (Collavo et al., 1974); this does not support the idea of a major role of these antibodies in the antitumor reaction, The role of neutralizing antibodies is not known in MSV tumors due to Gross-type MSV. Some weak neutralizing activity of the tumor-bearing mice has been mentioned (Kelloff et al., 1969). In conclusion, a strong antivirus immune protection exists in mice infected with MSV pseudotypes, with the possible exception of Gross-type MSV. In the example of M-MSV, the reaction certainly plays a major role in limiting virus replication and probably in helping the spontaneous elimination of nonautonomous tumors. However, a direct antitumor cell action of the neutralizing antibodies is unlikely in this system as well as their role in the cure of really autonomous tumors. Observations by Gadzar et al. (1973) with G,-MSV induced tumors are in perfect agreement with this conclusion.

IV. Analytical Study of the Antitumor Cell Reaction: Humoral Response

A. SEROLOGICALLY DEFINEDANTIGENSOF MSV TUMORCELLSURFACE

THE

1. Serological Studies of Virus-Producer Cells

The cells that produce MSV, produce at the same time a helper virus and most, if not all of the antigenic specificities are due to the helper


27

genome (see above Section II1,A). The MSV tumor antigens of the cell surface which can be studied by serological methods on producer cells, are therefore identical to those of murine leukemia, or MuLV-infected and producer cells. These antigens have been reviewed (Old and Boyse, 1965; Bauer, 1974; Levy, 1974), and they will not be extensively discussed here, since they are not specific to the MSV system. All available information in 1974 will be found notably in the exhaustive review of Bauer (1974). Five different kinds of cell surface antigens exist and can be detected simultaneously at the surface of murine C type oncornavirusinduced tumors (Table I ) . One can suppose that the situation could be the same on MSV-tumor cells, though relatively little information is available concerning this system. a. Viral Envelope Antigens ( V E A ) of the Budding Particles. VEA have been previously discussed ( Section III,A,2). As shown above, they are multiple on the same viron, and it is possible, while not definitely established, that they might have a weak effect in transplantation immunity (see Section II,C,l,a), They can be identified at the host cell surface by IEM (Aoki et al., 1970, 1974b), but apparently are not involved in complement-dependent cytotoxicity in the FMRGi system ( see Section I1I7C,2). Similarly, in the Gross system, the G-VEA and the Grossspecific cell-surface antigen, GSCA ( a ) are completely different ( Aoki et al., 1970). b. Viral Cell-Surface Antigens ( VCSA), VCSA are the specificities that can be detected at the cell surface, in morphologically normal areas, outside virus particles. They are independent of VEA, as discussed above. By serological methods, including in uitro complement-dependent cytotoxicity and immunofluorescence on living cells, two main specificities have been described in murine oncornavirus tumors, FMR (FriendMoloney-Rauscher ), also called FMRGi ( the same plus Graffi ), and GCSA ( a ) , which is specific to the Gross virus (Old et al., 1964, 1965; Aoki and Johnson, 1972). There is no doubt that FMRGi is present on M-MSV and H-MSV tumor cells, as shown by immunofluorescence with regressor or hyperimmune anti-MSV sera (Fefer et al., 1967a, 1967c, 1968a; Law et al., 1968a; Chuat et al., 1969; Koldovsky et aZ., 1969) or by complement-dependent cytotoxicity with regressor sera ( Lamon et al., 1973~).By contrast, the progresssively growing tumors of AKR mice (Chieco-Bianchi et al., 1974) and the FBJ tumors bear the GCSA ( a ) (Jones and Moore, 197413). Supposedly these antigens are also the TATAs defined by in uivo experiments (see Section II,C,l,a), and this assumption is supported by the parallelism of in uivo and in uitro detected specificities with clear separation between FMRGi and GCSA ( a ) systems. However, no direct evidence of the identity between GCSA ( a ) or

28


FMRGi and the TATA has been yet given, and the problem remains unsolved. More recently, it has been suggested that the FMRGi antigen could be an association of several different specificities (Ting and Herberman 1974), and even its autonomy has been discussed (Cerny and Essex, 1974). Group-specific cell-surface antigens exist in murine oncornavirusinduced tumors, in association with FMRGi or GCSA ( a ) (see Bauer, 1974). At least one specificity has been relatively well defined by serological procedures: the GCSA ( b ) antigen that rat typing sera can recognize on both GCSA ( a ) and FMRGi cells (Geering et al., 1966). It probably explains the cross-reactions observed between the two kinds of tumors with rat typing sera (Geering et al., 1966; Levy et d.,1969; Ferrer, 1973; Herberman, 1972). More recently, it has been shown that gsl and gs3 antigens, which are group specific and associated with the well defined internal viral protein p30, are also present at the cell surface as discussed in Section II,C,l,b, They could be identical to GCSA ( b ) . However, it seems unlikely that they could be the target of the antitumor cell reactions detected in the MSV tumors in mice since they are very poorly immunogenic for this animal. Finally, the possible expression of different viral polypeptides at the cell surface during virus replication including not only gp69/71 and p30, but also p15, and perhaps p12 (see, for review, August et al., 1974; Bolognesi et al., 1974; Lilly and Steeves, 1974), could account for the presence of several other antigenic specificities, the significance of which remains presently unknown in the antitumor response. c. Endogenous Virus-Directed Cell-Surface Antigens. Endogenous type C viruses can be derepressed in tumor cells, and their antigens would then be expressed at the cell surface among other virus-induced specificities. The recently described MEV. SA1 (Herberman et a!., 1974) has been suggested as the target of cell-mediated reactions, but it is not revealed by serology. However, the expression of endogenous viral antigens at the cell surface could explain some weak cross-reactions that have been sometimes observed in the mouse system (Klein and Klein, 1964; Rich et al., 1965; Levy et al., 1968) and especially the weak but definite capacity of M-MSV-induced tumors to absorb anti-GCSA ( a ) antibodies ( Chieco-Bianchi et al., 1974). d . Sarcoma-Specific Cell-Surface Antigen ( SCSA). As discussed in Section II,C,l,c, these antigens could exist in virus-producer cells, but they can be identified only in NP cells, with which they will be studied ( Section IV,A,B). e. Embryonic Antigens. The existence of these antigens has been demonstrated by transplantation immunity in MSV tumor cells (Section

+

29


11,C73).However, no information is available about a possible antibody response directed against these antigens in the MSV system. 2. Serological Studies of MSV-Transformed Nonproducer Cells The NP cells studied by transplantation immunity (Section II,C,2) were also used in serological experiments with the common goal to define a SCSA directed by the MSV genome, independently of any helper virus. In the serological approach, completely negative results have been regularly reported except in IEM. When any other serological methods were used, SCSA could not be detected on NP cells, even in the groups that reported positive results in transplantation immunity (see Section 11,C72).NP cells that do not induce the formation of neutralizing antibodies (McCoy et al., 1972b; Stephenson and Aaronson, 1972) are also unable to induce antitumor cell antibodies, detectable by immunofluorescence or complement-dependent cytotoxicity ( McCoy et al., 1972b, 1974; Chuat et al., 1969). Conversely, when anti-MSV or anti-MLV potent sera are tested by the same methods or by complement fixation against NP cells, they fail to detect any specific antigenicity (Chuat et al., 1969; Stephenson and Aaronson, 1972; McCoy et al., 1974). Even the very sensitive mixed hemadsorption test was negative with S Lcells, which possess the MSV genome and produce incomplete virions (Strouk et al., 1972). The rare positive reactions sometimes detected by complement fixation or immunofluorescence on living cells ( Stephenson and Aaronson, 1972; McCoy et al., 1974), are due to unrelated antigens, since the activity can be absorbed, not only by MSV-transformed, but also by normal 3T3 cells ( Stephenson and Aaronson, 1972). In IEM, hyperimmune sera, obtained by repeated and massive immunizations with mitomycin C-treated NP cells, have suggested the existence of SCSA (Aoki et al., 1973). Their specific activity can be absorbed by K-MSV- or M-MSV-transformed NP cells of rats or mice. Moreover, the same antigen is absent on Rauscher virus-induced leukemia cells. Unfortunately, no information about the existence of the SCSA in Moloney virus-infected cells has been reported. The SCSA persists in MSV-NP cells when they are superinfected by Rauscher virus to become producers, and it can be occasionally found in small areas of the virion envelope of a few virions, probably owing to the incorporation of antigens preexisting at the cell surface, as also demonstrated with H-2 antigens (Aoki and Takahashi, 1972). Specific absorptions allow the separation of SCSA from helper-induced antigens at the producer cell surface ( Aoki and Takahashi, 1972).

+

30


More recently, four different SCSA specificities have been described by the same group (Aoki et al., 1974a) and again by IEM. SCSA ( a ) would be independent of the MSV system, being specific to the woolly monkey sarcoma virus ( WMV)-transformed cells. The three other specificities were associated with K-MSV-induced tumors, but their significance is not clear. SCSA ( d ) could be identical to the previously described SCSA ( Aoki et al., 1973) since it exists in K-MSV rat and mouse NP cells. However, it is also present in rat cells transformed by the WMV, suggesting either that a common gene is shared by WMV and MSV, or that the SCSA might not be directed by the transforming genomes. The existence of the same SCSA ( d ) antigen in mouse cells transformed by the M-MSV would be valuable information to further clarify this problem and, at least, to say whether SCSA (Aoki et al., 1973) and SCSA ( d ) are the same or not. The two other specificities, SCSA ( b ) and SCSA ( c ) are still less understandable at present, since they are common to cells transformed by the WMV, and some, but not all, of the cells transformed by the K-MSV. They could be the result of various interactions of the viruses with different host cell genomes ( Aoki et al., 1974a), but their exact nature remains questionable. On the whole, it can be concluded that at least one SCSA directed by the MSV genome, could exist on the MSV-transformed cells. However, further studies are necessary to establish definitely that SCSA are undoubtedly due to the MSV transforming genome and could therefore be considered as the first example of a tumor-specific antigen in this system. It would be of special interest to determine the possible relationship of SCSA and the well-defined polypeptides of the virions. A recently described sarcoma-specific antigen of avian oncornavirusinduced tumors, could be an interesting model for the study of the SCSA in the MSV system (Gelderblom et al., 1972; Kurth and Bauer, 1972a,b), Whatever the nature of SCSA, the reasons why this antigen is not involved in tumor rejection in duo, as shown in Section II,C,2, are not known. This is a general problem, since tumor-associated cell surface antigens do not necessarily function as TATAs, especially in the case of oncornavirus-induced tumors ( Levy, 1974). In the present case, several hypotheses could be proposed: ( a ) An antigenic modulation could arise in uiuo, as for several other tumor antigens (Boyse et al., 1967; Aoki and Johnson, 1972; Ortaldo et al., 1974). Such a modulation is not necessarily detectable in in uitro experiments, conducted in a short period, as shown with GCSA ( a ) (Aoki and Johnson, 1972) and embryonic antigens of murine lymphomas (Ortaldo et al., 1974). ( b ) The antigens may be too scarce in N P cells to produce an efficient reaction. ( c ) SCSA could be


31

only weakly efficient in transplantation, with no possibility of showing this activity by the available in vivo method.

B. ANTITUMORCELL-REACTING ANTIBODIES Hyperimmune sera of highly immunized anti M-MSV animals (Pearson et al., 1973; Jones et al., 1974) as well as syngeneic regressor sera (Fefer et al., 1967a,b, Pearson et al., 1973) contain antitumor cell antibodies, which can be revealed by indirect immunofluorescence (IF) on living lymphoma cells of FMRGi antigenicity. Progressor mice produce only a low level of IF antibodies in comparison with regressor (Fefer et al., 1967c; Pearson et al., 1973). Newborns that become regressors have a detectable level of antibodies (Fefer et al., 1967b). The IF antibody production is not equivalent in all inbred strains; for instance, C57BL/6 are better responders than BALB/c, as was also found for neutralizing antibodies (Fefer et al., 1967b; Shachat et al., 1968). The kinetics of IF antibody synthesis parallels that of neutralizing antibodies. They are detectable 8 days after M-MSV inoculation and 100%of the recipients are positive on day 12 (Fefer et al., 1968a). However, the maximum titer seems to be detectable only later, and persists . for a long period, at least from 3 to 6 months (Lamon et al., 1 9 7 3 ~ )The activity has been detected only in the 7 S fraction (Lamon et al., 1 9 7 3 ~ ) . Complement-fixing cytotoxic antibodies have been also detected in the MSV system, but negative results are particularly frequent with this test, in our experience, as well as with other groups (Pearson et al., 1973). However, Lamon et al. ( 1 9 7 3 ~ )were able to detect a good level of cytotoxic antibodies in the response to primary-induced neoplasms in BALB/c. This activity was first detected around day 10, reaches a peak on day 15 past infection, at the moment of the maximum tumor size, and falls in early regressor sera to the level of controls. Thereafter, a progressive rise follows, reaching a good activity around day 30, and a maximum level of cytotoxic antibodies in long-range regressors (3-6 months). Both 7 S and 19 S fractions are cytotoxic all along the evolution (Lamon et al., 1 9 7 3 ~ ) . A cytotoxic activity of the regressor sera has also been detected during a 2-day in vitro culture of tumor cells (Chuat et al., 1969; Stephenson and Aaronson, 1972; Tamerius and Hellstrom, 1974). This method is probably particularly sensitive, and it would test an activity of 7 S immunoglobulins which is present at a high level, until 125 days after the virus infection. However, a certain degree of nonspecific reaction could also be involved in this test (Tamerius and Hellstrom 1974).

32


C. In Vivo ROLEOF

THE

ANTITUMORCELLANTIBODIES

As discussed in Section II,B,4,a, the serum of tumor regressor mice protects in vivo not only against MSV-virus but also against a tumor cell graft. However, the latter protection is relatively weak, suggesting that antitumor cell antibodies could play some role in tumor rejection, but that other, probably cell-mediated phenomenas, are more important. The kinetics of the antitumor cell antibodies is of poor help to solve this problem. They appear in parallel with neutralizing antibodies, during the antitumor response and no precise experiments have been reported in which any protective activity in vivo has been compared with the in vitro properties of the different Ig fractions. Tamerius and Hellstrom (1974) have shown that the titer of cytotoxic activity remains lower in progressor than in regressor sera, but the correlation is not clear enough to allow a definite conclusion. More suggestive evidence has been given by Gazdar et a2. (1973) since a good correlation exists in BALB/c, C3H, and (NZB X NZW)Fl inoculated with G,-MSV between the tumor evolution and the level of cytotoxic antibodies. Little information is available on the evolution of antitumor cell antibodies in immunosuppressed mice. X-Irradiation decreases the synthesis of IF antibodies, especially in the poor responder BALB/c mice (Fefer et al., 1967b,c), but curiously enough, cortisone acetate treatment, which enhances tumor growth (see Section II,B,2,a), did not decrease the level of IF antibody synthesis (Shachat et al., 1968). This does not support an important role of these antibodies in vivo. The observation that nude (nulnu) mice do not produce a significant level of IF antibodies, in contrast with 100%of the nu/+ controls (Davis, 1975), is more suggestive, but not conclusive. In fact, at the present time, a role played by cytotoxic or other antitumor cell antibodies in tumor rejection appears possible, but not fully demonstrated.

V. Analytical Study of the Antitumor Response: Cell-Mediated Immunity ( C M I )

A. DETECTION OF CMI BY THE COLONY-INHIBITION TEST(CIT) THE MICROCYTOTOXICITY ASSAY ( M A )

AND

The colony-inhibition test ( Hellstrom, 1967) and the related microcytotoxicity assay ( MA) (Takasugi and Klein, 1970) were the first methods to reveal a cell-mediated antitumor immunity, and the MSV system was one of their first applications (Hellstrom and Hellstrom, 1969;


33

Hellstrom et al., 1969a). From a historical point of view, these experiments opened a new era in tumor immunology, with the description of “cell-mediated antitumor reactions in experimental as well as in numerous human tumor systems. More recent experiments have shown that MA and CIT are complex methods that test simultaneously different parameters including nonspecific phenomena ( Seeger et al., 1974), and for this reason more-analytical methods will probably replace MA and CIT in future experiments. However much valuable information has been obtained by the MA, which was for several years the most extensively used in experimental and human tumor immunology. a. Kinetics of Lymphoid Cell Activity in MA. Spleen or lymph node cells of mice inoculated with M-MSV inhibit in uitro the growth of M-MSV tumor cells (Hellstrom et al., 1969a; Hellstrom and Hellstrom, 1969). The activity exists both in progressors and regressors, with the exception of very young animals less than 7 days old. The kinetics of the lymphoid cell activity is biphasic, with a first peak at the beginning of the tumor growth, followed by a fall at the moment when the tumor reaches its maximum tumor size, then by a second peak after tumor rejection (Lamon et al., 1972a; Plata et al., 1974). Later, the activity of the lymphoid cells decreases slowly. Other groups described different kinetics with only one peak (Herberman et al., 1975a). The reasons for this discrepancy are not known. b. Heterogenicity of Effector Cells. The effector cells in MA are multiple, involving T and non-T cells, T cells were found to be responsible for a part of the activity during the first peak and the beginning of the second one, but non-T cells appear as the only effectors at a later period, around day 2s past infection (Lamon et al., 1972b, 1973a; Plata et al., 1974). The nature of non-T cells is still debatable. From the results of Lamon et al. (1973a,b), which are also confirmed by Plata et al. (1974)) they are poorly phagocytic and nonadherent, so that it has been proposed that they could be B cells (Lamon et al., 1973a,b). However, in similar culture conditions it has been shown that macrophages could play a determinant role by a strong cytostatic action on tumor cells (Owen and Seeger, 1973; Seeger et al., 1974). The only possible conclusion at the present time is that different non-T cell populations are probably involved in the MA. c. Specific and Nonspecific Reactions. There is no doubt that the phenomena tested by MA are very complex, and the final results are due in large part to nonspecific reactions that involve macrophages and perhaps other non-T cells (Owen and Seeger, 1973; Seeger et al., 1974). However, the tumor specificity of the T-cell-mediated reactions appears

34


to be better in the same experiments. Nonproductive MSV-transformed cells (Strouk et al., 1972; McCoy et al., 1974) or cells that produce rare and defective virus particles ( Lamon et al., 197213) or methylcholanthrene induced sarcoma cells (Kall and Hellstrom, 1975) are not inhibited in the MA by anti-MSV immune lymphoid cells. This suggests that viral products of the tumor cell surface are specifically involved in the reaction. However, it cannot be excluded in such experiments that a specific immune reaction directed against virus particles would have induced an activation of non-T cells with cytostatic effect. In fact, the constant association of cytolysis and cytostasis in MA makes difficult any interpretation of results since cytostatic effects are mainly nonspecific (Seeger et al., 1974); see also Section V,B,l). In these conditions any specific immune reaction that could arise in the culture would perhaps be able to trigger the nonspecific cytostasis of tumor cells, even if the starter specific reaction is not related to an antitumor immune response. The cytolytic part of the phenomena revealed by MA is probably much more specific (Owen and Seeger, 1973; Seeger et al., 1974) as also demonstrated by more analytical methods that measure only tumor-cell cytolysis (see Section V,B,2). One can suppose that this part of the reaction is one of the possible specific triggering of the non-T cell-mediated cytostasis. In the usual conditions of the MA, specific and nonspecific reactions cannot be separated clearly enough to determine the significance and the quantitative level of specific phenomena. In fact, practically nothing is known about the probable specificity of T cell-mediated reaction in MA. For the same reason, the significance of the effector B cells described by Lamon et al. (1973a,b) cannot be definitely determined from the presently available data. The problem is especially difficult with these cells because, contrary to T-effector lymphocytes, they have not yet been detected by methods other than the MA. d. Significance of the Results Obtained in M A and in CIT. The main practical interest of the MA was to reveal the complexity of the anti-MSV immune response and the possibility of inhibiting the CMI in vitro by soluble factors (see Section V,C,l). However, this complexity and the association of specific and nonspecific reactions limit the value of this test so that more analytical methods are necessary to study the specific anti MSV-tumor response.

B. STUDYOF

THE

ANTITUMORCMI

BY

ANALYTICALMETHODS

We shall consider as “analytical methods” those that measure separately tumor-cell cytostasis or tumor-cell cytolysis. In addition we shall include in the same section the information obtained by measuring the


35

stimulation of lymphoid cells, or macrophage-migration inhibition by tumor antigens, since these two methods “analyze” an early event at the recognition stage of the immune response. 1. Measure of Tumor-Cell Cytostasis The cytostasis of tumor cells in absence of immune cytolysis has been studied by different procedures using sarcoma cells in monolayers (Owen and Seeger, 1973; Seeger et al., 1974) or lymphoma cells in suspension (Senik et al., 1974a,b; Kirchner et al., 1975a,b). From all the experiments, it can be concluded that during the coculture with immune lymphoid cells a very strong cytostatic activity appears progressively; it becomes maximum after 3-4 days in vitro and inhibits tumor-cell multiplication, This activity is due to non-T cells, which have all the characteristics of macrophages. Some tumor specificity of [’HI thymidine incorporation inhibition has been reported when very low lymphoid : target cell ratios were used (Senik et al., 1974a). However, most of the cytostatic activity is certainly nonspecific (Seeger et al., 1974; Senik et al., 1975a; Kirchner et al., 1975b). In addition it has been suggested that a decreased [3H]thymidine incorporation does not necessarily correlate the inhibition of tumor cell proliferation (Kirschner et al., 1975b). On the whole, these results are important since they emphasize the poor significance of any cytostatic phenomena for the study of cellmediated specific antitumor reactions. Their in vivo significance is not established: we do not know whether the same non-T cell-mediated reactions occur in vivo. They could play a nonspecific but determinant role inside the tumor in amplifying the specific reaction, but on the other hand, the same effector could also be immunosuppressive, counteracting the antitumor response (see Section V1,A). Finally, it is not completely excluded that such a strong activation of non-T cells could be due in large part to the in vitro experimental conditions with Iittle relevance in uiuo. However it is remarkable that in regressors the tumors are infiltrated by macrophages proving in vitro a nonspecific cytostatic activity (Holden et al., 1976). Macrophages are 4 to 8 times less abundant in progressor tumors and they remain mainly at the periphery (Russel et al., 1975). 2. Measure

of Tumor-Cell Cytolysis

a. Methods. The most widely used technique to obtain pure cytolysis of tumor cells by immune lymphocytes is the chromium release test (CRT) initially described in an allogeneic system by Brunner et al. (1968).

36


The target cells are uniformly labeled by the cytoplasmic marker, which is released during immune cytolysis. The results obtained in these conditions are very sensitive and clear-cut. The marker does not have to be incorporated into the DNA or nucleic acid metabolism, so that mitotic activity does not interact with the labeling. This method is especially convenient when highly sensitive lymphoma cells in suspension are used as targets, with short incubation periods. By contrast with less sensitive sarcoma cells grown in monolayers, the CRT is not as convenient, owing to the high level of spontaneous chromium release after 24 hours in uitro. The labeling by [3H]proline (Bean et al., 1973) is a better procedure when longer incubation periods are necessary, since the spontaneous elution of [3H]pr~lineis very weak during at least 48 hours, even with dividing cells. The labeling by [sH]proline having otherwise the same advantages over chromium labeling, the proline assay ( P A ) appears to be a good method to measure tumor-cell cytolysis with weakly sensitive target cells. A comparative study of PA and CRT shows that both tests measure probably the same phenomena involving a pure T-cell population (Gomard et al., 1976d). Other isotopic methods have been proposed (Seeger and Owen, 1973; Fossati et al., 1975). The results obtained in this tumor system by the different methods, which measure immune cytolysis, are summarized as follows. b. Kinetics of the in Viuo Cytolytic Actiuity. In CRT, the effector cells are detectable about day 8 after M-MSV infection, and their activity reaches a peak at about day 15, when the tumor size is at its maximum. After tumor rejection, the cytolytic activity decreases slowly to become very weak, sometimes undetectable, at about days 45 to 50 (Leclerc et al., 1972b; Lavrin et al., 1973). In the rare case where the tumors are spontaneously delayed and appear at about days 40 to 60 post inoculation, the activity is detectable at that time (Leclerc et al., 197213). After a booster virus inoculation, the cytotoxic activity in the spleen is detected only in the rare animals that develop a recurrent tumor (Leclerc et al., 197213). By contrast, the MSV-immune animals that resist a second virus challenge have no detectable splenic activity (Leclerc et al., 1972b). These observations could have cast some doubt about an in uiuo role of the cytolytic effector cells in antitumor immune protection. In fact, it seems that a high level of circulating cytolytic cells detectable in the spleen needs a very strong antigenic stimulation, as only produced by a large tumor-cell mass. Therefore, these high activities will be detected only when primary virus-induced tumors are palpable or after lymphoma cell grafts, which immediately stimulate a large number of immunogenic cells (Leclerc et al., 1972a; Holden et al., 1975). This situation is comparable to an allogeneic graft. By contrast, a second virus inoculation


37

does not induce a large number of tumor cells, owing notably to virus neutralization, so that activity will not be detectable in the spleen; but it could perhaps be detected around the site of virus inoculation, and in regional lymphoid organs, as shown after lymphoma cell booster (Holden et al., 1975). Furthermore, the fact that only weak cytolytic activity exists in strongly immune mice that have rejected the tumor is in agreement with the observation that cytotoxic T lymphocytes (CTL) lose their activity when not in contact with tumor cells (Senik et al., 1975a), but are rapidly restimulated by a second contact. The secondary CTL response may supposedly eliminate any reappearing cell clone without needing a large number of circulating CTL. A different kinetics of CTL during anti-MSV primary response has been reported (Gorczynski and Knight, 1975a), with a decrease in the activity after initiation of tumor rejection. These results are in opposition with those obtained in CRT against lymphoma cells in suspension, not only in the MSV system in mice and rats (Leclerc et al., 1972b; Lavrin et al., 1973; Veit and Feldman, 1975), but also in the rat anti-Gross response (Oren ct al., 1971; Shellam, 1974). This discrepancy could be due to the nature of target cells since Gorczynski and Knight used sarcoma cells instead of lymphoma cells. It is interesting to note that by MA using sarcoma cell monolayers, a fall of lymphocyte activity is observed at the maximum of tumor size in mice (Lamon et al.,1972a; Plata et al., 1974; see Section V,A,a). These results suggest, together with other arguments, that the nature of target cells could be a determining factor in the reaction (see Sections V,C,l and V,D). The CRT activity can be detected not only in regressor but also in progressor animals, which will die with growing tumors. In this case, it reaches a peak comparable to that of regressors, but falls acutely thereafter, whereas the tumors enlarge (Leclerc et aZ., 1972b). Only very young mice less than 10 days old do not develop this response (Leclerc et al., 1972a). Some fluctuations exist among the inbred strains of mice as far as the level of the immune response is concerned, C57BL/6 for instance, being very good responders whereas Balb/c provide weaker activities. An absence of CRT response of BALB/c, C3H, and (NZB NZW)Fl inoculated with G,-MSV has been reported (Gazdar et al., 1973); However, the target cells used were allogeneic in these experiments, which probably explains this result (see Section V,D,l). c. Nature of the Effector Cells. The effector cells in these reactions are T cells as shown by anti Thy.l.2 and complement treatment, and by the elimination of macrophages or other non-T cells from the effector suspensions (Herberman et al., 1973; Leclerc et al., 1973; Plata et aZ., 1973). In opposition with the complexity of the effector cells in MA (see

x

38


Section V,A,a), only T cells are involved in CRT all during the evolution of the anti MSV immune response (Plata et al., 1973). d . Nature of the Znteraction between CTL and Tumor Cells. The immune cytolysis due to the CTL seems to be completely independent of any antibody. They are probably mediated by structural specific receptors of the CTL membrane, which are still unknown, as for any other CTL-mediated reaction. The possible existence of T cells armed in uiuo by IgM cannot be completely excluded, since in uitro experiments by Lamon et al. (197513) have shown that not only B cells but also T cells and thymocytes become cytolytic after incubation with specific IgM. These results have still not been confirmed. In all the reported experiments, some cross antigenicity has been observed between murine oncornavirus-induced tumors, when tested in CTL-mediated reactions (Leclerc et al., 1972b; Herberman et al., 1974). Unrelated syngeneic tumors do not cross-react, which demonstrates the immunologic specificity. e. In Vitro Stimulation of CTL. The CTL activity can be induced in uitro when T lymphocytes are in contact with tumor cells. A primary response is obtained when normal lymphoid cells are cocultivated in the presence of MSV-syngeneic tumor cells, or related lymphomas (Plata et al., 1975). The activity which becomes detectable on day 5 and reaches a plateau on day 8, remains relatively weak. A secondary response can be demonstrated more quickly, beginning on day 3; it has maximum values on day 5, which are at least 10 times higher than during the primary response (Plata et al., 1975). The secondary response can also be detected by a simplified method, the “secondary CRT” (SCRT) (Senik et al., 1975a), which does not allow one to detect the primary in uitro response. In all these experiments, the effector cells of the reactions are only T cells and they are probably identical to the effector CTL produced after in uiuo sensitization. Macrophages are not involved in immune cytolysis, but they seem to play some role in the stimulation of CTL, since in their absence, the secondary response cannot be obtained in SCRT (Senik et al., 1975a). The primary stimulation of CTL in uitro provides for the first time a valuable method for study of the cytolytic immune response from the recognition to the effector stage. The secondary stimulation of CTL or their progenitors confirm that CTL can be rapidly reactivated at a high level by tumor cells, suggesting that they could be a determinant in uiuo in the elimination of any recurring tumor cell clone. In addition, these methods provide a convenient way to determine the antigenic specificities that are the targets of CTL as discussed in Section V,D. f. Significance of the Methods Measuring Pure Cytolysis by CTL. The


39

methods that reveal only the CTL activity allow an antitumor cell reaction which parallels the tumor evolution. They provide clear and quantitative results about a specific immunologic phenomenon that is probably important in vivo at the effector stage of the antitumor response. One can suppose that the same CTL-mediated phenomena are also involved in the CIT and in the MA, of which they could represent the specific part. However, they are not distinguishable in these tests from the nonspecific reaction (see Section V,B,l). The possibility of blocking the CTL reactions by soluble factors (Section V,C,l), the antigenic specificities involved ( Section V,D), and the in vivo relevance of the CTL (V,E) will be discussed elsewhere.

3. Study of the Recognition Stage of the Immune Response a. Primary Response. The mixed lymphocyte-tumor cell reaction (MLTR) fails to reveal any stimulation of nonimmune lymphocytes by lymphoma cells antigenically related to the MSV tumor (Senik et d., 1973). Similar negative results have also been reported, either in MSV or in the rat anti-Gross lymphoma system (Glaser et al., 1974; Knight and Gorczynski, 1975). These results are somewhat paradoxical since normal cells can be stimulated in vitro by tumor cells to produce CTL (Plata et al., 1975). One can suppose that they are due to a relatively poor sensitivity of the method as also suggested by the recent demonstration that soluble viral antigens can effectively stimulate normal, nonimmune, lymphocytes ( Gorczynski and Knight, 1975b). b. Secondary Response. The stimulation of lymphocytes of M-MSV infected mice by tumor cells was reported by Senik et al. (1973). Tumor cells stimulate [3H]thymidine incorporation in spleen or lymph node cells. A peak activity occurs from 3 to 7 days after MSV inoculation, before the appearance of the tumor. Thereafter, the activity rapidly decreases, to become undetectable despite the appearance of a high level of CTL in the same organs. Lymphocytes from progressors were negative (Senik et al., 1973). Therefore, in this experiment the activity was detected only before the appearance of the tumor or at its beginning. However, in other experiments, it was found at the peak of the tumor (Kirschner et al., 1978) or after tumor regression in the MSV system (Gorczynski, 1974b), as well as in the rat anti-Gross system (Glaser et al., 1974). The use of different experimental conditions could explain these discrepancies, and notably the fact that Senik et al. considered only very positive reactions, the results being interpreted according to restrictive stimulation indices, which take into account the very frequent nonspecific stimulation of mouse lymphocytes by irradiated tumor cells, or even by irradiated syngeneic lymphocytes. The stimulation described

40


by these indices is very specific, but the method is not fully sensitive. One can therefore suppose that after an initial peak at the beginning of tumor evolution, weaker activities could have been overlooked later on in regressors. By contrast, the use of soluble extracts as stimulators instead of tumor cells, and the choice of less restrictive indices, would have allowed the observation of these late stimulations in other experiments (Gorczynski, 1974b). In addition, the use of soluble extracts does not necessarily reveal the same phenomena found when whole tumor cells are used. The precise significance of lymphoid cell stimulation by tumor cells (Senik et al., 1973), or by M-MSV-transformed cell extracts, or by viral products (Knight and Gorczynski, 1975) is not fully understood. The responder cells are predominantly or exclusively T cells ( Gorczynski, 1974b; A. Senik, unpublished results), as also reported in rats reacting against a Gross virus-induced lymphoma (Glaser et al., 1974). From Senik et al. (1973), who observed an inverse relationship between responder cells in MLTR and killer cells in CRT, one could imagine that responders in MLTR and killers in CRT would represent two successive stages of an in uiuo CTL differentiation. The same inverse relationship has also been found in other tumor systems ( McKhann, 1971), as well as in allograft rejections in mice (see discussion in Senik et al., 1973). However, other experiments ( Gorczynski, 1974b; Knight and Gorczynski, 1975) show that the in uitro stimulation of regressor lymphocytes is a complex phenomenon involving at the same time several responses directed against different antigens, involving notably type-specific antigens, related to the gp69/71 viral glycoproteins, and group-specific viral antigens. The group-specific reaction has been observed both with whole tumor cells and with cell extracts as stimulators (Senik et al., 1973; Knight and Gorczynski, 1975). Therefore, one cannot exclude that different specific responses could be simultaneously elicited in such experiments, including antitumor-cell reactions and other reactions directed, for instance, against the viral particles. The secondary response of MSV-infected mice has also been studied by the macrophage-migration inhibition test ( MMIT ) , Peritoneal cells obtained from mice bearing M-MSV induced tumors have their migration in culture inhibited by soluble MSV-tumor tissue extracts, whereas extracts of unrelated, chemically induced, tumors are without any effect ( Halliday, 1971, 1972). Peritoneal cells from progressors are not inhibited. Similar results have been observed with spleen and lymph node cells (Gorczynski, 1974b). T cells are essential for this activity, as shown by its abolition after anti-Thy.l.2 and complement treatment ( Gorczynski, 1974b). The same suspensions that are responders in MMIT are also responders in stimulation, but two physically distinct populations respond in stimulation whereas only one of them releases factors able to cause


41

migration inhibition ( Gorczynski, 1974b). The blocking of this activity by soluble factors or by immunosuppressive cells are considered in Sections V,C,l, and VI,A. c. Significance of Stimulation and M M I Experiments in the MSV System, The stimulation of immune lymphocytes and the MMIT reveal mainly presensitization of hosts to tumor or virus-associated antigens. They test an early event of the secondary response, which is T-cell dependent. However, it is not possible to determine the exact significance of this event, since the final functions of the cells that are stimulated remain unknown in these tests. From the in vitro stimulation experiments with CTL generation (Plata et al., 1975), one can suppose that at least some of the stimulated cells are progenitors of effector cells in tumor rejection, but it is probable that other reactions, not necessarily related to tumor-cell rejection, are also measured by these tests. The development of methods allowing study of the recognition stage of the primary response would be especially interesting to determine the significance of these reactions, as well as the antigenic specificities involved and the possible induction of tolerance at this stage of the response.

C. INTERACXIONS OF EFFECTOR CELLSAND SOLUBLE FACXORS 1. Blocking of the Cell-Mediated Reaction by Soluble Factors The inhibition of cell-mediated reaction by soluble factors is a very important problem of general significance in oncology since, in theory, it could represent one of the main efficient mechanisms that enable tumors to escape immunologic control. In fact, several different kinds of “blocking” must be distinguished according to the stage of the response that is blocked: effector or recognition. In addition, at the final stage, the blocking can concern target or effector cells. a. Blocking at the Level of the Eflector Cells i. Observations in MA. In the MSV system, blocking factors were described in the progressor sera since the first reports of a cell-mediated antitumor reaction: they inhibit in vitro the antitumor effect of lymphoid cells (Hellstrom and Hellstrom, 1969). These experiments were done by the CIT, and then by MA. The blocking factors were first supposed to be antibodies acting at the level of the tumor cells; but they were rapidly demonstrated to be circulating antigen-antibody complexes acting at the effector cell surface (Sjogren et al., 1971). They are present in the serum of tumor-bearing mice during tumor growth and they disappear from the regressor sera taken several days after tumor rejection. By contrast, regressor sera are “unblocking,” which means that mixed with progressor

42

J. P. LEVY AND J . C. LECLERC

sera, they suppress the blocking effect (Hellstrom et al., 1969b; Hellstrom and Hellstrom, 1970). Other groups have observed in MA the same blocking ability of the serum of MSV-tumor-bearing animals (Skurzak et al., 1972; Plata and Levy, 1974), but the correlations with the progressor or regressor status of the donor were sometimes less clear (Pollack and Nelson, 1973), and contradictory effects of the same serum were found according to the dilution used (Skurzak et al., 1972). These observations suggest that the mechanism of the reaction tested in MA could be at least for one part an antibody-dependent cell-mediated cytotoxicity ( ADCC ) . This is confirmed by the demonstration that nonimmune lymphoid cells can be armed in vitro by some sera (Pollack et al., 1972; Pollack, 1973; Pollack and Nelson, 1974). One can suppose that the antibody is the determinant factor in arming of lymphocytes by antigenantibody complexes. By contrast, the determining element in blocking is an antigen, and the blocking can be obtained in MA not only by the serum of tumor-bearing mice, but also by soluble tumor antigens (Plata and Levy, 1974). In these conditions, it is easy to understand that the blocking factor would be present in excess in the serum of mice that bear a large tumor, whereas the “unblocking effect” of regressor sera would be due to an excess of free antibodies. The complexity of the MA (see Section V,A,) does not allow one to pinpoint what part of this reaction is concerned by blocking Several questions could be asked, notably: does blocking concern specific cytolysis or nonspecific cytostasis? Does it involve ADCC only or also direct CMI due to CTL? No definite answers can be given to these questions, but several points can be noted. It is probable that ADCC account in large part for blocking as discussed above, but one can hardly explain all the results by ADCC inhibition. It has been shown, for instance, that T cells are also blocked in MA (Plata and Levy, 1974) whereas ADCC in other systems that have been studied are regularly due to non-T cells (see, for review, MacLennan, 1972; Perlmann et al., 1972; Forman and Moller, 1973). However, these experiments do not specify whether the T cells that are blocked in MA are the specific CTL or other T cells responsible for other immune reactions. Comparative experiments would favor the second hypothesis (Plata and Levy, 1975), but recent results in blocking of CTL support the first hypothesis (see the next section). It has been shown that pure cytostatic reactions can also be blocked (Senik et al., 1974a) in the MSV system. Finally, it is probable that blocking concerns different reactions, including specific antitumor response, but the experimental conditions of MA limit the interpretation of results obtained in such experiments. ii. Blocking in CTL cytolytic reactions. The experiments that have


43

been done in CRT, with lymphoma cells as targets, failed to reveal any blocking of the CTL by soluble antigens (Leclerc et al., 1973; Plata et al., 1974), whereas aliquots of the same effector cell suspensions were blocked in MA (Plata and Levy, 1974). Therefore, it is possible that two different T-cell populations would be active in MA and in CRT. However another hypothesis would be that sarcoma cells, which are the target in MA, and lymphoma cells, which are the target in CRT, could be of partly different antigenicity: the blocking factor could be related to an antigen present on sarcoma but not on lymphoma cells. Furthermore, there is direct evidences that CTL can be blocked using sarcoma cells as targets in CRT (Gorczynski and Knight, 1975a). This blocking was obtained by serum of early regressors or by viral antigens including the p30 viral polypeptide. Similarly, recent experiments by our group indicated that the CTL reaction measured in the PA against sarcoma cells could be blocked, emphasizing once again the possible determining role of the target-cell nature ( Gomard et al., 1976d). Further arguments for the existence of in uitro blocking of CTL in oncornavirus-induced tumors can be drawn from the study of the rat anti-Gross lymphoma. This reaction which is due to T cells (Shellam, 1974; Djeu et al., 1974) follows the same kinetics as the anti-MSV reaction in mice (Oren et al., 1971; Shellam, 1974), and can be inhibited by progressor sera, by the whole virus, and by various soluble antigens, including the p30 polypeptide (Shellam, 1974; Shellam and Knight, 1974; Knight et al., 1975). However, rats react against antigens that are not recognized by mice (see Sections II,C,l,a and IV,A,l), so that conclusions obtained with rats are not necessarily valid for mice. Moreover, it would be important to specify in this system whether a direct CTL reaction or an ADCC is inhibited. A non-T-cell-mediated cytolysis is not excluded in rats (Veit and Feldman, 1975; Ortiz de Landazuri et al., 1974a,b,c), since an arming effect of progressor sera has been observed in CRT independently of the CMI. Blocking has also been observed with such sera by the same methods (Shellam and Knight, 1974; Knight et al., 1975), and further investigations are necessary to determine the nature of the reactions observed in rats. b. Blocking at the Level of the Target Cells. It must be emphasized that any blocking in the syngeneic MSV system, whether of ADCC or of CTL reactions, is due to antigens or antigen-antibody complexes. The blocking of the tumor-cell surface by antibodies competing with effector cells has never been documented in this system. Its in uiuo relevance appears very unlikely, notably because very high doses of reacting antibodies are necessary to obtain such blocking in vitro, whereas these doses are never observed in viuo during autochthonous tumor develop-

44

J. P. LEVY AND J.

C. LECLERC

ment. The in vivo sensitization of tumor cells by antibodies could in fact be a possible way to induced an ADCC ( see Section V,C,2). c. Blocking at the Recognition Stage. Halliday (1971, 1972) using M M I and Gorczynski et al. (1975) using stimulation tests reported that soluble antigens or sera of tumor-bearing mice inhibit the anti-MSV response. Such blocking at an early stage of the reaction can be determinant in viuo, but we do not know its real importance since the phenomena evinced in these two tests are still poorly understood (see Section V,B,3). A nonspecific blocking of recognition stage has been described in other systems, in which it involves an a-globulin (Glaser and Herberman, 1974), but it has not yet been documented in the MSV system. 2. Arming of Normal Lymphoid Cells The possible arming of nonimmune cells to produce an ADCC in the MSV system has been discussed above (see Section V,C,l,a). It has been demonstrated both in uitro (Pollack et al., 1972; Pollack and Nelson, 1974; Lamon et al., 1974, 1975a) and in viuo (Pollack, 1973). Progressor sera and, in some experiments, regressor sera, have been found to be arming. The cells involved include non-T cells, possibly B cells (Lamon et al., 1974), but T cells armed with IgM could also be included ( Lamon et al., 1975b). Most experiments have been done in MA, suggesting that cytostatic nonspecific reactions are the result of ADCC, as shown in experiments using purely cytostatic methods (Senik et al., 1974a). However, an ADCC can be also responsible for pure cytolysis in the MSV system (Harada et al., 1973, 1975), as is well known in other systems ( MacLennan, 1972; Perlmann et al., 1972; Forman and Moller, 1973). These results show that an ADCC can be associated with cytostasis or CTL-mediated cytolysis in any reaction, so that the nature of the effector cells must be precisely determined in each experiment, whatever the method used. 3. Conclusions on the Interactions of Soluble Factors with Effector Cells Both arming of nonimmune cells and blocking of various effectors can be observed in the MSV system, The ADCC induced by arming of normal cells represent a possible in uivo antitumor specific action of antibodies, since all the specificity of such reactions is supported by antibodies. However, the in vivo relevance of ADCC is still unknown: the relatively weak protection against tumor cells observed after in viuo transfer of immune sera (see Section II,B,4,a) suggests that its role is probably limited. It is highly probable that blocking by soluble antigens, or antigenantibody complexes with antigen excess, concern not only ADCC, but


45

also the cell-mediated reactions due to CTL. Precise experimental procedures, including the use of sarcoma cells as targets, seem to be necessary to reveal the CTL blockings, which are probably most important for the in uiuo reaction, according to their tumor-specificity ( see Section V,B,2). However, further experiments are still necessary to specify in uitro conditions of blocking and its in uiuo relevance. No experiments have been reported in the MSV system that suggest that in uiuo manipulations could act on the blocking factor to facilitate tumor rejection. Some limited observations have been made that suggest that injection in uiuo of regressor sera in progressor mice can increase the activity of CTL in the recipients (J. P. Levy et al., 1973), but no correlations have been made with tumor evolution.

D. ANTIGENICSPECIFICITIESINVOLVED IN CMI The antigens of the tumor-cell surface that react with CTL are still poorly known. It has not been definitely demonstrated in the MSV system whether they are related to TATAs and/or to serologically defined specificities. Even their number is not determined. A precise solution to these problems is a necessary prerequisite to establish, by comparison with the TATAs of the same tumors, what is the in uiuo significance of the in uitro observed CTL reactions. Direct cytolysis of tumor cells by CTL in CRT experiments does not provide conclusive evidence about the antigens involved, at least for the following two reasons: ( a ) All the cells are not valid targets, some of them being too sensitive, and many others too resistant, to immune cytolysis. Target cells in monolayers are especially resistant. ( b ) The CRT reaction is weak when CTL and tumor cells are allogeneic. It is therefore necessary to use indirect methods, and three possibilities are available : ( i ) the inhibition of CTL-mediated cytolysis by an excess of antigenically related tumor cells (Ortiz de Landazuri and Herberman, 1972); (ii) the in uitro secondary stimulation of CTL by tumor cells, in the SCRT (Senik et al., 1975b), or other related methods. Inhibition and stimulation of CTL study two different stages of their evolution. This is possibly the reason why results obtained by both methods are not identical, with notably better specificity and higher sensitivity in SCRT than in the inhibition test (Senik et al., 197513). (iii) The blocking of CTL reactions by well-defined soluble tumor-associated antigens is poorly documented, but its first results appear to be promising (Gorczynski and Knight, 1975a; Gorczynski et al., 1975). Comparison of results obtained by these different methods suggests that several antigens are simultaneously involved in CTL-mediated reactions,

46

J. P. LEVY A N D J. C. LECLERC

which is not surprising in view of the antigenic complexity of the MSV tumor-cell surface (see Section IV,A,l). In fact two different points must be considered: on the one hand, it is clear that the antigens reacting with syngeneic anti-MSV-CTL are in some part H-2 molecules altered by viral products. On the other hand, the nature of the viral antigen(s) involved in relationship with H-2 in these phenomena remains questionable. These two points will be studied successively.

1. Role of H - 2 Antigens in thc Interaction between Anti-MSV-CTL and Syngeneic Tumor Cells Anti-MSV-CTL from tumor-bearing mice is efficient mainly against syngeneic tumor cells (Ortiz de Landazuri and Herberman 1972; Senik et al., 1975b). Activity can be detected with allogeneic FMRGi ( + ) target cells only when highly potent CTL are used and it remains very weak in comparison with the syngeneic activity. The comparison of CTL from different inbred lines including congenic lines which differ only in the H-2 complex shows that this allogeneic restriction is due to H-2 antigens (Gomard et al., 1976a). The use of CTL coming from F1 hybrids immunized against FMRGi (+) parental tumor cells reveals that mice are sensitized by an “H-2 MSV modified antigen which is different in the two parental lines. Hybrids are able to lyse tumor cells sharing the same H-2 with the immunizing tumor and not those of the other parental line. Therefore the allogeneic restriction is due to the fact that CTL recognizes a complex structure on the tumor cell surface which involves both H-2 molecules and viral specificities. Anti-H-2 antibodies inhibit specifically the interaction of anti-MSV-CTL and syngeneic target cells (Gomard et al., 1976b). By comparing the reactivity of anti-MSV-CTL from different H-2 recombinants lines one can demonstrate that an identical H-2D subregion on CTL and the tumors cells is sufficient to allow a . of the K end of the good interaction (Gomard et al., 1 9 7 6 ~ )Products H-2 region are also concerned in most cases, but the respective roles of the K and I subregions have not been identified. On the whole, the most important point is that CTL does not recognize at the cell surface VCSAs identical to these reacting with antibodies, but something which associates H-2 and viral specificities (Gomard et aZ., 1975 a and c). 2. Role of Viral Specificities in the Interaction between Anti-MSV-CTL and Tumor Cells a. FMRGi or GCSA ( a ) Antigens and C T L Reactions. Cross reactivities have been occasionally observed in CRT between GCSA ( a ) ( +) and FMRGi ( +) lymphomas (Leclerc et al., 1972b), and they are also

MURINE SARCOMA VJRUS-INDUCED TUMOR

47

known in stimulation or MMI tests. However, systematic experiments have shown that other antigens than GCSA and FMRGi are involved. In fact, in most cases, anti-M-MSV CTL do not react against GCSA ( a ) ( ) and FRMGi ( -) cells. The inhibition test (Herberman et al., 1974) and the SCRT (Senik et al., 197513) confirm that FMRGi and GCSA ( a ) (+) lymphomas do not cross-react. Systematic SCRT experiments show that in uiuo or in uitro maintained tumor cells, as well as monolayers of productivity infected normal cells, are able to restimulate anti-M-MSV CTL provided they bear the FMRGi specificity (Senik et al., 1975b). Both allogeneic and syngeneic cells induce CTL stimulation in SCRT, in opposition with results of the inhibition test (Herberman et al., 1974; Senik et al., 197513). The stimulation by allogeneic cells remains very weak in comparison with syngeneic cells as shown by quantitative experiments (Plata et al., to be published), however it allows correlation of the secondary CTL response with FMRGi specificities. One can suppose that something is revealed in SCRT that is not detected in the inhibition test, owing to the different phenomena tested by both methods. Therefore, it may be concluded that FMRGi antigens, whatever their exact significance as discussed elsewhere (see Section IV,A,l), are probably among the main targets in CTL reactions. This conclusion supports the idea of an important role of CTL in uiuo, since the same pattern of specificity is found in transplantation immunity ( see Section II,C,l,a). It would be very interesting to know whether anti GCSA ( a ) CTLmediated reactions also exist in mice, in view of the evolution of the naturally occurring GCSA (+) sarcomas (see Section II,C,l). Unfortunately, this question remans undocumented. b. p30-Associated Antigens and C T L Reactions. The group-specific antigens associated with the p30 internal polypeptide of the virion have been discussed elsewhere in relation to in vim (see Section II,C,l,b) and serologically defined anti-MSV reactions (see Section IV,A,l). It has been recently reported that 1130 of the Moloney leukemia virus blocks the CTL anti-M-MSV reaction (Gorczynski and Knight, 1975a). The CTL blocking was not extensively documented in this report, but it shows that CTL anti-p30 probably exist in mice infected by the M-MSV. Their exact role is still not known. One can suppose that such CTL explain only a weak part of the CTL response, since it is mainly subgroup specific (see (Section V,D,2,a), whereas the major determinants of p30 are species and interspecies specific (August et al., 1974; Bolognesi et al., 1974; Lilly and Steeves, 1974). However, another determinant of p30 could also be involved. In addition, one cannot absolutely exclude the hypothesis that another polypeptides of the virion with subgroup specificity could have been present as a contaminant in the above-mentioned experiments ( Gorczynski and Knight, 1975a).

+

48


c. Type-Specific VEAs, SCSAs, a d CTL Reactions. The comparison of anti-Friend (Leclerc and Gomard, 1975) and anti-M-MSV CTL shows that ( a ) both react against the same FMRGi (+)lymphoma cells due to Friend, Moloney, Rauscher or Graffi viruses; ( b ) only anti-M-MSV CTL are able to lyse sarcoma cells in the PA (Gomard et aZ., 1976d). The latter result correlates with in uiuo experiments, since only anti-MMSV protect the recipients against M-MSV oncogenesis (Leclerc et al., 1976). Such a pattern of specificities cannot be explained by FMRGi, nor by p30-associated antigens. They suggest the possible role of a sarcoma-specific antigen (SCSA) as described in Sections II,C,2 and IV,A,2. However, SCRT experiments failed to reveal any sarcoma-cell specificity in the CTL secondary reaction (Senik et aZ., 1975b). Further experiments are necessary to define this point, and the possible intervention of type-specific VEAs in uitro as well in uiuo (see Sections II,C,l and 111,A). d . Antigens of Endogenous Type-C Viruses and CTL Reactions. Antigens of endogenous viruses can be expressed at the tumor-cell surface, whatever the tumor etiology. It has been proposed recently that they could be responsible for the CTL-mediated anti-M-MSV response (Herberman et al., 1974). This hypothesis was supported by two main arguments: ( i ) Allogeneic FMRGi ( +) cells do not inhibit anti M-MSV CTL. Opposite results found in SCRT (Section V,B,2a) do not support this argument, Moreover, the existence of an H-2 barrier to the expression of CTL-mediated cytolysis can better explain these results (Section V,D,l). (ii) Some cross reactivities are found, in the inhibition test, between tumor cells that produce endogenous virus without relationship with the FMRGi system, and M-MSV CTL. The latter observation must be carefully interpreted, since such cross-reactions are generally weak and are mainly observed at low CTL: inhibiting cell ratios. In these experimental conditions, nonspecific blocking is also frequent. Moreover, the same results were not obtained in SCRT: despite the very high sensitivity of this test, FMRGi (-) cells do not stimulate the activity of anti M-MSV CTL, even if they produce endogenous N or B tropic viruses (Senik et aZ., 1975b). On the whole, it appears that the role of endogenous virus-directed antigens requires further documentation. If their role was to be confirmed in CTL reactions, these antigens would possibly explain some cross-reactions between tumor cells that are not apparently related to other antigens. However, they would not explain the whole CTL reactivity, as confirmed by the fact that blocking is generally partial. Also, they would be of minor importance in uiuo since their in uitro defined pattern of specificities does not correlate with the in uiuo protection.

MUXINE SARCOMA VIRUS-INDUCED TUMOR

49

e. Conclusions. The CMI in the MSV system is probably directed against several antigens of the tumor-cell surface which are probably associated with H-2 modified antigens. Among them antigens related to the classical “FMRGi system” are certainly involved. In addition, typespecific antigens possibly related to the VEA, group-specific antigens associated to p30, and perhaps endogenous virus-directed specificities could be concerned, as well as possibly other antigens to be defined, including embryonic specificities. These suggest that different immune CTL subpopulations are simultaneously present. Moreover, other specificities could perhaps be involved in ADCC reactions, but practically nothing is known about them. The respective roles of the different antigens are not known, but the observation that FMRGi specificities are involved supports the idea of an in uiuo role of CTL, since the in uiuo reactions follow the same pattern of specificities. OF THE in Vitro E. In Viuo RELEVANCE DETECTED CELL-MEDIATED REACTIONS

Much is known about in uiuo tumor rejection and about the existence of in uitro detectable CMI. However, as a paradox, little information is available that would allow one to determine whether the in uitro phenomenas are relevant to in uiuo tumor rejection. Much of this has been discussed elsewhere in this review and will be merely summarized here. The main points are the following: 1. In uiuo transfer of immune lymphoid cells reveals that T cells are necessary for tumor rejection, but that CTL are probably not immediately efficient, in uiuo reactivation most probably being necessary and the role of other noncytolytic T cells not being excluded ( Section II,B,4,b). 2. The in uiuo kinetics of the CTL activity parallels primary tumor development and regression (Section V,B,l). 3. A long time after tumor rejection, the CTL of immune mice are only weakly efficient, but they are rapidly restimulated by contact of tumor cells, suggesting that a secondary response is able to reject any recurrent tumor clone in uiuo. Therefore, progenitors of CTL probably support in uiuo antitumor immunity (Section V,B,2,e). 4. Reactivity of anti-M-MSV CTL being regularly found against FMRGi (+) cells, whatever the possible role of other specificities, one can conclude that CTL display a similar pattern of antigenic specificities, which are also involved in in uiuo immune protection (see Sections II,C,l and V,D ) . 5. By contrast no argument exists that could support the hypothesis of a determinant role in uiuo of non-T cells in specific immune MSV-tumor

50


rejection. The transfer of non-T cells, including MA effectors and the cells responsible for cytostatic reactions, does not produce any in vivo protection. 6. T cells are found in regressing tumors in association with macrophages. These T cells evidence a typical CTL specific activity in uitro, whereas macrophages provide mainly nonspecific cytostatic reactions ( Holden et al., 1976),

VI. MSV Tumors and Immune Surveillance

A. MSV TUMORESCAPE The M-MSV tumors are not really good models for study of this problem since they are naturally rejected, MSV sarcomas of Gross-type antigenicity would be better models, since they effectively escape immune control, however, no analytical information is presently available on these systems. Some data concerning the M-MSV-induced tumors could be useful at least for a comparative study with other systems. They are as follows. 1. Blocking factors exist in the serum of tumor-bearing mice. They can inhibit not only ADCC and non-T cell reactions, but also CTL-mediated response, at least in vitro. Their in uiuo role is still not understood (see Section V,C,l); however, it is certain that they are not sufficient to induce tumor escape in the M-MSV system. 2. A nonspecific immunosuppressive effect of MSV has been reported (Chan et al., 1970), while undetectable in AKR (Chieco-Bianchi et al., 1974). It seems that it is not very strong, and even in nonregressive tumors of Gross antigenicity, it could hardly account for tumor escape. 3. The immunoselection of poorly immunogenic tumor cells could be an efficient way to escape tumor control. This mechanism could be involved in late-developing or recurring MSV tumors, but it has not been studied. The appearance of “Gross-type” MSV tumors in AKR after M-MSV infection is a variant of this phenomenon, owing to rescue of the MSV genome by endogenous viruses. 4. Immunosuppressive cells have been described in MSV-infected mice. These cells are possibly B cells (Gorczynski, 1974c; Kilburn et al., 1974) or macrophages (Kirchner et al., 1975a,b), but certainly not T cells. They inhibit in d t r o the stimulation of lymphoid cells by mitogens or antigens (Gorczynski, 1974c; Kirchner et al., 1974, 1975a; Gorczynski et al., 1975) and the MMI reaction in the presence of soluble antigens ( Halliday, 1972; Gorczynski, 1 9 7 4 ~ )These . effects are completely non-


51

specific, but they can also be detected against a specific antitumor reaction, with inhibition of the in vivo protective effect of transferred T cells ( Gorczynski and Norbury, 1974). In addition, a transient immunosuppression of the anti H-2 reaction is associated with a decrease of antiMSV response (Kiessling et al., 1974). It is remarkable that non-T cells with nonspecific cytostatic effects have been described in two different situations : “immunosuppression”in the above-mentioned experiments and “antitumor reactions” in previous reports ( see Section V,B,1). Their characteristics are identical in both cases (Kirchner et al., 1975a), so that one can suppose that the same cells are considered as “antitumor” or “immunosuppressive” according to the test used in vitro. One can hardly establish at the present time their in vivo role: tumor protection, or inhibition of the antitumor response, or both. It is not even excluded that they could be an in vitro artifact with little in vivo relevance. The immunosuppressive effect of these cells does not act on the effector cells in CRT (Kirchner et aZ., 19754. Moreover, both a strong nonspecific cytostatic activity, due to these cells, and a highly potent CTL-mediated specific cytolysis appear in cocultures of immune lymphocytes and related tumor cells (Senik et al., 1975a). Therefore, in this example, the supposed immunosuppressive cells are unable to inhibit the secondary antitumor response. Nevertheless, a suppressive effect on the secondary CTL response can be demonstrated by adding in vitro an excess of such immunosuppressive non-T cells ( Glaser et al., 1976).

B. APPEARANCEOF LEUKEMIAS AFTER REJECTION OF MSV TUMORS OF IDENTICAL ANTIGENICITY BALB/c that survive a M-MSV inoculation can become leukemic after several months. These leukemias are probably due to the Moloney virus. Knowing that all the serologically defined antigens are common to H- or M-MSV tumors and Moloney lymphomas (see Section IV,A,l) and that the TATAs are identical (see Section II,C,l,a), the development of these leukemias is not easily understandable. This model could provide valuable information about tumor escape and the relationships between solid tumors and leukemias, but it has still not been extensively studied. Some hypotheses can be proposed, without conclusions, to explain the escape of leukemic cells from antitumor surveillance : 1. The antigens important in tumor rejection are specific of sarcoma cells. 2. Poorly antigenic cells are selected by immunologic reactions. 3. The leukemia virus induces a specific tolerance, notably because T cells are their usual target,

52


4. The leukemic cells appear in protected areas without any contact with specific antibodies or CTL, notably in the thymus. On the whole, the M-MSV-induced sarcomas are possibly a good model of antigenic tumors that are spontaneously eliminated by immune surveillance (see Section VII,B), and the existence of phenomena that could counteract this surveillance are still not established, However such mechanisms could possibly exist, since leukemias of common antigenicity with MSV sarcomas can arise.

VII. General Comments

A. CONTRIBUTION OF THE MSV MODELTO THE UNDERSTANDING OF AN EFFICIENT ANTITUMORIMMUNE RESPONSE The MSV system, at least in the peculiar case of the strongly immunogenic pseudotypes, provides a remarkable model of an autochthonous tumor which is spontaneously rejected. This property is especially evident in the M-MSV-induced tumors, on which most of the studies concerning the antitumor response have been focused. The tumors induced in mice by infectious MSV inoculation are high virus producers, so that an antivirus and an antitumor-cell immune response occur simultaneously. The two responses are not easy to separate in this system. Moreover, it is probable that the antivirus reaction also has an antitumor effect in limiting tumor cell extension (see Section 111,C). The value of the MSV system as a model of immunity directed against a viral infectious disease is discussed in Section VI1,C. In the study of the antitumor immune response, the main interests of the MSV system are the following: 1. It is an autochthonous tumor that induces a strong antitumor immune response. It can be studied in syngeneic conditions, without any artificial procedures such as cell transplantation, which involves the use of in viuo or in uitro selected tumor cells especially adapted to the graft. In this system, it is possible to compare the response of various recipients, with different genetic backgrounds without histocompatibility problems. This allows genetic studies on the mechanisms of tumor induction and resistance. A similar situation rarely exists in experimental oncology and one can suppose that, for these reasons, the MSV system will be extensively studied in the future. 2. The MSV tumors possessing at their surface multiple kinds of antigens (see Section IV,A), the determination of the antigenic specificities involved in cell-mediated reactions in uiuo and in uitro is not easy. However, several different methods are now available, which allow one to


53

solve this problem (see Section V,D), so that the antigenic specificities of the in uitro reactions, and their in uivo relevance, become clearer and will most probably soon be completely determined. Then the MSV model will be especially valuable for study of the exact role of the different kinds of tumor antigens which are known: VEA and VCSA of exogenous or endogenous C-type viruses, sarcoma genome-specific antigens, embryonic antigens, etc. Moreover, comparison of producer and nonproducer MSV-tranformed cells provides a further possibility for separating these different antigens. Finally, the precise knowledge of antigenic viral polypeptides expressed at the cell surface in MSV-transformed cells, and the possible variations of this expression according to the host genetic background, will be of special interest in relation with tumor-cell immunogenecity. 3. In recent years, the study of anti-MSV tumor response has greatly contributed to the demonstration of the complexity of the cell-mediated immune response, and it has shown that the different methods supposed to test “cell-mediated immunity” do not measure, in fact, the same phenomena (see Sections V,A and B). From the experiments performed with the MSV model, we have learned, notably: the existence of effector cells of different nature, including T and non-T cells; the frequent association of specific and nontumor specific reactions; the different significance of cytolysis and cytostasis; and the possible coexistence of CTLmediated reactions with ADCC. The better understanding of the in uitro observed phenomena that results from these studies is of general significance in tumor immunology. It allows one to learn what is an efficient antitumor reaction, and what are the characteristics of a specific antitumor response, which must be looked for in animals as well as in human beings. At least the following conclusions can be drawn: a. The CTL-mediated reactions are probably determinant in tumor rejection and in tumor protection (see Section V,B,2). b. The role of antibodies is less clear (see Section IV,C), but they are probably much more protective than enhancing. No facilitating antibodies have been demonstrated (see Section V,C,l,a) , c. In addition to CTL and antibodies, other antitumor factors are detected in &TO. Non-T cells, including macrophages, are frequently involved in nonspecific reactions. Cytostatic phenomena, and perhaps cytolytic reactions of the ADCC type, are usual in tissue cultures, but their in uivo relevance is not known (see Sections V,A, and V,C,l,a). d. For these reasons, the reactions that measure a CTL-mediated cytolysis are probably the most interesting to study in tumor immunology, whereas any method that takes into account only nonspecific cytostasis, is less significant. In any case, whatever the tumor and the method, the nature of the effector cells, and as far as possible, their antigenic specifi-

54

J. P. LEVY AND J. C . LECLERC

cities must be determined before asserting that an “antitumor reaction” is really relevant to a specific antitumor immunity. This rule is certainly valuable in human tumors as recently shown by the demonstration that most of the previously described cell-mediated antitumor reactions reported in man, are in fact nonspecific (Takasugi et al., 1973; Bloom et al., 1974; De Vries et al., 1974; Peter et al., 1975). Even in experimental systems nonspecific cytolytic reactions have been detected in normal young mice ( Herberman et al., 1975b,c) or in preleukemic AKR (Gomard et al., 1974). A natural immunity possibly directed against C-type viruses cellular antigens has been also described in apparently normal mice (Kissling et al., 1975a,b). In both cases, they are mediated by non-T cells. e. Several different antigens are probably involved in tumor rejection in viuo (see Section II,C,l) as well as in CTL reactions in vitro (see Section V,B,2), All are viral products in the MSV system. f. Blocking of ADCC reactions is frequently observed, but is of probably poor in uiuo relevance in tumor evolution. The blocking of CTL, which could be more important, is still poorly documented and requires further studies. Its in uivo significance remains to be established (see Section V,B,l). In any case, soluble antigens are the determinant factors in blocking. g. Immunosuppressive cells of nonthymic origin with nonspecific activities have been described during tumor development. They are similar to the cells that are supposed to exert a cytostatic antitumor effect. The reality of their immunosuppressive role remains yet to be established, notably as far as specific antitumor response is concerned (see Section V1,A). h. CTL activity is dependent on the presence of tumor cells, and they lack efficiency when tumor cells are absent. However, a rapid restimulation of CTL occurs in the contact of antigenically related tumor cells (see Section V,B,2,e). These conclusions are probably valid not only for the MSV system, but also for other tumors including human tumors. As far as an experimental model can be useful to direct human research, points a-h above could help to avoid, in man, descriptions of “tumor-specific reactions” that would be, in fact, nontumor-specific phenomena.

B. GENERAL VALUE OF THE MSV MODEL IN TUMOR IMMUNOLOGY The MSV tumor is a good model for study of the mechanism of in vivo tumor cell elimination, as discussed above. However, it is a very peculiar


55

tumor, the value of which as a general model in tumor immunology remains questionable. In fact, if one considers the evolution of an MSVinduced tumor, it seems that different cases have to be considered. each one being possibly a different tumor model, of unequal value (see Section I1,A). a. The MSV as a Model of Antitumor Immune Surveillance. The inoculation of M-MSV in usual experimental conditions induces the rapid formation, in considerable numbers, of virus-infected and probably malignant cells. This situation is obviously different from the spontaneous appearance and slow progression of rare tumor cell clones. For this reason, the tumor becomes palpable and can reach a considerable size, even in resistant adults, before the fully developed immune response. The same would be true with Friend virus-induced disease or with large transplants of syngeneic tumor cells in mice or rats. In some way, this situation is similar to the one for allogeneic grafts of normal histoincompatible cells. In all these situations, the foreign cells are highly immunogenic and will be rejected, but the model is frankly artificial. It can hardly be considered as a model of the antitumor reaction in spontaneously growing and already detectable tumors, but perhaps only as a model of the early spontaneous rejection of highly antigenic tumor clones. However, in natural situations such clones would never produce macroscopic tumors before their rejection, and the very high level of antitumor response, particularly CTL mediated, which is observed in the case of MSV, would probably never occur in spontaneous conditions. On the whole, the reaction observed in this case is probably a considerable amplification of an event that arises spontaneously, but at a very low level, during immune surveillance. b. The MSV as a Model of Reaction against “Nonautonomous Tumors.” In addition to the antitumor response, an antiviral response occurs simultaneously that will limit the spreading of the virus, and therefore, the malignant transformation of surrounding normal cells. This procedure is the most important in anti-M-MSV tumor control, since it is possible (see Section I1,A) that most MSV-infected and transformed cells of these early tumors might not be really autonomous. Therefore, these tumors will be self-limited when the virus spreading stops. This part of the reaction cannot really be considered as a model in spontaneous tumor immunology, except for the rare cases of virus-produced tumors of limited malignant potency. c. The MSV as a Model of Reaction against Autonomous Tumors. Among the nonautonomous MSV-transformed cells, one can expect that really autonomous malignant clones will appear, as shown both in uiuo and in uitro (see Section 11,A). These clones would probably be the best

56


models for the study of naturally occurring tumors. In fact, three different situations exist in the MSV system: 1. Malignant clones arise inside the M-MSV or other MSV-producer and highly antigenic tumors. Even if they are not virus producers they will be superinfected and antigenically converted by the virus products. The in uivo immune response being essentially related to these virus products (see Section II,C,l), the immune CTL will probably rapidly destroy such clones. In this case, the antitumor CTL response is only one aspect of the antivirus-infected-cell immunity. In some way, this is an example of a spontaneous viral oncolysis. 2. The malignant clones do not produce Moloney helper virus, but the MSV genome is rescued by endogenous viruses, This situation, which apparently occurs naturally in uiuo, results in a progressively growing tumor, owing to the poor immunogenicity of ecotropic endogenous viruses. This would be the best model for the study of spontaneous tumors. Practically nothing is known about the possibility of an immune tumor rejection in this system and its hypothetical mechanism. Further studies of this model could be very interesting to determine the immune response against poorly antigenic tumors. 3. The malignant clones are completely nonproducers as observed with some in uitro transformed cells. As far as we know, such clones are very weakly immunogenic (see Section II,C,Z), and it is very difficult in this system to study an antitumor immune response that is extraordinarily weak, if not null. This does not exclude the fact that nonproducer clones of MSV-transformed cells could be a good model of natural tumors. In conclusion, the M-MSV tumor is a good model for study of the mechanism of destruction of highly antigenic transformed cells, in conditions that considerably amplify a probably natural phenomenon. This is a fascinating model for understanding the immune tumor-destruction, and the reason why an animal is protected against tumor recurrence. With this system, we probably, for the first time, have evidence that CTL are very important in the antitumor immune surveillance. However, to study the protection against the tumor that spontaneously becomes clinically detectable, other models like FBJ sarcoma could be better. At the present time, the data acquired in the M-MSV or H-MSV system must be generalized very carefully in tumor immunology. C. THE MSV SYSTEM:A MODELOF AN IMMUNE RESPONSE AGAINST A VIRALINFECTIOUS DISEASE The MSV tumor is at the same time a malignant disease and an infectious disease, with spreading of the virus from producer cells to the

MTJRINE SARCOMA VIRUS-INDUCXD TUMOR

57

surrounding normal cells. Two different procedures that tend to limit the virus replication exist in this system: ( a ) Neutralizing antibodies eliminate the free virus particles (see Section III), with transitory formation of circulating virus-antibody complexes ( Hirsch et d.,1969). ( b ) However, another procedure is necessary to eliminate the virus-producer cells: It involves the so-called “antitumor cell” antibodies and the CTL, since they react with viral products of the cell surface, which exist on all producer cells. The so-called “antitumor response” of the MSV system can be considered as an example of an antivirus-producer cell reaction. In fact, the existence of an antitumor reaction stricto sensu remains to be demonstrated in MSV-tumor-bearing mice. Therefore, the MSV disease very much resembles several other infectious viral diseases in which a carrier state of the virus can occur. In this case, the cells that are infected and virus producers, or at least producers of some viral cell-surface products, can be destroyed by an immune process ( see review, Doherty and Zinkernagel, 1974) . The demonstration that H-2 antigens modified by the MSV are the target antigens of antiMSV-CTL (see Section V,D,l) shows that H-2 antigens plays in this system the same role that was previously demonstrated with nononcogenic virus-infected cells ( Zinkernagel and Doherty, 1974, 1975; Doherty and Zinkernagel, 1975; Gardner et aZ., 1975). Alterations of H-2 molecules by viral product could be a determinant clue of the antiviral immune surveillance, by triggering the generation of specific CTL. In this way they allow the elimination of infected and tumor cells but, up to now, no argument has been given supporting the idea that nonvirus-producer tumor cells bear such modified H-2 and therefore would react with CTL.

