Advances in Cancer Research Volume 45

ADVANCESINCANCERRESEARCH VOLUME 45

Contributors to This Volume

Mauro Bendinelli

Donatella Matteucci

Brent H. Cochran

Evelyne Mougneau

Gisele Connan

Paul Neiman

FranGois Cuzin

Robert J. North

Herman Friedman

Minoo Rassoulzadegan

Nicolas Glaichenhaus

R. Schoental

Bernard S. Strauss

ADVANCES IN CANCER RESEARCH Edited by

GEORGE KLEIN Department of Tumor Biology Karolinska lnstitutet Stockholm, Sweden

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

Volume 45-1985

ACADEMIC PRESS, INC. Harcourt Brace Jovanovich, Publishers Orlando San Diego New York Austin London Montreal Sydney Tokyo Toronto

COPYRIGHT @ 1985 BY ACAD@MICPRESS. INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR T R A N S m D IN ANY FORM OR BY ANY m S . ELFXTRONIC OR MECHANICAL, INCLUDING PH-PY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM,WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, JNC.

orl.mb,Florida 32887

United Kingdom Edition ublished by ACADEMIC PRESS d C . (LONDON) LTD. 24-28 ovpl Road. London W 1 7DX

LIBRARY OF CONGRESS CATALOG CARD NUMBER:

ISBN 0-12-006645-9 PRINTED MTHB UNlTeD STATes O P A M W C A

0 5 8 6 0 7 00

9 0 1 6 5 4 3 2 I

52-13360

CONTENTS

CONTRIBUTORS TO VOLUME45 .............................................

ix

Down-Regulation of the Antitumor Immune Response ROBERT J . NORTH I . Introduction ...................................................... I1. Tumor Immunogenicity ............................................ 111. Concomitant Immunity as the Unsuccessful Response to Tumor Growth ............... IV . Evidence That Tumor Growth Induces Suppressor T Cells .......... V . Evidence That the Generation of Effector T Cells Precedes the Generation of Suppressor T Cells .................................. VI . Evaluation of the Evidence That Ly 1*.2. Suppressor T Cells Down-Regulate the Generation of Ly 1.. 2+ T Cell That Mediated Concomitant Immunity .............................. VII . Tumor-Induced Immunosuppression as the Explanation of Escape from Immunity ............................................ VIII . Suppression of Antitumor Immunity as an Example of Transplantation Tolerance ......................................... IX . Immunotherapeutic Significance of the Generation and Subsequent Decay of Concomitant Immunity ....................... x. Conclusion ........................................................ References ........................................................

1 2

5 7 12

15 22 25

26 38 40

Cellular Aspects of DNA Repair BERNARD S . STRAWS I. I1. 111. IV. V. VI . VII . VIII .

Introduction ...................................................... Recognition of Damage by Cells ................................... I n Vivo Aspects of Excision Repair ................................. Effects of the Local Environment on Reaction and Repair ............ Poly (ADP-Ribose) ................................................. The Repair of 0-Alkylated Sites .................................... Adaptive Response ................................................ Bypass of Lesions and Its Consequence ............................. V

45 46 50 66 70 74 78 80

vi

CONTENTS IX. Error-Prone Repair and Mutation ................................... X . Biological Role of Repair .......................................... XI . Appendix ......................................................... References ........................................................

84 90 93 95

The Blym Oncogenes PAULNEIMAN I . Definitions and Significance ....................................... I1. Oncogenes in Bursa1 Lymphomas ..................................

111. Blym-1 Oncogenes in Human Burkitt’s Lymphomas

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

IV. Issues for Continued Investigation ................................. References ........................................................

107 108 114 120 121

Retrovirus-Induced Acquired Immunodeficiencies MAUROBENDINELLI. DONATELLA M A ~ E U C CAND I . HERMAN FRIEDMAN I . Introduction ...................................................... I1. Retroviruses as Agents of Immunodeficiency ........................ 111. Immunodepressive Changes in Retrovirus-Infected Animals ......... IV. Functional Alterations of Immunocompetent Cells in Retrovirus-Infected Animals ............................... V. Mechanisms Leading to Immunocompetent Cell Alteration in Retrovirus Infections .................................. VI . Role of Retrovirus-Induced Immunodeficiency in Pathogenesis ...... VII . The Retroviral Etiology of AIDS ................................... VIII . Summary and Perspectives ......................................... References ........................................................

125 127 134 137 149 159 162 168 169

The Molecular Action of Platelet-Derived Growth Factor BRENTH . COCHRAN 1. Introduction ...................................................... I1. The Biology of Platelet-Derived Growth Factor ..................... 111. Biochemistry of PDGF ............................................ IV. The SisIPDGF Gene .............................................. V. The Biology of the Sis Oncogene ................................... VI . The PDGF Receptor .............................................. VII . Metabolic Effects of PDGF ........................................

183 184 188 192 194 195 197

VIII . IX. X. XI . XI1.

CONTENTS

vii

PDGF Modulation of the EGF Receptor ............................ EffectofPDGF on Ion Fluxes ..................................... PDGF-Stimulated Protein Phosphorylations ......................... Regulation of Gene Expression by PDGF ........................... Conclusion ........................................................ References ........................................................

200 201 202 204 209 211

Trichothecenes. Zearalenone. and Other Carcinogenic Metabolites of Fusarium and Related Microfungi R . SCHOENTAL I. I1. 111. IV . V. VI . VII . VIII . IX . X. XI . XI1 . XI11. XIV .

xv. XVI . XVII . XVIII . XIX .

Introduction ...................................................... Secondary Metabolites of Fusaria .................................. Epidemiological Considerations .................................... Occurrence and Pathological Effects of T-2 Toxin ................... Diacetoxyscirpenol (Anguidine. NSC-141537) ....................... The Antileukemic Baccharinoids and Other Macrocyclic Trichothecenes ........................................ Metabolism of T-2 Toxin ........................................... Effects of T-2 Toxin and Related Trichothecenes on the Immune System ................................................... Detection and Estimation of Trichothecenes ........................ Chemistry and Biological Activity of Zearalenone ................... Teratogenic Effects of Zearalenone and Bone Lesions ............... Metabolism of Zearalenone ........................................ Estrogenic Agents and Zeranol ..................................... Estrogenic Agents and the Development of Sex Organ Abnormalities and Tumors ............................... Carcinogenic Effects of Zearalenone ................................ Occurrence of Zearalenone and Distribution of Mycotoxins .......... Methods of Detection and Estimation of Zearalenone and Its Estrogenic Derivatives ......................................... Attempts at Detoxication of Fusarial Mycotoxins ..................... Conclusions ....................................................... References ........................................................

218 220 222 231 234 236 240 243 244 245 248 249 251 253 257 268 270 271 272 274

Cooperation between Multiple Oncogenes in Rodent Embryo Fibroblasts: An Experimental Model of Tumor Progression? NICOLASGLAICHENHAUS. EVELYNE MOUGNEAU.GISPLECONNAN. MINOORASSOULZADEGAN. A N D FRANCOIS CUZIN I . Introduction

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

I1. The Multiple Oncogenes of DNA Tumor Viruses

111. Cooperation between Cellular Oncogenes

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

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

291 292 294

...

Vlll

CONTENTS

IV . “Immortalization” by Genes of Group I: A Complex Phenotype ...... V. Changes in the Expression of Cellular Genes Induced by Group I Oncogenes .................................... VI . Early Stages of Transformation ..................................... References ........................................................ INDEX..................................................................... CONTENTS OF RECENTVOLUMES .............................................

296 300 301 303 307 321

CONTRIBUTORS TO VOLUME 45 Numbers in parentheses indicate the pages on which the authors’ contributions begin.

MAUROBENDINELLI(125), lnstitute of Epidemiology, Hygiene and Virology, University of Pisa, Pisa, ltaly BRENTH. COCHRAN (183), Center for Cancer Research and Department of Biology, Massachusetts Znstitute of Technology, Cambridge, Massachusetts 02139 GISELECONNAN (291), Unitt 273 de l’lnstitut National de la Santt et de la Recherche Mddicale, Centre de Biochimie, Universitd de Nice, 06034 Nice, France FRANCOIS CUZIN(291), Unite‘ 273 de I’Znstitut National de la Santd et de la Recherche Mddicale, Centre de Biochimie, Universitt de Nice, 06034 Nice, France HERMANFRIEDMAN (125), Department of Microbiology and Zmmunology, College of Medicine, University of South Florida, Tampa, Florida 33612 NICOLAS GLAICHENHAUS (291), Unitt 273 de 1’Znstitut National de la Santt et de la Recherche Mtdicale, Centre de Biochimie, Universitk de Nice, 06034 Nice, France DONATELLA MATTEUCCI(125), lnstitute of Epidemiology, Hygiene and Virology, University of Pisa, Pisa, ltaly EVELYNE MOUGNEAU(291), Unitt 273 de l’lnstitut National de la Santd et de la Recherche Mddicale, Centre de Biochimie, Universitt de Nice, 06034 Nice, France PAULNEIMAN (107), Fred Hutchinson Cancer Research Center, Seattle, Washington 981 04 ROBERTJ . NORTH (l), Trudeau Znstitute, lnc., Saranac Lake, New York 12983 MINOO RASSOULZADEGAN(291), Unitt 273 de l’lnstitut National de la Santt et de la Recherche Mddicale, Centre de Biochimie, Universitd de Nice, 06034 Nice, France R. SCHOENTAL (217), Department of Pathology, The Royal Veterinary College, University of London, London, NWl OTU, England BERNARD S . STRAUSS(45), Department of Molecular Genetics and Cell Biology, The University of Chicago, Chicago, Illinois 60637 ix

This Page Intentionally Left Blank

DOWN-REGULATIONOF THE ANTITUMOR IMMUNE RESPONSE Robert J. North Trudeau Institute. Inc.. Saranac Lake, New York

I. Introduction

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

11. Tumor Immunogenicity. . . . . . , . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Concomitant Immunity as the Unsuccessful Response to Tumor Growth..

IV. Evidence That Tumor Growth Induces Suppressor T Cells.. . . . . . . . . . . . A. Ly 1+,2- Suppressor T Cells Are the Major Obstacle to Adoptive Immunotherapy of Established Tumors. . . . . . . . . . . . . . . . . . . . . . . . B. T Cell-Mediated Suppression Is Specific . . . . . . . . . . . . . . . . . . . . . . . . . C. Kinetics of Generation of Suppressor T Cells. . . . . . . . . . . . . . . . . . . . . . V. Evidence That the Generation of Effector T Cells Precedes the Generation of Suppressor T Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Concomitant Immunity Is Mediated by Ly 1-,2+T Cells.. . . . . . . . . . B. Kinetics of Generation and Loss of the T Cells That Mediate Concomitant Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Evaluation of the Evidence That Ly 1+,2- Suppressor T Cells Down-Regulate the Generation of Ly 1-,2+ T Cell That Mediated concomitant Immuni............................................. A. Suppressor T Cells Suppress the Generation Rather Than the Function of Effector Cells. . . . .. .. ..... . . .. . . .. ...... B. Comparison with Other Models of Tumor-Induced Suppressor T Cells . VII. Tumor-Induced Immunosuppression as the Explanation of Escape from Immunity.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Suppression of Antitumor Immunity as an Example of Transplantation Tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX. Immunotherapeutic Significance of the Generation and Subsequent Decay of Concomitant Immunity. . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Cyclophosphamide as an Immunotherapeutic Agent . . . . . . . . . . . . . . . B. Ionizing Radiation as an Immunotherapeutic Agent . . . . . . . . . . . . . . . . C. Endotoxin as an Immunotherapeutic Agent . . . . . . . . . . . . . . . . . . . . . . . D. Tumor Regression Caused by Intralesional Adjuvants. . . . . . . . . . . . . . . X. Conclusion ..................................................... References ......................................................

.

..

.

.

.

.. ...

.. .. ..

.

1 2 5 7 7 10 11 12 12 14 15 15 19 22 25 26 27 30 33 36 38

40

I. Introduction

This article will discuss the immune response to chemically induced, transplantable tumors in syngeneic mice. Therefore, it will deal only with those tumors that are immunogenic by virtue of their possession of tumor-specific, transplantation rejection antigens. It will avoid dealing with the question of whether spontaneous human tu1 ADVANCES IN CANCER RESEARCH, VOL. 45

Copyright 8 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.

2

ROBERT J. NORTH

mors are immunogenic because it is evident from an ongoing published discussion (Hellstrom and Hellstrom, 1983; Mastrangelo et al., 1984)that this question cannot yet be answered. The current view in this laboratory is that the immune response to a progressive immunogenic tumor is a model with which to analyze an unsuccessful immune response to replicating antigens. If some or most human tumors prove to possess tumor-specific transplantation antigens, then the discussion that follows is relevant to the human disease. If not, the discussion still is relevant to the immune response to replicating antigens in general and will help to explain why certain infectious and parasitic diseases become chronic or progressive. It surely would be surprising, however, if it turned out that all human tumors were nonimmunogenic. The question of whether tumors that have been transplanted repeatedly over a number of years have remained truly syngeneic also will not be discussed. It will suffice to say that there is no evidence to the contrary. Indeed, it is apparent that the immunity that many of these tumors can evoke today is the same, in terms of type and strength, as the immunity they evoked when they were first induced. It is worth pointing out in this connection, moreover, that while there are those who argue (Hewitt, 1979) that tumors that are transplanted over many years are more allogeneic than syngeneic, there are others who argue (Uyttenhove et al., 1983)that the progressive growth of such tumors is possible only because of the survival and emergence of nonimmunogenic, antigen-loss tumor variants. Obviously both points of view cannot be correct. If the first notion were correct it should follow that practically all spontaneous and chemically induced tumors, whether immunogenic or not at the time of their emergence, should become increasingly immunogenic with time. On the other hand, if there were selective pressure for the preservation of nonimmunogenic variants during in vivo passage, as described for the P815 mastocytoma (Uyttenhove et al., 1983),then most transplantable tumors should rapidly become nonimmunogenic. In the absence of substantial evidence for either idea, it remains possible that most syngeneic tumors, although probably composed of cells of different antigenicity, may in fact be relatively stable immunogenically (Rogers, 1984). II. Tumor lmmunogenicity

An immunogenic tumor is one against which a syngeneic host can be immunized. The immunogenicity of chemically induced murine tumors was revealed first by Foley (1953),whose method of immuniz-

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

3

ing consisted of removing an established tumor by ligation. He showed that removal of tumor by this means left its host resistant to growth of a subsequent implant of cells of that tumor, but not to the growth of implants of cells of other tumors. The demonstration of the immunogenicity of chemically induced tumors and the specificity of the immunity they engender was soon confirmed by others who immunized against growth of an implant by injection of heavily X-irradiated, nonreplicating tumor cells (Revesz, 1960) or by repeated injection of subtumorigenic doses of replicating tumor cells (Old et al., 1962). It is apparent, however, that the most favored method of testing for tumor immunogenicity remains the one described by Prehn and Main (1957) which consists of testing for immunity to growth of a tumor implant after removing a primary tumor by surgery. This is not to say that this is the best test for immunogenicity, since it is apparent that a proper comparison of the available tests has not been made. On the contrary, it is safe to state that all tests for tumor immunogenicity have been empirically derived and that too much faith is placed in negative results obtained with them (Hewitt et al., 1976). Indeed, results of a recent study of the immunogenicity of two “spontaneous” guinea pig leukemias led Key et al. (1984) to stress the importance of employing optimum immunization procedures before concluding that a tumor is devoid of tramplantation rejection antigens. Be this as it may, the demonstrations of immunogenicity of chemically induced tumors were soon followed by attempts to determine whether the immunity they evoke is mediated by cells or antibody. It was shown (Old et al., 1962) that immunity to growth of a tumor implant is cell mediated in nature in that it can be passively transferred from immunized donors to normal recipients with lymphoid cells, but not with serum. It is known from the results of more recent in viva experiments (Fernandez-Cruz et al., 1979; Greenberg et al., 1980; Leclerc and Cantor, 1980; Berendt and North, 1980; North, 1984a)and in uitro experiments (Rouse et al., 1973; Plata et al., 1973; Burton et al., 1975; Wagner et al., 1980), moreover, that immunity to syngeneic tumors is mediated by T cells. Needless to say, the knowledge that syngeneic tumors can possess transplantation rejection antigens presents the problem of explaining how such tumors escape destruction by host immunity. Obviously, the fact that immunogenic tumors exist argues against theories of immunosurveillance, and it is not surprising that there have been a number of attempts to explain how such tumors avoid rejection. It has been suggested, for example, that tumors escape immune destruction by hiding their surface antigens, an explanation based on evidence

4

ROBERT J. NORTH

(Boyse et al., 1967; Hilgers et al., 1980) that tumor cells can modulate their surface antigens under immunological pressure. This explanation is related to a more recent one based on evidence (Bosslet and Schirnnacher, 1981; Uyttenhove et al., 1983) that progressive tumor growth, in the face of an antitumor immune response, is made possible by the emergence of stable clones of tumor cells that are antigenloss variants. Yet another suggestion for escape is that tumors avoid confrontation with host effector cells by secreting antiinflammatory factors that function to prevent host cells from migrating across vascular endothelium into the tumor mass (Fauve et d.,1974). This suggestion is supported by the findings (Pike and Snyderman, 1976) that implantation of certain tumors, or injection of small molecular weight extracts from them, can inhibit the entry of mononuclear cells, particularly macrophages, into peritoneal inflammatory exudates. It should be realized, however, that the idea of the secretion of antiinflammatory products by murine tumor cells is difficult to reconcile with the knowledge (Evans, 1973; Eccles and Alexander, 1974; Haskill et al., 1975; Dye and North, 1980) that solid and ascites tumors in mice contain very large numbers of host mononuclear cells, including macrophages. Another plausible explanation for the escape of immunogenic tumors is that the immunity these tumors evoke is too weak and is generated too late to reject an already established and rapidly growing tumor mass. It has been suggested (Old et al., 1962; Old and Boyse, 1964)that this allows the tumor to “sneak through” immune defenses, an idea in keeping with the knowledge that, whereas implantation of a large number of cells of a given tumor can result in early rejection of the tumor that emerges, implantation of a small number of tumor cells results in progressive tumor growth. Presumably the larger implant provides enough antigen to engender an immune response early enough to be effective. This “sneak through” hypothesis, because it is based on the weakness and inadequacy of antitumor immunity, is related to the most recent and popular explanation of escape which states that immunogenic tumors are able to escape immune defenses because they induce suppressor T cells. Evidence that suppressor T cells suppress antitumor immunity has been the subject of several articles and reviews (Naor, 1979; Greene, 1980; Schatten et al., 1984b), all of which make it clear that there is ample evidence for the presence of suppressor T cells in a tumor-bearing host. It is apparent, however, that some of the evidence is indirect and that a detailed hypothesis of tumor escape based on the negative regulatory function of suppressor T cells

DOWN-REGULATION OF ANTITUMOR

IMMUNE RESPONSE

5

has not been formally presented. There has been no suggestion in most cases as to the nature of immunity in the tumor-bearing host that suppressor T cells are supposed to suppress. Presumably, if suppressor T cells are responsible for tumor escape, then they must either prevent an antitumor immune response from being generated in the first place or they must function to down-regulate one that is in the process of being generated. Available evidence suggests that the second possibility is the more likely one, because it is well documented that progressive growth of tumors of proved immunogenicity evokes the generation of an immune response in the form of concomitant immunity. Therefore, before discussing the functional significance of suppressor T cells, it is first necessary, by way of introduction, to briefly discuss the evidence that antitumor immunity is generated. 111. Concomitant Immunity as the Unsuccessful Response to Tumor Growth

Concomitant antitumor immunity is a paradoxical state of acquired immunity that enables a host with a progressive tumor to neutralize the growth of an implant of cells of the same tumor given at a distant site. There is a relatively large literature on the subject of concomitant immunity (reviewed by Vaage, 1971; Gorelik, 1983; Tuttle et al., 1983) going back to the descriptions of it by Ehrlich (1906) and by Bashford et al. (1908)at the turn of the century. The interpretation that concomitant immunity serves no purpose, in that the tumor continues to grow unrestrictedly, has been negated by published evidence showing that failure of a host to generate concomitant immunity results in a much shorter survival time because of a faster dissemination of tumor cells and growth of tumor metastases (Milas et al., 1974). This has been observed in animals that fail to generate concomitant immunity because of having been immunodepressed by exposure to X-irradiation (Deodar and Crile, 1969; Yuhas et al., 1975) or by treatment with antilymphocyte serum. Faster development of systemic disease has also been observed in mice that have been made T cell deficient by thymectomy and lethal irradiation and restored with bone marrow (Kearny and Nelson, 1973). Again, there is evidence showing (Gershon and Kondo, 1971) that excision of a primary tumor can result in failure to generate concomitant immunity and consequently in earlier death of the host from the more rapid growth of seeded metastases. This last-mentioned finding presumably depended on the timing of tumor excision. The immunological consequences of tumor excision as it relates to concomitant immunity will be discussed later.

6

ROBERT J. NORTH

It is necessary to point out at this stage that the majority of published evidence shows, in agreement with the findings about tumor immunogenicity in general, that concomitant immunity is specific for the tumor that evokes its generation. There is some evidence, however, that concomitant immunity can be nonspecific, but only after the primary tumor becomes very large. Kearny and Nelson (1973) have shown, for example, that concomitant immunity to several chemically induced fibrosarcomas of recent origin is specific during early stages of tumor growth, but is nonspecific at later stages. It should be pointed out, however, that even the nonspecific phase of concomitant immunity might be specifically mediated in that it might depend on the immunologically mediated activation of a nonspecific defense mechanism such as activated macrophages. Alternatively, the nonspecific component may not be immunologically mediated, but may represent an additional antitumor mechanism that is superimposed on the specific mechanism. It was demonstrated that lymph node T cells from a tumor-bearing, concomitant immune donor can neutralize, essentially in a specific manner, the growth of an implant of tumor cells in a normal or irradiated recipient, although according only to the Winn neutralization assay (North and Kirstein, 1977). The T cell basis of concomitant immunity is further evidenced by the demonstration (Biddison et al., 1977; Ting et al., 1982; Tuttle et al., 1983) that its generation in response to the growth of certain tumors is associated with the acquisition of T cells that are specifically cytolytic for cells of these tumors in uitro. More will be said about cytolytic T cells later when the kinetics of the generation of concomitant immunity are discussed. It needs to be pointed out at this time, however, that an important aspect of concomitant immunity is that it can undergo rapid decay after the tumor reaches a certain critical size. This eclipse of concomitant immunity has been studied and discussed in some detail by Vaage (1971, 1973, 1977) and by Youn et al. (1973). There undoubtedly are some cases where concomitant immunity does not decay (Garelik, 1983).However, it needs to be determined whether the failure of concomitant immunity to undergo decay is more apparent than real in that the decay is masked by the late development of a mechanism of nonspecific resistance that is not T cell mediated. This would be evidenced by retention of nonspecific resistance to growth of a challenge implant in spite of the loss by the host of T cells capable of passively transferring specific immunity to appropriate recipients (Kearny et al., 1975).Be this as it may, examples of the rapid decay of specific concomitant antitumor immunity are important because they

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

7

provide a reason for postulating that suppressor T cells are generated in response to tumor growth. IV. Evidence That Tumor Growth Induces Suppressor T Cells

If the generation and subsequent loss of concomitant immunity is a common consequence of the growth of immunogenic tumors, this surely would need to be taken into account in the design of immunotherapeutic modalities because it would mean that any attempt to cause the regression of an immunogenic tumor by active or adoptive immunotherapy would represent an attempt to augment or superimpose an immune response either on an already developing concomitant immune response or on a concomitant immune response that is undergoing decay. It would be highly significant, moreover, if the decay of concomitant immunity proved to be an active process mediated by suppressor T cells, because a mechanism of active suppression of immunity might explain why it has proved so difficult to cause the regression of already established tumors by intralesional injection of immunoadjuvants or by the passive transfer of tumor-sensitized T cells from immunized donors. Indeed, we considered it highly likely that the presence of suppressor T cells was responsible for documented failures to cause tumor regression by adoptive immunotherapy (Rosenberg and Terry, 1977). It was reasoned, in turn, that if the presence of a mechanism of T cellmediated immunosuppression is responsible for the refractoriness of an immunogenic tumor to the antitumor function of passively transferred tumor-sensitized T cells, it should be possible to make the tumor susceptible to intravenously infused T cells by growing it in a recipient that has been rendered incapable of generating suppressor T cells. A. Ly 1+,2- SUPPRESSOR T CELLSARE THE MAJOR OBSTACLE TO ADOPTIVE IMMUNOTHERAPY OF ESTABLISHED TUMORS If the presence of tumor-induced suppressor T cells in a recipient animal is responsible for the refractoriness of its established tumor to the action of intravenously infused sensitized T cells from an immune donor, it should be possible to cause the regression of the tumor by preventing the production of suppressor T cells. This prediction was tested (Berendt and North, 1980) by determining whether passive transfer of tumor-sensitized T cells from immune donors would cause the regression of an established tumor growing in recipient mice that

8

ROBERT J. NORTH

were incapable of generating suppressor T cells because of having been made T cell deficient 6 weeks earlier by thymectomy and lethal y-radiation and protected with bone marrow (TXB mice). The donors of immune T cells were immunized 3 weeks earlier by the subcutaneous injection of an admixture of living tumor cells and Corynebacterium paruum (Dye et al., 1981). This method of immunization is known to leave the host specifically immune to the growth of an implant of tumor cells for many weeks and with splenic T cells capable of transferring this immunity to normal recipients. Experiments were performed with the nonmetastatic, methylcholanthrene-induced Meth A fibrosarcoma, syngeneic in BALB/c mice, and with the P815 mastocytoma syngeneic in DBA12. It was found (Berendt and North, 1980; Dye and North, 1981), in agreement with the general experience of others, that intravenous infusion of 1 organ equivalent of spleen cells from immune donors failed to have any effect on a tumor growing in immunocompetent recipients. In contrast, infusion of the same number of immune spleen cells caused complete regression of the same-sized tumor growing in TXB mice. It was apparent, therefore, that immunocompetent, tumor-bearing recipients possessed a T cell-dependent mechanism that blocked the capacity of passively transferred immune T cells to express their antitumor function. It was reasoned that if this were true, it should be possible to reveal the presence of this T cell-dependent mechanism of suppression by showing that it can be passively transferred. In other words, it was considered likely that passive transfer of spleen cells from an immunocompetent donor bearing a relatively large tumor should block the expression of antitumor immunity by passively transferred immune T cells in TXB recipients. This prediction proved correct in that passive transfer of 1 organ equivalent of spleen cells from immunocompetent donors bearing a 14- to 16-day (1 cm) Meth A tumor prevented 1 organ equivalent of immune spleen cells infused 3 hr earlier from causing regression of a 4-day tumor in TXB recipients. Moreover, because the same number of spleen cells from normal mice failed to prevent the expression of adoptive immunity in TXB recipients, it was concluded that the suppressor mechanism was tumor induced. The basic suppressor assay is depicted diagrammatically in Fig. 1. Evidence that the suppressor mechanism is T cell mediated came from experiments that determined whether the ability of suppressor spleen cells to prevent immune T cells from causing tumor regression in TXB recipients is abolished by treating the suppressor cells with monoclonal anti-Thy 1.2 antibody and complement. It was found (Berendt and North, 1980; Dye and North, 1981) that the suppressor

DOWN-REGULATION O F ANTITUMOR IMMUNE RESPONSE

Immunized donor (memory cells)

/

\

Immune lymphocytes

/

9

Tumor-bearing donor (Suppressor cell donor)

/

Suppressor lymphocytes

\ / €!El-

TXB Recipients

tumor rejected

tumor grows

FIG.1. Diagrammatic representation of the in uiuo suppressor assay. It measures the capacity of lymphocytes from a tumor-bearing donor to prevent lymphocytes from a preimmunized donor from causing regression of an established tumor in a test recipient made T cell deficient by thymectomy and irradiation.

capacity of the spleen cells was completely eliminated by treatment with anti-Thy 1.2 antibody and complement. These results left little doubt, therefore, that a progressively growing immunogenic tumor eventually evokes in its host the generation of a population of suppressor T cells. These findings were soon confirmed by others (Bonventre et al., 1982) who utilized essentially the same methods, except that athymic nude mice were employed, instead of TXB mice, as tumor-bearing test recipients. Additional experiments were performed to determine the Ly phenotype of the T cells that passively transfer suppression. The results of these experiments showed (North and Bursuker, 1984)that the ability of splenic T cells from donors with a 18day Meth A tumor to inhibit the expression of adoptive immunity against an established tumor in TXB test recipients was abolished by treating the suppressor spleen cells with anti-Ly 1 antibody and complement, but not by treating them with anti-Ly 2 antibody and complement. Therefore, the suppressor T cells that function in this model are of the Ly 1+,2- phenotype, a finding that makes them different from the suppressor T cells that function in other models of suppression of antitumor immunity (Schatten et al., 1984). However, experiments with the P815 masto-

10

ROBERT J. NORTH

cytoma revealed (North and Dye, 1985) that progressive growth of this tumor also evokes the generation of Ly 1+,2- suppressor T cells. B. T CELL-MEDIATED SUPPRESSION Is SPECIFIC The results of certain published studies have been interpreted as indicating that progressive growth of a tumor can result in a state of generalized immunodepression. This is said to occur, for example, in those cases where tumor growth evokes the production of suppressor macrophages. Indeed, there is evidence (Kirchner et al., 1974; Kruisbeek and Hees, 1977; Mitzushima et al., 1984) that animals bearing large tumors can possess macrophages capable of nonspecifically suppressing T cell responses in uitro. However, the interpretation that this represents evidence for nonspecific macrophage-mediated immunosuppression in uivo was recently challenged on the basis of results which show (Forni et al., 1982) that tumor-bearing mice that possess macrophages capable of inhibiting immune responses to certain antigens in uitro nevertheless are perfectly capable of mounting normal immune responses to antigens in uiuo. This surely indicates that caution should be exercised in postulating the existence mechanism of immunosuppression purely on the basis of in uitro evidence. The same can be said for theories of macrophage-mediated immunosuppression in animals chronically infected with pathogenic bacteria or parasites. Even if the magnitude of immune responses in such animals was greatly decreased according to in uiuo assays, the onus is on the experimenter to show that the reduced immune responsiveness is not the result of destruction of most of the antigen by a highly activated macrophage system generated in response to infection. There is evidence in this connection that a highly activated macrophage system can be generated in response to growth of immunogenic tumors, including the Meth A fibrosarcoma (North and Kirstein, 1977). However, no evidence was found in this laboratory to indicate that this tumor causes the generation of a state of generalized immunosuppression. For example, mice bearing Meth A tumors large enough to have induced suppressor T cells were shown to have retained a normal capacity to generate and express immunity to a tumor allograft (Berendt and North, 1980).Such mice also retained a normal ability to generate T cell-mediated immunity to infection with bacterial and viral pathogens (Bonventre et al., 1982). Indeed, reciprocal passive transfer experiments with the P815 mastocytoma and the syngeneic P388 lymphoma showed that T cell-mediated suppression of adoptive immunity to these tumors is specific (Dye and North, 1984). Thus,

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

11

whereas suppressor T cells from a donor bearing a 15-day P815 tumor were capable of suppressing the ability of passively transferred P815immune T cells to cause regression of a P815 tumor in TXB test recipients, these same suppressor cells were not capable of preventing P388-immune T cells from causing regression of the P388 tumor in TXB recipients. Reciprocally, suppressor T cells from donors with a 15-day P388 tumor were capable of inhibiting the expression of adoptive anti-P388 immunity, but not the expression of adoptive anti-P815 immunity. OF SUPPRESSOR T CELLS C. KINETICS OF GENERATION

Knowledge about the time that suppressor T cells are generated is important, because if they are generated late, it would indicate that they function to down-regulate an already ongoing concomitant immune response. Early generation of suppressors, on the other hand, would indicate that they prevent concomitant immunity from being generated. Needless to say, the first possibility seems more likely because it is in keeping with numerous published papers showing that concomitant immunity is, in fact, generated in response to tumor growth. However, it has been demonstrated with a syngeneic fibrosarcoma by one group of workers (reviewed by Greene, 1980) that suppressor T cells are generated in thymus, spleen, lymph nodes, and bone marrow as early as 24 hr after injecting tumor cells subcutaneously. This prompted a study in this laboratory of the kinetics of generation of suppressor T cells in response to growth of the Meth A and P815 tumors. Measuring the time of onset and the kinetics of generation of suppressor T cells involved the use of the standard in uiuo suppressor assay described above. It already had been shown with the P815 mastocytoma (Dye and North, 1981) that the degree to which adoptive immunity is suppressed by suppressor T cells is determined by the number of suppressor T cells infused into TXB test recipients relative to the number of immune T cells infused. Thus, the time course of generation of suppressor T cells was investigated by an experiment that involved infusing TXB recipient mice bearing a 4-day tumor with 1.5 X 10s immune spleen cells and dividing the mice into groups according to whether they were infused 1 hr later with 1.5 x 108 spleen cells from donors bearing a 3-, 6-, 9-, 12-, 15-,or 18-day tumor. It was found (North and Bursuker, 1984) that cells capable, on passive transfer, of inhibiting the expression of passively transferred immunity against an established tumor in TXB recipients were not acquired

12

ROBERT J. NORTH

until about day 9 of tumor growth when the tumor was about 8 mm in diameter. It was found, in addition, that suppressor cells progressively increased in number after this time until about day 18. Because these same kinetics were obtained with the P815 mastocytoma (North and Dye, 1985), it seems likely that this relatively late generation of suppressor T cells will be found to be a consequence of the growth of many immunogenic tumors, provided the same physiological assay is employed. Obviously, if the function of suppressor T cells is to down-regulate concomitant immunity, the cells that mediate concomitant immunity should be progressively lost after day 9 as suppressor cells are progressively acquired. It was necessary to measure next, therefore, the kinetics of generation of the cells that mediate concomitant immunity. V. Evidence That the Generation of Effector T Cells Precedes the Generation of Suppressor T Cells

A. CONCOMITANT IMMUNITY Is MEDIATEDBY Ly 1-,2+ T CELLS It was known from a previous study (Berendt and North, 1980) that growth of the Meth A fibrosarcoma evokes the generation of concomitant immunity. More recent experiments revealed (North and Bursuker, 1984) that concomitant immunity, as measured by resistance to growth of a standard Meth A challenge implant, is generated between days 6 and 9 of tumor growth and declines progressively thereafter, until about day 16 when immunity to growth of a challenge implant can no longer be detected. These results show, therefore, that concomitant immunity to the Meth A fibrosarcoma begins to undergo decay at about the time that suppressor cells begin to be generated, according to results described in the foregoing section. However, to further analyze the relationship between concomitant immunity and suppressor T cells, it was considered essential to develop an adoptive immunization assay that would enable concomitant immunity to be passively transferred. This would allow the cells that passively transfer immunity to be identified and their properties determined. A suitable adoptive immunization assay would also provide a means to measure, against time of tumor growth, changes in the relative number of the T cells that mediate concomitant immunity. Moreover, if it proved possible to passively transfer concomitant immunity systemically and have it expressed not only against growth of an implant, but against an

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

13

already established tumor, it would allow the expression if immunity to be measured immediately after sensitized T cells are passively transferred. However, to be able to cause the regression of a recipient’s established tumor by the passive transfer of lymphocytes from a donor that is incapable of causing the regression of its own tumor would seem an unrealistic expectation, to say the least. Even so, it was not possible until recently to demonstrate, in a routine fashion, that passive transfer of T cells, even from a hyperimmunized donor, can cause regression of an already established tumor in a recipient. This became possible only after it was realized that suppressor T cells generated in the recipient function to inhibit the antitumor function of passively transferred tumor-sensitized T cells (North et al., 1982). It was considered possible, therefore, that concomitant immunity might be expressed against an established tumor after passive transfer, provided the tumor-bearing recipient is rendered incapable of generating suppressor T cells and that its tumor is not too large. An attempt was made, therefore, to cause regression of a 3-day tumor in recipients by the passive transfer of spleen cells from donors bearing a 9-day tumor. The recipients were exposed to 500 rad of y-irradiation 1 hr before they were infused with donor spleen cells. The rationale for exposing the recipients of 500 rad of y-radiation was based on the results of a preceding study (North, 1984b) which showed that sublethal y-radiation facilitates the expression of immunity by passively transferred immune T cells by preventing the recipients from producing suppressor T cells. It was found (North and Bursuker, 1984) that 1organ equivalent of spleen cells from donors with a 9-day tumor caused, on passive transfer, complete regression of a 3-day tumor in y-radiated recipients, but not in normal recipients. It also was shown that the spleen cells that transferred this immunity were T cells in that they failed to cause tumor regression if they were treated with anti-Thy 1.2 monoclonal antibody and complement. Moreover, their antitumor function was totally eliminated by treatment with anti-Ly 2 monoclonal antibody and complement, but was not affected at all by treatment with anti-Ly 1monoclonal antibody and complement. Therefore, the cells that passively transfer concomitant immunity to the Meth A fibrosarcoma have the Ly 1-,2+ phenotype of cytolytic effector T cells. These same results were obtained when concomitant immunity to the P815 mastocytoma was investigated (North and Dye, 1985), and this is consistent with the demonstration by others (Tuttle et al., 1983)that the generation of concomitant immunity to the P815 mastocytoma is associated

14

ROBERT J. NORTH Lymph node or spleen cells J

Recipient with small tumor

Tumor-bearing donor with concomitant immunity to growth of implant

given 500 R before transfer

t

tumor rejected

FIG.2. Passive transfer assay for measuring concomitant immunity. Mice with a progressive tumor acquire lymphocytes that are capable, on passive transfer, of causing tumor regression in recipients given 500 rad of y-radiation 1 hr before passive transfer.

with the generation of T cells capable of lysing P815 targets in vitro. Indeed, it also recently was shown in this laboratory (North and Dye, 1985)that the presence in mice with a 9-day P815 tumor of Ly 1-,2+ T cells capable of causing regression of a tumor in y-irradiated recipients is associated with the presence of T cells that can lyse P815 tumor cells in uitro. The adoptive immunization assay for measuring concomitant immunity is shown diagrammatically in Fig. 2.

B. KINETICS OF GENERATION AND Loss OF

THE

T CELLS

THATMEDIATECONCOMITANT IMMUNITY To measure changes against time of tumor growth in the relative number of T cells capable of passively transferring concomitant immunity, 1 organ equivalent of spleen cells from donors bearing a 3-, 6-, 9-, 12-, 15-, or 18-day intradermal Meth A tumor was infused into yirradiated recipients bearing a 3-day intradermal tumor. It was found (North and Bursuker, 1984)that T cells capable of causing regression of the recipient tumor were generated first on about day 6 of tumor growth, reached peak number on day 9, and were then progressively lost until day 15 when their presence could no longer be detected. A more recent study revealed (North and Dye, 1985) that the generation and loss of T cells that passively transfer concomitant immunity to the

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

15

P815 mastocytoma have the same kinetics. This study with the P815 mastocytoma showed, in addition, that these T cells are generated and lost in concert with the generation and loss of T cells capable of lysing P815 tumor cells in uitro. It seems reasonable to predict that similar results will be obtained with other immunogenic tumors. V1. Evaluation of the Evidence That Ly 1+,2-Suppressor T Cells Down-Regulate the Generation of Ly 1-,2+ T Cell That Mediated Concomitant immunity

A. SUPPRESSOR T CELLSSUPPRESS THE GENERATION RATHER THANTHE FUNCTION OF EFFECTOR CELLS The evidence discussed in the foregoing sections would appear to leave little doubt that the progressive decay of concomitant immunity after day 9 of growth of the Meth A fibrosarcoma and P815 mastocytoma is caused by the negative immunoregulatory function of suppressor T cells that are acquired progressively from day 9 on. These data can be represented diagrammatically as shown in Fig. 3, where it can be seen that effector T cells do not disappear suddenly, but are progressively lost as suppressor T cells are progressively acquired. Therefore, the development of suppression is seen as a progressive increase in the ratio of suppressor T cells to effector T cells, until suppressor T cells become completely dominant in the adoptive immunization assay. On day 12 of growth of the Meth A tumor, for example, the spleen contains a mixture of effector and suppressor cells such that neither type of cell is totally dominant. This undoubtedly is the reason why treatment of spleen cells from tumor-bearing donors with anti-Ly 2 antibody and complement always appears to increase the level of suppression passively transferred, whereas treatment with anti-Ly 1 antibody and complement increases the Ievel of concomitant immunity transferred (North and Bursuker, 1984). It should be realized, however, that the aforementioned evidence for T cell-mediated suppression of concomitant antitumor immunity is correlative rather than causal in that the suppressor assay measures the ability of suppressor cells to inhibit the expression of passively transferred immunity from preimmunized donors rather than from concomitantly immune donors. Causal evidence would consist of the demonstration that an infusion of suppressor cells from donors with a 14- to 16-day Meth A tumor can prevent concomitant immunity from being generated. This recently was shown to be the case (Bursuker

.

16

ROBERT J NORTH

3

6

9

12

15

DAYS FIG.3. Diagrammatic representation of the kinetics of generation of effector T cells and suppressor T cells during growth of an immunogenic tumor. After day 9 of tumor growth there is a progressive increase in the ratio of suppressor to effector T cells.

and North, 1985)by an experiment that measured immunity to a challenge implant and the acquisition of T cells capable of passively transferring this immunity in mice that were infused with suppressor cells at the time their primary tumors were initiated. It was found that mice that were infused with suppressor cells failed to generate immunity to a challenge implant or to acquire T cells capable of passively transferring immunity to normal recipients. Thus, there is little doubt that suppressor T cells can block the generation of concomitant immunity. Presumably, the progressive blocking of further effector T cell production after day 9 of tumor growth results in decay of immunity

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

17

because there is inadequate replacement of effector T cells with a short functional life-span. The interpretation that suppressor T cells down-regulate concomitant immunity by inhibiting the generation rather than the function of effector T cells is supported by the results of another study that investigated the way that passively transferred suppressor T cells inhibit the expression of adoptive immunity against an established tumor in TXB test recipients. It was reasoned, in designing this study, that because there is invariably a 6-day delay before tumor regression begins in TXB recipients infused with tumor-sensitized T cells from preimmunized donors (not concomitantly immune donors), adoptive immunity cannot be expressed until an adequate number of cytolytic T cells are generated in the TXB recipients themselves. This reasoning proved to be correct in that it was found (Mills and North, 1983) that the onset of regression of a P815 tumor in TXB recipients infused with spleen cells from preimmunized donors was immediately preceded by the generation in the recipients’ draining lymph nodes of T cells capable of lysing P815 tumor cells in uitro. In these experiments, the donors of immune cells were mice that had been immunized several weeks earlier by injection of an admixture of living tumor cells and C. paroum. The immune cells were harvested, therefore, long after the donors had generated a cytolytic T cell response (Mills et al., 1981).Thus, the T cells that were passively transferred were memory cells that imparted to the TXB recipients the capacity to generate a cytolytic T cell response of sufficient magnitude to destroy the test tumor. Indeed, the adoptive, secondary cytolytic T cell response in the recipients was found to be similar in magnitude to that which was generated by the immunized donors under the augmenting influence of C. paroum. In view of the need for the generation of effector T cells by the TXB recipients themselves before passively transferred immunity can be expressed, it was logical to postulate that passively transferred suppressor T cells suppress the expression of adoptive immunity by inhibiting the ability of the recipients to generate effector T cells. This was found to be the case by an experiment which showed (Mills and North, 1983) that whereas infusion of tumor-bearing TXB recipients with immune cells alone resulted, after a 6-day delay, in an appreciable cytolytic response that immediately was followed by tumor regression, infusion of immune cells plus suppressor cells resulted in a cytolytic T cell response of very much lower magnitude and which was associated with a failure on the part of the recipient to cause

18

ROBERT J. NORTH

tumor regression. It seems reasonable to hypothesize, therefore, that suppressor T cells inhibit the expression of adoptive immunity by suppressing the capacity of the recipient to generate an adequate number of cytolytic effector T cells. This is in keeping with the additional finding that infusion of immune T cells into T cell-intact, tumorbearing recipients failed to result in a cytolytic T cell response or in tumor regression. A lesson to be learned from this study of suppression of adoptive immunity is that it is essential to know, before attempting to interpret the results of adoptive immunization experiments, the state of immunity possessed by the donors. Immune donors can be actively immune because they possess cytolytic effector T cells or they can be memory donors because they no longer possess cytolytic T cells. The properties that distinguish a state of active antitumor immunity from a state of memory immunity, and evidence that the speed at which adoptive immunity is expressed in tumor-bearing recipients is determined by the type of immunity passively transferred are the subjects of a recent publication by Dye and North (198413). This publication shows, among other things, that whereas passive transfer of active immunity results in immediate onset of regression of the recipient’s established tumor, passive transfer of memory immunity does not cause tumor regression until after an appreciable delay. It is apparent that most published studies of adoptive immunization against tumor syngrafts and allografts are based on results obtained with the passive transfer of sensitized Ly 1+,2- memory or helper T cells rather than on results obtained from the passive transfer of cytolytic T cells. It is not correct to conclude, on the basis of evidence that antitumor immunity can be passively transferred with Ly 1+,2- T cells at the exclusion of Ly 2+ T cells, that Ly 2+T cells play no role in the rejection process (Loveland et al., 1981).In the first place, because the donors that were employed almost certainly were memory immune, they probably no longer possessed specifically sensitized Ly 2+ cytolytic T cells capable of transferring immunity. In the second place, the TXB recipients that were routinely employed in published studies of adoptive immunization are known to possess precursors of cytolytic T cells that are capable of maturing into functional cytolytic T cells in the presence of helper T cells and antigen (Gillis et al., 1979; Duprez et al., 1982; LeFrancois and Bevin, 1984),a situation that exists in a tumor-bearing TXB recipient infused with tumor-sensitized helper T cells. Taken as a whole then, the evidence obtained in this laboratory with the Meth A fibrosarcoma and P815 mastocytoma supports the interpretation that suppressor T cells function to suppress the generation rather than the

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

19

function of effector T cells. This evidence is supported by evidence published by others which shows that a host bearing a large P815 mastocytoma (Takei et al., 1976,1977) or a large thymoma (Frost et al., 1982)possesses T cells that can specifically suppress the generation of T cells cytolytic for cells of these tumors in uitro.