D. SIMILARITIES OF ANTI-MSV AND ANTI-H-2 IMMUNE RFSPONSFS A remarkable point must be emphasized: the anti-MSV tumor-cell response is very similar to an immune response directed against an allogeneic graft of normal cells. Quantitative differences exist between these two reactions, but qualitatively no discrepancy has been reported. The appearance of cytotoxic or IF-detectable antibodies is very similar in both cases, but at a weaker level in the MSV system (see Section IV,B). The CTL activity is also weaker in anti-MSV than in anti-H2 response, but the same kinds of effector cells appear with similar kinetics (see Section V,B,2 and review by Cerottini and Brunner (1974). In uitro sensitization allows in both cases a high level of cytotoxic specific activity in secondary responses (see Section V,B,3 and Cerottini and Brunner, 1974). Finally, the in uiuo transfer of CTL suggests that they play a determinant role in uiuo in allograft rejection (Cerottini and

58


Brunner, 1974) as well as in MSV-tumor protection (see Section II,B,4,b). It must also be noted that a high level of CTL activity is detectable in vivo only in the tumor systems that are characterized by a strong immune tumor rejection, such as MSV, Friend, or Gross anti-C58 ( N T) D lymphomas in rats. All these models have, in common with allogeneic grafts, to be artificial systems in which a large number of strongly antigenic cells are inoculated or rapidly produced in the resistant recipients. The MSV system has fundamental interest in that it allows the study of CTL production and activity in a syngeneic genetic background involving no differences at the major histocompatibility complex or at any non-H-2-linked immune response gene. Nevertheless H-2 antigens altered by viral products appear as the main factor in MSV-tumor cell destruction (see Sections V,D,1 and VI1,C). Somatic modifications of H-2 are the trigger of CTL generation exactly like an in vivo transfer of H-2 genetically different cells. Therefore it is not surprising to observe a qualitatively identical immune response in both cases. The allogeneic graft of H-2 incompatible cells could be considered as a good model for the study of an antiviral immune response.

ACKNOWLEDGMENTS We wish to express our gratitude to Drs. E. Gomard and F. Plata for valuable discussions and help with the manuscript and to W. Aul for editorial assistance.

REFERENCES Aaronson, S. A ( 1971). Virology 44, 29-36. Aaronson, S. A., Jainchill, J. L., and Todaro, G. J. (1970). Proc. Natl. Acad. Sci. U.S.A. 66, 1236-1243. Allison, A. C . , Marga, J. N., and Hammond, V. (1974). Nature (London) 252, 746-747. Aoki, T. (1974). J. Natl. Cancer Znst. 52, 1029-1034. Aoki, T., and Johnson, P. A. ( 1972). J. Natl. Cancer Inst. 49, 183-193. Aoki, T., and Takahashi, T. (1972). J. Erp. Med. 135, 443-457. Aoki, T., Boyse, E. A,, Old, L. J., de Harven, E., Hammerling, U., and Wood, H. A. (1970). Proc. Natl. Acad. Sci. U.S.A. 65, 569-576. Aoki, T., Stephenson, J. R., and Aaronson, S. A. ( 1973). Proc. Natl. A c d . Sci. U.S.A. 70, 742-746. Aoki, T., Stephenson, J. R., Aaronson, S. A., and Hsu, K. C. (1974a). Proc. Natl. Acad. Sci. U S A . 71, 34453449. Aoki, T., Huebner, R. J., Chang, K. S., Sturm, M. M., and Liu, M. (1974b). J. Natl. Cancer Znst. 52, 1189-1197. August, J. T., Bolognesi, D. P., Fleissner, E., Gilden, R. V., and Nowinski, R. C. ( 1974 ) . Virology 60, 595-600.


59

Ball, J. K., McCarter, J. A., and Sunderland, S. M. (1973). Virology 56, 268-284. Bassin, R. H., Phillips, L. A,, Kramer, M. J., Haapala, D. K., Peebles, P. T., Nomura, S., and Fischinger, P. J. (1973). Bibl. Haematol. (Basel) 39, 272-280. Bauer, H. (1974). Ado. Cancer Res. 20, 275-341. Bean, M. A., Pees, H., Rosen, G., and Oettgen, H. F. (1973). Natl. Cancer Inst., Monogr. 37, 41-48. Bianco, A. R., Glynn, J. P., and Goldin, A. (1966). Cancer Res. 26, 1722-1728. Bloom, E. T., Ossorio, R. C., and Brosman, S. A. (1974). Int. J . Cancer 14, 326434. Bolognesi, D. P., Herper, G., Green, R., and Graf, R. (1974). Biochim. Biophys. Acta 355, 220-235. Boyse, E. A., Stockert, E., and Old, L. J. (1967). Proc. Natl. Acad. Sci. U.S.A. 58, 954-957. Brunner, K. T., Mauel, J., Cerottini, J. C., and Chapuis, B. (1968). Immunology 14, 181-196. Bubenik, J., and Turano, A. (1968a). Nature (London) 220,928-930. Bubenik, J., and Turano, A. ( 196813). Folia Biol. (Prague) 14, 433-439. Bubenik, J., Turano, A., and Fadda, G. (1969). Int. J. Cancer 4, 648-654. Burstein, N. A. (1970). Reu. Fr. Etud. Clin. B i d . 15, 873-875. Casazza, A. M., Di Marco, A., and Di Cuonzo, G. (1971). Cancer Res. 31, 19711976. Cerny, J., and Essex, M. ( 1974). Nature (London) 251,742-745. Cerottini, J. C., and Brunner, K. T. (1974). Adu. Immunol. 18, 67-132. Chan, S. P., Hook, W. A., Turner, W., and Chirigos, M. A. (1970). Infect. Immun. 1, 288. Chesterman, F. C., Harvey, J. J., Dourmashkin, R. R., and Salaman, M. H. (1966). Cancer Res. 26, 1759-1768. Chieco-Bianchi, L., Collavo, D., Biasi, G., and Colombatti, A. (1973). Br. J. Cancer 28, 238-244. Chieco-Bianchi, L., Colombatti, A., Collavo, D., Sendo, F., Aoki, T., and Fischinger, P. J. (1974). 1. Exp. Med. 140, 1162-1179. Chirigos, M. A., Perk, K., Turner, W., Burka, B., and Gomez, M. (1968). Cancer Res. 28, 1055-1063. Chuat, J. C., Berman, L., Gunven, P., and Klein, E. (1969). Int. J . Cancer 4, 465479. Collavo, D., Colombatti, A., Chieco-Bianchi, L., and Davies, A. J. S. (1974). Nature (London) 249, 169-170. Colombatti, A,, Collavo, D., Biasi, G., and Chieco-Bianchi, L. (1975a). Int. J . Cancer 16, 427-434. Colombatti, A., Collavo, D., Biasi, G., and Chieco-Bianchi, L. (197513). Int. J . Cancer 16, 435-441. Davis, S. (1975). J . Natl. Cancer. Imt. 54,793-794. Djeu, J. Y., Glaser, M., Kirchner, H., Huang, K. Y.,and Herberman, R. B. (1974). Cell. Immunol. 12, 164-169. De Vries, J. E., Cornain, S., and Rumke, P. (1974). Int. J . Cancer 14, 427-434. Doherty, P. C., and Zinkernagel, R. M. ( 1974). Tramplant. Reu. 19, 89-120. Doherty, P. C., and Zinkernagel, R. M. (1975). J. E x p . Med. 141, 503-507. East, J., and Harvey, J. J. (1968). Int. J. Cancer 3, 614-627. Eckner, R. J., and Steeves, R. A. (1971 ). Nature (London), New Bid. 229, 241-243. Eckner, R. J., and Steeves, R. A. (1972). J. Exp. Med. 136, 832-850. Epstein, L. B., and Knight, R. A. ( 1975). Br. J. Cancer 31, 499-512.

60

J . P. LEVY AND J. C. LECLERC

Fefer, A. (1969). Cancer Res. 14, 2177-2183. Fefer, A. (1970). Int. J . Cancer 5, 327437. Fefer, A., McCoy, J. L., and Glynn, J. P. (1967a). Cancer Res. 27, 962-967. Fefer, A., McCoy, J. L., and Glynn, J. P. (1967b). Cancer Res. 27, 1628-1631. . Res. 27, 2207-2211. Fefer, A., McCoy, J. L., and Glynn, J. P. ( 1 9 6 7 ~ )Cancer Fefer, A., McCoy, J. L., and Glynn, J. P. ( 1967d). Int. J. Cancer 2, 647-650. Fefer, A., McCoy, J. L., Perk, K., and Glynn, J. P. (1968a). Cancer Res. 28, 15771585. Fefer, A,, McCoy, J. L., and Glynn, J. P. (196813). Exp. Haematol. 15, 19-22. Ferrer, J. F., (1973). Int. J. Cancer 12, 378388. Finkel, M. P., Biskis, B. O., and Jinkins, P. B. (1966). Science 151, 698-701. Forman, J., and Moller, G. (1973). Transplant. Rev. 17, 108-149. Fossati, G., Holden, H. T., and Herberman, R. B. (1975). Cancer Res. 35, 2600-2608. Gardner, I., Bowern, N. A., and Blanden, R. V. (1975). Eur. I. Immunol. 5, 122126. Gazdar, A. F., Chopra, H. C., and Sarma, P. S. (1972a). Int. J . Cancer 9, 219-233. Gazdar, A. F., Sarma, P. S., and Bassin, R. H. (1872b). Int. J . Cancer 9,234-241. Gazdar, A. F., Russel, E. K., and Herberman, R. B. (1973). J. Natl. Cancer Inst. 50, 971-978. Geering, G., Old, L. J., and Boyse, E. A. (1966). J. Exp. Med. 124, 753-772. Geering, G., Hardy, W. D., Old, L. J., and de Harven, E. (1968). Virology 36, 678707. Gelderblom, H., Bauer, H., and Craft, T. (1972). Virology 47, 416-425. Giuliani, D., Casazza, A. M., and Di Marco, A. (1973). Biomedicine 18, 387392. Glaser, M., and Herberman, R. B. (1974). J. N d l . Cancer Inst. 53, 1767-1769. Glaser, M., Herberman, R. B., Kirchner, H., and Djeu, J. Y. (1974). Cancer Res. 34, 2165-2171. Glaser, M.,Kirchner, H., Holden, H. T., and Herberman, R. B. (1976). J. Natl. Cancer Inst. (in press). Gomard, E., Leclerc,-J. C., and Levy, J. P. (1973). J. Natl. Cancer Inst. SO, 955961. Gomard, E., Leclerc, J. C., and Levy, J. P. (1974). Nature (London) 250, 671-673. Gomard, E., Duprez, V., Henin, Y.,and Levy, J. P. (1976a). Ndure (London) 260, 707-709. Gomard, E., Duprez, V., Henin, Y.,and Levy, J. P. (1976b). Folio Btol. Prague (in press ) . J. Zmmunogenetics (in Gomard, E., Duprez, V., Henin, Y.,and Levy, J. P. ( 1976~). press). Gomard, E., Leclerc, J. C.,Henin, Y.,and Levy, J. P. ( 1976d). In preparation. Gorczynski, R. M. ( 1974a). J. Immunol. 112, 533-539. Gorczynski, R. M. ( 1974b). J. Zinmunol. 112, 1815-1825. Gorczynski, R. M. ( 1 9 7 4 ~ )J.. Immunol. 112, 182S1838. Gorczynski, R. M., and Norbury, C. (1974). Br. J. Cancer 30,118-128. Gormynski, R. M., and Knight, R. A. (1975a). Br. 1. Cancer 31, 387-404. Gorczynski, R. M., and Knight, R. A. (1975b). Eur. J . Zmmunol. 5, 148-155. Gorczynski, R. M., Kilburn, D. G., Knight, R. A,, Norbury, C.,Parker, D. C., and Smith, J. B. (1975). Nature (London) 254, 141-143. Greenberger, J. S., and Aaronson, S. A. (1973). J . Natl. Cancer Inst. 51, 1935-1938. Greenberger, J. S., Stephenson, J. R., Aoki, T., and Aaronson, S . A. (1974). Znt. J. Cancer 14, 145-152.


61

Habel, K., and Eddy, B. E. (1963).Proc. SOC. Exp. Biol. Med. 113, 1-8. Halliday, W. J. ( 1971). J . Immunol. 106, 855-857. Halliday, W. J. (1972). Cell. Immunol. 3, 113-122. Hanna, H. G., Tennant, R. W., and Coggin, J. H. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 1748-1752. Harada, M., Pearson, G., Pettigrew, H., Redmon, L., and Orr, T. (1973). Cancer Res. 33, 28862893. Harada, M., Pearson, G., Redmon, L., Winters, E., and Kasuga, S. (1975). J. Immunol. 114, 1318-1322. Hartley, J. W., and Rowe, W. P. (1966). Proc. Natl. Acud. Sci. U.S.A. 55, 780-786. Hartley, J. W., Rowe, W. P., and Huebner, R. J. (1970). J. Virol. 5, 221-225. Harvey, J. J. (1964). Nature (London) 204, 1104-1105. Harvey, J. J., and East, J. ( 1969). Int. J. Cancer 4 , 655-665. Harvey, J. J., and East, J. (1970). Immun. Tolerance Oncogenesis, Proc. Perugia Quadrenn. Int. Conf. Cancer, 4th, 1969 pp. 455-460. Harvey, J. J., and East, J. ( 1971). Int. Reo. Exp. Pathol. 10, 265-360. Hellstrom, I. ( 1967). Int. J . Cancer 2, 65-68. Hellstrom, I., and Hellstkim, K. E. ( 1969). Int. J. Cancer 4,587-600. Hellstrom, I., and Hellstkim, K. E. (1970).Int. J. Cancer 5, 195-201. Hellstrom, I., Hellstrom, K. E., Pierce, G. E., and Fefer, A. (1969a). Transplant. Proc. 1, 90-94. Hellstrom, I., Hellstrijm, K. E., Evans, C. A., Heppner, G. H., Pierce, G. E., and Yang, J. P. S. (1969b). Proc. Natl. Acad. Sci. U.S.A. 62, 362-368. Herberman, R. B. (1972).J. Natl. Cancer Znst. 48, 265-271. Herberman, R. B., ,Nunn, M. E., Lavrin, D. H., and Asofsky, R. (1973). J. N d l . Cancer Inst. 51, 1500-1512. Herberman, R. B., Aoki, T., Nunn, M. E., Lavrin, D. H., Soares, N., Gazdar, A., Holden, H., and Chang, K. S . S. (1974). J. Natl. Cancer Inst. 53, 1103-1111. Herberman, R. B., Kirchner, H., Holden, H. T., Glaser, M., and Bonnard, G. D. (1975a). Symp. ASM Tumor Virus Infect. lmmun., 1975. Herberman, R. B., Nunn, M. E., and Lavrin, D. H. ( 197513). In press. In Herberman, R. B., Nunn, M. E., Holden, H. T., and Lavrin, D. H. (1975~). press. Hirsch, M. S., Allison, A. C., and Harvey, J. J. ( 1969). Nature (London) 223, 739740. Hirsch, M. S., Proffitt, M. R., Tracy, G. S., and Black, P. H. (1972). J. Immunol. 108, 649-659. Holden, H., Kirchner, H., and Herberman, R. B. (1975). J. Immunol. 115, 327431. Holden, H. T., Haskill, J. S., Kirchner, H., and Herberman R. B. (1976). J. Immunol. ( in press ) . Hook, W. A., Chirigos, M. A., and Chan, S. P. (1969). Cancer Res. 29, 1008-1012. Huebner, R. J., Hartley, J. W., Rowe, W. P., Lane, W. T., and Capps, W. I. (1966). Proc. Natl. Acud. Sci. U.S.A. 56, 1164-1169. Hunsmann, G., Moenning, V.,Pister, L., Seifert, E., and Sch8er, W. (1974). Virology 62, 307-318. Igel, H., Huebner, R. J., Deppa, B., and Bumgarner, S . (1967). Proc. N d l . Acad. Sd. U.S.A. 58, 1870-1877. Ihle, J. N., Hanna, M. G., Jr., Roberson, L. E., and Kenney, F. T. ( 1974). J . Exp. Med. l s , 1568-1581.

62


Ishimoto, A., and Ito, Y. ( 1972). Cancer Res. 32,2332-2337. Jones, D. B., and Moore, M. (1973). Br. J. Cancer 27,415-426. Jones, D. B., and Moore, M. ( 1974a). Br. 1. Cancer 29,2140. Jones, D. B., and Moore, M. (1974b). Br. J . Cancer 29, 158-167. Jones, J. M., Jensen, F., Veit, B., and Feldman, J. D. (1974). J. Natl. Cancer Inst. 52, 1771-1777. Kal1,M. A., and Hellstrom, I. ( 1975). J . Immunol. 114, 1083-1088. Kelloff, G. J., Lane, W. T., Turner, H. C., and Huebner, R. J. (1969). Nature (London) 223, 1379-1380. Kiessling, R., Bataillon, G., Lamon, E. W., and Klein, E. (1974). Int. J. Cancer 14, 642-648. Kiessling, R., Klein, E., and Wigzell, H. (1975a). Eur. J. Immtmol. 5, 112-116. Kiessling, R., Klein, E., Pross, H., and Wigzell, H. (1975b). Eur. J. Immunol. 5, 117-121. Kilburn, D. G., Smith, J. B., and Gorczynski, R. M. (1974). Eur. J. Immunol. 4, 784-788. Kirchner, H., Chused, T. M., Herberman, R. B., Holden, H. T., and Lavrin, D. H. ( 1974). J. Exp. Med. 139, 1473-1487. Kirchner, H., Muchmore, A. V., Chused, T. M., Holden, H. T., and Herberman, R. B. (1975a). J. Immunol. 114, 208-210. Kirchner, H., Holden, H. T., and Herberman, R. B. (197513). J. Natl. Cancer Inst. 55, 971-975. Kirchner, H., Glaser, M., Holden, H. T., and Herberman, R. B. (1976). Int. J. Cancer 17, 362469. Kirsten, W. H., and Mayer, L. A. (1967). J . Natl. Cancer Inst. 39, 311-335. Klein, E., and Klein, G. ( 1974). J. Natl. Cancer Inst. 32, 547568. Knight, R. A., and Gorczynski, R. M. (1975). Int. J . Cancer 15, 48-58. Knight, R. A., Mitchinson, N. A,, and Shellam, G . R. (1975). Int. J . Cancer 15, 417-428. Koch, M. A., and Sabin, A. B. (1963). Proc. SOC. Exp. Biol. Med. 113, 4-11. Koldovsky, P., Turano, A,, and Fadda, G. ( 1969). Folia Biol. (Prague) 15, 224-225. Kurth, R., and Bauer, H. (1972a). Virology 47,426433. Kurth, R., and Bauer, H. (197213). Virology 49, 145-149. Lamon, E. W., Skunak, H. M., and Klein, E. (1972a). Int. J. Cancer 10, 581-588. Lamon, E. W., Skurzak, H. M., Klein, E., and Wigzell, H. (197213). J. Exp. Med. 136, 1072-1079. Lamon, E. W., Wigzell, H., Klein, E., Andersson, B., and Skurzak, H. M. (1973a). J. Exp. Med. 137, 1472-1493. Lamon, E. W., Wigzell, H., Anderson, B., and Klein, E. ( 1973b). Nature (London), New Blol. 244, 209-211. Lamon, E. W., Klein, E., Anderson, B., Fenyo, E. M., and Skurzak, H. M. ( 1 9 7 3 ~ ) . Int. J. Cancer 12, 637-645. Lamon, E. W., Andersson, B., Wigzell, H., Fenyo, E. M., and Klein, E. (1974). Int. J . Cancer 13, 91-104. Lamon, E. W., Skurzak, M.,Anderson, B., Whitten, H. D., and Klein, E. (1975a). J . Immunol. 114, 1171-1212. Lamon, E. W., Whitten, H. D., Skurzak, H. M., Anderson, B., and Lidin, B. (1975b). J. Immunol. 115, 1288-1294. Lapp, W. S., and Moller, C. ( 1969). Cell. Immunol. 17,3394344.


63

Lavrin, D. H., Herberman, R. B., Nunn, M., and Soares, N. (1973). J. Natl. Cancer. Inst. 51, 1497-1508. Law, L. W., and Ting, R. C. (1970). J. Natl. Cancer lnst. 44, 615-621. Law, L. W., Ting, R. C., and Stanton, M. F. (1968a). J . Natl. Cancer Inst. 40, 1101-1 112. Law, L. W., Ting, R. C., and Allison, A. C. (196813). Nature (London) 220, 611-612. Law, L. W., Chang, K. S. S., and Nakata, K. (1974). J. Natl. Cancer Inst. 52, 437-443. Leclerc, J. C., and Gomard, E. (1976). In preparation. Leclerc, J. C., and Gomard, E. (1975). Proc. Am. Assoc. Cancer Res. 66, 806. Leclerc, J. C., Gomard, E., Pavie, J., and Levy, J. P. (1972a). C. R . Hebd. Seances Acad. Sci. 274, 1233-1236. Leclerc, J. C., Gomard, E., and Levy, J. P. (1972b). lnt. J . Cancer 10, 589-601. Leclerc, J. C., Gomard, E., Plata, F., and Levy, J. P. (1973). Znt. J. Cancer 11, 426-432. Lee, J. C., and Ihle, J. N. (1974). J. Natl. Cancer lnst. 55, 831-840. Levy, J. A., Hartley, J. W., Rowe, W. P., and Huebner, R. J. (1973). J . Natl. Cancer lnst. 51, 525-533. Levy, J. P. ( 1974). Prog. lmmunol. 2,249-259. Levy, J. P., Leclerc, J. C., Varet, B., and Oppenheim, E. (1968). J. Natl. Cancer lnst. 41, 743-750. Levy, J. P., Varet, B., Oppenheim, E., and Leclerc, J. C. (1969). Nature (London) 224, 606-608. Levy, J. P., Leclerc, J. C., Gomard, E., Pavie, J., and Kourislky, F. (1973). Btbl. Haematol. ( Basel) 39, 689-697. Lilly, F., and Steeves, R. (1974). Biochem. Biophys. A d a 355, 105-118. Lo, A. C. H., and Ball, J. K. ( 1974). Virology 59, 545-555. McCoy, J. L., Fefer, A., McCoy, N. T., and Kirsten, W. H. (1972a). Cancer Res. 32, 343449. McCoy, J. L., Ting, R. C., Morton, D. L., and Law, L. W. ( 1972b). J. Natl. Cancer Inst. 48, 383-391. McCoy, J. L., Ting, R. C., McCoy, N. T., Reisher, J. I., Chan, S . P., and Law, L. W. (1974). Int. J . Cancer 13, 731-741. McKhann, C. F. ( 1971). Transplant. Proc. 3, 1233-1234. MacLennan, I. C. M. (1972). Transplant. Rev. 13, 67-90. Martin, J. W., and Martin, S. E. (1974). Nature (London) 249, 564-565. Martin, S. E., and Martin, J. W. (1975). lnt. J . Cancer 15, 658-664. Moller, G. ( 1971). lmmtrnology 20, 597-610. Moloney, J. B. (1966). Natl. Cancer lnst., Monogr. 22, 139-142. Nowinski, R. C., Fleissner, E., Sarkar, N. H., and Aoki, T. (1972). J . Virol. 9, 359366. Old, L. J., and Boyse, E. A. (1965). Fed. Proc., Fed. Am. SOC. Exp. Biol. 24, 1009-1017. Old, L. J., Boyse, E. A., and Stockert, E. (1964). Nature (London) 201, 777-779. Old, L. J., Boyse, E. A., and Stockert, E. (1965). Cancer Res. 25, 813-819. Oppenheim, E., Levy, J. P., and Leclerc, J. C. (1968). C. R . Hebd. Seances Acad. Sci. 268, 620-623. Oren, M. E., Herberman, R. B., and Canty, T. G. (1971). J . Natl. Cancer Inst. 46, 621-627.

64


Ortaldo, J. R.,Ting, C. C., and Herberman, R. B. ( 1974). Cancer Res. 34, 1366-1371. Oritz de Landazuri, M., and Herberman, R. B. ( 1972). Nature (London),New Biol. 238, 18-19. Ortiz de Landazuri, M., Kedar, E., and Fahey, J. L. (1974a). J . Natl. Cancer Inst. 52, 147-152. Ortiz de Landazuri, M., Kedar, E., and Fahey, J. L. (197413). Cell. Immunol. 14, 193-205. Ortiz de Landazuri, M., Kedar, E., and Fahey, J. L. ( 1 9 7 4 ~ ) J. . Immunol. 112, 2102-2110. Owen, J. T., and Seeger, R. C. (1973). Br. J. Cancer 28, Suppl. I, 26-34. Pantelouris, E. M. (1968). Nature (London) 217, 370-371. Pasternak, G. (1967). Nature (London) 214, 1364-1365. Pazimiho, N. H., an Yuhas, J. M. (1973). Cancer Res. 33,2668-2672. Pearson, G. R., Redmon, L. W., and Bass, L. R. (1973). Cancer Res. 33, 171-178. Perk, K., and Moloney, J. B. (1966). J. N d l . Cancer Inst. 37, 581-599. Perk, K., Moloney, J. B., and Jenkins, E. G. (1967). Int. J. Cancer 543-51. Perlmann, P., Perlmann, H., and Wigzell, H. (1972). Transplant. Reo. 13, 91-114. Peter, H. H., Pavie-Fisher, J,, Fridman, W. H., Aubert, C., Cesarini, C., Roubin, R., and Kourilsky, F. M. (1975). I. Immunol. 115, 539-548. Peters, R. L., Rabstein, L. S., Van Vleck, R.,Kelloff, G. J., and Huebner, R. J. (1974). J . Natl. Cancer Inst. 53, 1725-1729. Pierce, G. E. (1971). Int. J. Cancer 8, 22-31. Plata, F., and Levy, J. P. ( 1974). Nature (London) 249, 272-274. Plata, F., Gomard, E., Leclerc, J. C., and Levy, J. P. ( 1973). J. Immunol. 111, 667-671. Plata, F., Gomard, E., Leclerc, J. C., and Levy, J, P. (1974). J . Immunol. 112, 1477-1487. Plata, F., Cerottini, J. C., and Brunner, K. T. (1975). Eur. J. Zmmunol. 5, 227-233. Pollack, S. B. ( 1971).Int. J. Cancer 8, 264-271. Pollack, S . B. (1973).Int. J . Cancer 11,136-142. Pollack, S . B., and Nelson, K. ( 1973). J . Immunol. 110, 1440-1443. Pollack, S. B., and Nelson, K. ( 1974). Int. J. Cancer 14, 522-529. Pollack, S . B., Heppner, G., Brawn, R. J., and Nelson, K. (1972). Int. J . Cancer 9, 316-323. Price, C. H. G., Moore, M., and Jones, D. B. ( 1972). Br. J. Cancer 26, 15-27. Rich, M. A., Geldner, J., and Myers, P. (1965). J . Nutl. Cancer Inst. 35, 523536. Russel, S. W., Doe, W. F., and Cochrane, C. G. (1970). J . Immunol. 116, 164-166. Russel, S. W., and Cochrane, G. G. (1974). Int. 1. Cancer 13,5443. Russel, S . W., Doe, W. F., and Tozier, A. (1975). Fed. Proc., Fed. Am. Soc. Ezp. Biol. 34, 268. Salinas, F. A,, and Hanna, M. G. (1974). J . Immuwl. 112, 1028-1034. Schifer, W., Anderer, F. A., Bauer, H., and Pister, L. (1969). Virology 38, 387394. Schwartz, D. B., Zbar, R., Gibson, W. T., and Chirigos, M. A. (1971). Int. J. Cancer 8, 320-325. Schwartz, R. S., and Beldotti, L. (1965). Science 149, 1511-1514. Seeger, R. C., and Owen, J. J. T. (1973). Transplantation 15, 404-408. Seeger, R. C., Raper, S. A,, and Owen, J. J. T. (1974). Int. J. Cancer 13, 697-713. Senik, A., Gomard, E., Plata, F., Levy, J. P. (1973). Int. J. Cancer 12, 233-241. Senik, A,, De Giorgi, L., and Levy, J. P. (1974a). Int. 1. Cancer 14, 386-395.

MURINE SARCOMA VJRUS-INDUCED TUMOR

65

Senik, A., De Giorgi, L., Gomard, E., and Levy, J. P. (197413). Int. J. Cancer 14, 396-400. Senik, A,, Pozo-Hebrero, F., and Levy, J. P. ( 1975a). Int. J. Cancer 16, 946-959. Senik, A., Gisselbrecht, S., and Levy, J. P. ( 1975b). Int. J. Cancer 16, 960-972. Shachat, D. A., Fefer, A,, and Moloney, J. B. (1968). Cancer Res. 28, 517-520. Shellam, G. R. (1974). Int. J . Cancer 14,65-82. Shellam, G. R., and Knight, R. A. (1974). Ndure (London) 252,330-332. Simons, P. J. (1970). Aust. J. Exp. Biol. Med. Scl. 48, 105-114. Simons, P. J., and McCully, D. J. ( 1970). J. Natl. Cancer Inst. 44,1289-1303. Sjogren, H. O., Hellstrom, I., and Kelin, G. ( 1961). Cancer Res. 21, 329-333. Sjogren, H. O., Hellstrom, I., Bansal, S. C., and Hellstrom, K. E. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 137.2-1375. Skunak, H. M., Klein, E., Yoshida, T. D., and Lamon, E. W. ( 1972). J . Exp. Med. 135, 997-1002. Stanton, M. F., Law, L. W., and Ting, R. C. (1968). J . Natl. Cancer Inst. 40, 1113-1129. Steeves, R. A. (1968). Cancer Res. 28,338442. Steeves, R. A., and Axelrad, A. A. ( 1967). Int. I. Cancer 2, 235-244. Stephenson, J. R., Rand-Aaronson, S. A. (1972). J. Exp. Med. 135,503-515. Strand, M., and August, J. T. (1974). J. Virol. 13, 171-180. Strausser, H. R., and Bober, L. A. (1972). Cancer Res. 32,2156-2159. Strouk, V., Grundner, G., Fenyo, E. M., Lamon, E., Skurzak, H., and Klein, G. (1972). J. Erp. Med. 136,344-352. Stutman, 0. (1975). Nature (London) 253,142-144. Takasugi, M., and Klein, E. (1970). Transplantation 9, 219-227. Takasugi, M., Mickey, M. R., and Terasaki, P. J. (1973). Cancer Res. 33, 2898-2902. Tamerius, J. D., and Hellstrom, I. (1974). J. Immunol. 112, 1987-1996. Tennant, R. W., Hanna, M. G., and Farrely, J. G. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 316743171. Thomas, W. R., Aw, E. J., Papadimitriou, J. M., and Simmons, P. J. (1973). J. Natl. Cancer Inst. 51, 1541-1549. Ting, C. C., and Herberman, R. B. (1974). Cancer Res. 34, 1676-1683. Ting, C. C., Lavrin, D., Shiu, G., and Herberman, R. B. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 1664-1668. Ting, C. C., Shiu, G., Rodrigues, D., and Herberman, R. B. (1974). Cancer Res. 34, 1684-1687. Ting, R. C. (1967). Proc. Soc. Exp. Biol. Med. 126, 778-781. Ting, R. C. ( 1968). J . Virol. 2, 865-868. Trentin, J. J., and Bryan, E. (1966). Proc. Soc. Exp. Biol. Med. 121, 1216-1219. Turner, W., and Chirigos, M. A. (1969). Cancer Res. 29, 1956-1960. Turner, W., Ebert, P. S., Bassin, R., Spalin, G., and Chirigos, M. A. (1971). Proc. Soc. Erp. Biol. Med. 136, 1314-1318. Varet, B., Levy, J. P., Leclerc, J. C., and Senik, A. (1968). Int. J . Cancer 3, 727-733. Varet, B., Levy,-J. P., Leclerc,-J. C., and Kourilsky, F. M. (1971). Int. 1. Cancer 7, 313421. Varet, B., Cannat, A., FeingoId, N., Weschler, J., and Levy, J. P. (1973). Cancer Res. 33, 759-763. Veit, B. C., and Feldman, J. D. (1975). Int. J. Cancer 15,367-376. Weissman, I. L. (1973). J. Natl. Cancer Inst. 51,443-448.

66

J . P. LEVY AND J. C. LECLERC

Witter, R., Frank, H., Moennig, V., Hunsmann, G., Lange, J., and Schiifer, W. ( 1973). Virology 54, 330-345. Yoshiki, T., Mellors, R. C., and Hardy, W. D. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 1878-1882. Yoshiki, T, Mellors, R. C., Hardy, W. D., Jr., and Fleissner, E. (1974). J . Exp. Med. 139, 925-942. Zinkernagel, R. M., and Doherty, P. C. (1974). Natfrre (London) 248, 701-702. Zinkernagel, R. M., and Doherty, P. C. (1975). J . E x p . Med. 141, 1427-1436. Zisblatt, M., Goldstein, A. L., Lilly, F. and White, A. (1970). Proc. Natl. Acad. Sci. U.S.A. 66, 1170-1174.

ORGANIZATION OF THE GENOMES OF POLYOMA VIRUS AND SV40

Mike Fried and Beverly E. Griffin Imperial Cancer Research Fund Laboratories, Lincoln's Inn Fields, London, England

I. Introduction . . . . . . . . . . . . . . . . . 67 11. Action of Enzymes on the Viral DNAs and Construction of Physical Maps 69 111. Primary Sequence Studies . . . . . . . . . . . . . . 76 IV. Origin and Termination of Viral DNA Replication . . . . . . . 81 V. Location of Virus-Specific RNAs . . . . . . . . . . . . 83 VI. Virus-Induced Proteins . . . . . . . . . . . . . . 85 VII. Protein Binding Sites on Viral DNAs . . . . . . . . . . . 88 VIII. Genetic Mapping . . . . . . . . . . . . . . . . 89 IX. Essential and Nonessential Regions of the Viral Genomes . . . . . 94 X. Defective Viral DNAs . . . . . . . . . . . . . . . 99 XI. Comparison of the Polyoma Virus and SV40 Genomes . . . . . . 102 XII. Conclusion . . . . . . . . . . . . . . . . . . 105 References . . . . . . . . . . . . . . . . . . 107

I. Introduction

The major stimulus for the studies on two papovaviruses, polyoma virus ( Gross, 1953a) and simian virus 40 ( SV40) ( Sweet and Hilleman, 1960), has undoubtedly been their oncogenic potential. Both viruses will cause tumors in a variety of rodents in vivo (Gross, 1953b; Stewart et al., 1958; Eddy et al., 1961, 1962). In tissue-culture systems, polyoma virus will transform a number of different types of rodent cells, whereas SV40 can transform, in addition to rodent cells, monkey and human cells. These viruses do not, however, appear to be important tumor-causing agents in the wild. Only with large quantities of virus, inoculated into immunologically immature or deficient animals, can tumors be produced under experimental conditions. However, large numbers of transformed cells (cancer cells) can be produced in vitro in a simple and reproducible manner after viral infection. Thus, these viruses appear to offer a good model system for elucidating in vitro some of the events that lead in vivo to tumor formation. The vast amount of work of the past few years has pointed to remarkable similarities between the two viruses, as will be clearly evidenced from this review. In spite of overwhelming organizational similarities, polyoma virus and SV40 DNAs show very limited sequence homology 67

68

MIKE FRIED AND BEVERLY GRIFFIN

and different host-range properties. They differ not only with respect to the cells they transform, but also with respect to the so-called lytic properties of the viruses, that is, where viral infection leads ultimately to the production of more virus with death of the host cell. (Mouse cells are “permissive” for lytic infection with polyoma virus, whereas monkey cells are permissive for SV40.) In cells transformed with these viruses, viral DNA can be found covalently linked to host-cell DNA (Sambrook et al., 1 x 8 ) . In some cases, infectious viral genomes can be rescued from transformed cells after fusion with permissive cells (see Sambrook, 1972). The viral genome is therefore postulated to be responsible for initiating transformation, and the integrated viral DNA for maintaining the transformed state of the cell. Among the major unsolved problems in this field are those concerning the mechanisms by which viral genomes cause host cell transformation, and elucidation and/or location of the site( s ) of integration between the viral and host cell DNAs. It is still far from clear that the continued presence of the viral genome (or a part thereof) is necessary for the maintenance of transformation. In addition to probing “nonpermissive” cell types (cells that do not support virus growth but can be transformed), the behavior of polyoma virus and SV40 in permissive cells is being studied extensively in order to define the interaction between the virus and its host under conditions where the whole viral genome is replicated and expressed. Because these oncogenic viruses per se are very small and contain a limited amount of genetic information, the possibility of their thorough characterization presents an immediate challenge to molecular and cell biologists, virologists, and geneticists. The genetic information of both polyoma virus and SV40 resides in their DNAs, which have molecular weights of about 3.4 x lo* daltons (about 5200 nucleotide pairs). The DNAs exist in “native” form as covalently closed circular superhelical molecules with a coding capacity for about 200,000 daltons of protein. Three distinct genes have so far been identified. One can be correlated with an early function concerned with viral DNA synthesis and the other two with proteins found in the viral capsid. This review will deal mainly with the organization of the genetic information within the polyoma virus and SV40 genomes. Information on the structure and composition of polyoma virus and SV40, on their replication, their transcription and translation, on their genetics, and on the effect they have on various cell types may be found in Advances in Cancer Research (Sambrook, 1972)) the Molecular Biology of Tumour Viruses (Tooze, 1973), Selected Papers in Tumour Virology (Tooze and

POLYOMA AND SV&

GENOMES

69

Sambrook, 1974), and Cold Spring Harbor Symposium on Quantitative Biology, Vol. 39, 1974. Since the authors are actively engaged in research on polyoma virus, they confess to a natural bias toward this virus, which will doubtless be reflected in this review. II. Action of Enzymes on the Viral DNAs and Construction of Physical Maps

The prerequisite for the study of the virus and its interaction with host cells is at least a superficial understanding of the viral genome. Since the genetic information for polyoma virus and SV40 resides in their DNA, this means an understanding of the DNA either in terms of physical or genetic markers. Examination of the literature before 1970 indicates the handicaps experienced by investigators largely because such markers were not available. With polyoma virus and SV40, there are not even physical markers similar to those available for genomes of viruses such as adenovirus or lambda (i.e., the ends of the DNA), since the DNA of polyoma virus and SV40 is circular. Therefore, the discovery of restriction endonucleases that cleave double-stranded DNA at specific sites has been especially important for the study of the polyoma virus and SV40 genomes; these enzymes have allowed the DNAs to be characterized and have provided the physical markers hitherto lacking for biological definition of the viral DNAs. Restriction enzymes have been isolated from a number of microorganisms; some of them recognize specific sequences in double-stranded DNA and cleave both strands of the DNA. In any particular organism, the restriction endonucleases are part of a restriction-modification system, modification being used by the organism as a means of protecting its own DNA against cleavage by the enzyme. “Foreign” DNA, on the other hand, is cleaved by the restriction enzyme, To date, several classes of restriction endonucleases have been found: Some (class I ) are nonspecific in their cleavage. Others (class 11) cleave double-stranded DNA at the enzyme recognition sequence. Another type of enzyme represented by the enzyme from Haemophilw parahuemolyticus (Hph I, Kleid et d.,1976), cleaves at sequences other than those at the recognition site. Class I1 enzymes have been particularly useful in the study of viral DNAs. The substrates for a number of class I1 restriction endonucleases have been determined. They consist of symmetrical DNA sequences of 4 to 6 nucleotide base pairs (see Table I). Until recently, all evidence supported the specific requirement of these enzymes for double-stranded DNA. Two papers have recently appeared which suggest that some restriction endonucleases can cleave single-stranded DNA ( Horiuchi and Zinder, 1975;

70

MJKE FRIED AND BEVERLY GRIFFIN

Blakesley and Wells, 1975). Some will also recognize sites in which hydroxymethyluracil replaces thymine, whereas others will not ( Ito et al., 1975) . It would be superfluous in this review to consider restriction endonucleases in detail, most especially since some excellent general reviews have recently appeared. These cover not only the discovery and characterization of restriction enzymes, but also their uses in analysis and structuring of DNA molecules, and their potential uses in “genetic engineering.” For example, see “DNA Restriction and Modification Mechanisms in Bacteria,” Boyer ( 1971); “Restriction and Modification of DNA,” Meselson et al. (1972); “DNA Modification and Restriction,” Arber (1974); “The Primary structure of DNA,” Murray and Old (1974); “Restriction Endonucleases in the Analysis and Restructuring of DNA Molecules,” Nathans and Smith ( 1975); “Restriction Enzymes and the Cloning of Eukaryotic DNA,” Murray et al. (1975); and “Structural Mapping of the DNA of an Oncogenic Virus (Polyoma Viral DNA),” Griffin and Fried (1976). Studies on the cleavage of both polyoma virus and SV40 DNAs by a number of restriction endonucleases have been carried out (see Tablc I ) . Physical maps of the DNA of both viruses have been elucidated, based on their cleavage into discrete fragments by restriction endonucleases and the subsequent placing of the fragments in topographical order. For polyoma virus, the first physical map was based on the order of the eight fragments (Hpa I1 1 to 8, numbered according to decreasing size) produced by cleavage of the DNA with one of the enzymes froin Haemophilus parainfluenzae (Hpa 11); the two sites cleaved by an enzyme from Haeniophilus influenzae (Hind 111) and the single site cleaved by Escherichia coli RI endonuclease were also located on the Hpa II-restriction map (Griffin et al., 1974). For SV40, the first physical map was based on cleavage of the DNA with enzymes from Haemophilus influenzae (Hind I1 and 11) and H . parainfluenzae ( H p a I and 11). Eleven fragments produced by cleavage with Hind I1 and I11 ( A to K, lettered according to decreasing size) and four fragments produced by cleavage with Hpa I and I1 were ordered by Danna et nl. (1973); subsequent to this work, two additional Hind cleavage fragments ( L and M ) and an additional Hpa I fragment were observed and placed on the SV40 physical map (Fiers et al., 1974; Yang et al., 1975). Additional data on the cleavage of polyoma virus and SV40 DNAs are summarized on the physical map of each viral DNA (Figs. 1 and 2 ) and in Table I. For polyoma DNA, most of the data given correspond to studies on a large-plaque strain A-2 (Fried et al., 1974). Strain variability is discussed below and in Section IX.

POLYOMA AND

SV40

GENOMES

71

FIG.1. Restriction endonuclease cleavage maps of polyoma virus DNA ( A - 2 strain). The cleavage site for the Escherichia coli R I enzyme is designated to be at zero map unit, and other cleavage sites are given in map units (0-100). The sources used in constructing this map were: Eco RI, Griffin et al., 1974; Bam I, Griffin and Fried, 1976; Hae 11, Griffin and Fried, 1976; Hind 111, Griffin et al., 1974; Hind 11, Griffin and Fried, 1975; Kpn I, Crawford and Robbins, 1976; Hha I, Griffin and Fried, 1975; Bum I, M. Fried and B. E. Griffin, unpublished; Hga I, Shishido and Berg, 1976; Pst I, Crawford and Robbins, 1976; Hpa 11, Griffin et al., 1974; Hae 111, B. E. Griffin, unpublished; see also Summers, 1975. For further information about the restriction enzymes and their actions on polyoma virus DNA, see references to Table I.

From among the general methods used for deriving physical maps of polyoma virus and SV40 based on cleavage with restriction endonucleases (reviewed by Nathans and Smith, 1975; Griffin and Fried, 1976) the most useful have been: (1) digestion of DNA under limiting (incomplete) digestion conditions, followed by separation, isolation, and analysis of the partially digested products by further cleavage under limit

TABLE I CLEAVAGE OF POLYOMA VIRUSAND SV40 DNAs

BY

3

RESTRICTION ENDONUCLEASES* Number of sites

Microorganism Brevibaclerium albidum

Abbreviation

Bal 10

Haemophilus parainfluenzae

Hpa Id

Streptmnyces &bus G

sal I*

Bacillus amyloliquefaciens

Barn 1'

Escherichia coli RI

Haemophilus aegyptius

Em RI'

Hae IIp

Sequence

Hind IFJ

SV40

1

5f-CGGCCG35 3'-GCCGGGBf

T 1

GTTAACI CAATTG

T

-

1

GGATCCb CCTAGG

w Oi li

1'

1"

10.0

T

1

GAATTCm CTTAAG

T 1

RGCGCYq YCGCGR

T

Haemophilus influenzae-d

Polyoms

1'

1

7g.w

T

AAGCTTL TTCGAA

Haemophilus influenme-d Klebsiella pneumoniae O K 8

K p n In

Haemophilus huemolyticus

Hha Inn

t

1

NGCGCNb#nn NCG CGN

T

I 14

GTYRACu CARYTG

1

$

2.

1"

3'

2"

Brevibacterium umbra Haemophilus gallinarum Providencia stuartii 164

Bum I" Hga I& Pst I'

Haemophilwr parainfluenzae

Hpa IIdsdd

Haemophilw parahaemolyticus

Hph I=

Haemophilus aegyptius

Hae IIIff

4c 4" 5'

1

CCGGe GGCC

8f

T

1

GGCCOO CCGG

* NOTE:In addition, polyoma DNA has been found to lack recognition sites for a number of other restriction endonuzertses. These include: Serratiu marcescm (Sma I), Breuibacterium luteum (Blu I), Streptomyces achromogenes (Sac I and Sac 11), Xanlhomonas amaranthicola (Xam I), and Xanthomonas malvacearum (Xma I)c. [Xam I and Sma I are "isoschizomers," as are Xam I andSal I, and Hsu and Hind I11 (see below), respectively. Isoschicomer is a term used by R. Roberts to denote enzymes from different sources that recognize the same sequences.] Enzymes from Haemophilus suis (Hsu) and Anabaena uariubilis (Ava I) have each been reported to have two cleavage sites,lj.kkand enzymes from Haemophilus injZuenzae f (Hinf I), Moraxella bovis (Mbo I and Mbo 11) to have more than five sites.a Arthrobacter luteus (Alu I) and Escherichia coli R245 (RII) have a large number of cleavage sites in polyoma DNA.1.. For data on additional restriction endonuclease cleavage sites in SV40 DNA, see Fig. 2. a R. J. Roberts, personal communica0 Morrow and Berg (1972); Mulder and cc Shishido and Berg (1976) tion Delius (1972) d d Gromkova and Goodgal (1972) Murray el aZ. (1975) P Roberts et al. (1975) eeKleid et al. (1976) (cleavage site is c M . Fried and B. E. Griffin, unpubB. G. Barrel1 and P. M. Slocombe, 8-9 base pairs away from postulated lished personal communication recognition site) 7 Griffin and Fried (1975) Sharp et al. (1973) Middleton el al. (1972) Garfin and Goodman (1974) .Smith and Wilcox (1970); Smith 00 Murray and Old (1974) f Griffin et al. (1974) (1974) Ah Summers (1975) 0 Danna et al. (1973) Old et al. (1975) '* Subramanian et al. (1974) ; Lebowitz A J. R. Arrand, personal communication Kelly and Smith (1970) el al. (1974); Fiers et al. (1974) * Wilson and Young (1975) v Chen et al. (1975); Folk et aZ. (1975); 1 ) M. Vogt et al. (1976) (cleavage with 1 Griffin and Fried (1975); Yaniv et al. Lescure and Yaniv (1975) Hsu, an isoschizomeraof Hind 111) Yang et al. (1975) 5 1 L. V. Crawford, personal communi(1975) k J . Sambrook and M. Mathews, perGromkova et al. (1973) cation sonal communication Smith el al. (1976) il K. N. Subramanian, B. S. Zain, R. J. 1 Yoshimori (1971); Greene et al. (1974) Crawford and Robbins (1976) Roberts, and S. M. Weismann, unHedgpeth et al. (1972); Garfin et al. M. Mathews, personal communicapubhshed (1975) tion (Am Yang et aZ. (1976s) Robberson and Fried (1974) t+ Takanami (1974) nn Roberts et al. (1976) ff

50

E $

3K

e!

4

w

74


FIG. 2. Restriction endonuclease cleavage maps of SV40 DNA (modified from Fiers et ul., 1975). The cleavage site for the EschelSchia coli R I enzyme is designated to be at zero map unit, and other cleavage sites are given in map units (0-1.0). The sources used in constructing this map were: Eco RI, Morrow and Berg, 1972; Mulder and Delius, 1972; Bam I, J. Sambrook and M. Mathews, unpublished; Hpa 11, Sharp et al., 1973; Hae 11, Roberts et al., 1975; Hha I, K. N. Subramanian, S. Zain, R. J. Roberts, and S. M. Weissman, unpublished; Bum I, M. Mathews, unpublished; Hpa I, Danna et at., 1973; Sharp et d., 1973; Hind II/III, Danna et ol., 1973; Yang et al., 1975; Hin f, K. N. Subramanian, S. Zain, R. J. Roberts, and S. M. Weissman, unpublished; Eco RII, Subramanian et al., 1974; Hae 111, ibid; Yang et al., 1976a; Alu I, Yang e t al., 197613. For further information about the restriction enzymes and their action on SV40 DNA, see references to Table I.

( complete) digestion conditions. This method allows the identification of overlapping clusters of fragments and was used for obtaining the first physical maps of polyoma virus and SV40 DNA (Griffin et al., 1974; Danna et al., 1973, respectively); (2) successive cleavage with a number of restriction endonucleases (for references, see Table I and Figs. 1 and

POLYOMA AND S V N GENOMES

75

2). This has proved to be most useful where a physical map was already available and sites of cleavage by other enzymes were being studied. An alternative procedure that may be useful for fine-structure mapping was used by Summers (1975) to construct a partial physical map of polyoma DNA based on cleavage with Hae 111. This involves annealing a specific denatured fragment (primer) to wild-type single-stranded circles, then labeling nucleotide sequences adjacent to the primer using an alabeled deoxyribonucleoside triphosphate and DNA polymerase I. [The strain of polyoma virus studied by Summers appears to be of the A-3 type and his Hae I11 map therefore differs slightly from that shown (Fig. 1) for the A-2 strain (see below).] One of the interesting corollaries of the studies with restriction endonucleases has been the finding that restriction fragment patterns vary among strains. For polyoma DNA, the size of the fragment Hpa 11-5 varies in two large-plaque strains A-2 and A-3 (Fried et al., 1974), and a small-plaque strain has been found to have an additional Hpa IIcleavage site near the center of what is for most strains the largest Hpa I1 cleavage fragment (Fried et al., 1974; Crawford et al., 1974). Moreover, variability in size has also been found in fragment Hpa 11-3 (Fried et al., 1974; Crawford et al., 1974; Francke and Vogt, 1975; Vogt et al., 1976). In general, temperature-sensitive polyoma virus mutants have been shown to have the Hpa I1 cleavage pattern characteristic of the parental strain from which they were derived (Fried et al., 1974). For SV40, similar variabilities among strains have been observed using the restriction endonucleases Hae I11 (Huang et al., 1973) and Hind I1 and I11 (Nathans and Danna, 1972a; Nathans et al., 1973; Lai and Nathans, 1974c; Botchan et al., 1974). In addition to the work on the polyoma virus and SV40 genomes discussed above, several human papovaviruses (the genomes of some of which have been shown to have partial sequence homology with the SV40 genome ) have been analyzed with restriction endonucleases. These include a virus from a patient with progressive multifocal leukoencephalopathy (PML virus, Sack et al., 1973), a virus from the urine of an immunosuppressed renal transplant patient (BK virus, Howley et al., 1975a,b; Khoury et al. 197513) and viruses from patients with Wiskott-Aldrich syndrome, a rare immunodeficiency disease ( Howley et al., 1975b). Similar comparative studies between genomes of the human papovaviruses and that of polyoma virus have not yet been reported; in studies aimed mainly at comparing physical and biochemical properties of polyoma and BK viral particles, no cross reactivity between the capsid proteins responsible for the hemagglutination of red blood cells was detected (Seehafer et al., 1975). It should be noted that the DNAs of polyoma

76


virus and SV40 have been shown to have limited sequence homology ( Ferguson and Davis, 1975) (see Fig. 7). Studies on polyoma virus and SV40 DNAs with another type of enzyme have also been carried out. An enzyme known to cleave singlestranded DNA (nuclease S1 from AspergiUus o y z a e ) has been shown to cleave both polyoma virus and SV40 DNAs to full-length linear molecules ( Germond et aZ., 1974; Beard et al., 1973, respectively), The enzyme nicks DNA at a few sites (probably easily denatured sites), then cuts the intact strand at a site opposite to the nick. Because of apparent lack of specificity under most conditions, this enzyme has not been very useful for studying the wild-type genomes. (At high salt concentrations, some cleavage specificity has been reported, Beard et al., 1973). It has been used, however, with considerable success for mapping deletion and insertion mutants of SV40 DNA and some temperature-sensitive mutants ( see Sections VIII and IX). 111. Primary Sequence Studies

Very little is known at the moment about the primary sequences of polyoma virus DNA. A fingerprint map of the pyrimidine tracts has been obtained based on a two-dimensional separation of the products obtained after depurination of polyoma virus DNA (Griffin and Fried, 1976) ( Fig. 3). For double-stranded DNA (such as polyoma or SV40 DNA) such a “depurination fingerprint” provides a picture of the entire genome, since a tract of purines on one strand will have its complementary tract of pyrimidines on the other strand. Similar depurination maps have been made of each of the eight Hpa II-restriction fragments and of some of the Hae I11 fragments (Griffin and Fried, 1975; B. E. Griffin, unpublished). Unique tracts of pyrimidines have been found in seven of the eight Hpa I1 fragments (i.e., Hpa I1 fragments 1, 2, 3, 5, 6, 7, and 8) and these have been used for the characterization of polyoma virus variants and cloned defective species (see Sections IX and X). A tract of 7-8 thymidines in fragment Hpa 11-3 near the Hpa 11-33junction (Hae I11 fragment 14, Fig. 1) should potentially represent an easily denaturable region in polyoma DNA, although this is not one of the sites where T4 gene-32 protein has been found to bind (Monjardino and James, 1975; Yaniv e t al., 1975) (see Section VII). A pyrimidine tract approximately seventeen nucleotides long is found in Hpa 11-5 near the Hpa 11-3-5 junction (Hae I11 fragment 14’, Fig. 1).It has also been found in all the polyoma variants and strains examined to date, and appears to contain sequences essential to the polyoma genome (see Section IX). Considerably more information is available about primary sequences

POLYOMA AND

sv40

GENOMES

77

FIG.3. A depurination map of "P-labeled polyoma A-2 wild-type DNA (left) and a schematic diagram of this map (right). The separation and identification of the pyrimidine tracts were made as described by Ling (1972). Similar depurination maps of each Hpa I1 fragment have allowed unique pyrimidine tracts to be located within fragments on the polyoma physical map (Griffin and Fried, 1975; B. E. Griffin, unpublished). On the schematic diagram, the appropriate Hpa I1 fragment number (see Fig. 1 ) is put inside the circle representing the tract and indicates the fragment that contains the oligodeoxynucleotide. Thus, Hpa 11-1 can be seen to have four unique pyrimidine tracts; the largest tract can be found in Hpa 11-8, etc. of SV40 DNA; this comes mainly from two laboratories, that of Fiers in Ghent and that of Weissman at Yale. The studies of both groups have to date involved transcription of restriction enzyme fragments of SV40, and sequencing the resulting RNA. Weissman and collaborators (Zain et al., 1974; Dhar et al., 1974a,b,c) have determined the nucleotide sequence for most of fragment Hind G

78


and part of Hind B in the region near the Hind B-G junction where termination of DNA replication occurs in nondefective SV40 DNA (Danna and Nathans, 1972) (see Fig. 2) and the 3‘-ends of mRNAs have been mapped (see Section V ) , Subramanian et al. ( 1976) find a uridine-rich sequence from this region, GGUUUUUUAC, which resembles sequences found at the 3’-ends of several prokaryotic transcripts; uridine-rich sequences have been postulated to play a role in termination (Lebowitz et al., 1971; Dahlberg and Blattner, 1973; Sklar et al., 1975; Kleid et al., 1976; Rosenberg et al., 1976; Squires et nl., 1976). Notable also among the sequences in this region is a sequence of seventeen nucleotides, CAAUUGUUGUUGUUAAC, which can be seen to be a palindrome about a central nucleotide and to contain the transcript of the Hind cleavage site (underlined) (Dhar et al., 1974b). The significance of the 17 long palindromic sequence in the viral genome is not clear. It should be noted that the Hind B-G junction may be an intercistronic region in the genome (Fig. 7) and that a palindromic sequence in a genome puts a constraint on the structure of any RNA transcribed from it in that it prevents the RNA from assuming a hairpin structure (since such a structure would involve the pairing of like bases). It is possible to postulate from these observations that the palindromic sequence is part of a control element within the viral genome and may be important in the processing of the 3’ ends of the mRNAs. The significance of palindromes within a “coding” sequence has been discussed in detail by Pieczenik et al. (1974); it remains to be seen whether they have significance when they occur in regions that are not translated. Although it may have no relevance for a eukaryotic polymerase, it is nonetheless interesting to note that the region in Hind G near the Hind B-G junction has also been shown to contain a preferred initiation site for E . coli polymerase. Allet et al. (1974) have shown that the polymerase can protect a promotor region in a number of phage genomes (as well as a region in SV40 and adenovirus-2) from cleavage with the Hind restriction endonucleases. In further studies, Dhar et al. ( 1975) also report that the 5’ end of earl!/ cytoplasmic mRNA overlaps the 5’ end of the 19s late mRNA by more than 90 nucleotides (Fig. 7 ) . Primary sequence data have proved to be important in mapping the region of the genome transcribed early in infection ( Dhar et nl., 1975) (see Section V ) . Subramanian et al. (1976) have sequenced a transcript of the Eco RII-G restriction fragment which should contain the origin of DNA replication (see Section IV and Fig. 2 ) and the 5’ ends of the mRNAs (see Section V ) . The sequence they report allows for the RNA (and presumably complementary D N A ) to be folded into a number of

POLYOMA AND

sv40

GENOMES

79

hairpin loops with a central double-stranded region that contains a near-perfect palindrome composed of seventeen nucleotides, CCTCCAAAAAAGCCTCC. The palindromic sequence lies in Hind-C near the site proposed for the origin of DNA replication. Jay et al. (1976) have also determined by direct sequencing methods the sequence of nucleotides (28 long) around the Hind A/C junction. This sequence is identical to a sequence of 28 nucleotides found in the Eco RII-G restriction fragment by Subramanian et al. (1976), thus confirming not only the sequence itself but the use of transcription as a reasonable method for getting DNA sequences. Among the sequences in the 28 long oligomer is a nontandem duplication of an octamer, TTTGCAAA. Since this sequence comes from a region of the DNA that contains the origin of DNA replication (see Section IV) it is tantalizing to speculate that it is repeated because it is important in the replication of the viral DNA. At the moment, the data are not adequate, however, for much correlation to be made between sequences and function. Indeed, such correlations may only prove possible when sequences from comparable regions of the polyoma DNA are available for comparison with sequences from SV40 DNA. Fiers and collaborators have concentrated mainly on the small Hind restriction fragments, Hind F, H, I, J, and K (Van de Voorde et al., 1974; Fiers et al., 1974). The sequence of the transcript of Hind H (in the early region of the genome, where a number of A mutants have been mapped; see Section VIII) and a partial sequence for Hind K (in thc late region) have been reported (Fiers et al., 1975). For Hind H, using the transcribed strand with the same polarity as early cytoplasmic mRNA, the original sequence studies of Fiers et al. (1975) showed two of the reading frames to have several nonsense codons, and the third reading frame, which should then be the correct one, to have a UAA terminator codon near the end of the fragment. If the sequence reported is correct, the implication must be either that read-through of the UAA codon occurs, that UAA is not a terminator in all mammalian systems, or that the concepts of a single early gene product (see Section VIII) must be reconsidered. Although a number of studies have been carried out to show that transcription in vitro gives a faithful copy of the DNA (see, for example, Dhar et al., 1974c; Subramanian et al., 1976; Jay et al., 1976), the final answer to this enigmatic finding must probably await analysis by direct DNA sequencing methods. [W. Fiers (private communication ) believes that a transcription error was made; the sequence of Hind H is currently being checked by direct methods.] The sequences that code for the amino-terminal of the major SV40 capsid protein VP1 ( Ala-Pro-Thr-Lys-Arg-Lys-Gly-, Lazarides et al.,

80


1974) have been found in the transcription product of Hind K. Fiers et al. (1975) have shown that these coding sequences appear in only one of the reading frames, in Hind K near the junction with Hind E; the other two reading frames contain nonsense codons. In addition to being elegant work, these studies illustrate the usefulness of primary sequence studies for locating precisely biological markers on a genome. They also allow the amino acid sequence of part of VP-1 to be predicted prior to being determined by protein chemistry techniques. Since the sequence recognition sites of a number of restriction endonucleases are known, the presence (or the absence) of a recognition site can in itself be a limited source of information about primary sequences of the genome. For example, Hind I1 has been found to recognize a degenerate sequence

4

5’-GT RAC-3’ 3’-CARYTG-5’

t

where Y = pyrimidine, R = purine, and consequently to have a set of four possible sequence recognition sites (Kelly and Smith, 1970; Nathans and Smith, 1975); an endonuclease from H . pwuinfluenzue (Hpa I ) recognizes the specific sequence (Garfin and Goodman, 1974)

I

5’-GTTAAC-3’ 3’-CAATTG-5’

T

which is one of the four Hind I1 recognition sites. Polyoma DNA is not cleaved by Hpa I. Therefore from among the four potential cleavage sequences for Hind 11, the latter is clearly absent in polyoma DNA. For SV40 DNA there are seven Hind I1 sites and only four Hpa I sites (Danna et al., 1973; Yang et al., 1975). Therefore, at least two of the four possible sequences recognized by Hind I1 must be presented in SV40 DNA, and all four are possibly present. The restriction enzyme Hha I, which cleaves polyoma DNA at three sites (Griffin and Fried, 1975) and SV40 DNA at two sites (K. N. Subramanian, S. Zain, R. J. Roberts, and S. M. Weissman, unpublished), recognizes the sequences (Roberts et al., 1976) :

L

5‘-NGCGCN-3’ 3’-NCGCGN-5’

T

POLYOMA AND

sv40

GENOMES

81

where N is any nucleotide Hae I1 which cleaves polyoma and SV40 DNAs only once (Griffin and Fried, 1976; Roberts et al., 1975, respectively) recognizes the sequence (B. G. Barrel1 and P. M. Slocombe, unpublished) :

.Cr

5’-RGCGC -3’ 3’-YCGCGR-5’

t

where R = purine, Y = pyrimidine. It is clear therefore that, at least for one of these enzymes, sequence recognition involves more than the central four nucleotides and that Hha I must be recognizing at least two different sequences in both polyoma and SV40 DNAs, one of which is also recognized by Hae 11. As more data become available both with respect to the sequences involved in recognition by restriction endonucleases and their presence in polyoma virus and SV40 DNAs (see Figs. 1 and 2 ) the primary sequence information obtained by restriction enzyme analysis could become meaningful. IV. Origin and Termination of Viral DNA Replication

Restriction enzyme analyses combined with other techniques have helped define the intracellular replicative mode of polyoma virus and SV40 DNAs and to locate the origins of DNA replication. Polyoma virus and SV40 DNAs appear to replicate according to the Cairns model (Cairns, 1963) of DNA synthesis. By electron microscopic analysis of polyoma and SV40 replicative forms (Hirt, 1969; Bourgaux et al., 1969; Levine et al., 1970; Sebring et al., 1971), linearized by the restriction enzyme Eco RI, Fareed et al. (1972) and Crawford et al. (1973) concluded that replication was initiated from a unique origin site (or region) and that the two replication forks moved in opposite directions at approximately equal rates. The unique origin ( 0 )has been located on the polyoma physical map at 71 3 map units near the Hpa 11-3-5 junction (Fig. 1 ) by two methods, one based on an electron microscopic analysis (Fig. 4) of replicative forms cleaved with restriction enzyme Hind I11 (two cleavage sites) (Griffin et al., 1974), and the other on restriction enzyme analysis of pulse-labeled replicative form ( Crawford et al., 1974). A minor origin has been reported to exist at, or very close to, the Eco RI cleavage site (Robberson et al., 1975). Since replication appears to proceed bidirectionally at approximately equal rates, this would place the termination of DNA replication ( T ) 50

*

82


FIG. 4. Electron micrographs of ( A ) a replicating DNA molecule (less than 50% replicated) and ( B and C ) the smaller (Hind 111-2) and larger (Hind 111-1) fragments (see Fig. 1) obtained after cleavage of a replicating molecule with Hind 111. The nlidpoints of the bubbles in molecules represented by ( C ) were measured t o be 25 -C 3%from one end and 30 & 3%from the other. Replicating bubbles are indicated by arrows. ~ 4 5 , 0 0 0 .From Griffin et al. (1974).

map units away from the origin, at 21 =k 3 map units, near the Hpn II2-6 junction (Fig. 7 ) . In a similar fashion the unique origin for SV40 DNA replication has been located at 0.67 map units and termination at 0.14 map units on the SV40 physical map (Fig. 7 ) (Nathans and Danna, 1972b; Danna and Nathans, 1972). Termination of DNA replication of these viral DNAs is probably not sequence specific, but occurs about 180” away from the site of initiation on the circular DNA molecule. This is borne out by the efficient replication of cloned defective polyoma and SV40 variants (Griffin and Fried, 1975; Lee et al., 1975, respectively) which appear to lack the region of the molecule where termination normally occurs in nondefective molecules. Moreover, in cloned SV40 defective variants in which the “normal termination” sequences, although retained, are topologically displaced (and no longer 180” from the origin of replication), termination of DNA replication is not impaired. In these molecules, a region 180” from the origin of replication is utilized for termination (Brockman et al., 1975; Lai and Nathans, 1975a). I t is thought that the origin of DNA replication of these two viral DNAs is sequence specific. As a result, the nucleotide sequences around the origin of SV40 (Subramanian et al., 1976) and polyoma virus ( E . Ziff, unpublished results; B. E. Griffin, unpublished results) are currently being determined. Mutations in the regions (early) of the polyoma and SV40 genomes which are required for viral DNA synthesis are also required for the initiation of transformation (Fried, 1965, 1970; Eckhart, 1969; Franke and Eckhart, 1973; Di Mayorca et al., 1969; Tegtmeyer, 1972; Kimura and Dulbecco, 1973; Martin and Chou,

POLYOMA AND

sv40

GENOMES

83

1975) (see Section VIII). Evidence suggests that mutations in this early gene affect only the initiation of viral DNA synthesis; not elongation (Tegtmeyer, 1972, Franke and Eckhart, 1973) (see Section VIII). It is not clear what role the viral origin plays in initiating and/ or maintaining cell transformation, but it has been postulated that a virus-specified protein which binds to the origin and is needed for viral DNA replication is also active in transformed cells, stimulating host-cell DNA synthesis by means of the viral origin (which is now part of the host chromosome). A simple interpretation of this hypothesis fails, however, to explain the observation that under conditions where viral DNA synthesis is greatly inhibited (that is, at nonpermissive temperature, in cells infected with viral mutants which are temperature sensitive with respect to the only early viral function identified), the stimulation of host DNA synthesis is scarcely altered (Fried, 1970; Tegtmeyer and Ozer, 1971; Kimura and Dulbecco, 1972; Chou and Martin, 1974). A more detailed analysis of the replication of polyoma virus and SV40 DNA can be found in reviews by Salzman and Khoury (1974) and Levine ( 1974). V. location of Virus-Specific RNAs

In recent years numerous studies have been made on the transcription of the small papovaviruses in lytically infected and transformed cells (for reviews, see Sambrook, 1975; Salzman and Khoury, 1974). Attempts to map the position and polarity of the early (synthesized before viral DNA replication) and late (synthesized after viral DNA replication) viral specific RNAs in lytically infected cells have been made in a number of laboratories (Sambrook et al., 1972, 1973; Khoury et al., 1972, 1974, 1.975c, 1976; Kamen et al., 1974; Dhar et al. 1974a; May et al., 1975; Kamen and Shure, 1976). The general conclusions are that most of the early and late stable cytoplasmic messenger RNAs (mRNA) are synthesized from different contiguous portions of the genome and from different viral DNA strands; the early and late regions each comprise about 50% of the genome. Nuclear virus specific RNA, however, contains sequences not found in the cytoplasm; RNA sequences complementary to all of both DNA strands are probably present, but not in equal abundance ( Aloni 1972, Aloni and Locker, 1973; Kamen et al., 1974; Laub and Aloni, 1975). The 3’ ends of the stable mRNAs are polyadenylated (Weinberg et al., 1972a; Rutherford and Hare, 1974) whereas the 5’ ends are blocked with 7-methylguanosine ( Lavi and Shatkin, 1975). These modications at the RNA termini may be important for protection of the message (Marbaix et al., 1975) or for transport from the nucleus or they may play a role

84


in regulation of protein synthesis. The 19-20 S early messenger RNA appears to code for a viral protein that is immunologically related to Tantigen; the 16 S late mRNA appears to code for the major viral capsid protein (VP-1) and the 19 S late mRNA to code for the minor viral capsid proteins VP-2/VP-3 (see below). The stable early cytoplasmic viral messenger RNA is usually detected as a single species with a sedimentation coefficient of approximately 1920 S (Weinberg et al., 1972b, 1974; Weil et al., 1974; Kamen and Shure, 1976). This early 19-20 S RNA is complementary to about 50%of a continuous portion of the viral genome. In the case of SV40 it extends from near the origin of DNA replication through fragments Hind-A, H, I, and B to about the Hind-B-Hind-G junction (Dhar et al., 1974a; Khoury et d.,1976) (see Fig. 7 ) . In the polyoma virus genome the early 19-20 S mRNA extends from a site in Hpa 11-5 near the origin of DNA replication through fragments Hpa 11-5, 4, 8, 7 and 2 to a site in Hpa 11-6 ( Kamen and Shure, 1976; see Fig. 7 ). The stable late cytoplasmic mRNA is usually detected as two distinct species with sedimentation coefficients of 16 S and 19 S (Tonegawa et al., 1970; Weinberg et al., 197213, 1974; Buetti, 1974; Kamen and Shure, 1976). All the sequences found in the 16 S RNA are also found in the 19 S RNA (Weinberg et al., 1974; Kamen and Shure, 1976). The 19 S RNA is homologous to about 50%of the viral DNA and contains sequences not present in the 16 S RNA; the SV40 late 19 S RNA is thought to be a precursor of the 16 S RNA (Aloni et al., 1975). In the case of the SV40 genome the late 19 S mRNA extends from a site in Hind C through HindL, M, D, E, K, F, J, and G to a site near the Hind-G-B junction. The late 16 S mRNA extends from near the Hind-E-K junction through Hind-K, F, J and Hind-G to the Hind-G-B junction (Dhar et al., 1974a; May et al., 1975; Khoury et al., 1976) (see Figs, 1 and 7). For polyoma virus, the.late 19 S mRNA extends from a region in Hpa 11-3 through Hpa 11-1and into Hpa 11-6, whereas the late 16 S mRNA extends from a region in Hpa 11-1 (near the Hpa 11-1-3junction) into Hpa 11-6 (Kamen and Shure, 1976) (see Fig. 7). The 5’ ends of both 19-20 S early and 19 S late mRNAs have been mapped near the Hpa 11-3-5 junction for polyoma virus and near the Hind-A-C junction for SV40; the 3’ ends map in Hpa 11-6 and around the Hind-B-G junction, respectively (Kamen et al., 1974; Dhar et al., 1974a; Sambrook et al., 1973). For a review of transcription of polyoma virus and SV40 during lytic (productive) infection, see Acheson ( 1976). In both polyoma- and SV40-transformed cells most of the detected virus-specific RNA is transcribed from the same viral DNA strand as that transcribed early in the lytic cycle. This RNA includes sequences found

POLYOMA AND

sv40

GENOMES

85

in RNA made early in the lytic cycle, and in some cases sequences from the late region of DNA as well; the latter sequences have in a number of instances been shown to be transcribed from the strand opposite to that used in transcription of late RNA (the so-called anti-late RNA) (Sambrook et al., 1972; Khoury et al., 1973, 1975a; Ozanne et al., 1973; Kamen et al., 1974). Thus, if a viral specified protein is required for the maintenance of the transformed state, it is probably coded for by sequences from the early region of the viral genome. It is difficult to interpret the significance of viral transcription in transformed cell because it has not been shown that the transformed phenotypes of the cells still result from the original transformation by the viral genome. VI. Virus-Induced Proteins