B. COMPARISON WITH OTHERMODELSOF TUMOR-INDUCED SUPPRESSOR T CELLS It needs to be pointed out that the Ly 1+,2- suppressor T cells revealed by the physiological assays employed in this laboratory are different from tumor-induced suppressor T cells described by other workers. For instance, suppressor T cells generated by A/J mice in response to the syngeneic S 1509A sarcoma have been shown to be of the Ly 1-,2+ phenotype (reviewed by Greene, 1980). S1509A-induced suppressor cells, moreover, are generated in the thymus, spleen, lymph nodes, and bone marrow as early as 24 hr after implanting 1O'j S1509A cells subcutaneously. Thus, they are generated much earlier than are the suppressor cells being studied in this laboratory. S1509Ainduced suppressors are different also in terms of their distribution, because the results of an ongoing study (DiGiacomo and North, 1986) has revealed that suppressor T cells generated in response to the Meth A fibrosarcoma can be found only in the draining lymph node and the spleen, that is, only in .responding lymphoid tissue where the generation of effector T cells is known to occur. It surely seems reasonable to suggest that down-regulation of the production of immune effector cells is likely to take place at sites where effector cells are produced. Considering the very early generation and wide distribution of suppressor T cells generated in response to the S1509A tumor, it is obvious that they must mediate the suppression of a different type of immunity than that generated in response to growth of the Meth A fibrosarcoma and P815 mastocytoma. Whereas the suppressors generated in response to the Meth A and P815 tumors function to downregulate an already ongoing concomitant immune response, those generated in response to the S1509A sarcoma would appear to function to prevent any immunity from being generated in the first place. This would explain why the purported elimination of S 1509A-induced suppressor T cells by intravenous infusions of anti-I-J antibodies at the time of tumor implantation results in a reduced rate of growth of the tumor that emerges (Drebin et al., 1983). Indeed, it has been demonstrated on numerous occasions (Perry et al., 1978) that

20

ROBERT J. NORTH

S1509A-induced suppressor T cells are I-J+. Nevertheless, there is a need, in view of the importance of this suppressor model, to formally show that the decreased rate of growth of the S1509A tumor that occurs after infusion of anti-I-J antibody is caused by the elimination of suppressor T cells in the first place, and the consequential earlier generation of immune effector T cells, in the second. The ability of anti-I-J serum to slow the rate of growth of the S1509A tumor immediately after implantation would indicate an exceedingly early generation and expression of immunity in the absence of suppression. It might even be argued that this type of suppression is the reason why some tumors appear to be nonimmunogenic. However, the S1509A sarcoma is immunogenic, and this means that it probably evokes the generation of concomitant immunity. Another problem is that suppressor T cells with the same characteristics as those induced by the S1509A sarcoma have been described for the Meth A fibrosarcoma in another laboratory. According to the results of this other study, Meth A-induced suppressor cells are also 1-J+,Ly 1-,2+ T cells (Hawrylko, 1982; Hawrylko et al., 1982). Moreover, like those generated in response to the S1509A sarcoma, they are generated as early as 2 days after the tumor is initiated and are widely distributed. However, Meth A-induced suppressors are progressively lost after day 6 of tumor growth. Indeed, a closer examination of the evidence for these Meth A-induced, Ly 1-,2+ suppressor T cells shows that these cells are not defined in terms of their capacity to inhibit, on passive transfer, immunity to growth of a tumor implant in preimmunized recipients, but in terms of their ability to inhibit an intrafootpad delayed-type sensitivity reactions to injections of tumor cells. In fact, these suppressor cells appear not to suppress the expression of antitumor immunity in the same recipients. It is surely significant, therefore, that the 1-J+, Ly 1-, 2+ T cells generated in response to the S1509A sarcoma currently are also being studied in terms of their ability to inhibit delayed-sensitivity reactions to injections of S1509A cells in preimmunized recipients (Shatten et al., 1984a). The possibility must be considered, therefore, that there is no contradiction between the results obtained in this laboratory and those obtained elsewhere with the S1509A sarcoma and the Meth A fibrosarcoma. The Ly 1+,2- suppressor described in this laboratory are generated progressively in concert with the progressive decay of concomitant immunity and are defined in terms of their capacity to prevent the regression of an established tumor by a powerful mechanism of adoptive antitumor immunity that depends for its expression on the generation of effector T cells. In contrast, the Ly 1-,2+, I-J+suppressor cells described in other laboratories are generated before concomitant im-

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

21

munity and are defined mainly in terms of their capacity to suppress delayed-type sensitivity reactions in preimmunized donors. The problem that remains, therefore, is to determine the functional significance of delayed-type sensitivity, for although it is surely a correlate of immunity, its functional role in the immune response to progressive tumor growth remains unclear. On the basis of the timing of their generation, the physiological nature of the assay employed to reveal them, and the type of immunity they down-regulate, the Ly l+,2- suppressor T cells described with the Meth A fibrosarcoma and P815 mastocytoma in this laboratory bear little resemblance, then, to Ly 1-,2+ suppressor T cells induced by the S1509A sarcoma and Meth A fibrosarcoma as described in other laboratories. In fact, they bear a closer resemblance to the suppressor T cells generated by genetically susceptible BALBlc mice in response to progressive infection with the protozoan parasite, Leishmania tropica, for it has been shown (Liew et al., 1982) that BALBlc mice with progressive leishmaniasis generate a population of I-J-, Ly 1+,2- T cells that are capable, on passive transfer, of preventing immunized mice from inactivating a Leishmania challenge infection. The test recipients in this model of suppression are BALB/c mice that are sublethally X-irradiated to prevent them from generating suppressor T cells. The evidence obtained from this model of leishmaniasis has been interpreted to mean that the dominant negative regulatory function of suppressor T cells in BALBlc mice is the reason why these mice, in contrast to resistant strains of mice, permit the development of progressive disease. Insofar as the suppressor T cells generated in response to the Meth A fibrosarcoma and P815 mastocytoma being of the Ly 1+,2- phenotype is concerned, there are numerous other examples of immunosuppression mediated by T cells of this phenotype. These include suppression of in vitro immunity to minor (Macphail and Stuttman, 1982) and major (Holan and Mitchison, 1983) histocompatibility antigens, suppression of graft-versus-host reactions (Van Bekkum and KuaanShanzer, 1983; Rolink and Gleichmann, 1983),and suppression of H2-restricted interactions between helper T cells and B cells (Asano and Hodes, 1983). Indeed, Ly 1+,2- suppressor T cells recently have been shown to be generated in response to chronic UV-irradiation and to be capable, on passive transfer, of preventing recently reconstituted lethally X-irradiated or UV-irradiated mice from rejecting highly immunogenic UV-induced fibrosarcomas (Ulrich and Kripke, 1984). In this model of antitumor immunity, chronic UV-irradiation causes the emergence of fibrosarcomas which are too immunogenic to grow in normal mice, but which grow when transplanted to lethally X-irradi-

22

ROBERT J. NORTH

ated or UV-irradiated mice (Kripke, 1981). Although this model of suppression does not analyze the escape of immunogenic tumors in normal mice, mechanistically it is similar to the model employed in this laboratory. The possibility remains, nevertheless, that Meth A-induced and P815-induced suppression ultimately is mediated by Ly 1-,2+ T cells, because it has been shown in the case of certain models of suppression of DTH (Gershon, 1980; Benacerraf et aZ., 1982) that Ly 1+,2- T cells can function to induce the generation of Ly 1-,2+ T cells that presumably are the ultimate mediators of suppression. It should be pointed out, however, that if Ly 1-,2+ T cells are involved in the down-regulation of concomitant immunity to the Meth A fibrosarcoma and P815 mastocytoma, it is difficult to understand why treatment of spleen cells of suppressor donors with anti-Ly 2 monoclonal antibody and complement had not the slightest effect on the ability of the cells to suppress the expression of adoptive immunity. After all, the suppressor cells were harvested from donors in which concomitant immunity was being actively suppressed, and the tumor-bearing test recipients were TXB mice that had no capacity of their own to make suppressor T cells. Therefore, until evidence to the contrary is obtained, it must be assumed that Ly 1+,2- T cells are the actual suppressors of concomitant antitumor immunity to the Meth A fibrosarcoma and P815 mastocytoma.

VII. Tumor-Induced Immunosuppression as the Explanation of Escape from Immunity

Because all available evidence is consistent with the interpretation that tumor-induced suppressor T cells function, in the case of the Meth A and P815 tumors, to down-regulate an already ongoing concomitant immune response, it is logical to hypothesize that suppressor T cells are responsible for the escape of these and immunogenically similar tumors from immune destruction. This implies that if steps were taken to selectively eliminate suppressor T cells, the concomitant immune response would continue to generate effector T cells until enough of them were produced to cause regression of the tumor. If this were true, the selective elimination of suppressor T cells would be the major goal of immunotherapists. However, until it is shown that, in the absence of suppressor T cells, an immunogenic tumor can evoke the generation of a large enough number of effector T cells to cause tumor regression, it would be unwise to assume that successful

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

23

immunotherapy of all immunogenic tumors will automatically result from the selective elimination of suppressor T cells. The possibility remains that the immunogenicity of many tumors is too weak to evoke the generation of an adequate level of concomitant immunity, even in the absence of suppressor cells. It would seem safer to assume, until shown otherwise therefore, that successful immunotherapy of established tumors will also require the employment of agents that can directly augment the production of effector T cells. It needs to be brought to mind that the adoptive immunization assays employed in this laboratory and other laboratories are themselves successful models of adoptive immunotherapy in that they demonstrate that an infusion of an adequate number of tumor-sensitized, memory T cells can result, after a 6- to 8-day delay, in complete regression of an established rapidly growing tumor, provided the recipient is rendered incapable of generating suppressor T cells. This is an example of a situation, therefore, where adequate levels of immunity can be created experimentally at will to cause tumor regression. However, although these models make it clear that tumor-sensitized T cells can cause relatively large tumors to undergo complete and permanent regression, they cannot determine whether, in the absence of suppressor T cells, a tumor-bearing host can generate enough immunity of its own to cause complete tumor regression. In considering the design of experiments to test whether concomitant immunity does have the potential to develop to a high enough level to cause tumor regression in the absence of suppressor T cells, the possibility needs to be considered that some tumors may need to grow to a relatively large size before they represent enough immunogen to trigger an immune response. It has been shown in this laboratory (North and Kirstein, 1977), in this connection, that regardless of the number of tumor cells implanted and the duration of the latency period before the tumor emerges, concomitant immunity is not generated until the tumor grows to a certain size. It has been suggested that this type of evidence argues against a mechanism of immunosurveillance capable of destroying small numbers of tumor cells before they can form a palpable tumor (Lannon and McKhann, 1982). Indeed, the need for the Meth A and P815 tumors to grow progressively for about 6 days to a diameter of 4-5 mm before immune effector cells begin to be generated means that the immune response to these tumors has only 3 days to develop before suppressor T cells begin to be acquired. Consequently, just before suppression begins, the tumors are 8-10 mm in diameter and are growing rapidly. This represents a very large tumor mass for T cell-mediated immunity to reject. Experiments cur-

24

ROBERT J. NORTH

rently are in progress in this laboratory to determine whether in the absence of suppressor cells concomitant immunity can cause the regression of these immunogenic tumors. If so, then suppression of concomitant immunity represents the reason why these tumors continue to grow progressively. If not, then it is likely that the tumors grow progressively because the immunity they eventually evoke is of insufficient magnitude to destroy them. It is unlikely, on the other hand, that the Meth A and P815 tumors avoid immune destruction by way of the emergence of antigen-loss variants that cannot be recognized by an acquired immune effector mechanism. This can be stated because of the results of an ongoing study designed to measure changes in the ability of tumor cells harvested at progressive times of tumor growth to resist neutralization when implanted into mice previously immunized against the “parent” (To)tumor. Thus, mice were immunized by injecting them subcutaneously with an admixture of 2 x 106 tumor cells and 100 g of C. pamum, as described previously (Dye et al., 1981).Three weeks later, they were challenged with 106 tumor cells harvested from mice bearing a 3-, 6-, 9-, 12-, or 15-day intraperitoneal ascites tumor. The challenge tumors were initiated with lo6 tumor cells obtained from the same frozen stock as used for immunization. The challenge ascites contained over 108tumor cells on day 9 of tumor growth, and the mice were all dead by day 18. It was found that immunized mice were able to neutralize the growth of a lo6 tumor cell implant, regardless of whether the tumor cells were harvested at day 3 or day 15 of tumor growth (Fig. 4). This same result was obtained when immunized mice were challenged with tumor cells harvested from solid P815 and Meth A tumors at 3-day intervals over a 21-day period of tumor growth. Evidence will be discussed later which indicates that removal of suppressor T cells can result in enough immunity being generated to cause an appreciable degree of regression of the Meth A fibrosarcoma. That this need not be a general role, however, is indicated by results obtained from ongoing experiments with the SA1 sarcoma in this laboratory. The experiments thus far have failed to show the generation of suppressor T cells in the spleens of syngeneic A/J mice bearing a relatively large tumor. However, in spite of this, the level of concomitant immunity generated is insufficient to destroy the tumor. Failure to detect suppressor T cells in response to growth of the SA1 sarcoma is in keeping with a very slow decay of concomitant immunity to this tumor after day 9. However, because the decay of immunity does in fact occur, a slow acquisition of suppressor T cells cannot be ruled out at this time.

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

25

140

Immune recipients

120

;

100

v-

-a 0

0

80

f

3

60 Y

0 W

N

40

v)

20

0

I

1

5

10

I 15 DAYS

I

20

I

25 DAYS

FIG.4. Evidence that tumor cells harvested from a 15-day tumor are no more resistant to immunologically mediated destruction than tumor cells from a 3-day tumor. Shown are growth curves of tumor cells harvested from a 3-, 6-, 9-, 12-, and 15-day P815 mastocytoma growing as an intraperitoneal ascites (numbers on individual graphs) when implanted into normal recipients or recipients immunized against the original (To) tumor.

VIII. Suppression of Antitumor Immunity as an Example of Transplantation Tolerance

It may be stated with confidence that immunity generated in response to immunogenic tumors is T cell mediated rather than antibody mediated, and that the only generally accepted positive test for tumor immunogenicity at this time is the ability of a tumor to immunize a syngeneic host against growth of a subsequent implant of cells of that tumor. In fact, it appears to be the general experience that immunogenic tumors are difficult to raise antibodies against, a characteristic that likens tumor-associated antigens to minor histocompatibility antigens. Moreover, because tumor-associated antigens and minor histocompatibility antigens both are defined almost exclusively by transplantation rejection assays, they may be compared not only in terms of the immune responses they engender, but in terms of the types of suppressor cells they induce. The state of immunosuppres-

26

ROBERT J. NORTH

sion induced by histocompatibility antigens is referred to as transplantation tolerance, as defined as a long-lived state of T cell-mediated hyporesponsiveness to specific antigen. It is possible to argue, therefore, that progressive growth of a syngeneic tumor also should eventually supply enough antigen to induce a state of transplantation tolerance. To determine whether tumor-induced, T cell-mediated immunosuppression does, in fact, represent transplantation tolerance, it simply was necessary to determine whether a state of T cell-mediated suppression was retained long after excising a tumor that had grown large enough to have induced dominant suppression. The Meth A tumor was employed in this study because of the need to have a tumor that can grow to about 1.5 cm in diameter without metastasizing to distant sites. The experiments showed (Bursuker and North, 1984) that excision of a 16-day (1.2- to 1.5-mm) Meth A tumor left the host with a state of long-lived suppression to tumor-associated antigens, as evidenced by an inability, over many weeks, to mount a concomitant immune response to a second Meth A tumor. It was shown, in addition, that this long-lived state of suppression was associated with the persistence of a population of Ly 1+,2- T cells capable of passively transferring suppression to adoptively immunized TXB recipients, according to the standard suppressor assay. Moreover, because this state of T cell-mediated unresponsiveness was specific and lasted for more than a month after complete removal of the tumor, it was concluded that it represented a state of transplantation tolerance to tumor-specific, transplantation rejection antigens. Thus, it is suggested here that there is nothing peculiar or novel about tumor-induced immunosuppression. However, the fact that it can remain after removal of a solid tumor would need to be taken into account when considering whether it will be possible to cure animals of systemic diseases by administering immunotherapy after surgical debulking of a large primary tumor. IX. lmmunotherapeuticSignificance of the Generation and Subsequent Decay of Concomitant Immunity

Immunotherapy refers to a form of therapy that attempts to cause regression of an established tumor by augmenting the level of antitumor immunity. Obviously, because the goal of specific immunotherapy is to augment the level of specific antitumor immunity, it is not surprising that attempts to cause tumor regression by this means have relied almost exclusively on the use of immunoadjuvants, such as live BCG or Formalin-killed C. pamyurn, with a proved ability to augment immune responses in general. The foregoing discussion about con-

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

27

comitant immunity suggests, however, that there may be another way to augment the level of antitumor immunity, namely, by preventing the generation of suppressor T cells that function to down-regulate concomitant immunity. Indeed, because immunotherapy, as distinct from immunoprophylaxis, is aimed at the elimination of already established tumors, it follows that it involves attempts to augment an already ongoing concomitant immune response. Therefore, any rational approach to the design of immunotherapeutic modalities must take into consideration the kinetics of the generation and decay of concomitant immunity, since, according to the foregoing discussion, any attempt to cause the regression of an immunogenic tumor by immunotherapy represents an attempt to augment an already ongoing concomitant immune response that may be undergoing negative regulation. To ignore concomitant immunity would be to design immunotherapy by empirical procedure, a practice that probably was responsible for the disappointing results of immunotherapy trials in animals and humans over the past 15 years. It is the purpose of this section to discuss the immunotherapeutic effects of four different biological response modifiers. Two of these, cyclophosphamide and y-radiation, are generally considered to be immunosuppressive. The other two, bacterial endotoxin and Formalinkilled C. paruurn, are proved immunostimulators. It will be suggested that successful immunotherapy of established tumors is a realistic goal, provided the immunotherapeutic modality is designed on the basis of an adequate understanding of the regulation of the underlying antitumor immune response.

A. CYCLOPHOSPHAMIDE AS AN IMMUNOTHERAPEUTIC AGENT Although it is the cytotoxic action of the alkylating agent, cyclophosphamide, that is primarily responsible for its ability to cause the regression of most tumors, it has been shown repeatedly over a number of years that, in some cases at least, the antitumor effect of the drug partly depends on the acquisition and expression of host immunity. For example, a better therapeutic result from cyclophosphamide treatment has been described in animals that are preimmunized against their tumors (Moore and Williams, 1973; Chassoux et al., 1978). Conversely, the therapeutic effect of cyclophosphamide has been shown to be greatly diminished in animals that are immunodepressed by treatment with antithymocyte serum (Radov et al., 1976; Greenberg et al., 1980) or X-irradiation (Lubet and Carlson, 1978). There is mounting evidence, moreover, that cyclophosphamide treatment can aug-

28

ROBERT J. NORTH

ment the level of antitumor immunity by having a direct effect on immunoregulation. Indeed, it has been known for some time (reviewed by Goto et al., 1981; Turk and Parker, 1982) that treatment with cyclophosphamide can augment cell-mediated immune responses in general, particularly when the drug is given shortly before antigen. Pretreatment with cyclophosphamide has been shown (Glaser, 1979) to augment production of cytolytic T cells in response to injection of an SV40-induced tumor in syngeneic mice. It is significant that the administration of the drug after, instead of before, giving tumor cells failed to augment the generation of effector T cells. A similar augmenting effect of cyclophosphamide was described for the cytolytic T cell response to allogeneic cells (Rollinghoff et al., 1977). The additional finding in both studies that the immunoaugmenting effect of cyclophosphamide pretreatment could be negated by an infusion of T cells from normal donor mice was taken as evidence that the immunoaugmenting effect of the drug is based on its capacity to eliminate suppressor T cells. Presumably, the suppressor T cells that were eliminated in these studies must have been suppressor precursor cells because mature suppressor cells could not have been present before antigen was given. In any case, it needs to be pointed out that examples of enhanced immunity caused by pretreatment with cyclophosphamide represent examples of enhanced immunoprophylaxis rather than of immunotherapy. Convincing examples of the immunotherapeutic action of cyclophosphamide do, however, exist. One such example is the complete regression of the immunogenic MOPC-315 plasmacytoma of BALB/c mice which occurs after cyclophosphamide is given late in tumor growth (Hengst et at., 1980,1981; Mokyr et al., 1982). It was shown with this model that, whereas cyclophosphamide treatment fails to cause regression of a small MOPC-315 tumor, the same dose of the drug can cause complete regression of tumors as large as 20 mm in diameter. It further was shown that complete regression of large tumors could be prevented by infusion of antithymocyte serum, thereby providing convincing evidence that regression was immunologically mediated. Additional evidence that regression was immunologically mediated came from experiments which showed that although a small MOPC-315 tumor is refractory to treatment with cyclophosphamide, it undergoes complete regression if it is growing in an animal bearing a susceptible larger tumor on the contralateral side. In interpreting the reason for the susceptibility of large tumors, the authors suggest that tumor regression was not mediated by concomitant immunity, because the host did not display immunity to

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

29

growth of a challenge implant at the time of treatment. However, in view of the large size of the tumor at the time of treatment, it seems likely that concomitant immunity would already have decayed, and that cyclophosphamide functioned to cause a regeneration of the immunity by releasing it from suppression. This would explain why the onset of regression was delayed and why regression was associated with the acquisition by the host of a capacity to express immunity to a challenge implant. The well-documented knowledge that cyclophosphamide can destroy suppressor T cells provided the reason for predicting in this laboratory that treatment of a tumor bearer with the drug would enable passively transferred immune T cells to cause regression of the tumor. Experiments with the Meth A fibrosarcoma growing subcutaneously (North, 1982) showed that whereas injection of 100 mgkg of cyclophosphamide alone caused only partial regression of an established Meth A tumor and infusion of immune cells alone had no effect on tumor growth at all, infusion of immune celIs 1 hr after injecting cyclophosphamide caused complete and permanent regression of tumors in all mice. It was shown, in addition, that tumor regression caused by this combination therapy with cyclophosphamide and immune cells could be completely inhibited by intravenous infusion of T cells from donors bearing an established tumor, but not by T cells from normal donors. There can be little doubt, therefore, that cy-, clophosphamide facilitates adoptive immunotherapy of the Meth A fibrosarcoma by eliminating tumor-induced suppressor T cells that function to inhibit the expression of immunity passively transferred with immune T cells. A similar effect from combination chemotherapy and adoptive immunotherapy was reported earlier by Greenberg et al. (1981) for a disseminated murine leukemia. However, because in this case cyclophosphamide given alone greatly decreased the number of tumor cells, it was not possible to know the size of the tumor burden that passively transferred tumor-sensitized T cells were responsible for eliminating. It stands to reason in this regard that the best tumors with which to study the immunoaugmenting effects of cyclophosphamide are those that are totally resistant to the direct cytotoxic action of the drug. Such a tumor is the L5178Y lymphoma syngeneic in DBN2 mice, which is currently under study in this laboratory. Results obtained thus far with this tumor leave no doubt that the ability of a single 100 mgkg injection of cyclophosphamide to cause complete tumor regression results from the ability of the drug to greatly augment concomitant immunity. Among other things, the drug has no therapeutic effect whatever against an L5178Y tumor growing

30

ROBERT J. NORTH

in immunoincompetent mice. Therefore, the dramatic antitumor effect can be totally attributable to antitumor immunity.

B. IONIZING RADIATION

’

AS AN

IMMUNOTHERAPEUTIC AGENT

As discussed in a foregoing section, realization that the magnitude of an immune response is determined by the relative influences of helper T cells and suppressor T cells provided the reason for postulating that it should be possible to augment an already ongoing immune response by selectively removing suppressor lymphocytes. In support of this postulate, evidence was discussed previously which shows that properly timed treatment with cyclophosphamide, by removing suppressor T cells, can augment an immune response sufficiently to cause rejection of an already established immunogenic tumor. There is good reason to believe that whole-body ionizing radiation can augment antitumor immunity in the same way, perhaps in an even more convincing manner. It has been known for a number of years (reviewed by Anderson and Warner, 1976) that exposure to sublethal doses of X-irradiation can result in enhanced generation of humoral or cell-mediated immunity, providing exposure to irradiation is properly timed in relation to antigen. The fact that enhancement of an immune response can occur in the face of a substantial reduction in the cellularity of lymphoid tissue is in keepigg with the idea that enhancement can result from the removal of lymphocytes that function normally to inhibit the generation of immunity. Indeed, it is generally believed that suppressor T cells and their precursors are particularly sensitive to ionizing radiation, a situation that has been taken advantage of (sometimes unknowingly) to demonstrate successful adoptive immunotherapy of established tumor (Fernandez-Cruzet al., 1979; North et al., 1982; Eberlein et al., 1982; North, 198413). That sublethal X-radiation of a tumor-bearing recipient can eliminate suppressor T cells that function to inhibit the ability of intravenously infused, sensitized T cells to cause the regression of the recipients’ tumors is indicated by published results (Fernandez-Cruz et al., 1979) which show that sublethal (400 rad) X-radiation of recipient rats bearing an established Moloney virus-induced sarcoma enables passively transferred effector T cells generated in oitm to cause tumor regression. In this model of adoptive immunotherapy, effector T cells were generated from a 5day mixed lymphocyte-tumor cell culture. A more recent and detailed publication by North (1984b)provides additional evidence that the immunofacilitating action of ionizing radiation is based on its ability to eliminate suppressor T cells. This publication shows that

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

31

whereas intravenous infusion of immune spleen cells from immunized donor mice had no effect on the growth of an established tumor in immunocompetent recipients, infusion of the same number of immune cells caused complete and permanent regression of the samesized tumor in recipients that had received 500 rad of y-radiation 1hr before receiving spleen cells. It was shown, in addition, that tumor regression caused by combination therapy with y-radiation and immune spleen cells could be completely inhibited by passive transfer of splenic T cells from donors bearing a large tumor, but not from normal donors. This represents convincing evidence, therefore, that y-radiation facilitates the expression of adoptive immunity against an established tumor by eliminating the capacity of the tumor bearer to generate suppressor T cells. The study further revealed that tumorinduced suppressor T cells in the tumor-bearing donors were eliminated by the same dose of y-radiation that needed to be given to tumor-bearing recipients to facilitate regression of their tumors by passively transferred immune T cells. This ability of sublethal y-radiation to facilitate the expression of adoptive immunity was demonstrated with five different immunogenic murine tumors. It is apparent, therefore, that the success of other models of adoptive immunotherapy is the result of the ability of y-irradiation to remove suppressor cells from the tumor-bearing recipients. It may well be asked, in view of evidence that y-radiation eliminates suppressor T cells and their precursors, why there is any need to passively transfer immune T cells to cause tumor regression. According to a hypothesis developed in a preceding section, concomitant immunity might be expected to develop sufficiently in magnitude to cause tumor regression if suppressor T cells could be preferentially eliminated. The obvious reason why sublethal (400-500 rad) irradiation fails to result routinely in tumor regression is that it also destroys effector T cells. It is apparent, however, that X-radiation can preferentially eliminate suppressor T cells, provided it is given at certain stages of tumor growth. Indeed, it was revealed in an interesting paper by Hellstrom et d.(1978) that sublethal (400 rad), whole-body Xradiation of mice bearing established immunogenic fibrosarcomas resulted, after several days’ delay, in complete regression of the tumors in some mice and partial regression in the remainder. The role of antitumor immunity in this example of radiation-induced tumor regression was indicated by the additional findings that the tumor needed to be above a certain size to be susceptible. Moreover, because the therapeutic effect of X-radiation-could be prevented by infusing the X-irradiated tuAor bearers with T cells from normal mice, it

32

ROBERT J. NORTH

was suggested that radiation-induced tumor regression depends on the ability of X-radiation to eliminate suppressor T cells, thereby enabling an antitumor immune response to develop. It was further suggested that the degree of radiation-induced regression achieved depended on the level of immunity generated, and therefore on the immunogenicity of the tumor. Evidence in support of this explanation of radiation-induced tumor regression currently is being generated in this laboratory from studies with the SA1 sarcoma and Meth A fibrosarcoma (North, 1985). We have confirmed the results of Hellstrom et al. (1978) by showing that it is possible to cause complete regression of the SA1 tumor by exposing the hosts to 500 rad of y-radiation on day 6 of tumor growth. We also have revealed that y-radiation-induced tumor regression does not begin until after an appreciable delay, and that regression fails to occur if the tumor is growing in T cell-deficient mice (TXB mice) that are incapable of generating concomitant immunity. This later finding represents convincing indirect evidence for the view that radiation-induced regression is based on the development of an augmented immune response. Direct evidence for this view was obtained from an experiment that compared y-irradiated and normal tumor bearers in terms of their capacity to generate T cells capable of passively transferring concomitant immunity to y-irradiated recipients, according to the passive transfer assay described in a preceding section. It was found (North, 1985)that whereas the production of effector T cells in tumor-bearing control mice began on day 6 of tumor growth, peaked on day 9, and rapidly declined thereafter, the production of effector T cells in y-irradiated tumor bearers was sustained at peak level for at least an additional 10 days. In other words, whole-body, y-radiation resulted in a prolonged concomitant antitumor immune response that may also have been of greater than normal magnitude. Additional experiments that were designed to determine the effect of y-radiation on the generation of suppressor T cells showed that whereas control tumor bearers generated increasing numbers of suppressor T cells after day 9 of tumor growth, y-irradiated tumor bearers failed to show the presence of these cells at any time up to day 21 (North, 1985). There is little doubt from the results obtained with the in vivo suppressor assay, therefore, that correctly timed sublethal y-radiation of mice bearing a Meth A fibrosarcoma can result, by selectively eliminating suppressor T cell precursors, in a protracted concomitant antitumor immune response of s u e i e n t magnitude to cause complete tumor regression. Presumably, by day 6 of tumor growth, the inductive events needed for the generation of effector T cells become in-

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

33

sensitive to 500 rad of y-radiation. In this connection, there is evidence from studies of the antiallograft response (Kanawaga et al., 1983) that cytolytic T cell precursors, in contrast to helper T cells, are insensitive to 500 rad of y-radiation. It is possible, therefore, that by day 6 of tumor growth, radiosensitive helper T cells already have served their purpose in the inductive phase of the immune response, leaving committed radioresistant cytolytic T cell precursors to mature into functional effector cells. Needless to say, the foregoing discussion indicates that care should be taken in interpreting the results of adoptive immunization experiments that employ irradiated recipients. Depending on the timing of irradiation, the possibility exists that the immunity expressed by recipients of immune T cells partly consists of immunity generated by the recipients themselves.

C. ENDOTOXIN AS AN IMMUNOTHERAPEUTIC AGENT It was shown at the turn of the century by Coley (1909) that parenteral injection of filtrates of cultures of certain bacteria could result in regression of inoperable human sarcomas. These demonstrations of the therapeutic action of “bacterial toxins” are considered the first examples of tumor immunotherapy. It was not until sometime later, however, that similar experiments were performed with experimental animals and that examples of the therapeutic action of bacterial toxins began to accumulate. In the case of animal tumors, it became clear that it was the culture filtrates of gram-negative rather than grampositive bacteria that possessed therapeutic activity, and that the therapeutic factor common to gram-negative organisms was the lipopolysaccharide component (endotoxin) of their cell walls. It is generally not appreciated, in this regard, that “Coley’s toxins” were produced by gram-positive bacteria. Be this as it may, the genera1 findings about the antitumor properties of bacterial endotoxin were reviewed some years ago by Shear (1944), who pointed out that (1) endotoxin almost invariably causes necrosis in the center of tumors, but rarely causes their complete regression, (2) intradermal tumors are the most susceptible, and (3) the tumors need to be above a certain size in order to be susceptible. A role for antitumor immunity in endotoxin-induced regression was revealed first by Parr et al. (1973), who demonstrated that endotoxininduced regression of established tumors, in contrast to core-confined necrosis, fails to take place in animals that had been immunodepressed by treatment with X-radiation or antithymocyte serum. These authors also demonstrated that the tumors need to be above a certain

34

ROBERT J. NORTH

size to be susceptible and that tumors growing in the intradermal site were the most susceptible. Additional evidence for an immunological basis for endotoxin-induced tumor regression was supplied in publications by Berendt et al. (1978a), who showed that only immunogenic tumors are susceptible to endotoxin therapy and that regression fails to occur in TXB mice that are incapable of generating concomitant immunity. These authors also revealed (Berendt et aZ., 1978b) that immunogenic tumors do not develop susceptibility to endotoxin-induced regression until their hosts develop concomitant immunity to growth of an implant of tumor cells. This and other evidence provided the reason for suggesting that small tumors are refractory to endotoxin therapy, because they are not large enough to have evoked a concomitant immune response. However, this type of evidence for a role for concomitant immunity in endotoxin-induced regression is correlative rather than causal. Therefore, a study recently completed in this laboratory was designed to obtain direct causal evidence that endotoxin-induced tumor regression is based on the preceding generation of T cell-mediated concomitant immunity. It was anticipated, on the basis of evidence discussed in a preceding section, that endotoxin-induced tumor regression would be found to depend on the generation of Ly 1-,2+ effector T cells that passively transfer concomitant immunity. Thus, the methodology employed to demonstrate that concomitant immunity can be passively transferred with Ly 1-,2+ T cells and expressed against an established tumor (North and Bursuker, 1984) provided the means to determine directly whether endotoxin-induced tumor regression is dependent on concomitant immunity. The SA1 sarcoma was chosen for study because of its sensitivity to endotoxin therapy (Berendt et al., 1978a) and because of the demonstration (North, 1984c) that concomitant immunity to this tumor, like that generated in response to the Meth A fibrosarcoma and P815 mastocytoma, can be passively transferred with Ly 1-,2+ T cells and expressed against an established tumor in y-irradiated recipients. It was reasoned that whereas a 9-day tumor growing intradermally in a TXB host would be refractory to intravenous injection of endotoxin, because of the absence of concomitant immunity, the tumor would become sensitive to endotoxin after infusing the recipient intravenously with T cells from an immunocompetent, concomitantly immune donor bearing a 9-day tumor, that is, by infusing the TXB recipients with T cells from a donor with a susceptible SA1 tumor. This prediction proved to be correct (North, 1984d) in that injection of endotoxin followed 48 hr later by infusion of immune spleen cells from a 9-day tumor bearer resulted in com-

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

35

plete regression of large SA1 tumors in TXB recipients. In contrast, immune cells alone, or endotoxin alone, failed to cause any therapeutic effect. Because it was shown, in addition, that this priming effect of donor spleen cells was eliminated by treatment with'anti-Thy 1.2 antibody and complement, it was concluded that endotoxin-induced tumor regression is dependent on T cell-mediated immunity. It was surprising to find in addition, however, that the ability of donor T cells to prime a tumor in a TXB recipient for endotoxin-induced regression was completely eliminated by treating the cells with anti-Ly 1 antibody and complement, but not by treating them with anti-Ly 2 antibody and complement. Therefore, the T cells that are responsible, on passive transfer, for priming a tumor in a TXB recipient for the therapeutic action of endotoxin are different from the T cells that cause regression of a smaller tumor in the same type of recipient without the aid of endotoxin. The T cells that prime for susceptibility to endotoxin are of the helper or DTH phenotype, whereas those that express concomitant immunity without the aid of endotoxin are of the cytolytic phenotype. It surely is significant in this regard that in order for Ly 1+,2- T cells to prime a recipient tumor for endotoxin-induced regression, they needed to be infused into TXB tumor-bearing recipients 24-48 hr before endotoxin is given (North, 1984d). This suggests the distinct possibility that priming a tumor for endotoxin-induced regression involves the prior expression of a DTH reaction in the tumor bed. This suggests, in turn, that activated macrophages, rather than sensitized T cells, are the ultimate expressors of tumor cell destruction. It is predicted on the basis of results obtained thus far, therefore, that further analysis of endotoxin-induced regression of immunogenic tumors will provide evidence that lymphokine-primed, endotoxin-activated macrophages have a tumoricidal role to play in the in uivo setting. The possibility that a toxic molecule produced locally, like tumor-necrosis factor (Green et ul., 1976),ultimately is involved in the destruction of tumor tissue deserves serious consideration. However, even though it is likely that the ultimate effector mechanism of endotoxin-induced tumor regression is a nonspecific one mediated by macrophages, there should be no doubt, on the basis of the unequivocal evidence presented above, that tumor-sensitized T cells are required. Indeed, it has been suggested, on the basis of the need for tumor-sensitized T cells, that endotoxin-induced tumor regression is an example of immunofacilitation in that it facilitates the expression of an already acquired mechanism of antitumor immunity (North et ul., 1982). It should not be surprising, therefore, that the therapeutic response to endotoxin is determined by the level of concomitant anti-

36

ROBERT J. NORTH

tumor immunity that prevails in the host at the time of treatment. This means that a tumor becomes refractory to endotoxin treatment after concomitant immunity undergoes decay under the influence of suppressor T cells (North, 1984d). Presumably, then, successful immunotherapy with bacterial endotoxin depends on giving it before dominant numbers of suppressor T cells are generated.

CAUSED BY INTRALESIONAL ADJUVANTS D. TUMORREGRESSION It would be safe to state that the best documented and most convincing examples of successful immunotherapy of an established tumor is regression of the line 10 hepatocarcinoma of strain 2 guinea pigs caused by intralesional injection of an immunoadjuvant, such as BCG (Zbar et ul., 1971). The findings obtained with this model of immunotherapy have been the subject of a detailed review by Bast et aZ. (1976), who point out that there is good reason to believe that the antitumor action of intralesional immunoadjuvants is based on the ability of these agents to augment the level of antitumor immunity. This is evidenced by the findings that BCG- or C. pumum-induced tumor regression is associated with the acquisition by the host of immunity to growth of a tumor implant, and of lymphocytes capable of passively transferring this immunity to a normal recipient. However, in order for successful immunotherapy to occur, the tumor needs to be below a certain critical size (about 1cm in diameter) and to be growing intradermally. This last-mentioned requirement is not generally appreciated, but is in keeping with the knowledge (see Bartlett et aZ., 1976) that tumors appear most immunogenic when growing in the intradermal site. It is perhaps surprising that in spite of all that has been written about the therapeutic effects of intralesionally injected immunoadjuvants, it was not until fairly recently that it was shown that the therapeutic action of these agents is based on their ability to augment the level of antitumor immunity. Direct evidence that C. pamum injected intralesionally causes an augmentation in the generation of antitumor effector cells recently was supplied by experiments with the P815 mastocytoma (Dye et al., 1981; Mills et al., 1981).These experiments showed, in agreement with the general experience with murine tumors, that in order for intralesional C. pamum to cause regression of the P815 mastocytoma, it was necessary to inject C. pamum at the same time the tumor was initiated; that is, it was necessary to inject it as an admixture with tumor cells. Partial regression was obtained,

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

37

however, by injecting C. parvurn intralesionally on day 3 of tumor growth, but by day 6 the tumor was completely refractory. It was shown further that the onset of C. pawum-induced tumor regression did not begin until the tumor had grown progressively for 9 days. Regression was accompanied, moreover, by the acquisition of immunity to growth of a tumor challenge implant and T cells capable of passively transferring this immunity to a normal recipient. However, direct evidence that C. parvum causes the generation of a larger than normal number of effector cells had to await the results of an experiment that measured cytolytic T cell production in the lymph node draining the tumor. This experiment showed that whereas growth of an untreated tumor resulted in a cytol9tic T cell response of low magnitude (10-15% specific 51Crrelease at a 50: 1effector-to-target ratio), growth of the C. paruurn-containing tumor resulted in a cytolytic response of much larger magnitude (40-50% specific W r release at a 50 : 1 effector-to-target ratio) that peaked immediately before the onset of tumor regression. Indeed, a study still in progress in this laboratory (Johnson et al., 1986) has shown by limiting dilution analysis that the node draining a treated tumor generates 10 times more cytolytic T cells than the node draining an untreated tumor in control mice. It can be suggested, moreover, that because the cytolytic T cell response to an untreated tumor is a measure of concomitant immunity, the antitumor action of an intralesional adjuvant depends on its ability to augment concomitant antitumor immunity. Therefore, interpreting the results of immunotherapy with an intralesional immunoadjuvant requires a knowledge of the generation and decay of concomitant immunity. In this connection, it was shown convincingly in a preceding section that concomitant immunity to the P815 mastocytoma undergoes T cell-mediated immunosuppression after day 9 of tumor growth. Therefore, a negative regulatory influence of suppressor T cells can explain why a 9-day tumor is totally refractory to C. pawum therapy. However, suppressor T cells cannot be invoked to explain total refractoriness of a 6-day tumor because suppressor cells do not exist at day 6. However, there is evidence (Bast et al., 1976) that the immunoaugmenting action of an intralesional adjuvant almost certainly is based on the immune response to the adjuvant itself, which in the case of C. pawum takes 6-8 days to develop (Tuttle and North, 1975). Therefore, although suppressor T cells would not be present in the host treated with C. parvurn on day 6 of growth of the P815 mastocytoma, suppressor cells most certainly would be present by the time the response to C. pawurn developed sufficiently in magnitude to cause augmentation of antitumor immunity. The reason for drawing atten-

38

ROBERT J. NORTH

tion to results obtained with intralesional C. parvum is to justify suggesting that tumor-induced, T cell-mediated down-regulation of concomitant immunity needs to be considered when attempting to explain failures to obtain tumor regression by active immunotherapy with intralesional immunoadjuvants. It is apparent, at this stage of our knowledge, that the size beyond which a tumor becomes refractory to therapy with immunoadjuvants depends, at least in part, on the time of onset of generation of T cell-mediated immunosuppression. However, because the larger the tumor the larger the number of effector T cells required to cause its regression, the immune response may not be capable of supplying enough effector cells to cause regression, even under the influence of an immunoadjuvant in the absence of suppression . X. Conclusion

A framework of evidence has been presented which supports the hypothesis that down-regulation of antitumor immunity is the reason for the escape of at least some immunogenic tumors from immune destruction. Circumstantial evidence for this hypothesis consists of the demonstration that after a certain stage of tumor growth, concomitant antitumor immunity and the Ly 1-,2+T cells with the potential to express this immunity are progressively lost as the host progressively acquires Ly 1-,2+suppressor T cells. Direct evidence for the hypothesis consists of the finding that infusion of suppressor spleen cells from donors with a large tumor can prevent recipient mice from generating a concomitant immune response. Additional direct evidence is being supplied by an ongoing study which shows that complete or partial regression of the Meth A fibrosarcoma that results from appropriately timed exposure to sublethal, whole-body y-radiation is associated with a prolonged generation of effector T cells and an absence of suppressor T cells (North,1985). However, failure of some tumors to undergo complete regression in response to y-radiation in spite of the apparent elimination of suppressor T cells may mean that the negative regulatory influence of suppressor T cells need not explain all cases of immune escape. On the contrary, it is suggested that down-regulation of antitumor immunity by suppressor T cells can explain the escape only of those tumors that are immunogenic enough to evoke the generation of enough effector T cells to cause tumor regression in the absence of suppressor T cells. This implies that the immunity generated to some tumors is too little too late to cause regression, even in

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

39

the absence of the negative regulatory influence of suppressor cells. Therefore, successful immunotherapy of some established tumors, besides depending on the employment of agents capable of eliminating suppressor T cells, will also depend on the employment of agents capable of directly augmenting the generation of effector T cells. Be this as it may, all of the foregoing discussion would be purely academic if it were universally true that transplantable tumors avoid destruction by the immune system by generating antigen-loss variants that can multiply unrestrictedly in the presence of immune effector mechanisms generated against the original tumor. However, this cannot explain the escape of the P815 mastocytoma and Meth A fibrosarcoma from concomitant immunity, because it was shown in a foregoing section that cells harvested from late tumors were just as susceptible to immune destruction as cells harvested from early tumors when implanted into mice immunized to the “parent” tumor. Moreover, if the generation of antigen-loss variants were the universal mechanism of escape, the transplantable tumors employed routinely in this laboratory should have lost their immunogenicity many years ago. Indeed, the idea of selection for and outgrowth of nonimmunogenic antigen-loss variants relatively early during a single in vivo passage (Uyttenhove et al., 1983) is not in keeping with what appears to be the general experience that the immunogenicity of most transplantable tumors is surprisingly stable over many in vivo passages, even though the tumors may be composed of cells with different antigens (Rogers, 1984). Needless to say, the presence in a tumor of nonimmunogenic variant cells with a potential to survive an antitumor immune response also is not in keeping with the demonstration in this and other laboratories (reviewed by North, 1984a)that relatively large tumors can be caused to undergo complete and apparently permanent regression by adoptive immunization with sensitized T cells, provided, of course, the recipient tumor bearers are incapable of generating suppressor T cells. Most of the evidence (Weissman, 1973; Galli et al., 1982) argues against the possibility that in these models of immunotherapy small numbers of nonimmunogenic tumor cells in a solid or ascites tumor can be destroyed nonspecifically by a bystander effect. It is obviously important, therefore, to devote more experimental attention to investigating whether the survival and outgrowth of nonimmunogenic antigen-loss tumor cells is the reason for progressive growth of immunogenic tumors in general. However, results obtained with the exclusive use of in vitro assays that measure nonimmunogenic variants purely in terms of resistance to lysis by cytolytic T cell

40

ROBERT J, NORTH

clones cannot be considered conclusive. It is apparent that resistance to lysis can result from changes other than the loss of antigen (Greenberg et al.. 1981b). A hypothesis of tumor escape based on down-regulation of antitumor immunity and one based on the survival of nonimmunogenic variant tumor cells are not mutually exclusive. However, the former hypothesis currently is considered a more attractive one to test in this laboratory because it serves to explain the successes as well as the failures of active immunotherapy with intralesional immunoadjuvants and of adoptive immunotherapy with sensitized T cells.

ACKNOWLEDGMENTS This work was supported by Grants CA-16642 and CA-27794 from the National Cancer Institute and a grant from R. J. Reynolds, Inc.

REFERENCES

Anderson, R. E., and Warner, N. L. (1976).Ado. Immunol. 24,215-335. Asano, Y., and Hodes, R. J. (1983).]. Exp. Med. 158, 1178-1190. Badett, G., Kreider, J. W., and Purnell, D. S. (1976)./. Natl. Cancer Inst. 57, 12971303. Bashford, E. F., Murray, J. A., and Haaland, M. (1908). Rep. Imperial Cancer Res. Found. 3,262-268. Bast, R. C., Bast, B. S., and Rapp, H. J. (1976). Ann. N.Y. Acad. Sci. 277,60-92. Benacerraf, B., Greene, M. I., Sy, M.-S., and Dorf, M. E. (1982). Ann. N.Y. Acad. Sci. 392,300-309. Berendt, M. J., and North, R. J . (1980).]. E x p . Med. 151,69-80. Berendt, M. J., North, R. J., and Kirstein, D. P. (1978a).J. Exp. Med. 148, 1550-1559. Berendt, M. J., North, R. J., and Kirstein, D. P. (1978b)./. Exp. Med. 148, 1560-1569. Biddison, W. E., Palmer, J. C., Alexander, M. A., Cowan, E. P., and Manson, L. A. (1977).J. Immunol. 118,2242-2249. Bonventre, P. F., Nickol, A. D., Ball, E. J., Michael, J. G., and Bubel, H. C. (1982).J . Reticuloendothel. SOC. 32,25-34. Bosslet, K., and Schinmacher, V. (1981).J . E x p . Med. 154,557-562. Boyse, E. A., Stockert, E., and Old, L. J. (1967). Proc. Natl. Acad. Sci. U S A . 58,954959. Bursuker, I., and North, R. J. (1984).J. Exp. Med. 159, 1312-1321. Bursuker, I., and North, R. J. (1985). Cancer Immunol. Immunother., in press. Burton, R., Thompson, J., and Warner, N. L. (19'75).]. lmmunol. Methods 8,133-139. Chassoux, D. M., Gotch, F. M., and Maclennan, I. C. M. (1978). Br. J. Cancer 38,211218. Coley, W. B. (1909-1910). Proc. R. SOC. Med. (Surg. Sect.) 3, 1-48. Deodar, S. D., and Crile, G. (1969). Cancer Res. 29,776-779. DiGiacomo, A., and North, R. J. (1986). In preparation. Drebin, J. A., Waltenbaugh, G., Schatten, S., Benacerraf, B., and Greene, M. I. (1983).J. Zmmunol. 130,506-509.

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

41

Duprez, V., Hamilton, B., and Burakoff, S. J. (1982).J . Exp. Med. 156,844-859. Dye, E. S., and North, R. J. (1980).J. Immunol. 125, 1650-1657. Dye, E. S., and North, R. J. (1981).J.Exp. Med. 154, 1033-1042. Dye, E. S., and North, R. J. (1984a).J . Leukocyte Biol. 36,27-38. Dye, E. S., and North, R. J. (1984b). Transplantation 37,600-606. Dye, E. S., North, R. J., and Mills, C. D. (1981).J.E r p . Med. 154,609-620. Eberlein, T. J., Rosenstein, M., and Rosenberg, S. A. (1982).J.E r p . Med. 156,385-397. Eccles, S . A., and Alexander, P. (1974).Nature (London)250,667-669. Ehrlich, P. (1906).Arb. Inst. E x p . Ther. FrunkfurtMain 1, 65-80. Evans, R. (1973).Br. J . Cancer 28 (Suppl. l), 19-25. Fauve, R. M., Hevin, B., Jacobs, H., Gaillard, J. A., and Jacobs, F. (1974).Proc. Nutl. Acad. Sci. U . S A . 71,4052-4056. Fernandez-Cruz, E., Halliburton, B., and Feldman, J. D. (1979).]. Immunol. 123,17721781. Foley, E. J. (1953).Cancer Res. 13,35-37. Forni, G., Giovarelli, M., Lanfrancone, L., and Varesio, L. (1982). I n t . J . Cancer 29, 695-698. Frost, P., Prete, P., and Kerbel, R. (1982).Int. J . Cancer 30,211-217. Galli, S . J., Bast, R. C., Bast, B. S., Isomura, T., Zbar, B., Rapp, H. J., and Dvorak, H. F. (1982).J . Immunol. 129,890-899. Gershon, R. K. (1980).J. Allergy Clin. Immunol. 66, 18-25. Gershon, R. K., and Kondo, K. (1971).J.Natl. Cancer Znst. 46, 1169-1175. Gillis, S., Union, A. N., Baker, P. E., and Smith, K. A. (1979).J . E x p . Med. 149, 14601476. Glaser, M. (1979).J . E z p . Med. 149, 774-779. Gorelik, E. (1983).Adu. Cancer Res. 39, 71-120. Goto, M., Mitsuoka, A., Sugiyama, M., and Kitano, M. (1981).J.E x p . Med. 154,204-209. Green, S., Dobrjansky, A., Carswell, E. A., Kassel, R. L., Old, L. J., Fiore, N., and Schwartz, M. K. (1976). Proc. Natl. Acad. Sci. U S A . 73,381-385. Greene, M. I. (1980).Contemp. Top. Immunobiol. 11, 81-116. Greenberg, P. D., Cheever, M. A., and Fefer, A. (1980).Cancer Res. 40,4428-4432. Greenberg, P. D., Cheever, M. A., and Fefer, A. (1981a).J . E x p . Med. 154,952-963. Greenberg, P. D., Cheever, M. A., and Fefer, A. (1981b).j . Immunol. 126,200-203. Haskill, J. S., Proctor, J. W., and Yamamara, Y. (1975).]. Natl. Cancer Znst. 54,387-392. Hawrylko, E. (1982).Cell. Immunol. 66, 121-138. Hawrylko, E., Mele, C. A., and Stutman, 0. (1982). Cell. Immunol. 66, 139-151. Hellstrom, I., and Hellstrom, K. E. (1983).J . Biol. Resp. Mod$ 2,310-319. Hellstrom, K. E., Hellstrom, I., Kant, J. A., and Tamerius, J. D. (1978).J.Exp. Med. 148, 799-804. Hengst, J. C. D., Mokyr, M. B., and Dray, S. (1980).Cancer Res. 40,2135-2141. Hengst, J. C. D., Mokyr, M. B., and Dray, S. (1981).Cancer Res. 41,2163-2167. Hewitt, H. B. (1979).Clin. Radiol. 30, 361-369. Hewitt, H. B., Blake, E. R., and Walder, A. S. (1976). Br.]. Cancer 33,241-259. Hilgers, J., Sonnenberg, A., and Nusse, R. (1980). Br. J . Cancer 42,542-550. Holan, V., and Mitchison, N. A. (1983).E u r . J . Zmmunol. 13,652-656. Johnson, T. R., Chandler, J. P., and North, R. J. (1986).In preparation. Kanagawa, O., Louis, J. A., Eugers, H. D., and Cerottini, J.-C. (1983).J . Immunol. 130, 24-28. Kearney, R., and Nelson, D. S. (1973).Aust. J . E x p . Biol. Med. Sci. 51, 723-735. Kearney, R., Basten, A., and Nelson, D. (1975). Int. J . Cancer 15,438-450.

42

ROBERT J. NORTH

Key, M. E., Brandhorst, J. S., and Hanna, M. G. (1984).J. Biol. Resp. Modif.3,359-365. Kirchner, H., Chused, T. M., and Herberman, R. B. (1974)J. E r p . Med. 139,1473-1582. Kripke, M. L. (1981).Ado. Cancer Res. 34,69-106. Kruisbeek, A. M., and Hees, M. (1977).J. Natl. Cancer Znst. 58, 1653-1760. Lannon, D. R., and McKhann, C. F. (1982). Transplantation 33,99-100. Leclerc, J.-C., and Cantor, H. (198O).J. Zmmunol. 124,851-854. LeFrancois, L., and Bevin, M. J. (1984).J . E r p . Med. 159,57-67. Liew, F. Y.,Hale, C., and Howard, J. G. (1982).J. Zmmunol. 128, 1917-1922. Loveland, B. E., Hogarth, P. M., Ceredig, R. H., and McKenzie, 1. F. C. (1981).1. Exp. Med. 153,1044-1059. Lubet, R. A., and Carlson, D. E. (1978).J. Natl. Cancer Znst. 61,897-903. Macphail, S., and Stutman, 0. (1982).J. E r p . Med. 156,1398-1414. Mastrangelo, M. J., Berd, D., and Maguire, H. C. (1984). Cancer Treat. Rep. 68,207219. Milas, L., Hunter, N., Mason, K., and Withers, R. (1974). Cancer Res. 34,61-71. Mills, C. D., and North, R. J. (1983).J . E r p . Med. 157, 1448-1460. Mills, C. D., and North, R. J. (1985). Transplantation 39,202-208. Mills, C. D., North, R. J., and Dye, E. S. (1981).J . E r p . Med. 154,621-630. Mitzushima, Y., Wepsic, H. T., Yamumura, Y., and Desilva, M. A. (1984).Clin. E x p . Zmmunol. 57,371-379. Mokyr, M. B., Hengst, J. C. D., and Dray, S. (1982). Cancer Res. 42,974-979. Moore, M., and Williams, D. E. (1973). Znt. J . Cancer 11,358-368. Naor, D. (1979). Ado. Cancer Res. 29,45-125. North, R. J. (1982).J . E r p . Med. 55,1063-1074. North, R. J. (1984a).Ado. Zmmunol. 35,89-154. North, R. J. (1984b).Cancer Zmmunol. Zmmunother. 16,175-181. North, R. J. (1984~).Cancer Zmmunol. Zmmunother. 18,69-74. North, R. J. (1984d). Cancer Zmmunol. Zmmunother. 18,75-79. North, R. J. (1985). In preparation. North, R. J., and Bursuker, I. (1984).J . E r p . Med. 159,1295-1311. North, R. J., and Dye, E. S. (1985). Zmmunology 54,47-56. North, R. J., and Kirstein, D. P. (1977).J. E r p . Med. 145,275-292. North, R. J., Dye, E. S., Mills, C. D., and Chandler, J. P. (1982). Springer Semin. Immunopathol. 5,193-220. Old, L. J., and Boyse, E. A. (1964).Annu. Reu. Med. 15, 167-186. Old, L. J.. Boyse, E. A., Clark, D. A., and Carswell, E. (1962).Ann. N.Y. Acad. Sci. 101, 80-106. Parr, I., Wheeler, E., and AIexander, P. (1973). Br. J . Cancer 27,370-389. Perry, L. L., Benacerraf, B., and Greene, M. I. (1978).J . Zmmunol. 121,2144-2151. Pike, M. C., and Snyderman, R. (1976).J. Immunol. 117,1243-1249. Plata, F., Gomard, E., Leclerc, J.-C., and Levy,J. P. (1973).J. Zmmunol. 111,667-671. Prehn, R. T.,and Main, J. M. (1957).J . Natl. Cancer Znst. 18,769-778. Radov, L. A., Haskell, J. S., and Korn, J. H. (1976).Znt. J. Cancer 17, 773-779. Revesz, L. (1960).Cancer Res. 20,443-451. Rogers, M. J . (1984). Zmmunol. Today 5,167-168. Rolink, A. G., and Gleichmann, E. (1983).J . E r p . Med. 158,54&558. Rollinghoff, M., Staninski-Powitz, A., Pfizenmaier, K., and Wagner, H. (1977).J. Exp. Med. 145,455-459. Rosenberg, S. A., and Terry, W. (1977).Ado. Cancer Res. 25,323-388. Rouse, B. T., Rollinghoff, M., and Warner, N. L. (1973). Eur. J . Zmmunol. 3,218-222.

DOWN-REGULATION OF ANTITUMOR IMMUNE RESPONSE

43

Schatten, S., Drebin, J. A., Perry, L. L., Chung, W., and Greene, M. 1. (1984a). J . Zmmunol. 133,1064-1069. Schatten, S., Granstein, R. D., Drebin, J. A., and Greene, M. I. (1984b). C . R. C . Crit. Rev. Immunol. 4,335-379. Shear, M. J. (1944).J. Natl. Cancer Znst. 4,461-476. Takei, F., Levy, J. G., and Kilburn, D. G. (1976).J.Immunol. 116,288-293. Takei, F.,Levy, J. G., and Kilburn, D. G. (1977).J . Zmmunol. 118,412-417. Ting, C.-C., Rodrigues, D., and Nordon, R. (1982).Cell. Immunol. 66,45-58. Turk, J. L., and Parker, D. (1982).Zmmunol. Reu. 65,99-113. Tuttle, R. L.,and North, R. J. (19751.1.Natl. Cancer Inst. 55, 1403-1411. Tuttle, R. L.,Knick, V. C., Stopford, C. R., and Wolberg, G. (1983).Cancer Res. 43, 2600-2605. Ullrich, S. E.,and Kripke, M. L. (1984).J . Zmmunol. 133,27862790. Uyttenhove, C., Maryanski, J., and Boone, T. (1983).J . E x p . Med. 157,1040-1052. Vaage, J. (1971).Cancer Res. 31, 1655-1662. Vaage, J. (1973).Cancer Res. 33,493-503. Vaage, J. (1977).Cancer Res. 37, 1064-1067. Van Bekkum, D.W., and Kuaam-Shanzer, S. (1983).Transplant. Proc. 15,1451-1453. Wagner, H., Handt, C., Heeg, M., Rollinghoff, M., and Pfizenmaier, K. (1980).Nuture (London)284,278-280. Weissman, I. L. (1973).J . Natl. Cancer Znst. 51,443-448. Youn, J. K., LeFrancois, D., and Banski, G. (1973).J. Natl. Cancer Znst. 50,921-926. Yuhas, J. M.,Pazmino, N. H., and Wagner, E.(1975).Cancer Res. 35,237-241. Zbar, B., Bernstein, I. D., and Rapp, H. J. (1971).J. Natl. Cancer Znst. 46,831-839.

This Page Intentionally Left Blank

CELLULARASPECTSOFDNAREPAIR Bernard S. Strauss Department of Molecular Genetics and Cell Biology, The University of Chicago, Chicago, Illinois

I. Introduction .................................................... 11. Recognition of Damage by Cells . . . . . . . . . . . . HI. In Viuo Aspects of Excision Repair A. Measurement of Repair . . . . . . . . . . . . . . . . . B. Pool Size .................................................... C. Role of Different DNA Polymerases and Inhibitors . . . . . . . . . . . . . . . . IV. Effects of the Local Environment on Reaction and Repair.. . . . . . . . . . . . . V. Poly(ADP-Ribose) ............................................... VI. The Repair of 0-Alkylated Sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Adaptive Response. ............................ ... ... VIII. Bypass of Lesions and Its Consequence. .. ............. IX. Error-Prone Repair and Mutation. .................................. X. Biological Role of Repair. ......................................... XI. Appendix ....................................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

45

57

60 66 70 74 78 80 84 90 93 95

“For an object to be accessible to investigation, it is not sufficientjust to perceive it. A theory prepared to accommodate it must also exist. In the dialogue between theory and experience, theory always has the first word. It determines the form of the question and thus sets limits to the answer.” F . Jacob, “The Logic of Life.” Pantheon Books, New York, 1973

I. Introduction

Cells often require the accumulation of multiple insults before responding by mutation or death (Howard-Flanders and Boyce, 1966). Agents reacting with DNA to produce nucleotide changes which act as blocks to DNA synthesis are not usually lethal when only single adducts have been produced in the cellular nucleic acid (e.g., Strauss et al., 1979).In part, this may reflect the organization of the eukaryotic replication apparatus so that the sites of damage can serve as terminations within each replicon (Park and Cleaver, 1979a; Strauss et al., 1983), but it is also an indication of the presence in cells of “repair mechanisms,” metabolic systems which permit cells to remove or to bypass damage. In a strict sense, the term “repair” should be limited 45 Copyright 0 1985 by Academic Press, lac. ADVANCES IN CANCER RESEARCH, VOL. 45

All rights of reproduction in any form reserved.

46

BERNARD S. STFUUSS

to systems that remove damage and restore the DNA to its original structure (Hanawalt et al., 1979). Usage, however, requires that one consider as repair those mechanisms which permit cells to synthesize DNA past damage, often with mutagenic consequences. This review will therefore consider both the processes of excision repair and those that involve a more direct interaction with DNA synthesis. Since the mechanistic details of excision repair have been extensively reviewed recently (Hanawalt et al., 1979; Lindahl, 1982),including a review in this series (Teebor and Frenkel, 1983),this article will concentrate on the biological implications of the chemical data. In particular, I want to consider the interactions of damage to the DNA and the metabolic systems of the damaged cell. In this context, the major sort of damage to be considered is that in which a block to DNA chain elongation or initiation has been produced as a result of an alteration in a nucleotide so that it becomes noninstructive (or “pseudoinformational”; Walker, 1984) for Watson-Crick base pairing. Notice should be taken of two new journals with major interest in DNA repair. The new section of Mutation Research, “DNA Repair Reports,” is now (1984) in its second year, and the journal Carcinogenesis, now in Volume 5, devotes much attention to the relevant aspect of repair studies. In addition, the CANCERGRAMS on Molecular Biology (DNA), Mechanisms of Carcinogenesis (Macromolecular Alteration and Repair), Chemical Carcinogenesis (Nitroso Compounds), Mechanisms of Carcinogenesis (Oncogenic Transfomnation), and Short-Term Test Systems for Carcinogenicity and Mutagenicity contain abstracts of recent papers on the subject. CANCERGRAMS is published monthly and is distributed for the International Cancer Research Data Bank (ICRDB) by the National Technical Information Service (5285 Port Royal Road, Springfield, Virginia, 22161). ICRDB has also recently published a series of oncology “overviews” containing selected abstracts in a variety of the fields related to repair. As a result of the availability of this abstract series, I have restricted the review to selected topics and have not attempted an exhaustive survey of the literature. 11. Recognition of Damage by Cells

DNA synthesis can be studied in vitro by modification of the Sanger chain termination sequencing technique (see Smith, 1980) which permits determination of the exact nucleotide at which termination has occurred. In one published method (Moore and Strauss, 1979), a single-stranded DNA of known sequence, typically a circular viral DNA

DNA REPAIR

47

such as M13, is hybridized with a restriction fragment derived from the replicative form to give a primed template for DNA synthesis (Fig. 1). A model for synthesis on double-stranded templates employs a double-stranded DNA with a single nick, allowing either nick translation or strand displacement to occur (Fig. 1B). Although long stretches can be synthesized when undamaged DNA is used as the template, it is apparent that some stops or pauses appear at particular nucleotide sites with particular polymerases (Weaver and DePamphlis, 1982; Kaguni and Clayton, 1982; LaDuca et al., 1983; Hillebrand et al., 1984). These stops may be absolute terminations or may represent temporary pauses in synthesis: clearly, the distinction between pauses and stops is not absolute. Many (but not all) of the stops occur at regions in the DNA in which hairpin loops can be formed (Weaver and DePamphlis, 1982). Such stops can sometimes be overcome by addition of the Escherichia coli single-stranded DNA-binding protein (ssB) which can “melt out” regions of secondary structure (LaDuca et al., 1983). Other stops or pauses occur at places that are both sequence and polymerase specific (Kaguni and Clayton, 1982), although the rules that determine the importance of particular sequences have not been worked out. Lesions in single-stranded DNA produced by treatment with ultraviolet light (UV) (Rabkin et al., 1983), with N-acetoxyacetylaminofluorene (N-acetoxy AAF) (Moore et al., 1980), with benzo[a]pyrene diolepoxide (BPDE) (Moore and Strauss, 1979), with psoralen (Piette and Hearst, 1983), with hydroxyaminoquinoline l-oxide (Yoshida et al., 1984), or by removal of pyrimidines to make apyrimidinic (AP) sites (Sagher and Strauss, 1983) or of purines (Schaaper et al., 1983) block DNA synthesis. The block occurs so that synthesis terminates one nucleotide before the lesion (Moore and Strauss, 1979) (Fig. 2). Certain polymerases are able to synthesize to opposite the lesion, and this ability depends on the polymerase, the nature of the lesion, the site in the sequence, the cations present, and the dNTP concentration (Rabkin and Strauss, 1984). Termination is not due to 3’ > 5’ exonuclease activity (Moore et al., 1981), but insertion opposite lesions is strongly affected by this proofreading activity (Rabkin and Strauss, 1984). Some lesions block DNA synthesis at some sites but not at others: P-methylguanine appears to fit in this category (Toorchen and Topal, 1983). These methodologies depend on the use of natural templates. Although much useful information has been derived from the study of synthetic templates (e.g., Singer et al., 1984), the high probability of slippage during DNA synthesis on such templates makes it less certain that the experiments model in vivo synthesis. Synthesis by DNA polymerase I has been reported to continue on a

48

BERNARD S. STRAUSS

0

A

REACT * MEN1 FROM R.F. dATP dGTp (32P) dCTP dTTP

STAGE I

ainqle dNTP

R E STRICT

SECOND STAGE

(0

Denature ond Sequence

B

RESTRICT R F I

0 0 GAPPED

I

-

--

- restrict ---- -

- ---

purify template synthesis with

"P

dNTPs

Denature and Sequence

FIG. 1. (A) Scheme for study of DNA termination and bypass synthesis in uitro. Single-stranded circular viral DNA is reacted with mutagen and then annealed with an isolated restriction fragment prepared from viral replicative form. I n uitro DNA synthesis is then allowed to continue with a mixture of [32P]deoxynucleosidetriphosphates to form a first-stage product (Rabkin et al., 1983). This product is analyzed by treatment with the original restriction fragment and the position of termination of synthesis determined on a Sanger sequencing gel. Numerous variations of the basic technique are available. For example, a nonlabeled product may be 5' end labeled with =P after the reaction, obliterating the need for labeled triphosphates in the reaction mixture. Second-stage reactions are carried out by isolating the first stage and reacting a second time with any polymerase and single (nonlabeled) dNTPs. After restriction and analysis on sequencing gels, the identity of the nucleotide(s) promoting elongation can be determined. (B) Schematic diagram showing the preparation of double-stranded templates for the study of termination. Single-stranded viral DNA is annealed with denatured linear double-stranded DNA formed by cutting of viral replicative form with an enzyme making a single cut. After purification, the resulting nicked double strand can be used for replication by either strand displacement or nick translation. Either (+) or (-) strands can be reacted at will. A variant of this procedure results in molecules with gaps of desired length when the replicative form is treated with different restriction enzymes making single cuts or with one enzyme making two cuts. Hybridization is with the desired fragment purified by electrophoresis.

I

I AP

DNA

Icontrol

I

FIG.2. Termination of i n uitro DNA synthesis at or before AP (apyrimidinic) sites in DNA. DNA containing reduced AP sites was primed with a restriction fragment as shown in Fig. 1A and used as a substrate for synthesis using bacteriophage T4 DNA polymerase, E. coli DNA polymerase I with either Mgz+ or Mn2+as cofactors, AMV reverse transcriptase, or a DNA polymerase (z from human lymphoblastoid cells. The control uses DNA without AP sites as substrate. The figure shows termination before putative AP sites by T4 DNA polymerase as well as the effect of metal ion or of polymerase on the site of termination. [From Sagher and Strauss, 1983. Reprinted with permission fiom Biochemistry 22, 4518-4526. Copyright (1983) American Chemical Society.]

50

BERNARD S. STRAUSS

double-stranded template damaged with angelicin, even though synthesis on single-stranded templates is absolutely blocked (Piette and Hearst, 1983). It has been hypothesized that this is accomplished by use of the strand-displaced region as a template to permit the bypass (Lundquist and Olivera, 1982). Since this would be expected to be highly mutagenic, the hypothesis is open to test. Even lesions in single-stranded DNA can be bypassed in vitro given the correct conditions (Rabkin et al., 1983). Such bypass is determined both by the sequence distal to the lesion as well as by the particular polymerase, metal ions present, and concentrations of the deoxynucleotides. In one in uitro system polymerase p is more specific than polymerase a, the sequence TGCCCT is particularly effective as a site of bypass by dG, and at the site of an AAF-dG lesion, concentrations of over 0.5 mM dATP may still be effective in increasing the rate of bypass (Rabkin and Strauss, 1984).Notwithstanding our current ability to manipulate DNA synthesis catalyzed by known enzymes in vitt-o in model systems, there is no compelling evidence for direct bypass (translesion synthesis) in uiuo. One reason for supposing that more restrictive DNA synthesis control mechanisms operate in uiuo is the evidence that mammalian cells recognize in some way that their DNA has been damaged and halt DNA synthesis until repair processes take care of the damage (Painter 1981, 1983). Normal cells halt synthesis of DNA after treatment with ionizing radiation (Painter and Young, 1980; Houldsworth and Lavin, 1980), bleomycin (Waldman et al., 1983; Morris et al., 1983; Cohen and Simpson, 1982), or neocarzinostatin (Shiloh and Becker, 1982). Cells from individuals with ataxia telengiectasia, on the other hand, continue DNA synthesis at normal rates for some time after they have been damaged. The cells have been killed-ataxia cells are hypersensitive to radiation (Waldman et al., 1983)-and it must be concluded that the ability to turn off DNA synthesis after damage is vital to cell survival. The observations indicate that the initiation of DNA synthesis in vivo is carefully controlled and affected by damage to the molecule. The use of those biochemical mechanisms which permit bypass in the simple in uitro systems must therefore be restricted. The nature of the mediator (Painter, 1983) which controls synthesis is unknown. 111. In Viwo Aspects of Excision Repair

The general outline of the excision repair mechanism has been known for over 20 years since the original descriptions by Setlow and Carrier (1964) and Boyce and Howard-Flanders (1964). The likely

DNA REPAIR

51

sequence of reactions by which cells remove damage was supposed at that time to be (1)endonucleolytic action at the site of damage, (2) exonucleolytic excision of the damage, (3) repair synthesis and excision of contiguous nucleotides, (4) cessation of excision and completion of repair synthesis, and ( 5 ) joining of the newly synthesized strand to preexisting DNA (see Fig. 10 in Strauss, 1968). This sequence still describes our understanding of the reaction, although there is doubt about whether the excision and repair synthesis steps proceed in sequence or simultaneously (see below). Major advances have been made in the past few years in our understanding of the molecular basis of repair in prokaryotic systems. The history of studies on DNA repair indicates that understanding the bacterial mechanism invariably precedes similar understanding in eukaryotes. It is therefore necessary to briefly summarize the new advances. Such a restatement is useful because the discovery that there are major variations in what seemed to be a unitary bacterial mechanism provides a necessary cautionary note. Our present understanding of the mechanism of repair began with the discovery by Lindahl(l976) of the DNA N-glycosylases, enzymes which recognize particular purines or pyrimidines altered in specific ways (uracil can be considered an altered thymine) in DNA and cleave the N-glycosidic bond so as to liberate free bases, leaving the chain intact, but with an apurinic or apyrimidinic site (AP site) (Lindahl 1979, 1982). One repair pathway includes a specific DNA Nglycosylase to remove the damaged base, an AP endonuclease, one or more exonucleases, a DNA polymerase to insert a “patch” of new DNA, and a DNA ligase to seal the newly inserted patch to the old chain. Such systems have been reconstructed using enzymes from mammalian cells (Bose et al., 1978; Mosbaugh and Linn, 1983). The recognition that the UV endonucleases of bacteriophage T4 and of Micrococcus luteus had both glycosylase and endonuclease activity (Demple and Linn, 1980; Haseltine et al., 1980; Radany and Friedberg, 1980; Seawell et al., 1980; Lindahl, 1982) suggested that all excision repair proceeded via a glycolytic mechanism, even in eukaryotes, since it was known that intracellular treatment with the T4 UV endonuclease could bypass the incision defect in xeroderma pigmentosum cells (Tanaka et al., 1977). However, it appears that even in E. coli, the mechanism of excision is more complex. It had been known for some years that at least three gene products, the products of the uvrA, uvrB, and uvrC genes, are required for the excision of pyrimidine dimers in this organism (Seeberg et al., 1976). It was also known that ATP is required for the removal of the dimer (Seeberg and

52

BERNARD S. STRAUSS

Steinum, 1982). This mechanism has now been analyzed in detail. The combination of the three gene products with damaged DNA occurs in a stepwise fashion with the result being breakage of the damaged DNA at two sites, neither directly adjacent to the dimer (Sancar and Rupp, 1983).The net result is the removal of a patch of about 12 nucleotides. Excision repair in bacteria can then occur by two partially separate pathways: base excision proceeding by a glycolytic mechanism with the release of free bases (Lindahl, 1979)and nucleotide excision in which the initial step frees nucleotides. It seems likely that this dichotomy applies also to mammalian cells, even though there is as yet no enzymatic analysis of excision repair in mammalian cells comparable to that in E. coli. The enzymes of base excision have been studied in mammalian cells (e.g., Lindahl, 1982; Singer and Brent, 1981; Margison and Pegg, 1981),and there is every reason to suppose that the DNA N-glycosylase pathway is operative both for the removal of unwanted uracil from the DNA and for the repair of radiation- and alkylation-damaged bases. An additional method of dealing with alkylated bases by methyl transfer is discussed below. Our understanding of the detailed mechanism of excision in mammalian cells is limited by our ignorance of the enzymatic basis of the process and by our relative inability to apply genetic analysis to the system. Although serious attempts at mutant analysis are now under way (see below, Section X), to a large extent our understanding is based on inhibitor studies and on indirect methods of analysis. Insofar as many of the experiments have been done in tissue culture, the results may be (and in some cases are) dependent on whether the cells are growing or confluent, transformed or primary (Squires et al., 1982; Elliott and Johnson, 1983),as well as on the species (Hart and Setlow, 1974)and cell type (Liu et al., 1983).There is no clear evidence that repair activity is directly affected by passage number, by the age of the donor cells (Liu et al., 1982),or by the life-span of the species (Francis et al., 1981).The correlation between UV-induced repair activity and life-span found by Hart and Setlow (1974)for 7 species was not observed when 34 species were studied (Kato et al., 1980). If repair activity is affected by stage in the cell cycle, then the average repair activity in slowly growing cells will of course appear lower than in rapidly dividing cells. Some data indicate that the repair rate for UV damage in fibroblasts is the same for rapidly growing and confluent cells (e.g., Liu et al., 1982).This does not appear to be the case for lymphoblasts (Straws, 1984). Sirover and Gupta (1983)have summarized data showing a 2.3-fold greater rate of chemical adduct loss

DNA REPAIR

53

(acetylaminofluorene, bromomethylbenz[a]anthracene, and cis-Pt) in proliferating as compared to quiescent cells. In order to observe excision repair, a cellular system must be subject to an agent which damages or alters the DNA. Not all alterations, even though mutagenic, necessarily induce repair. Only those changes that inhibit DNA synthesis are likely to activate the excision repair mechanism. For example, substitution of BUdR for dT in DNA when not accompanied by the formation of apyrimidinic sites is not necessarily recognized as damage by the repair systems as is seen by the common use of BUdR incorporation for the measurement of repair (Friedberg and Hanawalt, 1981). The formation of 06-methylguanine (but not 06-ethylguanine) need not be a signal for the operation of the excision repair mechanism (Simon et al., 1981), although, as will be seen below, there are special proteins which recognize and eliminate this adduct. The transformation of cytosine into uracil by the action of the mutagenic agent bisulfite does not activate the (UV) nucleotide excision mechanism, although a special glycosylase does recognize and remove this base (see Lindahl, 1982; Arenaz and Sirover, 1983). In addition, methylation of cytosine by the DNA methylases produces altered bases that are not recognized by the excision repair mechanisms. In fact, in bacteria the system is arranged so that one of the methylated bases, 6-methyladenine, is actually the signal which protects sequences from the operation of a mismatch repair system (Lu et al., 1983; Radman and Wagner, 1984).

A. MEASUREMENT OF REPAIR The simplest method of measuring repair involves determination of the loss of adduct or, in the case of radiation-induced damage, of alteration in the DNA. Such determination is by difference, based on analysis of the DNA at different times. Demonstration of a differential loss of adduct not due to dilution by newly synthesized D N A is unequivocal evidence for repair. The sensitivity of the measurements has increased enormously with the introduction of immunoassays, particularly ELISA for the analysis (Table I). It is possible in the case of adducts which can be determined immunologically to determine the presence of just a few residues in a genome. The lower limit of detectability of benzopyrene adducts is 1 per 3 x 108 nucleotides (0.1 fmollpg DNA). This methodology has been used by Perrera et al. (1982) to attempt the detection of benzopyrene residues in lung. Only a small sample of tissue was available, but five lung cancer patients gave “presumptive” positive results. Such methodologies should be

TABLE I SOMEANTIBODY-BASED ANALYTICALTECHNIQUES FOR ALTEREDNUCLEOTIDES~ Adduct

Antibody

Method

Sensitivity

N-Acetyl-2-amino fluorene

PC

High-sensitivity ELISA

N-Acetyl-2-aminofluorene Aflatoxin B1

PC MC

USERIA USERIA

Ring-opened aflatoxin BI

MC

Competitive ELISA

0.1 finol/0.8 mg DNA at 0.1 pmol adductlmol DNA nucleotides 2 fmol 1 per 250,000 nucleotides, 15 fmol/lO ng DNA 1 per 300,000 nucleotides

Benzo[a]pyrene

PC

USERIA

Benzo[a]pyrene Benzo[a]pyrene @-Ethylguanine

PC MC MC

USERIA ELISA ISB

@-Ethylguanine OPMethylguanine

MC MC

Thymine glycol

MC

IEM RIA and chromatography ELISA

2 fmol

UV photoproducts (dimers)

MC

UV photoproducts (dimers)

PC

RIA

2.5 Jim2 = 1 per 1.3 X 104 Ts

13 finol/mg DNA; 1 per 250,000 nucleotides 3 fmol 19 fino1 0.3 b o l l 3 pg DNA inhibition of binding 1 adductllO0 pmol 1 per 1.5 x 107 dC

Reference Van Der Laken et al. (1982) Hsu et al. (1980) Haugen et al. (1981) Groopman et al. (1982); Hertzog et al. (1982) Hsu et al. (1981) Perrera et al. (1982) Santella et al. (1984) Adamkiewicz et al. (1985); Muller and Rajewsky (1980) Nehls et al. (1984) Wild et al. (1983) Leadon and Hanawalt (1983) Rajagopalan et al. (1984); Srickland and Boyle (1981) Mitchell and Clarkson (1981); Clarkson et al. (1983)

a PC, polyclonal; MC, monoclonal; RIA, radioimmunoassay; ELISA, enzyme-linked immunosorbent assay; USERIA, ultrasensitive enzymatic radioimmunoassay; ISB, immuno slot blot. General references include Poirier (1981), Muller and Rajewsky (1981), and Adamkiewicz et al. (1985).

DNA REPAIR

55

of use in population surveys (Perrera et al., 1982).Advances in chemical separation techniques and in the detection of adducts by fluorometry have increased the sensitivity of the nonimmunological, nonradioactive detection procedures by which loss of adducts can be measured (Swenberg, 1985). An analytical method of some utility depends on the recognition of pyrimidine dimers (perhaps the most studied lesion) by the M. Zuteus UV endonuclease [a detailed description of methodologies is to be found in Friedberg and Hanawalt (1981)l. Cyclobutane dimer-containing DNA strands are broken by treatment with this enzyme, with the number of breaks being equal to the number of available pyrimidine dimers. Sucrose gradient analysis of the product then permits calculation of the number of breaks from the molecular weight. The methodology does not measure removal of dimers, but only whether an incision has occurred, since a site at which such an incision had occurred but at which a dimer was still attached via one base would not register the presence of the dimer. This feature was not initially considered critical, but discovery that some repair proceeds via DNA N-glycosylase action makes it more important. Determination of the removal of an adduct does not give information on the mechanism of its loss. In order to demonstrate excision repair, one must show that new DNA has been inserted into an “old” strand and that the breaks which were made to permit this insertion have been healed. Insertion of new DNA into old strands is generally demonstrated by methods which involve the incorporation of radioactive thymidine or its halogenated analogs into DNA. The most established methodology is that devised by Pettijohn and Hanawalt (1964). DNA is prelabeled with [14C]thymidine, the cells are treated with some repair-inducing agent, and then are incubated in medium containing [3H]BUdR. Semiconservative synthesis results in formation of DNA of hybrid density whereas incorporation of small patches of BUdR does not alter the density sufficiently to result in a detectable change from that of the original. Coincidence between the patterns of 3H and 14C is observed after CsCl density gradient centrifugation. A variation of this technique makes it possible to determine the “patch size,” that is, the average number of nucleotides inserted for every lesion removed. Shearing of the repaired patches will increase the ratio of BUdWdT so that the average density of the molecules changes enough to be detectable in the gradients. This method gives values of 20-40 nucleotides replaced per UV-induced lesion removed (Edenberg and Hanawalt, 1972).A modification of the technique prepares DNA of uniform small size by staphylococcal nuclease digestion

56

BERNARD S. STRAUSS

of nuclei to yield nucleosomal DNA (Walker and Th‘ng, 1982). The uniform size of the fragment makes the density distribution easier to interpret (Th’ng and Walker, 1983). This methodology is to be compared to a method depending on flash photolysis. BUdR-containing DNA is sensitive to degradation by near-UV irradiation. Such irradiation sensitizes the DNA to alkali (probably the result of the formation of AP sites from the halogenated pyrimidine), and sucrose gradient centrifugation can be used to measure the resulting decrease in average molecular weight. Since the dose required to produce a break is inversely related to the size of a BUdR-containing patch, it is possible to calculate the patch size from the slope of dose-effect curves (Regan et al., 1971). The photolysis experiments show a wide range of average patch sizes, depending on the compound used to induce repair. Earlier values obtained by this method gave figures of about 140 for the repair patches resulting from UV damage and much lower patch sizes (approaching 1)for X ray- and methyl methanesulfonate (MMS)induced change (Regan and Setlow, 1974). The reason for the different results given by the density gradient and flash photolysis methodologies is not clear. Since many compounds induce a variety of lesions, the difference could be due to a different “mix” of products in the two cases. Recently, Snyder and Regan (198213) have found patch sizes of about 40 as one of the components of MMS-induced repair, a number comparable to that observed after UV-irradiation (Zelle et al., 1980). There probably is a “small patch” MMS component (Snyder and Regan, 1982b), but there now seems to be more general agreement that the size of the longer patch component is about 40 nucleotides. Very likely some technical component of the flash photolysis protocol is responsible for the adjustment in patch size value. However, until the enzymatic mechanism(s) in mammalian cells has been worked out, the theoretical significance of any particular experimentally determined patch size remains unknown. The finding in bacteria that nucleotide excision occurs as the result of double nicks -12 nucleotides apart on the same face of a DNA strand (Sancar and Rupp, 1983)provides a basis for a minimum patch size and a possible explanation of the difference in patch size between nucleotide excision-inducing agents such as UV and base excision inducers such as MMS, or at least its major component (Snyder and Regan, 1982a). However, the enzymatic studies in bacteria do not suggest why patch sizes of either 40 or 120 occur. It seems likely that the production of patches of this size requires 5’ > 3’ exonucleolytic action in addition to the initial incision and excision events. At this stage of our knowledge it is not clear how the final patch size is

DNA REPAIR

57

regulated or whether patch size is determined only by the competition between ligation and nuclease action. There is evidence that patch size as determined in cell cultures can be altered by changes in the biochemical environment, and increased patch size might result from what are essentially biochemical artifacts. For example, in the presence of the inhibitor hydroxyurea, the specific repair synthesis activity was doubled in a 2-hr period. Pool size effects (but not metabolic interconversion) were eliminated as an explanation, and it was concluded that patch size was increased (Clarkson, 1978), possibly because of a defect in ligation. Similarly, DNA repair synthesis is increased by some inhibitors of poly(ADP-ribose) synthetase, and it has been suggested that this occurs because of an inhibition of the ligation reaction resulting in more available time (before ligation) for the production of (longer) patches (James and Lehman, 1982; Cleaver et al., 1983a), although at the time of this writing no direct demonstration of an increased patch size had been reported. Collier et al. (1983) account for patch size and the distribution of different patch sizes by supposing that excision terminates when a specific nucleotide sequence is encountered. Although no biochemical evidence is available, their formal model predicts a heterogeneous distribution of patch sizes in qualitative agreement with the estimates of Edenberg and Hanawalt (1971) based on density distributions in CsCl gradients. B. POOLSIZE The methods described above for the measurement of excision repair depend on physical or biochemical techniques to separate repair from replicative incorporation of radioisotope into DNA. Additional methods of this type include the use of benzoylated naphthoylated DEAE-cellulose (BND-cellulose) for the separation of newly synthesized from repaired DNA (Scudiero et al., 1975; Straws, 1981). Since the DNA growing point contains single-stranded regions, it will adhere to BND-cellulose even at high (1.0 M) salt concentration. Repair synthesis occurs throughout the genome, so that many completed patches will be far from growing points in double-stranded DNA, which is eluted from BND-cellulose at 1.0 M salt. Repair specific activity can therefore be determined in the 1.0 M salt eluate from BND-cellulose columns. This methodology requires that chain elongation be restricted so that all semiconservative synthesis is close enough to the single-stranded DNA regions at the growing point (in the sheared preparations necessary when using mammalian DNA) to adhere to the resin. Hydroxyurea is reported to inhibit elongation but

58

BERNARD S . STRAUSS

not initiation so that short DNA fragments containing the origins accumulate (Vassilev and Russev, 1984). Such fragments have singlestranded regions and adhere to BND-cellulose in 1.0 M salt (Scudiero and Strauss, 1974; Hensen, 1978). It is therefore necessary to slow down semiconservative chain elongation, and this is done by the addition of an inhibitor, usually hydroxyurea. Under these circumstances, there is good separation of repair and replicative synthesis, and the repair activities are equivalent to those measured by the CsCl methodology (Scudiero et al., 1984a). BND-cellulose measurement is an example of the set of methods that depend on the incorporation of [3H]thymidine. Perhaps the most widely used and generally available method for the measurement of excision repair, the determination of unscheduled DNA synthesis (UDS), depends on the incorporation of r3H]thymidine followed by autoradiography. This method requires that the cytologist separate replicative from repair synthesis by observation of the density of silver grains over any particular nucleus. Quantitation of the method depends on estimating the amount of thymidine incorporated by grain counts. The interpretation of such measurements, as with any measurement based on the incorporation of [3H]thymidine, depends on the nucleotide pool size. As has been pointed out by several authors, any experimental treatment that results in changes in the nucleotide pools will have an effect on the repair measurements (Strauss, 1981; Cleaver et aZ., 1983~). Changes of 30 or 40% in “repair activities” can often be accounted for by slight changes in the available pools. Any process that suddenly interrupts DNA synthesis is likely to have at least a transient effect as the regulatory mechanisms adjust to the lessened demand for dNTPs. Treatment of mammalian (Das et al., 1983) and bacterial (Das and Loeb, 1984) cells with mutagens can have a dramatic, rapid, and differential effect. Treatment of Chinese hamster cells with N-methyl-N-nitro-N’-nitrosoguanidine (MNNG) or with UV results in a rapid expansion of the dATP and TTP pools within 2-4 hr. The levels of dGTP and dCTP remain relatively constant. The pool increases by a factor of two or three, depending on the time af3er treatment, and the level remains elevated for some time. Such changes in the pool could result in a serious underestimate of the patch size if not taken into account. In order to partially overcome the problem of pool size, certain simple strategems are available. One method involves calculation of a relative pool size based on the dilution of incorporated radioactivity resulting from the addition of a known amount of thymidine. From the amount of thymidine added and the ratio of radioactivity incorporated

DNA REPAIR

59

at the two specific activities, a relative pool size can be calculated and used to correct the results to some standard condition (Scudiero et al., 1975; Strauss, 1981; Clarkson, 1978).Since there is (presumably) little or no free thymidine in cells, such procedures yield only relative values. A direct method involves chromatographic determination of pool components. It is observed that treatments at high concentrations of MMS depress the dTTP pool (Bianchi et al., 1983). This depression occurs at concentrations of MMS which reduce the ATP content of cells (Sims et al., 1983; Strauss, 1984) and which would therefore prevent any phosphorylation (see below). Treatments at high doses of cytotoxic agents can result in unsuspected metabolic changes which affect repair values. Pool size changes can be particularly important when changes of repair activity up to %fold are encountered. A doubling of pool size would, for example, result in an apparent change in repair activity by a factor of three, a change that would be considered highly significant. Considering the numerous possible reasons for change, it would seem that while often of great qualitative importance (consider, for example, the demonstration that xeroderma cells of many complementation groups have low excision capabilities), the interpretation of smaller quantitative differences is difficult. Conclusions based on quantitative determinations of excision repair capacity should be checked by an independent methodology. This caution is particularly important in considering the possible role of the cell cycle in determining excision activity. It does seem established that a variety of repair-related enzymes vary in their activity in a regular way during the cycle. Uracil glycosylase and DNA polymerases appear to vary in a systematic manner (Gupta and Sirover, 1984). Excision repair, measured both by the incorporation of thymidine in the presence of hydroxyurea and repair synthesis as measured in CsCl gradients (Sirover and Gupta, 1983), also varies by about 4-fold. However, this last variation would be expected if the pool sizes were also to change. An interesting temporal aspect of the variation has also been reported with the peak in nucleotide and base excision repair coming immediately before the peak of DNA synthesis activity. In contrast, the peak of excision repair and synthesis coincided in Bloom’s syndrome cells (Gupta and Sirover, 1984). These data on variations in repair synthesis capability are to be contrasted with conflicting reports on the relative capacity of cells at different stages of growth to remove lesions. Growing and quiescent human skin fibroblasts are reported as equally competent in the removal of UV-induced pyrimidine dimers (e.g., Snyder and Regan,

60

BERNARD S. STRAUSS

1982a, Fig. 5). Kaneko and Cerutti (1980, Chart 3) report equal removal of AAF-DNA adducts by growing and confluent human fibroblasts. In contrast, Friedlos and Roberts (1978),studying 7-bromobenz[a]anthracene adducts in Chinese hamster cells, and Scudiero et al. (1976), using AAF adducts in human lymphocytes, observed a major difference in excision, comparing proliferating and nonproliferating cells. For lymphocytes at least, there is a correlation between changes in repair synthesis and excision. Sirover and Gupta (1983) have suggested a direct relationship between cell proliferation and excision repair capacity. AND INHIBITORS C. ROLE OF DIFFERENTDNA POLYMERASES

Cells contain several DNA polymerases with different properties and (presumably) different functions. In mammalian cells, two major polymerases, pol a and pol p have been reported to play a role in repair. Pol p is small, of -30,000-50,000 Da, and does not vary much in activity with changes in the cell cycle. DNA pol (Y is larger, about 120,000-220,000 Da (the exact value varies in different reports, possibly because of the occurrence of proteolytic degradation in the preparations). Pol a varies extensively in its activity throughout the cell cycle and is presumed to be the replicative polymerase in eukaryotic cells (for a general introduction to the polymerases, see Kornberg, 1980, 1982). The finding that aphidicolin-resistant Chinese hamster mutants have an altered pol a and a concomitant high spontaneous mutability (Liu et al., 1983) fits the concept of a as a replicative enzyme. It has been considered that pol p is more likely to be involved in eukaryotic repair (e.g., Hubscher et al., 1979) in supposed analogy with the system in bacteria in which pol I is the most attractive candidate repair enzyme and pol I11 the replication enzyme. However, both polymerases I1 and I11 of E. coli can participate in excision repair, since pol I-deficient mutants remain excision competent (see HanaWalt et al., 1979). Furthermore, recent studies with the inhibitors aphidicolin and with dideoxynucleosides indicate that in eukaryotes both a and p polymerases are involved in repair (e.g., Miller and Chinault, 1982). An interesting argument for the involvement of polymerase a in repair depends on the properties of cells with an aphidicolin-resistant, mutagenic polymerase (Y (Liu et al., 1984). Holding wild-type mutagen-treated cells so that they can repair but not carry out replicative DNA synthesis results in a dramatic decrease in the recovery of UV-induced mutations. Since such holding only increases the mutational yield in the aphidicolin-resistant cells, Liu et al.