The polyoma and SV40 viral capsids contain 3 major proteins that are thought to be virus coded (VP-1, VP-2, VP-3) and 4-5 histone proteins that are host coded (see Tooze, 1973). In some strains of SV40 a distinct VP-2 is not present in virus particles. The major capsid protein VP-1 (40,000-48,000 daltons) has been shown to be unrelated to VP-2 (30,00035,000 daltons ) or VP-3 (20,000-25,000 daltons ) by peptide mapping (Hewick et al., 1975; Gibson, 1974, 1975). In the case of polyoma virus, it has been shown that VP-2 contains almost all the peptides found in VP-3 in addition to a number of unique peptides (Hewick et al., 1975). It is not clear whether VP-3 and VP-2 are related in a biologically significant way in the cell or whether VP-3 arises as an artifact of proteolysis during isolation. In lytically infected cells or in transformed cells at least three new antigens have been detected. One is T antigen (Black et al., 1963; Takemot0 and Habel, 1965) (assayed by immunofluorescence or complement fixation), which is thought to be, at least in part, coded for by the virus. In both SV40 lytically infected and transformed cells a new protein has been detected, which has an apparent molecular weight of approximately 100,000 (which corresponds to that of T antigen) and reacts with antiserum containing antibodies against T antigen (Tegtmeyer et al., 1975). The SV40 T antigen has been shown to bind to double-stranded DNA from a variety of sources (Carroll et al., 1974; Jesse1 et al., 1975). This protein is phosphorylated and may be modified in other ways (P. Tegtmeyer, personal communication). The two other antigens, transplantation antigen ( Habel, 1961; Sjogren et al., 1961; Habel and Eddy, 1963; Khera et al., 1963; Koch and Sabin, 1963) and S-antigen (Tevethia et al., 1965; Malmgren et al., 1968; Irlin,

86

MME FRIED AND BEVERLY GRIFFIN

1967), are found at the cell surface. Transplantation antigen is assayed by the ability of animals immunized with live virus to reject cells transformed by that virus, and S antigen is assayed by immunofluorescence. In addition to the above three antigens, another antigen, U-antigen (Lewis and Rowe, 1971), is found in the cytoplasm of SV40-infected or transformed cells ( assayed by immunofluorescence). The BK papovavirus isolated from a human source contains a T antigen that cross-reacts with SV40 T antigen; capsid antigens are also immunologically related (Takemoto and Mullarkey, 1973). Nucleic acid hybridization studies have shown that SV40 and BK virus have detectable sequence homology only in their late region (Khoury et aZ., 1975b). The T antigen, S antigen, transplantation antigen, and capsid antigen are specific to the infecting virus; that is, there is no immunological cross-reaction between the antigens induced by polyoma virus and SV40. Needless to say, there is not enough coding information in the viral genomes for all these antigens. Thus, some may be different forms of the same protein or specifically induced or modified (e.g., by glycosylation) host proteins. Recently the use of in uitro protein synthesizing systems has allowed some of these proteins to be identified (at least in part) as being virus coded. By the use of defined viral mRNAs in these systems, the viral proteins have been assigned to regions on the DNA physical map (see Fig. 7). The addition of stable late 16 S mRNAs from SV40- or polyma virus-infected cells to in uitro protein-synthesizing systems leads to the production of a protein with electrophoretic mobility identical to the respective viral protein VP-1 (Prives et al., 1974; Smith et aZ., 1975). The addition of SV40 or polyma 19 S stable late mRNAs leads to the synthesis of a protein with electrophoretic mobility similar to the viral protein VP-2 (Prives et al., 1974; Smith et al., 1975). In all cases, the mRNAs were purified by means of their polyadenylated 3’ ends. The addition of in uitro synthesized polyoma complementary RNA (cRNA, made by E. coli RNA polymerase from the polyoma virus DNA strand which codes for late mRNA) to an fn uitro protein-synthesizing system results in the production of proteins with electrophoretic mobilities identical to both polyoma VP-1 and VP-2 (Smith et aZ., 1975). Furthermore, peptide maps of the proteins produced from viral RNAs synthesized either in uitro or in duo have showed these proteins to be indistinguishable from VP-1 and VP-2 (Prives et al., 1974; Smith et al., 1975; R. M. Hewick and W. F. Mangel, personal communication). The 19 S mRNA contains all the sequences present in the 16 S late mRNA and is thought to be the precursors of the 16 S RNA; the fact

POLYOMA AND

sv40

GENOMFS

87

that 19 S late mRNA does not induce VP-1 synthesis indicates that some sort of control must exist in the synthesis of the late proteins. Such control is not evident when in vitro synthesized polyoma late-strand cRNA is added to an in vitro protein-synthesizing system, since apparently both proteins VP-1 and VP-2 are produced. The difference between the translation of the in uitro and in vivo synthesized RNA may be a result of different promoters being used for RNA synthesis in the two cases. On the other hand, RNA made in vivo may contain some modification (e.g., methylation) which acts as a controlling element that is absent in the in vitro synthesized RNA. In any case it is clear that VP-2 and VP-3 (which are related) are encoded within a region of approximately one-quarter of the genome which begins near the origin of DNA replication, whereas VP-1 is encoded within a region of approximately one-quarter of the genome near the termination of DNA replication (see Fig. 7 ) . More precise data for the amino-terminal end of the SV40 VP-1 is available. The four N-terminal amino acids of this protein have been sequenced (Lazarides et al., 1974) and the corresponding DNA sequences that code for these amino acids have been determined and located on the physical map of SV40 DNA in the Hind-K fragment (Fiers et al., 1975; see Section 111). The studies on the protein (or proteins) made from viral early mRNA are less clear. When early mRNA made in viuo, or cRNA synthesized in uitro from the strand of SV40 DNA that codes for early mRNA, is added to an in vitro protein-synthesizing system, a protein (or proteins) is produced that reacts with antiserum containing antibodies to SV40 T antigen (Prives et al., 1974; Smith et al., 1975). The major immunoprecipitable protein made from SV40 cRNA migrates faster in acrvlamide gels than the protein made from RNA isolated in vivo (W. F. Mangel, personal communication), The latter (slower migrating) protein has an electrophoretic mobility similar to the protein( s ) isolated from SV40 transformed cells which reacts with antiserum containing antibodies to T antigen (R. B. Carroll and A. E. Smith, personal communication). The size difference observed for the proteins may point to a modification (e.g., methylation) in the RNA made in vivo that is not found in the cRNA synthesized in vitro and which prevents premature chain termination. Alternatively, the two RNAs may contain only partial sequence homology (e.g., the in vivo RNA could contain some host sequences or the in vitro RNA may contain only a portion of the sequences of the invivoRNA). Microinjection of either polyoma cRNA (Graessmann et al., 1975) or SV40 cRNA (Graessmann and Graessmann. 1976) into permissive cells induces new virus-specific nuclear antigens that can be detected by

88


immunofluorescence with antisera that contain antibodies to the appropriate T antigen. What is still not clear in all this work is whether the same antibodies in the antiserum containing activity against T antigen are also reacting both with those proteins made from in vitro RNA and in vivo RNA and those isolated from lytically infected and transformed cells. Immunological blocking experiments and detailed peptide maps of the various proteins should answer this question of identity. If one assumes that these proteins are identical (or related), T antigen must be at least in part coded for by the early region of the viral genome. VII. Protein Binding Sites on Viral DNAs

The T4 gene 32 protein binds to double-stranded DNA in regions that are easily denatured and usually rich in A and T; Delius et al. (1972) showed that under the appropriate conditions T4 gene 32 protein formed one denaturation loop per SV40 supercoiled molecule. Morrow and Berg (1972) showed that this denaturation loop was in a unique region of the SV40 DNA, and mapped this region 45% away from the single Eco RI cleavage site ( Morrow and Berg, 1973). By an electron microscopic analysis of gene 32-polyoma DNA complexes, cleaved with either Eco RI or Bam I (one cleavage site) or Hind I11 (two cleavage sites), Monjardino and James (1975) and Yaniv et al. (1975) were able to show binding of gene 32 at five or six different sites and to map them on polyoma DNA at 0, 8, 23, 47, 56, and 80 map units. The predominant binding site was found to be at 23 map units on the polyoma physical map, in Hpa 11-6 (an A,T-rich Hpa I1 fragment, Griffin et al., 1974) near the termination region for DNA replication (sce Fig. 7). The binding sites for gene 32 in polyoma virus and SV40 DNAs correlate very well with those regions of the DNAs that are most easily denatured by alkali (Mulder and Delius, 1972; Lescure and Yaniv, 1975). It has also been observed that both polyoma virus and SV40 DNAs isolated either from alkaline disrupted virus particles or intracellular pools have histones bound to them. These histones are arranged in an orderly manner on the DNA (Louie, 1974; Griffith et al., 1975); each group of associated histones (containing approximately 150-160 nucleotide pairs) are separated by regions of approximately 40-50 base pairs that contain no histones. Analysis of these histone-DNA complexes by restriction enzymes indicates that the histones are bound to different regions in different molecules of viral DNA ( McCann and Martin, 1975; Polisky and McCarthy, 1975). When SV40-histone complexes are treated

POLYOMA AND

sv40

GENOMFS

89

with high salt (to remove the histones) a viral capsid protein( s ) is found still associated with the viral DNA (Huang et al., 1972). PreIiminary experiments suggest that the protein is SV40 VP-1 (T. Maniatis and M. Botchan, personal communication). A protein appears to be bound at or near the origin of viral DNA replication (Griffith et al., 1975). A number of attempts have been made to determine whether the virally induced T antigen binds to a specific region of the viral DNA. Reed et al. (1975) by electron microscopy observed a slight preferential binding of semi-purified SV40 T antigen to supercoiled SV40 DNA at or near the origin of viral DNA replication. In other experiments, where restricted fragments of SV40 DNA have been utilized it has been found that semipurified SV40 T antigen binds to a number of different regions of the viral DNA (D . M. L. Livingston, personal communication); binding occurs preferentially with the largest restriction fragments. Since the T antigen used in these binding experiments is only semipurified, it may already contain DNA at its major binding site. Conclusive results on specific binding must await further purification of the protein. VIII. Genetic Mapping

A number of temperature-sensitive mutants of polyoma virus and SV40 have been isolated. Three distinct groups can be distinguished by complementation tests (Eckhart, 1969, 1974; Di Mayorca et al., 1969; Tegtmeyer and Ozer, 1971; Kimura and Dulbecco, 1972, 1973; Chou and Martin, 1974; Dubbs et al., 1974), although by other criteria the mutants have been divided into five groups. In one complementation group ( early mutants) physiological tests have shown that the mutation affects a function required for viral DNA synthesis; this group is represented by the A mutants of SV40 and by TS-A mutants of polyoma virus (Fried, 1965a, 1970; Eckhart, 1969; Di Mayorca et al., 1969; Francke and Eckhart, 1973; Tegtmeyer, 1972; Kimura and Dulbecco, 1973; Chou et al., 1974; Yamaguchi and Kuchino, 1975). The other two complementation groups contain mutants (late) that have defective viral capsid proteins; these are represented by the B and C mutants of SV40 and TS1260 and TSlO mutants of polyoma virus (Eckhart, 1969; Di Mayorca et al., 1969; Tegtmeyer and Ozer, 1971; Kimura and Dulbecco, 1972; Dubbs et al., 1974; Chou and Martin, 1974). Another group of mutants also have defective viral capsid proteins; these mutants complement the early mutants but do not complement either of the two late classes of mutants mentioned above and are represented by the BC mutants of SV40 (Tegtmeyer and Ozer, 1971; Chou and Martin, 1974). The polyoma virus late mutant TS-C (M. Fried, un-

90


published) appears to have similar complementation properties to the BC mutants of SV40 ( W. Eckhart, personal communication). A fifth group of temperature-sensitive mutants represented by the D mutants of SV40 and the TS-3 mutant of polyoma virus, do not complement any of the other groups. Physiological tests have failed to show any evidence of viral infection at the nonpermissive temperature when permissive cells are infected with mutant virus particles. On the other hand, such cells infected with the mutant viral D N A induce normal yields of infectious mutant particles at the nonpermissive temperature. N o early functions are expressed and no complementation takes place with these mutants (Eckhart and Dulbecco, 1974; Robb and Martin, 1972; Chou and Martin, 1974; Martin et aZ., 1974). The mutation appears to affect a protein in the viral capsid that is temperature sensitive in its release from the viral DNA; release is required for expression of the viral genetic information. Thus these mutants can be considered to be late mutants. The SV40 A mutants and the polyoma virus TS-A mutants are temperature sensitive for the initiation of transformation (Fried, 1965b; Eckhart, 1969; Di Mayorca et al., 1969; Tegtmeyer, 1972; Kimura and Dulbecco, 1973; Martin and Chou, 1975). Recently it has been reported that a number of phenotypic markers of transformation may be temperature sensitive in some cells transformed by SV40 A mutants (Kimura and Itagaki, 1975; Brugge and Butel, 1975; Tegtmeyer, 1975; Osborn and Weber, 1975; Martin and Chou, 1975). But this is not always the case with the same mutants either in the same or in different cell types; thus a more careful analysis will have to be performed before it can definitely be stated that the SV40 A function is required for the maintenance of the transformed state [see Nature (London) 255, 367-368 ( 1975) 1. So far no other strong evidence exists for the requirement of a viral gene product for the maintenance of transformation. Cells transformed by the polyoma TS-A mutants do not appear to be temperature sensitive for the maintenance of transformation (Fried, 1965b; Eckhart, 1969; Di Mayorca et al., 1969). A protein with a molecular weight of approximately 100,000 (which reacts with antiserum containing antibodies to SV40 T antigen) appears to turn over rapidly at the nonpermissive temperature in cells lytically infected or transformed with SV40 A mutants (Tegtmeyer et al., 1975). In lytically infected cells, one of the molecular forms of T antigen induced by A mutants has been reported to be more heat sensitive than the corresponding wild-type molecular form, as measured by complement fixation ( Kuchino and Yamaguchi, 1975). Also semipurified T antigen isolated from cells transformed with SV40 A mutants appears to be more heat labile than wild-type SV40 T antigen with respect to its ability to

POLYOMA AND

sv40

GENOMES

91

bind to double-stranded DNA (Tenen et al., 1975; Alwine et al., 1975). In mouse cells, either lytically infected or transformed by the polyoma TS-A mutant, the T antigen induced appears to be tempeiature sensitive, as assayed by immunofluorescence or complement fixation (Oxman et al., 1972; Paulin and Cuzin, 1975). The SV40 B, C, and BC mutants and their polyoma virus equivalents are not temperature sensitive either for the initiation or maintenance of transformation (Eckhart, 1969; Di Mayorca et al., 1969; Tegtmeyer and Ozer, 1971; Kimura and Dulbecco, 1972; Martin and Chou, 1975). The SV40 D mutants are temperature sensitive for transformation (or transform poorly at both permissive and nonpermissive temperatures depending upon the cell type) (Robb et al., 1972; Martin and Chou, 1975). This is consistent with the D mutants being impaired for expression of the viral DNA after infection with virus particles (see above). Cells transformed by SV40 D mutants do not appear to be temperature sensitive for most phenotypic markers of transformation although it has been reported that T antigen as assayed by immunofluorescence is temperature dependent in some mouse lines transformed by an SV40 D mutant (Robb, 1973). The polyoma mutant TS-3 does not appear to be temperature sensitive for the initiation of transformation as measured by the formation of transformed colonies in semisolid medium (Eckhart et aE., 1971). TS-3 transformed BHK cells do, however, appear to be temperature dependent with respect to a number of other phenotypic markers of transformation (Eckhart et al., 1971; Dulbecco and Eckhart, 1970). Unfortunately, only one TS-3-transformed BHK cell line has been studied with respect to temperature sensitivity for maintenance of transformation, leaving open the possibility that this observation was due to a cellular rather than a viral mutation. In addition to temperature-sensitive mutants, a number of plaque morphology variants exist for both polyoma virus and SV40. They consist of either small-, large- or minute-plaque types. A number of small-plaque variants of polyoma virus have been reported to have reduced transforming ability ( Gotlieb-Stematsky and Leventon, 1960; Sachs and Medina, 1960; Hare and Morgan, 1962). In the case of polyoma virus, host-range mutants exist. These mutants, exemplified by NG-18, have been reported to grow and plaque on certain cell types (usually transformed cells induced by a number of agents), but not on 3T3 cells (Goldman and Benjamin, 1975). Physiological tests indicate that the NG-18 type mutants stimulate host DNA synthesis and produce infectious viral DNA, but are inhibited in their ability to produce infectious virus and to expose lectin agglutinable sites in per-

92


missive cells (Benjamin and Burger, 1970; W. Eckhart, personal communication). The NG-18 mutants have also been reported to lack the ability to transform ( Benjamin, 1970). Unfortunately the characteristics of the small plaque wild-type parent of NG-18 (e.g., its transforming ability) remain obscure and work on these host range mutants has been restricted. A number of deletion and addition mutants (both infectious and noninfectious) of polyoma virus and SV40 exist. In the case of SV40, a number of general methods have been devised for the generation of these types of mutants (Mertz and Berg, 1974a; Mertz et al., 1974; Lai and Nathans, 1974a; Carbon et al., 1975; Shenk et al., 1976 see Section IX). Noninfectious deletion and addition mutants are grown in the presence of a helper virus, usually a temperature sensitive mutant, which is complemented by the intact gene present in the addition or deletion mutant ( Brockman and Nathans, 1974; Mertz and Berg, 1974b). A number of methods have been used to map mutations in polyoma virus and SV40 DNA. The method used most successfully to map temperature-sensitive and plaque-morphology mutants (which are probably due to single base changes) has been the marker rescue technique first used for mapping mutants of 4x174 (Weisbeek and Van De Pol, 1970; Edge11 et al., 1972; Middleton et al., 1972; Weisbeek et al., 1973). As presently used, this method involves the infection of cells with individual wild-type restriction enzyme fragments which have been hybridized to complete mutant single-stranded DNA, If the sequences present in the restriction enzyme fragment are almost entirely complementary to that part of the viral genome containing the mutant sequence, wild-type virus is produced. The mismatched base pair in the heteroduplex DNA may be repaired (Miller et al., 1976) using the wild-type sequence as template part of the time (to give wild-type virus) and the mutant sequence as template at other times (to give mutant virus). With this technique, 13 early mutants of the SV40 A group and the seven early mutants of the polyoma TS-A type have been mapped. Eight of the SV40 A mutants map in restriction enzyme fragment Hind-I, four in Hind-H, and one in Hind-B (Lai and Nathans, 1974b, 197513; Mantei et al., 1975). Thus the genetic lesions affecting the SV40 A function are found in sequences extending from 0.17 to 0.45 map units (see Fig. 7). Five of the polyoma TS-A mutants map between the restriction enzyme Hha I site in Hpa 11-2 and the Hha I site in Hpa 11-6, whereas the other two map between the Hind I11 site in Hpa 11-2 and the Hha I site in Hpa 11-2 (Miller and Fried, 1976a). The genetic lesions affecting the polyoma TS-A function are found in sequences extending from 1.0 to 26 map units on the physical map (see Fig. 7).

POLYOMA AND

sv40

CENOMES

93

It is not clear why all these TS mutants map in only a restricted part of the early gene region. One explanation is that there are actually two early proteins (see Section 111) and only one has so far been genetically located. Another explanation is that mutations may be restricted to a portion of a single early gene since only a limited portion of the early protein can accommodate amino acid changes and although inactivated at nonpermissive temperature still retain function at the permissive temperature. Similarly, six mutants of the SV40 B group have been mapped in restriction enzyme fragment Hind-F, two in Hind-K, and one in Hind-J. Four mutants of the SV40 C group have been mapped in Hind-J and one in Hind-K. Seven mutants of the SV-40 BC group have been mapped in Hind-G and one in Hind-J (Lai and Nathans, 1974b, 1975b; Mantei et al., 1975). The SV40 B, C, and BC late mutants are located in overlapping portions on the genome (between 0.95 and 0.17 map units, see Fig. 7 ) from which the 16 S late mRNA (which codes for SV40 VP-1, see Section VI) is derived. Similar observations have been made for equivalent polyoma late mutants. Polyoma mutants TS-1260, TS-10, and TS-C have all been mapped (Miller and Fried, 1976a) between the Hind I11 site in Hpa 11-1 and the Hpa I1 1-6 junction (between 27 and 46 map units, see Fig. 7 ) in the region of the genome from which the late 16 S mRNA (which codes for polyoma VP-1, see Section VI) is derived. Thus, the complementation observed between the SV40 B and C mutants (Kimura and Dulbecco, 1972; Chou and Martin, 1974) and polyoma mutants TS-1260 and TS-10 ( Eckhart, 1974) probably indicates intragenic complementation (within a gene), not intergenic complementation (between genes) as previously thought. A polyoma minute-plaque variant (208) derived from the large-plaque A-2 strain has also been mapped between the Hind I11 site in Hpa 11-1 and the Hpa I1 1-6 junction (Miller and Fried, 1976a); therefore this plaque morphology is a property of polyoma VP-1. A correlation has been made between difference in plaque morphology and a change in a single tryptic peptide of VP-1 (i.e., between the minute-plaque 208 and a largeplaque 208 virus isolated after marker rescue with A-2 restriction fragment Hpa 11-1 ( Hewick et al., 1977). Seven mutants of the SV40 D group have all been mapped in fragment Hind-E (Lai and Nathans, 1974b, 1975b; Mantei et al., 1975), a portion of the SV40 DNA that contains the genetic information for the minor related capsid proteins VP-2/VP-3 (see Section VI and Fig. 7 ) . This clearly demonstrates that the D function is a late function. It is not clear why these mutants are restricted to such a small region of the SV40 genome (0.86 to 0.95 map units). The TS-3 mutant of polyoma virus

94

MIKE FRIED A N D BEVERLY GRIFFIN

(Eckhart and Dulbecco, 1974), which is similar to the mutants of the SV40 D group, has not yet been mapped. From the SV40 data it would not be surprising to find that it is located between the Hind I11 site in Hpa 11-1 and the Hpa 11-35 junction on the polyoma physical map. Another technique that has proved successful in the mapping of some temperature-sensitive mutants has been developed by Shenk et al. ( 1975). This method utilizes the ability of the S1 nuclease to cleave at mismatched base pairs in heteroduplex molecules formed between a DNA strand of the mutant in question and the complementary DNA strand from a revertant of the mutant. The method has the advantage that it allows a mutation to be mapped more precisely than by the “markerrescue” method with less effort. It has the disadvantage that other mismatched base pairs (apparently silent mutations), as well as rich A,T regions, may also be sensitive to S1 cleavage and that not all mismatched base pairs may be efficiently cleaved by the action of the enzyme. Nonetheless, this method has proved to be useful for the mapping of small addition and deletion mutants (see Section IX). One SV40 A mutant and one SV40 D mutant have been mapped by this S1 method; the results are in full agreement with results obtained by the marker-rescue method. A third method that has been used for genetic mapping takes advantage of the fact that certain restriction enzyme fragments, owing to their cohesive ends, can be annealed and joined by DNA ligase. Thus, infectious hybrid molecules can be constructed in uitro from restriction enzyme fragments of DNAs with different genetic markers. Analysis of the phenotypes of the hybrid viruses allows the different markers to be located in specific regions of the genetic map. For example, the small-plaque morphology marker of the P16 strain of polyoma virus has been localized in the 44%Hind I11 fragment of polyoma virus DNA (Miller and Fried, 1976b). A fourth method that has been used for the location of large deletions and additions involves analysis of heteroduplex molecules by electron microscopy (Davis et al., 1971). It utilizes the hybridization of two different DNA molecules (of which at least one is frequently a restriction enzyme fragment) followed by inspection in the electron microscope in order to locate the addition, deletion, or substitution loops. This method has proved to be extremely valuable in the definition of defective DNAs where deletions, additions, and rearrangements are frequently present (Robberson and Fried, 1974; Mertz et al., 1974; Brockman et al., 1974; Khoury et al., 1974) (see Sections IX and X ) . IX. Essential and Nonessential Regions of the Viral Genomes

Two major approaches have been used to determine the regions of the viral genome essential for infection and/ or transformation of cells.

POLYOMA AND

sv40

95

GENOMFS

One has involved the study of variants containing additions or deletions in the viral genome. The other has involved assessing the biological activity of restriction enzyme fragments made from viral DNA. Certain regions of the polyoma virus and SV40 genomes can be modified ( either by deletions, additions, or alterations of nucleotide sequences) without loss of infectivity. For polyoma virus, a number of naturally occurring infectious variants have been isolated. One such variant contains a small deletion (about 0.3%) close to the origin of DNA replication in Hpa 11-5 (within 1.5%from the Hpa JI-3-5 junction), which does not affect either the infectivity or transforming activity of the virus (Fried et al., 1974). Another variant contains an addition of approximately 1.5%in Hpa 11-3 (Francke and Vogt, 1975), between 67 and 71 map units ( M . Fried, unpublished results); it also is fully active biologically, Two other polyoma viral variants contain an even greater number of extra sequences in the region of the origin of viral DNA replication. These two variants, which are related to each other, contain an additional Hpa I1 fragment which is either about 4.5 or 5.0%the length of the full-length genome and is located between Hpa 11-3 and Hpa 11-5 on the polyoma virus physical map, Analysis of the sequences present in the extra fragment of both variants has revealed that they are composed of the sequences from both Hpa 11-3 and 5 near the Hpa 11-3-5 junction (Fried and Griffin, 1977)(see Fig. 5 ) . Both of these variants contain two

A.2

251

FIG. 5. Physical maps of the polyoma A-2 wild-type DNA (see Fig. 1 ) and the polyoma infectious mutant 251. Compared with A-2 DNA, 251 contains an additional Hpa I1 restriction fragment (about 5% of the genome), which has sequences from Hpa 11-5 (71-72 map units) and Hpa 11-3 (67-71 map units) as shown. A related infectious variant (not shown) has a structure similar to 251 except that the extra Hpa I1 fragment is only 4.5%of the genome, having lost 0.5%of the Hpa 11-3 sequences between 69 and 70 map units (see Section IX). From Fried and Griffin, 1977.

96


Hpa 11-3-5 junctions, and both are infectious and transform normally. Similar repetitions of sequences from the Hpa 11-3-5 junctions are found in a number of defective polyoma variants (see Section X and Fig. 6). From these results it would appear that some additions and deletions around the origin of DNA replication of polyoma virus do not affect its biological activity. The NG-18 mutant of polyma virus (see Section VIII) contains a deletion of approximately 2.7% in Hpa 11-4 (J. Feunteun and T. L. Benjamin, personal communication), which is located between 80 and 85 map units on the physical map of polyoma DNA ( M . Fried, unpublished results) (see Fig. 1).It is not yet known whether this deletion is responsible for the mutant phenotype of NG-18. Another region of variation detected in the polyoma virus genome lies near the Hpa 11-14 junction (see Fig. 1).One variant has lost the Hha I site in Hpa 11-6 at 26 map units and another variant has lost an Hae I11 cleavage site in Hpa 11-1 at 28 map units (L. K. Miller, C. Barry and M. Griffiths, unpublished results). Whether these changes are due to single base changes or to small deletions is unknown but neither change affects the infectivity or transforming activity of these variants. The variations lie near the 3' ends of the stable mRNA transcripts and may be in regions that do not code for viral proteins. Variations in the apparent sizes of restriction fragments Hind-C, Hind-F, and Hind-A have been observed in different isolates of SV40 (Brockman and Nathans, 1974; Botchan et al., 1974; Y. Ito, personal communication). An SV40 variant has been isolated with a 2%addition to the genome, in Hind-C; the additional sequences have, however, not yet been defined (Brockman and Nathans, 1974). Another SV40 variant has lost the Hind cleavage site between Hind-A and Hind-C (Y. Ito, personal communication). All these variants are infectious, and as far as is known transform efficiently. A number of SV40 viable deletion mutants have been selected as a result of their loss of either the Hpa I1 restriction enzyme cleavage site at 0.72 map units (Mertz and Berg, 1974a) or the Hae I1 restriction site at 0.83 map units (C. Cole, T. Landers, and P. Berg, personal communication), Both these cleavage sites are in the late region of the genome, possibly in sequences coding for VP-2/VP-3. The deletions in the Hpa I1 resistant molecules occur in the region between 0.72 to 0.76 map units (see Fig. 2). Some of these deletions have a small plaque morphology, This can apparently be correlated with the number of nucleotides lost around 0.76 map units, In the Hae I1 resistant molecules between 10 and 230 base pairs can be deleted without loss of infectivity.

POLYOMA AND

SVM

GENOMES

97

Such deletions result in a minute plaque morphology. Both the Hpa I1 and Hae I1 viable deletion mutants do not appear to be affected in their transforming ability. Other SV40 viable deletion mutants have been isolated following random cleavage of the DNA, mild exonuclease digestion, and recyclization (Carbon et al., 1975; Shenk et al., 1976). As far as is known, these deletions do not dramatically affect the transforming activity of the virus nor the production of SV40 T antigen (Shenk et al., 1976). These deletions have been mapped in three areas of the SV40 genome: One is in the late region at 0.68 to 0.74 map units (similar to the Hpa I1 resistant mutants mentioned above). Another lies at 0.17 to 0.18 map units at the 3’ end of the stable mRNAs. The third deleted region lies in the early portion of the genome between 0.54 and 0.59 map units on the SV40 physical map ( 1343%from the origin of DNA replication) (Shenk et al., 1976); interestingly enough the deletion in the polyoma mutant NG-18 is located in a similar region relative to the origin of polyoma virus DNA replication ( see above). Nonviable SV40 deletion mutants have been isolated as a result of their loss of the Eco R I cleavage site (a t 0 map unit) and the Bam I cleavage site (0.14 map unit) ( S . Goff and P. Berg, personal communication); both sites lie in the late region of the genome in sequences that code for VP-1. Mutants with deletions from 0.3 map units to 0.14 map units which are missing an Eco RII restriction enzyme fragment also exist (Lai and Nathans, 1974a). When cells are infected with these deletion mutants in the presence of helper virus a protein smaller than SV40 VP-1 appears to be produced. SV40 noviable mutants have also been isolated with deletions in the early region of the genome. One contains a deletion at 0.48 map units and was selected by its resistance to S1 endonuclease (S. Manteuil-Brutlag, P. Berg, personal communication); another mutant is missing the Eco R I1 restriction enzyme fragment which contains Hind-H and Hind-I (Lai and Nathans, 1974a). The latter mutant appears to induce a protein ( P. Tegtmeyer, personal communication ) with molecular weight lower than that of the putative T antigen [i.e., the protein with an apparent molecular weight of 100,000 (Tegtmeyer et al., 1975)l. Since it is difficult to separate these nonviable deletion mutants from their helper virus, no adequate transformation analyses have been performed. Transformation has been achieved with cloned defective SV40 DNA molecules which contain contiguous sequences from 0.16 to 0.67 map units, and rearrangements of these viral sequences (W. Scott, W. W. Brockman, C. J. Lai, and D. Nathans personal communication). These

98


molecules contain the entire early region. Polyoma defectives that are missing approximately one-half of the early region, from 72 to 2 map units, but contain the entire late region (with some sequence rearrangements) do not appear to transform (Fried et al., 1974). Transformation has also been achieved with isolated restriction enzyme fragments of SV40 DNA. The smallest fragment reported to transform contains the entire early region and extends from the Bam I site (0.14 map units) to the Hpa I1 cleavage site (0.72 map units) ( Abrahams and van der Eb, 1975). Cells so transformed contain only those viral sequences that are found in the fragment used for transformation (J. Sambrook and A. J. van der Eb, personal communication). Transformation with fragments limits those regions of the viral genome which need to be linked to host DNA (integration sites) in order to produce the transformed cell. Recent results (Botchan et aZ., 1976) obtained by cleaving the DNA from a number of different SV40 transformed cell lines with a variety of restriction endonucleases (some of which do and some of which do not cleave within the viral DNA) showed that integration into the host cell DNA occurred at a number of sites. That is, that there was no single specific integration site within the viral DNA. In all the transformed cell lines, at least one copy of uninterrupted early viral sequences was present; in those instances where the integration site was found to lie within the early region, an additional (uninterrupted) copy of early sequences was found. When the DNA from SV40 transformed cells was cleaved with restriction endonucleases with no cleavage site in the viral DNA, fragments of different sizes containing viral DNA sequences were detected from different cell lines (Kettner and Kelly, 1976; Botchan, et al., 1976). These results suggest that there are a number of different sites within the host DNA where the SV40 viral DNA can be integrated. They do not, however, iule out the possibility that integration occurs at specific DNA sequences that are repeated in different regions of the host and viral genomes. In conclusion, the polyoma virus and SV40 genomes can tolerate limited additions and deletions in various portions of their DNA without biological activity being affected. Most interesting of these are the deletions in the early region which lie approximately 8-14% away from the origin of viral DNA replication and do not appear to affect transformation. To date all the evidence points to the early region only as being required for initiation of transformation. Since all transformed cell lines so far analyzed appear to contain an uninterrupted copy of sequences from the viral early region integrated into the host DNA, it seems likely that the early region is also needed for maintenance of transformation.

POLYOMA AND

SVM

GENOMES

99

X. Defective Viral DNAs

Passage of polyoma virus or SV40 at high multiplickies of infection (high virus to cell ratios) results in the formation of virus particles containing noninfectious ( defective ) supercoiled DNA molecules heterogeneous in size ( Thorne, 1968; Thorne et al., 1968; Blackstein et al., 1969; Fried, 1974, Yoshiike, 1968a,b; Brockman et aZ., 1973). With continuous high-multiplicity passage, particles that contain defective DNA can become predominant in virus stocks. Most defective DNAs are shorter than nondefective DNA and contain restricted portions of the viral genome. In some cases, defective DNAs may also contain host sequences covalently linked to viral sequences ( Lavi and Winocour, 1972, 1974). Although defective molecules always contain some (if not all) viral sequences they are not simply viral deletion mutants. As well as deletions, the defective molecules may contain rearrangements and reiterations of viral sequences and in some cases host sequences. Some of these variations confer a selective advantage, at least for replication, to the defective molecules so that they become predominant in the viral stocks. Simple viral deletions usually do not have such an advantage and therefore have to be maintained either separately or by selective pressure. With continued high-multiplicity passage, the type of defective that predominates in early passages may be replaced by different types of defective molecules ( Brockman et al., 1973; Fried, 1974). Although defective species isolated at various stages contain different viral sequences, there appears to be some pattern that may be indicative of a type of evolution. Thus the sequences retained in defective viral molecules, the sequences reiterated, and those rearranged may prove useful in understanding the evolution of the v i r u s itself and its role as a transforming agent, which affects both function and control within the host cell. Defectives also serve as a useful source of restricted portions of the viral genome. The host DNA found in certain defective species may indicate the region of the host genome where viral integration takes place. Alternatively, it may indicate those host sequences that confer a greater selective advantage to the defective molecules. A number of polyoma and SV40 defective species have been cloned in the presence of either wild-type or mutant viral helper molecules (Fried, 1974; Brockman and Nathans, 1974; Mertz and Berg, 19741, ) . Populations of naturally occurring polyoma defective molecules have been successfully cloned in the presence of polyoma infectious (helper) molecules; where studied, these have been found to be noninfectious, to interfere with production of infectious polyoma virus, and to be resistant to cleavage with the Eco RI restriction endonuclease (Fried, 1974).

100


An electron microscopic study of heteroduplexes between Eco RI linear nondefective DNA and circular defective molecules has shown that the cloned polyoma defectives were not simply deletion mutants, but contained regions of apparent nonhomology in addition to regions of homology to wild-type A-2 strain polyoma DNA (Robberson and Fried, 1974). More detailed studies using restriction enzymes [especially Hae I11 which makes many cleavages in polyoma DNA (Fig. l ) ] , and pyrimidine tract analysis (see Section I11 and Fig. 3) has helped elucidate the structure of a number of these cloned polyoma defective DNAs (Fried et al., 1974; Griffin and Fried, 1975; Lund et al., 1977). In the polyoma viral defectives studied to date, two closely spaced regions are always retained. One lies at 70-72 map units in Hpa 11-5 near the junction with Hpa 11-3where the origin of DNA replication has been mapped (see Section IV); the other lies about 67-69 map units in Hpa 11-3 (see Fig. 1). The sequences between these two regions are not always retained. Rearrangement of the viral sequences found in some defective species also involves these two regions, sometimes joined to each other and sometimes to another (probably specific) region of the genome (Fig. 6 ) . Sequences from Hpa 11-5 ( a t about 72 map units and including the Hpa 11-3-5 junction at 71 map units) are joined to sequences in Hpa 11-3at about 67 map units. These two regions have also been found joined to other parts of the polyoma genome. The most frequently observed rearrangement involves sequences at about 72 map units in Hpa 11-5 joined to a point in Hpa 11-2 at about 0.5 map units (between the Eco RI and Hind I11 cleavage sites) with the consequent loss of part of Hpa 11-5 and Hpa 11-2, including the Eco RI cleavage site, and all of Hpa 11-4, -8 and -7 (Lund et al., 1977). Why the sequences from these two regions are retained is not clear. Numerous hypotheses could be advanced. For example, one or both may be involved in DNA replication. Repeated origins of DNA replication would certainly confer a selective advantage to molecules containing them as compared with molecules that have only one origin. On the other hand, the region at 71 map units might be the origin of DNA replication and that at 68 map units important for ensuring encapsidation (i,e., serve as a focus where the viral capsid protein forms on the DNA) and therefore necessary for continued passage of defective virus particles from cell to cell. The physical map of one species of cloned defective DNA (D-50) is shown in Fig. 6. This species is composed of a 3-fold tandem repeat of a basic unit of DNA about 17%the size of the polyoma genome (Griffin and Fried, 1975). The 17% unit appears to be composed of a continuous portion of viral DNA extending from a point in Hpa 11-3 through Hpa

e42

W T M

+hal

-Hhal

Hha,

FIG.6. Physical maps of the polyoma A-2 wild-type DNA and the cloned defective DNAs D-92, D-75, and D-50. The origin of viral DNA replication ( 0 )is indicated on all the maps. D-92 (92% of the genome in length) contains the wild-type Hpa I1 restriction fragments 1, 6, and 3 and parts of fragments 2 and 5, but lacks completely fragments 4, 7, and 8 (from the early region of the genome). It also contains wild-type polyoma DNA sequences that have been rearranged: Sequences from Hpa 11-5 (near the origin of DNA replication) are found joined both to sequences in Hpa 11-2 and -3, as shown. Sequences from Hpa 11-3 (near the origin of DNA replication) in addition to being joined to sequences from Hpa 11-5 (see above) are also joined to sequences from Hpa 11-2 (Lund et al., 1976). D-75 (75%of the genome in length) has none of the Hpa 11 fragments of the wildtype strain although its sequences appear to be almost entirely (if not all) viral in origin. It is essentially made up of a 3-fold repeat of sequences from Hpa 11-3 (67-69 map units and 70-71 map units), from Hpa 11-5 (71-72 map units) and from Hpa 11-2 (from about 0.5-19.5 map units) as shown; v indicates the site of the sequences present in Hpa 11-3 from wild-type DNA but absent in the defective (6970 map units) ( Lund et al., 1976). D-50 (50% of the genome in length) is comprised of a 3-fold tandem repeat of a contiguous portion of the genome which lies between 67 and 84 map units. In addition to the 3-fold repeat of this segment of the ‘genome, molecules are also present in the D-50 DNA preparation, which contain 2-, 4-, 5-, and 6-fold repeats of this same region (Griffin and Fried, 1975).

102


11-5 into Hpa 11-4. The point where the sequences of Hpa 11-3 are joined to Hpa 11-4 lies around 67 map units, a region always maintained in defectives (see above), From the same DNA preparation which contained the 3-fold repeating unit (D-SO), molecules with 2-, 4-, 5-, and &fold tandem repeats of the 17%basic unit were isolated. It is interesting to note that the region containing the termination of DNA replication and the 3’ ends of the stable viral messenger RNAs, which have been mapped in Hpa 11-6 (see Sections IV and V), are not detected in these defective DNAs. It is not known how this type of defective molecule is generated, but it seems probable that the different species that contain tandem repeats of the 17%unit are the result of recombination events. A number of the genome reorientations found in the defective DNAs are also found in the polyoma infectious variants that have additional viral sequences ( see Section IX) , All the polyoma virus-defective species which have been cloned and subsequently studied have come from relatively early-passage virus stocks and have been found to contain only viral sequences. A number of cloned SV40 defective species have also been studied (Mertz and Berg, 1974b; Brockman et al., 1975; Lee et al., 1975). All of them retain the region around the origin of DNA replication at 0.67 map units. In some of these molecules it has been shown that more than one functional origin is present; in most defective species the Hind-A-C junction is reiterated, Fine-structure analyses have not been performed on these molecules to determine whether another closely spaced region is also always present, as is the case for the polyoma defective molecules. In various SV40-defective species, sequences from either Hind-A or Hind-C (around the origin of DNA replication) have been found to be joined to other (noncontiguous) regions of the genome ( Brockman et aZ., 1975; Lee et al., 1975). Some clonal isolates of late-passage SV40 defectives contain host DNA (Lee et al., 1975). In two species of this type, the only viral sequences present came from around the Hind-A-C junction. Analysis of these two species revealed that different sequences of host DNA were present; in one, reiterated host sequences were evident whereas in the other the host DNA was derived from unique sequences. It is not known how defective DNAs arise, but a likely possibility is that they are formed from the oligomeric viral DNA species found in virus producing cells (Cuzin et al., 1970; Martin et al., 1976). XI. Comparison of the Polyoma Virus and SV40 Genomes

In Fig. 7, the physical maps of polyoma virus and SV40 DNAs and the biological markers mapped on these DNAs are shown in such a way that

POLYOMA AND S V a GENOMES

103

.they can be related to each other. It can be seen at a glance that the organization of the information is very similar in both viral genomes. The early and late regions each account for approximately 50%of the genome. The 5’ ends of both the stable early 19-20 S and late 19 S mRNAs map near the origin of DNA replication, whereas the 3’ ends of these RNAs map relatively close to the terminus of DNA replication. With both polyoma virus and SV40 DNAs, the origins of replication appear to be rich in C and G (Mulder and Delius, 1972; Lescure and Yaniv, 1975). The termination of DNA replication, on the other hand, has been located in restriction enzyme fragments that are relatively rich in A and T (Danna and Nathans, 1972; Griffin et al., 1974) and very close to regions of the DNA that are known to be easily denatured (Beard et al., 1973; Monjardino and James, 1975; Lescure and Yaniv, 1975) . The sequences that code for the minor viral proteins (VP-2 and VP-3) of both viruses have been mapped toward the 5’ end of late mRNA, whereas the sequences that code for the major viral capsid protein (VP-1) have been mapped toward the 3’ end of late mRNA. Genetic mapping has shown that with both viral genomes intragenic complementation can take place between mutants with lesions in the sequences coding for VP-1. The early mutants of both polyoma and SV40 (the A group and TS-A type which are required at least for the initiation of transformation) have been found to map in less than 50%of the early region of the genome, in sequences found near the termination of DNA replication and the 3’ ends of the early stable mRNAs. For both viruses, deletions that do not affect viability have been found in the early regions of the DNAs, 8-14% from the origins of DNA replication and near the 5’ ends of the early stable mRNAs. With both viruses, mutations that do not impair biological functions have been detected around the origin of DNA replication either in the early or late regions of the genome and around the termination of replication ( see Section IX). Ferguson and Davis (1975) joined polyoma and SV40 DNAs covalently using the sequential action of Eco RI restriction endonuclease and E . coli DNA ligase. By an electron microscopic analysis, after denaturation and renaturation of this covalently linked hybrid, they were able to detect sequence homology regions between the two DNAs. These were divided into three contiguous regions with varying degrees of homology. The strongest region (about 75% homology) was mapped by Ferguson and Davis to lie between 0.93 and 0.98 map units on the SV40 physical map (in Hind fragments E and K ) and between 47 and 52%away from the polyoma Eco RI cleavage site (in Hpa 11-1) (see Fig. 7). It is interesting to note that this homology region may lie between the sequences that

104


EARLY REGION

LATE REGION T

9"

L

H

A

6

I

G

JI

I

D

FIK[ E

PR C

T

u)

a

>

a

-

EARLY REGION

'

'

20s EARLY mRNA

LATE REGION

P

10s LATE mRNA

=D

0

16s LATE mRNA

P

r0 , .

EARLY

PROTEIN 6)

VP-1

2

I

B

,

5

VP-2/VP-3,

a TSA MUTANTS

c

-

v)

TS,1260,10,C; MP208

1

00

op

p

1O ,

ZC!

map units

39

49

5,O

6.0

10

I1

FIG.7. Polyoma virus (bottom half of figure). The circular Hpa I1 physical map which consists of eight restriction fragments (Griffin et al., 1974), (see Fig. 1 ) has been linearized by breaking the molecule at 70.8 map units ( the Hpa 11-3-5 junction), where the origin of DNA replication ( O R ) has been mapped (Griffin d al., 1974). The termination of DNA replication ( T ) is also indicated. The single Eco RI restriction enzyme cleavage site is shown at zero map unit. The region in Hpa 11-1 indicated by the dotted lines is the portion of the polyoma virus genome which shows the strongest homology with the SV40 genome (Ferguson and Davis, 1975). Below the linear DNA map, the lines from top to bottom represent: ( 1 ) the early and late regions of the DNA (Kamen et al., 1974). ( 2 ) the regions of the DNA within which the viral mRNAs (early 20 S, late 19 S and late 16 S ) have been mapped (Kamen and Shure, 1976). ( 3 ) the regions of the DNA within which the viral capsid proteins (VP-1 and VP-2/VP-3) (Smith et al., 1975) and the early viral protein( s ) have been mapped. ( 4 ) the regions of the genome in which the early (TS-A type) and late (TS-1260, -10, and -C) temperature-sensitive mutants and the minute-plaque morphology mutant (208) have been mapped (Miller and Fried, 1967a). SV40 (top half of figure) : The circular Hind physical map which consists of thirteen fragments (eleven of which are shown) (Danna et d.,1973; Yang et al., 1975)

POLYOMA AND

SV40

GENOMES

105

appear to code for VP-1 and VP-2/VP-3 in both viruses, and may, a t least in part, represent a similar spacer region between the two late viral genes. Defective variants of both polyoma virus and SV40 retain the biologically important region of the viral genome that contains the origin of DNA replication and the region specifying the 5’ ends of stable mRNAs. This region is usually found in multiple copies in defective molecules. The retention of this region by these molecules designates the defectives as possible receptors for the cloning of genes in mammalian cells. This may be especially useful in the case of polyoma virus, which does not grow in human cells, and therefore may not present a biological hazard to man. XII. Conclusion

The organization of the genetic information of the two papovaviruses, polyoma virus and SV40, has been found to be very similar. The early and late regions each comprise about 50%of the genome, and the genes specifying the early and late proteins are distributed in a similar manner in both viral DNAs (Fig. 7 ) . The late region contains the genetic information for the structural capsid proteins and is probably not important either for the initiation or maintenance of transformation. The protein ( s ) specified by the early region of the genome has not yet been definitively identified, but T-antigen is probably coded for (at least in part) by the early region. It is interesting to note that all the early mutants that affect the initiation of transformation have been mapped in a restricted portion of the early region (50%of the early region near the termination of viral DNA replication; see Fig. 7 ) . The early region of the viral genomes is required for the initiation and possibly the maintenance of transformation has been linearized by breaking the molecule at 0.67 map units where the origin of DNA replication ( O R ) has been mapped (Nathans and Danna, 1972b; Danna and Nathans, 1972). The termination of DNA replication ( T ) is also indicated. The single Eco RI restriction enzyme cleavage site is shown at zero map unit. The region of Hind K and E indicated by the dotted lines is the portion of the SV40 genome that shows the strongest homology with the polyoma virus genome (Ferguson and Davis, 1975). Above the linear DNA map, the lines from bottom to top represent: ( 1) the early and late regions of the DNA (Khoury et al., 1 9 7 5 ~ ) ; ( 2 ) the regions of the DNA within which the viral mRNAs (early 19 S , late 19 S , and late 16 S ) have been mapped (May et al., 1975; Khoury et al., 1976); ( 3 ) the regions of the DNA within which the viral capsid proteins (VP-1 and VP2/VP-3) and the early protein( s ) have been mapped (Prives et al., 1974; Smith et al., 1975); ( 4 ) the regions of the genome in which the early ( A mutants) and late (B, BC, C, and D ) temperature-sensitive mutants have been mapped (Lai and Nathans, 1974b, 197515; Mantei et d., 1975).

106

MIKE FRIED A N D BEVERLY GRIFFIN

(see Section IX). How this is facilitated is not known, nor is it known how much of the early region is required, since there appear to be areas of the early region that can be deleted without affecting transformation (see Section IX).The relationship between the function of the early protein( s ) in viral DNA replication and its role in the transformation process is not clear. Recent advances have allowed the viral early protein( s ) to be made in uitro (from either viral mRNA or cRNA) under conditions where other viral and host proteins are not synthesized (see Section VI). These studies in combination with isolation of T antigen from infected and/or transformed cells, should allow for a more rigorous definition of the relationship of the viral early protein(s) and the transformation process. The nature of the integration of polyoma virus or SV40 DNA into host DNA has not yet been elucidated. Restriction enzymes which have been so valuable in determining the physical maps of the viral DNAs (see Section 11) are also being used to study the integration of the viral DNA into the host genome (Botchan and McKenna, 1973; Kettner and Kelly, 1976; Botchan et al., 1976). A number of regions with “regulatory” roles (e.g., origin of DNA replication, termini of mRNAs, regions that are not translated, intergenic regions, etc.) have been localized on the viral genomes. Fine structure analysis and DNA sequencing (see Section 111) may elucidate the relationship between nucleotide sequence and control functions. Since polyoma virus and SV40 are host dependent as well as host specific, the viral regions used for these regulatory functions may well prove to be similar to those utilized by the host cell. Thus by studying these viral genomes which contain a limited amount of genetic information (possibly only three genes), useful knowledge may be gained about the process of control in the more complex (thousands of genes) mammalian cell. The aim of this review has been to correlate the knowledge accumulated about these two viral genomes and to point the areas where knowledge is still needed in order to allow the viruses to be used as fine probes in understanding the host cell and how it can be changed from a “normal” to a “transformed cell. ACKNOWLEDGMENTS We wish to thank our many colleagues who have allowed us to use their scientific results prior to publication. We are also grateful to Drs. L. K. Miller and B. Ozanne for helpful criticisms of the manuscript. Furthermore, we would like to record our deep gratitude to Miss Sara Barlow for patient assistance in the preparation of the manuscript and Mrs. Moira Griffiths for help in preparing the diagrams.

POLYOMA AND

sv40

CENOMES

107

REFERENCES Abrahams, P. J., and van der Eb, A. J. ( 1975). J. Virol. 16, 206-209. Acheson, N. H. ( 1976). Cell 8, 1-12. Allet, B., Roberts, R. J., Gesteland, R. F., and Solem, R. (1974). Nature (London) 249,217-221. Aloni, Y. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 2404-2409. Aloni, Y.,and Locker, H. (1973). Virology 54, 495-505. Aloni, Y., Shani, M., and Reuveni, Y. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 2587-2591. Alwine, J. C., Reed, S. I., Ferguson, J., and Stark, G. R. (1975). Cell 6, 529534. Arber, W. ( 1974). Prog. Nucleic Acid Res. Mol. Biol. 14, 1 3 7 . Beard, P., Morrow, J. F., and Berg, P. (1973).J. Virol. 12, 1303-1313. Benjamin, T. L. (1970). Proc. Natl. Acad. Sci. U.S.A. 67, 394399. Benjamin, T. L., and Burger, M. M. (1970). Proc. Natl. Acad. Sci. U.S.A. 67, 929-934. Black, P. H., Rowe, W. P., Turner, H. C., and Huebner, R. J. ( 1963). Proc. Natl. Acad. Sci. U.S.A. 50,1148-1156. Blackstein, M. E., Stanners, C. P., and Farmilo, A. J. (1969). J. Mol. Biol. 42, 301-313. Blakesley, R. W., and Wells, R. D. ( 1975). Nature (London) 257, 421-422. Botchan, M., and McKenna, G. (1973). Cold Spring Harbor Symp. Quant. Biol. 38, 391-395. Botchan, M., Ozanne, B., and Sambrook, J. (1974). Cold Spring Harbor Symp. Biol. 39, 95-99. Botchan, M., Topp, W., and Sambrook, J. ( 1976). Cell (in press). Bourgaux, P., Bourgaux-Ramoisy, D., and Dulbecco, R. (1969). Proc. Natl. Acad. Sci. U.S.A. 64, 701-708. Boyer, H. W. ( 1971). Annu. Rev. Microbiol. 25, 153-176. Brockman, W. W., and Nathans, D. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 942-946. Brockman, W. W., Lee, T. N. H., and Nathans, D. (1973). Virology 54, 384-397. Brockman, W. W., Lee, T. N. H., and Nathans, D. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 119-127. Brockman, W. W., Gutai, M. W., and Nathans, D. (1975). Virology 66, 36-52, Brugge, J. S., and Butel, J. S. (1975). J. Virol. 15, 619-635. Buetti, E. (1974).J . Virol. 14, 249-260. Cairns, J. ( 1963). Cold Spring Harbor Symp. Quant. Biol. 28, 43-46. Carbon, J., Shenk, T. E., and Berg, P. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 13921396, Carroll, R. B., Hager, L., and Dulbecco, R. (1974). Proc. N d . A d . Sci. U.S.A. 71, 3754-3757. Chen, M. C. Y., Chang, K. S. S., and Salzman, N. P. (1975). J. Virol. 15, 191-198. Chou, J. Y.,and Martin, R. G. (1974). J. Virol. 13, 1101-1109. Chou, J. Y.,Avila, J., and Martin, R. G. (1974). J. Virol. 14, 116-124. Crawford, L. V., and Robbins, A. K. (1976). J. Gen. Virol. 31, 315-321. Crawford, L. V., Syrett, C., and Wilde, A. (1973). J. Gen. Virol. 21, 515-521. Crawford, L. V., Robbins, A. K., and Nicklin, P. M. (1974). J . Gen. Virol. 25, 133-142. Cuzin, F., Vogt, M., Dieckmann, M., and Berg, P. (1970). J. Mol. Biol. 47, 317-333.

108


Dahlberg, J. E., and Blather, F. R. ( 1973). Fed. Proc., Fed. Soc. Erp. Biol. 32, 664. Danna, K. J., and Nathans, D. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 30973100. Danna, K. J., Sack, G. H., and Nathans, D. ( 1973). J. Mol. Biol. 78,363376. Davis, R. W., Simon, M., and Davidson, N. ( 1971). In “Methods in Enzymology,” Vol. 21: Nucleic Acids, Part D ( L. Grossman and K. Moldave, eds. ), pp. 413-428. Academic Press, New York. Delius, H., Mantell, N. J., and Alberts, B. (1972).J. Mol. Biol. 67, 341350. Dhar, R., Subramanian, K., Zain, B. S., Pan, J., and Weissman, S. M. (1974a). Cold Spring Hurbor Symp. Quant. Biol. 39, 153-160. Dhar, R., Weissman, S. M., Zain, B. S., Pan, J., and Lewis, A. M. (1974b). Nucleic Acids Res. 1, 595-613. Dhar, R., Zain, S., Weissman, S. M., Pan, J., and Subramanian, K. ( 1 9 7 4 ~ )Proc. . Natl. Acad. Sci. U.S.A. 71, 371375. Dhar, R., Subramanian, K. N., Zain, B. S., Levine, A., Patch, C., and Weissman, S. M. ( 1975). Collog. Inst. Natl. Sante. Rech. Med. 47, 2 5 3 2 . Di Mayorca, G., Callender, J., Martin, G., and Giordano, R. (1969). Virology 38, 126-133. Dubbs, D. R., Rachmeler, M., and Kits, S. (1974). Virology 57, 161-174. Dulbecco, R., and Eckhart, W. (1970). Proc. Natl. Aoad. Sci. U.S.A. 67, 1775-1781. Eckhart, W. (1969). Virology 38, 120-125. Eckhart, W. ( 1974). Cold Spring Harbor Symp. Quant. Biol. 39,3740. Eckhart, W., and Dulbecco, R. (1974). Virology 60, 359360. Eckhart, W., Dulbecco, R., and Burger, M. M. (1971). Proc. Natl. Acad. Sct. U.S.A. 68, 283-286. Eddy, B. E., Borman, G. S., Berkeley, W. H., and Young, R. D. (1961). Proc. Soc. Exp. Biol. Med. 107, 191-197. Eddy, B. E., Borman, G. S., Grubbs, G. E., and Young, R. D. (1962). Virology 17, 65-75. Edgell, M. H., Hutchison, C. A., 111, and Sclair, M. (1972). 1. Virol. 9, 574-582. Fareed, G. C., Garon, C. F., and Salzman, N. P. (1972). 1. Virol. 10, 484-491. Ferguson, J., and Davis, R. W. (1975). J. Mol. Biol. 94, 135-150. Fiers, W., Danna, K., Rogiers, R., Van de Voorde, A., Van Herreweghe, J., Van Heuverswyn, H., Volckaert, G., and Yang, R. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 179-186. Fiers, W., Rogiers, R., Soeda, E., Van de Voorde, A., Van Heuverswyn, H., Van Herreweghe, J., Volckaert, G., and Yang, R. (1975). Fed. Eur. Biochem. Soc. Meet. [Proc.] 39, 17-33. Folk, W. R.,Fishel, B. R., and Anderson, D. M. (1975). Virology 64, 277-280. Francke, B., and Eckhart, W. (1973). Virology 55, 127-135. Francke, B., and Vogt, M. (1975). Cea 5,205-211. Fried, M. ( 1965a). Virology 25, 669-671. Fried, M. (1965b). Proc. Natl. Acad. Sci. U.S.A.53,486-491. Fried, M. (1970). Virology 40,605-617. Fried, M. (1974). J. Vtrol. 13, 939-946. Fried, M., and Griffin, B. E. ( 1977). Manuscript in preparation. Fried, M., Griffi, B. E., Lund, E., and Robberson, D. L. (1974). Cold Spring Hurbor Symp. Quant. Biol. 39,45-52. Gadin, D. E., and Goodman, H. M. (1974). Biochem. Biophys. Res. Commun. 59, 108-116.

POLYOMA AND

sv40

CENOMES

109

Garfin, D. E., Boyer, H. W., and Goodman, H. M. (1975). Nucleic Acids Res. 2, 1851-1865. Germond, J.-E., Vogt, V., and Hirt, B. ( 1974). Eur. J. Biochem. 43,591-600. Gibson, W. ( 1974). Virology 62, 319-336. Gibson, W. ( 1975). Virology 68, 539-543. Goldman, E., and Benjamin, T. L. ( 1975). Virology 66, 372484. Gotlieb-Stematsky, T., and Leventon, S. ( 1960). Br. J . Exp. Pathol. 41,507-519. Graessmann, M., and Graessmann, A. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 366-370. Graessmann, M., Graessmann, A., Niebel, J., Koch, H., Fogel, M., and Mueller, C. (1975). Nature (London) 258,756-758. Greene, P. J., Betback, M. C., Boyer, H. W., and Goodman, H. M. (1974). Methods Mol. Biol. 7, 87-111. Griffin, B. E., and Fried, M. ( 1975). Ndure (London) 256, 175-179. Griffin, B. E., and Fried, M. (1976). Methods Cancer Res. 12,49-86. Griffin, B. E., Fried, M., and Cowie, A. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 2077-2081. Griffith, J., Dieckmann, M., and Berg, P. (1975). J. Virol. 15, 167-172. Gromkova, R., and Goodgal, S. H. (1972). J. Bacteriol 109, 987-992. Gromkova, R., Bender, J., and Goodgal, S. (1973). J. Bacteriol. 114, 1151-1157. Gross, L. (1953a). Proc. SOC. Exp. Biol. Med. 83,414-421. Gross, L. ( 1953b). Cancer 6, 948-957. Habel, K. (1961). Proc. Soc. Exp. B i d . Med. 106,722-725. Habel, K., and Eddy, B. E. (1963). Proc. Soc. Erp. Biol. Med. 113,1-3. Hare, J. D., and Morgan, H. R. ( 1962). Virology 19, 105-107. Hedgpeth, J., Goodman, H. M., and Boyer, H. W. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 3448-3452. Hewick, R. M., Fried, M., and Waterfield, M. D. (1975). Virology 66, 408-419. Hewick, R. M., Waterfield, M. D., Miller, L. K., and Fried, M. (1977). Manuscript submitted for publication. Hirt, B. (1969). J. Mol. Biol. 40, 141-144. Horiuchi, K., and Zinder, N. D. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 2555-2558. Howley, P. M., Mullarkey, M. F., Takemoto, K. K., and Martin, M. A. (1975a). J . Virol. 15, 173-181. Howley, P. M., Khoury, G., Byrne, J. C.,Takemoto, K. K., and Martin, M. A. (1975b). J . Virol. 16, 959-973. Huang, E. S., Estes, M. K., and Pagano, J. S. (1972). J . Virol. 9, 923-929. Huang, E-S., Newbold, J. E., and Pagano:J. S. ( 1973). J. Virol. 11,508514. Irlin, I. S. ( 1967). Virology 32, 725-728. Ito, J., Kawamura, F., and Duffy, J. J. (1975).FEBS Lett. 55,278-281. Jay, E., Roychoudhury, R., and Wu, R. (1976). Biochem. Biophys. Res. Commun. 69, 678-686. Jessel, D., Hudson, J., Landau, T., Tenen, D., and Livingston, D. M. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 1960-1964. Kamen, R., and Shure, H. (1976). CeU 7, 361-373. Kamen, R., Lindstrom, D. M., Shure, H., and Old, R. W. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 187-198. Kelly, T. S., and Smith, H. 0. (1970). J. Mol. Biol. 51, 393409. Kettner, G., and Kelly, T. J. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 1102-1106.

110


Khera, K. S., Ashkenazi, A., Rapp, F., and Mellnick, J. L. (1963). J . Imrnunol. 91, 604-613. Khoury, G., Byme, J. C., and Martin, M. A. (1972). Proc. Natl. Acad. Sci. U S A . 6D, 1925-1928. Khoury, G., Byrne, J. C., Takemoto, K. K., and Martin, M. A. (1973). J . Virol. 11, 54-60.

Khoury, G., Fareed, G. C., Berry, K., Martin, M. A., Lee, T. N. H., and Nathans, D. (1974). J . M o l . Biol. 87, 289-301. Khoury, G., Martin, M. A., Lee, T. N. H., and Nathans, D. (1975a). Virology 63, 263-272. Khoury, G., Howley, P. M., Garon, C., Mullarkey, M. F., Takemoto, K. K., and Martin, M. A. (1975b). Proc. Natl. Acad. Sci. U.S.A. 72,2563-2567. Khoury, G., Howley, P., Nathans, D., and Martin, D. ( 1 9 7 5 ~ )J.. Virol. 15, 433437. Khoury, G., Carter, B. J., Ferdinand, F.-J., Howley, P. M., Brown, M., and Martin, M. A. ( 1976). J . Virol. 17, 832-840. (in press). Kimura, G., and Dulbecco, R. ( 1792 ). Virology 49,394-403. Kimura, G., and Dulbecco, R. (1973). Virology 52, 529534. Kimura, G., and Itagaki, A. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 673-677. Kleid, D., Humayun, Z., Jeffrey, A,, and Ptashne, M. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 293-297. Koch, M. A., and Sabin, A. B. (1963). Proc. SOC. Exp. Biol. M e d . 113,4-12. Kuchino, T., and Yamaguchi, N. (1975). J . Virol. 15,1302-1307. Lai, C.-J., and Nathans, D. (1974a). J . Mol. Blol. 89, 179-193. Lai, C.-J., and Nathans, D. (197413). Virology 60, 466-475. Lai, C.-J,, and Nathans, D. ( 1 9 7 4 ~ ) Cold . Spring Harbor Syrnp. Quant. Biol. 39, 53-60. Lai, C.-J., and Nathans, D. (1975a). J. M o l . Biol. 97, 113-118. Lai, C.-J,, and Nathans, D. (197513). Virology 66, 70-81. Laub, O., and Aloni, Y. (1975). J . Virol. 16, 1171-1183. Lavi, S., and Shatkin, A. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 2012-2016. Lavi, S., and Winocour, E. (1972). J. Virol. 9, 309-316. Lavi, S., and Winocour, E. (1974). Virology 57, 296-299. Lazarides, E., Files, J. G., and Weber, K. ( 1974). Virology 60, 584-587. Lebowitz, P., Weissman, S . M., Radding, C. M. (1971). J. Biol. C h e m . 246, 51205139. Lebowitz, P., Siege], W., and Sklar, J. (1974). J. Mol. Biol. 88, 105-123. Lee, T. N. H., Brockman, W. W., and Nathans, D. (1975). Virology 66, 53-69. Lescure, B., and Yaniv, M. (1975). J . Virol. 16, 720-724. Levine, A. J. (1974). Prog. M e d . Virol. 17, 1-37. Levine, A. J., Kang, H. S., and Billheimer, F. E. (1970). J . M o l . Biol. 50, 549568. Lewis, A. M., and Rowe, W. P. (1971). J. Virol. 7, 189-197. Ling, V. (1972). 1. M o l . Biol. 64, ,87-102. Louie, A. J. (1974). Cold Spring Harbor Symp. Qzrant. Biol. 39, 259-266. Lund, E., Fried, M., and Griffin, B. E. (1977). In preparation. McCann, P. P., and Martin, R. G. (1975). Biochim. Biophys. Acta 395, 280-283. Malmgren, R. A., Takemoto, K. K., and Carney, P. G. (1968). J. NatE. Cancer Inst. 40, 263-268. Mantei, N., Boyer, H. W., and Goodman, H. M. (1975). J . Virol. 16, 754-757. Marbaix, G., Huez, G., Burny, A., Cleuter, Y., Hubert, E., Leclercq, M., Chan-

POLYOMA AND

sv40

GENOMES

111

trenne, H., Soreq, H., Nudel, V., and Littauer, U. Z. (1975). Proc. Natl. Acud. Sci. U.S.A. 72. 3065-3067. Martin, M. A., Howley, P. M., Byrne, J. C., and Garon, C. F. (1976). Virology 71, 28-40. Martin, R. G., and Chou, J. Y. (1975).J. Virol. 15, 599-612. Martin, R. G., Chou, J. Y., Avila, J., and Saral, R. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 17-24. May, E., Kopecka, H., and May, P. (1975). Nucleic Acids Res. 2, 1995-2005. Mertz, J. E., and Berg, P. (1974a). Proc. Natl. Acad. Sci. U.S.A. 71, 4879-4883. Mertz, J. E., and Berg, P. ( 1974b). Virology 62, 112-124. Mertz, J. E., Carbon, J., Herzberg, M., Davis, R. W., and Berg, P. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 69-84. Meselson, M., Yuan, R., and Heywood, J. (1972). Annu. Reu. Biochem. 41, 447-466. Middleton, J. H., Edgell, M. H., and Hutchison, C. A., 111. (1972).J. Virol. 10, 42-50. Miller, L.K., and Fried, M. (1976a). J. Virol. 18, 824-832. Miller, L. K., and Fried, M. ( 1976b). Nature (London) 259,598-601. Miller, L. K., Cooke, B. E., and Fried, M. (1976). Proc. Natl. Acad. Sci. U.S.A. in press). Monjardino, J., and James, A. W. (1975). Nature (London) 255, 249-251. Morrow, J. F., and Berg, P. (1972). PTOC.Natl. Acud. Sci. U.S.A. 69, 3365-3369. Morrow, J. F., and Berg, P. (1973). J. Virol. 12, 1631-1632. Mulder, C,, and Delius, H. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 32154219. Murray, K., and Old, R. W. (1974). Prog. Nucleic Acid Res. Mol. Biol. 14, 117-185. Murray, K., Murray, N. E., and Brammar, W. J. (1975). Fed. Eur. Biochem. SOC. Meet. [Proc.] 39, 193-207. Nathans, D., and Danna, K. J. (1972a). J. Mol. Biol. 64, 515518. Nathans, D., and Danna, K. J. (1972b). Nature (London),New Biol. 236, 200-202. Nathans, D., and Smith, H. 0. (1975).Annu. Reu. Biochem. 44,273-293. Nathans, D., Adler, S. P., Brockman, W. W., Danna, K. J., Lee, T. N. H., and Sack, G. H. (1973). In "Virus Research" (C. F. Fox and W. S. Robinson, eds.), pp. 61-69. Academic Press, New York. Old, R., Murray, K., and Roizes, G. ( 1975). J. Mol. Biol. 92, 331-339. Osborn, M., and Weber, K. (1975).J. Virol. 15,636444. Oxman, M. N., Takemoto, K. K., and Eckhart, W. (1972). Virology 49, 675-682. Ozanne, B., Sharp, P. A., and Sambrook, J. (1973).J . Virol. 12,90-98. Paulin, D., and Cuzin, F. (1975). J. Virol. 15,393497. Pieczenik, G., Model, P., and Robertson, H. D. (1974). J . Mol. Biol. 90, 191-214. Polisky, B., and McCarthy, B. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 2895-2899. Prives, C. L., Aviv, H., Gilboa, E., Revel, M., and Winocour, E. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 309-316. Reed, S. I., Ferguson, J., Davis, R. W., and Stark, G. R. (1975). Proc. Nutl. Acad. Sci. U.S.A. 72, 1605-1609. Robb, J. A. (1973). J . Virol. 12, 1187-1190. Robb, J. A., and Martin, R. G. (1972).1. Virol. 9, 956-968. Robb, J. A., Smith, H. S., and Scher, C. D. (1972).J. Virol. 9,969-972. Robberson, D. L., and Fried, M. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 34974501. Robberson, D. L., Crawford, L. V., Syrett, C., and James, A. W. (1975). J. Gen. Virol. 26, 49-69.

112


Roberts, R. J., Breitmeyer, J., Tabachnik, N., and Myers, P. (1975). J . Mol. Biol. 91, 121-123. Roberts, R. J., Myers, P. A., Morrison, A,, and Murray, K. (1976). J. Mol. Biol. 103, 199-208. Rosenberg, M., de Chrombrugghe, B., and Musso, R. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 717-721. Rutherford, R. B., and Hare, J. D. (1974). Biochem. Biophys. Res. Commrrn. 58, 839-846. Sachs, L., and Medina, D. (1960). Nature (London) 187,715-716. Sack, G . H., Narayan, O., Danna, K. J., Weiner, L. P., and Nathans, D. ( 1973). Virology 51, 345350. Salzman, N. P., and Khoury, G. ( 1974). In “Comprehensive Virology” ( H . FraenkelConrat and R. R. Wagner eds. ), pp. 63-141. Plenum, New York. Sambrook, J. (1972). Adu. Cancer Res. 16, 141-180. Sambrook, J. F. (1975). In “Control in Virus Multiplication” (D. C. Burke and W. C. Russell, eds.), pp. 153-181. Cambridge Univ. Press, London and New York. Sambrook, J., Westphal, H., Srinivasan, P. R., and Dulbecco, R. (1968). Proc. Natl. Acad. Sci. U S A . 60, 1288-1295. Sambrook, J., Sharp, P. A,, and Keller, W. (1972). J . Mol. Biol. 70, 57-71. Sambrook, J., Sugden, B., Keller, W., and Sharp, P. A. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 37114715. Sebring, E. D., Kelly, T. J., Thoren, M. M.. and Salzman, N. P. (1971). J . Virol. 8, 478490. Seehafer, J,, Salmi, A,, Scraba, D. G., and Cotter, J. S. (1975). Virology 66, 192-205. Sharp, P. A,, Sugden, B., and Sambrook, J. (1973). Biochemistry 12, 3055-3063. Shenk, T. E., Rhodes, C., Rigby, P. W., and Berg, P. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 989-993. Shenk, T. E., Carbon, J., and Berg, P. (1976). J. Virol. 18, 664-671. Shishido, K., and Berg, P. ( 1976). J . Virol. 18, 793-798. Sjogren, H. O., Hellstrom, I., and Klein, G. (1961). Cancer Res. 21, 329-337. Sklar, J., Yot, P., Weissman, S. M. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 18171821. Smith, A. E., Bayley, S. T., Mangel, W. F., Shure, H., Wheeler, T., and Kamen, R. I. (1975). Fed. Eur. Biochem. SOC. Meet. [Proc.] 39, 151-164. Smith, D. I., Blattner, F. R., and Davies, J. (1976). Nucleic Acids Res. 3, 343-352. Smith, H. 0. (1974). In “DNA Replication, Methods in Molecular Biology” (R. B. Wickner, ed.), Vol. 7, pp. 71-85. Dekker, New York. Smith, H. O., and Wilcox, K. W. (1970). J . Mol. Biol. 51, 379-391. Squires, C., Lee, F., Bertrand, K., Squires, C. L., Bronson, M. J., and Yanovsky, C. (1976). J . Mol. Bid. 103, 351381. Stewart, S. E., Eddy, B. E., and Borgese, N. G. (1958). 1. Natl. Cancer Inst. 20, 1223-1236. Subramanian, K. N., Pan, J., Zain, S., and Weissman, S. M. (1974). Nucleic Acids Res. 1, 727-752. Subramanian, K. N., Dhar, R., Pan, J., Zain, B. S., and Weissman, S . M., (1976). ICN-UCLA Symp. Mol. Cell Blol. 5, in press. Summers, J. (1975). J. Virol. 15, 946-953. Sweet, B. H., and Hilleman, H. R. (1960). Proc. Soc. Exp. Biol. Med. 105, 420-427. Takanami, M. ( 1974). In “DNA Replication, Methods in Molecular Biology” (R. B. Wickner, ed. ), Vol. 7, pp. 113-133. Dekker, New York.