DNA REPAIR

61

(1984) conclude that mutant polymerase (Y must be involved in an “error-prone” repair synthesis. Dresler and Lieberman (1983) report that the relative use of pol a for repair induced by methylating agents, acetoxyacetylaminofluorene, or UV is greater at higher doses, a conclusion based on the greater sensitivity to aphidicolin of repair in human fibroblasts at high doses. Pol (Y requires gaps of about 20 nucleotides for its action, pol p acts at nicks (see Kornberg, 1980). It may be that what is observed in the laboratory when higher doses are used (as they ofien are because they make all the measurements easier) is a removal-patch-ligation mechanism that mimics in vivo repair, but which is not physiologically equivalent. Dresler and Lieberman (1983) conclude that much natural repair is by a p polymerase except in the case of repair of UV damage in the skin, which can receive (naturally) large doses of radiation. Of course, a major biological use of the repair mechanisms must be in the processing of UV-damaged DNA. An attempt is being made to understand the detailed processes of excision repair by the use of permeabilized cells and mutants, both induced and naturally occurring. Studies with repair-deficient xeroderma pigmentosum cells (Kraemer, 1983) are a standard control in determining the physiological relevance of particular in vitro systems. For example, the permeabilized system developed by Dresler et al. (1982) shows incorporation of dNTPs into repair patches when normal but not xeroderma cells are used. In this study, repair and replicative synthesis were separated by control of dNTP concentration and by the addition of salt. The K,,, (for dNTPs) for replicative synthesis was about 8 p M compared to -1 p M for repair synthesis. These results are useful, since they indicate how repair synthesis can continue in the presence of hydroxyurea, an inhibitor of ribonucleotide reductase which results in the lowering of nucleotide pools. The repair complex must have a higher affinity for dNTPs than does the replicating complex, the replisome, and therefore functions at lower concentration (but see below for an alternative interpretation). ATP is required for repair (Dresler and Lieberman, 1983), and the role of pol (Y is indicated by the inhibitory effect of aphidicolin but not dideoxytriphosphates in this permeabilized system. Kaufman et al. (1983)have used a similar permeabilized system to show that cytosine arabinoside is incorporated into the interior of DNA chains, at least in part, and does not necessarily terminate chains, since a large portion of the repair label becomes resistant to S1 and to exonuclease 111 digestion, indicating the formation of diester bonds. Studies with permeabilized cells confirm the finding with intact

62

BERNARD S. STMUSS

cells that xeroderma pigmentosum-derived systems are unable to repair UV-damaged DNA. Introduction of T4 UV endonuclease into the nuclei of such cells activates the repair process, indicating that it is the incision step which is defective (Tanaka et al., 1977).However, use of extracts has not yet provided any definitive evidence as to the molecular mechanism of the xeroderma defect. A report, now almost 10 years old, that extracts of xeroderma cells would incise UV-induced lesions in purified DNA, but not in chromatin (Mortelmans et al., 1976), has been confirmed by Kano and Fujiwara (1983),who report that extracts of group A, C, or G xeroderma cells became competent for excision after depletion of nonhistone proteins from the target DNA with 0.35 M NaC1. Extracts of group D xeroderma cells catalyzed the excision of intact chromatin. Surprisingly, in view of its demonstrated requirement for repair (Dresler and Lieberman, 1983),it was not necessary to add ATP to the reaction mixture. The results are also hard to understand in view of the experiments of deJonge et al. (1983), who microinjected extracts of HeLa and xeroderma pigmentosum complementation group C cells into group A, C, D, and F cells and measured unscheduled DNA synthesis (UDS). The extracts, including the group C extract, resulted in a (transient) restoration of activity in group A, but not in group C, D, or F cells. The “correcting factor” was sensitive to proteinase, making it seem likely that a typical protein gene product is involved. Equally “traditional” results have been obtained by Hellend et al. (1984), who report an endonuclease, deficient in xeroderma cells, whose activity is restored by mixing extracts from complementation group A and C cells. However, the situation in xeroderma must be more complex. Mansbridge and Hanawalt (1983) examined the residual repair activity of complementation group C xeroderma cells. They conclude that the residual repair is limited to particular regions of the DNA as though the bulk of the DNA in xeroderma cells is not accessible to repair enzymes. These data, plus the observations on the nonaccessibility of damage in chromatin discussed above, do not readily fit a traditional enzyme loss hypothesis unless some special interaction between amount of enzyme and sensitivity of damage in chromatin is postulated. Happily, progress on the cloning of the repair genes (see below) promises more mechanistic studies in the near future. Extensive genetic heterogeneity as observed in xeroderma pigmentosum with the numerous complementation groups discussed above is also seen in ataxia telangiectasia (Jaspers and Bootsma, 1982).Cells of individuals with this condition of hypersensitivity to ionizing radiation, bleomycin, and neocarzinostatin fall into (at least) three comple-

DNA REPAIR

63

mentation groups. In contrast, four Fanconi’s anemia patients fell into one complementation group, suggesting to Zarkzweski et al. (1983) that this “repair deficiency” disease is genetically homogeneous. Studies of excision repair have made extensive use of inhibitors of replication in order to lower the background of synthesis against which repair must be measured. A list of inhibitors effective or used in the study of DNA repair is given in Table 11. Of these inhibitors, hydroxyurea has been the most widely used as a suppressant of replicative DNA synthesis. The efficiency of hydroxyurea is due to its action as an inhibitor of ribonucleotide reductase, thereby lowering the pool of available deoxynucleoside triphosphates (Timson, 1975). Nonetheless, it is still a question as to why repair synthesis is relatively unaffected by concentrations of hydroxyurea, which have drastic effects on replicative synthesis. An argument based on the lower K, for repair (Dresler et al., 1982) is given above. However, it is not clear that this is the (only) explanation. Hydroxyurea causes the accumulation of UV- or MMS-induced strand breaks in confluent but not in dividing human fibroblasts (Snyder, 1984a), and a decrease in the purine but not the pyrimidine pools, even in nontreated cells (Snyder, 1984b). This accumulation of breaks is counteracted by addition of a mixture of deoxynucleotides. In contrast, hydroxyurea remains effective in inhibiting replicative synthesis even when the addition of high concentrations of exogenous deoxynucleosides has created large internal pools. Snyder (1984a) therefore suggests that dNTP pools for repair and replication are somehow compartmentalized. Hydroxyurea solutions can develop toxic materials which react with DNA (Rosenkranz and Rosenkranz, 1969). In addition, there is evidence that very high concentrations of hydroxyurea inhibit UV-induced DNA repair after incubation for extended periods (Francis et al., 1979). Concentrations of 2 mM and below are usually considered safe (Francis et al., 1979; Snyder, 1984a), and there does not seem to be an effect on base excision repair (Katz et al., 1983). There have also been reports of direct interaction between hydroxyurea and some mutagens, for example, MNNG (Irwin and Strauss, 1980).Even when not itself inhibitory to repair synthesis, hydroxyurea can potentiate the inhibitory effect of other compounds. For example, the chain terminator cytosine arabinoside (AraC) will induce breaks in DNA undergoing repair only in the presence of hydroxyurea (Dunn and Regan, 1979). However, hydroxyurea properly used remains a most useful tool for the study of repair processes. AraC is a chain terminator for E. coli pol I (Cozzarelli, 1977), but appears to be incorporated into phosphodiester bonds by mammalian

TABLE I1 COMMON INHIBITORS OF DNA SYNTHESIS USEDIN THE MEASUREMENTOF EXCISION REPAIR Mode of action

Comments

Hydroxyurea (HU)

Inhibits ribonocleotide reductase; reduces supply of deoxynucleotides

Arabinosyl cytosine (AraC), arabinosyl adenine

Inhibits DNA pol and DNA synthesis by inhibition of chain elongation

Aphidicolin

DNA polymerase a inhibitor

Dideoxythymidine

DNA polymerase p inhibitor

Novobiocin

DNA topoisomerase inhibitor; inhibits initiation of repair and replication

3-Aminobenzamide

Inhibitor of poly(ADPribose) synthetase

May inhibit UV repair at concentrations over 2 mm; base excision repair not affected; potentiates AraC effects Leads to accumulation of chain breaks in presence of HU at about 50% of sites of incorporation Leads to accumulation of breaks; inhibits repair synthesis by 20-80%; differential effect on log stationary cells Inhibits repair synthesis; 2 mm HU needed to suppress pools; requires phosphorylation High concentrations required; no accumulation of breaks; reported to inhibit UV-induced but not X rayinduced repair Preserves ATP levels at high alkylation doses; stimulates repair synthesis; increases SCE formation

Inhibitor

a

General references: Downes et al. (1983); Cozzarelli (1977).

Referencea Timson (1975); Francis et al. (1979); Katz et d. (1983); Snyder et al. (1981) Cozzarelli (1977); Snyder et al. (1981);Cleaver (1983) Ikegami et al. (1978); Cleaver (1982a); Snyder and Regan (1982a) Cleaver (1983)

Snyder et al. (1982); Cleaver (1983); Collins and Johnson (1979); Mattern and Scudiero (1981) Sims et al. (1983);Sims et al. (1982); Park et al. (1983)

DNA REPAIR

65

cells (see above, Cleaver, 1982a). Dunn and Regan (1979) report that although on removal of the inhibitor and addition of dC the number of DNA single-strand breaks is reduced, indicating the reversal of repair inhibition, the incorporated AraC is not removed. Cleaver ( 1982a, 1983) has studied the nuclease sensitivity of [3H]thymidine incorporated by repair synthesis in the presence of aphidicolin, an inhibitor of polymerase a or of AraC. He finds that the single-strand breaks induced by incubation with aphidicolin are repaired within 30 min after removal of the inhibitor as compared with the many hours required after the removal of AraC. The L3H1thymidineincorporated by repair synthesis was susceptible to degradation by exo I11 or S1 nuclease, but no more than about half the radioactivity could be removed with the nucleases. Cleaver concludes (1983), “that after growth in inhibitors there are two classes of repaired sites. A large fraction of r3H]thymidine in repaired patches is resistant to exonuclease 111 and S I because it is internally ligated; the other consists of patches with free 3’OH termini that are poorly hydrogen bonded because of the incorporation of base analogs.” Excision repair can be thought of as proceeding via either of two mechanisms. Either an oligonucleotide may be removed as in bacteria (Sancar and Rupp, 1983)and a patch inserted (“cut and patch” mechanism), or the two processes might proceed concurrently by a strand displacement mechanism in which an attack on a site by the incoming strand displaces a parental strand (Ganeson, 1974; Kornberg, 1980).In either case, the excision of lesions should be associated with the appearance of breaks in the DNA during the transition period before ligation of the repair “patch.” However, as compared with alkylating agents, treatments with relatively high doses of UV do not lead to the accumulation of breaks in the DNA (Cleaver et al., 1983b). This implies that either the initial incision step is rate limiting or, as has been suggested, that there is a DNA repair “complex” so that the initial incision is not made unless the rest of the enzyme system is ready to finish the repair event (Hanawalt et at., 1979).It is possible to accumulate repair-related breaks by incubating cells with DNA polymerase inhibitors. Incubation with AraC and hydroxyurea (Dunn and Regan, 1979; Snyder et al., 1981) or with aphidicolin (Snyder and Regan, 1981; Cleaver 1982a, 1983) results in abortive termination of about 50% of the repair events which appear as single-strand breaks (Cleaver, 1982a). The observation indicates that the analog is not an automatic DNA chain terminator in cells, but rather can be stably incorporated within DNA chains. Since the bacterial UV excision endonuclease system makes two

66

BERNARD S. STRAUSS

nicks about 12 nucleotides apart (Sancar and Rupp, 1983), it seems clear that in bacteria, at least part of the mechanism is “cut and patch” rather than strand displacement. The mechanism of excision repair in mammalian cells is not known. A priori, a strand displacement model implies that pol should be the primary repair enzyme, since /3 will act at single nucleotide nicks, whereas in vitro experiments with pol a suggest the need for single-stranded gaps of over 30 nucleotides (see Kornberg, 1980), and recent experiments show that a can replicate a whole M13 template starting from a primer (Grosse and Krauss, 1984). Cleaver suggests that pol a is involved in a “cut and patch” mechanism based largely on the finding that isolated DNA from aphidicolintreated cells cannot be ligated with T4 ligase, as would be expected if a “nick translation mechanism” operated. Clarkson et ul. (1983) showed that “excision of the (UV) lesions proceeds considerably faster than resynthesis” (Le., repair synthesis). These authors suggest that excision occurs early with resynthesis occurring afterward, a conclusion supported by Cleaver’s (1983) affirmation of the “cut and patch” model. The suggestion (Cleaver, 1983) that polymerase /3 is best able to repair the single base removal characteristic of base excision repair is reasonable but not definitive.

IV. Effects of the Local Environment on Reaction and Repair

Repair synthesis starts immediately upon the introduction of UVinduced damage (Ehmann et al., 1978),as does the removal of dimers as measured immun.dogically (Clarkson et al., 1983), implying that the DNA is under constant surveillance by the different repair enzymes. It is instructive to consider how many scanned “units” are present in the DNA in order to get an idea of the number of repair enzyme molecules that must be present. The rate of repair synthesis in HeLa cells saturates at about 30 J/m2 (Edenberg and Hanawalt, 1973),and this dose corresponds to the conversion of 0.18%of all Ts to dimers (Ehmann et ul., 1978). HeLa contains about 14.4 pg DNA per cell (Fasman, 1976), or 4.3 X mol of P, corresponding to 2.6 x lo7kb. The T content of human cells is 30 mol %, so that at 30 J/m2,1.4 X lo7 Ts are in dimers to give 7 x lo6 dimers per cell. There are therefore 3.6 kb/dimer at the point repair saturates. Cells contain enough enzyme to respond immediately to one lesion per 3.7 kb, or alternatively, damage at any of 7 million sites is immediately recognized by cells. Although it is convenient for some purposes to consider the DNA

DNA REPAIR

67

structure as uniform, this is clearly not so. DNA is organized into particular sequences making up the genes. Some special sequences are repeated many times. In primates, one particular sequence of 12 base pairs, the a sequence, accounts for about 1% of &e DNA and is present in tandem arrays in about 300,000 copies per (haploid) genome (Wu and Manuelidis, 1980; Lippke and Haseltine, 1981). Of interest to individuals studying the breakage and repair of chromosomes are the special repetitive sequences found at telomeres (Blackbum and Szostak, 1984). Recent studies have shown that the DNA helix is not always in the j3 configuration. Rather, its structure can switch to the “Z” form, particularly at runs of GC (Rich et al., 1984). Reaction of Gs with acetylaminofluorene seems to promote such transformation (Santella et al., 1981; Singer and Grunberger, 1983). In addition to sequence-imposed heterogeneity, there is a hierarchy of associations of DNA with protein. Specific nonhistone proteins bind to DNA, and it is assumed that some of this binding is part of a control system similar to the mechanisms observed in the prokaryotes (e.g., Hochschild et al., 1983). It is known that repressor proteins interact with the DNA in very specific ways-so specific that the interaction can be deduced from the change in the pattern of methylation observed with dimethylsulfate (Humayan et al., 1977). One can ask whether the association of nucleic acids with nonhistone protein or with chromatin (see below) protects it from reaction with mutagen (McGhee and Felsenfeld, 1979) and whether the location of a reacted site determines its susceptibility to repair. The DNA of eukaryotic cells occurs in organelles and in the nucleus. DNA repair (excision of pyrimidine dimers) does not occur in the major organelle, the mitochondrion, in yeast (Prakash, 1975; Magana-Schwencke et ul., 1982), or in mammalian cells (Clayton et ul., 1974). Most of the nuclear DNA in eukaryotic cells is organized into nucleosomes in which about 200 base pairs of DNA are wound around a complex of histone proteins in an organized manner (McGhee and Felsenfeld, 1980). Connecting the nucleosomes are linker regions of about 60 base pairs. A number of investigations have focused on comparisons of reaction and repair in nucleosomal and linker regions of the DNA. The major experimental tool for such studies is hydrolysis with staphylococcal nuclease. All “native” nucleosomal DNA is nuclease resistant, but it is not necessary that a11 nuclease-sensitive structures must be in linker DNA, since it is possible to partially or temporarily unwind nucleosomes, making portions of the DNA nuclease sensitive (Zolan et aZ., 1982b). One of the major experimental questions, therefore, is whether the transient micrococcal nuclease

68

BERNARD S. STRAUSS

sensitivity of nucleotides inserted by repair (Smerdon and Lieberman, 1978; Bodell and Cleaver, 1981; Oleson et al., 1979) is due to the temporary unwinding of DNA from the nucleosome during the repair process (Oleson et al., 1979), making it temporarily nuclease sensitive, or whether repair occurs first in the linker regions, which are then moved into nuclease-resistant nucleosomal structures. Thus, alternative explanations exist for the relative nuclease sensitivity of newly incorporated repair patches (Williams and Friedberg, 1979). Many but not all agents react differentially with nucleosomal and linker DNA. UV-induced pyrimidine dimers are found in equal frequencies in nucleosomal and linker DNA (Williams and Friedberg, 1979; Niggli and Cerutti, 1982). Many bulky chemical mutagens react preferentially with linker regions. Such compounds include N-acetoxy AAF (Kaneko and Cerutti, 1980), benzo[a]pyrene diolepoxide (Jahn and Litman, 1979),aflatoxin (Bailey et al., 1980),but not some of the coumarins, e.g., psoralen (Zolan et al., 1982a). Excision repair patches are initially nuclease sensitive, but with time become nuclease resistant (see above). It was initially supposed that the observations indicated a “sliding” of nucleosomes along the DNA strand so that patches were initially made in linker DNA and then incorporated into the nucleosomal structures (Smerdon and Lieberman, 1978).An alternative interpretation is based on the observation that even at the shortest labeling times, only about 70% of the repair label is sensitive to nuclease. It is argued that since this same pattern is seen during replicative incorporation of isotope, and since in uitro “sliding” occurs only at elevated salt conditions, repair is more likely to be associated with a transient change in the conformation of the nucleosome (Bodell and Cleaver, 1981; see also Sidik and Smerdon, 1984).While it is likely that this explanation is at least partly correct, there is also evidence suggesting that “nucleosome cores do not obligatorily reform at their original positions after DNA is repaired” (Zolan et al., 1982b). These last experiments take into account the changes in nuclease sensitivity expected as a result of replication (Smerdon and Lieberman, 1978). The question is not yet settled. Some changes in conformation or accessibility of nucleosomes must occur in uiuo. Methylation of DNA by dimethyl sulfate or with methyl nitrosourea results in a uniform distribution of N-7 guanine or N-3 adenine reaction product in nucleosomes and linker (McGhee and Felsenfeld, 1979). The 3-methyladenine, which is a major methylation product, is susceptible to only partial removal from nucleosomes when treatment is in uitro with either bacterial or a mammalian 3-

DNA REPAIR

69

methyladenine-DNA-N-glycosylase(Price et al., 1983).Nonetheless, the rate of removal of this adduct in uiuo is very rapid (Sklar and Strauss, 1981), indicating some very fast change in conformation to allow access of the repair enzymes. In a series of experiments, Zolan et al. (1982a,b) have studied the repair of damage in the a sequences of DNA after reaction with either furocoumarins, N-acetoxy AAF, or UV. Repair with UV occurs to an equal extent throughout the genome. With one of the coumarins [4’(aminomethyl)-4,5,8-trimethylpsoralen], repair in the LY sequence was only about 30% as great as in other portions of the genome. With acetyl A F adducts, repair was about 60% that in bulk DNA. More recent work (Zolan et al., 1984)has eliminated less interesting explanations such as differential reactivity to form special products. These workers conclude that some special interaction between these sequences and the chromatin results in an alteration of structure, making access by the repair enzymes more difficult. It has been a common belief that not all UV-induced dimers are removed from the DNA (see Ehmann et al., 1978).It is certainly true that many repair experiments have shown a fraction of the dimers remaining even after long periods of incubation. On the other hand, many such experiments have been carried out at high dose (high dose means a dose in which there is very low survival of the treated cells), and at such doses the meaning of a nonexcised lesion is not clear. There is a problem in measurements at low doses-most of the experiments were done before the availability of immunochemical analysis and resort was made to M . luteus nuclease sensitivity as a test for the presence of dimers. At low dose it appears that almost all dimers can be removed. However, what is true for pyrimidine dimers is not necessarily true for other lesions. Both N-acetoxy AAF and benzopyrene lesions are only partially removed from cells (Kaneko and Cerutti, 1980).In the case of AAF, reaction was initially higher in the linker regions. Only about half of the adducts were removed at an initial level of reactivity, which permitted a rapid increase in the number of cells. Adducts were preferentially excised from the linker region. That there should be a difference between UV and chemical lesions is peculiar, since the evidence from bacteria and from human (xeroderma) cells is that the same metabolic system repairs both types of lesion (Brown et al., 1979).However, some, but not all, reports indicate that UV and chemically induced lesions produce additive repair effects, indicating that there may be some difference in the repair pathways (e.g., Ahmed and Setlow, 1979).

70

BERNARD S. STRAUSS

V. Poly(ADP-Ribose)

Treatment of cells with many different genotoxic agents results in single- or double-strand breaks in the DNA. Ionizing radiation is the most obvious example. The alkylating agents provide a steady source of DNA, with breaks as a result of the constant glycosylase-induced and spontaneous depurination and depyrimidination that follows treatment with these agents. Such breaks serve as a signal that activates the enzyme poly(ADP-ribose) synthetase (Benjamin and Gill, 1980a). This enzyme catalyzes the reaction of NAD with acceptor protein to form ADP-ribosylated protein (Hayaishi and Ueda, 1977). Further addition of ADP-ribose to the established framework can create long chains of poly(ADP-ribose) attached to a protein receptor which is often the H2 histone protein (Benjamin and Gill, 1980b).The lengths of the chains may vary: Proteins may be “normally” ADP ribosylated with but single residues, but after treatment at high doses of a break-inducing agent, relatively long chains of poly(ADP-ribose) may result (Benjamin and Gill, 1980b; Surowy and Berger, 1983). Turnover of poly(ADP-ribose) is very rapid, both in uitro (Benjamin and Gill, 1980b) and in uiuo, where the polymer has been calculated to have a half-life of -6 min (Jacobson et al., 1983). Poly(ADP-ribose) synthetase has been isolated (Jump and Smulson, 1980; Slattery e t al., 1983).The protein from HeLa nuclei has a molecular weight of 112,000, an absolute requirement for DNA and histones, a K , for NAD of 46 p M , and a V,, of 1470 nmol/(min mg protein). The purified enzyme ADP-ribosylates histones only when reconstituted with oligonucleosomes. Binding of enzyme appears to occur in the linker regions, and the enzyme itself is a major (ADPribose) acceptor (Jump et al., 1980). Poly(ADP-ribose) synthetase has the interesting property of inhibiting the random DNase I-induced transcription of DNA by RNA polymerase 11, probably because of its ability to bind to nicked DNA (Slattery et al., 1983). Activation of the enzyme requires DNA with free ends. Supercoiled DNA will not activate. Nicked DNA is not as efficient as DNA with double-strand breaks (Benjamin and Gill, 1980b). ADP-ribosylation has important effects on protein activity, but these effects may be either stimulatory or inhibitory depending on the protein. DNA ligase activity is inhibited by histones, but this inhibition is reversed by the binding of poly(ADP-ribose) to the poly(ADPribose) synthetase (Ohashi e t al., 1983). An example of the complex result of ADP-ribosylation is the observation of the inhibition of a

DNA REPAIR

71

DNA topoisomerase copurifying with poly(ADP-ribose) synthetase when extensive poly(ADP-ribosylation) is allowed to occur (Ferno et al., 1983). One of the major difficulties in the study of poly(ADP-ribose) function is its determination. Conversion of radioactive NAD to an acidinsoluble form has been widely used (e.g., Berger and Sikorski, 1981; Berger et al., 1980), but this methodology is only available for in vitro systems, including permeabilized cells. Induction of massive poly(ADP-ribose) synthesis results in lowering of cellular NAD concentration, and enzymological analysis of NAD levels has been used to indicate poly(ADP-ribose) synthesis (Jacobson et al., 1980; Skidmore et al., 1979; McCuny and Jacobson, 1981; Jacobson et al., 1983). However, the doses required to induce this removal of NAD are relatively high. A direct method for poly(ADP-ribose) has been developed involving absorption of nuclease-digested material to dihydroxylborylSepharose columns, and specific elution and degradation to a fluorescing derivative, followed by HPLC separation (Sims et al., 1980). The evidence for a role of poly(ADP-ribose) synthesis in DNA excision repair reactions is first the link between the induction of DNA damage (single-strand breaks) and the induction of polymer synthesis. Inhibitors of poly(ADP-ribose) synthesis affect break rejoining and survival (Durkacz et al., 1980; James and Lehman, 1982). A wide variety of synthetase inhibitors is available, including some commonly used substances such as thymidine and caffeine (Sims et al., 1982). The most efficient class of poly(ADP-ribose) synthetase inhibitors includes benzamide and 3-aminobenzamide, a more water-soluble derivative (Purnell and Whish, 1980). Inhibition of poly(ADPribose) synthesis potentiates the killing activity of dimethyl sulfate (Durkacz et al., 1980) and MNNG (Nduka et al., 1980),but the effect is not great and corresponds to a dose reduction factor of two or three. The inhibitors prolong the time required for the rejoining of singlestrand breaks after treatment with alkylating agents (James and Lehmann, 1982; Morgan and Cleaver, 1983), and Creissen and Shall (1982) have suggested that poly(ADP-ribose) is required for efficient ligase action. Their suggestion that the compound regulates ligation in general is supported by the observations that the poly(ADP-ribose) inhibitors induce or enhance sister chromatid exchanges (Oikawa et al., 1980; Park et al., 1983). Lower doses of damaging agents have no effect on excision repair synthesis in the presence of poly(ADP-ribose) synthetase inhibitors (Durkacz et al., 1981; James and Lehmann, 1982). At higher doses, particularly of alkylating agents, an

72

BERNARD S . STRAUSS

increase in the measured excision repair synthesis is observed (James and Lehmann, 1982; Sims et al., 1982; Cleaver et al., 1983b). This increase in repair synthesis occurs without an increase in the measured removal of methylated bases, and it has been supposed that the inhibition of ligase activity permits larger repair patches to be inserted rather than leading to additional patches as might be expected if the role of the polymer had to do with the ligase step in repair (James and Lehmann, 1982; Cleaver et al., 1983b). No direct demonstration of increased patch size has yet been reported. It has been reported by some (Goodwin et al., 1978; Sims et al., 1983; Strauss, 1984) but not all investigators (e.g., Jacobson et al., 1980) that treatments with high concentrations of dkylating agent not only lowers the NAD content of cells, but also reduces the ATP level. So far, this reduction has been observed only in lymphoid cells. Under these conditions, synthetic processes in general (e.g., RNA and protein synthesis) are inhibited (Sims et d., 1983). Since ATP is required for excision repair (Dresler and Lieberman, 1983), repair synthesis is inhibited. Addition of poly(ADP-ribose) synthetase inhibitors blocks the degradation of NAD and of ATP (Sims et al., 1983; Strauss, 1984) (Fig. 3). A simple explanation of the effect of synthetase inhibitors on measured repair synthesis is that by blocking the withdrawa1 of ATP they overcome an inhibition rather than actually stimulating repair. Induction of poly(ADP-ribose) synthesis can be observed after UVirradiation of intact (Jacobson et al., 1983) and permeabilized cells (Berger et al., 1980; Berger and Sikorski, 1981). Human fibroblasts exposed to 5 J/m2 depleted their NAD pool and produced large increases in poly(ADP-ribose). Addition of hydroxyurea or AraC stimulates the accumulation of poly(ADP-ribose), presumably by increasing the lifetime of nonligated breaks (Jacobson et aZ., 1983). In these experiments 3-aminobenzamide failed to suppress the formation of poly(ADP-ribose) in the presence of AraC or hydroxyurea, and 3-aminobenzamide did not affect repair synthesis levels. Permeabilized xeroderma pigmentosum cells failed to respond to UV by producing poly(ADP-ribose), in contrast to the behavior of normal cell preparations (Berger et al., 1980),unless M . Zuteus UV endonuclease was first added (Berger and Sikorski, 1981). The ability of inhibitors of poly(ADP-ribose) synthetase to stimulate UV-induced repair synthesis appears to depend on the presence of inhibitors, such as hydroxyurea, which increase the steady-state level of DNA with broken ends (Cleaver et al., 1983a,b). Massive production of poly(ADP-ribose) and consequent lowering of ATP concentrations may well depend on cell

DNA REPAIR

73

I

T

"

"

I

"

'

"

'

"

'

I

"

'

'

- 10 -20 -30 PPM FIG.3. ATP levels in MMS-treated cells. Nuclear magnetic resonance measurements on intact cells of the human lymphoblastoid line GM606 reveal detectable nucleoside di- and triphosphate levels in control cells (A). MMS-treated (4mM, 60 min, 37°C) cells had no detectable nucleoside di- or triphosphates (B). MMS-treated cells incubated in 3-aminobenzamide had levels of nucleoside di- and triphosphates comparable to controls not treated with MMS (C) (data cited in Strauss, 1984; NMR measurements courtesy of D. Lynn and M. Chang, Department of Chemistry, University of Chicago). 10

0

type with lymphoid cells as used by Goodwin et at?.(1978),Sims et al. (1983),and Strauss (1984) giving more exaggerated responses than the fibroblasts used by Cleaver et al. (1983b)and by Jacobson et al. (1983). The rate of repair, at least in some cells, is determined by an interaction between the steady-state number of strand breaks, poly(ADPribose) formation, and the repair system(s). These factors may account for the observed inactivation of UV-induced excision repair by high MMS concentrations (Cleaver, 1982b; Park et al., 1981). It has been reported that this inhibition of repair synthesis is overcome by the

74

BERNARD S. STRAUSS

addition of 3-aminobenzamide (Strauss, 1984). Since the experiments were done in cells whose ATP level was depleted by treatment with alkylating agents (Fig. 3), it is possible that the inhibition is indirect and due to the MMS-induced depletion of ATP via poly(ADP-ribose) formation. These interactions occur at such high concentrations, e.g., at such extensive cell killing, that the role of ATP levels in repair at physiological conditions remains in doubt. It is possible that the polymer plays some role in regulating metabolism by control over ATP levels. More likely it is the regulation of ligase (Lehmann and Broughton, 1984)or of topoisomerase (see above) which is important. The discovery of poly(ADP-ribose) formation in yeast (Sugimura and Miwa, 1983)is important, since it opens up the possibility of genetic analysis of the system. VI. The Repair of 0-Alkylated Sites

Treatment of cells or of DNA with alkylating agents of the Sn-1 class (e.g., methyl nitrosourea, NMU) results in the production of a higher proportion of alkylated oxygens as compared with the alkylated N sites common with Sn-2 agents such as MMS or dimethyl sulfate (Singer and Grunberger, 1983; Lawley and Thatcher, 1970). The repair of 0-alkylated sites was ignored for many years (in part because of difficulties in analysis), but has recently been the study of intense investigation. Recent discoveries have cleared up some of the earlier confusing details while adding new problems. A key discovery was made by Samson and Cairns (1977), who showed that treatment of E. cob at low concentrations of MNNG induced a previously unknown repair pathway. Induction is rapid, quantitatively important, and is blocked by the protein synthesis inhibitor chloramphenicol. Addition of chloramphenicol along with MNNG greatly increases the yield of E. coli mutations (Sklar, 1978). Shortly after this discovery, the biochemical nature of the repair was worked out by Karran et al. (1979), who showed that induction involves the formation of a protein which removes the 06-methylguanine adduct. Later studies showed the protein to be a methyl acceptor which transfers methyl groups from 06methylguanine to one of its own cysteines (Olsson and Lindahl, 1980). The transfer inactivates the acceptor protein so that repair is stoichiometric; one molecule of protein is required for each guanine regenerated (Lindahl, 1982; Demple et al., 1982).The biology of the system is even more complicated because in bacteria the 06-methylguanine lesion is not the lethal adduct (Jeggo et al., 1978). Rather, induction

DNA REPAIR

75

involves several gene products, including a 3-methyladenine glycosylase (Karran et al., 1982). Bacteria unable to remove 3-methyladenine are very sensitive to the lethal effects of MMS (Evensen and Seeberg, 1982), an agent which produces about 10% of its products at this site (see Singer and Grunberger, 1983). There is now preliminary evidence that separate proteins may exist which remove O-alkylation products of thymine and cytosine (Ahmed and Laval, 1984; McCarthy et al., 1983, 1984). A distinct inducible methyltransferase repairs phosphotriesters produced by MNNG treatment of E. coli, and it has been suggested that the phosphotriesters may be one of the lethal alkylation lesions (McCarthey et al., 1983). Mutants of the inducible process, the ada genes are known (Jeggo, 1979)and have been cloned (Sedgwick, 1983). The response has been termed the “adaptive response” to distinguish it from inducible DNA repair under the control of Zex and rec genes (see Walker, 1984). Mammalian cells contain a protein similar in size (21,000-22,000 Da as opposed to 18,000-18,400 in bacteria) and kinetics to the protein in bacteria (Harris et al., 1983).Although not completely purified, the protein has been obtained from mouse (Bogden et al., 1981), from rat liver (Pegg et aZ., 1983), from human lymphoblastoid cells (Harris et al., 1983), and from human liver (Pegg et al., 1982) and placenta (Yarosh et al., 1984a). Part of the difficulty in understanding the removal process was and remains the cumbersome analysis for 06methylguanine and other minor O-alkylation products. This problem has been at least partly resolved by new chemical techniques (e.g., Reddy et al., 1984), sensitive HPLC methods for 06-methylguanine (Lawley and Warren, 1981), as well as by the development of antibodies specific for the 06-methylguanine adduct (e.g., Wild et al., 1983; and see Table I). In addition, knowledge of the repair mechanism has permitted development of a simple assay based on the transfer of the methyl group from nucleic acid to protein and dependent on the ability of proteolytic enzymes (Waldstein et al., 198%)to solubilize radioactivity bound to protein or on the resistance to acid hydrolysis of radioactivity bound to protein but not to DNA (Myrnes et al., 1984). Using these methods along with more traditional ones, it is possible to ask the following: (1)What is the distribution of acceptor activity in the different cells and organs? (2) Is reaction at the O6 position of guanine the only site critical for mutagenesis and carcinogenesis? (3) What is the lethal lesion in mammalian cells? (4) How is the level of acceptor protein regulated? (5)Is there an inducible process in mammalian cells paralleling the adaptive process in bacteria? Some of these questions arise from the results of in vivo whole-

animal studies (see Pegg 1977, 1983). It was observed by Goth and Rajewsky (1974) that treatment of rats with ethyl nitrosourea results in the development of brain but not of liver tumors. Examination of the repair of 7-N-alkyl and 6-0-alkyl lesions showed that although there was no differential effect for the 7-alkyl lesion, brain was relatively deficient as compared to liver in the removal of 06-ethylguanine. Since brain but not liver tumors developed after the regimen was used, this experiment was used as an argument demonstrating the key role of the O6 position in carcinogenesis. There is no doubt that the levels of acceptor protein do vary from organ to organ (see Pegg, 1983) and probably from individual to individual (Krokan et al., 1983; Myrnes et al., 1983),but the interpretation of such data is not simple. Every organ is made up of an assemblage of cells of different types, and the behavior of these cells may be rather different. It has been possible to separate the cell types from liver, for example, and to show that there is acceptor activity in hepatocytes but not in the nonparenchymal cells (Swenberg et al., 1982).The situation is even more complex since the nonparenchymal cells can be separated into Kupfer cells and sinusoidal epithelial cells. The Kupfer cells may remove some 06-methylguanine, but this is also accompanied by higher rates of replication (Lewis and Swenberg, 1983). It has been suggested that the factors of removal and replication interact, with replication while lesions are still present being a key in carcinogenesis (Lewis and Swenberg, 1980; Swenberg et al., 1982; Bedell et al., 1982). The acceptor protein has a restricted specificity. O6-EthyI derivatives are removed, but with lower efficiency (Sklar et al., 1981). The acceptor also seems to react with the 06-mono adduct stage of certain cross-linking agents, although it will not disrupt the cross-links themselves (Kohn et al., 1981).In fact, pretreatment of cells with MNNG to exhaust the supply of acceptor protein potentiates the ability of chloroethyl nitrosourea to form cross-links (Zlotogorski and Erickson, 1983,1984). The acceptor protein in mammals appears to be restricted in activity to the removal of 06-guanine alkylations, since there is an accumulation of the 04-thymidinederivative in liver under conditions where no accumulation of 06-alkylguanine occurs (Muller and Rajewsky, 1983; Swenberg et al., 1984). In this respect, the protein appears to differ from the bacterial protein which has been shown to accept both 06-guanine and 04-thymine methyl groups (McCarthy et al., 1984). The content of acceptor protein in human cells is related in some illdefined way to viral and malignant transformation. Some years ago it was reported that cells derived from an individual with xeroderma

DNA REPAIR

77

pigmentosum were deficient in their ability to remove 06-methylguanine from DNA (Goth-Goldstein, 1977; Altamirano et al., 1979). Although the specific observations were correct, it turns out that there is no relationship between the xeroderma characteristic and the presence of acceptor protein (Sklar and Strauss, 1981; Heddle and Arlett, 1980), but rather that the cells used were transformed either by SV40 or by Epstein-Barr virus. In some way transformation is often (but not always) associated with a deficiency of methyl acceptor protein (Day et al., 1980b; Yarosh et al., 1983; Harris et al., 1983). Cells with relatively “normal” content of acceptor protein have been called Mer+ (methyl repair +; Day and Ziolkowski, 1979, 1981) or Mex+ (methyl excision +; Sklar and Strauss, 1981). Deficient cells are Mer- or Mex-. About 20-30% of tumor lines are Mer- (Day et al., 1980a), and this ratio is about the same for EBV-transformed human lymphoblastoid lines and for SV40-transformed fibroblasts (Day et al., 1980b). On the other hand, tissues freshly obtained from human tumors uniformly contain significant amounts of 06-methylguanine acceptor protein, indicating that the deficient phenotype is not directly associated with the initial tumorigenic event (Wiestler et al., 1984). The difference between Mex+ and Mex- lines is likely epigenetic (Shiloh et al., 1983a; Sklar and Strauss, 1983).All normal human fibroblast lines reported have been Mex+ (see Shiloh et al., 1983a). Mex+ and Mex- cell lines can be obtained from the same male individual, indicating that an X-inactivation phenomenon is not involved (Sklar and Strauss, 1983). Crosses of Mex+ X Mex- lines indicate that the Mex- characteristic is recessive; the hybrid cells produce just as much acceptor protein (per cell) as do the Mex+ parents, indicating that no diffusible regulatory inhibitor protein is involved (Ayres et al., 1982). The data fit the idea that some nongenetic chromosomally autonomous change is involved. The experiments described above were carried out with human lymphoblastoid cells. Recent hybridization experiments with human tumor lines indicate a more complicated interaction between different strains (Yarosh et al., 198413). The loss of acceptor activity need not be absolute; many of the Mex- strains have some 06-methylguanine acceptor activity. One group of investigators has attempted to make sense of the situation by defining a category of intermediate cells that are “ T e r n + , mer-” (Scudiero et al., 1984a).The separation is partially based on the distinction between the susceptibility of cells to inactivation by M N N G and the capacity of cells for host-cell reactivation of adenovirus. The terminology refers to phenotype; no genetic separation of the characteristics has yet been reported. In most cases there is a correlation between

78

BERNARD S. STRAUSS

loss of methyl acceptor activity, or of 06-methylguanine removal activity, which is assumed to be the same, and the toxicity of MMNG, but this correlation is not absolute. Lines with considerable ability to remove O6-methy1guaninemay be as sensitive to MNNG as lines with much less activity, leading to the suggestion that 06-methylguanine may not be the lethal lesion (Ayres et al., 1982).The experiments have been done by studying both cell viability and the inactivation of adenovirus assayed on Mer+ and Mer- cells (e.g., Day and Ziolkowski, 1979). It may be that a lesion lethal for adenovirus is not lethal for cells; for example, lethal mutation is responsible for the decrease in viral viability. Alternatively, it is not excluded that the phosphotriester lesion may be involved in cell killing, as suggested by McCarthy et al. (1983). VII. Adaptive Response

In E. coli the adaptive response to MNNG treatment results in an increase in enzyme activity of over two logs (Lindahl, 1982). There is low but significant 06-methylguanine acceptor protein in noninduced cells (Mitra et al., 1982). Comparison of induction in bacteria with mammalian cells is therefore difficult, since most (all?) responses in mammals are of much smaller magnitude. An extensive series of observations indicates that treatment of rats with dimethylnitrosamine increases the 06-methylguanine removal activity in the liver (e.g., Montesano et al., 1983). With care, induction can be demonstrated in rat kidney as well (Pegg and Wiest, 1983). The increases are at best equivalent to a 3-fold increase in the amount of acceptor protein. For reasons that are not yet explained, this response to treatment with methylating agents is observed only in the rat (O’Connor et al., 1982). Furthermore, the increase can also be obtained by partial hepatectomy (Pegg et al., 1981),by hepatotoxins such as thioacetamide (Pegg and Perry, 1981), by toxic agents such as acetylaminofluorene (Charlesworth et al., 1981), and by hormones (Pegg et al., 1978; Pegg and Wiest, 1983).Although these agents do lead to cell proliferation, Pegg and Wiest (1983) conclude there is no “obligatory coupling” between the methyltransferase activity and growth rate. The effect of nonmethylating agents in increasing the level of methyl acceptor protein is particularly mysterious. The suggestion has been made that perhaps such treatments enhance endogenous methylation via Sadenosylmethionine, since it has been observed that compounds such as phenylhydrazine, which are not themselves methylating agents, can nonetheless induce the formation of methylated bases in whole

DNA REPAIR

79

animals (Becker et al., 1981). It has not been demonstrated that the nonspecific treatments do indeed lead to endogenous methylation. In addition, there is no information indicating why the phenomenon should be restricted to the rat. A significant number of studies report adaptive effects of treatment with methylating agents in mammalian cells (Laval and Laval, 1984; Waldstein et al., 1982a,b; Samson and Schwartz, 1980; Schwartz and Samson, 1983),but there is still no agreement on the generality, mechanism, or meaning of such observations (Karran et al., 1982; Foote and Mitra, 1984). The observations are also somewhat complicated by an uncertainty about the distinction between mutagenic and cytotoxic lesions and by the question of whether lesions inducing sister chromatid exchange (SCEs, see below) are the same as those inducing mutation. A simplifying assumption is that 06-methylguanine and possibly other O-alkylation sites are mutagenic, but that some other lesion(s) are cytotoxic (see Laval and Laval, 1984). In the case of the methylating agents, there is no need to succumb to the simplistic notion that only a single lesion is mutagenic or cytotoxic. Demonstration of an adaptive response (n mammalian cells appears to require rigorous adherence to a fixed protocol using low doses of MNNG or other stimulating agent. Samson and Schwartz (1980) showed that Chinese hamster cells responded with an adaptive response that protected against cytotoxicity and SCEs, but not mutation. It has been separately reported that Chinese hamster cells are unable to remove 06-methylguanine (Goth-Goldstein, 1980). In a series of investigations, Waldstein (1982a,b) reported that a variety of cells could be induced by a precise protocol, the induction being observed within 60 min and complete within 4 hr. The claim was made (Waldstein et al., 1982b) that Mer- cells had the same level of endogenous acceptor protein as did Mer+, but that the cell lines differed in their ability to regenerate protein via this inducible response. This claim has been disputed by at least three laboratories (Foote and Mitra, 1984; Yarosh et al., 1984a; Karran et al., 1982) who were unable to observe induction. A recent study (Laval and Laval, 1984) reports the development of an adaptive response in rat hepatoma cells 4 hr after a protocol involving nine pretreatments with MNNG given over a 48-hr period. Treatment with either MNNG or MMS protected against the cytotoxic effects of either alkylating agent. Only pretreatment with MNNG protected against the mutagenic effects of MNNG, and such pretreatment produced a 3-fold increase in the rate of 06-methylguanine removal. The data indicate a separation of mutagenic and cytotoxic lesions and an inducible response in these particular cells.

80

BERNARD S. STRAUSS

A simple view of the confusion in the literature is that failure to demonstrate the adaptive response is a result of slight deviations from a demanding protocol. However, the effects are at best 3-fold. Variations of this magnitude are observed in other enzymes throughout the cell cycle (Gupta and Sirover, 1980; Gombar et al., 1981). Although many of the investigations go to some lengths to exclude cycle variations, it is not obvious that the adaptive response observed is based on the same mechanism operative in bacteria. VIII. Bypass of Lesions and Its Consequence

For the most part, the removal of lesions by excision repair or by the action of an acceptor protein is without long-lasting consequence to the cell. As far as we can tell, the replication systems involved in the insertion of patches into DNA in non-S phase synthesis are not in any way more error prone than their normal replication analogs. Only in the case of excision repair carried out by cells with a mutant polymerase a does there appear to be an increase in mutation (Liu et al., 1984).The major biological consequence of the treatment of cells with mutagenic-carcinogenic agents appears to be the result of replication while the lesion remains in the DNA (Swenberg et al., 1982; Maher et al., 1979).A possible major factor in the induction of mutation is therefore the induction of replicative synthesis seen after treatment of quiescent cells with carcinogens-mutagens. Treatment may either induce amplification of inserted (viral) sequences (Lavi, 1981)or induce general cellular DNA replication (Cohn et al., 1984). Mutagenic lesions themselves can be classified as of two sorts. The first is instructive and directs Watson-Crick base pairing, albeit of the wrong sort. An example of such an instructive lesion is the presence of 2-aminopurine in the DNA or of 06-methylguanine produced by treatment ofcells with an Sn-1 alkylating agent (see above), at least at some sites (Toorchin and Topal, 1983; Eadie et al., 1984). Such lesions result in misreading and produce a huge excess of transition mutations when in the DNA (Coulondre and Miller, 1977; Strauss et al., 1982).A second class of lesion destroys the ability of the reacted nucleotide to base pair normally. Such noninstructive or pseudoinformational (Walker, 1984) lesions block DNA synthesis (Painter, 1981; Strauss et al., 1982). However, since the presence of such lesions is also mutagenic and carcinogenic, it is apparent that in some cases and in some cells the DNA replicative machinery must be able to overcome or bypass the lesion. This bypass can be detected experimentally. Lesions that are blocks to DNA synthesis in vitro (see above) can be

DNA REPAIR

81

detected in limited numbers in newly replicated DNA in vivo (e.g., Kaneko and Cerutti, 1980; Stacks et al., 1983). The question is how such replication occurs. Furthermore, in bacteria, replication past lesions is sometimes (but not always, see below) associated with mutation (Walker, 1984). Some bypass is therefore associated with what has been loosely called “error-prone synthesis.” The existence of bypass in the broad sense is indicated by the findings that limited numbers of lesions which block DNA synthesis in oitro can be found in progeny DNA (Cleaver, 1970; Meneghini et al., 1981). Such findings alone do not demonstrate translesion synthesis, since there are well-documented error-free mechanisms in bacteria by which sections of DNA of the proper polarity are spliced into gaps starting opposite the lesion by recombinational (recA dependent) processes (see Hanawalt et al., 1979). Recombinational repair occurs at most to only a limited extent in mammalian cells (Meneghini et al., 1981).Transfer of UV-induced lesions (pyrimidine dimers) to daughter DNA in the absence of recombination has been best demonstrated in SV40 virus-infected cells in which the presence of fully replicated viral molecules carrying lesions can be demonstrated (Stacks et al., 1983). Excision repair occurs in the absence of replication. Recombination repair in bacteria occurs after replication starting at distal initiation sites has passed over lesions to leave gaps which are filled in by recombination. This is “postreplication” repair as defined in bacteria (see Hanawalt et al., 1979). The results of studies on postreplication repair with mammalian cells can be briefly summarized as follows (extensive reviews are to be found in Meneghini et al., 1981; Meneghini, 1981): Addition of [3H]thymidine to UV-treated cells for short periods followed by lysis of the cells and alkaline sucrose sedimentation results in the appearance of label (newly synthesized DNA) in smaller pieces than that synthesized by control cells. When chased with nonlabeled thymidine, these pieces eventually increase in size until the size distribution is equivalent to the control pieces. Part of the experimental problem is the fact that mammalian DNA is so large that it is impossible to study “whole” DNA by readily available techniques. Many investigators therefore pretreat with X rays immediately before lysis of the cells on alkaline gradients to break up the DNA into manageable pieces (e.g., Cleaver et al., 1983~). This problem of size accounts in part for the great popularity of the “alkaline elution” technique (Kohn et al., 1981) in which rare numbers of breaks are recognized by their ability to release DNA from a meshwork formed by lysis of cells on a nitrocellulose filter.

82

BERNARD S. STRAUSS

The time required for newly synthesized pieces from damaged cells to attain control size is a function of the dose of mutagen-carcinogen. Studies with cells from the xeroderma variant lines show retarded ability of these pieces to attain the size of DNA from untreated or control, normal cells, indicating a deficiency in their ability to bypass When treated cells are incubated for damage (Cleaver et aE., 1983~). several hours following treatment and then given a pulse of radioactive thymidine, DNA of normal size is made, but total DNA synthesis remains inhibited as compared with untreated controls (Cleaver et al., 1983~). These observations have been interpreted as indicating that DNA synthesis is blocked on one strand when it reaches a lesion, but that synthesis continues on the other strand. At some point synthesis reinitiates on the blocked strand, leaving a single strand gap (for which there is good evidence in bacteria). Then a process of “postreplication repair” fills the gap and in some way adds a base opposite the lesion (see Doniger, 1978). Studies with SV40 confirm the hypothesis that any pyrimidine dimer prevents rather than “slows” DNA replication (Barnett et al., 1984). An alternative explanation is based on the special nature of DNA replication in eukaryotic cells (See Edenberg, 1983).The mammalian genome is so large and the rate of DNA chain elongation so slow (Kapp and Painter, 1982) that replication within the 8 hr or so of S period requires numerous independent units or replicons. Replication is bidirectional. Groups of replicons seem to be activated together. The initiation sequences have not been worked out, but there must be primary and secondary initiation sites, since the size of the replicons differs. For example, early in Drosophila development the entire complement of DNA can replicate in about 3 min compared to a much longer time in more mature cells, indicating an increase in the number of replicons (Rabinowitz, 1941; Kriegstein and Hogness, 1974). Human cells in culture have a replicon size of about 20 pm, rodent cells in general have an average replicon size of about 40-45 It is not certain whether there are fixed pm (Cleaver et al., 1983~). termination sites or whether synthesis continues until the extending replicons meet. It can also be surmised that some particular topoisomerase function must be involved at the joining of replicating units in order to maintain the appropriate coiling of the DNA. Cleaver and his co-workers, particularly Park (Park and Cleaver, ) ~ pointed out that there must be a 1979a,b; Cleaver et al., 1 9 8 3 ~have close connection between excision repair and DNA replication. Since noninstructive lesions are blocks to chain elongation, the removal of such lesions by excision repair will permit DNA synthesis to occur.

DNA REPAIR

83

There is also evidence that completion of some replicons is required for the initiation of new ones. Therefore, there is a relationship between the size of the replicons, the rate of excision repair, and the restoration of DNA synthesis. The behavior of xeroderma cells can be understood as due to a lowered rate of excision. Given these factors, it is not at all clear that a separate “postreplication repair” needs to be postulated, since the phenomena so classified can be understood as part of DNA synthesis in which replicons terminate at lesions (Park and Cleaver, 1979a; Strauss et al., 1983) (Fig. 4). Filling in of the one or two nucleotide gaps opposite a lesion could be accomplished by the in vivo analog of the process described in vitro (by Rabkin et al. (1983),aided by the higher dA pool size resulting from mutagen treatment (Das and Loeb, 1983). The “gaps” seen in sucrose gradient experiments can be explained as due to the termination of replicon synthesis at lesions. In a study of the replication of UV-damaged SV40 DNA, White and Dixon (1984) were unable to visualize singlestranded regions by electron microscopy in molecules retained by BND-cellulose (and presumed therefore to have single-stranded regions). Any single-stranded regions are therefore likely to be less than 150 bases in length. White and Dixon (1984) believe these to be replication and not repair intermediates and that they provide evidence that replication repair occurs by gap filling. Regardless of whether long-lived single-strand gaps do appear in DNA, there is a stage in which DNA synthesis must proceed past the lesion. Several models are available, including one in which a strand switching mechanism

:+

I

Bas. oddition 1 oppos~toadduct

I

- -1 FIG.4. Model of “bypass” as a result of termination of replicon progression at the site of lesions. Modified from a model suggested by Park and Cleaver (1979a). Left: One lesion per replicon showing termination before the lesion followed by eventual base addition opposite the lesion stablilized by ligation. Right: The situation with two lesions per replicon showing “frozen” growing points and an apparent gap in between (from Strauss et al., 1983). !k-L_l!!!?!c!i_on__

84

BERNARD S . STRAUSS

coupled with strand displacement and branch migration uses anondamaged strand to copy past the damage (Higgins et al., 1976). IX. Error-Prone Repair and Mutation

The prokaryotic systems provide models for our thinking about the situation in mammalian cells. A major difference between the experimental work with prokaryotic and (higher) eukaryotic systems has been the exploitation of genetics in the analysis of bacterial repair; the study of natural human mutants has not as yet helped clarify the situation. Even the availability of a large number of yeast mutants affecting radiation and chemical sensitivity and recombination (Haynes and Kunz, 1981) has not led to a greater understanding of the molecular biology of that system, notwithstanding the elegant methodology for genetic manipulation in yeast. It may be that the situation in eukaryotic systems is fundamentally different from that in bacteria, and our attempts to apply the details of the E. coli model have actually slowed progress. In E. coli the production of mutations from noninstructive lesions absolutely requires the participation of metabolic systems under gene control. [In addition to the primary sources in the field, the comprehensive reviews by Witkin (1976) and by Walker (1984) should be consulted for overviews of the situation in bacteria. J In the absence of these gene products mutation does not occur, although the cells may show greatly enhanced cytotoxicity after mutagen treatment. The critical gene products are under the control of the two adjacent genes, umuC and umuD, which have been cloned and sequenced (Elledge and Walker, 1983; Walker and Th’ng, 1982). These gene products allow mutations to occur, although the mechanism by which this occurs is not known. It had been suggested that error-prone translesion synthesis might occur as a result of inhibition of the 3’ > 5’ exonuclease activity which is part of prokaryotic DNA polymerase(s) (Villani et al., 1978), but although mutants in this activity are mutators (Echols et al., 1983),the umuC gene is not genetically linked to either the mutator or polymerase genes (Walker, 1984).We therefore do not know how the error-prone system works. It is likely that errors are made in the course of addition of nucleotides opposite the lesion in direct translesion synthesis, but this has not been unequivocally demonstrated in uiuo. A model system predicts the specificity of base substitution mutation based on the observation that polymerases add purines, particularly adenine, opposite noninstructional sites. Agents that react with pyrimidines should therefore lead to transitions, while

DNA REPAIR

85

reaction with purines should lead to transversions (Strauss et al., 1982). The predominance of G:C > T:A transversions resulting in E. coli from mutagenesis with benzo[aIpyrene, aflatoxin, and cis-Pt (Brouwer et al., 1981; Eisenstadt et al., 1982; Foster et al., 1983; Miller, 1983),all of which react primarily with G, lends support to this model and to one proposed by Schaaper et al. (1982), who suppose that many bulky mutagens act via apurinic sites (AP) as intermediates. Addition of A opposite such sites is the predominant reaction in vitro (Sagher and Strauss, 1983) and in vivo (Kunkel, 1984). The umuC gene product is inducible and is part of the “SOS” cascade of reactions which constitute the prokaryotic (i.e., E. coli) response to damage. Treatments that block DNA synthesis, including thymine starvation, activate the product of the recA gene, which in turn promotes the protease action of the lexA gene product (Little, 1984). This 1exA gene product is a repressor which controls the expression of many genes, including the A bacteriophage, the recA protein, the umuCD gene protein, and itself (Little and Mount, 1982). Inactivation of lexA therefore allows expression of the mutagenic function. One of the major indications of an SOS cascade is the occurrence of Weigle reactivation (Weigle, 1953). Irradiated bacteriophage shows increased survival when plated on previously irradiated bacteria which have been incubated after irradiation and before infection for a long enough period to permit synthesis of the proteins needed for the SOS reactions. Infection of unirradiated bacteriophage in such irradiated hosts yields an increased frequency of bacteriophage mutations. Such mutations are untargeted, since they occur without any lesion being present in the virus, presumably as a result of the general induction of the umuCD genes. The production of untargeted mutation as a result of some inducible process is the evidence for the induction of “error-prone” DNA synthesis. Untargeted mutation is to be contrasted with targeted mutation occurring at the site of damage in the DNA of treated cells. Most mutations observed are targeted (Miller, 1982). The untargeted “SOS-dependent” mutations are mostly G:C > T:A transversions leading to the suggestion that spontaneous apurinic-apyrimidinic sites may be the actual targets (Miller and Low, 1984). Viral DNA seems to be the most appropriate substrate for the demonstration of an effect of previous treatment on survival and mutation in mammalian cells (see Defais et al., 1983). Treatment of cells with small doses of mutagen, particularly UV-radiation, has been shown to increase the survival of herpesviruses, of adenoviruses, and of SV40 (see Table I1 in Defais et al., 1.983 for a complete listing). It has been

86

BERNARD S. STRAWS

shown with herpes (Bochstahler, 1981; Zurlo and Yager, 1984), with adenovirus, and with SV40 that pretreatment of cells with UV or other carcinogenic agents increases the survival of UV-treated virus (Defais et al., 1983).A period of time is required between infection and the observation of the effect which is abolished when the protein synthesis inhibitor, cycloheximide, is present. The enhancement of survival is similar for survival of SV40 (Sarasin and Hanawalt, 1978) or herpes (Das Gupta and Summers, 1978),being about 7-fold at a survival of about 1%. Although most of the data are given as ratios of survival in the range of 1%survival, the critical values are the numbers of lethal hits removed which can be obtained from the ratios of the slopes of the (assumed) exponential curves. Relative survivals of about 10 at 1% survival yield differences in lethal hits of about one, which is not much different from that observed with bacteria. In considering the results, it should also be realized that the extent of the SOS repair effect on survival is relatively small. Moore et al. (1982)have tabulated results in the literature indicating that no more than about two lethal hits can be repaired by the SOS system. It was Moore’s contention that the failure to report more relief from killing was due to the production of lethal mutations. This factor of lethality resulting from mutation may account for the finding that 06-methylguanine is apparently a lethal lesion for adenovirus (Day et al., 1984), but not necessarily for cells (Ayers et al., 1982; Yarosh, 1985).Not only survival, but also DNA replication itself is enhanced by pretreatment of cells shortly after infection with SV40 virus (Maga and Dixon, 1984). Treatment of monkey kidney cells with acetoxy acetylaminofluorene (AAAF) or UV prior to or just after SV40 infection led to enhanced viral DNA replication (measured by hybridization). Pretreatment of the host cells has also been shown to increase the mutability of treated SV40. Furthermore, pretreatment with AAAF is reported to result in an increase in mutation frequency of about 10fold, scoring reversion of an SV40 ts mutant (Sarasin et al., 1982).The elimination of a variety of explanations (multiplicity reactivation, host-cell reactivation-see Defais et al., 1983) makes it seem reasonable that a process similar to that of SOS repair does indeed occur in mammalian cells infected with extrachromosomal DNA. Analysis of the mutations produced has led one group to conclude that an inducible process leading to error-prone translesion synthesis can occur in mammalian cells (Sarasin et al., 1982),but there is some dispute as to how the results are to be interpreted (Cornelis et al., 1980; Bochstahler, 1981).Sarkar et al. (1984) found that pretreatment of monkey kidney cells with ethyl methanesulfonate (EMS) prior to transfection

~

DNA REPAIR

87

with a shuttle vector resulted in a 10-fold increase in the frequency of mutations observed in the bacteral supF gene carried by the vector. These authors conclude that an error-prone replication system has been induced. However, in the absence of genetic analysis pretreatment experiments are subject to different interpretations, of which induction is only one. For example, it is known that transfection with shuttle vectors results in high levels of spontaneous mutation in the genes carried by the vector (Lebkowski et al., 1984; Calos et al., 1983; Razzaque et al., 1983,1984). If the prior treatment with EMS in some way activated this system of spontaneous mutagenesis (e.g., by activating lysosomes), an “induced” mutagenesis would be observed, but without the operation of “error-prone replication” as usually understood. Chinese hamster cells of enhanced mutagenicity and sensitivity to toxic agents are known, but these may well be affecting the excision repair pathway (Thompson et al., 1982) or the polymerase in a way which does not involve an inducible effect (Liu et al., 1984). Several different kinds of investigations also make it more likely that mammalian cells have some adaptive response to damage, particularly ionizing radiation. The investigations of Kennedy and Little (Kennedy and Little, 1980; Kennedy et al., 1980, 1984) show that xirradiation of mouse C3H 10T1/2 cells [a partially transformed line still subject to contact inhibition (Reznikoff et al., 1973a,b)] produces some change in all cells which must be followed by a rare event to produce transformation. Schorpp et al. (1984) have just reported that UV treatment of human fibroblasts induces a group of proteins, including an extracellular protein which can induce the UV response in nonirradiated cells. The UV-induced proteins are the same as those induced by the tumor promoter, TPA (Mallick et al., 1982). Ben-Ishai et al. (Miskin and Ben-Ishai, 1981; Ben-Ishai et al., 1984) report that UV induces plasminogen activator activity in human fibroblasts and fetal cells, a result reminiscent of the protease activiation in SOS repair (Little, 1984). A recA protein-like activity, requiring ATP for recombinogenic activity, has been reported in human cells (Kenne and Ljungquist, 1984). There is no need to require inducibility for error production. One could certainly imagine a system that was constitutive and made errors. Notwithstanding this argument, most of the authors (e.g., Sarasin and Hanawalt, 1978; Das Gupta and Summers, 1978) attempting to understand the system in mammalian cells have opted to require the demonstration of a time lag before an effect is noticed and its inhibition when protein synthesis is inhibited with cycloheximide. How-

88

BERNARD S. STRAUSS

ever, eukaryotic cells (often) make use of small-molecular-weight compounds to regulate, whereas bacteria (often) solve regulatory problems by altering protein concentrations. In particular, many of the effects supposed to occur as a result of an SOS induction (e.g., Dinsart et aZ., 1984) could turn out to be due to changes in deoxynucleotide pools “induced” by blocks in DNA synthesis. Such changes have been shown to produce major changes in survival and mutation. Major effects of pool size on mutation and on cytotoxicity have been demonstrated after treatment with agents like MNNG which produce instructional lesions (Eadie et al., 1984). Changes in the dC :dT ratios in the medium result in changes in mutagenicity and cytotoxicity. Mutator mutants have been isolated which change pool sizes, and the mutator effects have been corrected by alterations of the medium (Meuth et al., 1979). The detailed nature of similar mutator mutants has been analyzed by Weinberg et al. (1981),who isolated two mutations affecting the M1 subunit of ribonucleoside diphosphate reductase and one mutation deficient in deoxycytidylate deaminase. These investigators determined mutation rates (as opposed to frequencies) by Luria-Delbruck variance analysis and showed a 30- to 40-fold increase. The pool sizes were shown to be very different from the wild type, and manipuIation of the pools with dC or dT (as has also been done by other investigators) partially restored the mutation rate toward the normal. Not only mutation rate, but also cytotoxicity can be controlled by pool size changes (Meuth, 1981),giving credence to the idea that at least a proportion of cytotoxicity is due to lethal mutation. Treatment with MNNG or UV results in changes in pool size (Das et aZ., 1983). Treatment of Chinese hamster cells with MNNG, mitomycin C, UV, or AraC led to a rapid expansion of the pools of dATP and TTP, but not of dGTP or dCTP. Pool sizes increased by a factor of three or four and the increase in Chinese hamster cells was complete by 4 hr after treatment and still persistent 8 hr afterward. Similar observations have been made in E. coZi strains in which increases in dATP and TTP pools are observed even in recA- and umuC- strains (Das and Loeb, 1984). Although nothing in these reports is inconsistent with the existence of an inducible SOS mechanism in mammalian cells, they do permit an alternative interpretation. Both mutations and cytotoxicity are changed by changes in pool size. Such changes occur over time periods similar to those found to be required for the assumed inducible effects in mammalian cells. Mutagens cause changes in pool size (see above). We do not know what the effect of cycloheximide, which not only inhibits protein synthesis but also very quickly inhibits DNA synthesis, may be. The magnitude of the pool size ef-

DNA REPAIR

89

fects on mutation and cytotoxicity is if anything greater than the reported inducible effects; as far as the mutation data are concerned, the effects are very close together in magnitude. Recent studies using in vitro models make the pool size changes seem even more significant. Insofar as the interactions between elongation and proofreading are important, factors which change the kinetic parameters of the DNA synthesizing system should have important effects on the misincorporation of bases. Systematic investigation of the in uitro effect of deoxynucleoside triphosphate concentrations on error frequency (Kunkel and Loeb, 1979; Fersht and Knill-Jones, 1981) indicates that the frequency of misincorporation depends both on the ratio of correct to incorrect nucleotide at the site of misincorporation and on the concentration of the next nucleotide to be added, since the probability of misincorporation is determined in part by the speed of polymerization, i.e., by the time that the proofreading system has to eliminate nucleotide before it is protected and removed from surveillance by chain elongation. An additional factor is involved when noninstructive lesions are present in the DNA, since i n uitro bypass depends on the concentration of dNTP present, the K , being about 500 times that for a normal complementary base on a good template (Rabkin and Strauss, 1984). This factor plus the observed preference of polymerization for adenine nucleotides when confronted with a noninstructional template (Strauss et al., 1982) makes the pool size factor particularly interesting. Such findings do not in any way show that mammalian cells do not have an error-prone, inducible system. They do indicate that the existence of such a system is not yet proved, nor has the supply of possible alternative explanations been exhausted. Not all bacteria have an SOS system. The refractory response of some bacteria to mutagens may be due to the lack of an operative SOS system or to its induction. In addition, the presence of an “errorprone” mode of replication is not necessarily inducible. Some systems inducible in E. coli are (partly) constitutive in other organisms, e.g., Bacillus subtilis, which has a constitutive level of 06-methylguanine methyltransferase 10 times higher than E. coli (Haddan et al., 1983). It is therefore not necessary that analogs of the umuCD genes need exist in eukaryotic cells, or if any exist, that they be inducible. The evolutionary argument that if constitutive, such systems would lead to unacceptably high levels of cell mutability is countered by the discovery that cells have backup systems to correct errors. Such systems must be operative in eukaryotes to counter the inherently greater infidelity of mammalian replication polymerases (Loeb and Kunkel, 1982). In bac-

90

BERNARD S. STRAUSS

teria, there is a methyl-directed mismatch repair system which depends on the time lag between synthesis of new DNA strands and the methylation of susceptible adenines in 5’-GATC-3’ sequences (Glickman and Radman, 1980). This repair system senses mismatches and uses the hypomethylation of one strand as a guide to the strand to be used as a template for mismatch correction (Radman and Wagner, 1984). The observation that a majority of the mutations observed in SOS repair are targeted (Miller, 1982) could be due in part to the operation of mismatch repair which removes errors opposite nontargeted sites, but which is unable to “decide” at targeted sites (B. Glickman, personal communication). The species-specific inequality of gene conversion toward one or another allele (Hastings, 1975) is evidence for the operation of a correction system in eukaryotes, biologically but not necessarily mechanistically related to mismatch repair in prokaryotes. Methylation plays an important role in eukaryotes. As an example relevant to the subject of this review, the hypomethylation of newly synthesized DNA in excision repair patches (Kastan et al., 1982) is a controlling factor in the induction of function of the metallothionin gene system in mammalian cells as the result of the induction of repair patches following mutagen treatment (Lieberman et al., 1983). X. Biological Role of Repair

Are excision repair, transmethylation, and translesion synthesis (bypass) important to the cell? In both bacteria and eukaryotic cells the evidence for the importance of these processes comes from the existence of mutants deficient in repair and hypersensitive to treatment with particular mutagens. In bacteria the number of pyrimidine dimers required to inactivate is determined by the genetic makeup of the treated strain. Uvr bacteria can tolerate about 60 dimers per lo7 nucleotides per lethal hit whereas uvr-rec- double mutants are killed by one or two such lesions (Howard-Flanders and Boyce, 1966). In humans, cells from xeroderma pigmentosum patients are hypersensitive to the effects of UV-irradiation (Kraemer, 1983), although even the most sensitive lines retain the capacity to replicate with some dimers remaining in the DNA (Cleaver, 1970). Data of Robbins et al. (1983)support the supposition that defective DNA repair is associated with premature death of neurons; at any rate, there is a correlation between the rate of UV-induced repair in cells and the development of neurological symptoms in the individual. Holding UV-irradiated fibroblasts in the confluent state so that DNA

DNA REPAIR

91

replication does not occur for periods of days decreases mutation frequency and increases survival (Maher et al., 1979). This decline in mutation frequency is consistent with the idea of an “error-free” process removing lesions which are mutagenic when encountered by the replication mechanism. Grosovsky and Little (1983) have confirmed these results, but have also demonstrated that short (0- to 6-hr) holding periods can increase mutation frequency, suggesting “the existence of a cell-mediated process which enhances the mutagenic potential of at least some UV-induced DNA adducts.” The orthodox view of cancer etiology is that a significant fraction of tumors are the result of mutation(s) induced by carcinogenic agents. This view has been challenged by Cairns (1981) (Kennedy et al., 1984), who points out that if chemical adducts are a direct cause of cancer, one would expect to find a higher incidence of internal tumors in xeroderma pigmentosum patients. On the basis of limited epidemiological evidence (Cairns, 1981),he concludes that this is not the case and that some other mechanism such as spontaneous transposition might be involved. Given the high incidence of skin tumors in xeroderma patients, Cairns admits that light-induced lesions are likely carcinogenic. The more recent paper is based on the experimental system of Kennedy et al. (1980, 1984), and the conclusions depend also on the nature of the mouse 10T1/2 cells used in those experiments (see above). Kraemer et al. (1984) have examined the literature on 726 xeroderma pigmentosum patients to test the Cairns conclusions. As expected, a huge increase in skin cancer was observed in these patients. In addition, a 12-fold increase in neoplasms at sites not exposed to radiation was observed, including a disproportionate increase in brain tumors, but not of lymphomas or endocrine system cancers. The Kraemer et al. (1984) data support the concept that cancer is induced by the kind of lesion that xeroderma patients are unable to repair in the case of skin and some other tumors. The Cairns (1981) view may be correct for the common internal neopIasms or, as Kraemer et al. (1984) suggest, alternative explanations based on the short and sheltered lives of xeroderma patients may account for the current lack of definitive data. A number of laboratories have isolated rodent lines sensitive to UVirradiation and to alkylating agents, and such strains are found to be deficient in excision repair activity (Thompson et al., 1981, 1983; Stefanini et al., 1982). Five different complementation groups of UVsensitive Chinese hamster mutants have been isolated, each of which showed low (less than 10%) incision rates at UV lesions, were hyper-

92

BERNARD S . STRAUSS'

mutable, and had greatly increased sister chromatid exchange rates in response to UV. The Chinese hamster system therefore resembles the human xeroderma system in the multiplicity of complementation groups (Thompson et al., 1982). Molecular cloning of a human repair gene ERCCl (Excision Repair Complementing defective repair in Chinese hamster cells) into the Chinese hamster mutants restores their resistance to UV and to mitomycin C (Rubin et al., 1983; Westerveld et al., 1984), indicating an interchangeability of repair genes in both species. Such data imply that rodent cells employ excision repair in duo, notwithstanding their lowered measured unscheduled DNA synthesis (UDS) and excision repair synthesis values. [The reverse is not necessarily true; a UV-resistant human cell mutant has been obtained without increased excision repair activity (Suzuki, 1984).] Revertants of xeroderma pigmentosum cells which have regained the normal resistance of human cells to UV-irradiation have a restored ability to remove pyrimidine dimers from their DNA (Royer-Pokora and Haseltine, 1984). Such results make it likely that excision repair

10-

~~

10

20

30

I

1

1

20

40

6(

uv Dose I J I ~ * ) FIG.5. (A) UV survival curves of mouse and human cells. (B) Host-cell reactivation of UV-irradiated herpes simplex virus in mouse and human cells. Closed circles, BALB/c mouse cells; closed squares, C3HIHe mouse cells; closed triangles, C57BU10 mouse; open circles, HK normal human skin fibroblasts; open squares, XP3OS xeroderma complementation group A fibroblasts. These cells gave the following UDS (unscheduled DNA synthesis) values (in grains per nucleus, controls in parentheses): BALB/c, 32.3 f 1.5 (10.1f 0.4); C3HMe, 22.5 f 1.2 (9.5 2 0.4); C57BU10,27.0 1.3(11.4 f 0.5); HK, 80.7 f 2.4 (12.8 f 0.5); XP3OS, 9.7 f 0.4 (9.3 f 0.4) (from Yagi, 1982).

*

DNA REPAIR

93

does play a major role in the protection of cells from cytotoxicity and from mutation, hence the idea that the process is “error free.” Additional factors must operate even in cell systems that do carry out (some) excision repair. Mouse cells have the same sensitivity to UV-irradiation as do human cells (Fig. 5; Yagi, 1982), but they have only one-fourth the UDS activity. Some of the variation reported in repair activity between cell passages in embryonic mouse cells (Peleg et al., 1976) is probably due to changes in pool size (see above), but the relative deficiency in excision repair activity of mouse cells as compared to humans is unmistakable, notwithstanding the similarity in killing curves. The difference is reflected in the relative survival of herpesvirus plated on mouse, human, and human xeroderma lines in which the survival of herpes plated on mouse cells is equivalent to that of herpes plated on xeroderma cells and significantly lower than when plated on normal human cells (Yagi, 1982). These data imply that there is some mechanism accessible to chromosomal DNA, but not to extrachromosomal (plasmid?) DNA, which permits the accommodation of UV lesions, although without necessarily removing them. The result is interesting in its own right and also because it clearly demonstrates that cellular metabolic systems can handle chromosomal and nonchromosomal DNA differently. XI. Appendix

The MEDLINE data base lists 637 papers, including 55 reviews in 1984 and 34 (3 reviews) so far (May, 1985) in 1985, containing material on DNA repair. Of interest is a new text devoted to the subject by Friedberg (1985).Particular progress has been made since this review was completed on studies of O-alkylation repair (see Yarosh, 1985) and on the problem of an inducible “SOS” system in eukaryotic cells. The role of poly(ADP-ribose) in repair remains a mystery (Berger, 1985; Cleaver et al., 1985). Further progress has been made toward the cloning of a human DNA repair gene (Rubin et al., 1985). Additional work on the use of antibodies to detect lesions in DNA has been published (Mitchell and Clarkson, 1984; Paules et al., 1985; Wani e t al., 1984), including a review by Strickland and Boyle (1984). Cohn and Lieberman (1984a,b) use antibodies to 5bromodeoxyuridine to isolate DNA sequences containing excision repair sites and show these to be nonrandomly distributed. Physiological and genetic factors affect the levels of repair, and there can be considerable interindividual and even intraindividual variation (Harris et al., 1984). Munch-Petersen et al. (1985) report a seasonal variation in repair ac-

94

BERNARD S. STRAUSS

tivities, with greater variation observed in summer months. Maslansky and Williams (1985) show that the total UV repair capacity of hepatocytes from five species of different longevity is identical, although the cells from longer lived species respond faster to low doses. They suggest the repair of low-level damage as playing a role in aging. The role of ligase activation in the action of poly(ADP-ribose) has been questioned in a paper showing that although increased repair replication occurs in human lymphoid cells inhibited with 3aminobenzamide, no increase in patch size is observed, as would be expected if ligase action were inhibited (Cleaver, 1985). The 06methylguanine acceptor protein in E. coli is synthesized as part of a 39-kDa complex which on isolation is readily degraded to a 19-kDa active protein (Demple et al., 1985). The larger complex removes methyl groups from phosphotriesters as well as acting on @-methylthymine. The inducible 3-methyladenine glycosylase in bacteria (the alkA product) also acts on O-methylated cytosine and on 02-methylthymine and 3-methylguanine residues. 3-Methyladenine induces SOS repair in uiuo (Boiteux et aE., 1984) and blocks DNA synthesis in uitro (Larson et al., 1985). It is possible that in mammalian cells, a product different from the 06-acceptor protein acts on 04-methylthymine adducts (Becker and Montesano, 1985). A stronger case can now be made for an inducible “SOS-like” system in eukaryotic cells. In yeast, the technique of Kenyon and Walker (1981) (see Walker, 1984) has been used to demonstrate damage-inducible genes (Ruby and Szostak, 1985).However, as a recent review by Siede and Eckardt (1984) concludes, “a general acting error-prone repair activity (in yeast) analogous to the SOS system of E. coli has not yet been demonstrated.” Herrlich et al. (1984) point out a variety of changes in human cells reminiscent of an SOS response. Additional demonstrations that pretreatment regimens increase the repair of a DNA (Leadon and Hanawalt, 1984) or the replication of damaged SV40 (Maga and Dixon, 1984) have appeared. There is now evidence for a type of mismatch repair in mammalian cells (Folger et al., 198510) and for a kind of gene conversion (Folger et al., 1985a).Of interest are recent studies on the specificity of mismatch repair in bacteria (Kramer et al., 1984; Dohet et al., 1985)which show that mismatches are repaired with different efficiency. T/G, C/A, and G/G mismatches are repaired with high efficiency; G/A, A/G, T/T, C/C, C/T, and T/C mismatches are repaired with low efficiency. Contrary to what might be expected, therefore, it is the potential transversions which escape correction by the mismatch repair system. These findings imply that some mutagen specificity results (e.g., Miller, 1983) might be the

DNA REPAIR

95

result of mismatch repair occurring after replication, so that what is seen is only the residuum of what is produced. Evidence for this view is found in a paper by Mattern et al. (1985).These investigators determined that mutations produced as the result of the operation of an SOS system were, for the most part, transversions. However, when the experiments were repeated in either E. coli dam or mutL strains, deficient in mismatch repair, an excess of transitions was again found.

ACKNOWLEDGMENTS The work reported from this laboratory has been generously supported by grants from the National Institutes of General Medical Sciences (GM 07816), the National Cancer Institute (CA 32436, CA 19265), and the Department of Energy (DE-AC02-76EV020).

REFERENCES Adamkiewicz, J., Ahrens, O., Eberle, G., Nehls, P., and Rajewsky, M. (1985). ZARC Monogr. Cyclic Nucleic Acid Adducts, in press. Ahmmed, Z., and Laval, J. (1984). Biochem. Biophys. Res. Commun. 120, 1-8. Ahmed, F., and Setlow, R. (1979). Cancer Res. 39,471-479. Altamirano-Dimas, M., Sklar, R., and Strauss, B. (1979). Mutat. Res. 60, 197-206. Arenaz, P., and Sirover, M. (1983).Proc. Natl. Acad. Sci. U S A . 80,5822-5826. Ayres, K., Sklar, R., Larson, K., Lindgren, V., and Strauss, B. (1982).Mol. Cell. Biol. 2, 904-9 13. Bailey, G., Nixon, J., Hendricks, J., Sinnhuber, R., and Von Holde, K. (1980).Biochemistry 19,5836-5842. Barnett, S., Landaw, E., and Dixon, K. (1984). Biophys. J. 46,307-321. Becker, R., and Montesano, R. (1985). Carcinogenesis 6,313-318. Becker, R., Barrow, L., and Shank, R. (1981).Carcinogenesis 2, 1181-1188. Bedell, M., Lewis, J., Billings, K., and Swenberg, J. (1982).Cancer Res. 42,3079-3083. Ben-Ishai, R., Sharon, R., Rothman, M., and Miskin, R. (1984).Carcinogenesis 5,357362. Benjamin, R., and Gill, D. M. (1980a).I . B i d . Chem. 255, 10502-10508. Benjamin, R., and GiII, D. M. (1980b).J . Biol. Chem. 255, 10493-10501. Berger, N. (1985). Radiat. Res. 101,4-15. Berger, N., and Sikorski, G. (1981). Biochemistry 20, 3610-3614. Berger, N., Sikorski, G., Petzold, S., and Kierohara, K. (1980). Biochemistry 19, 289293. Bianchi, V., Zantedeschi, A., and Levis, A. (1983). Chem. Biol. Interact. 43, 17-31. Blackburn, E. (1984). Cell 37, 7-8. Blackburn, E., and Szostak, J. (1984).Annu. Reu. Biochem. 53, 163-194. Bochstahler, L. (1981). Prog. Nucleic Acid Res. Mol. Biol. 26, 303-313. Bodell, W., and Cleaver, J. (1981).Nucleic Acids Res. 9, 203-213. Bogden, J., Eastman, A., and Bresnick, E. (1981).Nucleic Acids Res. 9, 3089-3103. Boiteux, S., Huisman, O., and Laval, J. (1984).EMBOJ. 3, 2569-2574. Bose, K., Karran, P., and Strauss, B. (1978). Proc. Natl. Acad. Sci. U.S.A. 75, 794-798.

96

BERNARD S. STRAUSS

Boyce, R.,and Howard-Flanders, P. (1964). Proc. Natl. Acad. Sci. U S A . 51,293-300. Bradley, M. O., and Tayler, 0. (1983). Carcinogenesis 4, 1513-1517. Brouwer, J., van de Putte, P., Fichtinger-Schepman, A., and Reedijk, J. (1981). Proc. Natl. Acad. Sci. U.SA. 78,7010-7014. Brown, A., Fickel, T., Cleaver, J., Lohman, P., Wade, M., and Waters, R. (1979).Cancer Res. 39,2522-2527. Cairns, J. (1981). Nature (London) 289,353-357. Calos, M., Lebkowski, J., and Botchan, M. (1983). Proc. Natl. Acad. Sci. U S A . 80, 3015-3019. Charlesworth, J., Chu, Y., O’Conner, P., and Craig, A. (1981). Carcinogenesis 2,329342. Clarkson, J. (1978).Mutat. Res. 52,273-284. Clarkson, J., Mitchell, D., and Adair, G. (1983). Mutat. Res. 112,287-295. Clayton, D., Doda, J.,and Friedberg, E. (1974). Proc. Natl. Acad. Sci. U S A . 71,27772781. Cleaver, J. (1970). lnt. J. Radiat. Biol. 18,557-565. Cleaver, J. E. (1982a). Carcinogenesis 3, 1171-1174. Cleaver, J. E. (198213). Cancer Res. 42,860-863. Cleaver, J. E. (1983). Biochim. Biophys. Acta 739,301-311. Cleaver, J. (1985). Cancer Res. 45, 1163-1169. Cleaver, J. E., Bodell, W. J., Borek, C., Morgan, W. F., and Schwartz, J. L. (1983a).Int. Symp. Princess Takamatsu Cancer Res. Fund 13,195-207. Cleaver, J. E., Bodell, W. J., Morgan, W. F., and Zelle, B. (1983b).J . Biol. Chem. 258, 9059-9068. Cleaver, J. E., Kaufinan, W., Kapp, L., and Park, S. (1983~). Biochim. Biophys. Acta 739, 207-215. CIeaver, J., Milam, K., and Morgan, W. (1985). Radiat. Res. 101, 16-28. Cohen, M., and Simpson, S. (1982). Enuiron. Mutagen. 4,27-36. Cohn, S . , and Lieberman, M. (1984a).J . Biol. Chem. 259,12456-12462. Cohn, S., and Lieberman, M. (1984b).]. B i d . Chem. 259, 12463-12469. Cohn, S., Krawisz, B., Dresler, S., and Lieberman, M. (1984). Proc. Natl. Acad. Sci. U S A . 81,4828-4832. Collier, I., Olson, W., and Regan, J. (1983).J . Theor. Biol. 102,487-499. Collins, A., and Johnson, R. (1979). Nucleic Acids Res. 7, 1311-1320. Cornelis, J., Lupker, J., and Van der Eb, A. (1980). Mutat. Res. 71, 139-146. Coulondre, C., and Miller, J. (1977).]. MoZ. Biol. 117, 577-606. Cozzarelli, N. (1977). Annu. Reu. Biochem. 46, 641-668. Cramer, P., and Painter, R. (1981). Nature (London)291,671-672. Creissen, D., and Shall, S. (1982).Nature (London) 296,271-272. Das, S . , and Loeb, L. (1984). Mutat. Res. 131,97-100. Das, S., Benditt, E., and Loeb, L. (1983). Biochem. Biophys. Res. Commun. 114,458464. Das Gupta, V., and Summers, W. (1978). Proc. Natl. Acad. Sci. U S A . 75,2378-2381. Day, R., and Ziolkowski, C. (1979). Nature (London)279, 797-799. Day, R., and Ziolkowski, C. (1981). Carcinogenesis 2,213-218. Day, R., Ziolkowski, C., Scudiero, D., Meyer, S . , Lubiniecki, A., Galloway, S., and Bynum, G. (1980a). Nature (London) 288,724-727. Day, R., Ziolkowski, C., Scudiero, D., Meyer, S., and Mattern, M. (1980b).Carcinogenesis 1,21-32. Day, R., Yarosh, D., and Ziolkowski, C. (1984). Mutat. Res. 131,45-52.

DNA REPAIR

97

Defais, M., Hanawalt, P., and Sarasin, A. (1983).Adu. Radiat. Biol. 10, 1-37. DeJonge, A., Vermeulen, W., Klein, B., and Hoejmakers, J. (1983).E M B O J . 2,637-641. Demple, B., and Linn, S. (1980). Nature (London) 287, 203-307. Demple, B., Jacobson, A., Olsson, M., Robins, P., and Lindahl, T. (1982).J.Biol. Chem. 257,13776-13780. Demple, B., Sedgwick, B., Robins, P., Totty, N., Waterfield, M., and Lindahl, T. (1985). Proc. N a t l . Acad. S c i . U S A . 82,2688-2692. Dinsart, C., Cornelis, J., Klein, B., van der Eb, A., and Rommelaere, J. (1984).Mol. Cell. Biol. 4,324-328. Dohet, C., Wagner, R., and Radman, M. (1985). Proc. Natl. Acad. Sci. U.S.A. 82,503505. Doniger, J. (1978).J. Mol. Biol. 120, 433-446. Downes, C., Collins, A., and Johnson, R. (1983).Mutat. Res. 112, 75-83. Dresler, S., and Lieberman, M. (1983).J. Biol. Chem. 258, 12269-12273. Dresler, S., Roberts, J,, and Lieberman, M. (1982). Biochemistry 21,2557-2564. Dunn, W. C., and Regan, J. D. (1979). Mol. Pharmacol. 15,367-374. Durkacz, B., Omidiji, O., Gray, D., and Shall, S. (1980).Nature (London)283,593-596. Durkacz, B., Irwin, J., and Shall, S . (1981). Biochem. Biophys. Res. Commun. 101, 1433-1441. Eadie, J., Conrad, M., Toorchen, D., and Topal, M. (1984).Nature (London) 308,201203. Echols, H., Lu, C., and Burgers, P. (1983).Proc. Natl. Acad. Sci. U S A . 80,2189-2192. Edenberg, H. (1983). Comments Mol. Cell. Biophys. 1, 383-347. Edenberg, H., and Hanawalt, P. (1972). Biochim. Biophys. Acta 272,361-372. Edenberg, P., and Hanawalt, P. (1973). Biochim. Biophys. Acta 324,206-217. Ehmann, U., Cook, K., and Friedberg, E. (1978).Biophys. ].22,249-264. Eisenstadt, E., Warren, A., Porter, J., Atkins, D., and Miller, J. (1982).Proc. Natl. Acad. Sci. U.S.A.79, 1945-1949. Elledge, S., and Walker, G., (1983).J . Mol. Biol. 164, 175-192. Elliot, G., and Johnson, R. (1983).J. Cell Sci. 60, 267-288. Evensen, G., and Seeberg, E. (1982).Nature (London) 296,773-775. Fasman, G. (1976). “Handbook of Biochemistry and Molecular Biology,” 3rd Ed., Vol. 11. CRC Press, Cleveland, Ohio. Ferno, A., Higgins, N. P., and Olivera, B. (1983).J . Biol. Chem. 258, 6000-6003. Fersht, A., and Knill-Jones, J. (1981). Proc. Natl. Acad. Sci. U S A . 78,4251-4255. Folger, K., Thomas, K., and Capecchi, M. (1985a).Mol. Cell. Biol. 5,59-69. Folger, K., Thomas, K., and Capecchi, M. (198513).Mol. Cell. Biol. 5, 70-74. Foote, R., and Mitra, S. (1984). Carcinogenesis 5, 277-281. Foster, P., Eisenstadt, E., and Miller, J. (1983).Proc. Natl. Acad. Sci. U.S.A. 80,26952698. Francis, A., Blevins, R. D., Carrier, W. L., Smith, D., and Regan, J. D. (1979).Biochim. Biophys. Acta 563,385-392. Francis, A., Lee, W., and Regan, J. (1981).Mech. Ageing Deu. 16, 181-189. Freidberg, E. (1985). “DNA Repair,” pp 614. Freeman, San Francisco, California. Friedberg, E., and Hanawalt, P. (1981). In “DNA Repair: A Laboratory Manual of Research Procedures” (E. Friedberg and P. Hanawalt, eds.), Vol. I, pp. xvii, 637. Dekker, New York. Friedlos, F., and Roberts, J. (1978). Nucleic Acids Res. 5,4795-4803. Ganeson, A. (1974).J. MoZ. Biol. 87, 103-119. Glickman, B., and Radman, M. (1980). Proc. Natl. Acad. Sci. U S A . 77, 1063-1067.