POLYOMA AND

sv40

GENOMES

113

Takemoto, K. K., and Habel, K. (1965). Proc. SOC. Exp. Biol. Med. 120, 124-127. Takemoto, K. K., and Mullarkey, M. F. (1973). J. Virol. 12, 625-631. Tegtmeyer, P. (1972). J. Virol. 10, 591-598. Tegtmeyer, P. (1975). J. Virol. 15, 613-618. Tegtmeyer, P., and Ozer, H. L. (1971).J. Virol. 8, 516524. Tegtmeyer, P., Schwartz, M., Collins, J. K., and Rundell, K. (1975). J. Virol. 16, 168-178. Tenen, D. G., Baygell, P., and Livingston, D. M. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 43514355. Tevethia, S., Katz, M., and Rapp, F. (1965). Proc. SOC. Exp. Biol. Med. 119, 896-901. Thorne, H. V . (1968). J. Mol. Biol. 35, 215-226. Thorne, H. V., Evans, J., and Warden, D. (1968). Nature (London) 219, 728-730. Tonegawa, S., Walter, G., Bernardini, A,, and Dulbecco R. (1970). Cold Spring Harbor Symp. Quant. Biol. 35,823-831. Tooze, J., ed. (1973). “The Molecular Biology of Tumour Viruses.” Cold Spring Harbor Lab., Cold Spring Harbor, New York. Tooze, J., and Sambrook, J., eds. (1974). “Selected Papers in Tumour Virology.” Cold Spring Harbor Lab., Cold Spring Harbor, New York. Van de Voorde, A., Rogiers, R., Van Herreweghe, J., Van Heuverswyn, H., Volckaert, G., and Fiers, W. (1974). Nucleic Acids Res. 1, 1059-1067. Vogt, M., Batcheler, L. T., and Boice, L. (1976). J. Virol. 17, 1009-1026. Weil, R., Salomon, C., May, E., and May, P. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 381-395. Weinberg, R. A., Ben-Ishai, Z., and Newbold, J. E. ( 1972a). Nature (London),New Biol. 238, 111-113. Weinberg, R. A., Warnaar, S. O., and Winocour, E. (1972b). J. Virol. 10, 193-201. Weinberg, R. A., Ben-Ishai, Z., and Newbold, J. E. (1974). J. Virol. 13, 1263-1273. Weisbeek, P. J., and Van De Pol, J. H. ( 1970). Biochim. Biophys. Acta 224, 328438. Weisbeek, P. J., Van De Pol, J. H., and Van Arkel, G. A. (1973). Virology 52, 408416. Wilson, G. A., and Young, F. E. (1975). J. Mol. Biol. 97, 123-125. Yamaguchi, N., and Kuchino, T. (1975). J. Virol. 15, 1297-1301. Yang, R., Danna, K., Van de Voorde, A., and Fiers, W. (1975). Virology 68,260-265. Yang, R. C.-A., Van de Voorde, A., and Fiers, W. (1976a). Eur. J. Virol. 61, 101-117. Yang, R. C.-A., Van de Voorde, A., and Fiers, W. (1976b). Eur. J. Virol. 61, 119-138. Yaniv, M., Chestier, A., Dauguet, C., and Croissant, 0. (1975). FEES Lett. 57, 126-129. Yoshiike, K. (1968a). Virology 34, 391-401. Yoshiike, K. (1968b). Virology 34, 402-409. Yoshimori, R. N. ( 1971). Ph.D. Thesis, University of California, Berkeley. Zain, B. S . , Weissman, S. M., Dhar, R., and Pan, J. (1974). Nucleic Acids Res. 1, 577593.


PrMICROGLOBULIN AND THE MAJOR HISTOCOMPAT1 BI L ITY COMPLEX Per A. Peterson, Lars Rask, and Lars Ostberg Institute of Medical and Physiological Chemistry, Biomedical Center, University of Upprala, Upprala, Sweden

I. Introduction . . . . . . . . . . . . . 11. Genetics of the Major Histocompatibility Complex . . . 111. Traits Associated with the Major Histocompatibility Complex IV. Isolation and Characteristics of p?-Microglobulin . . . V. Features of the Classical Transplantation Antigens . . . VI. Biochemical Properties of the Thymus-Leukemia Antigens VII. T-Locus Gene Products and p.-Microglobulin . . . . VIII. I-Region Defined Antigens and the Fc Receptor . . . IX. The S-Region and the Complement System . . . . . X. Conclusions and Speculations . . . . . . . . . References . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . . .

115 117 120 127 132 142 144 146 152

155 157

I. Introduction

Recently it has become apparent that the plasma membrane is much more than the rigid structure protecting the interior of the cell, which once was thought to be its only function. In fact, the plasma membrane probably fulfills a greater number of different tasks than any other cell organelle. Not only does it protect the integrity of the cell and allow the passage of specific nutrients, but it also receives and transduces numerous signals from other cells, directly or via mediators. These various functions require the cell membrane to contain a complex array of structurally distinct elements. It is likely that the majority of functions which require discriminatory biological specificity involve the participation of the membrane proteins. According to a generally accepted model, the plasma membrane is a fluid bilayer, in which the proteins, partly or completely surrounded by a lipid matrix, are allowed free lateral diffusion (Singer and Nicholson, 1972). Some of the proteins, which traverse the entire membrane, may be connected to the cytoskeleton of microfilaments and microtubules, thereby conferring structural links between the plasma membrane and other cell constituents. Other proteins may be more loosely attached to the outer surface of the membrane. The latter group of molecules comprises a number of receptors that transmit their signals via a number of different routes such as the activation of adenyl and 115

116

PER A. PETERSON ET AL.

guanyl cyclases, changes in the ion flux over the membrane, etc. Such receptors may normally be anchored to the cell surface through interactions with other macromolecules in or under the plasma membrane. Much work has been devoted to the study of cell membrane-bound receptors of various types. Exciting findings on the interaction of polypeptide hormones with their specific membrane receptors have opened up new avenues for further research (for review, see Cuatrecasas, 1974). The recognition of lymphocyte-bound immunoglobulin as the antigen receptor on cells precommitted to antibody synthesis promises to lead to the elucidation of fundamental questions concerning the triggering and maturation of the lymphocyte lineages ( see Coutinho and Moller, 1975). The plasma membrane also contains receptors for plasma proteins that transport trace nutrients like metals and some vitamins (see Putnam, 1975). In fact, several other plasma proteins may be metabolized following the binding of aged proteins to specific liver cell membrane receptors (Ashwell and Morell, 1974). A number of growth-promoting factors seem to convey their message via binding to specific receptors in the membrane (Cohen et al., 1974). Although these receptors are plasma membrane bound, not all cell types express them. This suggests that the various receptors are not obligate for the integrity of the membrane. However, a number of structural membrane proteins most probably also exist whose prime function may be to provide the framework for integration of various specific receptors. One would anticipate that such structural protein would be present in almost all cell membranes and that they would be “biologically inert” as compared to various receptors. The major histocompatibility antigens represent an intermediary group of cell surface glycoproteins. They are present on almost all cells, like most structural proteins, yet they participate in some highly specific biological reactions. Following grafting of nonsyngeneic tissue a vigorous rejection of the transplant occurs. This reaction involves the immune system at several levels, but the prime target molecules for the aggressive killer lymphocytes are the foreign major histocompatibility antigens of the graft. Although the true biological significance of this immunological reaction is poorly understood, much work has focused upon various aspects of these events for the obvious reason of their important consequences for organ transplantation and cancer treatment. As for hormone receptors, the progress in elucidating the structure of the histocompatibility antigens has met with difficulty mainly owing to technical obstacles. However, over the past few years these molecules have been isolated and characterized in some detail. It is the aim of this review to present some of the biochemical characteristics of the major histocompatibility antigens and of related proteins. The presentation will focus upon recent informa-

THE MAJOR HISTOCOMPATIBILITY COMPLEX

117

tion, since a number of excellent reviews on the biochemistry of histocompatibility antigens are already available ( Nathenson, 1970; Kahan and Reisfeld, 1973; Reisfeld and Kahan, 1970; Nathenson and Cullen, 1974; Vitetta and Uhr, 1975a; Moller, 1974). The histocompatibility antigens have been studied in a number of species. Genetic and biological knowledge is for obvious reasons far greater for the murine system than for any other system. Thus, this review will mainly deal with data pertinent to the mouse major histocompatibility complex. It is obvious that in the end information obtained with genetic, biological, and chemical techniques has to coalesce into a coherent picture. Such a full integration of knowledge from various disciplines may not be obtained in the immediate future. Accordingly, for the purpose of this presentation the intricate genetic and biological implications of the major histocompatibility antigen system will be touched upon only as a background insofar as they are directly related to our present knowledge of the structure of the antigens. II. Genetics of the Major Histocompatibility Complex

The discovery of the first trait related to transplantation is owed to Gorer (1937). By using serological techniques he defined a mouse blood group antigen, to be called H-2 ( Gorer, 1936), which was later recognized as being one of the gene products of the major histocompatibility complex. About a decade later Gorer et al. (1948) demonstrated that this trait (H-2) mapped in the ninth linkage group, which was subsequently identified with chromosome 17 (Miller et al., 1971). Soon it was realized that the H-2 system was very complex; it contained more than a single gene and exhibited a great number of alleles. A major step forward in studies of this system was taken by Snell (1958), who introduced the use of congenic mouse strains. These strains all have a constant genetic background on which various major histocompatibility complex haplotypes have been introduced from other mice. Recombinants of these congenic lines provided researchers with a tool to delineate the genetic fine structure of the major histocompatibility complex. This has led to the important conclusion that the genes for the classical transplantation antigens are linked to other kinds of traits. Most likely we have as yet witnessed only the beginning, so it should be realized that most of these newly discovered traits are classes that encompass multiple variants. The presently defined regions of the H-2 complex and some adjacent loci pertinent to the present discussion are depicted in Fig. 1. According to convention the chromosome region containing H-2K and H-2D, and the segment in between, is termed the H-2 complex. Information ob-

118

PER A. PETERSON ET AL. H-2 C 0

----

K

l

S

O

Tla

FIG. I . Loci present on the 17th mouse chromosome related to the major histocompatibility complex. C denotes the centromere.

tained from a number of informative crossovers has established four main regions of the H-2 complex, called K, I, S, and D. The recombin at'ion frequency between the K and D ends is about 0.5%.The TL and T regions are not included in the H-2 complex proper, but since they have an important relationship to the H-2 region they will be described in this context. Further subdivisions of this genetic segment is already possible from newly derived recombinants, and it is to be expected that the complexity will increase further in the near future. Although a sizeable number of recombinants have been analyzed and multiple crossovers have been found demarcating each one of the main regions, it is obvious that the boundaries between the regions are not sharply defined, since the traits analyzed as yet may occupy but a limited portion of each region. In addition, analyses depend on the existence of alternate genes, alleles, at defined loci. The success of the genetic mapping experiments, in spite of this limitation, gives a good idea of the tremendous polymorphism of the system. The serological definition of the H-2 alloantigens, after a period of confusion, has now emerged into a state of relative conformity. Early studies indicated the presence of several H-2 specificities shared among different alleles as well as those specific for a certain allele. This caused great problems in analyzing recombinants until it was realized that the specificities could be arranged into two groups. The unique specificities (prioate) were arranged into two allelic series ( Siiell et al., 1971a,b) Each H-2 haplotype (the particular set of alleles at all loci within the H-2 complex) could thereby be defined by two private serological specificities, which belonged to the K and the D regions, respectively. The broadly distributed specificities ( public antigens) could either represent similar antigenic determinants present on molecules distinguished by their private specificities or unique antigenic determinants on molecules of other genes located at different positions in the complex (Shreffler, 1970; Shreffler et al., 1971). The idea that K- and D-region molecules may share antigenic specificities led to the suggestion that genes in these regions may have evolved from a common ancestral gene (Shreffler et al., 1971). Most data on the serology of H-2 alloantigens is consistent with this duplication model ( Snell et al., 1973). Furthermore, biochemical studies and skin transplantations also corroborate important features of such a model. Although these observations are evidence for two major H-2


119

regions, K and D, which encompass structural genes for the classical transplantation antigens, they by no means rule out as yet the possibility that there are other loci in the same region which control some of the H-2 alloantigen reactivities. In fact, an H-2 G locus has been suggested to exist in the location between the S and D regions (Snell et al., 1973). Some of the so-called public H-2 specificities are probably unique antigenic determinants present on molecules of genes located in the I region (see Sachs and Cone, 1973). This was not realized until recently when several laboratories succeeded in raising antibodies against I-regiondefined antigens (see Shreffler and David, 1975). With use of I-region congenic mice already about 20 Ia-antigen specificities have been defined by serological techniques. This group of specificities will probably be subdivided into private and public specificities, as for H-2 alloantigens. It is apparent that the Ia-antigen specificities have allowed a further resolution of the genetic fine structure of the I region. Since the I region initially was defined by its control of several traits like regulation of the immune response, skin graft rejections, stimulation in the mixed leukocyte reaction, etc., it can now be anticipated that some traits, which have not been separated by crossovers, may be analyzed with use of serological reagents as to whether they are controlled by one and the same gene or by several genes. In view of the complexity of the I-region, it is conceivable that further genetic mapping studies correlated with serological analyses will not only increase the resolution but unravel new traits as well. The S region is defined by its control of two serum protein traits (Shreffler and Passmore, 1971). It is likely that the S region contains two or more genes. There are no reports demonstrating the existence of histocompatibility antigens reIated to the S region. However, the S-region traits introduced a marker of great value in early analyses of crossover locations in recombinant mice. The TL region encompasses the T2a locus which controls a series of alloantigens normally expressed only on thymocytes (Boyse et al., 1966; Boyse and Old, 1969). So far three alleles determining the presence or the absence of a particular set of antigenic specificities have been discovered. The location of the Tla locus is about 1centimorgan outside the D- end of the H-2 complex. Its close physical association with the H-2 complex, as well as its control of traits similar to those of the K and D regions makes the TL region important to consider in any discussion concerning the structure of H-2 region coded macromolecules. Toward the centromere on the 17th chromosome, another complex locus is located. This locus, T , contains a dominant gene T and a number of lethal or semilethal recessive alleles (t-alleles). The gene T interferes

120


with the development of the notochord at an early stage of development. The recessive alleles, almost all of which have been isolated from wild mice, produce morphologically normal heterozygotes but interact with T to confer taillessness to the progeny. Matings between animals carrying different t-alleles yield phenotypically normal progeny. This “complementation” suggests that the “alleles” do indeed belong to separate subloci of a complex locus. The genetics of this system is as yet not fully worked out (see Klein, 1974), but there is enough knowledge to suggest that the T-locus as a whole is at least as complex as the H-2 region. 111. Traits Associated with the Maior Histocompatibility Complex

An ever increasing number of traits have been shown to be associated with the major histocompatibility complex (summarized in Table I ) . Whether all these traits are controlled by separate genes is as yet unknown, but sufficient knowledge is available to suggest that the H-2 region must contain a minimum of about 10 separate genes. Shreffler and Klein (1970) have calculated from the recombination frequency in the H-2 complex that enough DNA is present to comprise about 500 cistrons. This calculation is based on the assumption that the number of recombinants observed are proportional to the physical length of the H-2 chromosomal segment. In view of the suppressed recombination frequency assoTABLE I TRarrs CONTROLLED B Y THE MAJORHISTOCOMPATIBILITY COMPLEX (H-2) OF THE MOUSE ~

~

~

~

Controlling sublocus Trait Serologically detected cell surface antigens Classical transplantation antigens Cell-mediated lympholysis target antigens Immune-response regulation Mixed-leukocyte reaction Graft-versus-host reaction T-B lymphocyte interactions Tumor virus susceptibility I-region associated (Ia) antigens Serum substance (Ss-protein) Sex-limited protein (Slp) Complement level Thymus-leukemia antigens (TLa)

K

I

S

D

(Tla)

THE MAJOH HISTOCOMPATIBILITY COMPLEX

121

ciated with some alleles of the T locus (Dunn, 1964) and its extension in wild mice into the H-2 region (Hammerberg and Klein, 1975), it is doubtful whether such calculations even approximate the truth. These calculations suggest, however, that there should be space enough for a great number of genes within the H-2 complex. The K and D regions contain at least one gene each. This has been convincingly demonstrated with the use of serological techniques. Biochemical work has nicely corroborated these findings ( Nathenson, 1970). The strong histocompatibility rejection associated with the H-2 complex has always been ascribed to the serologically detected antigens derived from the K and D regions. Later studies have extended this view, and it now appears to be established that also the I (Klein et al., 1974) and TL (Boyse et al., 1972) regions contribute to the graft rejection. When this complication was realized, several attempts were made to demonstrate unequivocally that graft rejection is indeed associated with H-2K and D antigens. Although as yet conclusive proof is largely lacking, circumstantial evidence supports the notion. The situation appears less complicated with regard to the cell-mediated lympholysis ( CML ) target antigens (see below), With few exceptions most experiments consistently show that there has to be a difference in the K- and/or D-region antigens between the stimulator and the responder cells to educate cells active in a lympholysis assay (for discussion, see Shreffler and David, 1975). Three traits are associated with the S region. The first one, discovered by Shreffler and Owen (1963) is a serum protein detectable by heteroantisera. The plasma concentration of this protein, the Ss (serum substance) protein, is controlled by a single gene with two alleles, Ssh and Ssz. The serum level is 5- to 20-fold higher in Ssh animals than in Ssz animals ( Shreffler and Passmore, 1971). The second one, the Slp (sex-limited protein) trait, involves the expression of a plasma protein that can be detected by alloantisera (Passmore and Shreffler, 1970). This protein is controlled by a single genetic locus, which so far has been impossible to separate from the Ss locus. The presence of the Slp antigen is determined by a dominant allele, Slp", which is under androgen-dependent regulation ( Passmore and Shreffler, 1971). Thus, females or castrated males of the Slp" genotype do not express the Slp alloantigen, whereas normal males and testosterone-treated females carry the trait. Animals with the recessive allele, Slpo, appear to lack the property to express the Slp antigen. Serological analyses have suggested that the Slp antigen is an allotypic variant of the Ss protein ( Shreffler and Passmore, 1971), These workers demonstrated, however, that all Ss molecules did not express the Slp antigen in Slp- positive sera.

122


Therefore, it is conceivable that the Ss-Slp traits involve two genes coding for very similar gene products. It is, however, impossible as yet to decide unambiguously whether the locus is structural or regulatory. The third trait was first described by D6mant et al. ( 1973). They made the important observation that the hemolytic complement level is controlled by the H-2 complex, in particular the S region. They noted that the complement level is well correlated to the Ss-protein concentration. This led to the suggestion that the Ss protein might be a complement component. Recent data have fully substantiated this idea (see below). The regulation of the complement level is the only functional property yet attributable to the S region. In spite of extensive analyses no convincing data have appeared that seem to involve the S region in histocompatibility reactions or immune response regulation. In keeping with this, there is no apparent serological relationship between the Ss-Slp proteins and the H-2 antigens. Furthermore, the Ss protein, in contrast to most other traits, is mainly not a cell surface-bound molecule, but a plasma protein. An interesting observation shows, however, that the Ss protein may be present on the surface of fibroblasts and other types of cells (Saunders and Edidin, 1974). In consideration of this finding it is possible that the association of the Ss-Slp traits with the H-2 complex does not represent an evolutionary “accident” but that the S region fulfills its function in concert with the other loci, which resemble one another rather strikingly (see below). The I region has turned into the most complex part of the MHC. It was early realized that this region had a key role in the graft-versus-host reaction ( GvHR). The development of the in vitro correlate to the GvHR, the mixed-leukocyte culture reaction ( M L R ) (Bain et al., 1963; Hirschhorn et al., 1963), has given impetus to a vast number of studies. With use of this technique, i.e., cuIturing of allogeneic lymphocyte populations in the presence of [ 3H]thymidine to measure the induction of proliferation, several laboratories have been able to narrow the responsible part of the H-2 complex. Consensus has now been reached that the I region is the most important (but not sole) part of the H-2 complex regulating the MLR response (Bach et al., 1972a,b; Meo et al., 1973a,b). In agreement with this observation it has also been documented that the GvHR is mainly controlled by the same region (Klein and Park, 1973; Livnat et al., 1973). The MLR is often accompanied by the education of some responder cells to become aggressor cells in the CML. Recently, Cantor and Boyse (1975a,b) were able to separate these two reactions. The responder cell population contains two subsets of T lymphocytes that are distinctive owing to their expression of certain cell-surface antigens, the Ly antigens.


123

One subset expresses the Ly-1 antigen whereas the other subset carries Ly-2,3 antigens. The Ly-l-positive T lymphocytes appear to recognize an I-region difference, manifest on the surface of the stimulator cells, and are thereby triggered to divide. Thus, if the stimulator and responder cells differ only with regard to the I region, a strong proliferative response is achieved. No cells, or very weakly active cells, in the CML assay will be obtained under these circumstances. The other subset of T lymphocytes expressing the Ly-2,3 antigens seem to recognize the difference in Kand/or D-region antigens. This difference by itself does not appear to provide enough of a proliferative stimulus to allow this subset of T lymphocytes to become active “killer” cells in the CML assay. However, an I-region-induced mitogenesis of the Ly-l-positive T-cells promotes the culture conditions, so that in the event of a simultaneous H-2K and/or D difference the second subset of T cells (Ly-2,3 positive) proliferates to become active “killer” cells. In contrast to the in vitro reaction, the GvHR seems to reveal the presence of histocompatibility antigens controlled by the I region. Klein et al. (1974) have shown that strong skin graft rejection can be obtained over an isolated I-region difference, although positive CML has not been recorded in any such combination. This discrepancy may merely reflect differences in sensitivity between in uitro and in uivo methods. Therefore, it seems reasonable to conclude that the I region determines (several) types of cell surface molecules, which are recognized by allogeneic T lymphocytes. Soon after I-region congenic strains had become available a number of laboratories were able to raise alloantisera directed against I-regionassociated antigens ( Ia antigens). An excellent review of these findings has recently appeared (Shreffler and David, 1975). Only a few salient features of such reagents will be noted here. Although it was generally expected that anti-Ia sera would react preferentially with T lymphocytes owing to the observed I-region-controlled immune response regulation, manifest at the T-cell level, most antisera did in fact react well with B lymphocytes and poorly if at all with T cells. It became apparent that the Ia sera generally contained a number of antibodies directed against distinct antigenic determinants, and some evidence has been presented that the Ia-antigenic specificities can be shown to belong to one of three groups: ( a ) those expressed exclusively on B lymphocytes, ( b ) those present only on T-cells, and ( c ) those shared between B and T lymphocytes (see Gotze, 1975). However, the Ia antigens are not exclusively confined to the lymphocyte lineages; at least some are present on macrophages, sperm, and epidermal cells, but none has been found on erythrocytes, brain, liver, or kidney cells ( McDevitt et aZ., 1974). It should

124


be realized that as yet almost only cytotoxic assays have been used to define the Ia-antigenic specificities. Despite this limitation, some 20 specificities have been defined. Some of these are clearly associated with different molecules (Cullen and Nathenson, 1974), but whether that will turn out to be the case for all specificities remains unsolved. That the I region codes for a great number of distinct molecules can, however, be inferred from the fact that recombinants have been obtained which allow a subdivision of the I region into possibly 5 separate entities. Since several Ia-antigenic specificities are attributable to one or another of these subregions, it appears reasonable to ascribe such specificities at least, to different molecules, Aside from the cytotoxically defined antigens, the antisera may contain a number of antibodies that for some reason do not induce cytotoxicity. Therefore, it is likely that, when additional techniques are introduced in analyzing the anti-Ia sera, their reaction pattern will become increasingly complex. One of the more exciting findings with regard to the H-2 complex is that this chromosomal segment controls the immune response to most if not all T-lymphocyte-dependent antigens (for reviews, see McDevitt and Benacerraf, 1989; Benacerraf and McDevitt, 1972). This led to the suggestion that T cells (in contrast to B cells) express an H-2 region coded antigen receptor ( Benacerraf and McDevitt, 1972). By analyses of several recombinant strains McDevitt et al. (1972) were able to map the immune response regulating gene for one particular antigen to the I region of the H-2 complex, and subsequent studies have corroborated and extended this finding. Current knowledge indicates that the I region contains at least two well defined loci, which are both engaged in controlling the response to several antigens, The mode of action of these two loci is probably very similar. Initial studies suggested that this regulation occurred only at the level of the T cell. Somewhat disappointingly, however, when analyses were made for the antigen-binding property of virgin T cells no clear-cut difference was noted between high-responder and low-responder animals ( Hammerling and McDevitt, 1974). An intriguing explanation for this unexpected finding is suggested from recent experiments by Munro and Taussig (1975). In analyzing the responsiveness of several inbred strains of mice to the synthetic polypeptide TGAL they obtained suggestive evidence demonstrating that two linked I-region genes are involved in the control of the immune response. One of these genes should be expressed in the T cell and display antigen-binding properties whereas the other gene product should be confined to the B cell constituting a receptor for the T-cell molecule. This would suggest


125

that the Hhmerling-McDevitt data conceivably could be explained on the assumption that these workers have examined strains of mice where the low responsiveness is due to a defect manifest not at the T-cell antigen-binding level, but at the level of the B-cell receptor. It should be pointed out, however, that the origin of the T-cell antigen receptor is still very controversial. Available information may therefore be interpreted in alternative ways, e.g., such that the I region does not confer the actual antigen-binding structure of the receptor (Binz et al., 1975) but merely regulates subsequent events in the immune response. It is already well documented, at least in one experimental system, that normal T-B cell cooperation, which usually is a prerequisite for antibody formation, is restricted to situations where I-region compatibility exists (Katz and Benacerraf, 1974). The necessity for I region identity between T and B lymphocytes in this type of collaboration does not appear to be antigen dependent. Recent data from two laboratories demonstrate, however, that cooperation in in vivo systems may occur over an I-region difference (von Boemer et al., 1975; Heber-Katz and Wilson, 1975). The hypothesis based upon the data by Katz and Benacerraf (1975) that complementarity in the I-region-coded cell-interaction molecules, which are postulated to be present on the surface of immunocompetent cells, is obligatory for the development of a normal antibody production may thus be not a qualitative, but rather a quantitative, requirement. In a number of cases it has been demonstrated that susceptibility to viral oncogenesis is associated with the H-2 complex (for reviews, see Lilly, 1972; Lilly and Pincus, 1973). The demonstration that susceptibility to Gross-virus-induced leukemogenesis is greatly dependent on a gene, Rgo-1, located in the H-2 complex (Lilly, 1966) has attracted much attention, and in a number of studies it has been documented that the same situation holds true for a number of other RNA tumor viruses. Using recombinant strains the Rgv-1 gene could be shown to map in the K-end region of the H-2 complex, and it has been suggested that Rgv-1 may in fact be an I-region gene (Lilly, 1972) that controls the immune response to virus-induced antigens. The Tlu locus is the predominant marker of the TL region. This locus controls the expression of allotypic variants of certain thymocyte surface antigens. It is likely that these antigens may fulfill some function in the T-lymphocyte lineage differentiation since the expression of these antigens is temporally restricted to the period when the T cell is located in the thymus (Boyse et d.,1966; Boyse and Old, 1969). Some of the TLantigen specificities may, however, be found also at peripheral sites, but only on leukemia cells.

126


An intriguing property of the TL antigens is that they undergo antigenic modulation in the presence of specific antibodies. Thus, treatment of TL-antigen-positive cells with an anti-TLa serum in vivo or in vitro converts the cell to a phenotypically TLa-negative state (Old et al., 1968). Concomitantly, the expression of H-D antigen molecules on the cell increases when the amount of TL antigen decreases, demonstrating that there is some interaction between these two antigens (Boyse et al., 1968). In contrast to H-2K and D antigens the TL-region does not control the expression of any antigens that may serve as target molecules in the CML ( Widmer et al., 1973), yet TL-region incompatibility frequently causes skin graft rejection (Boyse et al., 1972). This strongly suggests that a gene or genes, H ( Tla ), determining histocompatibility antigens is present in the TL region. No separation of the TLa and H( T L a ) traits has been documented as yet. It is therefore not possible to decide unequivocally whether the histocompatibility antigen and TL antigens may be the same despite the negative CML results and their apparently different tissue distribution. Also the T-locus gene products and the H-2K and D antigens display some interaction at the level of expression (Nicolas et al., 1975) as do TLa and H-2D antigens. The T locus appears to control several specific stages of the neuroectodermal development. The various t alleles produce well-defined and seemingly unique derangements of critical cell-specific or tissue-specific organizations in ontogenesis ( Bennett, 1964). The involvement of the T locus in embryogenesis, besides the morphological disturbances introduced by the t alleles, is also conspicuously demonstrated by the distorted transmission ratio of the t alleles from heterozygous males, but not from females, to an excess of their progeny. This observation, probably resulting from a superior fertilizing ability of t-allele-bearing sperm, led Bennett et al. (1972) to examine the presence of T-locus-controlled antigens on sperm. Their effort met with success, and several t-allele-specified antigens have now been demonstrated (Yanagisawa et al., 1974). One of these antigens (t"), coded for by a t allele, which manifests its action at the morula stage, has been shown to be present on certain pluripotent teratoma cells as well as on morulae ( Artzt et al., 1974). Under in vitro conditions some teratoma cell lines may differentiate. When this happens the cells gradually lose their t'? antigen but acquire H-2 antigens instead (Nicholas et al., 1975). These findings are very recent, but it appears to be already established that the T locus codes for a series of cell surface antigens that are temporally restricted in expression to certain stages of embryogenesis. Thus, although perhaps not yet conclusively established, there is strong presumptive evi-


127

dence that the T-locus-coded antigens are intimately involved in critical cell-to-cell recognition processes. IV. Isolation and Characteristics of /3,-Microglobulin

Several gene products coded for by loci located on the 17th mouse chromosome are composed of two types of polypeptide chains. The smaller subunit has been identified as p,-microglobulin ( see Section V ) . Various aspects of the biochemistry and biology of p,-microglobulin have recently been authoritatively reviewed by Poulik ( 1975). Human p,microglobulin was first isolated by Berggird and Bearn (1968). These workers examined the excretion of various macromolecules in urine of patients with various diseases. In certain cases with dysfunction of the proximal renal tubuli, Berggird found that one of the main macromolecules in urine was a small protein, termed /?,-microglobulin ( Berggird, 1964, 1975). Urinary protein from such patients was fractionated with use of a combination of ultrafiltration, zone electrophoresis, gel chromatography, and ion-exchange chromatographies. These purification steps provided an adequate scheme for the isolation of p,-microglobulin ( Berggird and Bearn, 1968). Recently, Berggird ( 1975) has described a simplified version of his original fractionation procedure, which allows rapid isolation in good yield of highly purified p,-microglobulin. p,-Microglobulin occurs in small amounts in body fluids like plasma, urine, saliva, cerebrospinal fluid, and colostrum (Evrin et al., 1971). It is now well established that p.-microglobulin in plasma is catabolized mainly by the kidney (Peterson et al., 1969; Bernier and Conrad, 1969). In turnover studies of p,-microglobulin in rats it was demonstrated that the half-life of the protein increased more than 5-fold on bilateral nephrectomy ( Fredriksson and Peterson, 1975; Bernier and Conrad, 1969). Under normal kidney conditions, the urinary excretion of p,microglobulin was very low, but in conditions where proximal tubular dysfunction had been induced experimentally, a substantial portion of the intravasally administered pr-microglobulin could ultimately be recovered in the final urine (Fredriksson and Peterson, 1975). These experimental findings have their precedents in various human disorders. In patients with impaired glomerular filtration rate, the serum level of p2microglobulin increases almost 2-fold with each 50% reduction of the glomerular filtration rate (Wibell et al., 1973). An increase in the urinary excretion of p,-microglobulin is, however, dependent on the glomerular function only to a moderate degree, but is almost exclusively regulated by the proximal tubular reabsorption of the filtered p,-microglobulin ( Fredriksson and Peterson, 1975). Patients with normal serum

128


level of the protein may, in the event of tubular damage, excrete 1000fold more µglobulin than usually noted in persons with normal kidney function (Peterson et al., 1969). Therefore, urine from patients with renal failure often provides an excellent source for the isolation of p,-microglobulin. By taking advantage of the knowledge that the urinary excretion of human µglobulin is greatly increased in tubular kidney disease, Smithies and Poulik (1927b) successfully isolated the dog homolog after induction of kidney damage by surgery and X-irradiation. A selective, reversible damage of the renal proximal tubules may also be accomplished by administration of, e.g., sodium chromate or sodium maleate. By using such salts to increase the excretion of low molecular weight proteins, it has been possible to obtain /µglobulin in highly purified form from rabbit (Berggird, 1974), guinea pig (Berggird, 1975) rat (Poulik et al., 1975; I. Berggird, personal communication), and mouse urine (Anundi et al., 1976; I. Berggird, personal communication). It appears that µglobulin is most easily obtained from urine, but other sources have also been explored. Rask et al. (1974a), Natori et al. ( 1975), and Anundi et al. (1976) have isolated the mouse protein from crude cell membrane fractions, and the human protein has been obtained from spent culture media as well (Nakamuro et al., 1973). Berggird and Bearn (1968) showed that p,-microglobulin is devoid of carbohydrate and has an apparent molecular weight of 11,800. It consists of a single polypeptide chain of 100 amino acid residues containing a single disulfide bridge. From its small size and electrophoretic mobility (,&-mobility on agarose electrophoresis at pH 8.6), they provisionally termed the protein p2-microglobulin.Almost 10 years after the initial discovery of p,-microglobulin ( Berggird, 1964), Smithies and Poulik (1972a) subjected the protein to analysis in an automatic sequencer. They came up with the exciting finding that a high degree of homology existed between the NH,-terminal portion of p,-microglobulin and part of the amino acid sequence of the immunoglobulin G heavy chains. This information was corroborated and extended by Peterson et al. (1972) and Cunningham et al. (1973), who determined the complete amino acid sequence of µglobulin. The homologies were also apparent in the COOH-terminal part of the molecule. It is particularly noteworthy that the two cysteine-residues in p,-microglobulin, which form a disulfide bridge, occupy the same relative positions as do the cysteines in all the immunoglobulin domains ( see Edelman and Gall, 1969) . Furthermore, the amino acid residues shared between ,&-microglobulin and the immunoglobulin homology regions are dispersed throughout the entire amino acid sequence. The homology is even more striking in view of


129

the fact that p,-microglobulin displays identical amino acid residues in 10 out of the 11 positions where all immunoglobuIin homology regions have identical residues, The homology between p,-microglobulin and the C, and C,,3 domains is somewhat greater (26%and 27%) than that between p2-microglobulin and the CH1 and CH2 domains (21%). The similarity with the latter two homology regions is less than that for the homology regions to one another (28%to 34%)but appears nonetheless to be highly significant. The immunoglobulin homology regions have been suggested to be independently folded into compact domains (Edelman and Gall, 1969; Edelman et al., 1969), which recently has been proved by X-ray crystallography (Poljak et al., 1973; Edmundson et al., 1974). These studies have shown that there is a “basic immunoglobulin fold,” which is an outstanding feature of all the immunoglobulin domains. It seems reasonable to predict that the same type of folding will be a feature even of the p,-microglobulin tertiary structure. The similarities between immunoglobulins and p,-microglobulin may suggest that p,-microglobulin represents a free immunoglobulin domain which has evolved from the same ancestral gene that has given rise to regular immunoglobulin light and heavy chains (Peterson et al., 1972). The partial amino acid sequences of p,-microglobulin from dog (Smithies and Poulik, 1972b) and rabbit (Cunningham and Berggbrd, 1974) have been determined. As expected, there is extensive homology between the protein from these two species and human p--microglobulin (80-9M of the residues are probably invariant). The amino acid sequence determinations of the homologs, although not yet complete, lend support to the view that during evolution the gene for p,-microglobulin has accumulated mutations at a rate only half that recorded for the constant domains of immunoglobulin light chains (see Dayhoff, 1972). The amino acid compositions of p,-microglobulin from mouse (Natori et al., 1975) and guinea pig ( Berggbrd, 1975) suggest that also in these species a high degree of conversation in the primary structure of p,-microglobulin will eventually be recorded (Table 11). It is particularly reassuring that p.-microglobulins from all species examined display but a single disulfide bridge, which most likely forms the typical immunoglobulin loop. The resemblance between p.-microglobulin and the immunoglobulin domains, documented at the level of primary structure, will most probably also be apparent in the tertiary structure (see above). Crystallographic data for p,-microglobulin have, however, not yet been published, but physical-chemical investigations are consistent with similarities in structure between this protein and the immunoglobulin domains (Table


130

AMINOA C I D CONTENT

OF

TABLE I1 &-MICROGLOBULIN FROM F O U ISI'l~X!Il*:S" ~ Source of &-microglobulin

Amino acid

Human*

Mousec

Rabbitd

Guinea pigd

Aspartic acid Threonine Swine Clutaniic acid Prolinc Glycine Alanine IIalf-cystine Valinc Metliionine Isoleucine Lcucinc Tyrosine Phcn ylalanine Lysine Histidine Ar g i n i n e Tryptophan

12 5 10 11

10 7 7 11 8

15 4 6 11 8

14 3

4

3

3 2 5

2 2 10

4

1

1

6

5

5 3

2 2 7 1 5 7 6 5 8

9

10 7 3 4 2 9

4

3 7

4

5

4

5 8

5

4 4 2

5

4 5

9 4 4

2

2

7 4 8 3 2

0 All values are expressed as residues per molecule, rounded off to the nearest integer. * From Cunningharn rt aE. (1973). c From Natori rl al. (1975). d From BcrggeLrd (197.5).

111). Karlsson (1974) has shown that p,-microglobulin has a low frictional ratio, like free constant and variable domains of immunoglobulin light chains (Karlsson et al., 1972), suggestive of a compact, almost spherical shape. Chiroptical methods reveal the presence of p-structure and absence of significant amount of a-helical structure in p,-niicroglobulin, which both are common properties of immunoglobulin domains ( Dorrington and Tanford, 1970). Circumstantial evidence for a similarity in tertiary structure between p,-microglobulin and immunoglobulin domains is suggested by the observations that p,-microglobulin binds to the lymphocyte Fc receptor (see Section VIII), a property attributed to the C,13domain of IgG (Yasmeen et al., 1973), and that it can interact with C1, the first component of complement (Painter et al., 1974). The latter property, which /?,-microglobulin shares with the CI,2 domain (Kehoe and Fougereau, 1969), may, however, be dependent upon the

THE MAJOR HISTOCOMPATLBILITY COMPLEX

131

TABLE I11 PHYSIC.IL-CHICMICAL Pao~icn~r~ OFe sHUMAN &-MICROGLOBVLIP Sedimentation coeficient, s ; , , ~ Stokes' molecular radius, ra Diffusion coefficient, D z o , ~ Frictional ratio, f/fo Molecular weight Partial specific volume, P Mean residue ellipticity at 218 nm Mean residue rotation a t 218 nm Molar extinction coefficient, E ? B O " ~ Tryptophan quantum yield, Exs8onm

1.6 S 16 13.3 X lo-? cm2sec-I 1.10 11,815 0.727 ml gm-l - 1900" em2 dmole-1 -550"cm2 dmole-1 19,850 0.09

Data from Bergghrd and Beam (1968) and Karlsson (1974).

primary structure rather than the tertiary structure ( Isenman et al., 1975). Only several years after the demonstration of P2-microglobulin in various body fluids it was realized that this protein was present on the surface of leukocytes (Peterson et al., 1972). Bernier and Fanger ( 1972) showed that lymphocytes synthesize p,-microglobulin, and this is a property both of T and B lymphocytes (Poulik, 1973). These findings together with the amino acid sequence information focused the attention on a possible common regulatory control mechanism for p--microglobulin and immunoglobulins. Two studies unequivocally showed that there is no linked synthesis of these two types of molecules (Hiitteroth et al., 1973; Nilsson et al., 1973). In fact, ,8,-microglobulin is synthesized by a great variety of nonlymphoid cells including those of mesenchymal and epithelial origin (Nilsson et al., 1974). To this date only two cell types have been shown to be devoid of p,-microglobulin. It has constantly been found that erythrocytes do not manufacture the protein (Evrin and Pertoft, 1973). The other cell type lacking ,&-microglobulin is an in vitro grown human lymphoma line ( Nilsson et al., 1974; Ostberg et al., 1975b ) , which displays a deletion in chromosome 15 (L. Zech, personal communication). This suggests that chromosome 15 may be involved in the control of the &-microglobulin synthesis, and recent data demonstrate that the structural gene for p,-microglobulin in fact resides on this chromosome ( Goodfellow et al., 1975). Although a wealth of data now support the view that p,-microglobulin is primarily a cell-surface protein (Peterson et al., 1972), all cell lines that synthesize the protein also secrete various amounts ( Nilsson et al., 1974). It is likely that this is not an in vitro artifact since ,&-microglobulin in vivo is widely distributed in biological fluids. Whereas p,-microglobulin

132


on the cell surface appears to be bound to other polypeptide chains (see Section V ) it occurs predominantly in free form extracellularly. This suggests either that &-microglobulin is synthesized in excess over its complementary chains or that the turnover of p,-microglobulin and the associated macromolecules occur via different routes. It should be of interest to examine the synthesis and degradation of p,-microglobulin in relation to H-2 and TL alloantigens, in order to get some information as to the regulation of the cell surface expression of these molecules. V. Features of the Classical Transplantation Antigens

As for all studies of cell membrane components, the progress in biochemical analyses of histocompatibility antigens has been hampered for two main reasons. First, owing to their very nature as cell membrane proteins the histocompatibility antigens are integrated into a lipid matrix. The cell membrane-integrated part of the molecules is therefore hydrophobic, and means have to be employed to release the histocompatibility antigens in water-soluble form. Second, although the classical histocompatibility antigens are expressed on almost all types of cells they are present in relatively low amounts. The scarcity of material is even more apparent considering the extensive polymorphism of the system. An advantage is, however, conferred to the system by its polymorphic nature. Combinations of mouse strains can be obtained that differ from each other only in one region of the H-2 complex. By reciprocal immunizations with cells or cell membrane fragments, such strain combinations will allow the production of seemingly monospecific antisera. Many alloantibodies directed against H-2 complex-coded antigens are cytotoxic and provide the basis for a sensitive method for the detection of the relevant alloantigens. Thus, the presence of alloantigens can be monitored by the inhibition of alloantibody-induced lymphocytotoxicity as measured by uptake of a vital dye, like trypan blue (Pincus and Gordon, 1972), or by the release of radioactive chromium from the target cells (Sanderson, 1964; Wigzell, 1!365). Several methods have been explored to solubilize histocompatibility antigens (see Nathenson, 1970). The two most prevalent methods used in studies of H-2 alloantigens are solubilization by means of controlled proteolytic digestion of crude cell membrane fractions ( Nathenson and Davies, 1966; Shimada and Nathenson, 1967) and solubilization of cell membrane molecules with the use of nonionic detergents (Kandutsch and Stimpfling, 1963; Schwartz and Nathenson, 1971a). The method involving proteolysis is believed to sequester the transplantation antigens into minor fragments. Since the assay for their detection is based upon


133

the antigen reacting with alloantibodies, only those fragments expressing the particular antigenic specificity recognized by the alloantiserum will be discovered. Solubilization with use of detergent probably releases from the cell membrane most if not all of the histocompatibility antigen molecule, Accordingly, it appears advantageous to use detergent for the solubilization but material solubilized in this manner is rather difficult to handle. The histocompatibility antigens form complexes with the detergent (Helenius and Simmons, 1974), which has to be present during the entire isolation procedure to minimize the tendency toward aggregation of the H-2 alloantigens. The presence of detergent renders lymphocytotoxic techniques hard to perform, since the detergent per se is cytotoxic. To circumvent the latter problem, Schwartz and Nathenson (1971a) resorted to an alternative isolation and assay method. They introduced radioactive label into the cell membrane molecules and precipitated the alloantigens, solubilized by treatment with nonionic detergent, with use of specific antisera. The antigen-antibody reaction occurs readily in the presence of nonionic detergents, so the specificity and rapidity of this technique has been of immense value in analyzing the various cell membrane components controlled by the H-2 complex. It should be realized, however, that this procedure will yield only limited information, since a prerequisite for further analyses is solubilization of the antigen-antibody precipitates, which usually has to be performed under denaturing conditions. Shimada and Nathenson ( 1969) established the glycoprotein nature of the serologically detected H-2 alloantigens. The classical histocompatibility antigens, solubilized by limited proteolysis with papain of a crude cell membrane fraction, were purified by a series of fractionation steps including ammonium sulfate precipitation, gel chromatography, and polyacrylamide gel electrophoresis. By this isolation procedure Shimada and Nathenson (1969) obtained two types of fragments, denoted class I and class 11, which had been purified to about a 500-fold higher antigen specific activity than the starting material. Class I fragments of both the H-2K and D regions displayed an apparent molecular weight of about 37,000. Some 10-20% of the weight was constituted by neutral sugars, glucosamine, and sialic acid. Class I1 fragments were smaller, apparent molecular weight 28,000, and contained percentagewise slightly more carbohydrate than the class I fragments. In view of the similarities in chemical composition between the two classes of fragments, it was suggested that the smaller component might have arisen from the larger fragment by action of the protease. These studies suggested that the transplantation antigens were composed of a single polypeptide chain. However, two laboratories demon-

134


strated that the human equivalent, the HL-A antigens, contained an additional subunit ( Cresswell et al., 1973; Tanigaki et aZ., 1973), After this observation Rask et al. (1974a) and Silver and Hood (1974) reexamined the polypeptide chain structure of H-2 alloantigens. Rask et al. (1974a) solubilized H-2 alloantigens by the procedure used by Shimada and Nathenson (1969) and purified the H-2 alloantigens by consecutive steps of CM-cellulose ion-exchange chromatography, Sephadex G-200 gel chromatography, and DEAE-Sephadex ion-exchange chromatography. In spite of several further attempts to achieve a higher purity (greater than about 50%),the charge heterogeneity of the H-2 alloantigens precluded the use of additional conventional physical-chemical separation methods. The final isolation step resorted to was, therefore, immunoprecipitation with specific alloantibodies. This purification procedure yielded H-2K as well as H-2D antigens which resolved into two components on sodium dodecyl sulfate ( SDS ) -polyacrylamide gel analysis of the dissolved immune precipitates. The predominating species displayed an apparent molecular weight of 37,000 both for H-2K and D antigens whereas the minor component had an approximate molecular weight of about 12,000. All the alloantigenic determinants were present on the larger polypeptide chain. Silver and Hood (1974) solubilized membrane macromolecules with use of a nonionic detergent and precipitated the H-2 alloantigens with specific antisera. H-2 alloantigens isolated by this procedure exhibited similar characteristics as the papainsolubilized fragments obtained by Rask et al. (1974a), but the major polypeptide chain had a higher molecular weight (see below). Rask et al. (1974b) went further and by peptide mapping experiments they could show that there exists homologies in primary structure between human p.-microglobulin and the small H-2 alloantigen subunit. This information together with a weak but highly significant immunological cross-reactivity between the small mouse H-2 alloantigen subunit and human p2microglobulin established the similarity between the two proteins ( Rask et nl., 1974a,b). Previous results had already documented that human p.-microglobulin is part of the HL-A antigens (Nakamuro et al., 1973; Peterson et al., 1974; Grey et al., 1973). The H-2K and D antigens are distinguished serologically by their particular private antigen specificity. Separation of K and D region antigens may accordingly be achieved with use of carefully selected alloantisera. It has, however, proved to be very difficult to achieve the same type of separation employing physical-chemical techniques. The size of H-2K and H-2D antigens, solubilized by limited proteolysis, is indistinguishable when examined by gel chromatography. Thc size of these alloantigenic fragments is very similar to that of albumin suggesting that

THE M A J O R HISTOCOMPATIBILITY COMPLEX

135

the molecules are either asymmetric and/or contain a relatively large amount of carbohydrate (see below). The size homogeneity of the alloantigens contrasts sharply with the charge heterogeneity. On DEAEcellulose ion-exchange chromatography at around physiological pH and ionic strength, H-2 alloantigens from the K region are eluted over a considerable part of the chromatogram. In similar analyses the H-2D alloantigen activity is distributed in a fashion resembling that noted for the K-region antigens. A number of explanations for this heterogeneity were explored. It could conceivably result from proteolysis occurring at different positions in the polypeptide chain, as evidenced by the occurrence of class I and I1 fragments (Shimada and Nathenson, 1969). Some H-2 molecules, but not all, may have lost p,-microglobulin. A variable content of sialic acids in the H-2 alloantigens would also induce charge heterogeneity. Available information ( Anundi et al., 1976) suggests that all these explanations may contribute to the charge heterogeneity. The single, most important cause for the heterogeneity, however, appears to be variations in content of sialic acid. Rechromatography of material originally subjected to ionexchange chromatography on a column of DEAE-Sephadex under identical conditions but after neuraminidase digestion shows that the H-2 alloantigens are eluted considerably earlier. This provides suggestive evidence for the importance of sialic acid for the observed heterogeneity, since after its removal the heterogeneity is drastically diminished. Similar findings have also been reported for papain-solubilized HL-A antigens (Parham et al., 1974). Although it is commonly believed that each private antigen specificity is expressed on molecules of identical primary structure, recent observations may argue against this notion (see below). Hess and Smith (1974) have suggested that heterogeneity is an inherent property of H-2K and D alloantigens, possibly owing to variations in amino acid sequence of H-2 alloantigens coded for by a single K or D region. This would mean that such a K or D region may compromise more than one structural gene (see below). Nathenson and co-workers have examined the carbohydrate structure of H-2 alloantigens in some detail. In agreement with studies on the reason for the H-2 alloantigen heterogeneity they found at least 3 sialic acid residues per H-2 molecule (Nathenson and Muramatsu, 1971). By using various radioactively labeled monosaccharide precursors, which were incorporated biosynthetically into H-2 alloantigens, Muramatsu and Nathenson ( 1970a,b, 1971) isolated glycopeptides from H-2 antigens. On gel chromatography and DEAE-Sephadex ion-exchange chromatography the glycopeptides were indistinguishable regardless of whether their origin was from normal spleen cells or from tumor cells (Muramatsu

136


and Nathenson, 1971) or whether the cells had been derived from mice of different H-2 haplotypes ( Muramatsu and Nathenson, 1970a). From these and other analyses Nathenson and co-workers have concluded that H-2 alloantigens display one, or at the most two, carbohydrate moieties. Furthermore, since the structure of the carbohydrate chain( s ) is so very similar for H-2 antigens of different haplotypes, it is unlikely that the carbohydrate portion contributes significantly to the antigenic characteristics of the H-2 alloantigens ( Nathenson and Muramatsu, 1971). Attempts have been made in Nathenson’s laboratory to disclose the detailed structure of the carbohydrate moiety, Enzymic digestions with use of various endo- and exoglycosidases combined with alkali and acid treatments of the carbohydrate portion have given substantial information. This has allowed Nathenson and Muramatsu (1971) to present a hypothetical model for the carbohydrate moiety of H-2 antigens. The carbohydrate-protein linkage is of the glucosaminyl asparagine type. The core sugars seem to include mannose in addition to glucosamine. Possibly three branches, terminating with sialic acid and containing galactose and glucosamine, are attached to the core. It is particularly interesting to note that the core sugars compromise most if not all of the mannose residues and a substantial portion of the glucosamine moieties, in view of recent findings with regard to the biosynthesis of cell membrane proteins. Leloir and his co-workers (see Behrens et al., 1973) have shown that there exist lipid intermediates, composed of dolichol and pyrophosphate, that sequentially accumulate monosaccharide units to build up oligosaccharides, that are transferred en bloc to some cell membrane proteins to become glycosylated. These oligosaccharide units appear to contain, e.g., a series of mannose residues in adjacent positions. It is, therefore, quite possible that H-2 alloantigens acquire a significant portion of their prosthetic groups in this manner. Scarcity of material and difficulties in obtaining H-2K and D antigens in a reasonably pure state have hampered progress in elucidation of the structure of the protein part of the molecules. Amino acid compositions have been determined for papain-solubilized H-2 antigens of different haplotypes (Shimada and Nathenson, 1971) but since no separation of H-2K and D region antigens was attempted such analyses give doubtful information. They show, however, that all the common amino acids are present and that cysteine and methionine are the least abundant. There are about 4 cysteine residues per papain-solubilized H-2K and D antigen, and they all seem to form disulfide bridges (Peterson et al., 1975a). The number of methionine residues is similar to that for cysteine as evidenced from cyanogen bromide fragmentation and the location in the primary structure of these residues appears relatively conserved both for H-2K


137

and D antigens ( Anundi et al., 1976). Several laboratories are engaged in work aiming toward the elucidation of the primary structure of H-2 alloantigens. Silver and Hood (1975) have devised a promising method for such determinations based on the selective labeling of one particular type of amino acid at a time followed by automatic sequence analysis of the entire H-2 antigen molecule. Those positions where the labeled residues appear are noted. By separately introducing label into each one of the various amino acids the complete structure may theoretically be obtained. It is doubtful whether the efficiency of this methodology will allow determination of the entire amino acid sequence, but large portions of the NH,-terminal end will certainly be determined. Therefore, it is to be expected that a great deal of information will appear shortly. The papain-solubilized H-2K and D alloantigens, although relatively convenient for biochemical work, represent but a portion of the intact molecule. To get an idea of the intact antigens, Schwartz and Nathenson (1971a) introduced the use of the nonionic detergent NP-40 for solubilization of cell membrane molecules. The H-2K and D antigens recovered with this procedure were, as expected, larger than their proteolytically derived counterparts. In a careful study Schwartz et al. (1973b) examined the molecular weight of such antigens. They repeatedly found a small but significant difference between H-2K and D antigens, The K antigens, with an apparent molecular weight of 47,000, contain a 4,000-dalton polypeptide chain portion more than the D antigens. Papain-digestion of H-2 alloantigens first solubilized with NP-40 gave rise to products very similar to those usually encountered on direct proteolytic treatment with papain of crude cell membrane fractions. This observation together with the fact that glycopeptides derived from papain-solubilized and Nonidet P-40 ( NP-40)-solubilized H-2 alloantigens are indistinguishable when compared by gel chromatography lend strong support to the view that the papain-solubilized fragments are indeed derived from the larger type of molecules obtained with NP-40 (Schwartz et al., 1973b). In the same study Schwartz et al. (1973b) made the important observation that a sizable fraction of the isolated H-2K and D antigens occurred as d i d fide-linked dimers. This result was recently corroborated by Peterson et al. (1975a), who demonstrated that the disulfide bridge (or bridges) is located in that part of the molecule in closest proximity to the cell membrane. In both studies, however, monomers of the H-2K and D antigens were encountered as well. Rask et al. (1974b) also obtained data consistent with a structure comprising two large chains for H-2 alloantigens. They did not, however, find any interchain disulfide bridges. These results suggest either that H-2 alloantigenic chains are normally disulfide-linked, but that on solubilization and further processing the

138


labile disulfide bridge( s ) is reduced in some molecules, or that the isolation conditions fortuitously introduce disulfide bonding in part of the antigens. At present this question remains unresolved. In our opinion, circumstantial evidence favors the view, however, that the H-2 alloantigens may comprise two alloantigen-carrying chains held together by noncovalent forces, at least (see Peterson et aZ., 1975b). A structure for HLA antigens encompassing two large chains has also been described (Strominger et aZ., 1974; Peterson et al., 1975a; Cresswell and Dawson, 1975). Recently, however, Snary et d. (1975) have cast doubt on the validity of these determinations. When deoxycholate was used to solubilize the HL-A antigens, no disulfide-linked dimers were obtained. The elution profile on gel chromatography of the solubilized HL-A antigens suggested, however, an anomalously high molecular weight. Snary et al. (1975) provide compelling evidence to suggest that the high molecular weight is due to detergent binding rather than to dimerization. It remains, however, to be established whether noncovalently linked HL-A dimers may sustain their protein-protein interactions in deoxycholate before the idea about a model for HL-A and H-2 antigens comprising two alloantigenic chains is abandoned. Physical-chemical examinations of H-2 alloantigens solubilized by NP-40 reveal that they are slightly smaller than immunoglobulin G. Since the molecular weight of these antigens is only about 120,000, the size characteristics signify an elongated shape provided the hydration is “normal.” The high frictional ratio is, however, also consistent with a globular structure comprising a significant amount of carbohydrate. The relatively low value found for the partial specific volume is in agreement with the latter view, From the physical-chemical properties, it is obvious that they are representative of a “typical” glycoprotein. The molecular weight estimations have been performed with the use of four different techniques, and since they are in good agreement they probably are reasonably accurate. From the weights of the constituent polypeptide chains, pz-rnicroglobulin, and the alloantigen-carrying chain, it is tempting to suggest that they occur in equimolar amounts (Peterson et aZ.; 1975a). The weight of p,-microglobulin is low, and minor errors in determinations of the weight of the NP-40-solubilized species may obscure the exact number of small chains in the molecule, It is, however, well documented in several studies that the papain-solubilized fragment comprises about 75%of the entire alloantigenic chain (Schwartz et aZ.,197313; Rask et aZ., 1974b) and a single molecule of µglobulin. Therefore, it is reasonable to assume that two alloantigenic chains and two p,-microglobulin molecules make up the intact H-2 alloantigen (Rask et aZ., 197413).


139

Limited proteolysis with papain of NP-40-solubilized H-2 alloantigens yields antigenically active molecules indistinguishable from the class I fragments (see above), Further digestion of the class I fragments preferentially splits the alloantigenic chain into two components. One component displays a molecular weight of about 13,000 whereas the other part is similar to the class I1 fragments. The latter component comprises somewhat more than 20,000 daltons of the alloantigenic chain and remains attached to p,-microglobulin ( Peterson et aZ., 1975a). Circumstantial evidence has been presented suggesting that each part of the papain-digested H-2 alloantigenic chain, which can be released on further proteolysis, contains a single disulfide bridge encompassing some 60 to 70 amino acid residues (Peterson et al., 1975a). These findings suggest that the heavy H-2 alloantigen chains are composed of tightly folded portions of the polypeptide chain, domains which are connected by exposed stretches of less tightly folded structures prone to be attacked by various proteolytic enzymes. Each domain may contain a single disulfide bridge of the immunoglobulin type, which is also present in ,&-microglobulin ( Peterson et al., 1972, 1975a). The two heavy chains may be linked by one or more disulfide bridges ( see above ) connecting the membrane-bound domains. Each heavy chain probably contains three domains. The outer domain, with an apparent molecular weight of somewhat more than 20,000, binds p?-microglobulin. Most if not all of the carbohydrate seems to be attached to the outer portion of the alloantigens so the polypeptide part of this domain comprises probably only about 15,000 daltons. The membranebound part of the molecule, which may comprise some 10,000-15,000 dalton of the polypeptide chain, most probably display a conformation which is different from the two other domains. Its integration into the membrane lipids, to ascertain anchorage of the H-2 antigens, most probably involve molecular interactions of a type not encountered by other parts of the molecule. The domainlike organization of the heavy polypeptide chain, the possible existence of immunoglobulin-like disulfide loops, the multichain structure, and the occurrence of the “free immunoglobulin domain,” µglobulin, as part of the H-2 alloantigen molecule are suggestive evidence for a common evolutionary origin of H-2 antigens and immunoglobulins. This will be discussed below (see Section X ) . The antigenic specificities of H-2K and D antigens are the markers recognized in the GvHR (Nathenson, 1970). Several studies have therefore been devoted to explore their nature. As discussed above, the carbohydrate moieties of H-2 alloantigens expressing different private and public antigenic specificities are so similar that it can almost certainly be excluded that they contribute significantly to the antigenic sites. Con-

140


sequently, removal of a substantial portion of the prosthetic group does not appear to diminish the antigen expression of the H-2 molecules ( NathenSon and Muramatsu, 1971 ) , Furthermore, isolated glycopeptides from H-2 alloantigens lack the property to inhibit alloantibody-induced lympholysis (Nathenson and Muramatsu, 1971). From these data it may be inferred that the antigen sites are created by the polypeptide portion of the alloantigens. Several lines of evidence support this notion. First, the antigenic sites are very sensitive toward denaturants known to cause derangement of the tertiary structure of proteins. Second, modification of one or more amino acids frequently results in diminished antigenicity. The latter is particularly obvious when c-NH,-groups are modified (Pancake and Nathenson, 1973)- Third, the expected allotypic variation is noted in the primary structure of H-2 alloantigens, as recently shown by Nathenson and Cullen ( 1974). From genetic data it is apparent that each one of the private specificities can be attributed to either the K or D region of a certain H-2 haplotype (see Section 11). This also appears to be true at the biochemical level. Cullen and Nathenson ( 1971 ) have presented convincing data along these lines. Using carefully selected alloantisera to precipitate NP40-solubilized, radiolabeled antigens they were able to show that precipitation of an H-2K antigen, specified by the private antigen specificity, exclusively removed that antigen but did not precipitate the D-region antigen, specified by another private specificity. Thus, the private antigenic specificities are readily separated and the inevitable conclusion is that they are associated with different molecules. This important observation was borne out not only in homozygous situations, but in the heterozygous state as well, Corroborating data have been obtained at the single cell level where it has been shown that each private H-2 antigen on the cell surface may redistribute independently of all other alloantigens present, which are specified by other private antigen specificites ( Neauport-SautCts et al., 1973). Cullen and Nathenson ( 1971) also demonstrated that some public antigen specificities coprecipitated with private antigens. The most straightforward explanation for this finding is that H-2K and D alloantigens may express, in addition to the particular private specificity, one or more public antigenic sites. This interpretation has to be viewed with caution, however, since solubilization and precipitation of H-2 molecules were performed under conditions that give rise to dimer formation. Thus, the isolated H-2 alloantigens comprised two antigen-carrying polypeptide chains. When this information is taken into account, one outstanding feature is obvious. Dimer formation, whether genuine or artifactual, occurs only between identical gene products from a single allele even in the heterozygous state. Al-


141

though a wealth of information convincingly documents the great similarity between K- and D-region allelic antigens the data of Cullen and Nathenson (1971) seem to exclude the existence of the “mixed” type of dimers. In consequence, dimer formation, if genuine, most probably arises at or shortly after polyribosomal release of nascent antigens. Another interpretation of these data is that there are more than one structural gene located in each K and D region. One gene may, thus, code for the polypeptide chain displaying the private antigen specificity whereas a second gene may code for a very similar polypeptide chain carrying the public specificities. This would mean that the K and the D region each contains at least two series of alleles. Indeed, some serological data with regard to the occurrence of public specificities is in agreement with this suggestion (see h e l l et al., 1973). In a recent study DBmant et al. (1975) have presented evidence for the two-gene concept. They found that for some haplotypes papain-solubilized H-2 antigens could be separated into those exhibiting private antigenic specificities and those displaying public antigens. Two types of H-2 alloantigens could not be distinguished by physical-chemical criteria, and both types of molecules comprised b,-microglobulin in addition to the antigenically active chain. Although quite exciting, available information is scarce and has to be corroborated and extended to several more public specificities before a revision of the H-2K and D region genetical fine structure seems implicated. Peptide mapping experiments with separately trypsin-digested H-2K and D antigens, labeled in lysyl and arginyl residues, suggest that no more than 20%to 30%of the resulting peptides are identical (Brown et al., 1974). The similarity is only slightly greater when allelic forms of the H-2 alloantigens are compared. The total number of peptides resolved in these experiments was low, however, so the experiments give but a rough idea of the relationship between the gene products. It should also be kept in mind that the type of analysis, ion exchange chromatography, overemphasizes differences. By using another approach, i.e., selecting for similarities by analyzing only small tyrosine-containing peptides obtained by combined peptic and tryptic digestion (see Cunningham-Rundles et al., 1975),Anundi et al. (1976) have demonstrated very similar peptide patterns for K- and D-region antigens. The latter observation is consistent with the finding that heteroantisera against H-2 alloantigens ( absorbed with pa-microglobulin) cross-reacts extensively with K- and D-region antigens (Anundi et al., 1976). Unfortunately, neither of these results answers unequivocally the question how large the differences are between allelic forms of the H-2 alloantigens, distinguished only by a private specificity. Neither do these data contribute

142


to resolve the question whether each alloantigen is composed of two dissimilar heavy chains ( DCmant et al., 1975). Since the private antigens seem to confer most of the biological spec?ficity to the H-2 alloantigens, (see Sections I11 and IV), it is of great importance to establish the location in the molecule of such antigenic sites. Extensive biochemical work along the lines presented, combined with serological analyses, promises to unravel these features in the near future. There is now general consensus that H-2 alloantigens are predominantly expressed on the plasma membrane. Attempts have been made to analyze the biosynthesis and cell-surface expression of the antigens in relationship to the cell cycle. Schwartz and Nathenson (1971b) used papain to destroy the H-2 alloantigenic sites on intact cells and measured the reexpression of the antigens. The antigens were quantitatively expressed within 6 hours of the digestion. This restoration occurred in a much shorter time than the cell division time. In another study they performed a “pulse-chase” experiment with radioactive amino acid precursors and measured the disappearance of labeled H-2 alloantigens from the cell surface (Schwartz et d.,1973a). The half-life of the labeled molecules was about 7 hours, which also in this case was much shorter than the cell division rate. The fate of the H-2 alloantigens in the turnover process is, however, unknown. The rather short half-life for H-2 alloantigens in in vitro cultured cell lines contrasts with the very slow H-2 antigen turnover noted for normal splenocytes (Vitetta and Uhr, 1975b). It is, therefore, possible that the synthesis and degradation of the H-2 antigens differ considerably among various cell types and even among subpopulations of cells. Studies on cloned cell lines and on subpopulations of splenocytes may thus yield interesting information. It would also be of interest to examine the synthesis and turnover of p2microglobulin, related to the H-2 antigens, particularly in view of the observation that p,-microglobulin, in contrast to H-2 antigens, may be secreted in free form (see Nilsson et al., 1974). VI. Biochemical Properties of the Thymus-Leukemia Antigens

The TL-region gene products display an organ-restricted expression. Normally these antigens are present only on the surface of thymocytes. Some leukemia cell lines may also express the TL antigens, and much work has been performed with such cells. Techniques for the solubilization and assay of H-2 alloantigens have been useful also for biochemical studies of TL antigens. Davies (1966) reported that T L antigens were solubilized by autolysis and were released from crude membrane fractions in a fashion similar to that for H-2 alloantigens. In subsequent


143

studies Davies et al. (1969) demonstrated that T L antigens are very similar to H-2 alloantigens by several criteria, e.g., size and charge. They could, however, achieve a partial separation of H-2 antigens and TL antigens by ion-exchange chromatography. This represented the first biochemical demonstration that the T L antigenic sites are present on molecules distinct from the classical H-2 antigens. TL antigens liberated from crude cell membrane fractions by limited proteolysis with papain yielded apparently homogeneous molecules (Muramatsu et al., 1974). On column chromatography, such molecules were of a size very similar, if not identical, to those of H-2 alloantigens. TL antigens from cells labeled with radioactive amino acid and sugar precursors were also examined by these authors. They could thereby establish that the papain-solubilized TL antigens, isolated by immunoprecipitation, contained carbohydrate like H-2 antigens. The glycopeptides obtained after Pronase digestion of isolated TL and H-2 alloantigens were, however, distinctly different. No details about the carbohydrate portion are as yet known. It is, however, likely that the charge heterogeneity of T L antigens, which again is similar to that observed with H-2 antigens, at least partly depends on variations in the content of sialic acid. Molecular weight estimation of the papain-solubilized TL antigens suggested the existence of a single polypeptide chain with the apparent molecular weight 38,000 (Muramatsu et al., 1973). It was recently shown, however, that TL antigens, like H-2 antigens, are composed of two types of polypeptide chains (Vitetta et al., 1975a; Ostberg et al., 1975a). The smaller subunit is identical to p,-microglobulin. Careful molecular weight determinations of papain-solubilized TL antigens under denaturing conditions established that the constituent polypeptide chains had molecular weights of 37,000 and 12,000, respectively ( Anundi et al., 1975), whereas the NP-40 solubilized antigens (Vitetta et al., 1972; Yu and Cohen, 1974) under identical conditions displayed two components with apparent molecular weights of about 50,000 and 12,000 (Vitetta et al., 1.975a; Ostberg et al., 1975a; Anundi et al., 1975). The demonstration that p,-microglobulin is part of the TL antigen molecule and the observation that the size of the alloantigenic polypeptide chain is very simiIar to that of H-2 antigens prompted an investigation of whether TL antigens, like H-2 antigens, display a tetrameric structure. Anundi et al. (1975) demonstrated that this indeed is the case. As for H-2 antigens Anundi and her colleagues could establish that the majority of NP-40-solubilized TL antigens contain two disulfide-linked heavy, alloantigenic polypeptide chains and two molecules of p,-microglobulin. It is noteworthy that a minor fraction of the molecules was not

144


held together by disulfide bonds. This observation, which has its precedence for H-2 antigens (see Section V ) , reinforces the striking similarity between the two types of molecules. It is likely that TL antigens will turn out to show domainlike structures like H-2 antigens. Some support for this notion was recently presented by Stanton et al. ( 1975), who noted that papain digestion of TL antigens liberates, in addition to the expected 37,000-dalton fragment, a minor component of the heavy chain with a molecular weight of about 20,000. These data are consistent with the view that there are at least two extended stretches of the TL-antigen polypeptide chain, which are available for cleavage by papain. The nature of the particular TL antigenic sites is not known with certainty. The demonstration by Muramatsu et al. (1973) that the carbohydrate portion of TL antigens is different from that of H-2 antigens suggests an intriguing possibility, considering the observation that H-2D antigens and TL antigens exhibit reciprocity in their expression at the single-cell level (Boyse et al., 1968). This could mean that the TL region controls somc glycosyl transferases, rather than structural genes for TL antigens, which provide H-2D antigens with a unique prosthetic group. The TL alloantigenicity could, thus, depend on the carbohydrate moiety. One would have to infer, however, that the addition of the carbohydrate structure obliterates the accessibility of antibodies for the particular H2D alloantigenic sites, which are known to be endowed in the protein part of the molecule. Therefore, the more conventional idea that the TL region codes for the TL antigens (Boyse and Old, 1971) seems more attractive. The latter notion, which is consistent with a proteinaceous nature for the antigenic sites, receives support from the observation that extremes of pH, protein denaturants, and iodine labeling of tyrosine residues completely abolish or greatly diminish the antigenicity of the TL molecules. The striking similarities in structure between TL and H-2 antigens have not yet been documented at the level of primary structure. It seems likely, however, that such investigations will establish regions of homology, since both types of antigens interact in a seemingly identical manner with p,-microglobulin. If this prediction is borne out, a reasonable conclusion is that the TL-region genes are derived from the same ancestral gene that evolved to become the K- and D-region genes. VII. T-Locus Gene Products and P2-Microglobulin

There is rapidly growing awareness that at least certain aspects of embryogenesis is controlled by genes that have products located on the cell surface. One of the first unambiguous demonstrations that this is