98

BERNARD S. STRAUSS

Gombar, C., Katz, E., Magee, P., and Sirover, M. (1981). Carcinogenesis 2,595-599. Goodwin, P., Lewis, P., Davies, M., Skidmore, C., and Shall, S. (1978). Biochim. Biophys. Acta 543,576-582. Goth, R., and Rajewsky, M. (1974). Proc. Natl. Acad. Sci. U S A . 71,639-643. Goth-Goldstein, R. (1977).Nature (London) 267,80-82. Goth-Goldstein, R. (1980). Cancer Res. 40,2623-2624. Groopman, J., Hauten, A., Goodrich, G., Wogan, G., and Harris, C. C. (1982). Cancer Res. 42,3120-3124. Grosovsky, A., and Little, J. (1983). Mutat. Res. 110,401-412. Grosse, F., and Krauss, G. (1984).Eur.]. Biochem. 141, 109-114. Gupta, P., and Sirover, M. (1980).Mutat. Res. 72,273-284. Gupta, P., and Sirover, M. (1984).Proc. Natl. Acad. Sci. U S A . 81, 757-761. Haddan, C., Foote, R., and Mitra, S. (1983)./. Bacteriol. 153,756-762 Hanawalt, P., Cooper, P., Ganeson, A., and Smith, C. (1979).Annu. Reo. Biochem. 48, 783-836. Harris, A., Karran, P., and Lindahl, T. (1983). Cancer Res. 43,3247-3252. Harris, C., Autrup, H., Vahakangas, K., and Trump, B. (1984). Banbury Rep. 16, 145154. Hart, R., and Setlow, R. (1975). Proc. Natl. Acad. Sci. U S A . 71,2169-2173. Haseltine, W., Gordon, L., Lindan, C., Grafstrom, R., Shaper, N., and Grossman, L. (1980).Nature {London) 285,634-641. Hastings, P. (1975).Annu. Reo. Genet. 9, 129-144. Haugen, A., Groopman, J., Hsu, I., Goodrich, G., Wogan, G., and Harris, C. (1981).Proc. Natl. Acad. Sci. U S A . 78,4124-4127. Hayaishi, O., and Ueda, K. (1977).Annu. Rev. Biochem. 46,85-116. Haynes, R., and Kunz, B. (1981). In “Molecular Biology of the Yeast Saccharomyces; Life Cycle and Inheritance” (J. Strathem, E. Jones, and J. Brocech, eds.), pp. 371414. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Heddle, J., and Arlett, C. (1980).Mutat. Res. 72, 119-125. Helland, D., Kleppe, R., Lillehaug, J., and Kleppe, K. (1984).Carcinogenesis 5, 833836. Hensen, P. (1978)./. Mol. B i d . 119,487-506. Herrlich, P., Mallick, U., Ponta, H., and Rahmsdorf, H. (1984). Hum. Genet. 67,360368. Hertzog, P., Smith, J.. and Gamer, R. (1982). Carcinogenesis 3,825-828. Higgins, N. P., Kato, K., and Strauss, B. (1976).J . MoZ. B i d . 101,417-425. Hillebrand, G., McCluskey, A., Abbott, K., Revich, G., and Beattie, K. (1984).NucZeic Acids Res. 12,3155-3171. Hochschild, A., Irwin, N., and Ptashne, M. (1983). Cell 32,319-325. Houldsworth, J., and Lavin, M. (1980).Nucleic Acids Res. 8, 3709-3720. Howard-Flanders, P., and Boyce, R. (1966). Radiat. Res. Suppl. 6, 156-184. Hsu, I., Poirier, M., Yuspa, S., Yolken, R., and Harris, C. (1980).Carcinogenesis 1,455458. Hsu, I., Poirier, M., Yuspa, S., Greenberger, D., Weinstein, I., Yolken, R., and Harris, C. (1981). Cancer Res. 41, 1091-1095. Hubscher, U., Kuenzle, C., and Spadari, S. (1979). Proc. Natl. Acad. Sci. U S A . 76, 2316-2320. Humayan, Z., Kleed, D., and Ptashne, M. (1977).Nucleic Acids Res. 4, 1595-1607. Ikegami, S., Taguchi, T., Ohashi, M., Oguro, M., Nagano, H., and Mano, Y. (1978). Nature {London) 275,458-460.

DNA REPAIR

99

Irwin, J.. and Straws, B. (1980). Enoiron. Mutogen. 2,381-388. Jacobson, E., Antol, K., Juarez-Salinas, H., and Jacobson, M. (1983).J.Biol. Chem. 258, 103-107. Jacobson, M., Levi, J., Juarez-Salinas, H., Barton, R., and Jacobson, E. (1980). Cancer Res. 40,1797-1802. Jahn, C., and Litman, G. (1979). Biochemistry 22,1442-1449. James, M., and Lehmann, A. (1982). Biochemistry 21,4007-4013. Jaspers, N., and Bootsma, D. (1982). Proc. Natl. Acad. Sci. U . S A . 79,2641-2644. Jeggo, P. (1979).J . Bacteriol. 139, 783-791. Jeggo, P., Defais, M., Samson, L., and Schendel, P. (1978). Mol. Gen. Genet. 162,299305. Jump, D., and Smulson, M. (1980). Biochemistry 19, 1024-1030. Jump, D., Batt, T., and Smulson, M. (1980). Biochemistry 19,1031-1037. Kaguni, L., and Clayton, D. (1982). Proc. Natl. Acad. Aci. U S A . 79, 983-987. Kaneko, M., and Cerutti, P. (1980). Cancer Res. 40,4313-4319. Kano, Y., and Fujiwara, Y. (1983). Carcinogenesis 4, 1419-1424. Kapp, L., and Painter, R. (1982). Int. Reo. Cytol. 80, 1-25. Kastan, M., Gowans, B., and Lieberman, M. (1982). Cell 30,509-516. Kato, H., Harada, M., Tsuchiya, K., and Moriwaki, K. (1980).Jpn.J . Genet. 55,99-108. Katz, E., Gupta, P., and Sirover, M. (1983). Mutat. Res. 112,345-358. Kaufmann, W., Bodell, W., and Cleaver, J. (1983). Carcinogenesis 4, 179-184. Kenne, K., and Ljungquist, S. (1984). Nucleic Acids Res. 12, 3057-3068. Kennedy, A., and Little, J. (1980). Carcinogenesis 1, 1038-1047. Kennedy, A., Fox, M., Murphy, G., and Little, J. (1980). Proc. Natl. Acad. Sci. U S A . 77, 7262-72166. Kennedy, A., Cairns, J., and Little, J. (1984). Nature (London) 307,85-86. Kenyon, C., and Walker, G. (1981). Proc. Natl. Acad. Sci. U.S.A. 77,2819-2823. Kohn, K., Eweg, R., Erickson, L., and Swelling, L. (1981). In “DNA Repair: A Laboratory Manual of Research Procedures” (E. Friedberg and P. Hanawalt, eds.), pp. 379-401. Dekker, New York. Kornberg, A. (1980). “DNA Replication,” pp. x, 724. Freeman, San Francisco, California. Kornberg, A. (1982). “1982 Supplement to DNA Replication,” pp. viii, 273. Freeman, San Francisco, California. Kraemer, K. (1983). In “Update: Dermatology in General Medicine” (T. Fitzpatrick, A. Eisen, K. Wolff, I. Freedberg, and K. Austen, eds.), pp. 113-142. McGraw Hill, New York. Kraemer, K., Lee, M., and Scotts, J. (1984). Carcinogenesis 5, 511-514. Kramer, B., Kramer, W., and Fritz, H. (1984). Cell 38, 879-887. Kriegstein, H., and Hogness, D. (1974). Proc. Natl. Acad. Sci. U S A . 71, 135-139. Krokan, H., Haugen, A., Myrnes, B., and Guddal, P. (1983). Carcinogenesis 4, 15591564. Kunkel, T. (1984). Proc. Natl. Acad. Sci. U S A . 81, 1494-1498. Kunkel, T., and Loeb, L. (1979).J. Biol. Chem. 254,5718-5725. LaDuca, R., Fay, P., Cheeang, C., McHenry, C., and Bambara, R. (1983). Biochemistry 22,5177-5188. Larson, K., Sahm, J., Shenkar, R., and Straws, B. (1985). Mutat. Res. 150, 77-84. Laval, F., and Laval, J. (1984). Proc. Natl. Acad. Sci. U.S.A.81, 1062-1066. Lavi, S. (1981). Proc. Natl. Acad. Sci. U . S A . 78, 6144-6148. Lawley, P., and Thatcher, C. (1970). Biochernistrv 116, 693-707.

100

BERNARD S. STRAUSS

Lawley, P., and Warren, W. (1981).l n “DNA Repair: A Laboratory Manual of Research Procedures” (E. Friedberg and P. Hanawalt, eds.), Vol I, part A, pp. 129-142. Dekker, New York. Leadon, S., and Hanawalt, P. (1983). Mutat. Res. 112, 191-200. Leadon, S., and Hanawalt, P. (1984). Carcinogenesis 5, 1505-1510. Lebkowski, J. J., DuBridge, R., Antell, E., Greisen, K., and Calos, M. (1984).Mol. Cell. B i d . 4, 1951-1960. Lehman, A., and Broughton, B. (1984). Curcinogenesis 5, 117-119. Lewis, J., and Swenberg, J. (1980).Nature (London) 288, 185-187. Lewis, J., and Swenberg, J. (1983).Carcinogenesis 4, 529-536. Lieberman, M., Beach, L., and Palmeter, R. (1983).Cell 35,207-214. Lindahl, T. (1976).Nature (London) 259,64-66. Lindahl, T. (1979).Prog. Nucleic Acid Res. Mol. B i d . 22, 135-192. Lindahl, T. (1982).Annu. Reo. Biochem. 51,61-87. Lippke, J., and Haseltine, W. (1981). In “DNA Repair: A Laboratory Manual of Research Procedures” (E. Friedberg and P. Hanawalt, eds.), Vol. 11, pp. 187-198. Dekker, New York. Little, J. (1984). Proc. Natl. Acad. Sci. U S A . 81, 1375-1379. Little, J., and Mount, D. (1982).Cell 29, 11-22. Liu, P., Parsons, C., and Hanawalt, P. (1982).J. lnoest. Dennutol. 79, 330-335. Liu, P., Parsons, S., and Hanawalt, P. (1983).J. Inoest. Dennutol. 81, 179s-183s. Liu, P., Chang, C., and Trosko, J. (1984). Somatic Cell Mol. Genet. 10,235-245. Loeb, L., and Kunkel, T. (1982).Annu. Reu. Biochem. 52,429-457. Lu, A., Clark, S., and Modrich, P. (1983). Proc. Nutl. Acad. Sci. U.S.A. 80,4639-4643. Lundquist, R., and Olivera, B. (1982). Cell 31,53-60. McCarthy, J., Edington, B., and Schendel, P. (1983). Proc. Natl. Acad. Sci. U . S A . 80, 7380-7384. McCarthy, T., Karran, P., and Lindahl, T. (1984).EMBO]. 3,545-550. McCurry, L., and Jacobson, M. (1981).J.Biol. Chem. 256,551-553. McGhee, J., and Felsenfeld, G. (1979).Proc. Nutl. Acad. Sci. U S A . 76, 2133-2137. McGhee, J., and Felsenfeld, G. (1980).Annu. Rev. Biochem. 49, 1115-1156. Maga, J., and Dixon, K. (1984).Photochem. Photobiol. 40,473-478. Magana-Schwenke, N., Henriques, J., Chanet, R., and Moustacchi, E. (1982).Proc. Natl. Acad. Sci. U S A . 79, 1722-1726. Maher, V., Dorney, D., Mendrala, A., Konze-Thomas, B., and McCormick, J. (1979). Mutat. Res. 62,311-323. Mallick, V., Rahmsdorf, H., Yamamoto, N., Ponta, H., Wegner, R., and Herrlich, P. (1982). Proc. Natl. Acad. Sci. U.SA. 79, 7886-7890. Mansbridge, J., and Hanawalt, P. (1983).In “Cellular Responses to DNA Damage” (E. Friedberg and B. Bridges, eds.), pp. 1-13. Liss, New York. Margison, G., and Pegg, A. (1981).Proc. Natl. Acad. Sci. U S A . 78,861-865. Maslansky, C., and Williams, G. (1985).Mech. Ageing Deo. 29, 191-204. Mattern, M., and Scudiero, D. (1981). Biochim. Biophys. Acta 653,248-258. Mattern, I., Olthoff-Smit, F., Jacobs-Meijsing, B., Enger-Valk, B., Pouwels, P., and Lohman, P. (1985). Mutat. Res. 148,35-45. Meneghini, R. (1981). Trends Biochem. Sci. 6,214-216. Meneghini, R., Menck, C., and Schumacher, R. I. (1981). Q. Reu. Biophys. 14, 381432. Meuth, M. (1981).Somatic Cell Genet. 7,89-102. Meuth, M., L’Heureux-Huard, N., and Trudel, M. (1979).Proc. Nutl. Acad. Sci. U S A . 76,6505-6509.

DNA REPAIR

101

Miller, J. (1982). Cell 31, 5-7. Miller, J. (1983).Annu. Reu. Genet. 17, 215-238. Miller, J., and Low, K. (1984). Cell 37, 675-682. Miller, M. R., and Chinault, D. N. (1982).J. B i d . Chem. 257, 10204-10209. Miskin, R., and Ben-Ishai, R. (1981). Proc. Natl. Acad. Sci. U S A . 78, 6236-6240. Mitchell, D., and Clarkson, J. (1981). Biochim. Biophys. Acte 655,54-60. Mitchell, D., and Clarkson, J. (1984). Photochem. Photobiol. 40, 743-748. Mitra, S., Pal, B. C., and Foote, R. S. (1982).J . Bacteriol. 152,534-537. Montesano, R., Bresil, H.,Planche-Martel, G., Margison, G., and Pegg, A. (1983). Cancer Res. 43,5808-5814. Moore, P., and Strauss, B. (1979). Nature (London) 278,664-666. Moore, P., Rabkin, S., and Strauss, B. (1980). Nucleic Acids Res. 8,4473-4484. Moore, P., Bose, K., Rabkin, S., and Strauss, B. (1981). Proc. Natl. Acad. Sci. U S A . 78, 110-114. Moore, P., Rabkin, S., and Strauss, B. (1982).I n “Molecular and Cellular Mechanisms of Mutagenesis” (J. Lemontt and W. Generoso, eds.), pp. 179-197. Plenum, New York. Morgan, W., and Cleaver, J. (1983). Cancer Res. 43,3104-3107. Moms, C., Mohamed, R., and Lavin, M. (1983). Mutat. Res. 112,67-74. Mortelmans, K., Friedberg, E., Slor, H., Thomas, G., and Cleaver, J. (1976). Proc. Natl. Acad. Sci:U.S.A. 73, 2757-2761. Mosbaugh, D. W., and Linn, S. (1983).J. Biol. Chem. 258,108-118. Muller, R., and Rajewsky, M. (1980). Cancer Res. 40,887-896. Muller, R., and Rajewsky, M. (198l).J. Cancer Res. Clin.Oncol. 102, 99-113. Muller, R., and Rajewsky, M. (1983). 2.Naturforsch. C : Biosci. 38, 1023-1029. Munch-Petersen, B., Wallevik, K., and Faber, M. (1985).Scand.J. Clin. Lab. Invest. 45, 37-44. Mymes, B., Giercksky, K., and Krokan, H. (1983). Carcinogeriesis 4, 1565-1568. Mymes, B., Norstrand, K.,Giercksky, K., Sjunneskog, C., and Krokan, H. (1984).Carcinogenesis 5, 1061-1064. Nduka, N., Skidmore, C., and Shall, S. (1980). Eur. J . Biochem. 105, 525-530. Nehls, P., Rajewsky, M., Spiess, E., and Werner, D. (1984). E M B O ] . 3, 327-332. Niggli, H., and Cerutti, P. (1982). Biochem. Biophys. Res. Commun. 105, 1215-1223. O’Connor, P., Chu, Y., Cooper, D., Maru, G., Smith, R., and Margison, G. (1982).Biochimie 64,769-773. Ohashi, Y., Ueda, D., Kawaichi, M., and Hayaishi, 0. (1983). Proc. Natl. Acad. Sci. U S A . 80,3604-3607. Oikawa, A., Toda, H., Kanai, M., Miwa, M., and Sugimura, T. (1980). Biochem. Biophys. Res. Commun. 97,1311-1316. Oleson, F., Mitchell, B., Dipple, A., and Liebeman, M. (1979). Nucleic Acids Res. 7, 1343-1361. Olsson, M., and Lindahl, T. (1980).J . Biol Chem. 255, 10569-10571. Painter, R. (1981). Mutat. Res. 84, 183-190. Painter, R. (1983). Radiat. Res. 95,421-426. Painter, R., and Young, B. (1980). Proc. Natl. Acad. Sci. U.S.A.77, 7315-7317. Park, S., and Cleaver, J. (1979a). Proc. Natl. Acad. Sci. U.S.A.76,3927-3931. Park, S . , and Cleaver, J. (1979b). Nucleic Acids Res. 6, 1151-1159. Park, S., Choi, K., Hong, S., and Cleaver, J. (1981). Mutat. Res. 82, 365-371. Park, S., Kim, C., and Kim, M. (1983). Enuiron. Mutagen 5, 515-525. Paules, R., Poirier, M., Mass, M., Yuspa, D., and Kaufman, D. (1985).Carcinogenesis 6, 193-198.

102

BERNARD S. STRAUSS

Pegg, A. (1977).Adu. Cancer Res. 25, 195-267. Pegg, A. (1983).Rev. Biochem. Toxicol. 5,83-133. Pegg, A., and Perry, W. (1981).Carcinogenesis 2,1195-1200. Pegg, A., and Wiest, L. (1983). Cancer Res. 43,972-975. Pegg, A., Hui, G., and Rogers, K. (1978).Biochim. Biophys. Acta 520,671-678. Pegg, A., Perry, W., and Bennett, R. (1981). Biochem. J. 197,195-201. Pegg, A., Roberfroid, M., Von Bohr, C., Foote, R., Mitra, S., Bresil, H., Likhachev, A., and Montesano, R. (1982).Proc. Natl. Acad. Sci. U S A . 79,5162-5165. Pegg, A., Wiest, L., Foote, R., Mitra, S., and Peny, W. (1983).J. Biol Chem. 258,23272333. Peleg, L., Roz, E., and Ben-Ishai, R. (1976).E x p . Cell Res. 104, 301-307. Perrera, F., Poirier, M., Vespa, S., Nakayama, J., Jaretzki, A., Curnen, M., Knowles, D., and Weinstein, I. (1982). Carcinogenesis 3, 1405-1410. Pettijohn, D., and Hanawalt, P. (1964).J . Mol. Biol. 9, 395-410. Piece, J., Heller, W., and Goldthwart, A. (1983). Carcinogenesis 4, 145-152. Piette, J., and Hearst, J. (1983). Proc. Natl. Acad. Sci. U S A . 80, 5540-5544. Poirier, M. (1981).]. Natl. Cancer Znst. 67,517-519. Prakash, L. (1975).J . MoZ. Biol. 98, 781-795. Price, J., Heller, E., and Goldthwait, D. (1983). Carcinogenesis 4, 145-152. Purnell, M., and Whish, W. (1980). Biochem. J . 185, 775-777. Rabinowitz, M. (1941).J . Morphol. 69, 1-49. Rabkin, S., and Strauss, B. (1984).J. Mol. B i d . 178, 569-594. Rabkin, S. D., Moore, P. D., and Strauss, B. S. (1983).Proc. NatZ. Acad. Sci. U S A . 80, 1541-1545. Radany, E., and Friedberg, E. (1980).Nature (London) 266, 182-185. Radman, M., and Wagner, R. (1984).Curr. Top. Microbiol. Zmmunol. 108,23-28. Rajagopalan, R., Melamede, R., Laspia, M., Wallace, S., and Erlanger, B. (1984).Radiat. Res. 97,499-510. Razzaque, A., Mizusawa, H., and Seidman, M. (1983).Proc. Natl. Acad. Sci. U S A . 80, 3010-3014. Razzaque, A., Chakrabarti, S., Joffe, S., and Seidman, M. (1984).Mol. Cell. Biol. 4,435441. Reddy, M., Gupta, R., Randerath, E., and Randerath, K. (1984).Carcinogenesis 5,231243. Regan, J., and Setlow, R. (1974).Cancer Res. 34,3318-3325. Regan, J., Setlow, R., and Ley, R. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 708-712. Reznikoff, C., Brankow, D., and Heidelberger, C. (1973a).Cancer Res. 33,3231-3238. Reznikoff, C., Bertrtam, J., Brankow, D., and Heidelberger, C. (1973b).Cancer Res. 33, 3239-3249. Rich, A., Nordheim, A., and Wagner, A. (1984).Annu. Rev. Biochem. 53,791-846. Robbins, J., Polinsky, R.,and Moshell, A. (1983).Ann. Neurol. 13(6),682-684. Royer-Pokora, B., and Haseltine, W. (1984).Nature (London) 311, 390-392. Rozenkranz, H., and Rozenkranz, S. (1969). Biochim. Biophys. Acta 195,266-267. Rubin, J., Joyner, A., Bernstein, A., and Whitmore, G. (1983). Nature (London) 306, 206-208. Rubin, J., Prideaux, V., Willard, H., Dulhanty, A., Whitmore, G., and Bernstein, A. (1985).Mol. Cell. B i d . 5, 398-405. Ruby, S., and Szostak, J. (1985).Mol. Cell. B i d . 5, 75-84. Sagher, D., and Strauss, B. (1983).Biochemistry 22,4518-4526. Samson, L., and Cairns, J. (1977).Nature (London) 267,281-283.

DNA REPAIR

103

Samson, L., and Schwartz, J. (1980).Nature (London) 287,861-863. Sancar, A., and Rupp, W. (1983). Cell 33,249-260. Santella, R., Grunberger, D., Broyde, S., and Hingerty, B. (1981).Nucleic Acids Res. 9, 5459-5467. Santella, R., Lin, C., Cleveland, W., and Weinstein, I. (1984). Carcinogenesis 5, 373377. Sarasin, A., and Hanawalt, P. (1978). Proc. Natl. Acad. Sci. U S A . 75, 346-350. Sarasin, A., Bourre, F., and Benoit, A. (1982). Biochimie 64,815-821. Sarkar, S., Dasgupta, U., and Summers, W. (1984). Mol. Cell. B i d . 4, 2227-2230. Schaaper, R., Glickman, B., and Loeb, L. (1982). Cancer Res. 42,3480-3485. Schaaper, R., Kunkel, T., and Loeb, L. (1983).Proc. Natl. Acad. Sci. U S A . 80,487-491. Scheurmann, R., Tam, S., Burgers, P., Lu, C., and Echols, H. (1983).Proc. Natl. Acad. Sci. U . S A . 80, 7085-7089. Schorpp, M., Mallick, U., Rahmsdorf, H., and Herrlich, P. (1984). Cell 37, 861-868. Schwartz, J., and Samson, L. (1983). Mutat. Res. 119,393-397. Scudiero, D., and Strauss, B. (1974).J. Mol. B i d . 83, 17-34. Scudiero, D., Henderson, E., Norin, A., and Strauss, B. (1975).Mutat. Res. 79,473-488. Scudiero, D., Norin, A., Karran, P., and Strauss, B. (1976).Cancer Res. 36, 1397-1403. Scudiero, D., Meyer, S., Clatterbuck, B., Mattem, M., Ziolkowski, C., and Day, R. S., 111. (1984a).Cancer Res. 44, 961-969. Scudiero, D., Meyer, S., Clatterbuck, B., Mattern, M., Ziolkowski, C., and Day, R. S., 111. (1984b). Cancer Res. 44,2467-2474. Seawell, P., Smith, C., and Ganeson, A. (1980).J. Virol. 35, 790-797. Sedgwick, B. (1983). Mol. Gen. Genet. 191,466-472. Seeberg, E., and Steinum, A. (1982). Proc. Natl. Acad. Sci. U S A . 79,988-992. Seeberg, E., Nissin-Meyer, J., and Strike, P. (1976).Nature (London) 263, 524-526. Setlow, R., and Carrier, W. (1964).Proc. Natl. Acad. Sci. U S A . 51,226-231. Shiloh, Y., and Becker, Y. (1982). Biochim. Biophys. Acta 721,485-488. Shiloh, Y., Tabor, E., and Becker, Y. (1983a).Mutat. Res. 112,47-58. Shiloh, Y., van der Schans, G., Lohman, P., and Becker, Y. (1983b). Carcinogenesis 4, 917-921. Sidik, K., and Smerdon, M. (1984). Carcinogenesis 5,245-253. Siede, W., and Eckhardt, F. (1984). Mutat. Res. 129, 3-11. Simon, L., Hazard, R., Maher, V., and McCormick, J. (1981).Carcinogenesis 2,567-570. Sims, J., Juarez-Salinas, H., and Jacobson, M. (1980).Anal. Biochem. 106,296-306. Sims, J., Sikorski, G., Catino, D., Berger, S., and Berger, N. (1982). Biochemistry 21, 1813-1821. Sims, J., Berger, S., and Berger, N. (1983). Biochemistry 22,5188-5184. Singer, B., and Brent, T. (1981). Proc. Natl. Acad. Sci. U.S.A.78,856-860. Singer, B., and Grunberger, D. (1983). “Molecular Biology of Mutagens and Carcinogens.” Plenum, New York. Singer, B., Abbott, L., and Spengler, S. (1984). Carcinogeesis 5, 1165-1171. Sirover, M., and Gupta, P. (1983). In “Human Carcinogenesis” (C. Harris and H. Autrup, eds.), pp. 255-280. Academic Press, New York. Skidmore, C., Davies, M., Goodwin, P., Halldorsson, H., Lewis, P., Shall, S., and Zia’ee, A. (1979).Eur. J . Biochem. 101, 135-142. Sklar, R. (1978).J . Bacteriol. 136,460-462. Sklar, R., and Strauss, B. (1981). Nature (London) 289,417-420. Sklar, R., and Strauss, B. (1983). Cancer Res. 43,3316-3320. Sklar, R., Brady, D., and Strauss, B. (1981). Carcinogenesis 2,1293-1298.

104

BERNARD S. STRAUSS

Slattery, Y., Dignam, J., Matsui, T., and Roeder, R. (1983).J . Biol. Chem. 258, 59555959. Smerdon, M., and Lieberman, M. (1978). Proc. Natl. Acad. Sci. U S A . 75,4238-4241. Smith, A. (1980). In “Methods in Enzymology” (G. Lawrence and K. Moldave, eds.), Vol. 65, pp. 560-580. Academic Press, New York. Snyder, R. D. (1984a). Mutat. Res. 131,163-172. Snyder, R. D. (198413). Biochem. P h u m c o l . 33,1515-1518. Snyder, R. D., and Regan, J. D. (1981). Biochem. Biophys. Res. Commun. 99, 10881094. Snyder, R. D., and Regan, J. D. (1982a).Biochim. Biophys. Acta 697,229-234. Snyder, R. D., and Regan, J. D. (198213). Carcinogenesis 3,7-14. Snyder, R., Carrier, W., and Regan, J. D. (1981).Biophys. J . 35,339-350. Snyder, R., van Houton, B., and Regan, J. (1982).Nucleic Acids Res. 10,6207-6219. Squires, S., Johnson, R., and Collins, A. (1982).Mutat. Res. 95,389-404. Stacks, P., White, J., and Dixon, K. (1983).Mol. Cell. Biol. 3, 1403-1411. Stefanini, M., Reccser, A., and Bootsma, D. (1982).Somatic Cell Genet. 8,635-642. Strauss, B. (1968).Curr. Top. Microbiol. Zmmunol. 44, 1-85. Strauss, B. (1981).In “DNA Repair: A Laboratory Manual of Research Procedures” (E. Friedberg and P. Hanawalt, eds.), Vol. l B , pp. 319-339. Dekker, New York. Strauss, B. (1984).Carcinogenesis 5,577-582.Strauss, B., Altamirano, M., Bose, K., Sklar, R., and Tatsumi, K. (1979).In “Carcinogens: Identification and Mechanisms of Action” (C. A. Griffen and C. Shaw, eds.), pp. 229-250. Raven, New York. Strauss, B., Rabkin, S., Sagher, D., and Moore, P. (1982).Biochimie 64, 829-838. Strauss, B., Ayres, K.,Bose, K., Moore, P., Rabkin, S., Sklar, R., and Lindgren, V. (1983). In “Chromosome Mutation and Neoplasia” (J. German, ed.), pp. 281-315. Liss, New York. Strickland, P., and Boyle J. (1981).Photochem. Photobiol. 34,595-601. Sugimura, T., and Miwa, M. (1983).Carcinogenesis 4,1503-1506. Surowy, C., and Berger, N. (1983).Biochim. Biophys. Acta 740,8-18. Suzuki, N. (1984).Mutat. Res. 125,55-63. Swenberg, J. (1985).“The Role of Cyclic Nucleic Acid Adducts in Carcinogenesis and Mutagenesis” (B. Singer and H. Bartsch, eds.). IARC Publ., in press. Swenberg, J., Bedell, M., Billings, K., Umbenhauer, D., and Pegg, A. (1982).Proc. Natl. Acad. Sci. U S A . 79,5499-5502. Swenberg, J., Dyroff, M., Bedell, M., Popp, J., Huh, N., Kirstein, V., and Rajewsky, M. (1984).Proc. Natl. Acad. Sci. U S A . 81, 1692-1695. Tanaka, K., Hayakawa, H., Sekiguchi, M., and Okada, Y. (1977).Proc. Natl. Acad. Sci. U S A . 74,2958-2962. Teebor, G., and Frenkel, K. (1983).Ado. Cancer Res. 38,23-59. Th‘ng, J., and Walker, I. (1983).Carcinogenesis 4, 975-978. Thompson, L., Busch, D., Brookman, K., Mooney, C., and Glaser, D. (1981).Proc. Natl. Acad. Sci. U S A . 78,3734-3737. Thompson, L., Brookman, K., Dillehay, L., Mooney, K., andcarrano, A. (1982).Somatic Cell Genet. 8,759-773. Thompson, L., Salazar, E., Brookman, K., and Hay, C. (1983).Mutat. Res. 112,329-344. Timson, J. (1975).Mutat. Res. 32, 115-132. Toorchen, D., and Topal, M. (1983).Carcinogenesis 4,1591-1597. Van der Laken, C., Hagenaars, A., Hermsen, G., Kreik, E., Kuipers, A., Nagel, J., Scherer, E., and Welling, M. (1982).Carcinogenesis 3,569-572.

DNA REPAIR

105

Vassilev, L., and Russev, G. (1984). Biochem. Biophys. Acta 781, 39-44. Villani, G., Boiteaux, S., and Radman, M. (1978).Proc. Natl. Acad. Sci. U.S.A.75,30373041. Waldmann, T., Miseti, J., Nelson, D., and Kraemer, K. (1983).Ann. Intern. Med. 99, 367-379. Waldstein, E., Cao, E., and Setlow, R. (1982a).Nucleic Acids Res. 10,4595-4604. Waldstein, E., Cao, E., and Setlow, R. (1982b). Proc. Nutl. Acad. Sci. U.S.A.79, 51175121. Waldstein, D., Cao, E., Bender, M., and Setlow, R. (1982~).Mutat. Res. 95, 405-416. Walker, G. (1984). Microbiol. Reu. 48,60-93. Walker, J., and Th’ng, J. P. (1982). Mutat. Res. 105, 277-285. Wani, A., Gibson-D’Ambrosio, R., and D’Ambrosio, S. (1984). Photochem. Photobiol. 40,465-472. Weaver, D., and DePamphlis, M. (1982).J . Biol Chem. 257,2075-2086. Weigle, J. (1953).Proc. Natl. Acad. Sci. U.SA. 39,628-636. Weinberg, G., Ullman, B., and Martin, D., Jr. (1981). Proc. Natl. Acad. Sci. U S A . 78, 2447-2451. Westerveld, A., Hoeijmakers, J., van Duin, M., de Wit, J., Odijk, H., Pastink, A., Wood, R., and Bootsma, D. (1984).Nature (London) 310,425-429. White, J., and Dixon, K. (1984). Mol. Cell. Biol. 4, 1286-1292. Wiestler, O., Kleihues, P., and Pegg, A. (1984). Carcinogenesis 5, 121-124. Wild, C., Smart, G., Saffhell, R., and Boyle, J. (1983). Carcinogenesis 4, 1605-609. Williams, 1.. and Friedberg, E. (1979). Biochemistry 18,3965-3972. Witkin, E. (1976). Bacteriol. Reu. 40, 869-907. Wu, J., and Manuelidis, L. (1980).J . Mol. Biol. 142, 363-386. Yagi, T. (1982). Mutat. Res. 96,89-98. Yagi, T., and Takebe, H. (1983).Mutat. Res. 112,59-66. Yarosh, D. (1985). Mutat. Res. 145, 1-16. Yarosh, D., Foote, R., Mitra, S., and Day, R. (1983). Carcinogenesis 4, 199-205. Yarosh, D., Rice, M., Day, R., Foote, R., and Mitra, S. (1984a).Mutat. Res. 131,27-36. Yarosh, D., Scudiero, D., Ziolkowski, D., Rhim, J., and Day, R. (1984b). Carcinogenesis 5,627-633. Yoshida, S., Koiwai, O., Suzuki, R., and Tada, M. (1984). Cancer Res. 44, 1867-1870. Zarkzewski, S., Koch, M., and Sperling, K. (1983). Hum, Genet. 64,55-57. Zelle, B., Berends, F., and Lohman, P. (1980).Mutat. Res. 73, 157-169. Zlotogorski, C., and Erickson, L. (1983). Carcinogenesis 4,759-763. Zlotogorski, C., and Erickson, L. (1984). Carcinogenesis 5,83-37. Zolan, M., Cortopassi, G., Smith, C., and Hanawalt, P. (1982a). Cell 28, 613-619. Zolan, M., Smith, C., Calvin, N., and Hanawalt, P. (1982b).Nature (London)299,462464. Zolan, M., Smith, C., and Hanawalt, P. (1984). Biochemistry 23,63-68. Zurlo, J., and Yager, J. (1984). Carcinogenesis 5, 495-503.

This Page Intentionally Left Blank

THE BLYM ONCOGENES Paul Neiman Fred Hutchinson Cancer Research Center. Seattle, Washington

I. Definitions and Significance ........................................ 11. Oncogenes in Bursal Lymphomas.. .................................. A. Pathogenesis of Bursal Lymphomas Induced in Chickens by Avian Leukosis Virus ................................ B. Detection of Candidate Oncogenes in Bursal Lymphomas . . . . . . . . . . . . C. CloningofChEZym-1 ............................................ D. Stage-Specific Activation of Oncogenes in Bursal Lymphomagenesis . . . 111. Blym-1 Oncogenes in Human Burkitt’s Lymphomas .................... A. Burkitt’s Lymphoma Contains Two Activated Oncogenes. . . . . . . . . . . . . B. Cloning of HuBZym-1 ........................................... C. Do EZym Oncogenes Play a Lineage and Stage-Specific Role?. ........ IV. Issues for Continued Investigation. .................................. References .......................................................

107 108 108 109 110 112 114 114 115 118

120 121

I. Definitions and Significance

In present usage, oncogenes are a set of cellular and viral genes which induce morphologic transformation and/or neoplastic changes in cells in which they are either activated or introduced in activated form. Their oncogenic properties are defined by the particular cells and biological test systems which are used to detect their presence. This review will summarize what is presently known about a single class of oncogenes designated Blym. These genes are the best characterized of a subset of oncogenes which were not initially discovered within the genomes of oncogenic viruses. They were first detected by the use of a transfection assay in which DNA from tumor cells transforms cultured NIH/3T3 mouse fibroblasts as illustrated in Fig. 1 (Lowy et al., 1978; Copeland et al., 1979; Goldfarb and Weinberg, 1979). Unlike other oncogenes which also transform NIH/3T3 cells, such as the M S family which has been found in association with tumors in a wide variety of cell lineages (Land et al., 1983; Cooper and Lane, 1984),Blym oncogenes have been detected so far only in B cell neoplasms in both animals (chickens) and humans. Although the characterization of Blym genes is not as advanced as that of a number of other oncogenes, their special significance lies in the accumulating 107 ADVANCES IN CANCER RESEARCH, VOL. 45

Copyright 0 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.

108

PAUL NEIMAN

Normal 3T3 Monolayer

Foci or Colonies of Transformed Cells

DNA ( Active-TransformingGene) CaC12 ppt

I

Nontumorigenic

c

Tumorigenic

FIG.1. The basic technique for detecting oncogenes by transfection of NIH/3T3 mouse cells. Monolayers of cultured NIH/3T3 cells are exposed to CaClzprecipitates of DNA containing an activated transforming gene (as described Lowy et al., 1978; Copeland and Cooper, 1980; Goldfarb and Weinberg, 1979). Foci of transformed cells or soft agar colonies appear in the cultures by 2-3 weeks after transfection. The presence of the transforming gene can be tested by blot hybridization analysis of DNA from the cultured transformed cells if appropriate hybridization probes are available. Spontaneous transformants can also usually be eliminated by secondary transfection assays, since DNA from such cells does not transform in this assay (Copeland et al., 1979; Cooper et nl., 1980). The tumorigenicity.of these cells can also be tested in uiuo, although often immunodeficient animals need to be used, and spontaneous transformants in control NIH/3T3 cell populations may present a serious problem in interpretation.

evidence that the role these genes play in tumorigenesis may be both lineage and stage specific.

II. Oncogenes in Bursa1 Lymphomas

A. PATHOGENESIS OF BURSAL LYMPHOMAS INDUCED IN CHICKENS BY AVIANLEUKOSIS VIRUS The induction by avian leukosis virus (ALV) of lymphomas in the bursa of Fabricius in chickens (Rubin et al., 1961; Fenner, 1976; Purchase and Burmester, 1978; Peterson et al., 1964,1966; Cooper et al., 1968) has been a revealing experimental system for the study of oncogenes. Infection of newly hatched chickens with this retrovirus results in the development of lymphomas which are fatal in 4-6 months. Serial histologic study of the pathogenesis of this tumor within the bursa has revealed several distinct stages. The earliest observable change is the proliferation of lymphoblasts within individual bursa1

THE BLYM ONCOGENES

109

follicles which is easily marked because their cytoplasm stains intensely red with methyl green pyronin (MGP) (Cooper et al., 1968; Neiman et aZ., 1980a). These abnormal follicles have been called transformed follicles. Serial sectioning indicates that maximally 0.020.2% of the bursal follicles is transformed in susceptible birds 4-8 weeks after hatching and infection with ALV (Neiman et al., 1980a). Because these proliferative lesions remain confined to follicles and appear to regress along with normal bursal follicles under the influence of normal physiologic controls, they have been felt to represent a preneoplastic stage in tumor development (Neiman et al., 1980a). One (or rarely two) of these preneoplastic-transformedfollicles will progress to form the next recognizable stage which is a discretely expanding bursal nodule (Neiman et al., 1980a). These first distinctly neoplastic lesions are clonal growths (Neiman et al., 1980b) composed of pyroninophilic lymphoblasts morphologically similar to transformed follicle cells. These discrete nodules appear as early as 10 weeks after infection and may remain discrete for many weeks before progressing to form lymphomas which metastasize widely within the bursa and to other organs. Mortality from these lymphomas in experimentally infected chicken flocks increases at about the sixth power of time similar to many other experimental and naturally occurring tumor systems (Neiman et al., 1980a). This virally induced neoplasm, therefore, demonstrates the usual features of multistage tumorigenesis. OF CANDIDATE ONCOGENES IN B. DETECTION

BURSALLYMPHOMAS

Bursa1 nodules represent the earliest stage of lymphomagenesis at which detailed molecular studies have been carried out. These lesions contain two activated oncogenes which are strong candidates to be responsible for some or all of the neoplastic phenotypes. Cooper and Neiman (1980) observed that DNA from all bursal nodules, bursal lymphomas, and a cell line derived from a bursal lymphoma called RP-9 which they studied transformed NIH/3T3 cells in transfection assays. DNA from adjacent normal bursal tissue and from other organs in infected birds did not transform in this assay. Blot hybridization analysis of secondary NIH/3T3 transformants failed to detect transfer of any known oncogene or any ALV-related sequences, including long terminal repeat (LTR) sequences which bear the proviral promoter and enhancer. The transforming gene involved, therefore, bears no obvious relationship to the second activated oncogene c-myc. This

110

PAUL NEIMAK

gene is the cellular homolog of the v-myc-transforming gene of the avian myelocytomatosis virus family of acute leukemia retroviruses. In over 90% of advanced bursal lymphomas, the c-myc locus was found to be transcriptionally activated as the result of a nearby integration of ALV proviral promoter and/or enhancer sequences (Hayward et al., 1981; Nee1 et ul., 1981; Payne et al., 1982). Subsequently, the same proviral integration events were detected in early bursal nodules (Cooper and Neiman, 1981). The precise role, if any, of activated c-myc- and the NIH/3T3-transforming gene in lymphomagenesis remains to be defined. It has been proposed that these genes might act at different stages of tumor development (Cooper and Neiman, 1981). C. CLONING OF ChBZym-1 A gene designated ChBlym-1 has recently been identified by molecular-cloning techniques which is believed to encode the NIH/3T3transforming activity detected in bursal lymphoma DNA (Goubin et aZ., 1983).The clone was derived from a charon phage library of DNA from NIH13T3 cells transformed by DNA from the RP-9 bursal lymphoma cell line. ChBZym-1 was identified by the technique of sib selection using the NIH/3T3 transfection assay. The identified clone transformed NIH/3T3 cells with an efficiency of several thousand foci per microgram of DNA. This efficiency is comparable to that of known viral oncogenes, such as src or rus, although the induced foci tend to be somewhat smaller and less refractile than those seen with src or rus (Fig. 2). In DNA blot hybridization analyses using a ChBZym-1 probe, the gene was found to be similar to other oncogenes in being highly conserved in vertebrate species, but, unlike most other oncogenes, appeared to belong to a family of closely related DNA sequences of about 5-10 copies (Goubin et al., 1983). An inconvenient partial homology was also detected between ChBlym-1 and a highly repeated mouse DNA sequence which prevented analysis of blot hybridization experiments with DNA from NIH/3T3 cells transformed by DNA from other bursal lymphoma cells. Therefore, it is not known whether the same transforming gene was present in all of the bursal lymphoma DNAs. Sequence analysis of the ChBlym-1 clone revealed a putative small polypeptide of 65 amino acids encoded in two exons preceded at the 5' end by typical promoter consensus sequences and followed at the 3' end by poly(A) addition sequences. A detailed sequence comparison with known oncogenes failed to reveal any significant homology.

THE BLYM ONCOGENES

111

FIG.2. Foci of transformed NIH/3T3 cells from transfection assays employing cloned HuBlym (A and C) and rasH (B and D) oncogenes. Foci at 12 days after transfection are ) Note the smaller size shown at low (A and B, 2 0 X ) and higher (C and D, 5 2 ~power. and less refractile nature of the B l y m foci. Photographs courtesy of Dr. Geoffrey Cooper.

112

PAUL NEIMAN

However, a broader search revealed significant homology between the amino acids encoded in the second and larger exon of the putative Blym protein and the amino-terminal sequences of the secreted forms of the transferrin family of proteins (Goubin et al., 1983).The degree of this homology is as strong between ChBlym-1 and transferrin as it is between this domain of transferrin and the other authentic members of the transferrin family (lactoferrin, ovotransferrin, and the melanoma-associated antigen, p97). The degree of divergence and the small size of the putative Blym protein relative to transferrins make it difficult to predict with certainty whether this apparent ancestral relationship to a transferrin domain has any functional significance. However, the accumulated evidence that transferrins are important growth factors for cells in culture (Barnes and Sato, 1980)and for proliferation during embryonic development (Ekblom et al., 1983), taken together with the observation that transferrin can serve as a lymphocyte mitogen independent of iron delivery (Tormey et al., 1972; Dillner-Centerlind et al., 1979; Trowbridge and Domingo, 1981; Trowbridge and Lopez, 1982; Trowbridge et al., 1982; Mendelsohn et al., 1983; Taetle et al., 1983), makes the speculation that Blym genes act via a pathway related to transferrin an attractive one.