145

the case is owed to Bennett et al. (1972). Having examined various morphological and genetic aspects of the T locus (see Dunn, 1964; Bennett, 1964), they turned their interest to the possible existence of T-locus-coded products on the cell surface. Since different t alleles often display distorted transmission rates (see Section I V ) it seemed likely that sperm expresses the t-allele phenotype. Bennett and her colleagues ( 1972) succeeded in raising alloantisera, which selectively recognized antigen T on sperm of mice of the appropriate genotype. They could not detect this antigen in adult tissues other than sperm. In subsequent reports antisera directed against products of other alleles specified by different complementation groups, have also been described ( Yanagisawa et al., 1974). All these antisera react with sperm, but usually only with sperm obtained from mice carrying the particular t allele under examination. It was pointed out above (Section IV) that the T locus obviously encompasses a number of linked genes rather than alleles. This implies that t mutants or their corresponding wild alleles, belonging to different complementation groups, may simultaneously be expressed on sperm. That this indeed is the case has been shown with serological methods; haplotypes of t mutants appear to express three or four antigenic specificities ( see Artzt and Bennett, 1975). Jacob and his associates raised autoantisera against an in oitro grown cell line derived from a mouse teratoma (Artzt et al., 1973). Teratoma cells are embryonic cells believed to display cell- as well as stage-specific antigens. In keeping with this hypothesis, early embryonic antigens most probably appear and disappear long before maturation of the immune system. No self-tolerance should thus be expected to such antigens which therefore would behave like autoantigens (see Alexander, 1972). The antiserum against the teratoma cells (anti-F9) reacted with early stages of the mouse embryo as well as with sperm (Artzt et al., 1973), but not with any other adult tissue, The tissue distribution and temporal expression of the antigen reacting with anti-F9 led Artzt et al. (1974) to examine whether this antigen was specified by the wild-type allele of one of the t mutants. Several elucidating experiments were all consistent with the view that the anti-F9 indeed recognized a wild-type allele product of the T locus. With use of the anti-F9 serum, this wild-type product of the T locus was isolated by Vitetta et al. (1975b). They labeled sperm surface antigens with radioactive iodine and a teratoma cell line with radioactive amino acid precursors, solubilized cell membrane molecules with NP-40, and purified the antigen by indirect immunoprecipitation. Analyses on SDS-polyacrylamide gels showed that the isolated components from sperm and teratoma cells were indistinguishable. The exciting observation was made that the antigen obviously is composed of two

146


types of subunits with apparent molecular weights of about 50,000 and 12,000. The small subunits are bound in the inta.ct molecule by noncovalent forces only, but the larger chains interact also via one (or more) disulfide bridge, Thus, this T-locus antigen displays a subunit structure and molecular weight characteristics which make it very similar to H-2 and TL antigens. Furthermore, like TL antigens, which have a reciprocal interaction with H-2D antigens on the cell surface, the wildtype allele product of the T-locus is expressed on cells (embryonic cells and germ cells) that do not carry H-2 alloantigens (Palm et al., 1971; Vitetta et al., 197513). Recent observations corroborate and extend these findings. Despite the lack of H-2 antigens on the blastocyst, /?,-microglobulin is present (Hikansson and Peterson, 1976). The same situation holds true for sperm. Preliminary studies of p,-microglobulin-associated polypeptide chains on sperm and blastocysts reveal striking resemblances with H-2, and TL-antigens (Peterson et nl., unpublished), It seems likely that at least some of these molecules will turn out to be coded for by the T locus. The various t-allele-specified products will undoubtedly provide an excellent basis for a genetic subdivision of the T locus, since these mutants may be recognized by selected antigenic specificities (see Artzt and Bennett, 1975). The wild-type allele products appear, however, to give rise to antibodies of less restricted specificity, since syngeneic antisera react with sperm from several species, including mail ( Buc-Caron et al., 1974; Jacob, 1975). Considering the particular nature of the syngeneic antisera it is astonishing that no cross-reactivity has been observed between the T-locus product and H-2 alloantigens in view of the reasonable assumption that the p,-microglobulin-binding structures of the two types of heavy most probably display common characteristics. Amino acid sequence determinations of the wild-type allele antigen of the T locus and H-2 antigens will undoubtedly establish the extent of this proposed similarity and give clues as to whether the two types of gene clusters ( T and H-2) are descendants from a common ancestral gene. VIII. I-Region Defined Antigens and the Fc Receptor

The finding that immunizations with lymphocytes over an isolated genetic disparity in the I region gave rise to cytotoxic antibodies (David et al., 1973; Hauptfeld et al., 1973; Sachs and Cone, 1973; Gotze et al., 1973; Hammerling et al., 1974) prompted investigations as to the properties of the molecules ( I a antigens) that displayed the antigenic sites. Cullen et al. ( 1974) labeled splenocyte macromolecules with radioactive leucine, solubilized cell membrane constituents by treatment with NP-40,


147

and isolated protein expressing the I-region-defined antigenic markers by immunoprecipitation with specific antisera. On SDS-polfacrylamide gel electrophoretic analyses the precipitates were shown to contain at least two molecular species with apparent molecular weights of about 60,000 and 30,000, respectively. On reduction of the 60,000-dalton component, the molecular weight was reduced to half, suggesting that it is composed of two similar polypeptide chains. By the same treatment of the 30,000dalton material it was demonstrated that it consists of a single chain. These and other data have made it highly likely that there are genes in the I region that code for polypeptide chains with apparent molecular weights of about 30,000 (Cullen et al., 1974; Cullen and Nathenson, 1974; McDevitt et al., 1974; Vitetta et al., 1974; Delovitch and McDevitt, 1975; David et al., 1975). In all these investigations, as well as in our own (Klareskog et al., 1976) heterodisperse patterns of the Ia antigens have been observed on SDS-polyacrylamide gel electrophoresis. One explanation for this may be that most antisera employed in these studies should theoretically ( and practically) recognize severd cytotoxically defined Ia specificities and it seems likely that at least part of the heterogeneity may be caused by genuinely different polypeptide chains. The heterogeneity is even more pronounced when doubly labeled Ia antigens are isolated. Cullen and Nathenson (1974), after having shown that Ia antigens are glycoproteins, labeled the molecules both in the amino acid and carbohydrate portions. Despite the fact that these workers used a combination of anti-Ia serum and splenocytes, which theoretically should precipitate molecules displaying but a single Ia-antigen specificity, heterogeneity in the isolated molecules was evident, suggesting the presence of glycoproteins differing in carbohydrate: amino acid ratios. In studies of recombinant strains carrying H-2 haplotypes, which arose by crossing-over within the I region, it has been possible to demonstrate unequivocally that most if not all of the Ia-antigen specificities can be arranged into at least three, possibly more, allelic series (see Shreffler and David, 1975). A single haplotype should accordingly be expected to express more than one antigenic specificity, which indeed has been documented at the serological level. This holds true even in biochemical analyses. Cullen et al. (1974) made the first observation that there could be more than one Ia molecule determined by a single haplotype. These findings were subsequently extended to several Ia specificities ( Cullen and Nathenson, 1974; McDevitt et al., 1974), and as yet no report has appeared demonstrating the presence of two antigenic determinants on the same Ia molecule. The ever-increasing genetic and serologic resolution of the I region has not yet been accompanied by the availability of antisera that have

148


allowed the isolation of seemingly homogeneous Ia antigens. Several explanations may be proposed to resolve this dilemma. Since Ia antigens are glycoproteins, containing at least fucose, mannose, galactose, and glucosamine ( Cullen and Nathenson, 1974 ) , posttranslational modification of a homogeneous polypeptide chain may give rise to the noted heterogeneity. It may also be possible that Ia antigens are highly susceptible to proteolysis, which is known to occur to some extent in solubilization with nonionic detergents. The heterodisperse pattern on SDSpolyacrylamide gel electrophoresis could, thus, result from partial fragmentation of a single molecule. Cullen et al. (1974) demonstrated that at least some Ia antigens are composed of two disulfide-linked chains. Accordingly, part of the heterogeneity may also be explained on the assumption that the chains, although similar, are nonidentical, The lack of homogeneity of the isolated Ia antigens may, however, suggest that available antisera, even those that appear monospecific serologically, recognize several I-region-defined gene products. There is already some precedence for this suggestion. Taussig and Munro have approached the question whether the I region codes for the antigen receptor on T lymphocytes. They have shown that mouse thymocytes, educated in uivo to a synthetic polypeptide antigen, under in uitro culture conditions release a soluble factor that is capable of restoring the thymus-dependent B-cell antibody response to that antigen in T-celldeprived mice (Taussig, 1974). The apparent molecular weight of this thymus-derived factor was estimated on Sephadex G-100 gel chromatography to be approximately 50,000. The factor may be absorbed out from a crude mixture by passage over a column containing covalently linked antibodies against Ia antigens (Munro et al., 1974). Available information suggests that the factor, which from genetic analysis appears to be coded for by the I region, does not carry any of the cytotoxically defined Ia-antigen specificities, yet Ia antisera contain antibodies directed against the factor. This result is in good agreement with the observation that Ia antigens, as measured by cytotoxic techniques, preferentially are detected on B cells, but corroborates studies demonstrating that I-regioncontrolled molecules may be expressed also on T lymphocytes (see Frelinger et al., 1974). Tada and his associates have examined an antigen-specific T-cell factor (Tada et al., 1975) that seems to share several features with the TaussigMunro factor. Thus, the molecule studied by Tada et al. ( 1975) displays a molecular weight in the 50,000 range and contains antigenic determinants apparently coded for by I-region genes. In contrast to the cooperating Taussig-Munro factor, the effect of the latter protein is quite the opposite, and it suppresses the help normally provided by T lymphocytes. The


149

suppressive factor has not yet been shown to carry any of the defined Iaantigen specificities, but tests to unambiguously exclude this property have not yet been reported. The supernatant of in vitro short-term, mixed-lymphocyte reactions between alloantigen-activated T cells and appropriate target cells contains a factor, produced by the T cells, that regulates triggering and differentation of B lymphocytes (Katz and Armeding, 1975). The factor has been shown to induce antibody production in B cells in vitro in the presence of antigen. In fact, it appears to replace the need for T-cell function (Armeding and Katz, 1974). There does not seem to be any antigen specificity requirement of the factor in contrast to the substance studied by Munroe and Taussig (1975). Since the factor is released from T cells on allogeneic stimulation, it appeared reasonable that it should display antigenic properties defined by the H-2 complex. Indeed, Armeding et al. (1974) could show that this allogeneic-effect factor (AEF) exhibits Ia-antigen determinants. Definitive data as to the expression of any given Ia-antigen specificity on AEF are not yet available, but since AEF is a T-cell product it is quite plausible that none of the known specificities will be associated with this molecule. Recently, Katz and his associates have presented some physical-chemical data for AEF. The factor is composed of two types of polypeptide chains with apparent molecular weights of about 40,000 and 12,000, respectively (Katz and Benacerraf, 1975). The minor subunit was recently identified as p,-microglobulin ( Armeding et al., 1975). It has not yet been rigorously established that it is the 40,000-dalton component of AEF that carries the Ia-antigenic determinants, but this may be inferred from previous studies where several laboratories have failed to demonstrate allotypes of p,-microglobulin. AEF represents the first I-region-coded molecule that is associated with P,-microglobulin. Most Ia antigens hitherto isolated are, however, derived from B-cells and the methods of their purification differ significantly from those used for AEF. It is, thus, much too early to exclude a more generalized association of p,-microglobulin with I-region products. Circumstantial evidence may suggest that there are other I-region defined cell surface molecules that contain p,-microglobulin. Several laboratories have reported that antibodies against p,-microglobulin inhibits the response in MLR (Bach et al., 1973; Solheim and Thorsby, 1974; Lindblom et al., 1974; McCalmon et al., 1975; Ostberg et al., 1 9 7 5 ~ ) . Ostberg and his colleagues have clearly shown that specific inhibition is effective only when antibodies or Fab’ fragments against p,-microglobulin bind to the responder cells ( Lindblom et al., 1974; Ostberg et al., 1 9 7 5 ~). It is thus possible that an I-region-coded receptor on T cells for I-regiondefined antigens on stimulator cells contains p,-microglobulin. Whether

150


this putative receptor has any resemblance to AEF or any of the other known factors remains to be established. Dickler and Sachs (1974) provided evidence that the Fc receptor on B lymphocytes expresses Ia-antigenic determinants. They and others showed that anti-Ia sera could abolish the binding of aggregated IgG or of antigen-antibody complexes to the Fc receptor (Dickler and Sachs, 1974; Basten et al., 1975). Out of a panel of antisera tested, this inhibition was noted only when the antisera had been raised over an I-region difference. All such antisera, however, may not block the accessibility of the Fc receptor for IgG, since Schirrmacher et al. (1975) were unable to corroborate the specificity of the inhibition. In a later study Dickler and his colleagues (1975) demonstrated that the Fc receptor carries other antigenic determinants, in addition to the Ia antigenic sites, which are not coded for by the H-2 complex, To resolve some of these discrepancies, Rask et aZ. (1975) isolated an Fc receptor from mouse splenocytes. The crude cell membrane macromolecules were liberated by treatment with an EDTA-containing buffer, and molecules with affinity for the FC portion of IgG were isolated on an affinity chromatography column containing covalently attached IgG. Usually Rask et aE. (1975) encountered a series of molecules with the apparent molecular weights 65,000, 18,000, and 15,000. Antisera were raised separately against the high molecular weight component and against a mixture of the smaller components. The latter antiserum cross-reacted with the 65,000-dalton component, and limited proteolysis of this component produced several distinct species with apparent molecular weights of about 45,000, 30,000, 18,000, and 15,000, From these data it was suggested that the 65,000-dalton component, on degradation by proteolysis, gave rise to the initially isolated, low molecular weight polypeptide fragments. However, solubilization of the Fc receptor by treatment of a crude membrane fraction with a nonionic detergent yielded two components that in preliminary analyses have the apparent molecular weights 50,000 and 25,000, respectively. Work in progress is aimed at elucidating the relationship between the various components. Although far from conclusive, available information may suggest that it is possible that on mouse splenocytes there are two distinct polypeptide chains, both similarly susceptible to limited proteolysis, which display affinity for the Fc portion of IgG. In preliminary investigations it has been documented that antisera against the various Fc-receptor fragments specifically abolish the binding of aggregated IgG to B-lymphocytes (Rask et aZ., 1975). With the use of such antisera it was also explored whether the Fc receptor carries Iaantigenic determinants. In spite of the fact that all Fc receptor on Blymphocytes could be redistributed to form polar caps with these antisera


151

subsequent analysis of the Ia-antigenic distribution showed that redistribution of the Fc receptor had not affected the Ia-antigens on the cell surface. Furthermore, isolated Fc-receptor had no inhibitory effect on anti-Ia serum-induced lymphocytotoxicity ( Rask et al., 1975). These experiments were performed with combinations of cells and alloantisera that theoretically should detect only a single Ia-antigenic specificity at a time. Therefore, it appears highly likely that the examined Fc receptor does not express any of the cytotoxically defined Ia-antigen specificities. The possibility that at least one Fc-receptor-binding polypeptide chain is coded for by the I region cannot be excluded on these grounds only. It is conceivable that an Fc-receptor may express I-region-defined antigenic determinants that usually fail to give rise to cytotoxic antibodies. Direct binding experiments with anti-Ia sera and isolated Fc-receptor will probably resolve this question. It is likely that I-region-defined molecules are engaged in MLR. Thus, binding of antibodies to the Fc receptor could conceivably impede the response of an allogenic MLR if it is borne out that this structure carries Ia-antigenic determinants. However, preliminary investigations with use of two rabbit antisera, raised against highly purified papain-solubilized mouse Fc receptor, do not give evidence for a role of the Fc receptor in MLR stimulation. Whether this result reflects only the particular specificity of the antisera employed, which may not be directed to the relevant part of the Fc-receptor molecule, or indicates that the Fc receptor is truly not involved in the MLR, remains to be elucidated. The known T-cell products which are coded for by the I region (see above) all seem to interact with other cells. The Taussig-Munro factor obviously triggers B lymphocytes as does AEF (Munro and Taussig, 1975; Katz and Benacerraf, 1975) and the suppressive T-cell factor probably interacts with T lymphocytes (Tada et al., 1975). The cooperating factor probably binds to B cells by a cell-surface receptor which itself also may be coded for by the I-region (Munro and Taussig, 1975). The chromosomal locations of the genes for the putative receptors for the other factors are unknown. It is interesting that the I region apparently codes for receptors that may recognize I-region-controlled antigens ( see Munro and Taussig, 1975). One may then ask if the Fc-receptor(s) under physiological conditions recognizes antigen-antibody complexes or if there exist other molecules with higher affinity for this receptor( s ) ? No such molecules have, of course, been identified but circumstantial evidence suggests that it may well be rewarding to keep this possibility in mind. L. Rask, L. Klareskog, and P. A. Peterson (unpublished) have obtained some support for the notion that intact IgG, (Fab’)2 and Fc fragments of IgG,

152


as well as p,-microglobulin, when separately attached to an insoluble matrix, bind Fc-receptorlike polypeptide chains about equally well, whereas several proteins, not involved in the immune system, display no measurable d n i t y for the receptor( s). From these observations it seems likely that the isolated Fc receptor(s) may not exhibit any high degree of specificity for the Fc part of IgG. In our opinion it is still open to question which is the specific ligand for the Fc receptor ( s ). IX. The S Region and the Complement System

The S region, defined by one or more genes for the Ss-serum protein and Slp allotype traits, has until recently received rather limited attention. All initial studies have been made by Shreffler and his associates (see Shreffler and Passmore, 1971). In preliminary investigations they examined some characteristics of the Ss-Slp proteins. On gel chromatography of mouse serum on Sephadex G-200 Shreffler and Passmore (1971) demonstrated that the Ss protein resolved into two well separated peaks. The earliest eluted component emerged from the column close to the void volume whereas the second peak appeared in a position indicating a size slightly larger than IgG. Both peaks contained material reactive with an anti& protein serum, but only protein in the position of the smaller component displayed Slp antigenic determinants. These data suggested that there might be a structural relationship between the Ss and Slp antigens. This could be shown to be the case by immunoprecipitations. Antiserum against the Ss antigen precipitated all Ss and Slp antigenic activity. However, antiserum against Slp determinants precipitated all Slp antigens but only a fraction of the ss protein (Shreffler and Passmore, 1971). Thus, a reasonable interpretation is that only some of the molecules displaying Ss antigenic determinants carry the Slp marker. The electrophoretic mobility of the Ss and Slp proteins is similar, and they migrate in the m2 to p position at pH 8.6 (Shreffler and Passmore, 1971) . Chaotropic ions will diminish the apparent molecular weight of the Ss-Slp proteins to about 75,000 from its original weight of about 150,000 or more. Both proteins exhibit similar stability toward extremes of pH, but the Ss antigenic determinants are retained at a higher temperature than that noted for the Slp determinants (Shreffler and Passmore, 1971). Capra et al. (1975) have described an isolation procedure for the Ss protein. They initially fractionated mouse serum by Sephadex G-200 gel chromatography. For further purification they choose the smaller of the two Ss-protein-containing peaks, which was pooled and subjected to


153

successive fractionation steps involving DEAE-Sephadex ion-exchange chromatography ( twice ) and Sephadex G-200 gel chromatography. By this procedure they obtained Ss protein, which seemed to be of high purity. The material was homogeneous on SDS-polyacrylamide gel electrophoresis and displayed an apparent molecular weight of about 120,000. Reduction of the protein, followed by separation on SDSpolyacrylamide gel electrophoresis, revealed the presence of four types of polypeptide chains with apparent molecular weights of 46,000, 35,000, 23,000, and 14,000, respectively. Occasionally, minor amounts of an 80,000-dalton species were also noted. Preliminary analyses suggested that the basic subunits were the two smaller ones, since under drastic conditions the 46,000- and 35,000-dalton components gave rise to the 23,000- and 14,000-daltonsubunits. L. Sandberg, B. Curman, and P. A. Peterson (unpublished) have also been able to isolate the Ss protein. Using the immunoassay for the Ss protein described by Shreffler and Owen (1963), they fractionated mouse plasma repeatedly on Sephadex G-200 gel chromatography (four times), followed by DEAE-Sephadex ion-exchange chromatography (twice), and preparative block electrophoresis. The isolated component was highly purified and antibodies raised against the protein reacted with a single protein in mouse serum. Mouse strains known to carry the Ssz genotype had only about one-tenth the serum level of the isolated protein compared to animals with the Ss” genotype. Furthermore, male animals had about twice the serum concentration of the purified protein as female mice of any given strain. The female serum level of the protein could, however, be increased to twice its normal level following testosterone administration. All these data strongly suggested that the protein isolated by Sandberg et al. was indeed identical with the Ss protein (see Shreffler and Passmore, 1971). Molecular weight determinations by a number of methods gave, on the average 140,000 for the isolated protein. In the presence of chaotropic ions, the weight was reduced to half, which again is strikingly similar to the results of Shreffler and Passmore (1971). On reduction and alkylation under denaturing conditions the molecular weight diminished further to 35,000. Thus, a single, apparently homogeneous polypeptide chain was detected on SDS-polyacrylamide gel electrophoresis and on gel chromatography in 6 M guanidine hydrochloride. Sandberg et al. compared the size on gel filtration of the isolated Ss protein and antigenically related material present in freshly drawn serum. It was consistently noted that the highly purified component was of a size slightly less than that of the minor serum peak containing Ss-antigen activity. Therefore, an alternative isolation scheme was devised since modifications of the Ss protein might have occurred during the previous,

154


rather tedious procedure, Specific antibodies against the highly purified Ss protein were covalently linked to an insoluble matrix and fresh plasma was passed over the column. Thus, highly purified Ss protein could be desorbed from the column (Curman et al., 1975). The Ss protein obtained by this procedure contained three types of polypeptide chains. The smallest one had a molecular weight of about 25,000 whereas the two larger chains occurred in elution positions corresponding to the 70,000-80,000 dalton molecular weight range on a Sepharose 6B column equilibrated with 6 M guanidine hydrochloride. In some cases several other Ss-reacting components were noted with apparent molecular weights of about 14,000, 17,000, 35,000, 45,000, 55,000, and 60,000, respectively. Although not yet consistent in detail, some data have been obtained that suggest that most if not all of these polypeptides represent fragments of the two larger chains, Some of these “fragments” are similar in size to those described by Capra et al. (1975). It should be pointed out that it is not yet possible to ascertain whether the Ss protein is very sensitive toward proteolysis or if, in fact, there are more plasma proteins than one controlled by the S region, However, available data may suggest that at least one protein regulated by the S region is composed of three types of dissimilar polypeptide chains ( Curman et al,, 1975). The two Ss-protein-containing peaks obtained on separating serum by gel chromatography are very similar immunologically. The larger component, which has a molecular weight of almost 1,000,000, may represent an aggregated form of the lower molecular weight component. It may, however, also represent a complex of the Ss protein with other, unrelated plasma proteins. This is rather likely considering the data by Dkmant et al. (1973) and Hansen et al. (1975) suggesting that the Ss protein may be involved in the complement cascade. A t least for human serum, Muller-Eberhard (1974) has shown that fluid phase C4,2,3 and C5,6,7,8,9 complexes may occur. Consequently, the known inhibition by EDTA of the initial events of the complement reaction is accompanied by a greatly diminished size of the high molecular weight Ss-containing component. Concomitantly, a notable increase in the relative amount of the lower molecular weight Ss protein was observed (Curman et al., 1975; Meo et al., 1975). A three-chain polypeptide structure for the Ss protein suggested that it might be similar to the human complement component 4. Curman et al. (1975) and Meo et al. (1975) examined the immunological cross-reactivity between the Ss protein and human C4. Antisera against human C4 reacted well with isolated Ss protein and vice versa. Lachmann et al. (1975) examined the role of the Ss protein in various complement assays. From that they concluded that the Ss protein is the murine equivalent of human C4. Curman et al.

THE MAJOR HLSTOCOMPATIBILITY COMPLEX

155

( 1975) and D. Capra (personal communication) corroborated this result in similar types of complement assays. Thus, consistent data from four laboratories strongly suggest that the Ss protein is complement component 4. In man it has been shown that a locus controlling C2-deficiency is closely linked to HL-A ( F u et al., 1974). Factor B of the alternate pathway of complement activation is also controlled by the human MHC region (Allen, 1974). In the latter case it has been ascertained that it is the structural locus for factor B, which is endowed in the MHC region ( Curman et al., 1976). Data with regard to the murine equivalents of C2 and factor B are not yet available. However, Ferreira and Nussenzweig (1975) have provided evidence which suggests that the C3 plasma level may in part be controlled by the H-2 complex. A further search for a relationship between other complement components and the MHC also in other species than man and mouse may be rewarding. This seems to be especially warranted for early-acting complement components, and it has already been shown that in addition to low C4 levels in mice of the Ss' genotype the activities of C1 and C2 are also lower than in Ssh genotype animals ( Goldman and Goldman, 1975). It is not known whether the Ss polymorphism is controlled by structural or regulatory genes. It is possible that only one of the Ss protein subunits is governed by the S region. This may hopefully be investigated by examining which of the chains carries the Slp allotype marker. Since this allotype is present on only part of the Ss molecules, it is conceivable that the S region encompasses two discrete Ss structural genes (Shreffler and Passmore, 1971), which, as for several other loci of the MHC, are isofunctional. The existence of several genes in the MHC regulating various complement components provides an additional link between humoral and cellular immunity mechanisms. X. Conclusions and Speculations

Our knowledge about various aspects of the major histocompatibility complex has expanded rapidly over the past few years. As yet the function( s ) of the gene products controlled by the MHC is largely unknown. A common denominator of in uiuo or in vitro reactions involving these glycoproteins is, however, their involvement in recognitive processes. It appears well established that the GvHR is triggered by cells interacting via possibly several MHC antigens. Furthermore, some I-region-defined products have already been implicated in cell communication events (Katz and Benacerraf, 1975; Munro and Taussig, 1975; Tada et al., 1975). In this context it also seems appropriate to examine the T locus, which

156


indeed is located on the same chromosome as the MHC. The various pseudoalleles of this locus, defined by the t mutants, most probably specify cell surface antigens which control precisely defined stages of the neuroectodernal differentiation ( see Sections IV and VII ) . The mode of action of these antigens is unknown, but it does not seem farfetched to suppose that they interact with cell-surface receptors of identical or complementary structure. In fact, it is already known that some of the MHC antigens may be recognized by complementary structures. H-2Kand D-region antigens recognized by T-killer cells certainly interact with receptors structurally distinct from the K and D antigens. This may represent an interaction brought about by the general antigen receptor on T lymphocytes, but since “killer” cells belong to that subpopulation of T cells which does not appear to encompass the “helper” cells (Cantor and Boyse, 1975a,b) the receptor for K and D antigens on “T-killer” cells may be quite specific. The Taussig-Munro antigen-specific, T-cell factor seems to be recognized by a B-cell surface molecule, specified by the I region (see Munro and Taussig, 1975). It is very likely that the B-cell receptor is different from the T-cell product since the two genes appear to be separable (Munro and Taussig, 1975). The existence of linked genes in the MHC region, which if generalized, conceivably could control the expression of ligand-receptor pairs, is an attractive hypothesis. This would be consistent with the apparent conservatism of the MHC region, which is observed even in distantly related species. Furthermore, the linked-gene concept would also contribute to explain the high degree of polymorphism and the linkage disequilibria noted in the system (see Bodmer, 1972). Several structural features of the MHC antigens are similar. Polypeptide chains coded for by the K, D, and TL regions are glycoproteins of similar size. At least one I-region product (Armeding et al., 1975) and one T-locus antigen (Vitetta et al., 197513) seem to share these properties. The association of all these polypeptide chains with p,-microglobulin suggests, furthermore, that at least those parts of the MHC and T antigens that are engaged in binding p2-microglobulin may be highly homologous. Thus, the known similarities for the MHC and T antigens are of a magnitude that makes it safe to predict that amino acid sequence determinations of the various proteins will reveal that they have had an interrelated evolution and that they most probably have arisen from a common ancestral gene. Whether this will turn out to be true for all MHC antigens remains doubtful, but the K, D, TL, and T antigens should probably be very similar. Unfortunately, it is not yet possible to know whether the structural similarity of these molecules is paralleled by their participation in func-


157

tional processes operative by similar modes of action. The T and TL antigens most probably fulfill their functions in differentiation processes. Since their cell surface expressions are reciprocal with regard to the H-2 antigens, one may speculate that the H-2 antgens also participate in similar, but permanent rather than transient, cell-to-cell recognition processes. If this is borne out, it might suggest that the MHC and T antigens are representative of a degenerate but selective cell differentiation system that may encompass also other loci located in the IX linkage group. The MHC region arose most probably at an early stage of evolution. Its ancestral gene may have duplicated and diversified to give rise to the genes located in the T, K, D, and TL regions. This set of genes, already governing various types of cell-to-cell recognitions, may have provided the embryo for the immune system (see Jerne, 1971; Bodmer, 1972; Gally and Edelman, 1973; Peterson et al., 1974, 1975a,b; Artzt and Bennett, 1975). The various immunological processes controlled by the MHC region may, thus, represent recent functions that have developed out of a preexisting recognition machinery. Moreover, some biochemical characteristics of the MHC antigens may be consistent with an evolutionary relationship between immunoglobulins and MHC antigens (Peterson et al., 1975a), and, if so, genetic material from a primitive 17th chromosome may have been translocated to other parts of the genome to ultimately give rise to the genes for the regular immunoglobulin polypeptide chains as well as for p,-microglobulin. The question whether the MHC region products and the regular immunoglobulins are evolutionarily related will, however, receive its unambiguous answer from amino acid sequence determinations, and available, indirect information is insufficient to fully elucidate the question.

ACKNOWLEDGMENTS We greatly appreciate the skillful secretarial assistance, cheerfully provided by Ms Christina Sjoholm. Work cited from the authors’ laboratory was supported by grants from the Swedish Cancer Society and the Swedish Medical Research Council.

REFERENCES Alexander, P. ( 1972). Nature (London) 235, 137. Allen, F. H. (1974). Vox Sang. 27, 173. Anundi, H., Rask, L., bstberg, L., and Peterson, P. A. (1975). Biochemistry 14, 5046. Anundi, A. H., Kvist, S., bstberg, L., and Peterson, P. A. (1976). In preparation. Armeding, D., and Katz, D. H. (1974). J . Exp. Med. 140, 19. Armeding, D., Sachs, D. H., and Katz, D. H. (1974). J . Exp. Med. 140, 1717.

158

PER A. PETERSON ET AL,

Armeding, D., Kubo, R. T., Grey, H. M., and Katz, D. H. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 4577. Artzt, K., and Bennett, D. ( 1975). Nature (London) 256, 545. Artzt, K., Dubois, P., Bennett, D., Condamine, H., Babinet, C., and Jacob, F. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 2988. Artzt, K., Bennett, D., and Jacob, F. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 811. Ashwell, G., and Morel], A. G. (1974). Ado. Enzymol. 41, 99. Bach, F. H., Widmer, M. B., Segall, M., Bach, M. L., and Klein, J. (1972a). Science 176, 1024. Bach, F. H., Widmer, M. B., Bach, M. L., and Klein, J. (197213). J . Exp. Med. 136, 1430. Bach, M. L., Huang, S. W., Hong, R., and Poulik, M. D. (1973). Science 182, 1350. Bain, B., Vas, M. R., and Lowenstein, L. (1963). Fed. Proc., Fed. Am. SOC. Exp. Biol. 22, 428. Basten, A,, Miller, J.F.A.P., and Abraham, R. (1975). J. Exp. Med. 141, 547. Behrens, N. H., Carminatti, H., Staneloni, R. J., Leloir, L. F., and Cantarella, A. I. ( 1973). Proc. Natl. Acad. Sci. U.S.A. 70, 3390. Benacerraf, B., and McDevitt, H. 0. (1972). Science 175, 273. Bennett, D. (1964). Science 144,263. Bennett, D., Goldberg, E., Dunn, L. C., and Boyse, E. A. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 2076. Berggird, I. (1964). Protides Biol. Fluids, Proc. Colloq. 12,285. Berggird, I. ( 1974). Biochem. Biophys. Res. Commun. 57, 1159. Berggird, I. ( 1975). Fed. Proc., Fed. Am. SOC. Exp. Biol. (in press). Berggird, I., and Bearn, A. G. (1968). J. Biol. Chem. 243, 4095. Bernier, G. M., and Conrad, M. E. ( 1969). Am. J . Physiol. 217, 1359. Bernier, G. M., and Fanger, M. W. ( 1972). J . Immunol. 109, 407. Binz, H., Kimura, A., and Wigzell, H. (1975). Scand. J. Immunol. 4, 413. Bodmer, W. D. ( 1972). Nature (London) 237, 139. Boyse, E. A., and Old, L. J. (1969). Annu. Rev. Genet. 3,269. Boyse, E . A., and Old, L. J. (1971). Transplantation 11,561. Boyse, E. A,, Old, L. J., and Stockert, E. (1966). Immunopathol., Int. Symp., 4th, 1965 p. 23. Boyse, E. A., Old, L. J., and Stockert, E. (1968). Proc. Natl. Acad. Sci. U.S.A. 60, 886. Boyse, E. A., Flaherty, L., Stockert, E., and Old, L. J. (1972). Transplantation 13, 431. Brown, L. J., Kato, K., Silver, J., and Nathenson, S. G. (1974). Biochemistry 13, 3174. Buc-Caron, M., Gachelin, G., Hofnung, M., and Jacob, F. ( 1974). Proc. Natl. Acad. Sci. U S A . 71, 1730. Cantor, H., and Boyse, E. A. (1975a). J . Exp. Med. 141,1376. Cantor, H., and Boyse, E. A. (1975b). J. Exp. Med. 141,1390. Capra, J. D., Vitetta, E. S., and Klein, J. (1975). J. Erp. Med. 142, 664. Cohen, S., Taylor, J. M., and Savage, R., Jr. (1974). Recent Prog. Horm. Res. 30, 533. Coutinho, A,, and Moller, G . ( 1975). Ado. Immunol. 21, 114. Cresswell, P., and Dawson, J. R. (1975). J . Immunol. 114, 523. Cresswell, P., Turner, M. J., and Strominger, J. L. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 1603. Cuatrecasas, P. ( 1974). Annu. Reo. Biochem. 43, 169.


159

Cullen, S. E., and Nathenson, S. G. (1971). J. Immunol. 107, 563. Cullen, S. E., and Nathenson, S. G. (1974). I n “The Immune System. Genes, Receptors, Signals” (E. E. Sercarz, A. R. Williamson, and C. F. Fox, eds.), p. 191. Academic Press, New York. Cullen, S. E., David, C. S., Shreffler, D. C., and Nathenson, S. G. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 648. Cunningham, B. A., and Berggird, I. (1974). Transplant. Rev. 21,3. Cunningham, B. A., Wang, J. L., Berggird, I., and Peterson, P. A. (1973). Biochemistry 12, 4811. Cunningham-Rundles, C., Jersild, C., Svejgaard, A., and Good, R. A. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 5081. Curman, B., Ostberg, L., Sandberg, L., Malmheden-Eriksson, I., Stklenheim, G., Rask, L., and Peterson, P. A. ( 1975). Nature (London) 258,243. Curman, B., Sandberg, L., and Peterson, P. A. (1976). In preparation. David, C. S., Shreffler, D. C., and Frelinger, J. A. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 2509. David, C. S., Cullen, S. E., and Murphy, D. B. (1975). J. Immunol. 114, 1205. Davies, D. A. L. (1966). Immunology 11, 115. Davies, D. A. L., Atkins, B. J., Boyse, E. A., Old, L. J., and Stocked, E. (1969). Immunology 16, 669. Dayhoff, M. O., ed. (1972). “Atlas of Protein Sequence and Structure,” Vol. 5, p. 182. Natl. Biomed. Res. Found., Washington, D.C. Delovitch, T., and McDevitt, H. (1975). lmmunogenetics 2, 39. DBmant, P., Capkovh, J., Hinzova, E., and Vorhcovh, B. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 863. Dkmant, P., Snell, G. D., Hess, M., Lemonnier, F., Neauport-Sautes, C., and Kourilsky, F. ( 1975). J. Immunogenet. (in press). Dickler, H. B., and Sachs, D. ( 1974). 1. Erp. Med. 140,779. Dickler, H. B., Cone, J. L., Kubiceck, M. T., and Sachs, D. H. (1975). J. Erp. Med. 142, 796. Dorrington, K. J., and Tanford, C. (1970). Ado. Immunol. 12, 333. Dunn, L. C. (1964). Science 144,260. Edelman, G. M., and Gall, E. (1969). Annu. Rev. Biochem. 38,415. Edelman, G. M., Cunningham, B. A., Gall, W. E., Gottlieb, P. D., Rutishauser, U., and Waxdal, M. J. (1969). Proc. Natl. Acad. Sci. U.S.A. 63,78 Edmundson, A. B., Ely, K. R., Girling, R. L., Abola, E. E., Schiffer, M., and Westholm, F. A. ( 1974). Prog. Immunol. 2, 103. Evrin, P. E., and Pertoft, H. (1973). 1. ImmunoE. 111, 1147. Evrin, P. E., Peterson, P. A., Wide, L., and Berggird, I. ( 1971). Scand. J . Clin. Lab. Invest. 28, 439. Ferreira, A., and Nussenzweig, V. (1975). J. E x p . Med. 140, 513. Fredriksson, A., and Peterson, P. A. (1975). Scand. I. Urol. Nephrol., Suppl. 26, 61. Frelinger, J. A. Neiderhuber, J., David, C. S., and Shreffler, D. C. (1974). J . Erp. Med. 140, 1273. Fu, S. M., Kunkel, H. G., Brusman, H. P., Allen, F. H., and Fotino, M. (1974). J . Ezp. Med. 140, 1464. Gally, J., and Edelman, G. M. (1973). Annu. Rev. Genet. 6, 1. Goldman, M. B., and Goldman, J. N. (1975). Fed. Proc., Fed. Am. SOC. Erp. B i d . 34, 979 (abstr. ).

160


Goodfellow, P. N., Jones, E. A., van Heyningen, V., Solomon, E., Bobrow, M., Migginano, V., and Bodmer, W. F. ( 1975). Nature (London) 254,267. Corer, P. A. (1936). Br. J . Exp. Pathol. 17, 42. Corer, P. A. (1937). J. Pathol. Baderiol. 44, 691. Corer, P. A., Lyman, S., and Snell, G. D. (1948). Proc. R. Soc. London, Ser. B 135, 499. Giitze, D. (1975). Zmmunogenetics 1, 495. Gotze, D., Reisfeld, R. A., and Klein, J. (1973). J. Exp. Med. 138, 1003. Grey, H. M., Kubo, R. T., Colon, S. M., Poulik, M. D., Cresswell, P., Springer, T., Turner, M., and Strominger, J. L. (1973). J . Exp. Med. 138, 1608. Hikansson, S., and Peterson, P. A. (1976). Transplantation 21, 358. Hammerberg, C., and Klein, J. ( 1975). Nature (London) 258, 296. Hammerling, G. J., and McDevitt, H. 0. (1974). J. lmmunol. 112, 1734. Hammerling, G. J., Deak, B. D., Mauve, G., Hammerling, U.,and McDevitt, H. 0. (1974). Immunogenetics 1, 68. Hansen, T. H., Shiu, H. S., and Shreffler, D. C. (1975). J . Erp. Msd. 141, 1216. Hauptfeld, V., Klein, D., and Klein, J. (1973). Transplant. Proc. 5, 1811. Heber-Katz, E., and Wilson, D. B. (1975). J . E r p . Med. 142, 928. Helenius, A,, and Simmons, K. ( 1974). Biochim. Biophys. A d a 415,29. Hess, M., and Smith, W. (1974). Eur. J . Biochem. 43,471. Hirschhorn, K., Bach, F. M., Kolodny, R. L., Firschein, I. L., and Hashem, N. ( 1963). Science 142, 1185. Hiitteroth, T. H., Cleve, H., Litwin, S. D., and Poulik, M. D. (1973). J . E r p . Med. 137, 838. Isenman, D., Painter, R. H., and Dorrington, K. J. (1975). Proc. Natl. A d . Sci. U.S.A. 72, 548. Jacob, F. ( 1975). Zn “Mammalian Early Development” (M. Balls and A. Wild, eds.). Cambridge Univ. Press, London and New York (in press). Jerne, N. K. ( 1971). Eur. 1. lmmunol. 1, 1. Kahan, B. D. and Reisfeld, R. A., eds. (1973). “Transplantation Antigens.” Academic Press, New York. Kandutsch, A. A., and Stimpfling, J. H. (1963). Transplantation 1, 201. Karlsson, F. A. ( 1974 ) , Zmmunochemistry 11, 11 1. Karlsson, F. A,, Peterson, P. A., and Bergglrd, I. (1972). 1. Biol Chem. 247, 1065. Katz, D. H., and Armeding, D. (1975). Proc. Leucocyte Cult. Conf., 9th, 1974 p. 727. Katz, D. H., and Benacerraf, B. (1974). In “The Immune System: Genes, Receptors, Signals” (E. E. Sercarz, A. R. Williamson, and C. F. Fox, eds.), p. 569. Academic Press, New York. Katz, D. H., and Benacerraf, B. (1975). Transplant. Reu. 22, 175. Kehoe, J. M., and Fougereau, M. ( 1969). Nature (London) 224,1212. Klareskog, L., Kvist, S., bstberg, L., and Peterson, P. A. (1976). In preparation. Klein, J. ( 1974 ), “Biology of the Mouse Histocompatibility-2 Complex.” Springer Verlag, Berlin and New York. Klein, J., and Park, J. M. (1973). 1. Exp. Med. 137, 1213. Klein, J., Hauptfeld, M., and Hauptfeld, V. (1974). Zmrnunogenetics 1, 45. Lachmann, P. J., Grennan, D., Martin, A,, and Xmant, P. (1975). Nature (London) 258, 242. Lilly, F. ( 1966). Genetics 53, 529. Lilly, F. (1972). J . Natl. Cancer Inst. 49,927. Lilly, F., and Pincus, T. ( 1973). Adu. Cancer Res. 17,231.


161

Lindblom, B., Ostberg, L., and Peterson, P. A. (1974). Tissue Antigens 4, 186. Livnat, S., Klein, J., and Bach, F. H. (1973). Nature (London),New Biol. 243, 42. McCalmon, R. T., Kubo, R. T., and Grey, H. M. (1975). J. Immunol. 114, 1766. McDevitt, H. O., and Benacerraf, B. (1969). Ado. Immunol. 11,31. McDevitt, H. O., Deak, B. D., Shreffler, D. C., Klein, J., Stimpfling, J. H., and Snell, G. D. (1972). J. Exp. Med. 135, 1259. McDevitt, H. O., Bechtol, K. B., Hammerling, G. J., Lonai, P., and Delovitch, T. L. (1974). I n “The Immune System. Genes, Receptors, Signals” (E. E. Sercarz, A. R. Williamson, and C. F. Fox, eds.), p. 597. Academic Press, New York. Meo, T., Vives, G., Miggiano, V. C., and Shreffler, D. C. (1973a). Transplant. Proc. 5, 377. Meo, T., Vives, G., Rijnbeck, A . M., Miggiano, V. C., Nabholz, M., and Shreffler, D. C. ( 1973b). Transplant. Proc. 5, 1339. Meo, T., Krasteff, T., and Shreffler, D. C. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 4536. Miller, D. A., Kouri, R. E., Dev, V. G., Grewal, M. S., Hutton, J. J., and Miller, 0. J. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 1530. Moller, G., ed. (1974). Transplant. Reo. 21, 1. Miiller-Eberhard, H. J. (1974). Prog. Immunol. ZZ, 1, 173. Munro, A. J., and Taussig, M. J. ( 1975). Nature (London) 256, 103. Munro, A. J., and Taussig, M. J., Campbell, R., Williams, H., and Lawson, Y. ( 1974). J. Exp. Med. 140, 1579. Muramatsu, T., and Nathenson, S. G. ( 1970a). Biochemistry 9,4875. Muramatsu, T., and Nathenson, S. G. (1970b). Biochem. Biophys. Res. Commun. 38, 1. Muramatsu, T., and Nathenson, S. G. (1971). Biochim. Biophys. Acta 241, 195. Muramatsu, .T., Nathenson, S. G., Boyse, E. A., and Old, L. J. (1973). I . Exp. Med. 137, 1256. Nakamuro, K., Tanigaki, N., and Pressman, D. (1973). Proc. Natl. Acad. Sci. U S A . 70, 2863. Nathenson, S. G. ( 1970). Annu. Reo. Genet. 4, 69. Nathenson, S. G., and Cullen, S. E. (1974). Biochim. Biophys. Acta 344, 1. Nathenson, S. G., and Davies, D. A. L. (1966). Proc. Natl. Acad. Sci. U.S.A. 56, 476. Nathenson, S. G., and Muramatsu, T. (1971). I n “Glycoproteins of Blood Cells” (G. A. Jamieson and T. J. Greenwalt, eds.), p. 254. Lippincott, Philadelphia, Pennsylvania. Natori, T., Tanigaki, N., Appella, E., and Pressman, D. (1975). Biochem. Biophys. Res. Commun. 65, 611. Neauport-SautCs, C., Lilly, F., Silvestre, D., and Kourilsky, F. M. (1973). 1. Exp. Med. 137, 511. Nicolas, J. F., Dubois, P., Jakob, H., Gaillard, J., and Jacob, F. (1975). Ann. Microbiol. (Paris) 126, 3. Nilsson, K., Evrin, P. E., Berggird, I., and PontCn, J. (1973). Nature (London) 244, 44. Nilsson, K., Evrin, P. E., and Welsh, K. I. (1974). Transplant. Reu. 21, 53. Old, L. J., Stockert, E., Boyse, E. A., and Kim, J. H. (1968). J. Exp. Med. 127, 523. Ostberg, L., Rask, L., Wigzell, H., and Peterson, P. A. (1975a). Nature (London) 253, 735. Ostberg, L., Rask, L., Nilsson, K., and Peterson, P. A. (1975b). Eur. 1. Immunol. 5, 462.

162


Ostberg, L., Lindblom, B., and Peterson, P. A. ( 1 9 7 5 ~ )Eur. . J. Immunol. 8, 108. Painter, R. H., Yasmeen, D., Assimeh, S. N., and Poulik, M. D. (1974). Immunol. Commun. 3, 19. Palm, J., Heyner, S., and Brinster, R. L. (1971). J. Exp. Med. 133, 1282. Pancake, S., and Nathenson, S. G. (1973). J . Immunol. 110, 1086. Parham, P., Humphreys, R. E., Turner, M. J., and Strominger, J. L. ( 1974). Proc. Natl. Acad. Sci. U.S.A. 71, 3998. Passmore, H. C., and Shreffler, D. C. (1970). Biochem. Genet. 4,351. Passmore, H. C., and Shreffler, D. C. (1971). Biochem. Genet. 5, 201. Peterson, P. A., Evrin, P. E., and BerggArd, 1. (1969). J. Clin. Inuest. 48, 1189. Peterson, P. A., Cunningham, B. A., Berggird, I., and Edelman, G. M. (1972). Proc. NatE. Acad. Sci. U.S.A. 69, 1697. Peterson, P. A., Rask, L., and Lindblom, J. B. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 35. Peterson, P. A,, Rask, L., Sege, K., Klareskog, L., Anundi, H., and Ostberg, L. (1975a). Proc. Natl. Acad. Sd.U.S.A. 74, 1612. Peterson P. A., Ostberg, L., and Rask, L. (197513). In “Second International Conference of Differentiation” ( N. Miiller-BBrat, ed. ), p. 529. Elsevier, Amsterdam. Pincus, J. H., and Gordon, R. 0. (1972). Transplantation 12, 509. Poljak, R. J., Amzel, L. M. Avey, H. P., Chen, B. L., Phizackerley, R. P., and Saul, F. ( 1973). Proc. Natl. Acad. Sci. U.S.A.70, 3305. Poulik, M. D. (1973). Immunol. Commun. 2, 403. Poulik, M. D. (1975). In “The Plasma Proteins” (F. W. Putnam, ed.), p. 433. Academic Press, New York. Poulik, M. D., Shinnick, E. S., and Smithies, 0.(1975). Fed. Proc., Fed. Am. SOC. Erp. Biol. 34, 4109 (abstr.). Putnam, F. W., ed. (1975). “The Plasma Proteins.” Academic Press, New York. Rask, L., Lindblom, B., and Peterson, P. A. (1974a). Nature (London) 249, 833. Rask, L., Ostberg, L., Lindblom, J. B., Fernstedt, Y., and Peterson, P. A. (197413). Transplant. Rev. 21, 85. Rask, L., Klareskog, L., Ostberg, L., and Peterson, P. A. (1975). Nature (London) 257, 231. Reisfeld, R. A., and Kahan, B. D. (1970). Ado. Immunol. 12, 117. Sachs, D. H., and Cone, J. L. (1973). J . Exp. Med. 138, 1289. Sanderson, A. R. ( 1964). Nature (London) 204, 250. Saunders, D. A., and Edidin, M. (1974). J . Immunol. 112,2210. Schirrmacher, V., Halloran, P., and David, C. S. (1975). J . E x p . Med. 141, 1201. Schwartz, B. D., and Nathenson, S. G. (1971a). 1. Immunol. 107, 1363. Schwartz, B. D., and Nathenson, S. G. (1971b). Transplant. Proc. 3, 180. Schwartz, B. D., Wickner, S.,Rajan, T. V., and Nathenson, S. G. (1973a). Transplant. Proc. 5, 439. Schwartz, B. D., Kato, K., Cullen, S. E., and Nathenson, S. G. (1973b). Biochemistry 12, 2157. Shimada, A., and Nathenson, S. G. (1967). Biochem. Biophys. Res. Commun. 29, 828. Shimada, A., and Nathenson, S. G. (1969). Biochemistry 8,4048. Shimada, A., and Nathenson, S. G. (1971). J. Immunol. 107, 1197. Shreffler, D. C. ( 1970). In “Blood and Tissue Antigens” (D. Aminoff, ed.), p. 85. Academic Press, New York. Shreffler, D. C., and David, C. S. (1975). Adu. Immunol. 20, 125.

THE M A J O R HISTOCOMPATIBILITY COMPLEX

163

Shreffler, D. C., and Klein, J. (1970). Transplant. Proc. 2, 5. Shreffler,D. C., and Owen, R. D. ( 1963). Genetics 48, 9. Shreffler, D. C., and Passmore, H. C. ( 1971). Immunogenet. H-2 Syst., Proc. Symp., 1970 p. 58. Shreffler, D. C., David, C. S., Passmore, H. C., and Klein, J. (1971). Transplant. Proc. 3, 176. Silver, J., and Hood, L. ( 1974). Nature (London) 249,764. Silver, J., and Hood, L. (1975). Nature (London) 256, 63. Singer, S. J., and Nicholson, G. L. (1972). Science 175, 720. Smithies, O., and Poulik, M. D. (1972a). Science 175,187. Smithies, O., and Poulik, M. D. (1972b). Proc. Natl. Acad. Sci. U.S.A. 69, 2914. Snary, D., Goodfellow, P., Bodmer, W. F., and Crumpton, M. J. (1975). Nature (London) 258, 240. Snell, G. D. ( 1958). J. Natl. Cancer Inst. 21, 843. Snell, G. D., Cherry, M., and Dhniant, P. (1971a). Transplant. Proc. 3, 183. Snell, G. D., Xmant, P., and Cherry, M. (1971b). Immunogenet. H-2 Syst., Proc. Symp., 1970 p. 2. Snell, G. D., Cherry, M., and S m a n t , P. (1973). Transplant. Rev. 15, 3. Solheim, B., and Thorsby, E. ( 1974). Tissue Antigens 4, 83. Stanton, T. H., Bennett, J. C., and Wolcott, M. (1975). J . Immunol. 115, 1013. Strominger, J. L., Cresswell, P., Grey, H., Humphreys, R. E., Mann, D., McCune, J., Parham, P., Robb, R., Sanderson, A. R., Springer, T. A., Terhorst, C., and Turner, M. J. (1974). Transplant. Reu. 21, 126. Tada, T., Taniguchi, M., and Takemori, T. (1975). Transplant. Reo. 26, 106. Tanigaki, N., Katagiri, M., Nakamuro, K., Kreiter, V., and Pressman, D. (1973). Fed. Proc., Fed. Am. Soc. Exp. Bid. 32, 1017 ( abstr. ). Taussig, M. ( 1974). Nature (London) 248,234. Vitetta, E. S., and Uhr, J. W. ( 1975a). Biochim. Biophys. Acta 415,253. Vitetta, E. S., and Uhr, J. W. (197513).J. Immunol. 115, 374. Vitetta, E. S., Uhr, J. W., and Boyse, E. A. (1972). Cell. Immunol. 4, 187. Vitetta, E. S., Klein, J., and Uhr, J. W. (1974). Immunogenetics 1, 82. Vitetta, E. S., Uhr, J. W., and Boyse, E. A. ( 1975a). J. Immunol. 114,252. Vitetta, E. S., Artzt, K., Bennett, D., Boyse, E. A,, and Jacob, F. (1975b). Proc. Natl. Acad. Sci. U S A . 72, 3215. von Boehmer, H., Hudson, L., and Sprent, J. ( 1975). J . Exp. Med. 142,989. Wibell, L. Evrin , P. E., and Berggird, I. ( 1973). Nephron 10,320. Widmer, M. B., Schendel, D. J., Bach, F. H., and Boyse, E. A. (1973). Transplant. Proc. 5, 1663. Wigzell, H. ( 1965). Transplantation 3, 423. Yanagisawa, K., Bennett, D., Boyse, E. A., and Dimeo, A. (1974). Immunogenetics 1, 57. Yasmeen, D., Ellerson, J. R., Dorrington, K. J., and Painter, R. M. (1973). J. Immunol. 110, 1706. Yu, A., and Cohen, E. P. (1974). 1. Immunol. 112, 1285.


CHROMOSOMAL ABNORMALITIES AND THEIR SPECIFICITY IN HUMAN NEOPLASMS: AN ASSESSMENT OF RECENT OBSERVATIONS BY BANDING TECHNIQUES Joachim Mark' Cytogenetic Laboratory, Department of Pathology, Central Hospital, Sktivde, Sweden

. .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I. Introduction . . . . . . . . . . . 11. Meningiomas . . . . . . . . . . . . . . . . A. Numerical Findings . . . . . . . . . . . . B. Karyotypic Findings by Conventional Staining Methods . . . C. Karyotypic Findings by Banding Methods . . . . . 111. Myeloproliferative Disorders. . . . . . . . . . . A. Phl-Positive Chronic Myeloid Leukemias . . . . . B. Ph'-Negative Chronic Myeloid Leukemias . . . . . . C. Chronic Myelomonocytic Leukemias . . . . . . D. Acute Myeloid Leukemias . . . . . . . . . E. Polycythemia Vera, Myelosclerosis, Essential Thrombocytosis, and Related Disorders . . . . . . . . . . . F. Miscellaneous Myeloproliferative Diseases . . . . . . IV. Lymphoproliferative Disorders . . . . . . . . A. Non-Burkitt Lymphomas. . . . . . . . . . . B. Burkitt Lymphomas . . . . . . . . . . . C. Myelomas . . . . . . . . . . . . . D. Louis-Bar Syndrome . . . . . , . . . . V. Concluding Remarks . . . . . . . . . . References. . . . . . . . . . . . . . . . Addendum . . . . . . . . . .

.

.

. . . . . . . .

165 166 166

168 171 174 174 183 185 185 191 194 196 197 205 208 210 212 215 222

I. Introduction

The innovation of chromosome banding methods (references in Yunis,

1974) has affected the entire field of cytogenetic tumor research. Because of the accuracy of these new methods, practically all previously studied neoplasms merit reexamination. Application of these techniques since their introduction has followed lines determined by methodological limitations. Therefore, interest has been focused upon such technically comparatively easy materials as neoplastic hematological diseases, malignant exudates, and cell lines. In this respect, the situation closely parallels what happened after the introduction of new methods in the early 1950s. Keeping in mind the fact that available data have been biased because of technical shortcomings, there exist several types or groups of neoplasms

' Present address: Department of Pathology, Central Hospital, 541 01 Skovde, Sweden. 165

166

JOACHIM MARK

or neoplastic conditions that have been investigated to such an extent as to enable interpretations and assessments. The disorders chosen for this purpose are meningiomas and some myeloproliferative and lymphoproliferative diseases. In addition, observations of a few other tumor types will be briefly summarized at the end of this chapter. The nomenclature used is that of the Paris Conference (1971). Other special terms used will be explained in the text. II. Meningiomas

Cytogenetically, the meningioma is currently the most thoroughly studied tumor type of all human solid neoplasms. The chromosomal findings in these benign tumors, mainly those obtained by conventional staining methods, were recently surveyed ( Mark, 1974). Since that survey a considerable amount of new data has become available, in particular data obtained by the new banding techniques. Owing to this, the remarkable specificity of the primary deviation in meningiomas, and also the specificity of the superimposed changes, there are reasons to summarize the available data from the period prior to banding studies and to give a survey of the results obtained by banding methods. The chromosomes have been studied in altogether 208 meningiomas, of which 142 belonged to a material investigated by a German group (Zankl et al., 1975a,b; Weiss et al., 1975; K. D. Zang, personal communications, 1975; references in Mark, 1974). Tumors T910, T1053, T1208, T1241, T1413, T1447, T1647, and T1753 of this material were not included in this review because either too few cells were studied or they lacked a definite stemline number and complete data for the stemline karyotype. Swedish material consisted of 63 tumors (Mark, 1969a, 1970b, 1973a,b,c, 1974; Mark et al., 1972a,b). Three additional meningiomas studied by an American group (Porter et al., 1969; Benedict et al., 1970; Paul et al., 1973) were included, bringing the total of meningiomas having complete, acceptable data to 200.

A. NUMERICAL FINDINGS The numerical findings in the 200 meningiomas are illustrated in Fig. 1, which shows the distribution of the stemline numbers. Of these, 60% were hypodiploid, 33%diploid, 4.5%hyperdiploid, and 2.5%hypotriploid. In the hypodiploid region the 45-chromosome stemlines outnumber the others. The complete absence of near-tetraploid and high-polyploid stemlines is another noticeable feature.

167

BANDING PATI’ERNS IN HUMAN NEOPLASMS

70

60

50

t 4c

-E, 0

.

c

31

n

5

z 2(

10

I

38

42

50

54

50

62

66

Chromosome number

FIG.1. The distribution of the stemline numbers in 200 meningiomas; filled areas = Swedish material; open areas = German material; hatched areas = American material.

The numerical pattern characterized above, and illustrated in the figure, shows that there is a pronounced tendency in meningiomas to lose chromosomes during their karyotypic evolution. A similar stemline pattern has not been observed in any other human tumor type that has been studied in detail. The observations in the German and the Swedish materials generally agree. The minor differences consist of a higher frequency of diploid stemlines and extreme hypodiploid stemlines occurring in the German material.

168

TOACHIM MARK

B. KARYOTYPIC FINDINGS BY CONVENTIONAL STAINING METHODS A total of 158 meningiomas have been studied by conventional staining methods. In this material 62 different stemline karyotypes were observed, but only two were outstanding in their frequency: one, characterized by monosomy G, which was found in 41 tumors, and another with a normal, diploid stemline, which was found in 46 neoplasms. In the latter group, however, the majority of the tumors contained either a sideline (i.e., an accessory, less frequent stemline) or variant cells with monosomy G. Conversely, many of the stemlines characterized by monosomy G had, in addition, a sideline or variant cells with a normal karyotype. These data were interpreted as unusually clear illustrations of the various discernible steps in typical, early karyotypic evolution in meningiomas: ( 1 ) Normal, diploid stemline ( S ) + ( 2 ) normal S variant cells with monosomy G + ( 3 ) normal S sideline ( s ) with monosomy G + ( 4 ) S with monosomy G normal, diploid s + ( 5 ) S with monosomy G variant cells with a normal karyotype + ( 6 ) S with monosomy G + ( 7 ) S with monosomy G and additional deviations. This step-by-step, clonal evolution demonstrates that a loss of one G chromosome, for unknown reasons, is advantageous during in uiuo conditions. During prolonged cultivation of meningioma cells, however, serial studies have shown a reverse population change, Thus, original heteroploid stemlines (mostly 45-chromosome stemlines with monosomy G ) were slowly overgrown by cells with a normal karyotype (Mark, 1974). These results demonstrate the importance of environmental conditions for the chromosomal evolution in meningiomas, and the potential risk of conclusions based solely on in uitro observations. In this context, it should be mentioned that it has been possible to obtain successful preparations from biopsies of a few meningiomas (Mark, 1970b). The findings in these cases agreed with observations in other tumors, which were all studied in preparations from short-term cultures. Stemlines with additional deviations or deviations other than monosomy G were found in 67 meningiomas. The distribution of the abnormalities on the various chromosome groups in these tumors is shown in Table I. The figures in the table, given as percentage, reveal the predominant involvement to be in groups G, D, and C. When the same calculations (not shown in the table) were made for those cases where the three types of A chromosomes were distinguished, and where El6 was separated from E17-18, it was found that the variation in these two groups affected mainly A1 and E17-18, respectively. All these data indicate a nonrandom pattern for the karyotypic evolution subsequent to the usual first steps, i.e., the development of a stemline with monosomy G.

+

+

+

+

BANDING PATTERNS IN HUMAN NEOPLASMS

INVOLVEMENT (%)

169

TABLE I VARIOUSCHROMOSOME GROUPSIN 67 STEMLINES ADDITIONAL CHANGES THAN MONOSOMY G

OF THE:

WITH OTHER OR

Chromosome groups

A

B

C

D

E

F

G

Markers

45

16

62

63

40

18

87

36

The high figures in the groups C and D are of special interest because these two groups were those usually involved in the few stemlines with no change in group G. Thus, it was suggested earlier (Mark, 1970b, 1974) that in rare cases the karyotypic evolution in meningiomas could originate with those steps that usually follow the primary G group change in the majority of the tumors. The involvement of the normal chromosome groups consisted of both gains and losses of chromosomes, but a loss predominated in all groups. This is an observation in agreement with the finding that the majority of the heteroploid stemlines were hypodiploid. Structural rearrangements accounted for another significant part of the variation. Thus, one or several marker chromosomes (the term is used for all structurally changed chromosomes, whether or not their origin is wholly or partly clarified) were included in the stemline karyotype of 24%of all meningiomas. Abnormally short or long acrocentrics were the most frequent morphological types, but also such rare markers as ring chromosomes (Fig. 2 ) and dicentrics (absent or seen only in a low frequency in other types of neoplasms) have been observed in some tumors. Among the markers whose origin could be traced with certainty, the majority were wholly or partly derived from the chromosomes in group G and the A1 chromosome. The most frequent structural change of the A1 chromosome was a short-arm deletion. A long-arm deletion in particular and deletions of both the short and the long arms were the two structural changes most prevalent in the G group chromosomes. Such anomalies and others of group G have been the only deviation from the normal in several tumors, and in a few additional neoplasms with a normal, diploid stemline there has been a sideline or variant cells with a changed G chromosome (Fig. 3 ) . These observations, and other findings discussed in detail earlier (Mark et al., 1972b; Mark, 1974), have raised the question whether a structural G chromosome deviation might be the initial, gross karyotypic change in

FIG.2. The karyotype of a Gbanded stemline cell in a meningioma with S = 42, XY, -1, -5, 9q+, -14, -18, -22, +l ring chromosome. ~ 2 8 0 0 From . Mark ( 1973a), with permission.


171

FIG. 3a, b, and c. The normal and the deleted (long-arm deletions) No. 22 in variant cells of three different meningiomas (M55,M59, and M62). G-band technique. ~2800. From Mark (1973b), with permission.

all meningiomas. If so, this structurally changed G chromosome has rapidly been lost through mitotic disturbances in most tumors. C. KARYOTYPICFINDINGS BY BANDINGMETHODS Soon after the introduction of the banding techniques, it was shown by Q-banding methods that the chromosome type involved in both the numerical and the structural, primary change in group G was consistently one No. 22 (Mark et al., 1972a,b; Zankl and Zang, 1972). These results have been confirmed by continued studies consisting of analyses by G(Figs. 2 and 3) and Q-banding methods (Mark, 1973a,b,c, 1974; Zankl et al., 1975a,b; Weiss et al., 1975; K. D. Zang, personal communications, 1975). It was also mentioned in the early reports (Mark et al., 1972a,b) that the affected No. 22 usually was completely lost from the tumor cells; i.e., there was no evidence of a regular translocation of the missing No. 22, or part of it, onto any other chromosome type. This conclusion has recently been corroborated by Zankl et al. ( 1975b). Currently banding data are available from 43 meningiomas, 27 of these belonging to the Swedish material and 16 to the German. The observations in the Swedish material (with exception of one hyperdiploid and two hypotriploid tumors) were recently presented in a survey whose aim was to clarify in particular the characteristics of the deviations often seen in addition to the involvement of No. 22 (Mark, 197313). In order to evaluate this pattern, all karyotyped cells in a certain tumor were used to calculate the mean frequency of involvement (numerical and/or structural) of each chromosome type in that neoplasm; then, the figures for all tumors were used to determine the same mean frequencies for the whole material. Table II-a somewhat modified version of these previous calculations-shows the mean values for those 25 Q- and G-banded meningiomas which had their stemline in the diploid-near-diploid region ( SU-

172

JOACHIM MARK

TABLE I1 MEANDISTRIIIUTION (%) OF CHROMOSOMAL DEVIATIONS DIFFERENT CHROMOSOME TYPES" Chromosome groups and numbers

Swedish material

ON

Swedish and German material

~

A B

C

D

E F G

1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Y X

Markers Number of tumors:

11.7l 0.8' 0.7' 1.3' 4.01 4.2' 4.38 15.417 11.00

4.22 4.5' 10.3' 4.1' 9.17 6.48 2.9' 4.46 5.04

14.4'0 0.5'

0.44 5.78

4.92 2.6' 7.56 9.317 14.012 9.96 7.66 6.34 5.06

10.4O 16.11' 4.2' 7.68 7.9'

4.78 5.4' 3.56 73.024 0.3' 1.32 24.612

8.28 4.57 81.0'1 20.14 5.74 25.416

25

41

5.30

eSuperior figures indicate the number of tumors with involvement of a certain chromosome type.

perior figures indicate the number of tumors with an involvement of a certain chromosome type). The high value for No. 22 and the consistency of this involvement is the most obvious feature in Table 11. Conversely, the other G chromosomes, i.e., No. 21 and the Y chromosome, were affected in only a small number of tumors and usually in a low frequency of their cells. A second conspicuous feature in the table is the preferential pattern in group C, namely the high value for No. 8 and the fairly high value for No. 9. An analogous situation prevails for No. 1 in group A and possibly also for


173

Nos. 14 and 15 in group D. The figures in the other normal chromosome groups, i.e., B, E, and F, are too low to permit conclusions. The values for the abnormal chromosomes, the group termed markers, reemphasize the prominence of structural rearrangements in the karyotypic evolution of meningiomas. A direct comparison with the 16 banded tumors belonging to the German material is difficult because data are regularly available only for stemline cells, not for sidelines or variant cells. When calculations, as those outlined above, are applied to such a material, they produce too high values for those chromosome types affected and too low values for the Others. With these reservations in mind the values in the German material have been added to those in my own material (third column in Table 11). As far as Nos. 21 and 22 in group G, markers, and group A are concerned, there are no significant changes; neither are there changes in groups By E, and F. In group C the involvement of No. 9 gains in importance, but No. 8 is still the chromosome type affected in most tumors. In group D there is a similar slight change, the values for No. 15 now being higher than those for No. 14. For the sex chromosomes the values are much higher, especially for the Y chromosome. These results are due to the fact that the German material included several cases selected ( according to analyses of conventional karyotypes, satellite associations, and also the frequency of Y- and Barr bodies) for banding just to show that deviations affecting the sex chromosomes may occur also in meningiomas (Zankl et al., 1975a). If stemlines with an involvement of the gonosomes were as frequent as suggested by the German authors, one should have expected at least some cases to have appeared in the Swedish material. Actually, not a single case was seen in a series of more than 25 consecutive tumors. Accordingly, definitive evaluation of the role of the gonosomes in the chromosome variation in meningiomas has to await studies of many more cases, preferably those of prospective, not retrospective, series. The available data from the banding analyses suggest that Nos. 1 and 22 are the chromosome types most frequently involved in the structural rearrangements. The morphological characteristics of the resulting marker types, however, vary from case to case, and so far, there are no indications that certain specific chromosome regions (“hot spots”) would be particularly liable to breakage. As shown in previous paragraphs the karyotypic evolution in meningiomas almost always takes place within the diploid region, and usually the hypodiploid zone. In rare cases, however, there is a development of near-triploid stemlines. The mechanisms for such an evolution is unknown. Data from two Q-banded hypotriploid tumors (Mark, 1973c; these cases

174

JOACHIM MARK

were not included in the calculations above because almost all chromosome types were involved in the variation), however, indicate that these mechanisms favor a progression toward a balanced triploid complement. Probably the most important reason for this development is an improved viability. In this context, it is of interest to note that some cases of the benign neoplasm mola hydatidosa show a triploid constitution (references in Mark, 1969a, 1973~). In summary, the results of the banding studies of meningiomas reveal a consistent involvement of No. 22 in group G, and, as to the superimposed changes, a nonrandom pattern in group A (No. l ) ,group C (Nos. 8:and 9), and group D (Nos. 14 and 15); structural changes were found to be fairly common, and they most frequently affected Nos. 1 and 22, but did not lead to consistent marker types. As to the nonrandom pattern for the superimposed changes, two explanations have been previously advanced ( Mark, 1973b): ( 1) These features are related to properties of chromosome organization and, accordingly, should be found in neoplasms of various types, ( 2 ) These features are peculiar to meningiomas and represent additional, less specific, expressions of influence of the oncogenic agent(s) on the hereditary apparatus. At that time, when practically no data were available for other human tumors, the latter explanation seemed most plausible. But, as will be shown in the sections to follow, this assumption now appears doubtful, and it seems most likely that both explanations are partly correct. 111. Myeloproliferative Disorders

The chronic myeloid leukemia (CML) is a natural starting point for a review of the cytogenetical findings in neoplastic-preneoplastic hematological disorders. This is because of the manifold connections between CML and other disorders of the group of myeloproliferative diseases, and also because of the great amount of well-documented chromosomal facts that have been accumulated about CML in the era prior to banding studies. These latter facts have been considered in detail in several recent reports (Sandberg and Hossfeld, 1970, 1974; Muldal and Lajtha, 1974; Stryckmans, 1974). Thus, all these results will not be iterated here, but some of the more important observations will be commented on.

A. PH'-POSITIVECHRONIC MYELOIDLEUKEMIAS The first result of banding studies of Phl-positive CML revealed that the Philadelphia chromosome actually represented a chromosome No. 22 with a deletion involving approximately the distal half of its long arm


175

( Caspersson et aZ., 1970; ORiordan et al., 1971). These investigators also clarified that Down's syndrome was characterized by trisomy 21, not trisomy 22. The well-known increased frequency of leukemia (as a rule ALL and AML) in Mongolism could thus no longer be ascribed to cytogenetic factors directly related to those in CML. The next step in the banding studies concerned the additional chromosomal changes often seen in connection with blastic transformation of CML. One particular rearrangement, the loss of one E l 7 or E l 8 and the gain of one metacentric marker of C group size, had previously been observed in a considerable number of cases (de Grouchy et d.,1968; Bauke, 1973; and others). Using G-banding technique, Lobb et al. (1972) studied 3 such cases and found the metacentric marker to correspond to an isochromosome for the long arm of No. 17, i.e., i( 17q). These results agreed with earlier conclusions about the origin of the metacentric marker, and later on, the findings were also supported by observations by the Q-banding technique reported by Rowley (1973a). At the same time Rowley demonstrated the occurrence of a second consistent chromosomal abnormality in Ph'-positive CML, namely one chromosome 9q+. The amount and the staining properties of the extra material located at the end of the long arm of No. 9 accorded well with the missing distal part (about 40%) of the long arm of the Ph' chromosome. Thus, the observations indicated a translocation of the missing part of one No. 22 onto one No. 9 (Fig. 4). Rowley's important discovery, t ( 9q+;22q- ), was soon confirmed by others (Petit and Cauchie, 1973; van den Berghe, 1973; Berger, 1973; Dinauer and Pierre, 1973; Whang-Peng et al., 1973, 1974; Raposa et al., 1974; and others), and it was also reaffirmed by extended studies by Rowley (1973b). Whang-Peng et al. ( 1974) made detailed banding studies of the Ph' chromosome and the chromosome 9q+ in 5 cases of

FIG.4. Partial karyotype showing the normal and the structurally changed Nos. 9 and 22 he., t(9q+; 22q-)] in a case of Ph'-positive CML; G-band technique. x2800. From J. Mark, unpublished observations, 1976.

176

JOACHIM MARK

CML; the authors concluded that the deleted segment of one No. 22 was, in fact, translocated onto one No. 9, and with regard to the breakpoints they suggested the formula t( 9;22) (q34;q12). However, in an extensive study comprising 42 cases of CML, Hayata et al. (1975) found the breakpoint in No. 22 to be somewhat more proximal, and they proposed the formula t( 9;22) (q34;qll). In this context, it should be mentioned that it is still an open question whether or not the translocation is reciprocal ( Mayall et al., 1974; Muldal et al., 1975). Soon after the publication of Rowley’s original report (1973a) some exceptions to the t(9q+;22q-) pattern were described. Hayata et al. (1973) first reported a case with a translocation onto the end of the long arm of one No. 2 instead of one No. 9. Then reports followed of single cases with translocation of the deleted part of No. 22 onto the long arm of one No. 19 (Gahrton et al., 1974b), the long arm of one No. 21 (Bottura and Coutinho, 1974), the long arm of the second No. 22 (Foerster et al., 1974), the short arm of one No. 13 (Hayata et al., 1975), the short arm of one No. 11 ( Muldal et al., 1975), and the deleted (loss of band q22) long arm of one No. 9 (Hayata et al., 1975). Even cases without any translocation were claimed to exist ( Mitelman, 1974b). In 10 additional cases with a deviating translocation pattern (peculiar to each case), the rearrangements were more complex and involved 3 or more chromosomes or at least 3 different chromosome arms (Engel et al., 1974; Ishihara et al., 1974; Potter et al., 1975; Nowell et al., 1975; Hayata et al., 1975). Thus, excluding the somewhat doubtful case described by Mitelman ( 1974b), altogether 17 different variant translocations are on record. The usual t(9q+;22q-) pattern has been found in all the other published cases, which, to this author’s knowledge, now amount to 158 (cf. below). According to these data the frequency of the variant cases is about 10% This figure is probably too high, since it seems likely that the material is biased in favor of exceptional cases. At present there are no indications that the clinical course in cases with variant translocations differs from that in cases with the typical pattern. In this context, it is of interest to note that a common denominator for all patterns of translocation is the preservation of the deleted segment of No. 22. This remarkable feature suggests that genes of importance for viability are located in this distal part of the long arm of No. 22. It is also possible that the translocation leads to some sort of a position effect influencing factors such as viability, growth control, growth rate, etc. Soon after their introduction, banding techniques were also used to clarify pathogenic characteristics of CML. Thus, using the Q-banding technique, Gahrton et al. (1973, 1974a) studied the heteromorphic region, i.e., the satellites, of chromosomes No. 22 in 8 CML-patients and


177

their parents. The results showed that the Ph' chromosome was as likely to be derived from the maternal as the paternal No. 22. In addition, the findings demonstrated that CML is an acquired disease. The latter fact has been clear also from earlier observations, which include among others studies of monozygotic twins where only one twin in each pair had CML (Jacobs et al., 1966; Goh et al., 1967; and others). Although Phl-positive CML is an acquired disease, there are several observations which indicate that overt clinically manifested leukemia may be preceded by a fairly long symptom-free period. Thus, some cases are on record in which only a portion of the marrow cells-years before apparent disease-contained the Ph' chromosome ( Canellos and WhangPeng, 1972; Baccarani et al., 1973; Verhest and van Schoubroeck, 1973). The observations by Gahrton et al. ( 1973, 1974a) also indicated a monoclonal origin of the Ph' chromosome, as did the G-banding studies by Hossfeld (1975) in a case of CML with heteromorphic chromosomes No. 9. Convincing evidence for a clonal evolution was earlier presented by Fialkow et al. (1967) from investigations of CML patients heterozygous for the X-linked enzyme G-6PD ( glucose-6-phosphate dehydrogenase ). Studies of CML patients with constitutional, gonosomal mosaicism (Fitzgerald et al., 1971; Moore et al., 1974) have strengthened this conclusion. At the time of diagnosis, or at some time during the course of the disease, a small group of male CML patients (estimated to about 4%of all Phi-positive cases of CML; Garson and Milligan, 1972) showed the marrow karyotype 45,X0,Ph1. Recently, Shiffman et al. (1974) summarized the findings in 22 such cases on record. To this series can be added 12 more cases, 1 reported by Motomura et al. (1973), 5 published by Lawler et al. (1974), and 2 each described by NoweII et al. (1975), Whittaker et al. (1975), and Hayata et al. ( 1975). According to Shiffman et al. (1974) the 45,X0,Ph1 karyotype is presumably developed by a stepby-step, clonal evolution, i.e. 46,XY + 46,XY, Ph' + 45,XO, Ph'. The loss of the Y chromosome does not protect the patients from blastic transformation. The survival time of these Y-negative cases, however, often seems to be longer, sometimes much longer (Nowell et d., 1975), than that of the other CML patients ( Whang-Peng et al., 1968; Ezdinli et al., 1970). This interpretation has been doubted by Lawler et al. ( 1974) and by Whittaker et al. (1975), but it has gained support from observations made by many other investigators (Pedersen, 1968; Garson and Milligan, 1972; Sandberg and Sakurai, 1973; Shiffman et al., 1974; Nowell et d.,1975; and others). A loss of the Y chromosome in the marrow cells of elderly malesusually aged 60 or over-has been found with considerable frequency

178

JOACHIM MARK

in several studies (O’Riordan et al., 1970; Walker, 1971; Pierre and Hoagland, 1972). This anomaly occurred in patients without any obvious disease, hematologic or otherwise, and it has been interpreted as an expression of natural aging processes. This explanation, however, cannot be immediately applied to the Y-negative CML patients because most of these men were less than 60 years old, The observations in healthy, elderly males and in Y-negative CML patients can nevertheless be related to each other. Thus, certain normal cellular mechanisms or processes, such as aging, might be to some extent retained in cells of some neoplasms; owing to factors such as intensified cell divisions or the action of neoplastic agent( s ) , the aging processes might then be triggered much earlier than in normal cells. The same explanation (Mark, 1974) was previously proposed for the similarities between some karyotypic deviations in meningiomas and those observed during the degenerative phase of cultures of normal human cells (Saksela and Moorhead, 1963). At present, there is sooner or later an inevitable transformation of almost all cases of CML to one of three different terminal patterns (Vallejos et al., 1974) : ( a ) blastic crisis (the hematological picture similar to that in AML), ( b ) a phase with increased resistance to all standard therapeutic agents (high white cell count and thrombocytopenia but no overt elevation of the blast forms in the marrow), or ( c ) proggressive myelofibrosis ( accompanied by anemia, thrombocytopenia, and a high white cell count). In most cytogenetic reports these patterns are not separated, and therefore the term blastic crisis or transformation is used for all of them in the discussion below. In approximately one-third of the cases of CML the Ph’ chromosome, and thus the translocation t( Qq+;22q- ) , is the only detectable change throughout the course of the disease (Stryckmans, 1974). In the remaining two-thirds of the cases the blastic transformation is associated with additional, superimposed changes ( Baserga and Castoldi, 1972; cf. also Table 111). In this latter group hyperdiploidy by far outnumbers hypodiploidy. It is also well known that among the chromosomes contributing to hyperdiploidy, those in groups 6-X-12, 19-20, and 21-22 predominate (de Grouchy et al., 1968; Berger, 1970; Crossen et al., 1971a; Pedersen, 1973, 1975; and others). As mentioned above the losses mostly affect the chromosomes E17-18. It is also an experience that a duplication of the Ph’ chromosome often heralds a blastic transformation. There seems to be an exception of special character to this rule. Thus, Whang-Peng et al. (1973) described 5 cases of CML with a dicentric Ph’ chromosome; this “duplication” of the Ph‘ chromosome did not produce any significant change in the mean survival or the prognosis. This is another indication that position effects may play an important role also in tumor cytogenetics.


179

Recently, Muldal et al. (1975) advanced the idea that blastic transformation, and also the common duplication of the Ph’ chromosome at this stage of the disease, is associated with a further structural change of the long arm of the Ph’ chromosome, namely, a loss, or “use up,” of the terminal band q12. Further detailed analyses are necessary to test this interesting hypothesis. Analyses employing banding methods have already upgraded understanding of the additional changes usually related to or preceding the blastic transformation. These deviations, and also those sometimes found already in the chronic phase, are characterized in Table 111. This table shows the absolute number of cases with numerical and/or structural involvement of the different chromosome types in 176 banded cases. Of these, 174 were collected from the literature available, the most informative reports being those of Rowley (1973a), Gahrton et al. (1974c), Hossfeld ( 1974b), Fleischman and Prigogina ( 1975), Prigogina and Fleischman ( 1975b), Nowell et al. ( 1975), and Hayata et al. ( 1975). One male and one female case (in the chronic stage and showing the “standard” translocation) from my own material were also included (J. Mark, unpublished observations, 1976). The calculations were made in relation to the basic karyotype 46, 22q-[not 46, t(9q+;22q-)] because of the occurrence of variant translocations; the normal No. 22 and the 22qwere separated in order to make particularly clear the involvement of the Ph’ chromosome. The high values for No. 9 (Table 111) are almost completely caused by the usual long-arm translocation, and they reflect the consistency of this structural deviation. As seen from the table, almost all the other chromosome types are involved in the further superimposed variation. However, 4 different chromosome types are preferentially affected, namely the Ph’ chromosome, No. 17, No. 8, and No. 19. The involvement of the Phl chromosome, especially in the later stages of the disease, mainly corresponds to a duplication, rarely triplication, of this marker; the remaining part of the variation is mostly related to the formation of a dicentric Ph’ chromosome. The involvement of No. 17 consists of the formation of an isochromosome for its long arm in approximately twothirds of the cases. The remaining part of the variation comprises cases with a gain or a loss of one No. 17 and a few cases with various deletions or translocations. The involvement of No. 8 and No. 19 consists almost entirely of cases with a gain of usually only one of these chromosomes. The figures discussed above clearly indicate a nonrandom pattern also for several of the superimposed changes in CML. The sequence for the appearance of these deviations, however, shows considerable variation from case to case. From their studies of 15 cases, Prigogina and Fleisch-

TABLE 111 DISTRIBUTION OF CHROMOSOMAL DEVIATIONS (NUMERICAL AND/OR STRUCTURAL) ON DIFFERENT CHROMOSOME TYPESIN 176 PHLPOSITIVE CML Chromosome types Stage

Sex

1

2

3

4

5

6

7 8

9

27 1 - 3 6 - - - 3 - 46 -

-

1 - - 2 1 1 1 3 - - 1 -------

___---------1 0 9 1 - - 3

1

1

5 -

-

20 1 2 1 - 1 - 3 3 6 4 2 1 1 3 1 2 - - - - - - -

1 1

1

2 4

2

1

6

6 47 2 8 72 4 2 48

2 2

1 - 2 1 4 3

4 16 167 6

4

2

1 3

-

3 13 2 - 1 0 1

-

1

5

2

_ _ -

31

3

40

5

-

1 - 9 - 2 1 - 15 - -

3 24 - 11

1

20 37

2

6 - 2 5 1 2

. . . ._ ._ ._ ._ . . . _. _. _. _. _. _. _. . _. -. - .4

1 1

0 b

1 - 2

115 7 5 1 I - - - -

_58_5_ 4_ 2- -1 _4 - 1- -3 - 1-9 --1 0 Chronic,ter- XX 2 - 1 minal,and X Y - 1 1 ? _ _ _ blastic Grandtotal: 2 1 2

Number of X cases

22 10 11 12 13 14 15 16 17 18 19 20 21 22 Ph' Y

1 1

9 - 2 18 5 -

2

27

- _- - -4

5

2

51 77 48

176

P

g


181

man (1975b) have suggested that one of the patterns of karyotypic evolution in CML is as follows: (1) t(9q+; 22q-) + ( 2 ) t(9q+;22q-), -17, f i ( 1 7 q ) + (3) t(9q-;22q-), -17, +i(17q), +8+ ( 4 ) t(9q+; 22q-), -17, +i( 17q), +8, +19. In addition, the authors suggested that a duplication of the Ph’ chromosome emerged at different stages of such a pattern (and others) of karyotypic evolution. To test this hypothesis (and its clinical implications) and to reveal other common evolutionary patterns many more patients must be studied, preferably by serial analyses throughout the course of the disease. The karyotypic observations in CML, discussed above, constitute an interesting parallel to previous chromosomal findings in some thoroughly studied animal tumors, namely, the Rous sarcomas in the mouse and in the rat ( Mark, 196913, 1970a; Mitelman, 1974a). A predetermined, stepby-step evolution was especially obvious in the rat sarcomas, and in both species the progression was simultaneous with an increasing dedifferentiation of the neoplasms. In both CML and the Rous sarcomas the determination of the karyotypic evolution is strongest for the first step in the chain, and in late stages the picture becomes less clear due to a variety of superimposed changes that are often peculiar to each case or tumor. If we assume that most human carcinomas have been studied in such a late phase, or possibly more advanced stages, it could explain why it has been difficult to detect consistent changes or even distinguish a preferential pattern of variation. The factors responsible for the blastic transformation of CML are inadequately understoood, but recently Hossfeld and Schmidt ( 1973) presented chromosomal data suggesting a primary role for the spleen, not only in the pathogenesis of CML, but also in the blastic transformation. The latter idea was reemphasized by Baccarani et al. (1974). They suggested that the spleen plays an active role in the transformation “either as a source of more malignant blast cells, or as an abnormal haemopoietic environment which offers a selective growth advantage to abnormal variants coming from outside the organ.” The cytogenetic basis for this hypothesis was the finding that the leukemic cells in the spleen, compared with those in the marrow, were more often pseudodiploid or aneuploid and more often contained 2 Ph’ chromosomes. In a later study by these investigators ( Zaccaria et al., 1975), such chromosomal deviations in the cells of the spleen were reported to occur more often than in the marrow at diagnosis, thus probably long before blastic crisis. Observations, essentially similar to those outlined above, were also described by Mitelman et al. (1974c, 1975a) in two cases subjected to serial chromosome analyses of bone marrow cells and spleen cells. In their last report the authors suggested a metastatic spread to the bone marrow of an abnormal clone

182

JOACHIM MARK

developed in the spleen, an assumption on line with ideas earlier proposed by Sandberg and Hossfeld (1974). Studies of 2 cases with socalled “blastic tumours,” by Hossfeld et ul. (1975a), have provided additional data with a bearing on this problem, Neither these latter observations nor those in the other studies, however, have given a decisive answer to the question of whether blastic transformation is an event initiated in the marrow or in extramedullary sites, particularly in the spleen, At present it seems to be obvious only that the extramedullary sites usually represent a more favorable environment for the growth of leukemic clones with superimposed chromosomal deviations. Although not conclusive, the cytogenetic observations summarized above automatically lead to the question of the effects of splenectomy on the course of CML and the incidence of blastic crisis. The observations in a clinical study of a series of CML patients, splenectomized comparatively earIy in the course of the disease (as soon as the first remission had been obtained by chemotherapy), were reported by Schwarzenberger et al. (1973). The results suggested that blastic transformation niight be postponed but not prevented by splenectomy. Thus, the incidence of blastic crisis was of the same order in the splenectomized group as in the reference group, i.e., 13/15 patients and 19/24 patients, respectively. Similar, but somewhat more promising, results were recently reported by Spiers et al. (1975) in a preliminary communication dealing with 26 splenectomized patients. The only way to prevent blastic crisis still seems to be an eradication of the Phi-positive cells. This has been possible in a few patients subjected to either only intense chemotherapy ( references in Stryckmans, 1974) or chemotherapy plus splenectomy (Bull, 1975). In the former group the return to a Ph’-negative state was usually neither complete nor permanent, but was often associated with a long survival. In the other group, consisting of 4 patients, all became Phi-negative. Only one of these patients, however, remained Ph’-negative. Owing to a short observation period, it is too soon to tell whether this state will last. Out of the other 3 patients, with a recurrence of the Ph’ chromosome, 2 have relapsed, and one of them is in clinical remission. Finally, it should be mentioned that the Ph’ chromosome (or at least a marker with similar morphological characteristics ) has been observed in one or a number of cases of most diseases belonging to the group of myeloproliferative disorders ( references in Sandberg and Hossfeld, 1970, 1974; Hossfeld et al., 1975b). Though questioned, these findings seem logical for several reasons: (1) The disorders of the myeloproliferative group are closely related and often change or merge into another. ( 2 ) The Ph’ anomaly is known to affect a stem or precursor cell (and thus

BANDING PATI'ERNS IN HUMAN NEOPLASMS

183

it is found in all the different marrow systems, i.e., the erythrocytic, the myelocytic and the megakaryocytic systems), but this does not necessarily imply that the transformed stem cell is always completely undifferentiated. ( 3 ) The phenotypic variation (of Phl-positive diseases) could in some cases be related to peculiarities of the genotype or of the immunological mechanisms of the host, simultaneous mutational events in the affected stem cell, and other inherited or acquired factors. The Ph' chromosome has also been described in some cases of lymphocytic-lymphoblastic leukemia types ( Propp and Lizzi, 1970; Schmidt et al., 1975; van Biervliet et al., 1975; Bloomfield et al., 1975). These observations, however, need to be confirmed before any assessments can be made.

B. PH'-NEGATIVE CHRONIC MYELOIDLEUKEMIAS According to the findings in several large series, about 10% of all patients with CML are Ph'-negative (references in Sandberg and Hossfeld, 1970, 1974). Clinically, the negative and positive groups differ in several respects. For example, the Phl-negative patients are predominantly males, they are older, have lower white blood cell and platelet counts, and have elevated serum and urine muramidase. Above all, the patients respond poorly to chemotherapy, the estimated mean survival time being 8-15 months as compared to 31-40 months for the positive group ( Whang-Peng et al., 1968; Ezdinli et al., 1970). Because of these and other clinical differences, some investigators have regarded Ph'negative CML as a different disease and have proposed other names for it, e.g., subacute myelocytic leukemia and acute leukemia of elderly (Ezdinli et al., 1970). Most of the Ph'-negative cases studied with conventional methods have shown seemingly normal marrow karyotypes (Cervenka and Koulischer, 1973; and others). Out of the 5 cases studied in blastic phase, one remained diploid and the other 4 displayed hyperdiploid karyotypes ( case-references in Sandberg and Hossfeld, 1974). Results with the new methods were first reported in 1974, when Rowley ( 1974b), using Q-banding technique, described her findings in 4 cases, and Hossfeld, using G-banding methods, described his observation in 3 cases. The karyotypic findings in these patients, and in 8 other cases hitherto reported in the literature, plus one case from my own material (J. Mark, unpublished observations, 1976), are listed in Table IV (together with a few clinical data supporting some of the facts discussed above). As seen from the table, about two-thirds of the cases showed a completely normal marrow karyotype. In these and the other cases, special attention was usually paid to the morphology of the chro-

184

JOACHIM MARK

THEK;tRYOTYPE

TABLE IV CLINICAL FINDINGS I N 16 PH~-NEQATIVE SIUDIED WITH BANDING METHODS

AND SOME

PATIENTS

Patients ~

Survival from Age at month of diagnosis diagnosis No. (years) (months) 58 40 72 73 ? ?

69 8 9 10 11

12 13 14 15 16

0

71 68 ? 72 76 335 ?

7 27 4 ? ?

Marrow karyotype 46,XY 46,XX 46,XX 46,XX 46,XY 47,XY, 20q--, +21 (60%) 46,XY, 20q--, (40%) 45,XO, -Y

? I

>22

?

4 ?

14 ? 12

1

? ?

72

3

46,XYa 47,XX, +13 46,? 46,XYn 46,XY 45,XO, - Y 46,? 47,?, +9" 46,XX"

References Rowley (1974b) ltowley (1974b) ltowley (1974b) ltowley (1974b) Hossfeld (1974a) Hossfeld (1974a) Hossfeld and Wendehorst (1974) Gahrton et al. (1974~) Hsu et al. (1974a) Ishihara et al. (1974) Mitelman el al. (1974b, 1975b) Mitelman et al. (1974b) Hays et al. (1975) Gahrton et al. (1975) Gahrton et al. (1975) J. Mark (unpublished observations, 1976)

Studied in blastic phase.

mosomes Nos. 9 and 22, but no abnormalities could be found, The autosoma1 changes, recorded in 3 cases, were different; the scanty data indicate an evolutionary pattern different from that in Phl-positive CML. The gonosomal changes, found in 2 cases, consisted of the loss of the Y chromosome. In the adult case (Hossfeld, 1974a; Hossfeld and Wendehorst, 1974), the unusually benign course was suggested to be due to this feature, analogous to ideas put forward earlier for Y-negative Ph'positive cases. In the child with juvenile CML (Hays et al., 1975), however, the clinical course was as malignant as that usually observed in other children with the same type of leukemia. This casts doubts on the idea that there is a direct beneficial effect of the loss of the Y chromosome.

BANDING PATTERNS I N HUMAN NEOPLASMS

185

C. CHRONIC MYELOMONOCYTIC LEUKEMIAS The chronic myelomonocytic leukemia ( CMML ) is predominantly found in elderly males. This disease usually starts with weakness and long-standing anemia. Either in the beginning or at some time during the course, monocytosis and often thrombocytopenia, granulocytopenia, or granulocytosis are added to the symptoms. Death from intercurrent infections or the terminal transition to acute leukemia is common. Six patients with CMML were studied cytogenetically by Hurdle et al. (1972). Three cases showed a normal marrow karyotype, and 3 patients had a 45,XO,-Y karyotype ( confirmed by fluorescence methods). Marrow cultures from one of the latter cases were also studied by Moore and Metcalf (1973), and the in vitro analyses showed the same karyotypic abnormality. The Y chromosome was missing in all cells in the 3 cases reported by Hurdle et al. (1972). This is a difference in comparison with the varying percentage of cells with Y chromosome loss, found in the marrow of normal males aged 60 or more ( ORiordan et al., 1970; Pierre and Hoagland, 1972). Again (cf. CML above and the acute leukemias below), there are observations suggesting a relationship between loss of sex chromosomes and leukemias. Studies of female cases with CMML, and detailed studies of cases with a transition to acute leukemia, will be of interest to reveal the cytogenetic characteristics of this debated disease (Dameshek and Gunz, 1964). D. ACUTEMYELOID LEUKEMIAS The first analyses of the chromosomes in acute myeloid leukemias (AML) were carried out by Ford et a2. (1958). Since then more than 600 cases have been studied. These extensive investigations, performed with conventional staining methods, have clarified some basic cytogenetic characteristics of this type or, as proposed by Flandrin and Bernard (1975), “these types” of leukemia. Thus, according to a recent survey by Sandberg and Hossfeld (1974), nearly 60%of the cases have shown a seemingly normal diploid stemline; almost without exception, the remaining cases have had their stemline in the diploid-near-diploid region. The karyotype analyses of pseudodiploid and heteroploid stemlines have not revealed any consistent numerical or structural abnormality. It has often been noticed, however, that chromosomes in groups C and G, and possibly also D and E, are most commonly involved in the deviations. Especially Trujillo et al. ( 1974 ) have also emphasized the comparatively high frequency of abnormal stemlines with the complex karyotype 46,-C,+D,+E,-G.

186

JOACHIM MARK

Sandberg et al. (1973) and Sakurai and Sandberg (1974) have pointed out certain prognostic implications of the karyotypic findings in acute myeloid leukemias. They found that the median survival (after initiation of antileukemic therapy) of patients with only abnormal karyotypes in their marrow was significantly shorter than that of those with either only normal or a mixture of normal and abnormal karyotypes (except for patients over 70 years of age in both the last-mentioned groups). The poor prognosis of patients with only abnormal karyotypes in their marrow was presumed to be due to the absence of normal cells that could repopulate the marrow after treatment. Whether this reasoning holds true, or the poor prognosis is merely a reflection of a more advanced stage in the progression, remains to be clarified. Before discussing results obtained with the new methods, some terms and abbreviations for these will be introduced: AML = acute myelocytic and myeloblastic leukemia; AMML = acute myelomonocytic and myelomonoblastic leukemia; APL = acute promyelocytic leukemia; EL = erythroleukemia. AML, AMML, and APL are closely related types of acute leukemias but also the various subgroups of E L are nowadays most often regarded as variants of AML, and they usually transform into AML during the course of the disease. Sandberg and Hossfeld (1974) recently summarized the cytogenetic observations in 64 cases of E L reported in the literature; all of these were studied with conventional methods. The findings were similar to those in AML except for a more common occurrence of hypodiploid stemlines and a higher frequency of structural abnormalities leading to the formation of such fairly uncommon marker types as ring chromosomes and dicentrics. Below, the available banding data for all these types of acute leukemias will be considered together, but for completeness, and to allow comparisons, the observations in each type are separated in the tables. Investigations with banding techniques have been rewarding. These studies were initiated by Rowley (1973d). In 2 female patients with AML she found the same stemline karyotype, viz. 45,XO,-X,t(8;21) ( q22;q22), Soon afterward Sakurai et al. ( 1974) reported two additional cases of AML with apparently the same structural rearrangement: one female with the karyotype 46,XX,t(8;21) (q22;q22) and one male with the karyotype 45,XO,-Y,t(8.21) ( q22;q22), This highly suggestive picture, however, has been substantially modified by the continued studies, which now encompass more than 90 cases. The detailed karyotypic findings in all available cases of AML, AMML, APL, and E L were recently summarized by Rowley and Potter (1976). The survey included 19 cases collected from the literature, 18 cases ob-

BANDING PAmERNS IN HUMAN NEOPLASMS

187

tained by personal communications, and 50 cases belonging to the authors’ own series (some of them published earlier), The data of their compilation will be used here with the following modifications: the case reported by Kohn et al. (1975), with a de novo appearance of a Ph’ chromosome in a marrow with monosomy 6, is excluded owing to the uncertainty as to the real nature of the disease; 2 of the authors’ cases are excluded because of insufficient banding data; 4 female and 1 male cases of AML with a normal marrow karyotype reported by Mitelman and Brandt, 1974) are added. The resulting total material comprises 89 cases, 53 AML, 26 AMML, 2 APL, and 8 EL. The karyotypic findings in this material are characterized in Table V, which demonstrates the distribution of numerical and/or structural deviations on the different normal chromosome types. In the next to the last column of Table V, it is shown that 30 (16 13 1) of the 89 cases had a normal, diploid stemline. This figure is almost certainly too low because of the tendency to report only cases with abnormal karyotypes. A frequency of about 50%,as found by Rowley and Potter (1976) in their series of 48 cases, is probably more correct. No fewer than 39 different karyotypes were found among the 59 cases with abnormalities. As seen from Table V, these deviations affect almost all chromosome types. There is, however, a preferential involvement of two chromosome pairs, namely No. 8 and No. 21. The involvement of No. 8 consisted of a gain of one such chromosome in 13 of the 22 cases; the remaining part of the variation was mainly related to 8/21-81 17 translocations. The involvement of No. 21 was due to a gain or a loss of one chromosome in 7 of the 14 cases; the other part of the variation was completely related to 8/21, 21/21, and 17/21 translocations, in particular the 8/21 translocations, Chromosome No. 7 was also frequently affected, and the involvement consisted of a loss of one No. 7 in 9 cases and a short-arm deletion [de17(p13-14)] in the remaining 2 cases. The fairly common occurrence of monosomy 7 in the acute myeloid leukemias is further corroborated by a recent brief report of 2 additional cases of AML with this deviation (Gahrton et al., 1975; Zech et al., 1975). The involvement of Nos. 9 and 17 comprised a variety of numerical and structural changes; among the latter, however, several identical rearrangements were found in different cases (see below). The variation affecting the gonosomes, finally, was wholly related to the loss of one X chromosome in the female cases and to the loss of the Y chromosome in the male cases. A second indication of nonrandom patterns in the acute leukemias was the occurrence of the same stemline karyotype in 2 or more patients. Eight such karyotypes are listed in Table VI. As seen from this table,

+

+

TABLE V THEDISTRIBUTION OF DEVIATIONS (NUMERICAL AND/OR STRUCTURAL) ON DIFFERENT CHROMOSOME TYPES IN 89 ACUTENONLYMPHOCYTIC LEUKEMIAS

Type of leukemia AML AMML APL Total: EL

Chromosome types 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Y X

4 1 1 2 4 4 6165 1 4 12---3 1 1 2 - 1 - 2 3 2 - 1 - 1 - - 1 2

2 1-10-2 1--3-3

4 1

Abnormal karyotype, number of females/ males

Normal karyotype, number of females/ males

18/20 4/8 1/1 23/29

10/5 6/8 16/13

53 26 2 81

2/5

1/0

8

-

Total number of cases

cc

8

E z

E

189


the karyotype with an extra No. 8 predominated. It is of considerable interest, however, that also more complex karyotypes and some specific structural rearrangements, in particular the 8/21 translocation, were found more than once. The 8/21 translocations might actually be more important than is clear from Table VI. Thus, Sakurai et al. (1974)) when (which reconsidering old data, found the karyotype -C,+D,+E,-G could be expained by a 8/21 translocation) or closely related karyotypes in 9 of more than 110 patients belonging to their own series; Rowley and Potter (1976) also mentioned that this same karyotype (-C,+D,+E, -G) was found in 12 of 69 patients reported by Trujillo et al. (1974). The frequent loss of sex chromosomes together with 8/21 translocations (Table VI) is a noticeable feature. Additional studies are necessary to clarify whether there is a relationship between these two changes or their association is merely due to coincidence. In support of the latter idea, Rowley and Potter (1976) described a female patient, with a normal karyotype initially, who developed a 8/21 translocation (of the above-mentioned type) during the course of the disease but no accompanying loss of an X chromosome, A third indication of nonrandom patterns was the karyotypic observations in cases with clonal evolution. Convincing evidence of the occurrence of such an evolution also in acute leukemias was recently presented by Rowley and Potter (1976) and by Golomb et al. (1976). These investigators reported the findings in 8 cases, 5 having a normal karyotype and 3 an abnormal karyotype initially. As many as 5 of the 6 patients with an additional chromosome showed a gain of one No. 8 during progression; a seventh patient developed an 8/21 translocation (cf. previous paragraph). In 3 cases there was a gain of one No. 18, and 2 TABLE VI RECURRENT KARYOTYPES I N THE ACUTEMYELOID LEUKEMIAS Number of cases Karyotype

AML

AMML

APL

EL

45,XO) - Y 45,XX or XY, -7 4.5,XO)-X, t(8;21)(q22;q22) 45,XO, - Y, t(8;21)(q22;q22) 46,XX, t(6;9)(p23;q34) 46,XX or XY, de1(17)(qllq21) 46,XX or XY, i(17q) 47,XX or XY, +8

-

2 -

-

2 1

4 2 2 2 2 9

-

2 -

190

JOACHIM MARK

of these patients, in addition, had an extra No. 21. In at least one case with a normal karyotype initially, the extra No. 8 was the first change from the normal. All these data agree with the other observations in the acute leukemias and the importance of a supernumerary No. 8 is reemphasized. The findings with banding methods in the various forms of acute leukemias are listed separately in Tables V and VI. Excepting APL (where only 2 cases have been studied), there is at present little to suggest that the different forms are distinguished by karyotypic deviations peculiar to each type. In spite of this, there are already results strongly indicating important clinical implications of the chromosomal findings in the various types of acute myeloid leukemias. Thus, Golomb, together with Vardiman and Rowley (1976) have recently presented data concerning the clinical course of the patients belonging to the abovementioned series of 50 acute nonlymphocytic leukemias (Rowley and Potter, 1976). The authors found that the patients with normal chromosomes initially (group I ) had a median survival of 10 months, whereas those with abnormal chromosomes initially (group 11) had a median survival of only 2 months. Complete remission was induced in 11 of 21 treated group I patients (median complete remission 18 months) and in 6 of 22 treated group I1 patients (median complete remission 10 months ) . However, the 10 treated group I1 patients, who had only abnormal metaphases in their marrow samples, had a poorer response (10%complete remission) and a shorter median survival ( 2 months) than the group I1 patients, who had at least one normal metaphase (42%complete remission and a median survival of 9 months). There was no difference in either the complete remission rate or the median survival between the patients with hypodiploid, pseudodiploid, and hyperdiploid stemline karyotypes, respectively. Treated patients with AML and AMML had a similar median survival time. However, when the AML patients were subdivided in those with and those without a chromosome abnormality, the median survival times were clearly different, i.e., 2 months versus 18 months, respectively. (In addition, the AML patients with normal chromosomes showed a complete remission rate of 81.5%and those with abnormal chromosomes only 20%) For unknown reasons, the pntients with AMML did not show a similar difference in median survival times (as the AML cases) when subdivided into groups I and 11. The 2 cases of APL (both having an apparently identical long-arm deletion of one No, 17) had similar clinical histories with disseminated intravascular coagulation; none of them responded to therapy. The picture in EL cases remains t o be clarified. The cited clinicopathological correlations of data from chromosome

BANDING PAlTERNS IN HUMAN NEOPLASMS

191

banding holds promise for the future classification and handling of the acute myeloid leukemias. These results also support the ideas earlier advanced by Sandberg et al. (1973) and mentioned in the beginning of this section. The karyotypic findings in the acute leukemias represent to a considerable extent a parallel to the observations in Ph’-positive CML. In both cases there is an obvious nonrandom involvement of several chromosome types, and the sequence of the superimposed changes seems to vary from case to case, perhaps in both cases owing to the existence of many different evolutionary patterns. One important difference, however, is the absence of a primary unifying change, like the Ph* chromosome [or rather t ( 9q+;22q- ) ] in the acute leukemias. If such a basic primary change really exists, the banding results show that it must be extremely small or rather it would correspond to one or several point mutations. The characteristics of the nonrandom karyotypic patterns in the acute leukemias, and also Phl-positive CML, are of considerable interest in view of recent observations in other myeloproliferative disorders and some types of refractory anemias. The latter two groups of diseases have been linked closer to the leukemias by a recent investigation which provided direct evidence for an involvement of erythropoietic precursors in the leukemic process in AML, too ( Blackstock and Garson, 1974).

THROMBOCYTOSIS, E. POLYCYTHEMIA VERA,MYELOSCLEROSIS, ESSENTIAL AND RELATED DISORDERS The present group of myeloproliferative diseases includes polycythemia Vera ( PV ) , myelosclerosis ( M ), essential thrombocytosis ( ET ), and some allied disorders of ill-defined types. These diseases are all closely related and they may merge or change into another or into AML, CML, or other more special types of leukemias. Apparently, there are good reasons to regard these disorders as preneoplastic conditions or variants of a basically neoplastic ( leukemic) process. Reports describing a Phl-like chromosome in some cases of PV, M, ET, and related conditions are in keeping with this idea (references in Sandberg and Hossfeld, 1974; Hossfeld et al., 197513; see also Phl-positive CML). Sandberg and Hossfeld ( 1974) have recently surveyed the cytogenetic observations obtained by conventional staining methods in 135 cases of PV (the majority were treated cases), 65 cases of M, and 11 cases of ET. A normal picture occurred in more than two-thirds of the cases of PV and ET and in more than half of the cases of M (Phl-positive cases excluded). Gains and/ or losses of C group chromosomes were the deviations encountered most frequently in M and ET. In PV there was also a

192

JOACHIM MARK

common involvement of group C, especially in untreated cases; in addition, however, a deleted F group chromosome was seen in several cases in different series, but as a rule only in those cases treated with chemotherapy and/or P3* (Lawler et al., 1970; Visfeldt, 1971; Shiraishi et aZ., 1975). Observations with banding methods were first presented by Reeves et al. (1972) in 4 cases of PV. These results and those in 13 additional cases of PV are listed in Table VII together with the findings in 2 cases of M, 1 case of ET and 5 cases of ill-defined myeloproliferative disorders. The data, in spite of their scarcity, strongly indicate a nonrandom involvement of the C group chromosomes. Thus, a gain of one or two Nos. 8 andlor one No. 9 was found in 7 of the 17 cases of PV, in all cases of M and ET and in 3 of the 5 cases of the ill-defined type. This involvement is cytogenetical support for the close relation between the disorders of the present group. As all the diseases can change into acute leukemias, it is most interesting to note that not only No. 8 but also No. 9 were frequently involved in the acute myeloid leukemias (cf. Table V). A change into AML was actually described for the case with a 51-chromosome stemline reported by Rowley ( 1975a). The conclusions above are strengthened by a brief report by Gahrton et (1.1. (1975). The authors studied 10 cases of PV and found supernumerary Nos. 9, 21, and 8 in at least some of the cells in 4, 2, and 1 cases, respectively. Among the cases of PV listed in Table VII, there is a deviating group, which usually does not show an involvement of No. 8 and/or No. 9 but, instead, a variety of often complex structural rearrangements. Out of these, 3 types deserve notice: (1) 5 cases with a deleted No. 20, probably d e l ( 2 0 ) ( q l l ) in all cases; ( 2 ) 2 cases with a deleted No. 7, i.e. de1(7)(q22); and ( 3 ) 2 cases with a 1/15 translocation, probably t(lq;l5q) in both of them. The reappearance of the same structural change in 2 or more cases in 3 different instances is a remarkable feature. As all cases were treated, it is quite possible that the therapy, in particular the treatment with 32P,is related, in some way or another, to the appearance of these structural anomalies. Their specificity and nonrandom character, however, can hardly be explained by the therapy given, unless there exist certain “weak points” (“hot spots”) in the complement. This question remains to be clarified, as does the relationship between the various deviations in PV and the transition into acute leukemia. Currently, it is known only that chromosomal deviations in PV are almost invariably present in cases that have transformed into leukemia (Lawler et al., 1970). Recently, Rowley (1975a) reported 3 cases of myeloproliferative dis-

TABLE VII KARYOTYPIC OBSERVATIONS WITH BANDING METHODSI N POLYCYTHEXIA VERA, MYISLOSCLEROSIS, ESSF:NTI.IL THROMBOCYTOSIS, A N D RELATED MYELOPROLIFERATNE DISORDERS OF ILLDEFINEDTYPE

Diagnosis Polycythemia Vera

Karyotype 46,?, 20q? ,?, 20q -, other deviations not reported 46,XX, del(20)(q11) 46,XY, l l q - , 13q-/46,XY, 11q-/46,XY, 13q46,XY, del(7) (q22)/46,XY 46,XY, del(7) (q22)/46,XY, del(7) (q22), inv(l4) ( p l l :q24) 46,XX, -15, +t(lq;l5q) 46,XY, t(l2;17)(q13;~11)/47,XO,-Y, +t(Y;l)(q12;q21), 4-9 47,XX, + l p 47,XX, +8/46,XX ? ,? , +8, ? ? ,? , +9, ? 5l,XX,+t(l;l5)(pl?;ql?), t(2;11)(p13;q21),’ +3, +8?, +8?, +C, -15,16p--, +19 47,? , +9 48,XX, +8or +lo, +9

+

M yelosclerosis Essential thrombocytosis Ill-defined myelo- 46,XX,/47,XX, +9 proliferative 47,XY, +8 other deviations not reported disorders ? ,? , +8, 47,XYq -, +1 46,XY/47,XY, +mar

+

a

Constitutional anomaly. Changed into A m .

Number of cases

References

3 1 1

Reeves et al. (1972) Reeves et al. (1972) Shiraishi et al. (1975) Shiraishi et al. (1975) Rowley (1973e) Tsuchimoto et al. (1974) Wurster-Hill et al. (1976) Rowley (1973c, 19758) Wurster-Hill et al. (1976) Hsu et al. (197413) Rowley, cited by Hsu et al. (1974b) Wurster-Hill, cited by Hsu et al. (1974b) Rowley (1975a)

2 1

Davidson and Knight (1973) Rowley (1973~)

2 2 1 1 1 1 1 1 1 1

Knight et al. (1974) Hsu et al. (1974b) Whaun et al. (1974) Warburton and Bluming (1973) Brandt et al. (1975)

X

C

Fz

194

JOACHIM MARK

eases ( 2 cases of PV, Table V; 1 case of AML with the karyotype 48,XX,+1,-5,+11,+mar) which were trisomic for either the entire No. 1 or for its long arm. The author referred to the similar findings in 3 additional cases collected from the literature ( 2 cases of PV reported by Wurster-Hill et al., 1976, and 1case of an ill-defined disorder described by Warburton and Bluming, 1973; see Table VII); these findings indicated a possible nonrandom involvement of No. 1 in myeloproliferative diseases, too. Although the findings in AML (Table V) support this idea, many more cases must be studied to allow definite conclusions to be drawn. Nevertheless, the observations available are of interest in that the similarities to the variation found in the meningiomas are increased (cf. Sections I1 and V, and Tables I1 and XII).

F, MISCELLANEOUS MYELOPROLIFERATIVE DISEASES Banding methods in past years have also been used for studies of a variety of partly ill-defined hematological conditions with a recognized or possible propensity for neoplastic transformation. The investigations comprise 4 sideroblastic anemias, 3 cases of a special type of refractory, dysplastic anemias, 2 cases of pancytopenias, and 1 case of granulocytopenia and thrombocytopenia. The cytogenetic observations in these conditions are listed in Table VIII. The chromosomal findings in sideroblastic or sideroachrestic anemias were recently surveyed by Cohen et al. (1974) and by Sandberg and Hossfeld (1970, 1974). The 27 cases studied, using conventional methods, have usually shown diploid-near-diploid chromosome numbers in their marrow cells. The karyotypes were said to be essentially normal in about 40%of the cases whereas the others have displayed a variety of disturbances, such as nonspecific heteroploidy, slight increase in polyploidy, and deletions or pericentric inversions affecting an F-group chromosome (the F abnormalities were, as a rule, seen only in a small number of the cells; de Grouchy et al., 1966). The observations in the 4 banded cases (Table VIII) support some of these previous results. What is more stiiking, however, is the similarity to the findings in polycythemia Vera (and related disorders; cf. Table VIII). This is of considerable interest in view of the known propensity for both diseases to transform into acute leukemia. According to a recent study comprising 20 patients with sideroblastic anemia (Granberg et al., 1975), there were increased frequencies of breakage, structural rearrangements, and satellite associations during the phase clinicopathologically corresponding to a sideroblastic type of anemia. Both during the phase of transformation into leukemia and after-

TABLE V I I I CHROMOSOMAL FINDINGS I N ‘‘PRELEUKEMIC” OR POTENTIALLY Diagnosis Sideroblastic anemia

Refractory, dysplastic anemia Pancytopenia Granulocytopenia a

+ thrombocytopenia

Transformed into acute myeloid leukemia.

Karyotype 45,?,-7” 46,XY,2Oq47,XX, +8/46,XX 47,XY, +8 46,XY, del(5) (q12q23) 46,XX, dc1(5)(q12q23) 47,XY, +8 acentric fragment 45,XY, -7, 47,XX, +8

+

“PRELEUKEMIC”

Number of cases

CONDITlONS

El

.3 References Gahrton et al. (1974d) Cohen et al. (1974) Jonasson et al. (1974) Hellstrom et al. (1971) van den Berghe et al. (1974) van den Berghe et al. (1974) dc la Chapelle et al. (1972) Rowley (1973e) de la Chapelle et al. (1972)

$

5 2!

5

‘ 3

8 s c (

v1

E

z

UI

196

JOACHIM MARK

ward, the frequency of breakage and satellite associations was decreased, but the frequency of marker chromosomes increased. If these data are confirmed, it might be possible to exactly determine the crucial cytogenetical steps in the transformation process. The specificity of the anomaly in the refractory, dysplastic anemia, reported by van den Berghe et aZ. (1974), is remarkable. The authors referred to one possible further case in the literature (Fitzgerald and Hamer, 1971). A critical evaluation of these data, however, cannot be done until additional cases with the same or similar anomalies have been described, and also the possible propensity for a neoplastic conversion has been investigated, The abnormalities found in the 2 cases of pancytopenia and in the single case of granulocytopenia and thrombocytopenia agree with the observations in AML and those in previous group (cf. Tables V and VII ) The significance of these deviations, however, are difficult to evaluate, partly because of the few observations, but especially because the clinical course in these cases is unknown. In this context, there are reasons to mention some of the concluding remarks in Nowell’s report ( 1971 ) about chromosomal deviations in “preleukemic” conditions ( several cases of pancytopenias, granulocytopenias, and thrombocytopenias were included in this material). It was found that the risk of developing clinical leukemia within a few months is great if a chromosomal marrow abnormality is detected in a nonirradiated “preleukemic” patient; patients demonstrating such development were believed to have subclinical leukemia already at the time of the cytogenetical study. Conversely, it was suggested that “preleukemic” patients, not showing such development within 3 months (but having a chromosomal marrow abnormality), are perhaps thereafter at no greater risk than comparable patients without a chromosomal anomaly in their mrrow. The latter assumption, however, was questioned by Krogh Jensen and Philip (1973), who believed that an abnormal stemline in the marrow was synonymous with a neoplastic clone. The long survival in some documented cases with an abnormal stemline in the marrow was ascribed to a probable, low degree of malignancy of some neoplastic clones and also to a follow-up that was not long enough.

.

IV. Lymphoproliferative Disorders

Lymphoproliferative disorder is a descriptive term used here to denote a variety of usually closely related neoplastic and preneoplastic diseases originating from cells of the lymphoid tissues. The disorders to be discussed comprise non-Burkitt lymphomas, Burkitt lymphomas, myelomas,

BANDING PAlTERNS I N HUMAN NEOPLASMS

197

and Louis-Bar syndrome (and together with this syndrome some observations in routine lymphocytic cultures ) . The chronic lymphocytic leukemias are not included in this survey, the reason being that the chromosomal findings with conventional methods (references in Sandberg and Hossfeld, 1974), as well as those with banding methods (Crossen, 1975), most likely do not reflect the chromosomal constitution of the neoplastic cells, Also, there are no reliable methods for isolating the leukemic cells from the normal ones, or for stimulating the neoplastic cells to divide. Neither are the acute lymphocytic or lymphoblastic leukemias included. Thus, the findings with conventional methods were considered in detail earlier (Sandberg and Hossfeld, 1974; Muldal and Lajtha, 1974) and only a few cases have been studied with the new techniques (Steel, 1971; Huang et al., 1975; Garson and Milligan, 1974; Schmidt et al., 1975; van Biervliet et al., 1975; Bloomfield et al., 1975). One observation in the report by Huang et al. (1975), however, is mentioned because of the findings discussed in the following sections. This is the occurrence of a translocation of almost the whole long arm of one No. 14 onto the short arm of one No. 11 in a cell line established from a patient with acute lymphoblastic leukemia (leukemic cells with T-cell characteristics). An ideal basis for a discussion of the cytogenetic findings in lymphoproliferative disorders would have been a subdivision of the material according to B- and T-cell characteristics (see Seligmann, 1974). Unfortunately, this is not yet possible, since data about these features are missing in almost all reports.

A. NON-BURKITT LYMPHOMAS In comparison with leukemias, cytogenetic data about malignant nonBurkitt lymphomas ( non-BL ) have been accumulated slowly during the past 15 years. The number of cases reported in the literature, or cited in reviews, include about 180 tumors. Complete data by banding methods are available for only 17 of these lymphomas (Fleischmann et al., 1972; Reeves, 1973; Lawler et al., 1975; Prigogina and Fleischman, 1975a; Boecker et al., 1975; Mark, 1975b; Zech et al., 1976), if cell lines are excluded. A detailed cytogenetic survey of the non-BL is seriously hampered by inadequate documentation of many cases, especially older ones, but even some recently reported ones. Information about one or several of the following characteristics is often missing: the histologic subtype, the modal number, the stemline karyotype, and the basis for the observations (preparations from biopsies, short-term cultures, or both sources).

198

JOACHIM M A R K

Because of such reasons the following principles were adhered to in the present survey: ( 1 ) Histopathologically the tumors were divided into 3 groups: ( a ) cases of Hodgkin’s disease, ( b ) histiocytic lymphomas or reticulosarcomas ( including pure histiocytic types and undifferentiated types), and ( c ) lymphocytic lymphomas (including pure lymphocytic types and mixed lymphocytic-histiocytic types). ( 2 ) The level of ploidy was generally used instead of the modal number. ( 3 ) Karyotypic findings in cases studied with conventional methods were discussed only with regard to some consistent features.

1. Numerical Findings Recently Reeves (1973) and Lawler et al. (1975) summarized the predominant chromosome numbers of the presumed tumor cells in 129 cases of non-BL (46 cases from the Royal Marshden Hospital and 83 cases collected from the literature). Table IX shows a modified version of these data supplemented with 57 additional cases. These were as follows: 1 lymphosarcoma and 1 reticulosarcoma studied by Wisniewski and Korsak (1970); 17 cases of Hodgkin’s disease, 13 reticulosarcomas and 8 lymphosarcomas reported by Fleischman et al. (1974); 1 case of Hodgkin’s disease and 1reticulosarcoma described by Nassar and Khouri (1974); 2 cases of lymphosarcomatous tumors studied by Prigogina and Fleischman (1975a); 2 cases of Hodgkin’s disease investigated by Boecker et al. (1975); 1 case of Hodgkin’s disease and 6 reticulosarcomas analyzed by this author ( Mark, 197513, unpublished observations, 1976) ; finally, 3 lymphosarcomas ( 2 of these cell lines) and 1 cell line from a case of Hodgkin’s disease reported by Zech et al. ( 1976). Table IX demonstrates a striking difference between the Hodgkin lymphomas and the non-Hodgkin cases in their distribution of the modal

LEVEL OF PLOIDY OF

THE

TABLE IX MODAL POPULATION I N 186 NON-BURKITT LYMPHOMAS“ Non-Hodgkin lymphomas

Level of ploidy

Hodgkin’s disease

Histiocytic

Lymphocytic

2n 3n 4-6n Number of tumors

29 42 29 66

78 3 19 59

90 3

7 61

Tho figures in each histological group are given as percentages (except for the bottom row). 0


199

populations. In the former group almost half of the tumors have their mode in the triploid region and the remaining cases are equally distributed on the diploid and the tetraploid-high polyploid regions. Conversely, modes in the triploid region are rarely seen in the non-Hodgkin lymphomas. Instead, more than three-quarters of both tumor types in this group have their modal populations in the diploid region. The higher frequency of tetraploid-high polyploid modes in the histiocytic lymphomas, as compared to the lymphocytic tumors, might reflect a true difference. The different pattern of modality in the Hodgkin’s lymphomas, compared with the other two lymphoma groups, can hardly be attributed to a deviating composition of the tumor cell population. Other factors, such as immunological mechanisms, however, may be responsible for the differences in the evolutionary patterns. Thus, the Hodgkin’s group might include many lymphomas whose early growth at the diploid level is often strongly counteracted by (possibly changed) immunological mechanisms. A stemline shift, originating from doubling products (known to occur in a somewhat varying frequency in most neoplasms) of neardiploid tumor cells would be one way to increase viability and, at least partly, escape the action of immunological forces. This suggestion is in agreement with experiences from animal model systems. Thus, Hauschka and Levan as early as 1953 demonstrated an inverse relationship between chromosome ploidy and host-specificity in a series of 16 transplantable mouse tumors (4 lymphomas were included in this material). Tumors with diploid chromosome modality were found to be completely strainspecific whereas neoplasms with a near-tetraploid constitution exhibited a varying degree of genetic indifference regarding their host-requirements. The pattern of modality in the Hodgkin’s lymphomas was also recently discussed by Miles (1973). The author emphasized the common occurrence of two, presumably neoplastic populations in Hodgkin’s disease, one diploid-near-diploid and one polyploid. This feature was thought to indicate a relatively earlier stage of cytogenetic progression than is usually found in most epithelial cancers. This assumption might be correct, when a comparison is made with carcinomas; but it seems to be an unlikely explanation for the differences in pattern of modality between Hodgkin lymphomas and the histiocytic-lymphocytic lymphomas. Two very special types of lymphomatous conditions were not included in the above survey, namely mycosis fungoides and SBzary’s syndrome. Only 13 cases of mycosis fungoides have been studied cytogenetically (Spiers and Baikie, 1968; Obara et al., 1970; Erkman-Balis and Rappaport, 1974). The modal population in these cases was usually in the diploid region, most often in the hyperdiploid zone. These results concur

200

JOACHIM MARK

with the close relationship, histiopathologically, between mycosis fungoides and histiocytic lymphomas. The so-called SBzary’s syndrome is a much debated disorder, which is regarded by some investigators as an erythrodermic variant of mycosis fungoides, and considered by others to be a different entity clinicopathologically. The cases of SBzary’s syndrome differ cytogenetically from mycosis fungoides by the PHA stimulation of the circulating abnormal cells and the high frequency of polyploid, usually near-tetraploid, stemlines ( Crossen et al., 1971b; Lutztner et nl., 1973; Dewald et al., 1974; Ding et al., 1975).

2. Karyotypic Findings The karyotypic data that have been accumulated about non-BL during the prebanding era comprise a great number of more or less complex numerical and structural abnormalities. An effort to compile these data and to find consistent features was initially made by Spiers and Baikie (1966, 1968, 1970). The authors introduced the hypothesis that anomalies of the chromosomes Nos. 17 and 18 played a special role in the genesis and/ or evolution of some reticuloendothelial neoplasms. They pointed out that aberration of Nos. 17 and 18 exceeded in variety the reported anomalies of any other chromosome or chromosome group, and that both the frequency of these aberrations and their nature made them most unlikely to be caused by chance. The discussed anomalies consisted of numerical changes, as monosomy or trisomy for Nos. 17-18, and of structural changes, as long- and short-arm deletions, isochromosomes for the long and short arms, and various translocations. The hypothesis proposed by Spies and Baikie (1966, 1968, 1970) gained support from investigation performed by Millard ( 1968). The results of her own studies and a reconsideration of previously reported cases indicated that chromosome No. 18 was the particular pair most often involved in the structural rearrangements. Short-arm deletions of No. 18 (the so-called “Melbourne” chromosome) and long-arm deletions of this pair together make up the majority of cases with structural anomalies in group E. All such cases on record were recently reviewed by Reeves ( 1973). These observations, supplemented with one case of mycosis fungoides (Erkman-Balis and Rappaport, 1974), are summarized in Table X. It shows that deletions of No. 18 (177) have been found in all types of non-BL and that long-arm deletions predominate. Out of the 20 non-BL hitherto studied by banding methods, however, only one tumor, a lymphocytic lymphoma (Reeves, 1973), has contained cells with a deleted No. 18 ( 18q-). The rarity of non-BL cases with deletions of No. 18, even in banded materials, throws doubt upon the specificity of these particular changes. Further


201

TABLE X LYMPHOXAS WITH DICLETIONS O F CHROSOME No. 18 (17?) Type of aberration Tumor type

IIodgkin’s disease Histiocytic lymphomas Lymphocytic lymphomas Mycosis fungoides Total numbcr:

P2 1 1 -

4

g-

1 3 6 -1 11

studies, however, are necessary to completely evaluate the original hypothesis of Spiers and Baikie ( 1966,1968,1970). Later Fleischmann et al. (1972) made a new try at finding consistent cytogenetic features in the non-BL group. The authors surveyed chromosome data from 36 different lymphomas and tabulated 33 different karyotypes. Although a very complex picture emerged, some suggestive features were seen: i.e., losses of Nos. 17-18 in 12 cases and gains in only 1; in 8 of the 33 karyotypes changes, mostly gains, of chromosomes NO. 16; C group changes in 18 karyotypes, with predominance of losses in stemlines with numbers below 48, and generally gains in stemlines with 48 or more chromosomes; G-group changes, mostly gains in 15 karyotypes. Marker chromosomes of various types, and as a rule with unknown derivation, were observed in almost 60%of the cases. No marker type was found to be characteristic of the entire group of non-BL. Fleischmann et al. (1972) in their review, as in a previous report (Fleischmann et al., 1971), paid special attention to a marker type of medium size and with identical arms. Using the Q-banding technique, the authors had observed one or several of these metacentric markers in 1 completely analyzed lymphosarcoma and in 4 additional, only partly analyzed, lymphomas. Three additional cases with similar markers were collected from the surveyed material (cases of non-BL studied with conventional methods). To these 8 cases can be added a recently reported ninth case (Mark, 197513). In this histiocytic lymphoma, the metacentric marker was probably an isochromosome for the long arm of another marker ( a 13q-) included in the stemline karyotype. The origin of the metacentric markers in the other 8 lymphomas was obscure. In the above-mentioned 9 cases, the morphology of the metacentric marker agreed with that of an isochromosome. Judging from published karyotypes in the banded cases, however, these markers do not look quite the same in all tumors. There are often small differences in both

202

JOACI-IIM M A R K

size and banding between the metacentric markers in different cases; furthermore, in cases with several such markers there are also differences between them. Therefore, contrary to the idea of Fleischmann et nl. (1972), it is suggested here that the metacentric markers actually represent a variety of different isochromosomes (or possibly in some cases products of deletions and/or centric fusions ). The early, scanty fluorescence studies of non-BL cases have been subsequently supplemented with investigations by G-banding technique. Using this method Reeves (1973), and Lawler et nl. (1975), described the cytogenetic findings in a series of 10 non-BL ( 3 cases of Hodgkin’s disease, 3 histiocytic, and 4 lymphocytic lymphomas). In this series, the authors found no numerical or structural change common to all tumors or any of the histological subtypes. (The data, in particular the deviations affecting normal chromosome types, however, are very difficult to evaluate. This is because a definite stemline karyotype is absent or not given in many cases, and in some other cases only a few cells are analyzed. ) Among the many structural changes recorded, however, several occurred in two or more tumors, and in some cases the breakpoints were even identical, for instance 2 cases with a chromosome lp- and breakpoint in p22, and 3 cases with a chromosome 9q- and breakpoint in q22. By plotting all available breakpoints in the entire material, Reeves (1973) obtained additional data supporting the idea of a preferential pattern of the structural rearrangements in non-BL. One important aspect in this pattern, emphasized by the author, was that almost without exceptions the breakpoints appeared to be in the light-staining bands and at the centromere. The last few years 10 further cases of non-BL have been investigated with the new techniques: G-banding methods were used in the studies of 2 lymphosarcomatous tumors (Prigogina and Fleischman, 19754, 2 cases ( b u t only one completely analyzed) of Hodgkin’s disease (Boecker et al., 1975), and 2 histiocytic lymphomas (Mark, 1975b), and the Q-banding technique was utilized for the analyses of 3 Iymphosarcomas (1 biopsy and 2 cell lines) and 1 cell line derived from a case of Hodgkin’s disease (Zech et al., 1976). Except for 1 case of Hodgkin’s disease, all these lymphomas had their mode in the diploid-near-diploid region and a11 of them had one, or more often several, marker chromosomes included in their stemline karyotype. These structural deviations and those recorded by Fleischmann et al. (1972) and by Reeves (1973) are summarized in Table XI. The following principles were used for the tabulation: cIarified translocations were denoted with a p- or q- for the donor chromosome and a p+ or q+ for the receptor chromosome; only a p+ or q+ was used in cases where the origin of the extra material

STRUCTUHAL

TABLE XI ABERRATIONSRECORDED Ipi 21 B A N D E D NON-BIJRKITT LYMPIIOM.48

td

2

Ez 0

Chromosome groups and chromosome types

'j

B

A Type of aberration

1

2

3

4

C 5

6

7

8

9

D 10 11

12

I3

14

E 1.5

16

17

F

G -~ 18

19

20

21

22

Y

X

5 M

m

5

204

JOACHIM MARK

was uncertain or unknown; a few complicated rearrangements (as a rule with questionable interpretation ) were excluded, as were also the cases with uncertain derivation of the marker chromosomes. The figures in Table XI give an approximate idea of the complexity of the structural changes and also an estimation of their distribution on the various chromosome types. As to the latter feature, it is obvious that certain chromosome types are affected more often than the others, namely Nos. 1 and 14, in particular, but also Nos. 3, 9, 6, and 12. The involvement in these cases (as usually also of the other chromosome types) as a rule consists of a variety of changes. Exception to this rule are the deviations concerning pair No. 14. In this case, all changes involve the long arm, and with one exception (an inversion involving the long arm; Zech et al., 1976), they lead to the formation of a 14q+ marker. In most of the lymphomas with this marker, it had a morphology similar to the one shown in Fig. 5. The high frequency of non-BL cases with a marker chromosome 14qf (found in 9 of 21 banded cases) is of great interest with regard to recent observations in other lymphoproliferative disorders ( cf. below and cf. introduction to the present section), A numerical karyotypic deviation, also worth noticing in this context, is the occurrence of trisomy 7 in 5

FIG.5. Partial karyotype of a histiocytic lymphoma showing the normal No. 14 and the 14q+ marker in two stemline cells; G-band technique. x2800. From Mark (197Sb), with permission.


205

non-BL cases (Fleischmann et al., 1972; Zech et al., 1976). So far, this is the only change among the normal chromosome types, which has been observed in a considerable proportion of the banded tumors. 3. Angioimmunoblastic Lymphadenopathy

Angioimmunoblastic lymphadenopathy (AIL ) is an only recently recognized clinicopathological entity. It is thought to be a benign disease. The chromosomes of lymph-node-derived cells have been studied in 2 cases of AIL by Hossfeld et al. (1976). In one case, investigated only with conventional methods, 61 out of 75 analyzed cells were abnormal and they had pseudodiploid and hyperdiploid karyotypes. These showed one Bq+ marker and/or a gain of one C group chromosome. In the other case, studied with banding methods, 22 out of 89 analyzed cells were abnormal, and they also had pseudodiploid-hyperdiploid karyotypes. Most of these deviating cells showed only a gain of one No. 3; the others (i.e., those with a consistent pattern), in addition to this change, displayed gains and/or losses affecting Nos. 8,9,20, and 21. As pointed out by Hossfeld et al. (1976), the occurrence of cytogenetically abnormal clones in these cases casts doubt on the assumption that AIL is, in fact, a benign disease; the chromosomal findings rather support the view that AIL might be considered a ‘prelymphoma.” Longitudinal studies will answer these clinicopathological questions. Further cytogenetic analyses will also be helpful for a classification of the entity AIL and for a determination of its relationship to other types of truly malignant non-BL.

B. BURKITTLYMPHOMAS The cytogenetic studies of the Burkitt lymphomas (BL) were initiated by Jacobs et al. ( 1963). They described the findings in a series of 19 tumors investigated in biopsies; of these cases 13 yielded adequate or partly adequate preparations. Later, this study has been supplemented with a few additional cases also analyzed in biopsies (Stewart et al., 1965; Gripenberg et al., 1969; Manolov et al., 1970) and by investigations of a great number of BL cell lines and clones of such cell lines, derived from different lymphomas (Stewart et al., 1965; Cooper et al., 1966; Chu et al., 1966; Rabson et al., 1966; Miles and O’Neil, 1967; Bishun and Sutton, 1967; Kohn et al., 1967, 1968; Toshima et al., 1967; Tomkins, 1968; Kurita et al., 1968; Hughes, 1968; Tough et al., 1968; Huang et al., 1969, 1970; Nadkarni et al., 1969; Macek and Benyesh-Melnick, 1969; Zajac and Kohn, 1970; Whang-Peng et al., 1970; Ikeuchi et al., 1971; Macek et al.,

206

JOACHIM MARK

1971). These studies from the prebanding period showed a pattern of the modal numbers similar to that in the lymphocytic lymphomas, i.e., a predominance of diploid-near-diploid modes. Several BL cases had a seemingly normal, diploid stemline karyotype. The majority of the others had pseudodiploid or hyperdiploid stemlines with inconsistent numerical deviations in the normal chromosome groups; an involvement of groups C (mostly gains) and G (mostly losses), however, were found comparatively often, particularly in cell lines. In many cases marker chromosomes of various types were detected. Some of these marker types, however, were found in several cases. Thus, a big acrocentric marker, believed to be derived from the long arm of a No. 2, was observed in at least 7 of the tumors studied in biopsies and in several cell lines. A C group chromosome (sometimes more than one), believed to represent a No. 10, with a constriction at the end of its long arm, was found in many of the BL cell lines, but in varying percentages of the metaphases. The application of banding methods to the BL material changed the picture radically. Thus, Manolov and Manolova ( 1972), using the Q-banding technique, reported a marker band in one chromosome No. 14 in 10 out of 12 tumors studied. The anomaly was seen in all analyzable cells in 5 out of 6 tumor biopsies and in 7 out of 9 BL cell lines examined. The marker band was located at the end of the long arm of one homolog No. 14. The anomaly was considered a reliable marker for the Burkitt lymphomas. The origin of the extra material at one No. 14, however, could not be determined. The occurrence of some cases without the 14q+ marker was proposed to be due to a disappearance during tumor development or perhaps to the existence of two different types of BL. These explanations are difficult to assess, partly because no complete stemline karyotypes were presented. An enigma in connection with this report is the complete absence of the marker 14q+ in a previously published series OF 12 Q-banded BL ( 7 biopsies and 11 cell lines from altogether 12 different BL) ( Manolov et nl., 1971 ). The marker chromosome 14q+ has also been observed in the BL cell lines JIYOYE (Steel, 1971) and P3J.KRlK (Petit et al., 1972). More recently, the same marker was found in 7 out of 7 BL cell lines briefly described by Jarvis et al. (1974). The complete stemline karyotype in the cell line reported by Petit et al. (1972) was as follows: 47, XY, 7qh, 9qh, 13q+, 14q+, -21, +2 unidentified C-group chromosomes with a variable morphology. Thus, there were 2 different C-group chromosomes with a secondary constriction, but none of then1 was a No. 10. A chromosome No. 10 with a secondary constriction, however, was seen in the material studied by Manolov and Manolova (1972) although no details were given. Two C-group chromosomes with a secondary constriction

BANDING PA’ITERNS IN HUMAN NEOPLASMS

207

have been observed (Miles and O’Neil, 1967; and others), and they have usually been considered homologous. The observations by Petit et al. (1972) cast doubt on this conclusion, The banding pattern in the BL has recently been reevaluated by Zech et al. (1976). They found a 14q+ marker in biopsies of 10 out of 10 tumors and in 3 out of 4 BL cell lines. Furthermore, a small segment at the end of the long arm of one No. 8 (i.e., 8q-) was missing in all 14qf positive cases with metaphases of good technical quality (10 out of the 13 cases). The size and stainability of the missing segment of one No. 8 agreed with the corresponding characteristics of the extra material of one No. 14. Thus, the authors concluded that the marker 14q+ most likely had resulted from a translocation between Nos. 8 and 14, i.e., t ( 8q--;14q+). This important discovery represents a parallel to the recent observations in Phl-positive CML (i.e., the translocation t ( 9q+; 22q- ) (cf. Section II1,A). As in the latter condition, it is not presently known whether the translocation in the BL is reciprocal. The abnormality t ( 8q-;14q+) was not seen in 8 lymphoblastoid cell lines ( 3 of them derived from BL cases and 5 from non-BL cases), nor in the lymphocytes of the peripheral blood of 5 BL patients. Thus, Zech et al. (1976) (cf. also Jarvis et al., 1974) found no direct relationship between the Epstein-Barr virus and the occurrence of the specific translocation. The role played by this lymphotropic herpesvirus in the development of BL remains unclear (cf. the recent survey by Klein, 1975). In their material of BL cases, Zech et al. (1976) observed trisomy 7 in 2 biopsies and in 1 cell line; in an additional cell line, however, one No. 7 was involved in a translocation, Steel (1971) also reported trisomy 7 and partial trisomy 7 in the BL cell lines RAJI and JIYOYE, respectively. Except for the translocation t ( 8q-;14q+ ), the involvement of No. 7 is so far the only additional change observed with noticeable frequency. This involvement gains in importance by the occurrence of trisomy 7 in as many as 5 of the 20 banded non-BL cases (cf. Section IV,A), and it is also of interest in view of recent findings in Louis-Bar syndrome and some observations in routine lymphocyte cultures (cf. Section IV,D). The specific 8 to 14 translocation and also the involvement of No. 7, reported by Zech et al. (1976), are important observations, which, however, need confirmation by other investigators. Further studies of BL cases are also necessary to clarify several other questions: (1) The distribution, frequency, and real nature of the secondary constrictions in the group C chromosomes. This special anomaly has attracted much interest in the past; a few banded cases of BL with these constrictions are on record, but data about this feature are completely missing in some series (Jarvis et al., 1974; Zech et al., 1976). ( 2 ) The characteristics of the

208

JOACHIM MARK

involvement of group G in BL. A loss of G-group chromosomes has often been found in BL cell lines studied in the prebanding era; with the exception of the cell line studied by Petit et al. (1972), however, these observations have not hitherto been confirmed by the studies with the new techniques. (3) The derivation of the big acrocentric marker found in many of the biopsies of BL which were studied with conventional methods (Jacobs et al., 1963). This abnormality has not been redetected in the banded cases, and, because of the big size of the marker, it seems unlikely that it would correspond to the 14q+ marker detected at first with banding methods. ( 4 ) The further chromosomal evolution in cases of BL without the specific translocation t(8q-;14q+). Zech et al. (1976) found one BL cell line without this translocation, and Manolov and Manolova (1972) mentioned 2 cases that lacked the 14q+ marker (and thus, probably also lacked the 8q- marker). These findings might be due to a disappearance of the markers during the progression or the existence of two different types of BL, as suggested by Manolov and Manolova (1972). However, there is also the possibility that the specific translocation could evolve later during progression or, as has been noticed in some meningiomas (cf. Section 11), that the evolution in some BL cases could originate with steps that usually represent late, superimposed deviations. ( 5 ) The cytogenetic picture in BL versus non-BL. This is no doubt the most important question to be made clear by the future studies. Thus, the translocation t(8q-;14q+) was seen in one of the non-BL cases studied by Zech et al. 1976), but only 4 cells could be completely analyzed in this biopsy from a lymphosarcoma. The 8 to 14 translocation was not observed in the 16 banded non-BL investigated by others (Fleischmann et al., 1972; Reeves, 1973; Prigogina and Fleischman, 1975a; Boecker et al., 1975; Mark, 1975b), but the marker 14q+ was seen in 6 cases. It is possible that the small 8q- change was overlooked in some or all of the cases exhibiting a chromosome 14q+; however, it is also conceivable that the 14q+ marker has a different and/or varying origin in different types of lymphoproliferative disorders. C. MYELOMAS

The term myeloma is used here for plasmocytoma, multiple myeloma, and also plasma-cell leukemia. The cytogenetic studies of myelomas, performed with conventional methods, are very difficult to evaluate because short-term cultures have been used to a great extent in many or most of the reported larger series. According to several recent studies and reviews (Sandberg and Hossfeld, 1970, 1974; Dartnall et al., 1973; Anday et al., 1974; Krogh Jensen et al., 1975), some type of chromosomal abnormali-


209

ties has been found in most cases. An abnormal stemline, however, was seen in only about one-fourth of the cases (Krogh Jensen et ul., 1975). In these cases the mode was usually in the diploid region, most often in the hyperdiploid zone. The karyotype analyses have not revealed any consistent numerical or structural change. It should be mentioned, however, that big acrocentric markers, bigger than the D group chromosomes, have been found in many cases in at least one or a few cells. This is of interest with regard to some observations with banding methods. The technique used in the recent chromosome studies of myelomas has involved various G-banding methods. Wurster-Hill et al. (1973) was first to give a brief description of the findings in 2 cases, one with multiple myeloma and one with plasma-cell leukemia. In the former case there was a bimodal chromosome count of 46 and 42. The 46-chromosome cells analyzed had a normal karyotype. All of the twenty-one 42-chromosome cells studied, however, contained a marker chromosome 14q+ (replacing a normal No. 14) with two extra bands at the end of its long arm. Other markers, which were not described, were also present, and the origin of the extra segment in one No. 14 could not be determined. An apparently identical 14qf marker was also seen in 6 out of the 35 karotyped cells in the other case (the modal number = 46). Additional markers were found in this case too, and the origin of the extra bands in one No. 14 remained obscure. Observations with banding methods in 2 additional cases, both of them having multiple myeloma, were reported in brief by Philip and Drivsholm ( 1974) and by Philip ( 1975a,b). In one of these cases (Philip, 1975a), 6 out of 14 kaiyotyped cells exhibited a 14q+ marker. Its morphology was similar to that in the above-mentioned myelomas, described by Wurster-Hill et al. (1973), and, as in these cases, the origin of the extra material on No. 14 could not be established. In the cells with a 14q+ there was also another, bigger marker chromosome, which was formed by a 1 to 3 translocation (breakpoint in No. 3 = q27-29). No 14q+ marker was seen in the other case of multiple myeloma ( Philip and Drivsholm, 1974; Philip, 1975b). Instead, these tumor cells contained 3 other big marker chromosomes with dissimilar morphology. It is of interest that a chromosome No. 3 participated in the formation of one of these markers and furthermore, that the breakpoint in No. 3 was the same as that in the case discussed above, namely q27-29. Although undoubtedly scanty, the summarized banding data from studies of myelomas do suggest that the marker 14q+ is a common cytogenetic abnormality also in this group of disorders. The high frequency of cases with big acrocentric markers among the myelomas studied with conventional methods (cf. above), is a finding in agreement with this

210

JOACHIM MARK

assumption. The involvement of chromosome No. 3 in at least 2 out of the 4 banded cases of myeolmas is also a feature of interest because this particular chromosome type was one of those most frequently affected in the structural deviations recorded in non-BL cases (cf. Table XI).