D. STAGE-SPECIFIC ACTIVATIONOF ONCOGENES IN BURSAL LYMPHOMAGENESIS The observation of activated oncogenes in bursal lymphomas leaves open the question of whether they are actually involved in the initiation or maintenance of these tumors. In a formal sense, this issue remains to be resolved for the tumor-associated oncogenes in general (Duesberg, 1983; Albino et al., 1984).One strategy which would more firmly establish their role would be the experimental introduction of these genes into biological systems which would allow the observation of their effects on tumor development. It is possible to exploit this strategy in the chicken bursa system. The work of Eskola and Toivanen (1974) established conditions for the ablation of embryonic bursal lymphocytes with the alkylating agent cyclophosphamide. It is possible to reconstitute normal bursal follicles by injection of small numbers of normal embryonic bursal lymphocytes into the chorioallantoic vein, and Neiman et al., (1985) have recently conducted studies in which cyclophosphamide-ablated embryonic bursas were reconstituted either with normal bursal stem cells or with the same cells infected with an avian myelocytomatosis virus called HB1 which carries an active v-myc oncogene (Ramsay et al., 1982; Enrietto et al.,

THE BLYM ONCOGENES

113

1983).When normal stem cells were transplanted, large pyroninophilic lymphoblasts morphologically simiIar to transformed follicle cells were first observed to accumulate near the basement membrane delimiting the lymphocyte-depleted follicular medulla. These cells appeared to serve as precursors for the smaller and less pyroninophilic lymphocytes which repopulate first the follicular medulla and then the cortex. When the bursal inoculum was infected with HB1, the same large pyroninophilic cells first appeared near the medullary basement membrane. These v-myc-bearing cells failed to differentiate and appeared simply to proliferate within both the medullary and cortical regions of the bursal follicles forming structures which were indistinguishable from the transformed follicles observed during the first stages of lymphomagenesis in ALV-infected birds. The large number of transformed follicles obtained by this technique, up to 100% by 2-4 weeks after hatching, provided the first opportunity to study these preneoplastic cells. Southern blot analysis showed that D N A from these transformed follicle cells contained the HB1 myc gene, but failed to transform NIH/3T3 cells in transfection assays (Neiman et al., 1985).Therefore, these apparently preneoplastic lymphoblasts lacked an activated Blym-1 or similar NIH/3T3-transforming gene, but contained a v-myc gene which, typically, scores poorly if at all in the NIH/3T3 assay (Copeland et al., 1979).A conservative conclusion from these studies to date is that the phenotype of bursal lymphocytes containing at least one type of activated myc gene (the HB1 v-myc gene) and lacking a Blym-l-like gene is similar or identical to the pyroninophilic lymphoblasts of preneoplastic-transformed follicles. This conclusion suggests that the role of Blym-1 or other NIH/3T3-transforming genes, if any, lies in progression toward the neoplastic bursal nodule stage of lymphomagenesis. The concept that activated myc genes are not sufficient for full neoplastic change received further support from the observation that a constitutionally deregulated myc gene introduced into transgenic mice acts as a heritable predisposing factor requiring further transforming events for full development of cancer (Stewart et al., 1984). It is curious that, like other v-myc-containing viruses, HB1 induces fairly acute neoplasms in other organs over the same time period during which it induces only preneoplastic changes in the bursa (Enrietto et al., 1983; Neiman et al., 1985).It is possible that activated myc genes can induce different degrees of neoplastic transformation in different cell lineages. Alternatively, second oncogenes may also play a role in the more advanced tumors associated with infection by v-myc-containing viruses.

114

PAUL NEIMAN

These transplantation experiments with the chicken bursa were initial efforts and are still subject to significant limitation in interpretation. For example, the v-myc gene of HB1 is structurally somewhat different than the LTR-activated c-myc locus observed in ALV-induced lymphomas and could, in principle, act in a different fashion. The HBl v-myc-induced transformed follicles have not yet been shown to progress to bursal lymphomas in a manner analogous to that observed in ALV-induced disease with the appearance of activated Blym-1 or other NIH/3TS-transforming genes in more advanced neoplasms. Nevertheless, these early experiments provide a basis for an attractive working hypothesis which can be tested in this model system. In its simplest form, this notion predicts that transcriptional activation of myc in a particular type of bursal lymphoctyte, either by LTR insertion or by transduction as part of a retroviral genome, results in the proliferation of preneoplastic pyroninophilic lymphoblasts which retain their ability to home to bursal follicles and are limited in their growth to the follicular architecture. Progression occurs when Blym becomes activated in one of these transformed follicle cells leading to clonal neoplastic growth. Since these early neoplasms appear to be discrete premetastatic growths (bursal nodules), activation of additional genes may be required for further progression to fully malignant-metastasizing tumors. One intriguing possibility is that further mutations at the Blym and/or myc locus might play a role in progression. This possibility is suggested by sequence analysis of LTR-activated c-myc genes cloned from ALV-induced lymphomas (Westaway et aZ., 1984) which detected the presence of potentially significant mutations. Studies involving the introduction into bursal stem cells of various oncogenes in retroviral constructs may provide further insights into the validity of this model.

111. Slym-1 Oncogenes in Human Burkitt’s Lymphomas

A. BURKITT’S LYMPHOMA CONTAINS Two ACTIVATED ONCOGENES

The observation that ChBlym-1 sequences were evolutionarily conserved in humans (Goubin et al., 1983) provided the opportunity to search for activation of homologs of this transforming gene in human neoplasms. Burkitt’s lymphoma was an attractive tumor to investigate in this regard for two general reasons. Like avian bursal lymphomas,

THE BLYM ONCOGENES

115

Burkitt’s lymphomas are composed of pyroninophilic B lymphoblasts at an intermediate stage of differentiation which express surface immunoglobulin (Klein et al., 1968). Also like bursal lymphomas, Burkitt’s lymphomas contain an altered m y c locus. This alteration results not from a retroviral LTR insertion, but rather from a chromosomal translocation in which the c-myc region on chromosome 8 recombines with an immunoglobulin gene locus on chromosome 14, or less frequently on chromosomes 2 or 22 (Dalla-Favera et al., 1982; Taub et al., 1982; Nee1 et al., 1982; Marcu et aZ., 1983). While the details of myc activation in Burkitt’s lymphoma are the subject of intense investigation which is beyond the scope of this review, the evidence suggests that the m y c alleles in the rearranged chromosomes of these tumor cells are regulated abnormally (Nishikura et al., 1983; ar-Rushdi et al., 1983; Maguire et al., 1983). These characteristic recombination events and their effects on m y c gene expression occur in both Epstein-Ban- virus (EBV) positive and negative lymphomas. Transfection assays of DNA from Burkitt’s lymphoma cell lines with characteristic cytogenetic changes regularly demonstrate transformation of NIH/3T3 cells, and the transforming activity of DNA from different cell lines shows the same pattern of sensitivity to inactivation by digestion with a battery of restriction endonucleases (Diamond et al., 1983). Transforming activity is present in DNA from both EBV positive and negative tumor cell lines. Blot hybridization analysis of DNA from NIH/3T3 cells transformed by Burkitt’s lymphoma DNA failed to detect the transfer of either EBV-related sequences or human c-myc of any other known oncogene. In contrast, DNA from EBV-immortalized human umbilical cord blood B lymphocytes does not contain NIH/3TS-transforming genes (Diamond et al., 1983). Therefore, like chicken bursal lymphomas, Burkitt’s lymphomas contain at least two distinct tumor-specific oncogenes which are strong candidates for a role in tumorigenesis. OF HuBZym-1 B. CLONING

The homology between ChBlym-1 probes and human DNA sequences made it possible to determine whether the transforming genes transferred from Burkitt’s lymphoma cells to NIH/3T3 cells in transfection assays were related to ChBZym-1. Using a ChBZym-1 probe to screen a lambda phage library of DNA from the Burkitt’s lymphoma line CW678, 15 cross-hybridizing clones were selected. One of these clones contained DNA which transformed NIH/3T3 cells with an efficiency comparable to that of ChBlym-1 and other onco-

116

PAUL NEIMAN

genes which are highly transforming for NIH/3T3 cells (Diamond et al., 1983).This clone is designated HuBZym-1. The transforming gene resides on a 0.95-kb EcoRI DNA fragment which cross hybridizes with ChBZym-1 probes. This cross-reaction is relatively weak, however, and requires relatively low-stringency conditions (Diamond et al., 1983). Under these low-stringency conditions, HuBZym-1 probes cross-react with a middle repetitive family of sequences in human DNA, but, under conditions of high stringency, a single HuBEym-1 locus is identified which does not appear grossly rearranged in Burkitt’s lymphoma DNA. Zn situ hybridization studies using the cloned transforming DNA fragment as a probe mapped HuBZym-1 to lp32 on the short arm of chromosome 1 (Morton et aZ., 1984). This single copy character is in contrast to ChBZym-1 which, as mentioned, appears to belong to a small family of conserved sequences which can be detected in human and chicken DflA under stringent conditions. Taken together with the relatively weak hybridization between human- and chicken-transforming ChBZym clones, these observations suggest that the transforming gene activated in CW678 cells, while clearly related to ChBZym-1, is not among the most closely related Blym sequences in human DNA. Additional evidence supports the conclusion that HuBZym-1 is the NIHI3T3-transforming gene detected in transfection experiments with Burkitt’s lymphoma DNA from various cell lines (Diamond et al., 1983). Blot hybridization analysis of NIH/3T3 transformants induced by transfection with DNA from CW678 and from a second lymphoma cell line EW36 demonstrated the transfer of a 5-kb Hind111 DNA fragment identical with that detected by HuBlym-1 probes in DNA from both normal and lymphoma DNA from the several Burkitt’s lymphoma cell lines. The minimal conclusion from these studies is that NIH/3T3-transforming genes belonging to the B l y m family are somehow activated in B cell neoplasms composed of cells at about the same state of differentiation in both chickens and man. In man, this appears to be a frequent if not uniform event, although the studies with established cell lines need to be confirmed with DNA from fresh tumor tissue, and the presence of an activated oncogene from the M S family has also been described in a Burkitt’s lymphoma cell line (Murray et al., 1983).I n chickens, activation of an NIHi3T3-transforming gene is also a frequent or uniform event, although it remains to be determined whether it is often or always a Blym gene. Determination of the sequence of HyBZym-1 provided further details of the relationship between the human- and chicken-transforming genes (Diamond et uZ., 1984).A comparison of the putative struc-

117

THE BLYM ONCOGENES

ture of the two genes and the polypeptides that they appear to encode is shown in Fig. 3. As was observed with ChBZym-1, the coding regions of HyBlym-1 are distributed in two exons flanked at the 5’ end by a strong “TATA” homology suggesting a promoter and at the 3’ end by a poly(A) addition sequence. The overall DNA sequence homology between HuBlym-1 and ChBZym-1was 50-60% with several short stretches approaching 75% homology. This degree of sequence matching is consistent with the previously mentioned requirement for low-stringency conditions in cross hybridization reactions. The HuBZym-1protein inferred form the DNA sequence is composed of 58 amino acids (compared to 65 for ChBlym-1) of which 17 (30%)are shared with the putative ChBZym-1 protein. Both Blym proteins appear to be small basic polypeptides containing 21% arginine plus lysine residues. As shown in Fig. 3, the homology observed between ChBZym-1 and the amino-terminal sequences of the transferrin family is also preserved in the HuBZym-1 protein. Furthermore, the protein sequence homology between the Blym genes tends to overlap that between the Blym proteins and the transferrins in that, of the 17 amino acids shared between the putative Blym proteins, seven are shared with the transferrins. Finally, the segment of amino-terminal IBlym-1 : CCAAAT

AATATTAAA

AATAAA

rBlym-1: CCAT CCAT AATATTTA I

!

$ATAAA

I

MTLRGLRLQWREOLKM!

ARPCLL~K_R_~KIVSYI~FLL~~LKGTLAIDSLYSLOFAGGN

FIG.3. Comparative structure of the ChBlym-1 and the HuBlym-1 oncogenes. The numbering indicates nucleotide positions for ChBlym-1 (Goubin et al., 1983) and HuBlym-1 (Diamond et al., 1984).The open boxes represent the putative exons of the genes and the darkened segments represent the protein-coding sequences. The polypeptides inferred from the sequence are shown in single letter amino acid codes. Amino acids shown in boldface are shared between the two putative Blym proteins. Amino acids underlined are shared with one or more members of the transferrin family.

118

PAUL NEIMAN

sequences in transferrin which contains the amino acids shared with Blym is also relatively conserved among the other authentic members of the transferrin family, ovotransferrin, lactoferrin, and p97 (Goubin et al., 1983; Diamond et al., 1984). These observations suggest not only an ancestral relationship, but also the possibility of similar selection pressures in the evolution of the Blym genes and the transferrins. This possibility makes it more attractive to search for a functional relationship between the growth factor properties of the transferrin proteins and the mechanism of action of Blym genes.

C. Do Blym ONCOGENES PLAYA LINEAGE AND STAGE-SPECIFICROLE? The discovery of the activation of unique and closely related oncogenes in similar B cell neoplasms of two different species raises the possibility that Blym genes may play a role in tumorigenesis only in B cells at specific stages of differentiation. This possibility seems to be in striking contrast to the implication of other oncogenes, such the TUS family and even myc, in a wide spectrum of tumor types (reviewed in Land et al., 1983; Cooper and Lane, 1984). Data from transfection experiments support the notion of the existence of oncogenes of this type. DNAs from breast cancers and from lymphomas and from leukemias representing discrete stages of B and T lymphocyte differentiation have been found to transform NIH/3T3 cells. Restriction endonuclease inactivation studies suggest that either the same or a very similar transforming gene is present in tumors of the same type, regardless of the species of origin or the mechanism of induction (viral, chemical, or spontaneous), while distinct genes are activated in the tumors representing different lineages and stages of differentiation (Lane et al., 1981, 1982a,b). These genes remain to be characterized further, although a unique NIH/3TS-transfonning gene has recently been cloned from T lymphoma cells (Lane et aE., 1984).Another possible example may be a locus named bcl-1 composed of sequences near the chromosomal breakpoint of the 11:14 translocation in human chronic lymphocytic leukemia of the B cell type (Tsujimoto et al., 1984). Whether such a restricted class of oncogenes is actually instrumental in in vivo tumorigenesis remains to be determined with certainty. The similarity of Burkitt’s lymphoma and avian retrovirus-induced bursa1 lymphomas supports the inclusion of the Blym genes in such a putative class. The pyroninophilic lymphoblasts which make up these tumors bear a striking morphological similarity, and both tumors carry

THE BLYM ONCOGENES

119

principally monoclonal IgM inserted in the cell membrane (Cooper et al., 1974; Gunven et al., 1980). Burkitt’s lymphoma, however, occurs in humans in both children and adults and can be causally associated with a herpesvirus, EBV (for review see Nilsson and Klein, 1982), which does not appear to play a necessary role in the activation of either c-myc or Blym. Bursa1 lymphomas are developmentally restricted tumors which can only be induced by infection of chickens with ALV within a few weeks of hatching (Fadly et al., 1980). Presumably, there is a class of target cells which is present in appreciable numbers for a limited period of time in the developing bursa. These cells could represent a population which is peculiarly susceptible to assuming the phenotype of preneoplastic-transformed follicle cells when subjected to the action of c-myc genes transcriptionally deregulated as a consequence of nearby proviral LTR insertions. The B cellimmortalizing functions of EBV (Miller and Yale, 1971; Rosen et al., 1977; Einhorn and Ernberg, 1978) and/or other viral genes might serve the same role as the putative developmental gene(s) in bursa1 cells by sensitizing human B cells to the transforming effects of c-myc genes deregulated as a result of a chromosomal translocation. In this speculative multistep model, Blym genes may be peculiarly susceptible to mutation in c-myc-altered B cells perhaps because of a normal role in B cell development, proliferation, or function, and/or because of some specific activity of the myc gene. At least one type of Blym alteration appears to cause the gene to acquire NIH/3TS-transforrning activity, a fact which leads to the suggestion that these activated Blym genes mediate progression of these B cell neoplasms (Neiman et al., 1985). While the model described seems to fit the data acquired so far, there clearly are alternative explanations. The essential features of this idea, however, should be testable experimentally. Germane to the issue of lineage specificity of the Blym genes is a survey of biological systems in which their transforming activity can be detected. It might be argued that the fact that Blym oncogenes can be scored in a fibroblastic line like NIH/3T3 mitigates such lineage restriction, at least in an absolute sense. It should be pointed out, however, that other oncogenes that appear lineage restricted in uiuo, such as the v-abl gene of Abelson leukemia virus which induces B cell neoplasms in mice (Abelson and Rabstein, 1970), transform NIH/3T3 cells in uitro. The precise relationship between morphological transformation of cultured fibroblasts and in v i m multistage tumorigenesis has yet to be defined. In this connection, it is interesting that a new technique for detecting the activity of oncogenes employing cultured human skin fibroblasts also detects the induction of anchorage-inde-

120

PAUL NEIMAN

pendent growth by HuBlym-1 (B. Sutherland, personal communication). One principal reason to explore the stage-specific activity of the Blym and myc oncogenes in the development of these B cell neoplasms was the presence of more than one candidate oncogene in the same tumor. It may be relevant to note that the tumor systems discussed in detail in this article are not the only instances in which there is evidence for the activity of two distinct oncogenes. Pre-B cell lymphomas induced in mice by the v-abl gene of Abelson leukemia virus contain a distinct NIH13T3-transforming gene (Lane et at., 1982a). Lymphomas induced by murine leukemia virus and carcinomas caused by mouse mammary tumor virus also contain distinct NIH/ 3T3-transforming genes which are not closely linked to the putative oncogenes identified at proviral integration sites (Lane et al., 1981, 1982b, 1984; Nusse and Varmus, 1982; Peters et al., 1983; Corcoran et al., 1984; Cuypers et al., 1984; Dickson et al., 1984). Although the NIH/3T3-transforming genes associated with these tumor systems are not yet further characterized, there appears to be a number of opportunities to explore the role of oncogenes in multistage carcinogenesis in several experimental systems. IV. Issues for Continued Investigation

A major unresolved issue which applies not only to the Blym oncogenes, but to all the tumor-associated oncogenes is the unequivocal demonstration that they play a central role in the initiation, progression, and/or maintenance of the neoplasms in which they have been identified. Formal genetic experiments have proven such a role for the prototypic retroviral oncogenes of acute transforming viruses, and similarly convincing experimental strategies are needed for tumor systems in which cellular oncogenes are activated in the absence of such viral genomes. Useful strategies directed toward that goal will probably require the introduction of gene constructs into germ cells or somatic tissue stem cells. Some promising early experiments of this type involving the Blym and myc oncogenes are reviewed in the preceding sections. A second general strategy is to advance our knowedge of the normal cellular functions of these genes and their perturbation in neoplastic transformation. Such additional knowledge might make the role of oncogenes in cancer obvious. With respect to the Blym oncogenes, there are major gaps in our knowledge for which conclusive data are not yet available. The normal cellular homologs of these transforming

THE BLYM ONCOGENES

121

genes have not been identified nor have the transcriptional or translational products been identified and definitively characterized. These analyses are complicated by the presence of a number of copies of cross-reacting sequences at the DNA (Goubin et al., 1983), RNA, and protein levels (Neiman et al., unpublished). Therefore, at this writing, there are no unambiguous data which establish the mechanism of either activation or operation of the Blym oncogenes. For all of the reasons reviewed in this article, it can be submitted that the continued investigation of these and other aspects of the biology of the Blym genes will provide further insights into the molecular mechanisms underlying neoplastic change. ACKNOWLEDGMENTS The authors thank Dr. Geoffrey Cooper for helpful suggestions and critical reading of the manuscript and Kay Shiozaki for typing the manuscript. Research work by the author described in the article was supported by Grants CA28151 and CA20068 from the National Cancer Institute.

REFERENCES Abelson, H. T., and Rabstein,’L. S. (1970).Cancer Res. 30,2213-2222. Albino, A. P., Le Strange, R., Oliff, A. I., Furth. M. E., and Old, L. J. (1984).Nature (London)302,69-71. ar-Rushdi, A., Nishikura, K., Erikson, J., Watt, R., Rovera, G., and Croce, C. M. (1983). Science 222,390-393. Barnes, D., and Sato, G. (1980).Cell 22, 649-655. Cooper, G. M., and Lane, M.-A. (1984).Biochim. Biophys. Acta Reo. Cancer 738,9-20. Cooper, G. M., and Neiman, P. E. (1980).Nature (London)278,656-659. Cooper, G . M., and Neiman, P. E. (1981).Nature (London) 292,857-858. Cooper, G. M., Okenquist, S., and Silverman, L. (1980).Nature (London)284,418-421. Cooper, M. D., Payne, L. N., Dent, P. B., Burmester, B. R., and Good, R. A. (1968).J . Natl. Cancer Inst. 41,373-378. Cooper, M. D., Purchase, G. H., Bockman, D. E., and Gathings, W. E. (1974).]. Immunol. 113, 1210-1215. Copeland, N. G., and Cooper, G. M. (1980).1.Virol. 33, 1199-1202. Copeland, N. G., Zelenetz, A. D., and Cooper, G. M. (1979).Cell 17,993-1002. Corcoran, L. M., Adams, J. M.,Dunn, A. R., and Cory, S . (1984).Cell 37, 113-122. Cuypers, H.T., Selten, G., Quint, W., Zilstra, M., Maandag, E. R., Boelens, W., van Wezenbeek, P., Melief, C., and Berns, A. (1984).Cell 37, 141-150. Dalla-Favera, R., Bregni, M.,Erickson, J., Patterson, D., and Gallo, R. C. (1982).Proc. Natl. Acad. Sci. U S A . 79,7824-7827. Diamond, A., Cooper, G. M., Ritz, J., and Lane, M. -A. (1983).Nature (London)305, 112-116. Diamond, A., Devine, J. M., and Cooper, G. M. (1984).Science 225,516-519. Dickson, C., Smith, R., Brookes, S., and Peters, G. (1984).Cell 37,529-536.

122

PAUL NEIMAN

Dillner-Centerlind, M. L., Hammarstrom, S., and Perlmann, P. (1979).Eur.J. Zmmunol. 9,942-948. Duesberg, P. (1983).Nature (London) 304,219-226. Einhorn, L., and Ernberg, 1. (1978). Znt. J. Cancer 21, 157-164. Ekblom, P., Thesleff, I., Saxen, L., Miettinen, A., and Timpl, R. (1983).Proc. Natl. Acad. Sci. U S A . 80,2651-2655. Enrietto, P. J., Payne, L. N., and Hayman M. J. (1983). Cell 35,369-379. Eskola, J., and Toivanen, P. (1974). Cell. Zmmunol. 13,459-471. Fadly, A. M., Purchase, H. G., and Gilmour, D. G. (1980).Ado. Comp. Leukemia Res. Proc. Int. Symp. 9th, Pitsundu, Georgia 1979. pp. 41-42. Fenner, F. (1976). Virology 71, 371-378. Goldfarb, M. P., and Weinberg, R. A. (1979).J. Virol. 32,30-39. Goubin, G., Goldman, D. S., Luce, J., Neiman, P. E., and Cooper, G. M. (1983).Nature (London) 302,114-119. Gunven, P., Klein, G., Klein E., Norris, T., and Singh, S. (1980).Znt. J . Cancer25,711719. Hayward, W. S., Neel, B. G., and Astrin, S. M. (1981). Nature (London) 290,475-480. Klein, E., Klein, G., Nadkarni, J. S., Nadkami, J.J., Wigzell, H., and Clifford, P. (1968). Cancer Res. 28, 1300-1310. Land, L., Parada, L. F., and Weinberg, R. A. (1983). Science 222,771-778. Lane, M.-A., Sainten, A., and Cooper, G. M. (1981). Proc. Natl. Acad. Sci. U S A . 78, 5185-5189. Lane, M.-A., Neary, D., and Cooper, G. M. (1982a). Nature (London) 300,659-661. Lane, M.-A., Sainten, A., and Cooper, G. M. (1982b). Cell 28,873-880. Lane, M.-A., Sainten, A., Doherty, K. M., and Cooper, G. M. (1984). Proc. Natl. Acad. Sci. U S A . 81,2227-2231. Lowy, D. R., Rands, E., and Scolnick, E. M. (1978).J. Virol. 26,291-298. Maguire, R. T., Robins, T. S., Thorgeirsson, S. S., and Heilman, C. A. (1983).Proc. Natl. Acad. Sci. U S A . 80,1947-1950. Marcu, K. B., Harris, L. J., Stanton, L. W., Erikson, J., Watt, R., and Croce, C. M. (1983). Proc. Natl. Acad. Sci. U S A . 80,519-523. Mendelsohn, J., Trowbridge, I. S., and Castagnola, J. (1983). Blood 62,821-826. Miller, G., and Yale, J. (1971). Biol. Med. 43,358-365. Morton, C. C., Taub, R., Diamond, A., Lane, M.-A., Cooper, G. M., and Leder, P. (1984). Science 223,173-175. Murray, M. J., Cunningham, J. M., Parada, L. F., Dautry, F., Lebowitz, P., and Weinberg, R. A. (1983). Cell 33, 749-757. Neel, B. G., Hayward W. S., Robinson, H. L., Fang, J., and Astrin S. M. (1981). Cell 23, 323-334. Neel, B. G., Jhanwar, S. C., Chaganti, R. S. K., and Hayward, W. S. (1982). Proc. Natl. Acad. Sci. U S A . 79,7842-7846. Neiman, P. E., Jordan, L., Weiss, R. A., and Payne, L. N. (1980a). Cold Spring Harbor Con$ Cell Proliferation 7,519-528. Neiman, P. E., Payne, L. N., and Weiss, R. A. (1980b).J . Virol. 34, 178-186. Neiman, P. E., Wolf, C., and Cooper, G. M. (1985).Proc. Natl. Acad. Sci. U S A . 82,222226. Nilsen, K.,and Klein, G. (1982).Ado. Cancer Res. 37,319-380. Nishikura, K., ar-Rushdi, A., Erikson, J., Watt, R., Rovera, G., and Croce, C. M. (1983). Proc. Natl. Acad. Sci. USA. 80,48224826. Nusse, R., and Varmus, H. E. (1982). Cell 31,99-109.

THE BLYMONCOGENES

123

Nusse, R., van Ooyen, A., Cox, D., Fung, Y. K. T., and Varmus, H. (1984). Nature (London) 307, 131-136. Payne, G. S., Bishop, J. M., and Varmus H. E. (1982). Nature (London) 295,209-213. Peters, G., Brookes, S., Smith, R., and Dickson, C. (1983). Cell 33, 369-377. Peterson, R. D. A., Burmester, B. R., Fredrickson, T. N., Purchase, H. G., and Good, R. A. (1964).J. Natl. Cancer Znst. 32, 1343-1354. Peterson, R. D. A., Purchase, H. G., Burmester, B. R., Cooper, M. D., and Good, R. A. (1966). J . Natl. Cancer Znst. 36,585-598. Purchase, H. G., and Burmester, B. R. (1978). Zn “Diseases of Poultry” (M. S. Hofstad, B. W. Calneck, C. F. Helmboldt, W. M. Reid, and H. W. Yoder, Jr., eds.), 7th Ed., pp. 418-468. Iowa State Univ. Press, Ames. Ramsay, G. M., Enrietto, P. J., Graf, T., and Hayman, M. J. (1982). Proc. Natl. Acad. Sci. U S A . 79,6885-6889. Rosen, A., Gergely, P., Jondal, M., Klein, G., and Britton, S. (1977). Nature (London) 267,5244. Rubin, H. L., Cornelius, A., and Fanshier, L. (1961). Virology 17, 143-156. Stewart, T. A., Pattengale, P. K., and Leder P. (1984). Cell 38,627-637. Taetle, R., Honeysett, J. M., and Trowbridge. I. S. (1983). Znt.]. Cancer 32,343-349. Taub, R.4 Kirsch, I., Morton, C., Lenoir, G., Swan, D., Tronick, S., Aaronson, S., and Leder, P. (1982). Proc. Natl. Acad. Sci. U S A . 79, 7837-7841. Tormey, D. C., Imrie, R. C., and Mueller, G. C. (1972). E x p . Cell Res. 74, 163-169. Trowbridge, I. S., and Domingo, D. L. (1981). Nature (London) 294, 171-173. Trowbridge, I. S., and Lopez, F. (1982). Proc. Natl. Acad. Sci. U S A . 79, 1175-1179. Trowbridge, I. S., Lesley, J., and Shulte, R. (1982).J. Cell. Physiol. 112,403-410. Tsujimoto, Y., Yunis, J., Onorato-Showe, L., Erikson, J., Nowell, P. C., and Croce, C. M. (1984). Science 224,1403-1406. Westaway, D., Payne, G., and Varmus, H. E. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 843-897.

This Page Intentionally Left Blank

RETROVIRUS-INDUCED ACQUIRED IMMUNODEFICIENCIES Mauro Bendinelli. Donatella Matteucci. and Herman Friedman' Institute of Epidemiology. Hygiene and Virology. University of Pisa. Pisa. Italy. and 'Department of Microbiology and Immunology College of Medicine. University of South Florida. Tampa. Florida

.

I . Introduction . . . . . . . . . . . . ..................................... I1. Retroviruses as Agents of I ................... A . Endogenous Retroviruses ...................................... B. Exogenous Retroviruses ........................................ I11. Immunodepressive Changes in Retrovirus-Infected Animals ............ IV . Functional Alterations of Immunocompetent Cells in Retrovirus-Infected Animals ....................................... A . TLymphocytes ............................................... B. B Lymphocytes ............................................... C Accessory Cells ............................................... D Nonspecific Defense Mechanisms ............................... E Recapitulation ................................................ V. Mechanisms Leading to Immunocompetent Cell Alteration in Retrovirus Infections .................................. A . Retrovirus Replication within Immunocompetent Cells . . . . . . . . . . . . . B . Immunodepressive Virion Components ........................... C. Perturbation of Normal Immunoregulatory Circuits . . . . . . . . . . . . . . . . . D. Tumor-Dependent Mechanisms ................................. VI . Role of Retrovirus-Induced Immunodeficiency in Pathogenesis . . . . . . . . . A . Influences on the Inducing Infection and on the Ensuing Disease . . . . B. Reduced Resistance to Superinfection ............................ VII The Retroviral Etiology of AIDS ................................... VIII Summary and Perspectives ........................................ References ......................................................

. . .

. .

125 127 129 132 134 137 139 141 143 145 148 149 149 152 154 158 159 159 161 162 168 169

1. Introduction

Investigations of oncogenic viruses have. understandably. always centered on the neoplasia-inducing properties and. initially. the study of the immunodepressive effects of retroviruses (RV) was no exception . Following reports in the early sixties that certain murine leukemia viruses (MuLV) could adversely affect host's immune competence (Old et al., 1960; Peterson et al., 1963). investigation of the phenomenon received impetus from the belief that it might explain why many highly immunogenic RV-induced tumors are not rejected by host's defenses. and by the correlative hope of broadening the understanding of neoplastic progression in general . As a conse125 ADVANCES IN CANCER RESEARCH. VOL. 45

Copyright 0 1985 by Academic Press. Inc. A11 rights of reproduction in any form reserved.

126

MAURO BENDINELLI ET AL.

quence, when in the mid-seventies the climate of opinion regarding the theory of immune surveillance of tumors became unfavorable, studies in the area underwent a parallel low tide of popularity. Indeed, as pointed out by Weiss (1984), for years in the laboratory the immunodeficiencies caused by oncogenic RV [oncovimses (OV)] were regarded as nothing but a nuisance because infected animals died of superinfections before they could develop tumors. However, RV are not only agents of neoplasia. Indeed, there are numerous RV that have never been associated with cancer. Some are agents of chronic degenerative diseases of animals, while others have never been linked to any form of disease and may be entirely apathogenic. Furthermore, OV elicit a vast array of nonneoplastic manifestations which can accompany the tumorous diseases or occur as independent pathological entities. Examples of nonproliferative pathology given by OV are anemia, thrombocytopenia, bone marrow atrophy, alterations of metabolism, stunting, obesity, ataxia, and osteopetrosis caused by viruses of the avian leukemia virus (ALV) and reticuloendotheliosis virus (REV) groups (Smith, 1982; Smith and Schmidt, 1982; Weiss et al., 1982; Carter et al., 1983; Carter and Smith, 1984). Severe erythroid hypoplasia, abortions, runting, and glomerulonephritis are frequently observed in cats infected with feline leukemia virus (FeLV) (Hardy, 1980,1982).Nonneoplastic disorders associated with murine OV include the hematocrit changes caused by Friend leukemia complex (FLC) and Rauscher leukemia complex (RLC) (Ostertag and Pragnell, 1981),the abnormal gait and paraplegia caused by Abelson murine leukemia virus (Ab-MuLV) (Rosenberg, 1982), and the neurogenic paralysis associated with MuLV infection of wild mice (Fredrickson et al., 1972; Gardner, 1978). FLC has also been associated with runting, autoimmune hemolytic anemia, and glomerulonephritis in rats (Kuzumaki et al., 1974). Among such nonneoplastic manifestations, immunodeficiency holds a very prominent position. In fact, though immunodepressive symptoms are a frequent occurrence in viral infections in general (Specter and Friedman, 1978; Bendinelli, 1984),they are induced by so many members of the RV family and may be so long lasting and devastating that they have come to be regarded as an independent problem. Thus, for example, it is generally accepted that FeLV-infected cats may suffer generalized immunodeficiencies that represent a serious threat to life, independently of the neoplasia that may also ensue (Olsen, 1979; Hardy, 1982). Since recent results showing that RV are involved in the etiology of the recently recognized acquired immunodeficiency of humans

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

127

known as AIDS and related pathology have led to a resurgence of interest in the immunodepressive properties of oncogenic and nononcogenic RV, a critical appraisal of our current understanding of the mechanisms whereby RV immunodepress seems timely. Previous reviewers have clearly pointed out how RV can impair the immune system both directly and via the neoplastic diseases they may induce (Salaman, 1969; Dent, 1972; Friedman, 1974; Stutman, 1975; Specter and Friedman, 1978; Cerny and Essex, 1979; Friedman et al., 1979; Olsen, 1979; Bendinelli, 1981; Hardy, 1982).This article deals primarily with the former type of mechanisms, but it is necessary to emphasize that, even in well-studied models, distinguishing between tumorunrelated and tumor-dependent immunodepressive events may be exceedingly difficult. The subject has been a major focus of our research for a number of years. This, hopefully, will excuse us for making frequent reference to our own results. II. Retroviruses as Agents of Immunodeficiency

For a comprehensive description of the extraordinarily large number of RV that have been demonstrated in laboratory and free-living animals since the pioneering work of ROUS,Ellermann, Bang, and Gross, the reader is referred to the numerous reviews and books that have recently appeared in all areas of retrovirology (Hanafusa, 1977; Klein, 1980,1982a; Stephenson, 1980; Brahic and Haase, 1981; Hooks and Detrick-Hooks, 1981; Bishop, 1982; Weiss et al., 1982; Enrietto and Wike, 1983; Vande Woude et al., 1984). The few points briefly touched upon here focus on their natural history and other aspects that may help in understanding the information discussed in the following sections. The family Retroviridae classifies all RNA viruses known to replicate by way of a DNA intermediate integrated in the host cell genome. The members studied in some detail have exhibited essentially homogeneous morphological, structural, biochemical, and physical properties, as well as similar genetic organizations and molecular strategies of replication. However, biological cycles and pathogenetic properties of individual members differ widely. Indeed, it is on the latter properties that the presently accepted subdivision of RV in subfamilies is essentially based (Matthews, 1982).Additional criteria of subdivision are natural and experimental host range (in vitro ecotropic RV grow effectively only on cells of the species of origin, xenotropic RV only on cells of species other than that of origin), mode of transmission, location of the nucleoid within the virion (central in type C particles,

128

MAURO BENDINELLI ET AL.

eccentric in B and D particles), maturation characteristics, antigenic markers, cross-interference, and others. Oncogenic RV are grouped in the subfamily Oncovirinae. A key distinction within this group is based on the presence or absence in the viral genome of specific transforming genes (v-onc) originally transduced from homologous sequences of normal cells. Viruses lacking a v-onc cause tumors after a latency of several months or more (chronic OV), are replication independent, regularly tumorigenic only in immunologically deficient hosts, and do not cause readily detectable transformation in uitro. Some are highly oncogenic under natural conditions and are leading causes of death in animal species ranging horn chicken to cattle. Thus, for example, virus-induced leukemia and lymphoma are relatively frequent in cats, their incidence being 5- to 10-fold greater than in humans (Olsen, 1979). In contrast, the majority of onc-carrying RV induce macroscopic neoplasms within a few days or weeks (acute OV), are tumorigenic also in immunologically competent hosts, and transform cells efficiently in uitro. Most lack essential vegetative sequences and depend for replication on concomitant infection of the cell by a related nondefective helper virus. Certain OV, however, do not appear to contain a cell-derived oncogene and, nevertheless, induce acute leukemias. An example is FLC, whose rapid leukemogenic activity has been related to products of viral structural genes (Clark and Max, 1983; Ruta et al.,

1983). There are about 30 acute OV of birds, rodents, cats, and primates that have been characterized. Although isolates obtained from primary tumors exist, the best known ones have been derived following serial passage of slow OV or transplantable tumors either in the natural host or in heterologous species. Thus, it is generally believed that they are generated by a rare recombination event(s) between retroviral genomes and cellular DNA and that their normal destiny is to perish together with the tumorous host unless timely rescued by the laboratory. The subfamily Lentivirinae encompasses a number of RV associated with the etiology of “slow” diseases of animals, such as visnamaedi of sheep, arthritis-encephalitis of goat, and possibly equine infectious anemia. RV classified in the subfamily Spumavirinae persist in infected hosts in the absence of any apparent pathological damage and are unmasked in cultured cells by the formation of vacuolated syncitia. Recognized species include the simian foamy virus and the bovine and feline syncitial viruses.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

129

A. ENDOGENOUS RETROVIRUSES Endogenous RV are ubiquitous in nature in the form of proviruses stably integrated in the host’s chromosomes. They have been identified in so many and widely divergent species that they may be common to all animals, are present at varying numbers of copies in all the cells of the host, and are transmitted vertically through the germ line as inherited genes. Some, spontaneously or following activating treatments, become expressed at relatively high leveI. When this occurs, endogenous RV may be transmitted to the progeny extragenetically (through the milk, for example), but horizontal transmission seems on present evidence to occur very rarely, if ever. Thus, endogenous RV may be regarded as cellular genes with the unique ability to generate virions. However, they share many important characteristics of and may be progenitors of exogenous RV. Most endogenous RV studied so far are xenotropic and have yet to be shown to possess pathological potential. Actually, their remarkable evolutionary conservation has led to the postulate that they are beneficial to the host (Aaronson and Stephenson, 1976). However, when repeatedly passed in suitable hosts, endogenous RV may present a varied degree of oncogenicity, as exemplified by a number of leukemogenic viruses isolated from X-irradiation and chemically induced tumors. Moreover, some ecotropic types are certainly a significant cause of slowly arising spontaneous tumors in particular inbred strains of animals; a classical example is the endogenous virus of AKR mice that is expressed at high levels throughout life and is associated with a high incidence of spontaneous T cell lymphocytic lymphoma and leukemia which develop by one year of age. Indeed, the relationships of endogenous RV and their recombinants to various naturally occurring neoplastic and nonneoplastic (e.g., autoimmune) affections remain a promising avenue of research (Ulrich and Nex& 1985). Despite interesting speculation (Phillips et al., 1975; Kassan and Chused, 1977; Moroni and Shumann, 1977; DeLamarter et al., 1979; Hellman et al., 1980; Wecker and Horak, 1982) and scattered results showing that certain endogenous RV may be inhibitory for immunocompetent cell functions in vitro (Table I), little is known about the effects of endogenous retroviral infections on the immune homeostasis of the host. In an analysis of 59 inbred mouse strains, Heiniger et aZ. (1975) detected no correlations between phytohemagglutinin (PHA) responsiveness and expression of endogenous RV, while Crittenden et aZ. (1982) found that the effects of endogenous retroviral

TABLE I RETROVIRUSES ETIOLOGICALLY ASSOCIATEDWITH ACQUIREDIMMUNODEFICIENCY IN NATURALLY AND EXPERIMENTALLY INFECTED HOSTS" Virus Host

Chronically tumorigenic

Chicken

Avian leukosis (ALV) REV associated

Mouse

Murine leukemia (MuLV): Friend- (F-), Rauscher (R-), Gram-, Gross-, Moloney(Mo-), radiation- (Rad-), dimethylbenzanthracene(DMBA-), urethane-, 334CMammary tumor (MuMTV)

Acutely tumorigenic Avian erythroblastosis Avian myeloblastosis Avian myelocytomatosis Avian sarcoma (ASV) Reticuloendotheliosis (REV-T) Friend leukemia complex (FLC) Rauscher leukemia complex (RLC) Moloney-murine sarcoma (Mo-MuSV)

Nontumorigenic

c

k2

Rat Cat Sheep Horse Baboon Macaque

Mo-MuLV Feline leukemia (FeLV)

Gibbon Man

Gibbon ape 1eukemiab.h

“Modified from Bendinelli et al. (1985). vitro evidence only. %tiff and Olsen (1983). dSvennerholm et al. (1978). eBanks and Henson (1973). fDenner et al. (1980); Weislow et al. (1981). gHooks and Detrick-Hooks (1981). hWainberg et al. (1983a).

RD-114b,c Visna-maedid Infectious anemia= Endogenousbf Mason-Pfizer (MPMV), related type-D, foamyg T-lymphotropic-3 (HTLV3) Lymphadenopathy-associated (LAV)

132

I

MAURO BENDINELLI ET AL.