D. LOUIS-BAR SYNDROME Louis-Bar syndrome ( LB ), or ataxia telangiectasia, is a rare autosomal recessive disorder characterized by progressive cerebellar ataxia, oculocutaneous telangiectases, and immunodeficiency of poorly understood nature. LB also belongs to the so-called chromosome instability syndromes, and the patients show a predisposition to cancers, in particular development of lymphoid neoplasia. The chromosomes in short-term cultured lymphocytes from LB patients have been intensely studied during the last few years (references in the review by Harnden, 1974). The results are unusually consistent, and they are of great interest with regard to the chromosomal findings in malignant lymphomas. Thus, it was recently shown (McCaw et al., 1975) that the lymphocyte cultures from all the 9 patients with LB, studied with banding methods, contained anomalous clones. The proportion of cytogenetically abnormal lymphocytes varied from case to case, but, as a rule, there was an increase with time in cases studied longitudinally. The anomalous clone in all of the 9 cases exhibited a structural rearrangement affecting one No. 14, furthermore the long arm of No. 14. Seven cases were characterized by translocations (14 to 14 in 3 cases, 14 to 7 in 2 cases and 14 to 6 and X, respectively, in one case each), one by a ring 14 (thought to be formed by deletions of the ends of both the short and the long arms of No. 14), and one by a chrommome 14q+ (the origin of the extra material in this case was unknown). Analyses of the breakpoints in chromosome No. 14, in cases with translocations, revealed a clustering in band q12 (cf. below, observations in routine lymphocytic cultures). McCaw et al. (1975) also tabulated data from 13 additional cases of LB with comparable chromosome abnormalities, studied by others and with conventional methods. All but 2 cases (where a D involvement, however, remained a possibility) contained an abnormal clone with rearrangements affecting a D group chromosome. The observations with the old and new techniques have been corroborated by the findings in some additional, recently reported cases of LB (Rary et al., 1975; Hayashi and Schmidt, 1975; Hook et al., 1975). One of the LB patients studied by McCaw et al. (1975) had later developed chronic lymphocytic leukemia. The leukemia cells showed the 14 to 14 translocation observed in lymphocytes before the onset of the disease. In addition, the neoplastic cells displayed further changes, and


211

the stemline karyotype can be written as follows: 41, XX, lq+, 1Oq+, -13, -14, 14q+, -15, -16,18q+, -20 ( 3 translocations: 1/13,14/14, 15118, respectively). Unfortunately, only the abnormal lymphocyte clones, not the so-called “lymphoreticular neoplasias,” were studied in the LB cases reported by Bochkov et al. (1974). Studies of the lymphocytes and the lymphoid malignancies in additional cases will be of utmost importance in the future, and such investigations must be performed before the suggestive findings outlined above can be transformed to any statements. Some recent cytogenetic findings with banding techniques in routine lymphocyte cultures (the cultures were generally harvested after 72 hours, and they were established from individuals without any known chromosome breakage syndromes, malignancies, or any unusual X-ray or drug exposure) have a bearing on the above-mentioned results. The observations reported from three centers ( Beatty-DeSana et nl., 1975; Welch and Lee, 1975; Hecht et al., 1975), concerned the characteristics of de moo structural rearrangements, in particular translocations. The findings by the three groups of investigators agreed, showing a nonrandom pattern for the translocations. Thus, the chromosomes Nos. 7 and 14 were preferentially involved, and furthermore, the breakpoints in these chromosome types were found to be clustered in specific regions, namely 7q3(3-6), 7 ~ 1 3 and , 14ql(1-2) (cf. the findings in LB, above). The biological meaning of these results is uncertain at the present time. Two of the groups working with these problems (Welch and Lee, 1975; Hecht et al., 1975), however, mentioned in particular the resemblance between the nonrandom translocations discussed above and those recorded in LB patients. It will be of great interest to investigate whether the translocations in the former cases are also related to an increased propensity for development of lymphoid malignancies. The common occurrence of 14q+ markers, or other structural rearrangements of No. 14, in various lymphoproliferative disorders, as also the numerical and/or structural deviations affecting No. 7 in some diseases, gains in importance by the cytogenetic findings in LB and routine lymphocyte cultures. Two other observations of interest in this context are: (1) the translocation t ( llp+;l4q-) in the cell line established from a patient with acute lymphoblastic leukemia (Huang et uZ., 1975; cf. introduction to the Section IV,A) and ( 2 ) the constitutional anomaly, i.e., the reciprocal 4/ 14 translocation, in another patient who had developed acute lymphoblastic leukemia ( Garson and Milligan, 1974). All these findings, when put together, indicate (as was pointed out earlier, Mark, 1970a) that tissue specificity, not only the oncogenic agent( s )

212

JOACHIM MARK

(Mitelman et al., 1972), is probably also an important factor in the determination of the chromosomal patterns in neoplasms. V. Concluding Remarks

The cytogenetic observations, particularly those obtained by banding methods, in three types or groups of neoplastic-preneoplastic conditions have been reviewed: ( 1) meningiomas, a benign tumor type; ( 2 ) myeloproliferative disorders, including several clearly malignant diseases, some disorders with a questionable malignancy, and some diseases with known or probable preneoplastic character; ( 3 ) lymphoproliferative disorders, including several malignant tumor types and a hereditary syndrome with a known propensity for development into malignancy, in particular lymphoid malignancies. Table XI1 is a condensed survey of the cytogenetic findings in the diseases mentioned above. The tentative assessment of these data is obvious from the disposition of the table, and some of the implications are commented on below. A definite specific anomaly, aIso primary in the evolutionary chain, is, as seen from the table, found in 3 neoplastic conditions, i.e., meningiomas, Phl-positive CML, and Burkitt lymphomas; such a change is also seen in two conditions with a known or possible propensity for malignant development, i.e., Louis-Bar syndrome and a special type of refractory, dysplastic anemia. In the first three neoplastic diseases mentioned, one, or more often several, chromosome types participate in a chromosomal variation of a secondary or superimposed nature. A corresponding nonrandom pattern is also seen in most of the other preneoplastic-neoplastic conditions, and these chadges, as shown in the table, are for the time being interpreted as superimposed deviations. Further extensive studies are undoubtedly necessary before we can decide whether such an interpretation is correct or the varying initial gross deviations recorded (in, for instance, AML) are related to a multiplicity, or perhaps less specific action, of the oncogenic agents. In several disorders the gonosomes comprise the chromosome types, which are possibly also involved in the nonrandom superimposed deviations. This is of interest because studies of the occurrence of Barr bodies and Y-bodies in tumor cells of mammary carcinomas, bladder carcinomas, lung carcinomas and a variety of other neoplasms (Gropp et al., 1967; Atkin and Petkovi6, 1973; Vass and Sellyei, 1973; Sellyei and Vass, 1975; and others) have indicated a fairly common loss of sex chromosomes in these cases. The reasons for this possible common loss of sex chromosomes in various neoplasms are at the present time completely obscure. The data in Table XII, the findings cited in detail in previous sections

213


TABLE XI1 CONDENSED SURVEY OF THE MAJORFINDINGS WITH B A N D I NMETHODS G IN MENINCIOMAS A N D SOMEM Y E L O P B O L I F ~ R AAT N IDVLYMPHOPROLIFICRATIVIC I~ DISORDERS

Type of preneoplastic-neoplastic disorder Meningioma

Other chromosome Chromosome types types possibly preferentially involved in Specific and affected by the nonrandom, primary change superimposed changes secondary changes -22 (or 22q -)

Myeloproliferative disorders Ph1-positive CML

22q - and usually t (9q ;22q -) Ph1-negative CML ? CMML - Y (or - X ) ? AML ? ? PCV, M, E T and related disorders Dysplastic anemia del(.i)(q12q23) Sideroblastic ? anemia Pancytopenia, ? granulocytopenia, and thrombocytopenia Ly mphoproliferativc disorders Non-Burkitt change of No. 14? lymphoma Burkitt lymphoma t(8q - ;14q +) Myeloma 14q ? Louis-Bar change of No. 14 syndrome

+

+

1,8,9, 14, 15

x, y

8, 17, 19, 22q -

Y

0

Y

?

?

7, 8, 9, 17, 21 1, 8, 9, 20

x, y ?

? 8, 20?

?

8

?

1,.3, 7, 9, 14, 17-18

?

?

7 ? 71, second No. 14

and those to be mentioned below, when put together, strongly indicate that some autosomal chromosome types are also involved more frequently than others in the deviations in various types of neoplasms, namely the chromosomes Nos. 1, 8, 14, and 22. In this context it is worth noting that also several very specific structural aberrations ( affecting the autosomes) have been found in completely different neoplastic conditions, i.e., a 9qf marker in the majority of Ph*-positiveCML (cf. Section III,A), in a cell line of a lymphosarcoma (Zech et al., 1976), and in a meningioma (Mark, 1973a); a 20q- marker in a meningioma (K. D. Zang, personal com-

214

JOACHIM MARK

munications, 1975), in a case of Ph'-positive CML (Prigogina and Fleischman, 1975a), in a case of Ph'-negative CML (Table IV), in sideroblastic anemias, and in pdycythemia Vera ( Tables VII and VIII ) ; a 22q- marker in a cell line of a lymphosarcoma (Zech et al., 1976), in several meningiomas, and in Phl-positive CML (cf. Sections II,C and II1,A). Thus, there are observations indicating that both the numerical and structural variation among the autosomes in different neoplasms have features in common. An understanding of these intriguing findings (anticipated in the earlier discussion of the cytogenetic findings in meningiomas) is not to be expected until more is known about the differences in structural and functional organization between the various chromosome types, and until knowledge about the factors controlling the spindle function has increased. An understanding of the genotypic changes correlated with the phenotypic changes termed cell differentiation is also believed to be important (though hitherto mostly neglected) for any interpretation of similarities or dissimilarities between the chromosomal patterns in neoplasms developed from different tissues. In addition to the observations summarized in previous sections of this chapter, there are a number of other reports concerning the banding patterns in a variety of human tumors. In brief, some of these results are as follows: ( 1 ) Bladder carcinomas: G-banding analyses of 6 hypotetraploid tumors showed consistent banding patterns in only a few normal chromosome types (Nos. 2, 4 and 13) and the origin of the markers (found in all cases) remained uncertain (Falor and Ward, 1973). ( 2 ) Colonic polyps of sporadic and hereditary types: G-banding analyses have shown that the previously established C-D-pattern ( usually losses or gains in these two groups) was related to deviations preferentially affecting chromosomes Nos. 8 and 14 (Mark et al., 1972c; Mitelman et al., 1974a). ( 3 ) Malignant astrocytic gliomas: G-banding of 3 different established cell lines of these gliomas showed very complex numerical and structural changes which particularly involved C and D group chromosomes and most frequently Nos. 7 and 14 (Mark et al., 1974a,b,c). ( 4 ) Mammary carcinomas: two G-banded pseudodiploid primary tumors showed extensive numerical and structural deviations, which, however, in both cases involved Nos. 4 and 16 (the groups B and E, in particular E16, were found to be involved frequently in the variations recorded in published cases studied with conventional methods) ( Mark, 19751). ( 5 ) Malignant exudates from 5 ovarian carcinomas ( Ticpolo and Zuffardi, 1973; Berger and Lacour, 1974; Kakati et al., 1975), 2 lung carcinomas and 1 breast carcinoma (Kakati et al., 1975) and 1 gastric carcinoma (Granberg et al., 1973) have been studied with various banding methods. In this heterogeneous group, the chromosomes Nos. 1,3, 7, and 8 seemed to be most frequently affected by the usually extensive numer-

BANDING PATI'ERNS I N HUMAN NEOPLASMS

215

ical structural remodeling of the karyotypes. There are also many reports dealing with banding analyses (usually only partial) of both old and newly established cell lines (Chen and Shaw, 1973; Lin and Goldstein, 1974; Elliot et al., 1974; and others). The observations in the studies mentioned last are often too fragmentary and/or too disparate to permit any conclusions. In several of the preceding cases, however, there are findings supporting the conclusions outlined above, and which also hold promise for the future search for specific changes or patterns in different tumor types. ACKNOWLEDGMENTS I am indebted to Drs. Janet Rowley, Lore Zech, Avery Sandberg, Klaus Zang, and their collaborators for giving me access to some of their data prior to publication. I also want to thank Lena Strindmar for help with the references; Margareta Anderberg and Inger Carlgren for secretarial assistance; and Rigmor Dahlenfors and Allan Garsater for help with the photographic material. The present work was supported by grants from the Swedish Cancer Society, the Central Hospital of Skovde, and the Carlsson family.

REFERENCES Anday, G. J., Fishkin, B., and Gabor, E. P. (1974). J. Natl. Cancer Inst. 52, 10691079. Atkin, N. B., and Petkovip, I. (1973). J. Clin. Puthol. 26, 126-129. Baccarani, M., Zaccaria, A., and Tura, S. ( 1973). Lancet 2,1094. Baccarani, M., Zaccaria, A., and Tura, S. (1974). Lancet 2,401. Baserga, A., and Castoldi, G. L. (1972). Haematologia 57, 621-640. Bauke, J. (1973). Dtsch. Med. Wochenschr. 98, 1956-1959. Beatty-DeSana, J. W., Hoggard, M. J., and Cooledge, J. W. (1975). Nature (London) 255, 242-243. Benedict, W. F., Porter, I. H., Brown, C. D., and Florentin, R. A. (1970). Lancet 1, 971-973. Berger, R. (1970). Reo. Eur. Etud. Clin. Biol. 15, 1000-1007. Berger, R. (1973). Nouu. Presse Med. 2, 3121. Berger, R., and Lacour, J. (1974). Pathol. Biol. 22, 603-606. Bishun, N. P., and Sutton, R. N. P. (1967). Br. J . Cancer 21, 675-678. Blackstock, A. M., and Carson, 0. M. (1974). Lancet 2, 1178-1179. Bloomfield, C. D., Peterson, L. C., and Brunning, R. D. (1975). Proc. Am. Soc. Hematol., p. 58 (to be published in Blood, 1976). Bochkov, N. P., Lopukhin, Y. M., Kulesbov, N. P., and Kovalchuk, L. V. ( 1974). Humangenetik 24, 115-128. Boecker, W. R., Hossfeld, D. K., Gallmeier, W. M., and Schmidt, C. G. (1975). Nature (London) 258, 235-236. Bottura, C., and Coutinho, V. ( 1974). Blut 29, 216-218. Brandt, L., Mitelman, F., and Panani, A. (1975). Scand. 3. Huematol. 15, 187-191. Bull, J. ( 1975). Blood Cells 1, 161-162 (quoted by Brecher). Canellos, G. P., and Whang-Peng, J. ( 1972). Lancet 2, 1227-1229. Caspersson, T., Gahrton, G., Lindsten, J., and Zech, L. (1970). E r p . Cell Res. 63, 238-240.

216

JOACHIM MARK

Cervenka, J., and Koulischer, L. (1973). “Chromosomes in Human Cancer.” Thomas, Springfield, Illinois. Chen, T. R., and Shaw, M. W. (1973). Cancer Res. 33,2042-2047. Chu, E. W., Whang, J. J. K., and Rabson, A. S. (1966). J . Natl. Cancer Inst. 37, 885-891. Cohen, M. M., Ariel, and Dagan, J. ( 1974). Isr. J . Med. Sci. 10, 1393-1396. Cooper, E. H., Hughes, D. T., and Topping, N. E. (1966). Br. J . Cancer 20, 102-113. Crossen, P. E. (1975). Humangenetik 27, 151-156. Crossen, P. E., Mellor, J. E. L., Vincent, P. E., and Gunz, F. W. (1971a). Cytobios 4, 29-48. Crossen, P. E., Mellor, J. E. L., Finley, A. G., Ravich, R. B. M., Vincent, P. C., and Gunz, F. W. (1971b). Am. I. Med. 5 0 , 2 4 4 4 . Dameshek, W., and Gunz, F. W. ( 1964). “Leukemia,” 2nd ed. Grune & Stratton, New York. Dartnall, J. A., Mundy, G. R., and Baikie, A. G. (1973). Blood 42, 229-239. Davidson, W. M., and Knight, L. A. ( 1973). Lancet 1, 1510. de Grouchy, J., de Nava, C., Zittoun, R., and Bousser, J. (1966). Nouo. Reu. Fr. Hemtol. 6, 367388. de Grouchy, J., de Nava, C., Feingold, J., Bilski-Pasquier, G., and Bousser, J. (1968). Eur. 1. Cancer 4,481492. de la Chapelle, A,, Schrijder, J., and Vuopio, P. (1972). Clin. Genet. 3, 470-476. Dewald, G., Spurbeck, J. L., and Vitek, H. A. (1974). Mayo Clin. Proc. 49, 553557. Dinauer, M. C., and Pierre, R. V. (1973). Lancet 2,971. Ding, J. C., Adams, P. B., Patison, M., and Cooper, I. A. (1975). Cancer 35, 13251332. Elliot, A. Y., Cleveland, P., Cervenka, J,, Castro, A. E., Stein, N., Hakala, T. R., and Fraley E. E. ( 1974). J . Natl.Cancer Inst. 53, 1341-1349. Engel E., McCee, B. J., Flexner, J. M., Russel, M. T., and Myers, B. J. (1974). N . Engl. J . Med. 291, 154. Erkman-Balis, B., and Rappaport, H. (1974). Cancer 34,626-633. Ezdinli, E. Z., Sokal, J. E., Crosswhite, L., and Sandberg, A. A. (1970). Ann. Intern. Med. 72, 175-182. Falor, W. H., and Ward, R. M. (1973). J . Am. Med. Assoc. 226, 1322-1327. Fialkow, P. J., Gartler, S. M., and Yoshida, A. (1967). Proc. Natl. Acad. Sci. U.S.A. 58, 1468-1471. Fitzgerald, P. H., and Hamer, J. M. ( 1971). Blood 38,325-335. Fitzgerald, P. H., Pickering, A. F., and Eiby, J. R. (1971). Br. J . Haematol. 21, 473-480. Flandrin, G., and Bernard, J. ( 1975). Blood Cells 1,7-15. Fleischman, E. W., and Prigogina, E. L. (1975). Humangenetik 26, 335-342. Fleischman, E. W., Prigogina, E. L., Platonova, G. M., Chudina, A. P., Kruglova, G. V., and Kaverzneva, M. M. (1974). Neoplasm 21,51-61. Fleischmann, T., Hikansson, C. H., and Levan, A. (1971). Hereditas 69, 311314. Fleischmann, T., HHkansson, C . H., Levan, A., and Moller, T. (1972). Hereeditas 70, 243-258. Foerster, W., Medau, H. J., and Uffler, H. (1974). KZh. Wochenschr. 52, 123-126. Ford, C. E., Jacobs, P. A,, and Lajtha, L. G. (1958). Nature (London) 181, 15651568. Gahrton, G., Lindsten, J., and Zech, L. (1973). Exp. CeU Res. 79, 246-247. Gahrton, G., Lindsten, J., and Zech, L. (1974a). Blood 43, 837-840. Gahrton, G., Zech, L., and Lindsten, J. (1974b). Erp. Cell Res. 86, 214-216.

BANDING PAnERNS IN HUMAN NEOPLASMS

217

Gahrton, G., Lindsten, J., and Zech, L. ( 1 9 7 4 ~ )Acta . Med. Scand. 190, 355460. Gahrton, G., Lindsten, J., and Zech, L. (1974d). Proc. SOC. Med. Suec., HE 3, 119. Gahrton, G., Franzh, S., Killander, D., Lindsten, J., and Zech, L. (1975). Suen. Ldk-siillsk. Handl. 84, HE 20, 118. Garson, 0. M., and Milligan, W. J. (1972). Scand. J. Haemutol. 9, 186-192. Garson, 0. M., and Milligan, W. J. (1974). Scand. J. Haematol. 12, 256-262. Goh, K. O., Swisher, S. N., and Herman, E. C. (1967). Arch. Intern. Med. 120, 214-219. Golomb, H. M., Vardiman, J., and ltowley, J. D. (1976). Bbod (submitted for publication). Granberg, I., Gupta, S., and Zech, L. (1973). Hereditus 75, 189-194. Granberg, I., Hast, R., Reizenstein, P., and Skirberg, K.-0. (1975). Sven. Ldksallsk. Handl. 84, 4, HE 7, 112. Gripenberg, U., Levan, A,, and Clifford, P. (1969). lnt. J. Cancer 4, 334-349. Gropp, H., Pera, F., Lohmann, H., and Wolf, U. (1967). 2. Krebsforsch. 69, 326-334. Harnden, D. G. (1974). In “Chromosomes and Cancer” (J. German, ed.), pp. 619636. Wiley, New York. Hauschka, T. S., and Levan, A. (1953). Exp. Cell Res. 4,457-467. Hayashi, I., and Schmid, W. (1975). Humungenetik 30, 135-141. Hayata, I., Kakati, S., and Sandberg, A. A. ( 1973). Lancet 2, 1385. Hayata, I., Sakurai, M., Kakati, S., and Sandberg, A. A. (1975). Cancer 36, 11771191. Hays, T., Humbert, J. R., Peakman, D. C., Hutter, J. J., Morse, H. G., Robinson, A,, and August, C. S. (1975). Humangenetik 29, 259-264. Hecht, F., McCaw, B. K., Peakman, D., and Robinson, A. ( 1975). Nature (London) 255, 243-244. Hellstrom, K,, Hagenfeldt, L., Larsson, A., Lindsten, J,, Sundelin, P., and Tiepolo, L. ( 1971). Scand. J. Haemutol. 8,293406. Hook, E. B., Hatcher, N. H., and Calka, 0. J. (1975). Humangenetik 30, 251-257. Hossfeld, D. K. ( 1974a). Nature (London) 249, 864. Hossfeld, D. K. (1974b). Humongenetik 23, 111-118. Hossfeld, D. K. (1975). Z. Krebsforsch. 83, 269-273. Hossfeld, D. K., and Schmidt, C. G. (1973). In “Chemotherapy of Cancer Dissemination and Metastasis” (S. Carattini and G. Franchi, eds.), pp. 223-234. Raven, New York. Hossfeld, D. K., and Wendehorst, E. ( 1974). Acta Haematol. 52,232-237. Hossfeld, D. K., Bremer, K., Meusers, P., Wendehorst, E., and Reis, H. E. (1975a). 2. Krebsforsch. 84, 4957. Hossfeld, D. K., Torniey, D., and Ellison, R. R. (1975b). Cancer 36, 576-581. Hossfeld, D. K., Hijfien, K., Schmidt, C. G., and Diedrichs, H. (1976). Lancet 1, 198. Hsu, L. Y. F., Papenhausen, P., Greenberg, M. L., and Hirschhorn, K. (1974a). Acta Haematol. 52, 61-64. Hsu, L. Y. F., Alter, A. V., and Hirschhorn, K. (1974b). C h . Genet. 6, 258-264. Huang, C. C., lmamura, T., and Moore, G. E. (1969). J. Natl. Cancer Inst. 43, 11291146. Huang, C. C., Minowada, J., Smith, R. T., and Osunkoya, B. 0. (1970). J. Natl. Cancer lnst. 45, 815-829. Huang, C. C., Hou, Y., Woods, L. K., Moore, C. E., and Minowada, J. (1975). J . Natl. Cancer Inst. 53, 655-660. Hughes, D. T. (1968). Nature (London) 217,518-523.

218

JOACHIM MARK

Hurdle, A. D. F., Carson, 0. M., and Buist, D. G. P. (1972). Br. J. Haematol. 27, 773-782. lkeuchi, T., Minowada, J., and Sandberg, A. A. (191). Cancer 28, 499-512. Ishihara, T., Kohno, %-I., and Kumatori, T. ( 1974). Br. J . Cancer 29, 340442. Jacobs, E. M., Luce, J. K., and Cailleau, R. (1966). Cancer 19, 869-876. Jacobs, P. A., Tough, I. M., and Wright, D. H. (1963). Lancet 2, 1144-1146. Jarvis, J. E., Ball, G., Rickinson, A. B., and Epstein, M. A. (1974). lnt. J. Cancer 14, 716-721. Jonasson, J., Gahrton, G., Lindsten, J., Sinionsson-Lindemalm, C., and Zech, L. (1974). Blood 43, 557-563. Kakati, S., Hayata, I., Oshimura, M., and Sandberg, A. A. (1975). Cancer 36, 1729-1738. Klein, G. (1975). N . Engl. J. Med. 293, 1353-1357. Knight, L. A., Davidson, W. M., and Cuddigan, B. J. ( 1974). Lancet 1, 688. Kohn, G., Mellnian, W. J., Moorhead, P. S., Loftus, J., and Henle, G. (1967). J. Natl. Cancer lnst. 38, 209-222. Kohn, G., Diehl, V., Mellnian, W. J., Henle, W., and Henle, G. (1968). J. Natl. Cancer lnst. 41, 795-804. Kohn, G., Manny, N., Eldor, A., and Cohen, M. M. (1975). Blood 45, 653-657. Krogh Jensen, M., and Philip, P. (1973). Acta Med. Scand. 193, 353-357. Krogh Jensen, M., Eriksen, J., and Djernes, B. W. (1975). Scand. J. Huematol. 14, 201-209. Kurita, Y., Osato, T., and Ito, Y. (1968). J. Natl. Cancer Inst. 51, 1355-1366. Lawler, S. D., Millard, R. E., and Kay, H. E. M. (1970). Eur. J. Cancer 6, 223-233. Lawler, S. D., Lobb, D. S., and Wiltshaw, E. (1974). Br. J. Haematol. 27, 247-252. Lawler, S. D., Reeves, B. R., and Hamlin, I. M. E. (1975). Br. J. Cancer 31, 162167. Lin, C. C., and Goldstein, S. ( 1974). J . Natl. Cancer lnst. 53,298-304. Lobb, D. S., Reeves, B. R., and Lawler, S. D. (1972). Lancet 1, 849-850. Lutzner, M. A,, Emerit, I., Durepaire, R., Flandrin, G., Crupper, C., and Prunieras, M. ( 1973). J. Natl. Cancer lnst. 50, 1145-1162. McCaw, B. K., Hecht, F., Hamden, D. G., and Teplitz, R. L. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 2071-2075. Macek, M., and Benyesh-Melnick, M. (1969). Bacteriol. Proc. p. 154. Macek, M., Seidel, E. H., Lewis, R. T., Brunschwig, J. P., Wiberly, I., and BenyeshMelnick, M. ( 1971). Cancer Res. 31, 308321. Manolov, G., and Manolova, Y. ( 1972). Nature (London) 237,3334. Manolov, G., Levan, A,, Nadkami, J. S., Nadkarni, J., and Clifford, P. (1970). Hereclitas 66, 79-100. Manolov, G., Manolova, Y., Levan, A,, and Klein, G. (1971). Hereditas 68, 160-163. Mark, J. ( 1969a). Acta Neuropathol. 14, 174-184. Mark, J. ( 1969b). Eur. J . Cancer 5,307-315. Mark, J. (1970a). Hereditas 65, 59-82. Mark, J. ( 1970b). Eur. J. Cancer 6, 489-498. Mark, J. ( 1973a). Acta Pathol. Microbiol. Scand., Sect. A 81, 588-590. Mark, J. ( 1973b). Hereditas 75,213-220. . Neuropathol. 25, 46-53. Mark, J. ( 1 9 7 3 ~ )Acta Mark, J. ( 1974). I n “Chromosomes and Cancer” (J. German, ed.), pp. 497-517. Wiley, New York. Mark, J. (1975a). Eur. J. Cancer 11, 815-819.


219

Mark, J. (1975b). Hereditus 81, 289-292. Mark, J., Levan, G., and Mitelman, F. ( 1972a). Hereditas 71, 163-168. Mark, J., Mitelman, F., and Levan, G. (1972b). Acta Pathol. Microbiol. Scand., Sect. A 80, 812-820. Mark, J., Mitelman, F., Dencker, H., Norryd, C., and Tranberg, K.-G. ( 1 9 7 2 ~ )Acta . Pathol. Microbiol. Scand., Sect. A 81, 85-90. Mark, J., Pontkn, J., and Westermark, B. (1974a). Humangenetik 22, 323-326. Mark, J., Pontkn, J., and Westermark, B. (1974b). Acta Neurqpathol. 29, 223-228. 78, 304-308. Mark, J., Pontkn, J., and Westermark, B. ( 1 9 7 4 ~ )Hereditus . Mayall, B. H., Carrano, A. V., and Rowley, J. D. (1974). Clin. Chem. 20, 10801085. Miles, C. P. (1973). Nutl. Cancer Inst., Monogr. 36, 197-201. Miles, C. P., and ONeill, F. ( 1967). Cancer Res. 27, 392402. Millard, R. E. (1968). Eur. J . Cancer 4, 97-105. Mitelman, F. ( 1974a). In “Chromosomes and Cancer” (J. German, ed.), pp. 675-693. Wiley, New York. Mitelman, F. ( 1974b).Hereditas 76, 315-316. Mitelman, F., and Brandt, L. (1974). Scand. 1. Haemutol. 13,321-330. Mitelman, F., Mark, J., Levan, G., and Levan, A. (1972). Science 176, 1340-1341. Mitelnian, F., Mark, J., Nilsson, P. G., Dencker, H., Norryd, C., and Tranberg, K.-G. ( 1974a). Hereditus 78, 63-68. Mitelman, F., Brandt, L., and Nilsson, P. G. (197413). Hereditus 78, 302404. . J. Haemutol. 13, 87-92. Mitelman, F., Brandt, L., and Nilsson, P. G. ( 1 9 7 4 ~ )Scand. Mitelman, F., Nilsson, P. G., and Brandt, L. (1975a). J. Natl. Cancer Inst. 54, 13191321. Mitelman, F., Levan, G., and Brandt, L. (1975b). Hereditus 80, 291-293. Moore, M. A. S., and Metcalf, D. (1973). Int. 1. Cancer 11, 143-152. Moore, M. A. S., Ekert, H., Fitzgerald, M. G., and Carmichael, A. (1974). Blood 43, 15-22. Motoniura, S., Ogi, K., and Horie, M. ( 1973). Acta Haematol. 49, 300-305. Muldal, S., and Lajtha, L. C. (1974). I n “Chromosomes and Cancer” (J. German, ed. ), pp. 451480. Wiley, New York. Muldal, S., Mir, M. A,, Freeman, C. B., and Geary, C. G. (1975). Br. J. Cancer 31, 364-368. Nadkarni, J. S., Nadkami, J. J., Clifford, P., Manolov, G., Fen@, E. M., and Klein, E. (1969). Cancer 23, 64-79. Nassar, V. H., and Khouri, F. P. ( 1974). Arch. Pathol. 98, 367-369. Nowell, P. C. (1971). Cancer 28, 513-518. Nowell, P. C., Jensen, J., and Gardner, F. (1975). Humangenetik 30, 13-21. Obara, Y., Makino, S., and Mikuni, C. (1970). Proc. Jpn. Acad. 46, 561-566. O’Riordan, M. L., Berry, E. W., and Tough, I. M. (1970). Br. J. Haematol. 19, 8390. O’Riordan, M., Robinson, J. A., Buckton, K. E., and Evans, H. J. (1971). Nature (London) 230, 167-168. Paris Conference. (1971). Birth Defects: Orig. Artic. Ser. 8, No. 7. Paul, B., Porter, I. H., Benedict, W. F. (1973). Humangenetik 18, 185-187. Pedersen, B. (1968). Acta Pathol. Microbiol. Scand. 72, 360-366. Pedersen, B. ( 1973). Eur. J. Cancer 9,509-513. Pedersen, B. (1975). Blood Cells 4,227-234. Petit, P., and Cauchie, C. (1973). Lancet 2, 94.

220

JOACHIM MARK

Petit, P., Verhest, A., Lecluse van der Bilt, F., and Jongsma, A. (1972). Pathol. Eur. 7, 17-21. Philip; P. ( 1975a). Hereditus 80, 155-156. Philip, P. (197513). Hereditus 81, 124-125. Philip, P., and Drivsholm, A. ( 1974). Biomedicine 21,42!3-430. Pierre, R. P., and Hoagland, H. C. ( 1972). Cancer 30, 889494. Porter, I. H., Benedict, W. F., Brown, C. D., and Paul, B. (1969). Exp. Mol. Pathol. 11, 340467. Potter, A. M., Sharp, J. C., Brown, M. J., and Sokol, R. J. (1975). Humangenetik 29, 223-228. Prigogina, E. L., and Fleischman, E. W. (1975a). Humangenetik 30, 109-112. Prigogina, E. L., and Fleischman, E. W. (1975b). Humungenetik 30, 113-119. Propp, S., and Lizzi, F. A. ( 1970). Blood 36,353-360. Rabson, A. S., O’Conor, G. T., Baron, S., Whang, J. J., and Legallais, F. Y. (1986). Int. 1. Cancer 1, 89-106. Raposa, T., Natarajan, A. T., and Granberg, I. ( 1974). J. Natl. Cancer Inst. 52, 1935-1 938. Rary, J. M., Bender, M. A., and Kelly, T. E. ( 1975). J. Hered. 66, 33-35. Reeves, B. R. ( 1973), Humungenetik 20,231-250. Reeves, B. R., Lobb, D. S., and Lawler, S. D. (1972). Humaqenetik 14, 159-161. Rowley, J. D. ( 1973a). Nature (London) 243,290-293. Rowley, J. D. (1973b). N. Engl. J. Med. 288,220-221. Rowley, J. D. ( 1 9 7 3 ~ )Lancet . 2,390. RowIey, J. D. (1973d). Ann. Genet. 16, 109-112. Rowley, J. D. ( 1973e). Lancet 2, 1385-1386. Rowley, J. D. ( 1974a). Lancet 2,835-836. Rowley, J, D. (1974b). J. Med. Genet. 11,166-170, Rowley, J. D. (1975a). Cancer 36, 1748-1757. Rowley, J. D. (1975b). Proc. Natl. Acud. Sci. U.S.A. 72, 152-156. Rowley, J. D., and Potter, D. (1976). Blood 47, 705-721. Saksela, E., and Moorhead, P. S. (1963). Proc. Nutl. Acad. Sci. U.S.A. 50, 390-395. Sakurai, M., and Sandberg, A. A. ( 1974). Cancer 33, 1548-1557. Sakurai, M., Oshimura, M., Kakati, S., and Sandberg, A. A. ( 1974).Lancet 2,227-228. Sandberg, A. A,, and Hossfeld, D. K. (1970). Annu. Reo. Med. 21, 379-408. Sandberg, A. A., and Hossfeld, D. K. (1974). In “Handbuch der allgemeinen Pathologie” (W. Altman et al., eds.), Vol. 6, Part 5, pp. 141-287. Springer-Verlag, Berlin and New York. Sandberg, A. A., and Sakurai, M. ( 1973). Lancet 1,375. Sandberg, A. A., Sakurai, M., and Hossfeld, D. K. (1973). Proc. Natl. Cancer Conf., 7th, 1973 pp. 333-343. Schmidt, R., Dar, H., Santorineou, M., and Sekine, I. ( 1975). Lancet 1, 1145. Schwarzenberg, L., Math& G., Pouillart, P., Weiner, R., Locour, J., Genin, J., Schneider, M., de Vassal, F., Hayat, M., Amiel, J. L., Schlumberger, J. R., Jasmin, C., and RosenfeId, C. ( 1973). Br. Med. J. 1,700-703. Seligmann, M. (1974). N. Engl. 1. Med. 290, 1483-1484. Sellyei, M., and Vass, L. (1975). Lancet 1, 1041. Shiffman, N. J., Stecker, E., Conen, P. E., and Gardner, H. A. (1974). Can. Med. ASSOC. J . 110, 1151-1154. Shiraishi, Y., Hayata, I., Sakurai, M., and Sandberg, A. A. (1975). Cancer 36, 199202.


221

Spiers, A. S. D., and Baikie, A. G. (1966). Lancet 1,506-510. Spiers, A. S. D., and Baikie, A. G. ( 1968). Cancer 22, 193-217. Spires, A. S. D., and Baikie, A. G. ( 1970). Br. J . Cancer 24, 77-91. Spiers, A. S. D., Baikie, A. G., Galton, D. A. G., Richards, H. G. H., Wiltshaw, E., Goldman, J. M., Catovsky, D., Spencer, J., and Peto, R. (1975). Br. Med. J . 1, 175-179. Steel, C. M. (1971). Nature (London) 233,555556. Stewart, S . E., Lovelace, E., Whang, J. J., and Ngu, V. A. (1965). J. Nutl. Cancer Inst. 34, 319327. Stryckmans, P. A. (1974). Semin. Hemutol. 11, 101-127. Tiepolo, L., and Zuffardi, 0. ( 1973). Cytogenet. Cell Genet. 12,8-16. Tomkins, G. A. (1968). lnt. J., Cancer 3, 644-653. Toshima, S., Takagi, N., Minowada, J., Moore, G. E., and Sandberg, A. A. (1967). Cancer Res. 27. 753-771. Tough, I. M., Hamden, D. G., and Epstein, M. A. (1968). Eur. J. Cancer 4, 637646. Trujillo, J. M., Cork, A., Hart, J. S., George, S. L., and Freireich, E. J. (1974). Cancer 33, 824-834. Tsuchimoto, T., Buhler, E. M., Stalder, G. R.,Mayr, A. C., and Obrecht, J. P. (1974). Lancet 1,566. Vallejos, C. S., Trujillo, J. M., Cork, A., Bodey, G. P., McCredie, K. B., and Freireich, E. J. (1974). Cuncer 34, 1806-1812. van Biervliet, J. B., van Hemel, J., Geurts, K., Punt, K., and de Boer-van Wering, E. (1975). Lancet 2, 617. van den Berghe, H. (1973). Lancet 2, 1030-1031. van den Berghe, H., Cassirnan, J.-J., David, G., Fryns, J.-P., Michaux, J.-L., and Sokal, G. ( 1974). Nature (London) 251,437-438. Vass, L., and Sellyei, M. (1973). Lancet 1, 550-551. Verhest, A,, and van Schoubroeck, F. ( 1973). Lancet 2, 1386. Visfeldt, J. ( 1971). A d a . Puthd. Microbiol. Scand., Sect. A 79,513523. Walker, L. M. S. (1971). Br. J . Huematol. 21, 455-461. Warburton, D., and.Bluming, A. ( 1973). Blood 42,799-804. Weiss, A. F., Portmann, R., Fischer, H., Simon, J., and Zang, K. D. (1975). Proc. Natl. Acud. Sci. U.S.A. 72, 609-613. Welch, J. P., and Lee, C. L. Y. ( 1975). Nature (London) 255,241-242. Whang-Peng, J., Canellos, G. P., Carbone, P. P., and Tjio, J. H. (1968). Blood 32, 755-766. Whang-Peng, J., Gerber, P., and Knutsen, T. (1970). 1. Natl. Cancer lnst. 45, 831839. Whang-Peng, J., Knutsen, T. A., and Lee, E. C. ( 1973). J . Natl. Cancer lnst. 51, 2009-2012. Whang-Peng, J., Lee, E. C., and Knutsen, T. A. ( 1974). J. Nut. Cuncer Inst. 52, 1035-1036. Whaun, J. M., Lin, C. C., Dundas, J. B., and Cornish, S. (1974). Proc. Congr. lnt. SOC. Hematol., 15th, 1974, Abstract, p. 166. Whittaker, J. A., Davies, P., and Khurshid, M. (1975). A d a Huematol. 54, 350-357. Wisniewski, L., and Korsak, E. ( 1970). Cancer 25, 1081-1086. Wurster-Hill, D. H., McIntyre, 0. R., Cornwell, C. G., 111, and Maurer, L. H. (1973). Lancet 2, 1031.

222

JOACHIM MARK

Wurster-Hill, D., Whang-Peng, J., McIntyre, 0. R., Hsu, L. Y. F., Hischhorn, K., Modan, B., Pisciotta, A. V., Pierre, R., Balcerzak, S . P., Weinfeld, A., and Murphy, S. (1976). Semin. Hematol. 13, 13-32. Yunis, J. J., ed. (1974). “Human Chromosome Methodology,” 2nd ed. Academic Press, New York. Zaccaria, A,, Baccarani, M., Barbieri, E., and Tura, S . (1975). Eur. J. Cancer 11, 123-128. Zajac, B. A., and Kohn, G. ( 1970). J. Natl. Cancer lnst. 45,399-405. Zankl, H., and Zang, K. D. ( 1972). Humangenetik 14,187-189. Zankl, H., Seidel, H., and Zang, K. D. (1975a). Humangenetik 27, 119-128. Zankl, H., Weiss, A. F., and Zang, K. D. (1975b). Humangenetik 30,343448. Zech, L., Lindsten, J., Udkn, A.-M., and Gahrton, G. (1975). Scand. J. Haematol. 15, 251-255. Zech, L., Haglund, U., Nilsson, K., and Klein, G. (1978). lnt. J . Cancer 17, 47-56. Addendum

The work reported herein was completed in January 1978. Since then some further reports concerning banding observations in human neoplasms have been published (for instance, a few cases of Ph’-positive CML with translocations other than the “standard type, some cases of AML, AMML, EL, and dysplastic anemias, a few non-BL, single cases of malignant exudates, and some old or newly established cell lines). These contributions mostly confirm or support conclusions outlined in this chapter. The new observations do not change any of the interpretations or .assessments presented here, and, thus, these data have not been included.

TEMPERATURE-SENSITIVE MUTATIONS IN ANIMAL CELLS

Claudio Basilico Deportment of Potholow, New York University School of Medicine, New York, New York

. . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

I. Introduction 11. ts Growth Mutants . . . . . . . . . . . . . . . A. Methods of Induction and Selection B. Frequency and General Behavior . . . . . . . . . . C. Genetic Analysis D. Characterization 111. ts Mutations Affecting the Expression of Specialized Functions Mutations Affecting the Expression of Neoplastic Transformation in Vitro IV. Nature of the ts Mutations Described in Animal Cells V. Conclusions References Note Added in Proof . . . . . . . . . . . . . .

. . . . . .

. . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

223 225 226 230 232 236 253 255 261 262 264 266

1. Introduction

The natural occurrence of temperature-sensitive ( ts ) phenotypes has been known for decades. All organisms have an optimum temperature for growth and proliferation, and this varies widely from species to species. For example, avian cells are capable of sustained multiplication at rather high temperatures ( -42OC) while amphibian cells do not grow well above 3OOC. This type of natural temperature sensitivity has not, however, been exploited in biological research since it is likely to involve several different processes, and within a species there is no temperatureinsensitive wild-type (wt) phenotype to which the ts can be compared. In contrast, ts mutations are unusual occurrences; they result in a ts phenotype in an organism which in the wt state is capable of normal function at the temperature used. A temperature-sensitive mutation can be defined as a mutation that imposes a restriction in the temperature of growth or normal function in an organism which in the wt state does not exhibit such a restriction. This is generally caused by an impairment, expressed only at the restricted temperature, of a specific cell function. The most common ts mutations are heat sensitive: that is the defect is expressed only at high temperature. Thus, the use of the term ts generally implies heat-sensitive mutations. Although this is not formally correct, I will continue to use it here. Ts mutations that restrict growth at low, but 223

224

CLAUD10 BASILICO

not at high, temperature are also known. They are generally called cold sensitive.’ A ts mutant then is defined by two sets of conditions. A low, permissioe temperature at which growth or other cell function is normal, and a high, nonpermissive temperature at which either growth or a specific cell function is defective. The wt cell behaves normally under both sets of conditions. The occurrence of ts mutations in bacteria and Neurospora has been known for at least 25 years, Horowitz (1950) isolated a large number of mutants of Neurosporu, in which the requirement for growth of a given metabolite or amino acid was expressed only at high temperature. Therefore, these mutants belong to the class of ts auxotrophs. He also was able to isolate similar mutants from Escherichia coli K12, and used these mutations to gain basic genetic information, bearing in particular on the one gene-one enzyme hypothesis (Horowitz and Leupold, 1951). Maas and Davis (1952) obtained E . coli ts mutants that required pantothenate for growth only at high temperature, and they were able to show that the pantothenate-synthesizing enzyme derived from these cells was extremely heat labile. Thus the hypothesis that this type of mutation entailed at least in some cases the formation of an altered protein that could still be active, but was more heat labile than the wt product, was already generally accepted when Campbell (1961) published his studies on “sensitive” mutants of h, a classification under which he grouped mutants of the ts, suppressorsensitive, and pH-sensitive type. The use of ts mutations for the study of the genetics of microorganisms did not, however, become widespread until about 1963, when Epstein, Edgar, and their colleagues published their papers on the study of conditional lethal mutations in T4 phage (Epstein et al., 1963; Edgar et al., 1964; Edgar and Lielausis, 1964). Since then, ts mutations have become one of the standard tools of molecular genetics, and much about their nature is known. The defect in ts mutants is usually attributable to a single amino acid substitution in a protein (Drake, 1970), although ts mutations resulting from alterations of transfer RNA have been described (Yamamoto et al., 1972). As a result of such a mutation, the affected protein apparently retains at least partial function at low temperature, but loses its functionality at certain high temperatures that do not affect the wt gene product. The experimental evidence for this is limited, since only in a The isolation of cold-sensitive mutants of mammalian cells has recently been reported (Farber and Unrau, 1975). Six cold-sensitive variants of Chinese hamster cells have been isolated; one of them has been characterized in some detail, and shows a rapid cessation of DNA and protein synthesis upon shift from the permissive (39”) to the nonpermissive (33”) temperature. The defect of this mutant does not appear to involve ribosomal subunits assembly,

TEMPERATURE-SENSITIVE CELL MUTATIONS

225

few cases has the gene product affected been isolated and characterized in oitro (see Drake, 1970). However, there are no compelling reasons to doubt the general validity of these findings. The usefulness of ts mutations for a variety of studies stems mainly from the following: (1) They belong to the class of conditional lethal mutations, i.e., mutations expressed only under conditions that the experimenter can control. Thus, this type of mutation can be used to study essential functions, whose defectiveness would be lethal in a nonconditional state. ( 2 ) They have been shown to map all over the genome of the organisms studied, and therefore, they can be used to study a variety of cell functions (Edgar and Lielausis, 1964). ( 3 ) The switch from the permissive to the nonpermissive condition (and vice versa) can easily be achieved by manipulating the temperature of incubation; this permits a direct internal control of the experiments and provides a sensitive and precise way of testing the involvement of a specific gene-product in determining a given phenotype, In the past 12 years conditional lethal ts mutations have been used to attack fundamental problems of genetics, physiology, and molecular biology. After the initial work on bacteriophages, ts mutants of bacteria, yeast, and drosophila have been obtained. Ts mutations in animal viruses have also been obtained and vigorously exploited. Recently, several investigators have tried to develop a system of ts mutations in somatic animal cells, in order to study their genetics, physiology, and growth regulation. It is the purpose of this chapter to review and evaluate the present state of affairs in this field. In general, one can distinguish two types of ts cell mutations. The first type is in a function that is essential for growth and cell division. Thus such mutants are incapable of growth at the nonpermissive temperature. The second type of mutation affects specialized or salvage functions. The ts mutation does not interfere with growth at high temperature, but will inhibit the expression of one or more specialized functions, resulting in an altered, but not lethal phenotype. The first type of mutants has been studied more extensively, but a few examples of the second type are also known. I wish to point out that these cell variants may not represent mutations in a strict sense in all cases. For the sake of simplicity, however, I will refer to them generally as “mutants.” Whether this is justified will be discussed later. II. ts Growth Mutants

The first ts growth mutants of animal cells were isolated from BSC-1 monkey cells (Naha, 1969) and from the L line of mouse cells

226

CLAUD10 BASILICO

(Thompson et al., 1970). Subsequently ts mutants have been isolated from a variety of cell lines including Chinese hamster cells, Syrian hamster cell lines, and mouse cells. To my knowledge, ts mutants of human cells have not yet been obtained, although one sees no reason why this should not be possible. A. METHODS OF INDUCTION AND SELECTION Temperature-sensitive mutations are believed to be missense mutations. Although this has not been conclusively demonstrated for animal cell mutants, there is no compelling reason to doubt this assumption. Consequently, most investigators have used substances that have been shown to act as missense mutagens in microorganisms to induce ts mutations in animal cells. In practice, only two mutagens have been generally used, namely N-methyl-N-nitrosoguanidine ( NG) and ethane methyl sulfonate (EMS ) . Most authors have used these mutagens at doses that cause a lethality of 70-903, followed by a variable period of mutation “fixation” time at the permissive temperature (from 3 to 8 cell generations) (for a review of the more technical aspects of this subject, see Thompson and Baker, 1973; Basilico and Meiss, 1974). Very little information is available on the effectiveness of the mutagenic treatment. Most workers have not bothered to compare frequency of mutants in treated vs untreated populations, mainly because of the difficulties in quantitating the yield of mutants in terms of mutation rate, etc. (see later). Thus it cannot be stated with any certainty whether the mutagens used were effective in inducing ts mutations, or whether the mutants isolated preexisted in the population, or were in any other way spontaneous. Good evidence in favor of the effectiveness of the mutagens is the fact that, in most cases tested (Meiss and Basilico, 1972; Thompson et al., 1973) mutagens were definitely effective in inducing reversion of the ts mutations (Table I ) . However, Smith and Chu (1973) examined the frequency of ts mutants in Chinese hamster cells with or without exposure to EMS. They used a replica plating method that has the advantage of eliminating the possible losses of mutants produced by selection, The disadvantage of the method is that the size of the cell populations that can be examined is quite limited. No significant increase in the yield of ts mutants was found after EMS treatment. However, the frequency of ts mutants that these authors obtained was very high, and most of the mutants were leaky. It is probably impossible to base any conclusion on the isolation of leaky, or density-dependent (ts phenotype suppressed at high cell densities) ts mutants, since their genetic nature is unknown. If one takes

227


TABLE I EFFECTOF NITROSOGLJANIDINE O N THE REVERSION FREQUENCY OF ts BHK C E L L S ~ J

ts strain 4223

AF8 BCH

Treatment None Nitrosoguanidine None Nitrosoguanidine None Nitrosoguanidine

Number of cells plated a t 39°C Number of ( X lo6) revertants 86 86 10 15 10 10

3 40 6

54 7 50

Reversion frequency ( X 10-6) 0.035

0.46 0.6 3.6

0.7 5

From Meiss and Basilico (1972), with permission. Cells growing a t 33°C were exposed t o a nitrosoguanidine concentration of 1 pg/ml for 16 hours, which, in these experiments, resulted in -60% lethality. They were incubated for another 2 days a t 33”C, then replated a t 39°C a t 3 to 4 X lo8 cells/ 100-mm petri dish. Cells were fixed and stained after 14 days. Control cells were carried through the same procedure without addition of the mutagen. (I

into account only “good ts mutants, the data from Smith and Chu show that they obtained two mutants from EMS-treated cultures and none from the untreated. Clearly this matter needs further experimentation. After a mutagenesis and mutation “fixation” time, most authors have used a method for selection of ts mutants from the mutagenized cell population. Selection has the theoretical advantage of enriching the cell population for rare or specific mutants, but it has the disadvantage of complicating the interpretation of the results in terms of mutation rate, preferential sites for mutagenicity, etc. This disadvantage is negligible if one is interested only in obtaining mutants. The selection methods used have in general been based on the establishment of conditions that are lethal to wild-type cells, but not to the mutants. Conditions can be devised for killing wt cells at the nonpermissive temperature, taking advantage of the fact that they would multiply normally, while the ts mutants are arrested in growth and division. The method generally used consists of shifting the cells to the nonpermissive temperature and exposing them to agents that are lethal for dividing cells, or for cells performing certain essential functions (e.g., DNA synthesis). If all goes well, this results in killing the wt population, but not the mutant, which at this temperature is not dividing. After the treatment, the cells are shifted back to the permissive temperature to allow growth of the ts mutants (for more details, see Thompson and Baker, 1973; Basilico and Meiss, 1974).

228

CLAUD10 BASLLICO

The following killing agents have been used: 1. DNA synthesis inhibitors, such as 5-fluorodeoxyuridine ( FUdR ) and cytosine arabinoside (Thompson et al., 1970; Meiss and Basilico, 1972; Wang, 1974). These drugs act by inhibiting DNA synthesis (Graham and Whitmore, 1970; Cohen et al., 1958) and are lethal for DNA-synthesizing cells in exponential growth ( Heidelberger, 1965; Pollack et al., 1968; Graham and Whitmore, 1970). Their mechanism of killing is not well understood. Since these drugs are generally not harmful to non-DNA-synthesizing cells (Pollack et al., 1968), they have been used in the hope of selecting mutants that do not synthesize DNA at the nonpermissive temperature (DNA-), Their main disadvantage is that they require rather long times (12-24 hours) to produce efficient cell killing, and as such they are not very useful for selections of mutants that do not withstand long exposures to the nonpermissive temperature. 2. DNA “poisons,” such as aH-labeled thymidine or 5-bromodeoxyuridine ( BUdR), followed by exposure to light (Thompson et al., 1970; Naha, 1969; Liskay, 1974; Wittes and Ozer, 1973; Roufa and Reed, 1975). They are incorporated into the DNA of DNA-synthesizing cells, and lethality of [3H]thymidine, is due to the internal radioactive decay of the 3H atoms (Person, 1963), whereas lethality of BUdR is probably due to the fact that the bromouracil-substituted DNA is much more sensitive to UV or near-UV light than ordinary DNA. Death then results from the formation of strand breaks and chromosomal aberrations (Puck and Kao, 1967; Chu et al., 1972). These substances have also been used to select for DNA- ts mutants. 3. Incorporation of tritiated amino acids into proteins, to select for protein synthesis, or tRNA synthetase mutants (Thompson et al., 1975). This method is quite effective, especially if accompanied by a relative shortage of amino acids in the medium (see later and Fig. 4 ) . 4. Combinations of the above treatments, such as [ 3H]thymidine incorporation followed by exposure to cytosine arabinoside ( Thompson et al., 1970; Wang, 1974). 5. A different type of selection, not based on the preferential killing of wt cells, has been used by Roscoe et al. (1973a). These authors have used a brief treatment of hydroxyurea, which blocks DNA synthesis and prevents cycling wt cells from reaching mitosis, followed by mitotic detachment, and plating of the mitotic cells at low temperature. This method was used in the hope’of enriching for mutants blocked in mitosis at the nonpermissive temperature. In general, all these methods of selection have yielded mutants, although not always of the predicted type. The first types of selection used methods that should have selected for DNA- mutants, but a number


229

of limitations were imposed on the selection procedure. The following in my opinion are the most important: ( a ) a generally long “mutation expression time” was allowed at the nonpermissive temperature before addition of the selective agent, thus favoring mutants that had delayed cessation of growth. ( b ) The total time of exposure to the nonpermissive temperature was quite long, favoring mutants with high survival at 39OC (although some mutants isolated were not of this type), ( c ) Recycling was often used. These factors make it quite difficult to assess the efficacy of the selection systems used when measured as yield of specific type of mutants. A quick survey of the mutants obtained with selections designed for DNAmutants shows a number of surprising results: a large number of mutants that do not have a DNA- phenotype (Meiss and Basilico, 1972; Basilico and Meiss, 1974; Thompson and Baker, 1973); a mutant affected in tRNA synthetase, which has a DNA- phenotype, but dies extremely fast at 39OC, and which therefore should not have survived the selection technique (Thompson et d.,1973); and only a minority of DNAmutants. Thus, a skeptical reader may tend to doubt the value of any such selection, and wonder whether random testing of mutagenized cells might not have given the same results. Such a pessimistic view is, however probably unjustified, although there are two reports of substantially high frequencies (lo-? to 10-3)of ts mutants isolated after random testing of mutagenized cells (Smith and Chu, 1973; Thompson et al., 1971). No good characterization of these mutants has ever been published, however, and in most cases they are reported to be not very “clean” (see the discussion of the nature of the ts mutations). Furthermore, in the few cases tested, reconstruction experiments showed that the selection system was effective in enriching for mutants, even for mutants whose phenotype in principle should not have been selected for (Basilico and Meiss, 1974). This suggests that many unknown factors may determine the ability of cells to survive certain types of selection, and that it is our ignorance of these factors that makes us consider some of the results obtained with a given selection procedure unexpected. It has also been observed that recycling through the selection method is effective in enriching for “good mutants (i.e., with a clear-cut ts phenotype) even if the frequency of total mutants does not increase appreciably from one to three cycles (Meiss and Basilico, 1972; Basilico and Meiss, 1974). Since the leaky, or high reversion, mutants are generally of little value, it would appear that the selection procedure used have been effective in selecting for mutants with a clear ts phenotype. The main problem that remains to be solved is that of good specificity,

230

CLAUD10 BASILICO

and most investigators are now addressing their efforts to this purpose. Preliminary results obtained by using synchrony methods, coupled with short exposure of the cells to high temperature, suggest that different classes of mutants can be isolated. These mutants have low survival at 39OC, and appear to have a DNA- phenotype (H. K., Meiss, personal communication). If these results were confirmed, they would provide a plausible answer to the question of why DNA- mutants were rarely isolated in the previous attempts. Other more complicated methods of selection are now being used in the attempt, for example, to select for specific cell cycle defects. It is hoped that all these efforts will produce satisfactory selection systems. As a conclusion, it is quite evident that it would be fruitless to attempt interpretations of mutation rates and mutation induction on the basis of the frequency of the mutants isolated, The selection techniques used so far have undoubtedly resulted in the killing of a good proportion of ts mutants, and the only two experiments carried out without selection, of which I am aware (Thompson et al., 1971; Smith and Chu, 1973), yielded a number of mutants too low to permit statistical evaluation. AND GENERAL BEHAVIOR B. FREQUENCY

As discussed before, the frequencies of ts mutants obtained by different laboratories vary widely. Reported figures range from for nonfor different types of selecselective isolation procedures to 10-u to tion (Thompson and Baker, 1973; Basilico and Meiss, 1974). Apart from the frequencies, there seems to be general consensus that all ts clones isolated fall into four broad categories (Table 11): 1. “Good mutants: These cells plate at the permissive temperature with an efficiency similar to that of wt cells, and their growth rate is normal. At the nonpermissive temperature, these mutants do not form colonies or multiply extensively irrespective of the size of the inoculum plated. Growth is arrested after a variable time following shift to the nonpermissive temperature, but in general these mutants do not perform more than one average division at the high temperature. Reversion, as measured by the ability to form colonies at the nonpermissive temperature, is low ( - 0.05.

allow correlation of DNCB reactivity with prognosis or clinical stage. It is apparent that additional studies are required in melanoma patients. Careful quantitation, with biopsy of equivocal responses, is necessary if we are to accurately assess the biological and clinical significance of these data. 2. Response to Recall Antigens Unlike DNCB sensitization, interpretation of the results of testing with a single battery of recall microbial antigens is complicated by the variability of exposure to these antigens. Further comparison of data between studies is difficult because of the variation in criteria for a positive test and variability in the number and kinds of antigens used. Nonetheless, ready availability and ease of application have led to fairly extensive testing in melanoma patients. Results are summarized in Table XI. Seigler et al. (1972) evaluated skin test reactivity to 23 microbial antigens in 20 patients with metastatic melanoma who had undergone extensive prior therapy ( combinations of surgery, irradiation, limb perfusion, and chemotherapy). Reactivity was assessed by determining the geometric mean cutaneous response to the 23 antigens in each cancer patient. These data were compared to those obtained from 138 nonanergic controls. Reactivity in melanoma patients was significantly depressed at all time intervals from 0.25 to 72 hours. At 48 and 72 hours, no melanoma patients fell within the normal range. However, only 6/20 patients failed to manifest delayed cutaneous hypersensitivity reactions to all 23 test antigens. This comprehensive study clearly demonstrates that reactivity to recall antigens is depressed in melanoma patients with diffuse metastases and extensive prior treatment.

319

HUMAN CUTANEOUS MALIGNANT MELANOMA

TABLE X I RESPONSEOF MELANOMA PATIENTS TO SKIN TESTING WITH MICROBIAL ANTIGENS Edber and Morton (1970)"

Operable and/or disease free 6 months 21 Positive of 6

26/47

(55%)

Inoperable and/or recurred within 6 months 6/29 (20%)

p

=

0.01

p

=

NS'

GTOSS and Eddie-Quartey (1976)*

No recurrence 1 year Mean No. positive k 1 S.D.

2 . 2 7 f 1.02

Recurrence

< 1year

1 . 8 0 k 1.34

Zieglor et al. (1969)

Stage I

2 1 Positive of 5

(100%)

8/8

Stage I1 and I11 8/11 (73%)

p = NS

Golub et al. (1974)

Melanoma pts.d

2 2 Positive of 4

24/28 (86%)

Seigler et al. (1978)

Melanoma pts. Controls (138) G M R D (mm)

< O . 10

0.25

Regressors

Progressors

Gutterman et al. (1974)

2 2 positive of 6 2 3 positive of 6

17/21 11/21

(81%) (52%)

23/35 (66%) 10/35 (28%)

p = NS p = NS

Mastrangelo ct al. (1976a)

Mean No. positive of 5

Regressors (5)

Nonresponders (10)

1.4

1.5

18/100 patients with melanoma. 12/26 patients with melanoma. Number of patients in group. d Pts. = patients. c GMRD, geometric mean reaction diameter (mm) to 23 antigens. I NS = p > 0.05. a

b

p = NS

320

WALLACE H. CLARK ET AL.

Ziegler et al. ( 1969) found reactivity to recall antigens in 161 19 patients ( 16 with local or regional disease). Golub et al. ( 1974) elicited reactivity to recall antigens in 24/28 melanoma patients. In these two studies, patients with regional and disseminated disease were not compared for reactivity. However, it is unlikely that a significant difference could be demonstrated in view of the overall high level of reactivity, Although it would appear from the studies of Ziegler et al. and Golub et al. that skin reactivity to recall antigens is preserved in melanoma patients with lesser tumor burdens and/or less extensive prior treatment, the lack of comparison with a normal control may have precluded detection of a more modest impairment. Several investigators have attempted to correlate response to recall antigens with prognosis. Eilber and Morton (1970) noted that 55%of patients operable and disease-free 6 months were positive to one or more microbial antigens on preoperative skin testing, whereas only 20% of those who were inoperable or recurred early reacted. When comparing these data with those obtained with DNCB in the same patient population, the investigators concluded the DNCB reactivity more closely predicted the subsequent clinical course of the patients as only 8%of the patients who were disease-free were DNCB negative, but 45%of these patients were negative to all microbial antigens. However, the data for melanoma patients were not presented separately. Gross and EddieQuartey (1976) evaluated reactivity to six recall microbial antigens in 26 patients (14 lung cancer, 12 melanoma) clinically free of disease and correlated these data with the subsequent clinical course. No significant differences were noted in the mean number of positive skin responses to recall antigens among controls and patients with favorable and unfavorable clinical outcomes. Again, data for melanoma patients were not presented separately. Several investigators have attempted to correlate skin test reactivity to recall antigens with response to therapy. Gutterman et al. (1974) noted that patients experiencing objective tumor regression following DTIC BCG therapy were more likely to be responsive than patients demonstrating tumor progression. However, Mastrangelo et aZ. ( 1976a) found no difference in response to recall antigens between responders and nonresponders to intralesional BCG therapy. In summary, the carefully controlled study of Seigler et al. (1972) clearly demonstrated that reactivity to recall antigens is depressed in melanoma patients with diffuse metastases and extensive prior treatment. Surprisingly little information is available regarding correlation of reactivity to recall antigens and prognosis, clinical status, or response to therapy, For a comprehensive evaluation of the current status and future

+


321

prospects for delayed hypersensitivity reactions in cancer patients, the reader is referred to Burdick et al. ( 1975).

3. Lymphocyte Transformation ( L T ) Lymphocyte transformation to mitogens and various antigens has also been extensively evaluated. In a study by Seigler et al. (1972) 10 of 21 patients evaluated for LT to PHA showed reduced activity when compared with normal controls. None of these ten patients had diffuse tumor involvement. Golub et aZ. (1974) assessed the in uitro LT to various mitogens in 29 patients with malignant melanomas: 2 had no evidence of disease, 15 had regional disease, and 12 had distant metastases. These melanoma patients as a group, when compared with normal controls, had significantly decreased levels of LT to PHA, pokeweed mitogen ( P W M ) and concanavalin A (Con A ) , In brief, Golub et al. ( 1974) and Seigler et al. (1972) noted a decreased LT to PHA in patients with largely regional or advanced disease as compared with normal controls. Ziegler et al. (1969) were unable to demonstrate a difference in L T to PHA between patients with localized disease ( 8 ) and those with more advanced melanoma ( 9 ) . A comparison with normal controls was not attempted, and an insufficient number of patients had visceral metastases to warrant separate analyses. DeGast et al. (1975) conducted a more extensive study assessing LT to PHA and antigens (diphtheria, tetanus toxoid, hemocyanin) in 61 melanoma patients (31 localized disease, 13 regional metastases, 10 distant nonvisceral metastases, 7 visceral metastases). Test results were correlated with clinical stage and the course of disease over a subsequent 6-month period. The seven patients with visceral disease, all of whom experienced progression, had diminished LT to PHA when compared with normal controls and melanoma patients of all other stages. The remaining 54 melanoma patients could not be distinguished from normal controls or each other on the basis of clinical stage or disease course over the brief 6-month observation period. LT to test antigens showed wide variation when evaluated by stage, but LT to all three was uniformly low in patients with visceral metastases. Further, a correlation between LT to test antigens and subsequent clinical course was noted: 9/14 (648) patients who failed to react to all 3 antigens progressed as compared with 1/40 (3%)patients who responded to one or more of the test antigens. These data support the prior observations of Golub et al. (1974) and Seigler et at. (1972) that melanoma patients with more advanced disease demonstrate decreased LT to PHA when compared with controls. These data are also in agreement with the observation of Ziegler et al. (1969),

322


who were also unable to distinguish between patients with local disease and those with more advanced (not terminal) disease on the basis of LT to PHA. DeGast et al. (1975) provided new data in demonstrating that patients with local or regional disease cannot be distinguished from normal controls. These results are in agreement with those of Lui et al. (1975), who noted that with optimal concentrations of PHA, LT was depressed only in preterminal melanoma patients. However, with threshold concentrations of PHA, impaired responses were regularly associated with disseminated disease. That a single determination of LT to PHA is not predictive of clinical course is in agreement with the observations of Gross and Eddie-Quartey (1976) in a mixed group of patients (lung cancer and melanoma). The correlation of LT to a complex of test antigens and clinical course is exciting and warrants further confirmation. A thorough assessment of the ability of LT to PHA and test antigens to indicate prognosis must await serial studies in an adequate group of surgically cured melanoma patients who are at high risk for recurrence. Several investigators have attempted to correlate LT to PHA with the patients’ response to therapy. Cheema and Hersh (1971) studied LT to PHA and streptolysin “0” in 40 patients with a variety of mixed solid tumors (18 melanoma) before and after chemotherapy and reported that a rapid recovery and overshoot of this index following chemotherapy was associated with a more favorable prognosis. Gross and Eddie-Quartey (1976) noted that LT to PHA was impaired, when compared to that of normal controls, in patients (14 lung cancer, 12 melanoma) with both favorable and unfavorable clinical courses and that the two groups could not be distinguished on the basis of a single determination. However, after 3 months of BCG therapy, patients who ultimately had a favorable clinical course demonstrated increased LT to PHA. Patients with an unfavorable clinical course did not, In these two studies, data for melanoma patients were not presented separately, Inasmuch as melanoma patients constituted a reasonable segment of both study populations, the behavior of the groups most likely accurately reflects the behavior of the melanoma patients. However, before firm conclusions can be drawn, data from the melanoma patients alone must be assessed. Lieberman et al. (1975) evaluated LT to PHA in seven patients undergoing intralesional BCG therapy. All four responders demonstrated a marked increase in LT to PHA following treatment, whereas the nonresponders did not. No effort was made to assess pretreatment values regarding their ability to predict response. Roth et al. (1975) evaluated LT to PHA in 40 stage I and I1 melanoma patients and noted no difference between the BCGtreated group and the control group. Serial studies in individual patients were not done. Lui et al. (1975) failed to note an increase in LT to PHA


323

after BCG immunotherapy. These two groups did not attempt to correlate test results with response to therapy or clinical course. In summary, patients with advanced disease have depressed LT on exposure to PHA. Patients with more modest tumor burdens cannot be distinguished from controls. Further, a single determination of LT to PHA is not predictive of prognosis. The observation of DeGast et al. (1975) that LT to a complex of antigens correlates with prognosis is encouraging and warrants further study. The data regarding the effect of therapy on this assay are conflicting. This picture may clarify when therapy data are also analyzed with respect to clinical response and clinical course. Taken as a whole, these studies suggest that melanoma patients, especially those with advanced disease, are less easily sensitized to DNCB, have lower delayed hypersensitivity responses to a variety of antigens, and display impaired transformation responses to PHA and various antigens. The nonspecific immunologic reactivity of individual patients, however, varies from normal to severely depressed. It would not be surprising if nonspecific immunologic reactivity did indeed correlate with specific immune responses to melanoma antigens. But because this correlation has not been demonstrated, it would seem preferable to directly measure the specific responses rather than to draw inferences from the nonspecific ones.

E. THERAPEUTIC ASPECTSOF

THE

IMMUNE RESPONSE

1. Zmmunotherapy Perhaps the most important question regarding the immunology of melanoma is whether immunologic responses can be manipulated to produce regression. The immunotherapy of malignant melanoma and the use of nonspecific immunostimulants in the treatment of human malignancy have been recently reviewed ( Mastrangelo et aZ., 1976b,c). Only nonspecific immunostimulation with BCG ( Bacillus Calmette-Guerin ) has been studied in sufficient detail to warrant specific presentation. BCG has been used as a postoperative adjuvant in patients with stage I1 and I11 melanoma and also intralesionally in patients with surgically incurable disease accessible to injection, Eilber et al. (1976) and Gutterman et al. (1973) have reported improved disease-free intervals and/or survivals for stage I1 and 111 melanoma patients treated postoperatively with BCG (either alone or in combination with tumor cells) when these patients are compared with historical or concurrent nonrandomized con-

324


trols. These data appear to indicate that BCG has a growth-retarding effect on micrometastases of malignant melanoma. The validity of these data can be questioned because of the use of historical or nonrandomized concurrent controls but are ( nonetheless) encouraging. Randomized prospective trials will be required before postoperative adjuvant BCG can be recommended as standard therapy. More convincing evidence can be derived from trials using intralesional BCG in patients with accessible but surgically incurable melanoma. Many investigators have reported on this modality of therapy. For references, the reader is referred to Mastrangelo et aZ. (1976b,c). Regression of injected lesions occurred in 65%of patients and regression of uninjected lesions, in 21% of patients. Complete tumor regression (almost always skin lesions) was achieved in 25% of patients. In our own experience (Mastrangelo et aZ., 1976c), regression of injected as well as uninjected lesions was most likely to be seen in dermal metastases. Patients with visceral metastases have responded poorly. The mean duration for response in our series for complete responders was 12.6f months with one patient in remission in excess of 41 months. Relapse has occurred in 50%of patients in unmaintained complete remission. Regression of injected lesions is not surprising in view of the intense inflammatory response that follows BCG injection. It has not been determined whether regression of uninjected lesions results from the generation of an inflammatory response by the presence of occult BCG organisms or from an augmentation by BCG of the patient’s immunologic response to tumor antigens, The first hypothesis is supported by the fact that uninjected lesions that regress are usually located in the same area (e.g., a single extremity) as the injected lesions which raises the possibility of migration of BCG organisms into these lesions from injected sites. We studied two patients with disseminated satellitosis without demonstrable visceral disease and observed no regression of uninjected distant dermal metastases. Proponents of an immunologic mechanism for tumor regression find support in the following observations: ( a ) BCG organisms have not been demonstrated in regressed uninjected lesions. ( b ) Tumor regression is generally restricted to subjects who are capable of an immune response to BCG. ( c ) Regression of distant visceral metastases (liver, lung) has been reported in two cases. The most clearly documented of these was the regression of a solitary pulmonary metastasis concurrent with regression of injected and uninjected dermal melanoma metastases. BCG immunotherapy, then, is effective when given intralesionally and may also work when used systemically as postoperative adjuvant treat-


325

ment. Given the current state of our understanding of the mechanism of action of BCG, it seems reasonable to assume that these therapeutic effects are indicative of an augmentation of the patient’s immunologic response against his melanoma.

2. Spontaneous Regression It seems appropriate to review the subject of spontaneous regression of melanoma since this phenomenon was reviewed by Everson (1964) and more recently by Cole (1974), who reported 17 cases. Since that time an additional four cases have been reported (Doyle et al., 1973; Maurer et al., 1974; Bulkley et al., 1975; Bodurtha et al., 1976). In the majority of cases (11) dermal metastases regressed. In only one instance was a visceral (liver) metastasis reported to regress. Patients presenting with metastatic melanoma without a known primary constitute from 2.4 to 8.7%of cases in most large series (Baab and McBride, 1975; Smith and Stehlin, 1965; Pack et al., 1952; Das Gupta et al., 1963). In a large number of these cases, historical and physical evidence suggested that the primary cutaneous melanoma spontaneously regressed prior to the development of metastatic lesions. The mechanism of tumor regression is unknown, but there is some evidence that immunologic factors may play a role. Sumner and Foraker (1960) treated a patient with lymph node and subcutaneous metastases with 250 ml of compatible blood from a donor who had experienced spontaneous regression of melanoma. The recipient also experienced complete tumor regression of 5 months’ duration. The patient reported by Bulkley et al. ( 1975) demonstrated delayed cutaneous hypersensitivity to 2/4 KCl extracts of allogeneic human melanoma cells. The lymphocytes of this patient were not cytotoxic against allogeneic tissue culture melanoma cell lines, but these studies were performed 11 years after tumor regression. Maurer et al. (1974) reported a gradual decrease in lymphocyte transformation to autologous tumor which correlated with clinical disease progression in a patient who had previously experienced spontaneous tumor regression. However, significant lymphocyte cytotoxicity ( W r release from melanoma cells) was maintained over this same interval. Bodurtha et a2. ( 1976) demonstrated significant lymphocyte cytotoxicity (which increased with time) in a patient experiencing complete spontaneous regression of dermal melanoma metastases. The rarity of spontaneous regression phenomena in patients with metastatic melanoma virtually precludes a systematic study of mechanism. Perhaps a greater understanding of the immunology of melanoma will lead to an explanation of this strange phenomenon.

326


F. CONCLUSIONS Although the data are not as clear and unequivocal as one would like, the studies reported here provide support for the three originally stated hypotheses regarding the immunotherapy of melanoma. First, the existence of tumor-specific antigens has been established by the demonstration of such antigens in melanoma cell extracts and by the presence in patients’ sera of antibody that is specifically directed toward melanoma cell surface determinants. Second, it seems clear that patients generate humoral ( cytotoxic antibody) and cellular (lymphocyte cytotoxicity, lymphocyte transformation, and lymphokine production) immunologic responses to melanoma antigens, even though the cellular responses may not be directed against tumor-specific antigens. An impact of these immunologic responses on the course of the disease has been suggested by the correlation of serum blocking factors with poor prognosis and cytotoxic antibody with early stage tumors. Third, that immunologic responses can be therapeutically manipulated to produce tumor regression has been shown by the limited but well documented success with intralesional BCG. In which directions should research in the immunology of melanoma now proceed? Clearly, additional effort should be focused on the isolation, purification, and characterization of melanoma antigens. This could lead to a better understanding of what determines whether an immunologic response is tumor-inhibiting or tumor-enhancing and would help to standardize in uitro assays in which antigen is used, We need better in uitro tests for tumor-directed cell-mediated immunity; i.e., tests that are technically simple, reliable, and relevant to in vitro events. This objective could be pursued by improving presently used tests, such as lymphocyte cytotoxicity, by standardizing techniques and gathering more data on specificity, or by developing new assays. In particular, assays for antitumor effects of lymphocytes and macrophages within tumor masses would seem worthwhile in view of the interesting animal data on this subject (Evans, 1972). More attention should be focused on tumordirected humoral immunity. Cytotoxic antibody should be studied in large numbers of patients and serially measured to confirm the seemingly excellent correlation with clinical course, Almost nothing is known about antibodies with other activities, such as those cytophilic for macrophages, in melanoma patients, Finally, and quite obviously, we need to develop more effective immunotherapy for melanoma patients. Carefully designed and well controlled surgical adjuvant studies using BCG and perhaps tumor cell vaccines are warranted since small, clinically undetectable tumor burdens should be more susceptible to immunologic destruction


327

than large, clinically apparent tumor burdens. Other immunostimulants, such as Coynebacterium parvum and levamisole are also worth testing. Most important, new ideas of immunotherapy must be generated as our understanding of the basic immunologic mechanisms increases. Most of the research cited in this review was carried out in the past 10 years. This was a period of great enthusiasm for melanoma1 immunology, and human tumor immunology in general, and has increased our knowledge but has also produced much data that are tenuous and confusing. The next 10 years should be devoted to a perhaps less enthusiastic but more analytical approach to this subject that has so much potential importance for human cancer biology. V. Fine-Structural Studies

The cells of malignant melanomas are quite different from normal melanocytes at the fine-structural level. The abnormalities involve the nucleus, the melanosomes, and, according to Klug and his associates (Klug and Gunther, 1971a; Klug, 1972a), the mitochondria. Various virus-like profiles have been described and have been mentioned elsewhere in this review. The following discussion is a general overview of the observations on melanoma fine structure. While some attempt has been made to correlate fine-structural abnormalities with biologic behavior, such correlations, to be mentioned later, are speculative at best.

A. MELANOSOMAL ABNORMALITIES There seem to be three major patterns of melanosomal structure in tumor cells. Melanosomes which clearly show helical melanofilaments exhibiting cross-linkage and significant melanin deposition may be the commonest form in some melanoma cells. Melanosomes of this form go through the same general maturation stages that characterize normal melanogenesis (Clark and Bretton, 1970, 1971; Toda et al., 1968), but, as a rule, the internal structure of the melanosome is not completely obliterated by melanin synthesis. The general appearance of the cytoplasm of melanoma cells having cross-linked melanofilaments is quite different from a normal melanocyte actively synthesizing melanin. First, the melanosomes are much more numerous in melanoma cells than normal cells. Second, the melanosomes show significant variation in size; some are much larger than normal, others normal, and still others shorter. Klug and his co-workers (Klug and Gunther, 1971a, b, 1972; Klug, 197213) have called this kind of melanoma cell type A and have observed that

328


such cells have mitochondria that are longer than normal and have numerous cristae. Clark et al. (1972) have called such cells type I cells and have shown that they are the dominant ones found in the epidermis of the radial growth phase of lentigo maligna melanoma. Cesarini has made similar observations ( Cesarini et al., 1969; Cesarini, 1971 ) , Cells similar to the type A cells of Klug are also noted in invasive melanomas (Klug and Gunther, 1972), but, in our experience, invasive melanoma cells, especially those in the vertical growth phase of any of the forms of melanoma, tend to show other melanosomal abnormalities, to be presently described. There are melanoma cells with melanofilaments in the melanosomes which may be readily distinguished from the type A cells of Hug. These cells have a cytoplasm crowded with spherical melanosomes; other cytoplasmic organelles are sparse. The melanosomes show only 1 or 2 melanofilaments, and cross-linkage is rarely observed. Melanin synthesis per melanosome (not necessarily per cell) is greatly diminished, and Complete obliteration of melanosomal structure by melanin is not observed. Cells having the foregoing appearance have been called type I1 cells by Clark et al. (1972). As such cells still have helical melanofilaments, they may be regarded as still having the distinctive structural phenotype of a melanocyte, i.e., a melanosome with melanofilaments. The common, large, epithelioid melanocyte so distinctive for many melanomas, has the foregoing fine structure. The cell is especially common in the radial growth phase of malignant melanoma of the superficial spreading type. The second major melanosomal abnormality seen in melanoma cells shows loss of the melanofilament matrix. The cells show numerous granular organelles, but the cytoplasm is not crowded with the structures, which are 250 nm to 500 nm in width. Melanin is deposited within these structures in a coarsely granular fashion. Klug has stated that such cells have spherical mitochondria with few cristae and has called the cells type B cells (Klug and Giinther, 1972). Hirone has also observed and clearly described these granular melanosomes (Hirone et al., 1971). Clark et al. (1972) called the foregoing cells type I11 melanoma cells and noted that the cytoplasm also contained spherical organelles with a lamellar internal structure. The lamellar organelles may show some form of pigment, but extensive melanization, as may occur on the granular melanosome, is not observed. Cesarini has made the important observation that amelanotic melanoma cells may have numerous granular organelles without pigment deposition ( Cesarini and Clark, 1972; Cesarini, 1971). There are other forms of cells associated with diminished melanin synthesis, and these will be mentioned presently.


329

The correlation of granular melanosomes with biologic behavior has not been carried out precisely, Such cells are commonly seen in the vertical growth phase of superficial spreading melanoma, in nodular melanoma, and in metastatic melanoma, but statements of behavior potential beyond the general observation are not at present warranted. The final fine-structural abnormality related to melanosomes is the apparent cessation of melanosomal synthesis. Occasionally, in the vertical growth phase of superficial spreading and nodular melanomas, cells are seen whose cytoplasm is crowded with mitochondria and free ribosomes, but melanosomes cannot be identified. Obviously, there is a sampling problem here, and it cannot be stated that an absolute cessation of melanosomal synthesis has occurred. However, the evidence suggests that, in association with lesional progression in malignant melanoma, the most distinctive phenotype of a melanocyte-melanosomal synthesis-is no longer expressed. Such cells have been termed type IV cells (Clark et al., 1972).

B. AMELANOTICMELANOMA Cesarini (1971) has pointed out that there may be more than one form of cell which fails to synthesize melanin. Amelanotic cells may have melanosomes showing melanofilaments with cross-linking, but no pigment is deposited on the melanosomal matrix. He has also shown that the granular melanosome may fail to synthesize melanin. Virtually by definition, the cell without melanosomes just described does not synthesize melanin. We have also seen cells with only 1 or 2 melanofilaments (type 11, above) without apparent melanin synthesis. Thus, most, if not all, melanoma cell varieties seen in electron micrographs may occur as amelanotic variants. C. NUCLEARABNORMALITIES The nuclei of melanoma cells are quite abnormal (Klug, 1972a, 1974). Nucleoli are invariably altered. They may be compact and much larger than normal. Commonly, the nucleolar material is dispersed and complexly ramified, virtually bridging the nucleoplasm from nuclear envelope to nuclear envelope. Klug ( 1974) described numerous small vacuoles as being commonly present. Large cytoplasmic inclusions may also occur within nuclei. There is no nuclear abnormality as striking as the nuclear envelope. There are deep indentations forming a complexly convoluted structure, having a great increase in nuclear surface.

330


VI. Summary and Conclusions

Malignant melanoma is an uncommon tumor derived from pigmentsynthesizing cells; it is seen in all races of man and in many mammals as well as lower animals. The inductive circumstances operative in this tumor must be widespread in nature and, in part, are heritable cellular susceptibilities to neoplastic transformation. In addition to these generally operative inductive circumstances, which doubtless do not involve exposure to light in any fashion now known, it has been established that there has been a significant increase in incidence of melanoma in individuals born since 1923. The increasing incidence affects the back and arms of both sexes, and the legs of women. The demographic and epidemiologic data strongly indicate that light plays a major role in the increasing incidence of melanoma; however, the role of light has many paradoxes; it does not appear to act as a dose-related carcinogen in most forms of melanomas. It is possible that light interacts with other factors, such as special melanocytic moles (which, in some instances, may be heritable) that act as specific cellular targets for the action of light. Studies over the past 10 years have delineated the developmental biology of the primary lesions of the various melanomas. The developing lesions precisely follow the rules of tumor progression put forth by Leslie Foulds. A variety of anatomic parameters, such as levels of invasion and thickness of primary tumors have been used as mensural data which parallel tumor progression. Such data permit rather accurate prediction as to when a primary tumor acquires competence for metastasis, and the data may then serve as a guide for surgical treatment, chemotherapy, and immunotherapy. The cytoplasm of melanoma cells, especially the fine structure of melanosomes, is characteristically and progressively altered with the neoplastic process. This alteration parallels in a crude way the cellular events of indirect tumor progression. The established lesions of melanoma, both primary and metastatic, are associated with various immunologic phenomena. Apparently tumorspecific melanoma antigens have been demonstrated; patients may generate humoral and cellular immunologic responses to these antigens, and, to some extent, the neoplastic systems may be manipulated by immunologic means. Finally, one cannot view this system at the present time without suggesting that the immediate future may see its inductive circumstances clearly defined. One may be able to understand, at a cellular level what is “competence for metastasis,” and, if this be understood, one can rationally approach the control of the disease. It seems reasonable that the dis-


331

orderly life form that is cancer is probably illustrated rather completely by malignant melanoma, the one complete neoplastic system that forms and evolves on the body surface, where it may be studied with precision.

REFERENCES Ainsworth, A. M., Clark, W. H., Jr., Mastrangelo, M. J., and Conger, K. (1976). Cancer 37, 1928-1936. Alexander, P. ( 1974 ) . Cancer Res. 34,2077-2082. Allen, A. C. ( 1949 ) . Cancer 2, 28-56. Allen, A. C., and Spitz, S. ( 1953). Cancer 6, 1-45. Anderson, D. E. ( 1971). Cancer 28,721-725. Anderson, D. E., Smith, J. L., and McBride, C. M. (1967). J . Am. Med. Assoc. 200, 74 1-746. Anderson, N. G., Holladay, D. W., Caton, J. E., Candler, E. L., Dierlam, P. J., Eveleigh, J. W., Ball, F. L., Hollerman, J. W., Breillatt, J. P., and Coggin, J. H., Jr. ( 1974). Cancer Res. 34, 2066-2076. Andrews, J. C. (1968). Arch. D e m t o l . 98,282-283. Arrington, J. H., Reed, R. J., Ichinose, H., and Krementz, E. T. (1976). Submitted for publication (used with permission). Avis, P., and Lewis, M. G. ( 1973). J . Natl. Cancer Inst. 51, 1063-1065. Baab, G. H., and McBride, C. M. (1975). Arch. Surg. (Chicago) 110, 896-900. Balda, B. R., Hehlmann, R., Cho, J. R., and Spiegelman, S. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 36974700. Baldwin, R. W. ( 1975). J . Natl. Cancer Inst. 55, 745-748. Basombrio, M. A. (1973). J . Oral Pathol. 2,231-253. Bast, R. C., Jr., Bast, B. S., and Rapp, A. J. (1976). Ann. N.Y. Acad. Sci. (in press) Bean, M. A., Bloom, B. R., Herberman, R. B., Old, L. J., Oettgen, H. F., Klein, G., and Terry, W. B. (1975). Cancer Res. 35, 2902-2913. Beardmore, G . L. (1972). In “Melanoma and Skin Cancer” (W. H. McCarthy, ed.), pp. 39-64. Government Printer, Sydney. Berkelhammer, J, (1974). Semin. Oncol. 1, 397-408. Berkelhammer, J., Mastrangelo, M. J., Laucius, J. F., Bodurtha, A. J., and Prehn, R. T. (1975). Int. J . Cancer 16, 571-578. Bingham, D. L. C., Chadsey, L. C., Ramchand, S., and Doris, P. J. (1971). Can. Med. Assoc. J . 105, 607-610. Birkmayer, G . D., Balda, B. R., Miller, F., and Braun-Falco, 0. ( 1972). Naturwissenschaften 59, 369-370. Bluming, A. Z., Vogel, C. L., Ziegler, J. L., and Kiryabwire, J. W. M. ( 1972). I. Natl. Cancer Inst. 48, 17-24. Boddie, A. W., Jr., Urist, M. M., Chee, D. O., Holmer, E. C., and Morton, D. L. (1975). Int. J . Cancer 16, 1035-1041. Bodurtha, A. J., Chee, D. O., Laucius, J. F., Mastrangelo, M. J., and Prehn, R. T. (1975). Cancer Res. 35, 189-193. Bodurtha, A. J., Berkelhammer, J., Kim, Y. H., Laucius, J. F., and Mastrangelo, M. J. ( 1976). Cancer 37,735-742. Brent, L., Brown, J., and Medawar, P. B. (1958). Lancet 2, 561-564. Breslow, A. ( 1970). Ann. Surg. 172, 902-908. Breslow, A. (1975). Ann. Surg. 182, 572-575.

332


Bulkley, G. B., Cohen, M. H., Banks, P. M., Char, D. H., and Ketcham, A. S. (1975). Cancer 36,485-494. Burdick, J. F., Wells, S. A., Jr., and Herberman, R. B. (1975).Surg., Gynecol. Obstet. 141,779-794. Burkitt, D. P. (1969).J . Natl. Cancer Inst. 42, 19-28. Canevari, S., Fossati, G., Della Porta, G., and Bdzarini, G. P. (1975).Int. J . Cancer

16, 722-729.

Carrel, S., and Theilkaes, L. ( 1973).Nature (London) 242, 609-610. Catalona, W.J,, and Chrbtien, P. B. ( 1973).Cancer 31, 353356. Cawley, E.P. (1952).AMA Arch. Dermatol. 65,440450. Cesarini, J. P. (1971).Rev. Etud. Clin. Biol. 16, 316-322. Cesarini, J. P., and Clark, W. H., Jr. (1972).Proc. Int. Cancer Conf., 1972 Abstract, pp. 121-122. Cesarini, J. P., Bonneau, H., and Calas, E. (1969).Bull. Soc. Fr. Dermatol. Syphiligr.

76, 479-482.

Cheema, A. R., and Hersh, E. M. ( 1971).Cancer 28,851-855. Clark, W. H., Jr., and Bretton, R. (1970).Fed. Proc., Fed. Am. Soc. E r p . Biol. 29, 758 ( abstr. ), Clark, W. H., Jr., and Bretton, R. (1971).Monogr. Pathol. 10, 197-214. Clark, W.H., Jr., and Mihm, M. C. ( 1969).Am. J . Pathol. 55,3947. Clark, W.H., Jr., and ten Heggeler, B. (1972).J. CeU B i d . 55,45a,Abstr. No. 90. Clark, W. H., Jr., Pathak, M. A., Szabo, G., Bretton, R., Fitzpatrick, T. B., and El-Mofty, A. M. (1988). Prog. Photobiol., Proc. Int. Congr., 5th, Abstracts, p. 57. Clark, W. H., Jr., From, L., Bemardino, E., and Mihm, M. C., Jr. (1969).Cancer Res. 29, 705-726. Clark, W. H., Jr., ten Heggeler, B., and Bretton, R. (1972).In “Melanoma and Skin Cancer” (W. H. McCarthy, ed.), pp. 121-141. Government Printer, Sydney. Clark, W. H., Jr., Ainsworth, A. M., Bernardino, E. A., Yang, C. H., Mihm, M. C., and Reed, R. J. (1975).Semin. Oncol. 2,83-103. Cochran, A. J., Jehn, U. W., and Guthoskar, B. P. (1972).Lancet 1, 1340-1341. Cochran, A. J., Mackie, R. M., Thomas, C. E., Grant, R. M., Cameron-Mowat, D. E., and Spilg, W. G . S. ( 1973).Br. J . Cancer 28, Suppl. 1, 77-82. Cole, W. H. (1974).Ann. N.Y. Acad. Sd.230,111-141. Cornain, S., DeVries, J. E., Collard, J,, Vennegoor, C., VanWingerden, I., and Rumke, P. ( 1975).Int. J . Cancer 16,981-997. Currie, G. A., and Basham, C. ( 1972), Br. J . Cancer 26,427-438. Currie, G. A,, Lejeune, F., and Hamilton-Fairley, G. (1971).Br. Med. J. 2, 305410. Das Gupta, T.,Bowden, L., and Berg, J. W. (1963).Surg., Cynecol. Obstet. 117,

341-345.

David, J. R., Al-Askari, S., Lawrence, H. S., and Thomas, L. (1964).J . Immunol.

93, 264-273.

DeGast, G . C., The, T. H., Koops, H. S., Huiges, H. A., Oldhoff, J., and Nieweg, H.0.( 1975).Cancer 36,1289-1297. DeVries, J. E., Rumke, P., and Bernheim, J. L. (1972).Int. J . Cancer 9, 567-576. DeVries, J. E., Cornain, S., and Rumke, P. ( 1974).Int. J. Cancer 14,427434. DeVries, J. E.,Meyering, M., Van Dongen, A., and Rumke, P. (1975).Int. J. Cancer

15, 391400.

Doyle, J. C., Bennett, R. C., and Newing, R. K. (1973).Med. J . Aust. 2, 551552. Eilber, F. R., and Morton, D. L. ( 1970).Cancer 25,362367.


333

Eilber, F. R., Morton, D. L., Holmes, E. C., Sparks, F. C., and Ramming, K. P. (1976). N . Engl. J . Med. 294,237-240. Elliott, P. G., Thurlow, B., Needham, P. R. G., and Lewis, M. G. (1973). Eur. J . Cancer 9, 607-610. Elwood, J. M., and Lee, J. A. H. (1974). Can. Med. Assoc. J. 110, 913-915. Elwood, J. M., and Lee, J. A. H. (1975). Semin. Oncol. 2,149-154. Elwood, J. M., Lee, J. A. H., Walter, S. D., Mo, T., and Green, A. E. S. (1974). Int. J . Epidemiol. 2, 325-332. Evans, R. ( 1972). Transplantation 14,468-473. Everson, T. C. (1964). Ann. N.Y. Acad. Sci. 114, 721-735. Falk, R. E., Mann, P., and Langer, B. ( 1973). Arch. Surg. (Chicago) 107, 261-265. Fass, L., Herberman, R. B., Ziegler, J. L., and Kiryabwire, J. W. M. ( 1970). Lancet 1, 116-118. Fitzpatrick, T. B., and Szabo, G. ( 1959). J . Inoest. Dermatol. 32, 197-209. Fossati, G., Colnaghi, M. I., Della Porta, G., Cascinelli, N., and Veronesi, U. (1971). Int. J . Cancer 8, 344-350. Foulds, L. (1954). Cancer Res. 14, 327-339. Foulds, L. ( 1965). Cancer Res. 25, 1339-1347. Foul&, L. (1969). “Neoplastic Development,” Vol. 1. Academic Press, New York. Fraser, D. G., Bull, J. G., and Dunphy, J. E. (1971). Am. J . Surg. 122, 169-173. Frenk, E., and Schellhom, J. P. (1969). Dermutologica 139, 271-277. Fritze, D., Kern, D. H., Drogenmuller, C. R., and Pilch, Y. H. (1976). Cancer Res. 36, 458-466. Gellin, G. A., Kopf, A. W., and Garfinkel, L. (1966). Ado. B i d . Skin 7, 329. George, M. and Vaughn, J. H. (1962). Proc. SOC. Exp. Biol. Med. 111, 514-521. Gercovich, F. G., Gutterman, J. U., Mavligit, G. M., and Hersh, E. M. (1975). Med. Pediatr. Oncol. 1, 277-287. Golub, S. H., O’Connell, T. X., and Morton, D. L. (1974). Cancer Res. 34, 18331837. Greaves, F. M., Brown, G., and Rickinson, A. B. (1975). Clin. Immunol. Immunopathol. 3, 51-24. Greeley, P. W., Middleton, A. G., and Curtin, J. W. (1965). Plast. Reconstr. Surg. 36, 26-37. Gross, N. S., and Eddie-Quartey, A. C. ( 1976). Cancer (in press). Gutterman, J. U., Mavligit, G., McBride, C., Frei, E., Freireich; E. J., and Hersh, E. M. (1973). Lancet 1, 1208-1212. Gutterman, J. U., Mavligit, G., Gottlieb, J. A., Burgess, M. A., McBride, C. E., Einhom, L., Freireich E. J,, and Hersh, E. M. (1974). N. Engl. J. Med. 291, 592597. Gutterman, J. U., Mavligit, G., Reed, R., Richman, S., McBride, C. E., and Hersh, E. M. (1975). Semin. Oncol. 2, 155-174. Hansen, M. G., and McCarten, A. B. (1974). Am. J. Surg. 128,5574561. Hartmann, D., and Lewis, M. G. (1974). Lancet 1, 1318-1320. Hayami, M., Hellstrijm, I., Hellstrom, K. E., and Lannin, D. R. (1974). Int. J . Cancer 13, 43-53. Hellstrom, I., and Hellstrom, K. E. (1973). Fed. Proc., Fed. Am. SOC. Exp. Biol. 32, 156-159. Hellstrijm, I., Hellstriim, K. E., Sjogren, H. O., and Warner, G. A. (1971a). Int. J. Cancer 7, 1-16.

334


Hellstrom, I., Sjogren, H. O., Warner, G., and Hellstrom, K. E. (1971b). Znt. J. Cancer 7,226-237. Hellstriim, I., Hellstrijm, K. E., Sjogren, H. O., and Warner, G. A. ( 1 9 7 1 ~ )Int. . J. Cancer 8, 185-191. Hellstriim, I., HellstFijm, K. E., Sjogren, H. O., and Warner, G. A. (1973a). Int. I. Cancer 11, 116-122. Hellstrijm, I., Warner, G. A., Hellstrijm, K. E., and SjBgren, H. 0. (197313). Int. J . Cancer 11, 280-292. Hellstrom, I., Hellstrijm, K. E., and Warner, G. A. ( 1 9 7 3 ~ )Int. . J . Cancer 12, 348353. Henle, G., and Henle, W. (1970). J . Infect. Dis. 121, 303-310. Henle, C., Henle, W., Clifford, P., Diehl, V., Kafuko, G., Kirya, B. G., Klein, G., Morrow, R. H., Munube, C. M. R., Pike, P., Tukei, P. M., and Ziegler, J. (1969). J . Natl. Cancer Inst. 43, 1147-1157. Henle, G., Henle, W., Klein, C., Gunven, P., Clifford, P., Morrow, R. H., and Ziegler, J. L. (1971). J. Nutl. Cancer Inst. 46, 861-871. Heppner, G. H., Stolbach, L., Byme, M., Cummings, F. J., McDonough, E., and Calabresi, P. (1973). Int. J . Cancer 11, 245-260. Herberman, R. B., and Oldham, R. K. (1975). 1. Nat2. Cancer Inst. 55, 749-753. Hirone, I., Nagai, T., Matsubara, T., and Fukushiro, T. (1971). In “Biology of Normal and Abnormal Melanocytes” (T. Kawamura, T. B. Fitzpatrick, and M. Seiji, eds.), p. 329. Univ. Park Press, Baltimore, Maryland. Ho, J. H. C. ( 1972). Ado. Cancer Res. 15,57-92. Hollinshead, A. C. (1975). Cancer 36, 1282-1288. Hollinshead, A. C., Herberman, R. B., Jaffurs, W. J., Alpert, L. K., Minton, J. P., and Harris, J. E. ( 1974). Cancer 34, 1235-1243. Holmes, E.C., Roth, J. A., and Morton, D. L. (1975). Surgery 78, 160-164. Hutchinson, J. ( 1890). Arch. Surg. (London) 2, 83-86. Hutchinson, J. ( 1892a). Arch. Surg. (London) 3,315-322. Hutchinson, J. ( 1892b). Arch. Surg. (London) 4, 61-65. Hutchinson, J. ( 1894). Arch. Surg. (London) 5, 253-256. Hutchinson, J. (1896). Arch. Surg. (London) 7, 297-317. hie, K., Irie, R. F., and Morton, D. L. (1974). Science 186, 454-456. Jarvis, J. E., Ball, C., Rickinson, A. B., and Epstein, M. A. (1974). Int. 1. Cancer 14, 716-72 1. Jehn, U. W., Nathanson, L., Schwartz, R. S., and Skinner, M. (1970). N . Engl. J. Med. 283, 329-333. Kaplan, E. N. (1974). Plart. Reconctr. Surg. 53,421-428. Ketcham, A. S., and Chrbtien, P. B. ( 1975). Panminerua Med. 17, 174-178. Klein, G. (1975). N. Engl. J . Med. 293, 1353-1357. Klug, H. ( 1974). Acta Morphol. Acad. Sci. Hung. 22,321-326. Klug, H., and Ciinther, W. ( 1971a). Dematologicu 143, 84-94. Klug, H., and Giinther, W. ( 1971b). Arch. Geschwulstforsch. 37, 368-386. Klug, H. (1972a). Zentralbl. Allg. Pathol. Pathol. Anat. 115,265-271. Klug, H . (197213).Zentralbl. AUg. Puthol. Pathol. Anat. 116,316-324. Klug, H., and Giinther, W. (1972). Br. J . Dermutol. 86, 395-407. Knudson, A. G. (1975). Cancer 35, 1022-1026. Knudson, A. G., Strong, L. C., and Anderson, D. E. (1973). Prog. Med. Genet. 9, 113-158. Lancaster, H. 0. (1956). Med. J . Aust. 1, 1082-1087.


335

Lancaster, H. O., and Nelson, J. (1957). Med. J. Aust. 1,452456. Lee, J. A. H. ( 1973). J. Inuest. Dermatol. 59, 445448. Lee, J. A. H. (1975). P r q . Clin. Cancer 6,151-161. Lee, J. A. H., and Carter, A. P. ( 1970). J . Natl. Cancer Inst. 45,91-97. Lee, J. A. H., and Merrill, J. M. (1970). Med. J. Aust. 2, 846-851. Levy, N. H. ( 1973). Natl. Cancer Inst., Monogr. 37, 85-92. Levy, N. L., Seigler, H. F., and Shingleton, W. W. ( 1974). Cancer 34, 1548-1557. Lewis, M. G. (1967). Lancet 2, 921-922. Lewis, M. G., and Phillips, T. M. (1972). Int. J. Cancer 10, 105-111. Lewis, M. G., Ikonopisov, R. L., Naim, R. C., Phillips, T. M., Hamilton-Fairley, G., Bodenham, D. C., and Alexander, P. ( 1969). Br. Med. J . 3,547-552. Lewis, M. G., Phillips, T. M., Cook, K. B., and Blake, J. ( 1971). Nature (London) 232, 5 2 5 4 . Lewis, M. G., Avis, P. J. G., Phillips, T. M., and Sheikh, K. M. A. (1973a). Yale J. Biol. Med. 46, 661-668. Lewis, M. G., McCloy, E., and Blake, J. (197313). Br. J . Surg. 60, 443-446. Lewis, M. G. Hartmann, D., and Jerry, L. M. ( 1976). Ann. N.Y. Acad. Sci. (in press). Lieberman, R., Wybran, J., and Epstein, W. (1975). Cancer 35,756-777. Lindahl, T., Klein, G., Reedman, B. M., Johansson, B., and Singh, S. (1974). Int. J. Cancer 13, 764-772. Lui, V. K., Karpuchas, J., Dent, P. B., McCulloch, P. B., and Blajchman, M. A. ( 1975). Br. J . Cancer 32, 323430. Lynch, H. T., Frichot, B. C., Lynch, P., Lynch, J., and Guirgis, H. A. (1975). Surg., Gynecol. Obstet. 141, 517-552. Lynch, H. T., and Krush, A. J. (1968). Can. Med. Assoc. J . 99, 17-21. McCoy, J. L., Jerome, L. F., Dean, J. H., Perlin, E., Oldham, R. K., Char, D. H., Cohen, M. H., Felix, E. L., and Herberman, R. B. (1975). J. Natl. Cancer Inst. 55, 19-23. McGovern, V. J. ( 1970). Pathology 2, 85-98. McGovern, V. J., Mihm, M. C., Bailly, C., Booth, J. C., Clark, W. H., Cochrm, A. J., Hardy, E. G., Hicks, J. D., Levene, A., Lewis, M. G., Little, J. H., and Milton, V. W. ( 1973). Cancer 32, 1446-1457. Mackie, R. M., Spilg, W. G. S., Thomas, C. E., and Cochran, A. J. ( 1972). Br. J . Dermatol. 87, 523-528. Magnus, K. ( 1973). Cancer 32, 1275-1286. Maguire, H. C., Jr. (1975). Int. J. Dermatol. 14, 3-11. Manolov, G., and Manolova, Y. ( 1972). Nature (London) 237,3334. Mark, G. J., Mihm, M. C., Litepolo, M. G., Reed, R. J., and Clark, W. H., Jr. (1973). Hum. Pathol. 4, 395-418. Mastrangelo, M. J,, Sulit, H. L., Prehn, L., Bomstein, R. S., Yarbro, J. W., and Prehn, R. T. (1976a). Cancer 37, 684-692. Mastrangelo, M. J., Bellet, R. E., Laucius, J. F., and Berkelhammer, J. (1976b). In “Oncologic Medicine” (P. F. Engstrom and A. I. Sutnick, eds.), pp. 71-93. Univ. Park Press, Baltimore, Maryland. Mastrangelo, M. J,, Berd, D., and Bellet, R. E. ( 1 9 7 6 ~ )Ann. . N.Y. Acad. Sci. (in press). Maurer, L. H., McIntyre, 0. R., and Rueckert, F. (1974). Am. J . Surg. 127, 397403. Mavligit, G. M., Ambus, U., Gutterman, J. U., Hersh, E. M., and McBride, C. M. ( 1973a). Nature (London) 243, 188-190.

336


Mavligit, C. M., Cutterman, J. U., McBride, C. M., and Hersh, E. M. (1973b). Natl. Cancer Inst., Monogr. 37, 167-176. Mavligit, C. M., Hersh, E. M., and McBride, C. M. ( 1 9 7 3 ~ )J.. Natl. Cancer Inst. 51, 337-343. Mavligit, G., Gutterman, J. U., McBride, C., and Hersh, E. M. (1974a). Prog. Exp. Tumor Res. 19, 222-252. Mavligit, G . M., Hersh, E. M., and McBride, C. M. (1974b). Cancer 34, 1712-1721. Metzgar, R. S., Bergoc, P. M., Moreno, M. A., and Seigler, H. F. (1973). J. N d l . Cancer Inst. 50, 1065-1068. Miller, T. R., and Pack, G. T. ( 1962). AMA Arch Dermutol. 88,3539. Mishima, Y . (1960). J. Invest. Dermutol. 34,361-375. Mishima, Y . (1966a). Cutb 2,588-591. Mishima, Y. (196613).Adu. Biol. Skin 8, 509. Mishima, Y. ( 1967). Cancer 20, 632-649. Mitchell, R. E. (1963). J . Inoest. D e m t o l . 41, 199-212. Morton, D. L. (1971). J. Reticuloendothl. SOC.10, 137-160. Morton, D. L., Malmgren, R. A., Holmes, E. C., and Ketcham, A. S. (1968). Surgery 84, 223-240. Morton, D. L., Eilber, F. R., Malmgren, R. A., and Wood, W. C. (1970). Surgery 88, 158-184. Moschella, S . L. ( 1981). Arch. Dermatol. 84, 1024-1025. Mukherji, B., Vassos, D., Flowers, A., Binder, S. C., and Nathanson, L. (1975a). Cancer Res. 35, 37214730. Mukheji, B., Vassos, D., Flowers, A., Binder, S. C., and Nathanson, L. (1975b). Int. J . Cancer 18, 971-980. Muller, C., and Sorg, C. (1975).Eur. J . Immunol. 5, 175-178. Musher, D. R., and Linder, A. E. (1974). Dig. Dis. ID, 855-859. Nagel, G. A., St. Arneault, G., Holland, J. F., Kirkpatrick, R., and Kirkpatrick, D. (1970). Cancer Res. 30, 1828-1832. Nagel, G. A., Piessens, W. F., Stilmant, M. M., and Lejeune, F. (1971). Eur. J . Cancer 7, 4 1 4 7 . Nairn, R. C., Nind, A. P. P., Guli, E. P. G., and Davies, D. J. (1972). Med. J . A&. 1, 397403. Nelson, R.A,, Jr. ( 1953). Science 118, 733-737. Nicholls, E. M. (1973). Cancer 32, 191-195. Niederman, J. C., Evans, A. S., Subrahmanyan, L., and McCollum, R. W. ( 1970). N . Engl. J . Med. 282, 361-365. Nonoyama, M., Huang, C. H., Pagano, J. S., Klein, G., and Singh, S. (1973). Pfoc. N d l . Acad. Sci. U.S.A. 70, 3265-3268. Oettgen, H. F., Aoki, T., Old, L. J., Boyse, E. A,, deHarven, E., and Mills, G. M. (1968).J. Natl. Cancer Inst. 41,827-843. Oettgen, H. F., Bean, M. A,, and Klein, G. (1972). Cancer Res. 32, 2845-2853. Oldham, R. K., Djeu, J. Y., Cannon, G. B., Siwarski, D., and Herberman, R. B. ( 1975). J. Natl. Cancer Inst. 55, 1305-1318. ONeill, P. A,, and Romsdahl, M. M. (1974). Immunol. Commun. 3, 427438. O’Neill, P. A., Mackler, B. F., and Romsdahl, M. M. (1976). J. Natl. Cancer Inst. (in press ). O’Toole, C. ( 1973). Nut. Cancer Inst., Monogr. 37, 19-24. Pack, G. T., Gerber, D. M., and Scharnagel, I. M. (1952). Ann. Surg. 138, 905-911. Parks, L. C., Smith, W. J., and Williams, G. M. (1974). Surgery 78, 43-49.


337

Parsons, P. G., Gross, P., and Pope, J. H. (1974). Int. J. Cancer 13, 606-618. Pavie-Fischer, J., Kourilsky, F. M., Picard, F., Banzet, P., and Puissant, A. (1975). Clin. Exp. Immunol. 21, 430-441. Pers, M. (1963). Ugeskr. Laeg. 125, 613. Peter, H. H., Diehl, V., Kalden, J. R., Seeland, P., and Eckert, G. ( 1975). Behn’ng Inst. Mitt. 56, 167-177. Phillips, T. M., and Lewis, M. G. (1970). Reo. Eur. Etud. Clin. Biol. 15, 1016-1020. Pierce, G. E., and DeVald, B. L. ( 1975). Cancer Res. 35,2729-2737. Pinsky, C. M., Hirshaut, Y.,and Oettgen, H. F. (1973). Natl. Cancer Inst., Monogr. 39, 225-228. Powell, A. E., Sloss, A. M., Smith, R. N., Makley, J. T., and Hubay, C. A. (1975). Int. 1. Cancer 16, 905-913. Reed, W. B., Becker, S. W., Sr. and Becker, S. W., Jr. (1965). Arch. Dermutol. 91, 100-119. Robins, R. A., and Baldwin, R. W. (1974). Int. J. Cancer 14, 589-597. Roenigk, H. H., Jr., Deodhar, S. D., Krebs, J. A., and Barna, B. (1975). Arch. Dermutol. 111, 720-725. Rogni, M. U., and Tubon, H. ( 1974). Obstet. Gywcol. 43,658-664. Romsdahl, M. M., and Cox, I. S. ( 1969). Surg. Forum 20, 126-128. Romsdahl, M. M., and Cox, I. S. (1970). Arch. Surg. (Chicago) 100, 491-497. Romsdahl, M. M., and Cox, I. S. ( 1973). Yale J . Biol. Med. 46,693-701. Rosenberg, S.A. (1973). Natl. Cancer Inst., Monogr. 37, 139-140. Roth, J. A., Golub, S. H., Holmes, E. C., and Morton, D. L. (1975). Surgery 78, 66-75. Roth, J. A., Holmes, E. C., Reisfeld, R. A., Slocum, H. K., and Morton, D. L. (1976). Cancer 37, 104-110. Russel, J. L.,and Reyes, R. G. ( 1959). J . Amer. Med. Assoc. 171,2083-2086. St. Arneault, G., Nagel, G., Kirkpatrick, D., Kirkpatrick, R., and Holland, J. F. ( 1970 ) . Cancer 25, 672-677. Schoch, E. P. ( 1963). Arch. Dermatol. 88,445-455. Segall, A,, Weiler, O., Lacour, J., and Lacour, F. ( 1972). Int. J . Cancer 9, 417-425. Seigler, H. F., Shingleton, W. W., Metzgar, R. S., Buckley, C. E., Bergoc, P. M., Miller, D. S., Fetter, B. F., and Phaup, M. B. (1972). Surgery 72, 162-174. Silverstone, H. ( 1964). In “Skin Cancer in Queensland, Australia” (H. F. Blum and F. Urbach, eds.), Report of the Airlie House Conference on Sunlight and Skin Cancer, p. 61. Natl. Inst. Health, Bethesda, Maryland. Smith, F. E., Henly, W. S., Knox, J. M., and Lane, M. (1966). Arch. Intern. Med. 117, 820-823. Smith, J. R.,Jr., and Stehlin, J. S., Jr. (1965). Cancer 18, 1399-1415. Sorberg, M., and Bendixen, G. (1967).Actu Med. Scand. 181, 247-256. Spider, L. E., Levin, A. S., and Wybran, J. (1976). Cell Immunol. 21, 1-19. Stevenson, G. T., and Laurence, D. J. R. ( 1975). Int. 1. Cancer 16,887-896. Stewart, T. H. M. (1969). Cancer 23, 1368-1379. Sumner, W. C., and Foraker, A. G. (1960). Cancer 13,79-81. Szabo, G. ( 1967). Philos. Trans. R . SOC. London, Ser. B 252,447485, Takasugi, M., Mickey, M. R., and Terasaki, P. I. (1973). Cancer Res. 33, 28982902. Takasugi, M., Mickey, M. R., and Terasaki, P. I. ( 1974). J . Natl. Cancer Inst. 53, 1527-1538.

338


Toda, K., Hori, Y., and Fitzpatrick, T. B. (1968). Fed. Proc., Fed. Am. Soc. Exp. Biol. 27, 722. Trozak, D. J., Rowland, W. D., and Hu, F. (1975). Pediatrics 55, 191-204. Turkington, R. W. (1965). J. Am. Med. Assoc. 192, 77-82. Unsgaard, B., and O’Toole, C. (1975). Br. J . Cancer 31, 301316. Urbach, F. (1969). In “The Biologic Effects of Ultraviolet Irradiation” (F. Urbach, ed. ), pp. 635-650. Pergamon, Oxford. Veronesi, U., Cascinelli, N., Fossati, G., Canevari, S., and Balzarini, G. (1973). Etir. J. Cancer 9, 843-846. Visa, D., and Phillips, J. (1971). Reu. Inst. Pasteur Lyon 4,339342. Visa, D., Phillips, J. (1975). Int. J . Cancer 16, 312-317. Wallace, D. C., Exton, L. A., and McLeod, C. R. C. (1971). Cancer 27, 1262-1266. Wallace, D. C., Beardmore, C. L., and Exton, L. A. (1973). Ann. Surg. 177, 15-20. Wanebo, H. J., Woodruff, J., and Fortner, J. G. (1975). Cancer 35, 666-676. Whimster, I. W. (1965). Ann. Ital. DermutoZ. Clin. Sper. 19, 168-191. Whitehead, R. H. ( 1973). Br. J . Cancer 28,525529. Whitehouse, J. M. A. (1973). Br. J . Cancer 28, Suppl. 1 , 170-174. Wood, G. W., and Barth, R. F. (1974). J . Nutl. Cancer Imt. 53,309316. Zech, L., Haglund, U., Nilsoon, K., and Klein, C. ( 1976). Int. J . Cancer 17, 4 7 5 6 . Ziegler, J. L. (1973). In “Cancer Medicine” (J. F. Holland and E. Frei, 111, eds. ), pp. 1321-1330. Lea & Febiger, Philadelphia, Pennsylvania. Ziegler, J. L., Lewis, M. G., Luyombya, J. M. S., and Kiryabwire, J. W. M. (1969 ). Br. 1. Cancer 23,729-734. zurHausen, H., Schulte-Holthauser, H., Klein, G., Henle, W., Henle, G., Clifford, P., and Santesson, L. ( 1970). Nature (London) 228, 1056-1058.

SUBJECT INDEX

Brevibacterium albidum enzyme, effect

A

on viral DNAs, 69 Acute myeloid leukemia, chromosome studies on, 185-191 Amelanotic melanoma, fine-structure studies on, 329 Anemias, chromosome studies on, 194-196 Angioimmunoblastic lymphadenopathy, chromosomal abnormalities of, 205 Animal cells, temperature-sensitive mutations in, 223-266 Antigens, of malignant melanoma, 287416 Antimitotic agents, effect on murine sarcoma virus tumors, 9 Antithymocyte serum, effect on murine sarcoma virus tumors, 9-10 Arthrobacter luteus, enzyme, effect on viral DNAs, 73

Brevibacterium umbra enzyme, effect on viral DNAs, 73 Burkitt lymphoma chromosome abnormalities in, 196, 205-208 inductive factors in, 269-270 C

B

Bacillus amyloliquefaciem enzyme, effect on viral DNAs, 69 Bacillus Calmette-Cuerin (BCC), use in melanoma immunotherapy, 323325 Banding studies, on human neoplasms, 165-222

339

Capsid antigen, of polyoma and SV4Q viruses, 86 Cell-mediated immunity in malignant melanoma, 288-307 to murine sarcoma virus, 3 2 5 0 analytical studies, 34-41 antigenic specificities in, 45-49 Chromosomal abnormalities human tumors and, 185-222 in lymphoproliferative disorders, 196-212 of meningiomas, 168-174 of myeloproliferative disorders, 174196 Chronic myeloid leukemia, chromosome studies on, 174 Chronic myelomonocytic leukemia, chromosome studies on, 185 Colony-inhibition test, in detection of cell-mediated immunity to murine sarcoma virus, 32-34

340

SUBJECT INDEX

Complement system, S region and, 152-155 Connective tissue, changes of, in malignant melanoma, 281 Cortisone, effect on, sarcoma virus tumors, 8-9 Cytotoxic antibody, of malignant melanoma, 314-318

D

Delayed cutaneous hypersensitivity, of malignant melanoma, 288-289 DNA( s ) synthesis of, temperature-sensitive mutants affecting, 237-245 viral, enzyme effects on, 89 Dinitrochlorobenzene, reactivity of malignant melanoma patients to, 316-318 Down’s syndrome, trisomy 21 in, 175

E

Effector cells, in blocking of murine sarcoma virus tumor cells, 41-45 Embryonic antigens, of murine sarcoma virus tumors, properties, 4 Endonucleases, viral DNA cleavage of, 72-73 Enzymes, effect on SV40 and polyoma virus DNAs, 69-78 Epidermis, changes in, in malignant melanoma, 281 Escherichia coli R245, effect on viral DNAs, 73 Eschen’chia coli RI, enzyme, effect on viral DNAs, 72

F

Fc receptor, I region defined antigens and, 146-152 Finkel murine sarcoma virus, 2 Friend virus, tumor-associated transplantation antigens of, 18

G

Gazdar murine sarcoma virus, 2 Genetics of major histocompatibility complex, 117-120 of malignant melanomas, 272-274 of ts mutations, 232-236 GraE virus, tumor-associated transplantation antigens of, 16 Granulocytopenia, chromosome studies on, 194-198 Cross virus, tumor-associated transplantation antigens of, 18

H

H-2 antigen( s ) MSV antibodies and, 57-58 as part of histocompatibility complex, 117-120 role in murine sarcoma virus specificity, 48 Harvey murine sarcoma virus, 2 antigens of, 21 Helper virus, murine sarcoma virus TATAs due to, 17 Hemophilus aegyptius enzyme, effect on viral DNAs, 72, 73 Hemophilus gallinarum enzyme, effect on viral DNAs, 73 Hemophilus hemolyticus enzyme, effect on viral DNAs, 72 Hemophilus influenzae enzyme, effect on viral DNAs, 89, 73 Hemophilus parainfluenxae enzyme, effect on viral DNAs, 72, 73 Hemophilus parahaemolyticus enzyme, effect on viral DNAs, 89, 73 Hemophilus suis enzyme, effect on viral DNAs, 73 Histiocytic lymphomas, chromosomal abnormalities, 201 Histocompatibility complex, 115-183 genetics of, 117-120 traits associated with, 120-127 Hodgkin’s lymphomas, chromosome abnormalities in, 199-200, 201

SUBJECr INDEX I

Immune adherence assay, of melanoma antigens, 313-314 Immune surveillance, murine sarcoma virus tumors and, 5 0 5 2 Immunobiology, of malignant melanoma, 285-327 Immunofluorescence, in studies of melanoma patients, 309-313 K

Kirsten murine sarcoma v i m , 2 Klebsa'ella pneumoniae OK8 enzyme, effect on viral DNAs, 72 1

Lentigo malignant melanoma, 279 Leukemias, appearance of, after rejection of murine sarcoma virus tumors, 5 1 5 2 Leukocyte migration inhibition assay, of melanoma antibodies, 308 Light, role in malignant melanoma induction, 270-271, 280-281 Louis-Bar syndrome, chromosome abnormalities in, 197, 210-212 Lymphocytes transformation response of, in malignant melanoma antigens, 303-306 transformation studies on, in melanoma patients, 321-323 Lymphomas, chromosome abnormalities in, 196-208 Lymphoproliferative disorders, chromosomal abnormalities in, 198-212 M

Macrophage-migration inhibition test, in studies of murine sarcoma virus, 40

Malignant melanoma, 267-338 amelanotic, 329 antigens of, 287-316 cell-mediated immunity in, 287316

341

antibody against, 307416 congenital nevus and, 276277 connective tissue changes in, 281 cytotoxic antibody of, 314-316 delayed cutaneous hypersensitivity studies on, 288-289 developmental biology of, 278-285 dinitrochlorobenzene reactivity in, 318-318 familial aspects of, 272-274 fine-structural studies on, 327-329 immunobiology of, 285427 immunofluorescence studies of, 309313 immunotherapy of, 323-325 increase of, 268 induction of, 268-278 intralesional transformation of, 280 lentigo maligna type, 279, 280-282 lymphocyte-mediated cytotoxicity studies on, 289-303 lymphocyte transformation response studies on, 303-306, 321-323 melanosomal abnormalities in, 327329 metastasis of, 284-285 moles and, 277-278 nuclear abnormalities of, 329 radial growth phase of, 279-280 recall antigen response to, 318-321 spontaneous regression of, 325 sunlight role in, 270-271 superficial spreading type of, 282-285 target cells of, 268-269, 274-278 types of, 279-280 vertical growth phase of, 280 virus role in, 271-272 Melanocytes, changes in, in malignant melanoma, 281 Melanocytic system congenital malformations of, 276-277 epidermal malignant melanoma and, 275-276 Melbourne chromosome, in lymphomas, 200 Meningiomas, chromosomal studies on, 166-174 Metastases, of malignant melanomas, 284-285

342

SUBJECT INDEX

&Microglobulin, 115-163 amino acid content of, 130 isolation and characteristics of, 127132 T-locus gene products and, 144-146 Microcytotoxicity assay, in detection of cell-mediated immunity to murine sarcoma virus, 32-34 Moloney virus, 2 pseudotypes of, 21 tumor-associated transplantation antigens of, 15-16 Murine sarcoma virus (MSV) antigens of, 21-22 antitumor cell-reacting antibodies of, 31 in vivo role, 32 cell-mediated immunity to, 32-50 cells transformed by, serology of, 29-31 neutralizing antibodies of, 24-25 in vivo role of, 25-26 pseudotypes of antigens, 21-22 Murine sarcoma virus-induced tumor, 1-66 antibodies to, anti-H-2 immune responses and, 5 7 5 8 antigens of, 4 antitumor cell reaction of, 26-32 antivirus immune response of, 21-26 cell-surface antigens of, 28 embryonic antigens of, 28-29 enhancement of, in immunosuppressed animals, 8-1 1 evolution of, 3 5 immune surveillance and, 50-52 immunological rejection of, 3-20 immunological studies on, 1-66 inhibition of, by stimulation of immune functions, 11 isolates of, 2 as model of antitumor immune response, 5154 of immune response against viral disease, 56-57 in tumor immunology, 54-56 sarcoma-specific cell-surface antigens of, 28

tumor-associated transplantation antigens of, 14-20 viral cell-surface antigens of, 27-28 viral envelope antigens of, 27 Mycosis fungoides, chromosome studies on, 199, 201 Myelomas, chromosome abnormalities in, 196, 208-210 Myeloproliferative disorders, chromosome studies on, 174-196 Myelosclerosis, chromosome studies on, 191-194 N Nevi, melanocytic, malignant melanomas from, 276-277 Nodular melanoma, 279 Non-Burkitt lymphomas, chromosome abnormalities in, 197-205

P

p30-associated antigens, of murine sarcoma virus, 47 Pancytopenia, chromosome studies on, 194-196 Philadelphia chromosome, in chronic myeloid leukemia, 174-178 Plasma-cell leukemia, chromosome abnormalities of, 208 Plasmocytoma, chromosome abnormalities in, 208 Polycythemia Vera, chromosome studies on, 191-194 Polyoma virus DNA of defective, 99-102 enzyme effects on, 69-76 primary sequence studies, 76-81 protein binding sites, 88-89 replication, 81-83 genetic mapping of, 89-94 genomes of, 67-113 SV40 compared to, 102-105 proteins induced by, 85-88 virus-specific RNAs and, 83-85

343

SUBJECr INDEX

Protein synthesis, temperature-sensitive mutants affecting, 248-251 Prouidencia stuartii 164 enzyme, effect on viral DNAs, 73 R

Rauscher virus, tumor-associated transplantation antigens of, 16, 18 Refractory dysplastic anemias, chromosome studies on, 194-196 RNA synthesis, temperature-sensitive mutations affecting, 245-248 RNA tumor virus, in malignant melanomas, 271 S

Temperature-sensitive mutations affecting division process, 236-237 affecting DNA synthesis, 237-245 affecting protein synthesis, 248-251 affecting RNA synthesis, 245-248 in animal cells, 223-266 characterization of, 236 frequency and general behavior of, 230-232 genetic analysis of, 232-238 growth type, 225-253 list, 238-239 induction and selection methods, 226230 nature of, 261-262 neoplastic transformation and, 255260 in specialized functions, 253-260 Thrombocytopenia, chromosome studies on, 194-196 Thrombocytosis, chromosome studies on, 191-194 Thymus-leukemia antigens, biochemical properties of, 142-144 T-locus gene products, &-microglobulin and, 144-146 Transformation, neoplastic, temperaturesensitive mutations and, 255-261 Transplantation antigens of polyoma and SV40 viruses, 85-86 properties of, 132-142 Tumor-associated transplantation antigens (TATA), of murine sarcoma virus, 14-20

S antigen, of polyoma and SV40 viruses, 86 S region, complement system and, 152155 Sarcoma cell-surface antigens, properties of, 4 SQary’s syndrome, chromosome studies on, 199, 200 Sideroblastic anemias, chromosome studies on, 194-196 Skin, malignant melanoma in, 267438 Streptomyces albus G enzyme, effect on viral DNAs, 72 SV40 DNA of defective, 99-102 enzyme effects on, 69-76 primary sequence studies, 76-81 protein binding sites, 88-89 replication, 81-83 genetic mapping of, 89-94 genomes of, 67-113 essential and nonessential regions, 94 polyoma virus compared to, 102105 proteins induced by 85-88 virus-specific RNAs and, 83-85

Viral cell-surface antigens, of murine sarcoma virus tumors, properties, 4 Viral envelope antigens, of murine sarcoma virus tumors, properties, 4 Viruses, role in malignant melanoma, 271-272 VP proteins, of polyoma and SV40 viruses, 85

T

X

T antigens, of polyoma and SV40 viruses, 85

V

X-rays, effect on murine sarcoma virus growth, 8-9

CONTENTS OF PREVIOUS VOLUMES Carcinogenesis and Tumor Pathogenesis 1. Berenblum Electronic Configuration and Carcino- Ionizing Radiations and Cancer genesis Austin M. Brues C. A. Coulson Survival and Preservation of Tumors in Epidermal Carcinogenesis the Frozen State E. V. Cowdy James Craigie The Milk Agent in the Origin of Mam- Energy and Nitrogen Metabolism in mary Tumors in Mice Cancer L. Dmochowski Leonard D. Fenninger and G . Burroughs Mider Hormonal Aspects of Experimental Tumorigenesis Some Aspects of the Clinical Use of T. U.Gardner Nitrogen Mustards Calvin T. Klopp and Jeanne C. BateProperties of the Agent of Rous No. 1 man Sarcoma R. J. C. Harris Genetic Studies in Experimental Cancer L. W. Law Applications of Radioisotopes to Studies of Carcinogenesis and Tumor Me- The Role of Viruses in the Production of tabolism Cancer Charles Heidelberger C. Oberling and M. Guerin The Carcinogenic Aminoazo Dyes Experimental Cancer Chemotherapy James A. Miller and Elizabeth C. C. Chester Stock Miller AUTHOR INDEX-SUB JECT INDEX The Chemistry of Cytotoxic Alkylating Agents Volume 3 M. C. 1. Ross Etiology of Lung Cancer Nutrition in Relation to Cancer Richard Doll Albert Tannenbaum and Herbert Silverstone The Experimental Development and Metabolism of Thyroid Gland Plasma Proteins in Cancer Tumors Richard J . Winder Harold P. Morris AUTHOR INDEX-SUB JECT INDEX Electronic Structure and Carcinogenic Activity and Aromatic Molecules: Volume 2 New Developments A. Pullman and B. Pullman The Reactions of Carcinogens with MacSome Aspects of Carcinogenesis romolecules P. Rondonl Peter Alexander Chemical Constitution and Carcinogenic Pulmonary Tumors in Experimental Animals Activity Michael B. Shimkin G. M. Badger

Volume 1

344

CONTENTS OF PREVIOUS VOLUMES

Oxidative Metabolism of Neoplastic Tissues Sidney Weinhouse AUTHOR INDEX-SUB JECT INDEX

Volume 4

Advances in Chemotherapy of Cancer in Man Sidney Farber, Rudolf Toch, Edward Manning Sears, and Donald Pinkel The Use of Myleran and Similar Agents in Chronic Leukemias D. A. G. Gdton The Employment of Methods of Inhibition Analysis in the Normal and Tumor-Bearing Mammalian Organism Abraham Goldin Some Recent Work on Tumor Immunity P. A. Gorer Inductive Tissue Interaction in Development Clifford Grobstein Lipids in Cancer Frances L. Haven and W. R. Bloor The Relation between Carcinogenic Activity and the Physical and Chemical Properties of Angular Benzacridines A. Lacassagne, N. P. BuuHoZ, R. Daudel, and F. Zajdela The Hormonal Genesis of Mammary Cancer 0. Miihlbock AUTHOR INDEX-SUB JECT INDEX

Volume 5

Tumor-Host Relations R. W. Begg Primary Carcinoma of the Liver Charles Bemwn Protein Synthesis with Special Reference to Growth Processes both Normal and Abnormal P. N. Campbell

345

The Newer Concept of Cancer Toxin Waro Nakahara and Fumiko Fukuoka Chemically Induced Tumors of Fowls P. R. Peacock Anemia in Cancer Vincent E. Price and Robert E. GreenFeld Specific Tumor Antigens L. A. Zilber Chemistry, Carcinogenicity, and Metabolism of 2-Fluorenamine and Related Compounds Elizabeth K. Weisburger and John H. Weisburger AUTHOR INDEXSUB JECr INDEX

Volume 6

Blood Enzymes in Cancer and Other Diseases Oscar Bodansky The Plant Tumor Problem Armin C. Braun and Henry N. Wood Cancer Chemotherapy by Perfusion Oscar Creech, Jr. and Edward T. Krementz Viral Etiology of Mouse Leukemia Ludwick Gross Radiation Chimeras P. C. Koller, A. J. S. Davies, and Sheila M. A. Doak Etiology and Pathogenesis of Mouse Leukemia J. F. .4. P. Miller Antagonists of Purine and Pyrimidine Metabolites and of Folic Acid G. M. Timmis Behavior of Liver Enzymes in Hepatocarcinogenesis George Weber AUTHOR INDEXSUB JECT INDEX

Volume 7

Avian Virus Growths and Their Etiologic Agents J. W. Beard

346


Mechanisms of Resistance to Anticancer Agents R. W . Brockman Cross Resistance and Collateral Sensitivity Studies in Cancer Chemotherapy Dorris J. Hutchison Cytogenic Studies in Chronic Myeloid Leukemia W . M . Court Brown and Ishbel M . Tough Ethionine Carcinogenesis Emmanuel Farber Atmospheric Factors in Pathogenesis of Lung Cancer Paul Kotin and Hans L. Folk Progress with Some Tumor Viruses of Chickens and Mammals: The Problem of Passenger Viruses G. Negroni AUTHOR INDEX-SUB JECT INDEX

Volume 8

The Structure of Tumor Viruses and Its Bearing on Their Relation to Viruses in General A. F. Howatson Nuclear Proteins o f Neoplastic Cells Harris Bzisch and William J. Steele Nucleolar Chromosomes : Structures, Interactions, and Perspectives M . J. Kopac and Gladys M . Mateyko Carcinogenesis Related to Foods Contaminated by Processing and Fungal Metabolites H . F. Kraybill and M . B. Shimkin Experimental Tobacco Carcinogenesis Ernest L. Wynder and Dietrich H o f man AUTHOR INDEX-SUB JECT INDEX

Volume 9

Urinary Enzymes and Their Diagnostic Value in Human Cancer Richard Stambaugh and Sidney Weinhouse

The Relation of the Immune Reaction to Cancer Louis V . Caso Amino Acid Transport in Tumor Cells R. M . Iohnstone and P. G. Scholefield Studies on the Development, Biochemistry, and Biology of Experimental Hepatomas Harold P. Morris Biochemistry of Normal and Leukemic Leucocytes, Thrombocytes, and Bone Marrow Cells I . F . Seitz AUTHOR INDEX-SUB JECT INDEX

Volume 10

Carcinogens, Enzyme Induction, and Gene Action H . V . Celboin I n Vitro Studies on Protein Synthesis by Malignant Cells A. Clark Grifin The Enzymatic Pattern of Neoplastic Tissue W . Eugene Knox Carcinogenic Nitroso Compounds P. N. Magee and I. M . Barnes The Sulfhydryl Group and Carcinogenesis s. Harrington The Treatment of Plasma Cell Myeloma Daniel E. Bergsagel, K . M . Crifith, A. Haut, and W . J. Stuckley, Jr. AUTHOR INDEX-SUB JECT INDEX

Volume 1 1

The Carcinogenic Action and Metabolism of Urethan and N-Hydroxyurethan Sidney S. Mirvish Runting Syndromes, Autoimmunity, and Neoplasia D. Keast Viral-Induced Enzymes and the Problem of Viral Oncogenesis Saul Kit

347


The Growth-Regulating Activity of Polyanions: A Theoretical Discussion of Their Place in the Intercellular Environment and Their Role in Cell Physiology William Regelson Molecular Geometry and Carcinogenic Activity of Aromatic Compounds. New Perspectives Joseph C . Amos and Mary F. Argus AUTHOR INDEX-SUB JECT INDEX

CUMULATIVE INDEX

Volume 1 2

Antigens Induced by the Mouse Leukemia Viruses G . Pasternak Immunological Aspects of Carcinogenesis by Deoxyribonucleic Acid Tumor Viruses G . I . Deichman Replication of Oncogenic Viruses in Virus-Induced Tumor Cells-Their Persistence and Interaction with Other Viruses H. Hanafusa Cellular Immunity against Tumor Antigens Karl Erik Hellstrom and lngegerd Hellstrom Perspectives in the Epidemiology of Leukemia lrving L. Kessler and Abraham M . Lilienfeld AUTHOR INDEX-SUB JECT INDEX

Volume 13

The Role of Immunoblasts in Host Resistance and Immunotherapy of Primary Sarcomata P . Alexander and 1. G . Hall Evidence for the Viral Etiology of Leukemia in the Domestic Mammals Oswald Jarrett

The Function of the Delayed Sensitivity Reaction as Revealed in the Graft Reaction Culture Haim Ginsburg Epigenetic Processes and Their Relevance to the Study of Neoplasia Gajanan V. Sherbet The Characteristics of Animal Cells Transformed in Vitro lan Macpherson Role of Cell Association in Virus Infection and Virus Rescue I. Svoboda and I. Hlofhnek Cancer of the Urinary Tract D. B. Clayson and E . H . Cooper Aspects of the EB Virus M. A. Epstein AUTHOR INDEX-SUB JEC T INDEX

Volume 14

Active Immunotherapy Georges Mathk The Investigation of Oncogenic Viral Genomes in Transformed Cells by Nucleic Acid Hybridization Ernest Winocour Viral Genome and Oncogenic Transformation: Nuclear and Plasma Membrane Events Georges Meyer Passive Immunotherapy of Leukemia and Other Cancer Roland Mottn Humoral Regulators in the Development and Progression of Leukemia Donald Metcalf Complement and Tumor Immunology Kusuya Nishioka Alpha-Fetoprotein in Ontogenesis and Its Association with Malignant Tumors G . 1. Abeler Low Dose Radiation Cancers in Man Alice Stewart AUTHOR INDEX-SUBJECT

INDEX

348


Volume 15

1,3-Bis(2-chloroethyl )-1-nitrosourea (BCNU) and Other Nitrosoureas in Cancer Treatment: A Review Stephen K. Carter, Frank M. Schubel, Jr., Lawrence E. Broder, and Thomas P. Johnston

Oncogenicity and Cell Transformation by Papovavirus SV40: The Role of the Viral Genome J. S. Butel, S. S. Teoethia, and 1. L. Melnick AUTHOR INDEX-SUB JECT INDEX Nasophatyngeal Carcinoma (NPC) 1. H. C. Ho Transcriptional Regulation in Eukaryotic Volume 17 Cells A. J. MacGillioray, J. Paul, and G. Polysaccharides in Cancer: Glycoproteins and Glycolipids Threlfall Viiai N. Nigam and Antonio Cantero Atypical Transfer RNA’s and Their OriSome Aspects of the Epidemiology and gin in Neoplastic Cells Etiology of Esophageal Cancer with Ernest Borek and Sylvia 3. Kerr Particular Emphasis on the Transkei, Use of Genetic Markers to Study Cellular South Africa Origin and Development of Tumors Gerald P. Warwick and john S. Harin Human Females lngton Philip J. Fialkow Genetic Control of Murine Viral LeuElectron Spin Resonance Studies of Carkemogenesis cinogenesis Frank Lilly and Theodore Pincus Harold M. Swartz Some Biochemical Aspects of the Rela- Marek‘s Disease: A Neoplastic Disease of Chickens Caused by a Herpestionship between the Tumor and the virus Host K. Nazerian V. S. Shapot Mutation and Human Cancer Nuclear Proteins and the Cell Cycle Alfred G. Knudson, Jr. Gary Stein and Renato Bmerga Mammary Neoplasia in Mice AUTHOR INDEX-SUB JECT INDEX S. Nandi and Charles M. McGrath AUTHOR INDEX-SUB JECT INDEX

Volume 16

Polysaccharides in Cancer Vijai N. Nigam and Antonio Cantero Antitumor Effects of Interferon lon Gresser Transformation by Polyoma Virus and Simian Virus 40 Joe Sambrook Molecular Repair, Wound Healing, and Carcinogenesis : Tumor Production a Possible Overhealing? Sir Alexander Haddow The Expression of Normal Histocompatibility Antigens in Tumor Cells Alena ‘Lengeiool

Volume 18

Immunological Aspects of Chemical Carcinogenesis R. W. Baldwin Isozymes and Cancer Fanny Schapira Physiological and Biochemical Reviews of Sex Differences and Carcinogenesis with Particular Reference to the Liver Yee Chu Toh Immunodeficiency and Cancer John H. Kersey, Beatrice D. Spector, and Robert A. Good


Recent Observations Related to the Chemotherapy and Immunology of Gestational Choriocarcinoma K. D. Bagshawe Glycolipids of Tumor Cell Membrane Sen-itiroh Hakomori Chemical Oncogenesis in Culture Charles Heidelberger AUTHOR INDEX-SUB JECT INDEX

Volume 19

Comparative Aspects of Mammary Tumors

J. M. Hamilton The Cellular and Molecular Biology of RNA Tumor Viruses, Especially Avian Leukosis-Sarcoma Viruses, and Their Relatives Howard M . Temin Cancer, Differentiation, and Embryonic Antigens: Some Central Problems J. H. Coggin, Jr. and N. C. Anderson Simian Herpesviruses and Neoplasia Fredrich W . Deinhardt, Lnwrence A. Falk, and Lauren C. Wolfe Cell-Mediated Immunity to Tumor Cells Ronald B. Herbennan Herpesviruses and Cancer Fred Rapp Cyclic AMP and the Transformation of Fibroblasts Ira Pustan and George S . Johnson Tumor Angiogenesis Judah Folkman SUBJECT INDEX

Volume 20

Tumor Cell Surfaces: General Alterations Detected by Agglutinins Annette M. C. Rapin and Mar M . Burger Principles of Immunological Tolerance and Immunocyte Receptor Blockade C.I . V. Nossal

349

The Role of Macrophages in Defense against Neoplastic Disease Michael H. Levy and E. Frederick Wheelock Epoxides in Polycyclic Aromatic Hydrocarbon Metabolism and Carcinogenesis P. Sims and P. L. Grover Virion and Tumor Cell Antigens of C-Type RNA Tumor Viruses Heinz Bauer Addendum to “Molecular Repair, Wound Healing, and Carcinogenesis: Tumor Production a Possible Overhealing?” Sir Alexander Haddow SUBJECT INDEX

Volume 21

Lung Tumors in Mice: Application to Carcinogenesis Bioassay Michael B. Shimkin and Gary D. Stoner Cell Death in Normal and Malignant Tissues E. H . Cooper, A. J . Bedford, and T . E. Kenny The Histocompatibility-Linked Immune Response Genes Baruj Benacerraf and David H. Katt Horizontally and Vertically Transmitted Oncornaviruses of Cats M . E ssex Epithelial Cells: Growth in Culture ‘of Normal and Neoplastic Forms Keen A. Raferty, 17. Selection of Biochemically Variant, in Some Cases Mutant, Mammalian Cells in Culture G. B. Clements The Role of DNA Repair and Somatic Mutation in Carcinogenesis James E. Trosko and Ernest H . Y. Chu SUBJECT INDEX

350


Volume 2 2

Renal Carcinogenesis J. M . Hamilton Toxicity of Antineoplastic Agents in Man: Chromosomal Aberrations, Antifertility Effects, Congenital Malformations, and Carcinogenic Potential Susan M . Sieber and Richard H . Adamson Interrelationships among RNA Tumor Viruses and Host Cells Raymond V . Gilden Proteolytic Enzymes, Cell Surface Changes, and Viral Transformation Richard Roblin, Zih-Nan Chou, and Paul H . Black Immunodepression and Malignancy Osias S t u t m n SUBJECT INDEX

Volume 23

The Genetic Aspects of Human Cancer W . E . Heston

The Structure and Function of Intercellular Junctions in Cancer Ronald S. Weinstein, Frederick B. Merk, and Joseph Alroy Genetics of Adenoviruses Harold S. Ginsberg and C. S. H . Young Molecular Biology of the Carcinogen, 4-Nitroquinoline 1-Oxide Minako Nagao and Takashi Sugimura Epstein-Barr Virus and Nonhuman Primates: Natural and Experimental Infection A. Frank, W . A. Andiman, and G. Miller Tumor Progression and Homeostasis Richmond T. Prehn Genetic Transformation of Animal Cells with Viral DNA or RNA Tumor Viruses Miroslav Hill and Jana Hillova SUBJECT

INDEX

A 8 7 C 8 D 9

E O F

1

G 2

H 3 1 4 1 5

Advances in Cancer Research, Volume 24

Advances in Cancer Research

Advances in Botanical Research, Volume 24

Advances in Food Research, Volume 24

Advances in Food Research, Volume 24

Advances in Cancer Research Volume 51

Advances in Cancer Research Volume 43

Advances in Cancer Research, Volume 91

Advances in Cancer Research Volume 42

Advances in Cancer Research Volume 41

Advances in Cancer Research Volume 78

Advances in Cancer Research, Volume 106

Advances in Cancer Research, Volume 84

Advances in Cancer Research, Volume 20

Advances in Cancer Research Volume 76

Advances in Cancer Research Volume 65

Advances in Cancer Research Volume 40

Advances in Cancer Research Volume 63

Advances in Cancer Research Volume 35

Advances in Cancer Research Volume 75

Advances in Cancer Research Volume 49

Advances in Cancer Research Volume 32

Advances in Cancer Research Volume 82

Advances in Cancer Research, Volume 57

Advances in Cancer Research, Volume 5

Advances in Cancer Research Volume 70

Advances in Cancer Research Volume 60

Advances in Cancer Research Volume 47

Advances in Cancer Research, Volume 30

Advances in Cancer Research Volume 52

Advances in Cancer Research Volume 54

Advances in Cancer Research, Volume 24