I

genes on t e immunologic and pathologic response of chickens to exogenously inoculated RV were largely specific for structurally related agen . Althoug it is generally agreed that the immune capabilities of the leukemia- rone AKR mouse decline markedly concomitantly with the developme t of neoplastic lesions (Collavo et al., 1975; Roman and Golub, 19 6), the immunoresponsiveness of preneoplastic animals has been e subject of conflicting reports (Metcalf and Moulds, 1967; Dor6 et d.,~ 1969; Perkins et al., 1971; Martig and Tribble, 1974; Zatz,

t

1975; Hatt n and Dunton, 1978), possibly due at least in part to difficulties in nding appropriate uninfected controls. Thus, while the influences bf endogenous RV on the immune system might receive e use of mouse substrains carrying deliberately endoviruses (Jaenish, 1976), the present unproperties of RV must almost controlled exogenous infections. B. EXOGE~OUS RETROVIRUSES Exogenops RV are spread among susceptible hosts horizontally and/ or vertically in a manner identical to other infectious agents. Many are broadly diqtributed, infecting numerous individuals of their respective host slpecies, even though they generally cause disease in a minority. Thqs, for example, by l year of age at least 50% of pet cats living in suburban areas have been exposed to FeLV (Blasecki, 1981). The routes of transmission are multiple. FeLV is transmitted via saliva or transplacentally (the common outcome of naturally acquired infection is a transient infection which is rapidly terminated by the development of ‘mmunity, but 1-2% of infected cats become persistently viremic an excrete high titers of virus). The gibbon ape leukemia is excreted th ough saliva and urine. The murine mammary tumor virus (MuMTV) is usually passed via lactation but is also passed by mating (and select d strains are passed to the progeny genetically). Also the bovine leu emia virus is shed with milk, though it spreads more effectively horiqontally. ALV are transmitted through the egg albumen and, less fdequently, by contact, while MuLV are passed only vertically and, at least within laboratory strains, do not appear to circulate horizontally. Transspecies transmission of selected RV has also been documented (Todaro, 1980). Modalities of infection are known to influence the consequences of RV infection in terms of host invasion, tumorigenesis, induction of immunity, and epizoology (Herberman,

1

,k

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

133

1982). The possibility should be considered that they also influence the immunodepressive effects of the virus. The ability to immunodepress the host is shared by so many exogenous RV infecting such widely divergent species (Table I) that the property might be regarded as having significant survival value for such viruses. However, for the historical reasons mentioned in Section I the immunodepressive properties of oncogenic types have been the subject of special attention. From the literature it would also appear that acute OV are more profoundly immunodepressive than other RV (Bendinelli and Nardini, 1973b; Rup et al., 1982), but any generalization of this kind should be accepted with reserve because the already mentioned difficulties in discriminating tumor-related and tumor-unrelated changes are obviously more stringent in the case of rapidly oncogenic viruses. On the other hand, the histotype of the tumor(s) caused by the virus does not appear to be critically important. Oncoviruses exhibiting powerful immunodepressive potential encompass members which produce T lymphocytic [Gross-MuLV, dimethylbenzanthracene MuLV (DMBA-MuLV), Moloney MuLV (MoMuLV), radiation MuLV (Rad-MuLV), FeLV], B lymphocytic [ALV, reticuloendotheliosis (REV-T), Friend MuLV (F-MuLV), and Rauscher MuLV (R-MuLV)], monomyelocytic (avian myeloblastosis and myelocytomatosis viruses, Graffi-MuLV), and erythroid (FLC and RLC) leukemias, as well as sarcomas [Moloney murine sarcoma virus (MO-MuSV)]. FLC has represented a strong attraction for research workers interested in RV-induced immunodeficiency (Bendinelli, 1981). In fact, this viral complex has several advantages, not the least of which is the rapid appearance and the severity of the immunodepressive symptoms in infected hosts. A further reason for interest in this model has been the possibility of directly comparing the immunodepressive effects of the entire rapidly leukemogenic complex with those of the slowly tumorigenic helper, F-MuLV. As suggested by data obtained in various models (Bendinelli and Nardini, 1973b; Smith and van Eldik, 1978; Rup et al., 1982), helper viruses contribute significantly to the overall immonodepressive activity of oncoviral complexes. Since helper viruses are usually in great excess over the defective components, the immunological effects produced by small doses of such complexes may actually be entirely due to the helpers (Kateley and Friedman, 1976; Bendinelli, 1977). Interestingly, in the avian system the immunodepressive properties have been shown to vary with subgroup specificity of chronic leukemia viruses. Spleen cells of chickens infected with ALV of subgroup B

134

MAURO BENDINELLI ET AL.

did not respond well to concanavalin A (Con A), PHA, and pokeweed mitogen, while spleen cells from chickens injected with subgroup A viruses responded normally (Rup et al., 1979,1982; Fadly et al., 1982; Price and Smith, 1982). Although the reasons for this discrepancy have not been investigated, the fact is of interest because it indicates that the surface properties of the virion may influence the immunodepressive effects of RV. It is important to consider at this point that the remarkable genetic plasticity of RV and the consequent propensity to generate variants (Weiss et al., 1982) allow one to hypothesize the existence within retroviral preparations of viral subpopulations that are responsible for most if not all of their immunodepressive activity. Consistent with this possibility, (1)viruses such as FLC and 334C-MuLV have been found to lose and reacquire the ability to immunodepress depending on whether they had been propagated in vitro or in vivo (Eckner, 1975; Robinson et at., 1980);(2)variants of Rad-MuLV and 334C-MuLV exhibited distinguishable immunosuppressive activities (Haran-Ghera et al., 1977; Robinson et al., 1980; Yefenof and Zilcha, 1982); and (3) different clones of F-MuLV have been seen to diverge in the ability to suppress lymphocyte blastogenesis in vitro (Cerny and Essex, 1979). Since dissecting retroviral variants is an area of current research in many laboratories, the possibility should soon become accessible to more detailed verification. It must also be recalled that exogenous RV may become contaminated by unrelated immunodepressive agents (e.g., the lactic dehydrogenase virus) which may contribute to or generate inconsistency or conflicting data (Rowson and Mahy, 1975). 111. lrnrnunodepressive Changes in Retrovirus-Infected Animals

For a detailed account of the vast array of immune functions that have been found adversely affected in naturally and experimentally RV-infected animals the reader is referred to previous reviews (Dent, 1972; Specter and Friedman, 1978; Cerny and Essex, 1979; Friedman et al., 1979; Bendinelli, 1981; Hardy, 1982). Only a few points are summarized here and very briefly. 1. The spectrum of humoral cellular parameters affected as well as rapidity and degree of depression vary considerably depending on the virus-host combination considered. For example, ALV have been typically associated with a decreased humoral responsiveness and a variable effect on cell-mediated immunity. Similarly, a number of studies have established that cell-mediated responses are disrupted in

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

135

TABLEI1 IMMUNERESPONSESFOUND DEPRESSED IN SUSCEPTIBLE BALBIc MICEINFECTED WITH THE FRIEND LEUKEMIA COMPLEX“ Background production of natural antibody Primary antibody responsiveness Secondary antibody responsiveness Immunological maturation Delayed hypersensitivity Contact sensitivity Graft rejection T cytotoxic responsiveness Resistance to superinfection Resistance to tumor grafts ~

~~~~

“From Bendinelli (1981).

FLC-infected mice (Table 11),but it is generally agreed that this occurs at more advanced stages of infection and to a lesser extent as compared to humoral immunity. In contrast, decreased cellular immune responses were reported in mice infected with Gross-MuLV and in cats infected with FeLV, but reports regarding antibody responses have been variable (Trainin et al., 1983). 2. Occurrence and extent of immunodepression by a given RV are dependent on the host’s genotype. There is often a close parallelism between susceptibility to the immunodepressive and to the tumorigenic actions of the virus (Ceglowski and Friedman, 1969). However, the parallelism is not absolute. Decreased antibody responsiveness has been repeatedly observed in leukemia-resistant MuLV-infected mice, although the impairment was generally transient and mild as compared to the effects in leukemia-susceptible animals. Cell-mediated responses appear particularly refractory to inhibition by MuLV in resistant mice, though short-lasting depressions have been reported (Mortensen et al., 1974). Genetic and phenotypic bases of resistance to RV-induced immunodepression are essentially unknown. Mouse susceptibility to immunodepression by FLC is controlled not only by loci that regulate leukemogenesis but also by independent genes (Kumar et al., 1978; Meruelo and Bach, 1983). In view of the key role played by direct interaction of the virus with immunocompetent cells in the genesis of imrnunodepression (Section V), restriction of viral replication is probably a very important determinant. 3. Infection must precede antigenic challenge in order for immunodepression to be most evident. Though in certain models some degree of depression was seen even when the infection was initiated a few days after antigen inoculation, ongoing responses and established im-

136

MAURO BENDINELLI ET AL.

munity are generally spared. Increased levels of preexisting natural antibody-forming cells have been noted in the spleen of FLC-infected mice (Hirano et al., 1971; McAlack et al., 1971; Bendinelli et al., 1975~1, but it is not known whether this reflects a true potentiation of natural antibody synthesis, disturbances of lymphocyte circulation through the organ (Bainbridge and Bendinelli, 1972), or enhanced lymphocyte blastogenesis due to the virus. In uitro a similar effect was observed when fresh normal spleen cells were exposed to FLC (Bendinelli and Friedman, 1980) but not when the cells were derived from infected mice (Bendinelli et al., 1978b). A transient but significant stimulation of the anti-sheep red cell (SRC) antibody response has also been observed following RV inoculation into genetically resistant animals (Okunewick et al., 1978). 4. Although immunodepressive changes may develop soon after infection and, in the case of OV infections, well in advance of recognizable tumors, they become more severe with increasing time from infection. This fact is of interest because of the characteristically long latency of AIDS in humans (Curran et al., 1984). Notable exceptions are the immunodepressive effects of MuMTV (Tagliabue et al., 1981) and F-MuLV. In the latter model, the depression of antibody responses peaks within 2 weeks from infection and then gradually subsides to leave the responsiveness at about half the normal level for the life-span of the infected animal. In this regard the immunodepressive effects of F-MuLV differ considerably from those of the entire FLC (Fig. 1). 5. The extent of immune inhibition may differ markedly with the type of antigen as well as other variables concerning immunization protocol. This, together with the information in point 3 above, suggests that the events most intensely affected by RV are those related to the inductive phase of responses. Were the deficit located in the subsequent steps, such as the antigen-driven proliferation and differentiation of lymphocytes, one would expect the degree of depression to be independent on the immunogen. As discussed in Section IV,C, such contention has found considerable support in the study of the cellular basis of immunodepression. 6. In certain models a compartmentalization of immunodepression is clearly evident. For example, mice infected with FLC or F-MuLV a few days before immunization with SRC produce one-tenth or less of the normal number of splenic antibody-forming cells both in uivo and in vitro, whereas lymph node responsiveness is little changed for much longer periods (Bendinelli, 1973). Similarly, the inhibitory ef-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

loo

137

I

DAYS

WEEKS

MONTHS

FIG.1. Antibody responsiveness of BALB/c mice infected with FLC (0)or with FMuLV (0).The animals were infected at 6 weeks of age with 104 PFU (S+ L-) of virus. At the indicated time intervals from infection their spleen cells were cultured in uitro (10' per well) and stimulated with SRC (lo6per well). Direct antibody-forming cells percentage of the average were assayed 5 days later. The results are e:pressed:Fs response of uninfected conk01 mice of the sax713 age (4-6 mice per group).

fect of FLC on alloreactive cytotoxic T lymphocyte generation was maximal in the spleen, intermediate in peritoneal cells, and nil in lymph node cells (Genovesi et al., 1982). The spleen is the major target for leukemogenesis by these viruses, but the difference is evident before the appearance of neoplastic changes. IV. Functional Alterations of lmmunocompetent Cells in Retrovirus-Infected Animals

Retroviral infections may induce significant tumor-independent histopathology of lymphoid tissues. Examples are (1)the cellular depletion observed in the thymus and/or the peripheral lymphoid organs of cats infected with FeLV (Anderson et al., 1971; Hoover et al., 1973; Pack and Chapman, 1980), of birds infected with selected members of the REV group (Purchase and Witter, 1975), and of monkeys injected

138

MAURO BENDINELLI ET AL.

with Mason-Pfizer monkey virus (MPMV) (Fine et al., 1975) or affected by a spontaneous immunodeficiency syndrome that has been linked to MPMV-related viruses (King et al., 1983; Daniel et al., 1984; Marx et al., 1984; Stromberg et al., 1984), (2) the lymphoid hyperplasia induced by F-MuLV (Carter et aE., 1970; Bendinelli and Nardini, 1973a) and by MAV-2-0 (Smith and Ivanyi, 1980), a myeloblastosisassociated virus of subgroup B that induces osteopetrosis (Banes and Smith, 1977), and (3) the depressed peripheral T lymphocyte counts observed in FeLV-infected cats (Cerny and Essex, 1979; Rojko et al., 1979). Nonetheless, RV-induced immunodeficiency is not merely a reflection of morphologically appreciable changes, as shown by its occurrence also in animals which evidence no lymphoid injury by conventional pathological analysis and by the fact that immunocytes of infected animals continue to exhibit substantial functional derangements once transferred in uitro. Information on the cellular events responsible for RV-induced immunodeficiency has been obtained by comparing the functional behavior of immunocytes from infected and control animals in a number of tests in viuo and in uitro. The majority of available data derives from few extensively explored murine models. A further limitation of these data is that one cannot be certain of whether a given change is due to intrinsic defects of the cell type apparently affected or rather mirrors alterations of interreacting elements in effector and/or regulatory circuits. It is a general problem of cellular immunology that one cannot exclude that indirect influences are at work in a phenomenon unless monotypical cell populations, and preferably clones, are investigated. Thus, reduced responsiveness of lymphoid cells to polyclonal B and/ok T mitogen stimulation in vitro has been described in a variety of RV-infected mammals and birds (Toy and Wheelock, 1975; Cockerel1 et al., 1976; Tagliabue et al., 1980; Price and Smith, 1982; Rup et al., 1982; Daniel et al., 1984; Marx et al., 1984; Stromberg et at., 1984), but in several systems the ceIlular defect responsible has been located outside the responder cells (Rudczynski and Mortensen, 1978; Herberman et al., 1980; Tagliabue et al., 1980; Price and Smith, 1982). For example, the splenic lymphocytes of MuSV-infected mice showed markedly inhibited responses, but removal of macrophages resulted in a cell population fully competent to evince T cell reactions in vi tro (Herberman et al., 1980). Despite such limitation, we will review the information dealing with different immunocompetent cell classes separately, in an attempt to pinpoint likely targets for RV immunodepressive activity.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

139

A. T LYMPHOCYTES An impairment at the level of T cells, “the prime movers of the immunological universe” (Klein, 1982b), might explain many of the immunological deficits observed in RV-infected mice. Here we shall deal with T cells responsible for the antigen-specific portion of cellmediated reactions (Te)and with T cells programmed to facilitate and/ or amplify the function of other lymphocytes (Th) or their precursors. The role played by suppressor T cells (Ts) in the generation of RVassociated immunodeficiency is discussed in Section V,C. As yet, the changes undergone by the various T cell subpopulations defined by surface antigens have not been attentively scrutinized in RV-infected animals. In certain RV-infected hosts-typically in FeLV-infected cats-the general pattern of immunodeficiency is strongly suggestive of a selective defect of T cell-mediated immune functions. Unfortunately, in such systems there is little or no information on whether the deficit results from an intrinsic impairment of T cells. Lymphocytes of FeLVinfected animals showed a reduced motility of surface membrane Con A receptors (Dunlap et d.,1979). However, in a recent study peripheral blood lymphocytes of normal and FeLV-infected preleukemic cats were comparable in receptors for guinea pig erythrocytes (a marker of feline T cells), interleukin 2 (IL-2) dependency, proliferative activity and life span in vitro as well as in the ability to develop PHA-induced nonspecific cytotoxicity (Grant et al., 1984). The system where the functional behavior of T cells has been most extensively studied is FLC infection of mice. In this model, FLC depression of cell-mediated reactions is accompanied by parallel modifications of in vitro correlates such as migration inhibitory factor production and T-dependent cytotoxicity, indicating that Te may be less functional in infected animals (Bendinelli, 1981). However, like cell-mediated responses of the intact mouse, in vitro tests were affected relatively late in the course of infection (Genovesi et al., 1982). For example, splenic T cells showed a reduced capability to respond to mitogens and to mediate allogeneic effects considerably late as compared to depression of antibody responsiveness, and lymph node, thymus, and peripheral blood cells revealed no loss of responsiveness for a much longer time (Dracott et aZ., 1977; Genovesi et aZ., 1982). Depression of cell-mediated responses was a late effect, relative to depression of humoral responses, also in mice infected with MoMuLV (Cerny and Essex, 1979) and RLC (Meredith et d.,1978). Con-

140

MAURO BENDINELLI ET AL.

tact sensitivity was not decreased in F-MuLV-infected mice or in leukemia-resistant mice infected with FLC (Bendinelli et al., 1975a). With few exceptions (Mortensen et aZ., 1974), in the above studies retroviral infection resulted in diminished cell-mediated responses only if it preceded exposure to antigen. It seems, therefore, reasonable to assume that already sensitized Te are not markedly affected by RV and that the events responsible for depression occur at the level of their inductiodmaturation. The inductive stages of cellular responses are little understood, but that they may, indeed, be inhibited by RV infections is illustrated by the reduced ability of FLC-infected spleen cells to generate specifically cytotoxic lymphocytes when stimulated in vitro with allogeneic (Garaci et al., 1981),Friend leukemia (Plata et uZ., 1980), or unrelated tumor (Genovesi et at., 1982) target cells. Adoptive cell-transfer experiments performed in viuo have shown that T cells of MuLV-infected mice may (Bennett and Steeves, 1970; Ceglowski and Friedman, 1970) or may not (Haran-Ghera et at., 1977) be able to help B cells of MuLV-infected mice may (Bennett and Steeves, 1970; Ceglowski and Friedman, 1970) or may not (HaranGhera et al., 1977) be able to help B cells normally in the generation of T-dependent antibody responses. However, such data are now questionable because it has been recognized that carryover virus-producing cells initiate a rapid infection in the recipient irradiated animals and immunodepress by mechanisms different from those operative in virus-infected intact animals (Toniolo et aZ., 1984), that infected lymphocytes acquire abnormal homing characteristics (Bainbridge and Bendinelli, 1972; Gillette and Fox, 1977), and that uninfected recipients may react against grafted virus-infected syngeneic cells (Casali and Trinchieri, 1984). On the other hand, in the only published experiments designed to directly monitor the Th function of RV-infected animals, T lymphocytes from 4-day FLC-infected animals provided a normal help to uninfected T cell-depleted lymphoid cultures in the antibody response to SRC, even though the T cell donors showed greatly reduced antibody responses (Dracott et al., 1977). The spleen of FLC and F-MuLV-infected mice presents a prompt decrease in the percentage of cells that bind anti-Thy-1.2 antibody (Kateley et al., 1974; Friedman and Kateley, 1975; Bendinelli and Friedman, 1976). The change has been attributed to blocking by masking substances (Mascio and Ceglowski, 1978) or to a reduced expression of the relevant antigen possibly related to a rapid fall in circulating thymic factor(s) (Tonietti et al., 1983), but its significance is unclear. T cells carry out some of their functions through the secre-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

141

tion of diffusible products. There are few studies exploring the secretory funtions of T cells in RV-infected animals, though the aspect would seem to deserve attention in view of recent results showing that in vitro a purified lymphokine obtained from L1210 leukemia cells greatly enhanced the antibody response of FLC-infected spleen cells (Butler and Friedman, 1979) and that human peripheral blood leukocytes when exposed in vitro to avian, murine, feline, and primate RV become inhibited in the production of and in the response to IL-2 (Wainberg and Margolese, 1982; Copelan et d.,1983; Wainberg et al., 1983b). In preliminary experiments, spleen cells of FLC- or FMuLV-infected mice showed no changes in IL-2 production and responsiveness (unpublished results).

B. B LYMPHOCYTES Reduced antibody responsiveness is often a prominent manifestation of RV-induced immunodepression. Since the ability to initiate and develop an antibody response is related to the number and normal functioning of B lymphocytes, these cells were initially considered a likely target of immunodepression (Dent, 1972). A model where B lymphocyte functionality has been attentively monitored is, again, FLC infection of mice. Early adoptive cell transfer experiments had initially indicated that FLC infection might result in depletion of potential antibody-forming cells in the spleen and bone marrow or directly interfere with their maturation and proliferation (Bennett and Steeves, 1970; Ceglowski and Friedman, 1970; Shearer et aZ., 1973), but the limitations inherent to this experimental approach have already been discussed. In fact, there is now substantial evidence that B cells and their precursors are not grossly impaired by FLC quantitatively or qualitatively. The binding of erythrocyte antigen by such cells was normal (Halasa et al., 1972). Cells involved in ongoing responses continued to secrete normal amounts of antibody (Koo et al., 1971). Responses to antigens wholly or partly independent of T and accessory cell help, such as bacterial lipopolysaccharide (LPS) and pneumococcal capsular polysaccharide, were considerably less affected than responses to other antigens (Dracott et al., 1978; Bendinelli et aZ., 1978a). Costimulation with LPS increased the response to a T-dependent antigen both in vitro and in vivo (Bendinelli et d., 1976a, 1978a; Dracott et d.,1978; Butler and Friedman, 1979), so that the depression was markedly reversed, though not always entirely overcome. Increased numbers of antibody-forming cells were achieved after addition of uninfected macrophages or macro-

142

MAURO BENDINELLI ET AL.

phage-produced factors (see below). Infection failed to affect the residual antibody responsiveness of T-depleted mice (Dracott et al., 1977). Finally, mitogen-driven blastogenesis of B lymphocytes in uitro was relatively spared by FLC infection; even late in the disease it was only slightly depressed in the spleen and peripheral blood, and normal in lymph nodes (Genovesi et al., 1982). The above results, coupled with aforementioned evidence that RV tend to primarily affect the early inductive steps of immune responses, indicate that the key deficit(s) responsible for inhibition of humoral responses by FLC is not located in the B lymphocytes, but should rather be searched in the collaborative interactions that trigger them to differentiate into antibody-secreting cells. This does not, however, imply that B cells remain absolutely normal responsive. For example, FLC and F-MuLV infections were shown to cause a rapid decline in the proportion of spleen lymphocytes expressing surface immunoglobulins and in the motility of lymphocyte membrane as evidenced by capping anti-Ig sera (Kateley et al., 1974; Friedman and Kateley, 1975; Bendinelli and Friedman, 1976). Early changes were also observed in the external membrane architecture of putative splenic B lymphocytes obtained from FLC-infected mice (Farber et al., 1978). Furthermore, antibody responses to antigens that stimulate B cells in the relative absence of T lymphocytes and macrophages were depressed by FLC, though less severely than thymus-dependent responses (Bendinelli et al., 1978a). It seems, therefore, feasible that B cells of FLC-infected mice may present a somewhat increased threshold for activation, but that the deficit is easily overcome, or escapes detection, following stimulation by strong signals. Since anamnestic responses have proved especially sensitive to depression, committed and memory B cells might be preferentially vulnerable in this respect (Specter and Friedman, 1978; Cerny and Essex, 1979). It is also likely that more substantial damage of the B cell lineage occurs during the late tumorous stages of FLC infection. There are few investigations aimed at monitoring specific B cell properties in other RV infections. Nevertheless, the general impression is that these cells as such are not much impaired. For example, antibody secretion and virus replication were seen to Foexist in individual celIs of Mo-MuLV-infected mice (Celada and Asjo, 1973) and rats (Cremer et al., 1969) and antibody responses to T-independent antigens were little or not influenced by Rad-MuLV (Haran-Ghera et al., 1977) and F-MuLV (Bendinelli et al., 1978a). Similarly, T-independent responses were only marginally, if at all, affected by RLC and Mo-MuLV (Meredith et al., 1978).

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

143

C. ACCESSORYCELLS The induction of most, possibly all, specific immune responses-be they humoral or cell-mediated-requires accessory functions provided by cells variously classed as macrophages, macrophage-like cells, dendritic cells, etc. (Unanue, 1981; Bellanti and Herscovitz, 1984). Chronologically, the first indication that impairment of such cells can contribute to the genesis of RV-associated immunodeficiencies was the observation that addition of macrophage-rich peritoneal exudate cells or adherent spleen cells from normal syngeneic donors restored the ability of F-MuLV-infected murine lymphoid cultures to mount near-normal antibody responses to SRC. The cells responsible for the effect were more than 95% macrophages by criteria such as morphology, adherence, phagocytosis, presence of Fc receptors, and absence of lymphocyte-specific markers, and various other cell types proved ineffective. Since infected spleens showed no numerical deficiencies of total adherent phagocytic cells, these results clearly suggested the existence of a functional deficit of accessory cells (Bendinelli et al., 197513). This conclusion was corroborated by the observation that F-MuLVinfected animals responded well to antigens less dependent on macrophage help than SRC (Bendinelli et al., 1978a) and developed highly enhanced antibody responses to SRC when injected with appropriate numbers of syngeneic macrophages together with the antigen (Bendinelli et al., 197613). The anti-SRC responses achieved by administering optimal numbers of exogenous macrophages averaged about 80% in vitro and 40% in vivo as compared to the uninfected counterparts, suggesting that inefficient help by accessory cells is an important, though probably not the only, cause of F-MuLV immunodepression. Essentially similar evidence was obtained in FLC-immunodepressed mice. However, in this case the restoration effected by macrophages in vitro was of low degree, and all the attempts to alleviate the immunodepression by injecting exogenous macrophages directly into infected animals were unsuccessful (Ceglowski and Friedman, 1975; Specter et al., 1978). It was concluded that in FLC infection additional pathways to immunodepression not only exist but are largely predominant. To what extent such additional mechanisms are linked to the rapid leukemogenic action of FLC remains to be established. Additional evidence for inefficient accessory cell function as a determinant of RV-induced immunodepression comes from other mu-

144

MAURO BENDINELLI ET AL.

rine models. Normal macrophages reversed the suppression of in uitro lymphocyte responsiveness produced by established lines of tumors (Ting and Rodrigues, 1979) whose immunodepressive potential appears to be due to release of MuLV-related products (Bluestone and Lopez, 1979). Accessory cells of adult mice neonatally infected with Mo-MuLV proved unable to support the generation of secondary cytotoxic T lymphocytes in vitro by uninfected spleen cells depleted of Ia-positive cells. Most interestingly, however, in this study the macrophage deficit was selective for the cell-associated antigens induced by the infecting virus (Biasi et al., 1983). Possibly relevant are also findings showing that adherent spleen cells of NZB mice, a strain known to harbor a highly expressed endogenous virus (Morse and Hartley, 1981), failed to cooperate normally with lymphocytes in the generation of antibody responses (McCombs et al., 1975). Further supportive evidence comes from the avian system. Chickens infected as embryos with the osteopetrosis virus MAV-2-0 exhibit a dramatic loss of lymphoid organ mass, reduced IgM antibody production, and depressed responses to various mitogens (Smith and van Eldik, 1978; Hirota et al., 1980).However, in cell-mixing experiments the blastogenic response of infected lymphoid cells to Con A was effectively restored by macrophage-rich adherent cells of blood or spleen from normal chickens. Since adherent cells were numerically normal in the lymphoid tissues of infected birds and did not present gross alterations in Fc-receptor content, Fc-dependent phagocytosis, and nonspecific esterase staining, these results clearly showed that the T lymphocytes of MAV-2-0-infected chickens were intrinsically normal responsive and that the deficit was located at the accessory cell level (Price and Smith, 1982). The functions attributed to accessory cells in the induction of immune responses are many fold, including promotion of lymphocyte viability and proliferation, antigen processing and/or presentation, and mediation of T cell-B cell cooperation. Indications that the accessory cell deficit of RV-infected animals may be related to the preparation of antigen for recognition by lymphocytes were obtained in the FMuLV model: (1)the number of macrophages required for restoration was low (0.5-1% responder cells); (2) the degree of macrophage activation was critical and the nature of the in uiuo activating stimulus was also important; (3) ultraviolet (UV) irradiation of macrophages at doses known to block cell division did not reduce, and in certain instances enhanced, macrophage activity, whereas doses known to interfere with protein synthesis were inhibitory; (4) it was crucial that macrophages were added during the early phases of the response; (5)

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

145

agents known to replace the lymphocyte viability-proliferation promoting activity of macrophages, such as 2-mercaptoethanol and macrophage-conditioned medium, failed to enhance the response; and (6) conditions that reportedly curtail the need for the antigen-presenting function of macrophages, such as increasing SRC antigen dosage and adding Escherichia coli or Serratia marcescens LPS as well as allogeneic-conditioned medium, partially circumvented the need for added macrophages (Bendinelli et al., 1976a; Butler et al., 1980). Clearly, a close look should be given to the expression of la antigen by the macrophages of RV-infected animals. Recent results have shown that FLC-infected mice produce reduced m o u n t s of soluble factors needed for optimal antibody responses both in basal conditions and upon LPS stimulation. Furthermore, supplementation with such factors brought the antibody responsiveness of spleen cells from grossly FLC-leukemic spleens to normal or even slightly elevated levels (Butler et al., 1983). These findings suggest that the production by macrophages of diffusible products which deliver activating signals to lymphocytes such as IL-1 may also become deficient in the course of RV infections.

D. NONSPECIFICDEFENSEMECHANISMS RV infections may result not only in impairment of cells needed for mounting specific immune responses, but also in deficit of mechanisms that participate in host’s defenses either as constitutive broadrange effectors or because recruiteaarmed by antibody and other antigen-specific mediators. It seems likely that such deficits contribute significantly to the general immunodeficiency picture that eventuates in infected hosts. Table I11 summarizes present knowledge on the effects of RV infections on macrophage functions other than those involved in the induction (Section IV,C) and regulation (Section V,C) of immune responses. Reports have described substantial alterations of carbon clearance in mice infected with FLC and RLC (Old et al., 1960; Seidel and Nothdurft, 1976). It seems, however, likely that the viral preparations used in such studies were contaminated by the lactic dehydrogenase virus, which is known to impair reticuloendothelial system function selectively, or were secondary to tumor development because investigations in vitro have consistently shown Fc-dependent and Fc-independent phagocytosis to be normal or near normal in productively infected and/or transformed macrophages of RV-infected chickens and mice.

TABLE111 EFFECTOF RETROVIRUSINFECTIONS ON MACROPHAGE FUNCTIONS Functions

Change

Virus

Phagocytosis

Normal or slightly reduced

RLC ALV Abelson-MuLV F-MuLV AMV

Expression of Fc receptors

Normal Reduced Normal Reduced Reduced

FLC F-MuLV RLC Gross-MuLV

Cox and Keast (1973) Gazzolo et al. (1975) Raschke et al. (1978) Bendinelli et al. (1980) Durban and Boettiger (1981b) Price and Smith (1982) Bendinelli et al. (1980) Price and Smith (1982) Friedman and Ceglowski (1971) Bendinelli (1973)Komitowski and Kurba 1974) Normann et al. (1981)

Reduced Reduced Enhanced Reduced Reduced Reduced

Several RLC MuMTV M 0-M u SVa FLC Abelson-MuLV

Cianciolo et al. (1980) Cox and Keast (1973) Tagliabue et al. (1980) Taniyama and Holden, 1979a) Genovesi et al. (1982) Broxmeyer et al. (1980)

Enhanced

Several

Greenberger et al. (1980)

Enhanced

FLC

Johnson et al. (1984)

M AV-2-0

F-MuLV M AV-2-0

Motility in uitro Accumulation at sites of inflammation Killing of ingested bacteria Spontaneous cytotoxicity Antibody-dependent cytotoxicity Responsiveness to soluble mediators Production of granulocyte-macrophage colony-stimulatingfactor Stimulation of erythropoiesis Progressing tumors only.

References

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

147

However, since the perturbations of macrophages (Table 111) and their myelomonocytic precursors (Lieberman and Sachs, 1979; Klein et al., 1982) described in RV infections involve functions considered of paramount importance in host defenses, it is not surprising that evidence suggestive of a correlation between macrophage impairment and progression of RV infection and RV-induced tumors has repeatedly been reported (Gazzolo et al., 1979; Marcelletti and Furmanski, 1979; Hoover et al., 1981; Wainberg et al., 1983a). Although large numbers of mononuclear macrophages have been shown to infiltrate OV-induced tumors from the early phases of growth (Metcalf et al., 1967; Stanton et al., 1968; Siegler, 1970; Russell et al., 1976; Puccetti and Holden, 1979),their functionality as effectors of resistance may be lower than normal, especially in progressing tumors (Russell and McIntosh, 1977; Taniyama and Holden, 1979a; Wainberg et al., 1983a). The data dealing with the behavior of other nonspecific defense mechanisms are few and almost exclusively restricted to the effectors of cytolysis. Natural killer (NK) and natural cytotoxic (NC) activities were shown to be depressed in FLC- and MuSV-infected mice. The residual NK activity of FLC-infected mice was boosted by interferon, but remained low as compared to similarly treated controls. The mechanisms active in the regulation and suppression of natural cytolysis are still poorly resolved. However, the depression induced by FLC was not due to impaired binding to the target cells and appeared to involve macrophage-like suppressor elements (Bendinelli et al., 1981; Migliorati et al., 1983; Lust et al., 1984; Moody et al., 1984). Similarly, in the MuSV model NK activity assayed in situ in progressively growing sarcomas was low or undetectable but could be markedly increased by removal of adherent phagocytic cells (Gerson et al., 1981). The ability to effect antibody-dependent cell cytotoxicity was also decreased after the second week of FLC infection. The reduction was evident in the spleen, peripheral blood, and peritoneal cells using both erythrocytic and tumor targets, thus indicating that different effector cells were probably suppressed (Genovesi et al., 1982). A marked neutropenia has been observed in FeLV-infected kittens (Rojko et al., 1979) and in monkeys inoculated with MPMV (Fine et al., 1975).An adrenal-independent inhibition of eosinophil exudation at sites of secondary antigenic challenge was described in FLC-infected mice but, since the influx of other granulocytes was normal, the effect was considered secondary to the depression of immune functions (McGarry et al., 1978). Inflammatory responses to epicutaneous

148

MAURO BENDINELLI ET AL.

irritants (Bendinelli et al., 1975a) and complement levels (Bendinelli et al., 1974) were normal in FLC-infected mice. Even though RV induce interferon (Blank and Murasko, 1980), the production of inducer-stimulated interferon was found to be reduced in MuLV-infected mice (De Maeyer-Guignard, 1972).

E . RECAPITULATION The gist of the preceding discussion is that a variety of potentially important functional aberrations of immunocompetent cells have been described in RV-infected mice. However, many key questions remain unresolved. It has already been noted that it is impossible to be sure that a given change is intrinsic to the cell type apparently affected. Suffice it to add that in no model has it yet been possible to determine which of the changes found pertain in a critical way to the genesis of immunodepression rather than being nonessential side effects, let alone to arrange them into a causally linked chain. This is not surprising since we are just getting the first glimpses of the high level of interdependence existing between immunocompetent cells and of how such interdependence can be disturbed in physiological and pathological situations. The problem is probably further compounded by the fact that the cellular pathways leading to RV-induced immunodeficiency are multiple and their relative impact varies depending not only on the infecting virus but also on the time interval from the initiation of infection and other variables. Such contention is suggested by results showing that (1)in FLC-infected mice the cell types which mediate suppression change with time (Bendinelli et al., 1979);( 2 )the immunocompetent cell classes impaired by Rad-MuLV and FLC differ depending on mouse strain and preexposure to X-rays (Haran-Ghera et al., 1977; Bendinelli et al., 1978);and (3)lethally irradiated mice reconstituted with F-MuLV-infected spleen cells receive no beneficial effects from procedures that reproducibly restore the antibody responsiveness of intact F-MuLV-infected animals (Toniolo et al., 1984). On the other hand, among the multiple alterations of immunocytes observed in RV-infected hosts, the impairment of macrophages seems to hold a very prominent position. It is tempting to speculate that the alterations of these cells and especially of their accessory functions represents a unifying determinant in the genesis of many manifestations of RV-induced immunodeficiency.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

149

V. Mechanisms Leading to lmmunocompetent Cell Alteration in Retrovirus Infections

Subversion of host’s immune functions by viral infections in general is believed to result either from virus invasion of variable numbers and subsets of immunocytes or via less direct means (Bendinelli, 1984). The many unanswered questions regarding the cellular events responsible for RV-induced immunodeficiency render the identification of the precise mechanisms that initiate and sustain such changes highly problematic.

A. RETROVIRUS REPLICATIONWITHIN IMMUNOCOMPETENT CELLS Few infectious agents proliferate and persist in their hosts to the extent shown by RV. This is certainly due to the unequaled genetic intimacy these viruses establish with host cells and to compatibility of their replication with cell survival, but presumably also reflects the remarkable ability of RV to proliferate in cells of the lymphocyte and macrophage lineages. Although in vitro many RV grow in a large spectrum of cell types, in vivo the majority of, possibly all, RV become first evident in organs which are lymphoreticular in nature (Weiss et al., 1982; MerueIo and Bach, 1983). Extensive replication in such organs may actually be a prerequisite for tumorigenesis by chronic OV (Hayward et al., 1982). The class(es) of lymphocytes supporting RV growth varies with the virus-host combination and cannot be inferred from the histotype of the tumor that the virus may eventually induce. For example, the Tlymphoma inducers Gross-MuLV, Mo-MuLV, and FeLV proliferate effectively in many cell types including B lymphocytes throughout the preleukemic period (Gisselbrecht et al., 1978; Horak et al., 1981; Rojko et al., 1981); viruses of the ALV group grow extensively in other lymphoid cells before bursa1 tumors start to develop (Beard, 1980), and even MuMTV, whose transforming activity is extremely restricted, has been seen to invade lymphoid tissues (Hilgers and Bentvelzen, 1978). Also the cells of the monocyte-macrophage lineage become readily infected by RV, as shown by the wide spectrum of RV that have been seen to replicate in such cells (Table IV). Macrophage permissiveness to FeLV was actually higher than that of lymphocytes which represent the transformation target of this virus (Hoover et al., 1981; Rojko et al.,

150

MAURO BENDINELLI ET AL.

TABLE IV RETROVIRUSES THATREPLICATE IN CELLS OF THE MONOCYTE-MACROPHAGE LINEAGE Host Chicken

Virus Avian erythroblastosis ALV Avian myeloblastosis Avian myelocytomatosis

Quail Mouse

Cat Goat Sheep Horse

ASV Avian erythroblastosis Avian myelocytomatosis ASV Abelson-MuLV Cas-MuLV FLC F-MuLV Gross-MuLV Mo-MuLV FeLV Arthritis-encephalitis Visna-maedi Infectious anemia

Reference Durban and Boettiger (1981a) Gazzolo et al. (1975) Moscovici and Gazzolo (1982) Moscovici and Gazzolo (1982) Rangan and Bang (1967) Moscovici et al. (1981) Linial (1982) Yamanouchi et al. (1979) Raschke et a2. (1978) Fredrickson et al. (1984) TonioIo et al. (1980) Marcelletti and Furmanski (1978) Godard et al. (1983) Biasi et al. (1983) Hoover et al. (1973) Narayan et al. (1983) Narayan et al. (1982) Henson and McGuire (1974)

1981). Moreover, evidence from the avian (Gazzolo et al., 1979; Wainberg et al., 1983a),murine (Marcelletti and Furmanski, 1979; Johnson et al., 1980), and feline (Hoover et al., 1981) systems suggests that replication in the monocyte-macrophage lineage may be a critical event for the initial invasion of the host and for persistence by RV, even though these cells usually maintain the nontransformed phenotype. Interestingly, Gazzolo et a2. (1974, 1975) noticed variation in susceptibility of macrophages to ALV and avian sarcoma virus (ASV) belonging to different subgoups of envelope specificity, a variation that seems to parallel the different immunodepressive properties of such subgroups in infected chickens (Rup et al., 1979, 1982; Fadly et al., 1982; Price and Smith, 1982). In accordance with the requirement for ongoing cellular DNA synthesis in order that RV replicate (Temin, 1967),the ability of lymphocytes to support RV replication is greatly enhanced in activated as compared to resting cells. Reports are numerous that RV replication and endogenous provirus expression is augmented following mito-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

151

genic stimulation of lymphocyte cultures (Ruddle et al., 1976; Cerny and Isaak, 1979; Rojko et al.; 1982; Ewert et al., 1983) and in animals undergoing intensive immune responses (Phillips et al., 1975). The permissiveness of macrophages for RV has also been shown to vary depending on the state of activation (Pessano et al., 1979; Narayan et al., 1983) and other cellular variables (Hoover et al., 1981). Clearly, collectively these findings imply that during RV infections the immunocytes actively engaged in immune responses are bound to become preferentially invaded by the infecting virus. With few notable exceptions (Weiss et al., 1982; Popovic et al., 1984), RV replication does not lead to extensive cytopathology. When certain specific conditions pertaining to the virus, to the cell, and to the microenvironment occur, infected cells transform. Otherwise there often are no obvious deleterious effects on cell physiology and behavior, while the amount of viral synthesis may vary considerably under the influence of genetic and epigenetic factors. However, even apparently normal RV-infected cells differ from uninfected cells in several aspects, including the expression of new surface antigens (Weiss et al., 1982). On the other hand, Oldstone et al. (1984) and Casali et al. (1984) have recently drawn attention to the fact that viruses may perturb the fine specialized functions of infected cells owing to subtle effects on cellular synthesis and metabolism without appreciable morphological injury. Thus, it would certainly not be surprising if RV replication within immunocompetent cells did not permit the normal development of the highly integrated and coordinated cellular activities presumably needed during the inductive phases of immune responses. Although FLC and Mo-MuLV replicating within B lymphocytes have proved compatible with the secretion of antibody (Koo et al., 1971; Celada and Asjo, 1973),RV replication in immunocytes might also be expected to cause unbalanced production of some among the numerous soluble mediators which effect communication within the immune system. The level of circulating thymic factor(s) underwent a dramatic decrease following FLC infection, possibly related to a wave of viral replication in the thymus (Tonietti et al., 1983). That incapacitation of immunocompetent cells during RV infection results, at least in great part, from direct interaction with the infecting virus is strongly suggested by several forms of evidence. First, a linkage between malfunctioning of and productive RV replication in macrophages and other immunocytes has repeatedly been noticed (Marcelletti and Furmanski, 1978; Cerny and Essex, 1979; Greenberger et al., 1980; Biasi et al., 1983). Also, the organ compartmentalization of FLC-induced immunodepression has been attributed to regional dif-

152

MAURO BENDINELLI ET AL.

ferences in lymphocyte susceptibility to the virus (Isaak et al., 1979). Second, the suppression of normal lymphocyte functions exerted by MuLV-infected cells in uitro was evident, at least in certain instances, also when suppressing and responder cells were separated by cellimpermeable membranes (Friedman et al., 1979) and could be abrogated by virus-specific antibody (Kamo et al., 1975; Bendinelli et al.,

1979). Third, substantial inhibition of normal responder lymphoid cultures was effected also by the addition of cell-free preparations of RV. Acellular preparations of a number of RV have been shown to inhibit the proliferative responses of lymphocyte cultures to conventional antigens, alloantigens, T mitogens and, less efficiently, to B mitogens (Phillips et al., 1975; Specter et al., 1976a; Cockerel1 and Hoover, 1977; Cerny and Essex, 1979; Kumar and Bennett, 1981). Furthermore, partially purified FLC was found to inhibit the antibody response of murine spleen cells without altering their viability, and the effect was abolished by preincubating the viral inoculum with antiviral antibody (Specter et al., 197613) (Fig. 2). A further line of evidence is discussed in the following section. VIRIONCOMPONENTS B. IMMUNODEPRESSIVE In a number of studies involving a variety of RV-cell combinations, suppression of normal lymphoid responder cultures was also observed by the use of UV- or detergent-inactivated viral preparations (Fowler et al., 1977; Hebebrand et al., 1977; Nichols et al., 1979; Denner et al., 1980; Weislow et al., 1981; Copelan et al., 1983; Langweiler et al., 1983; Stiff and Olsen, 1983). This was not due to cytotoxicity and has been explained either by postulating that RV are structured in such a fashion as to abort immunocyte functions by simple physical contact or with the presence in RV particles of structural components endowed with potent immunomodulatory properties. Consistent with the former idea, the effect could be reproduced with nonviral membranous vesicles that approximated viruses in size and also by the use of lymphoid cells lacking receptors for the added virus, thus showing that viral entry may be irrelevant to immunosuppression (Israel et al., 1979; Wainberg and Israel, 1980). The data supporting the second explanation are numerous. (1)The major envelope glycoprotein g p 70 of MuLV was shown to possess a strong affinity (Delarco et al., 1978; Choppin et al., 1981) and to be mitogenic for murine lymphocytes (Bubbers et al., 1980);( 2 )disaggregated virions and semipurified components of RLC suppressed mito-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

153

DILUTION OF SEMIPURIFIED FLC

FIG.2. Effect of a cell-free preparation of FLC on the antibody responsiveness of normal lymphoid cultures. Spleen cells of BALBlc mice were supplemented with the indicated dilution of virus (plasma from bweek infected mice partially purified by differential centrifugation) and then processed as described in Fig. 1. Note that doses of ultraviolet rays (104 erg) that completely inactivated virus infectivity only had a minor effect on the immunodepressive properties.

gen-driven blastogenesis (Hellman et al., 1980) as well as the spontaneous cytolytic activity of mouse spleen cells (Hartzfeld et al., 1981); (3)nanogram amounts of extracts of the purified 15,000-Da envelope protein p15E of selected MuLV were seen to inhibit macrophage accumulation at sites of inflammation in mice (Cianciolo et al., 1980); and (4) the p15E protein of FeLV was shown to inhibit antigen- and mitogen-induced blastogenic responsiveness of normal feline (Mathes et al., 1978,1979) and human (Hebebrand et al., 1979; Copelan et d., 1983) lymphocytes as well as the proliferation of bone marrow hematopoietic precursor cells (Wellman et al., 1984). Although the relevance of these highly artifactual experiments to what occurs in vivo is unclear, these findings imply that RV virions and subvirion components might not need to be produced inside immunocompetent cells or organs to effect significant immunosuppression. It is also noteworthy that the suppressive effects of certain trans-

154

MAURO BENDINELLI ET AL.

plantable murine tumors in uitro were shown to be mediated by RV-related products (Bluestone and Lopez, 1979; Cianciolo et al., 1983)and that effusions of human cancers were reported to contain an inhibitor of monocyte chemotactic responses which reacted with monoclonal antibodies to the p15E protein of selected murine RV (Cianciolo et al., 1981; Snyderman and Cianciolo, 1984).

c. PERTURBATION OF NORMAL IMMUNOREGULATORY CIRCUITS

1

The normal equilibrium of the immune system is maintained by the concerted action of antigen-specific and nonspecific homeostatic control reactions. Although the evidence discussed so far indicates that active infection of immunocytes is a crucial determinant in the generation of RV-induced immunodeficiencies, an aberrant activation or inappropriate tuning of regulatory circuits might be an important contributing factor and also explain the development of autoreactive cytotoxic effectors and other immunological dysfunctions observed in RV pathogenesis (Cox and Keast, 1973; Cerny and Essex, 1979; Schenk and Howe, 1979; Ihle and Lee, 1982). Among possible causes of derangement of immunoregulatory mechanisms is RV replication in immunocytes itself. Theoretically, in the immune network, any virus-induced modification of immunocytes, even if compatible with a normal functioning of the infected cells, is liable to be amplified by multiple cascades of effects. Thus, for example, the display of virion components onto the plasma membrane-an essential step in RV maturation-might activate interacting cells along improper paths (Greenspan and Wainberg, 1981). The polyclonal activation of lymphocytes effected by certain RV (Bendinelli and Friedman, 1980; Bubbers et al., 1980) might also reflect itself on many immune functions. Accordingly, even the seemingly straightforward suppression effected by the direct addition of acellular RV preparations to lymphoid cultures appeared to result via more complex avenues than a direct action on responder cells, often involving the activation of suppressor mechanisms (Israel et al., 1980; Israel and Wainberg 1981; Kumar and Bennett, 1981; Langweiler et al., 1983; Copelan et al., 1983; Wainberger et al., 1983b). There are quite a few parallels between RV-induced immunodeficiency and antigenic competition (Dracott et al., 1977). Indeed, there are no solid grounds to dismiss antigenic competition as a route to virus-induced immunodepression in general. Since the dominance of an immunogen on another depends on relative doses and timing, and

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

155

viruses as replicating agents provide an extremely large supply of antigenic material, the differences observed between the effects of infectious and inactivated virus preparations by no means exclude a role for antigenic competition. Circulating virus-antibody complexes have been documented in several RV infections (Oldstone, 1975). Being capable of influencing the lymphoid system in many ways (Casali and Oldstone, 1983; Virgin and Unanue, 1984), immune complexes might be important contributors to deregulation, especially when produced in proximity to or at the surface of immunocompetent cells (Oldstone et al., 1984), as conceivably occurs in most RV infections. On the other hand, observations that antisera directed against MuLV can modulate murine lymphocytes (Lee and Ihle, 1976; Schumann and Moroni, 1978; Moroni et al., 1980; Langdom and Shellam, 1981; Senn and Papoian, 1983) suggest further mechanisms of immunological derangment by anti-RV immune responses. Of course, many more and sophisticated mechanisms whereby RV might disrupt the normal immunoregulatory mechanisms could be proposed. However, as already discussed, our limited understanding of the cellular basis of RV-associated immunodepression hampers further meaningful speculation on the subject.

1. Nonspecific Suppressor Cells in RV-Znfected Hosts Suppressor cells and circuits capable of nonspecifically diminishing the normal lymphocyte responses via diffusible products or by direct cell-to-cell contact are currently viewed as critical determinants of immunodeficiency (Klein, 198210).To address the question of whether such mechanisms participate in RV-associated immunodepression, several in uitro assay systems have been developed in which normal lymphocyte functions are monitored in the presence of cocultured RV-infected lymphoid cells (Cerny and Stiller, 1975; Kateley et al., 1975; Toy and Wheelock, 1975). Among cells with immunoregulatory functions, nonspecific Ts appear to play a pivotal role because they are elicited both physiologically during the normal development of immune responses and in a variety of pathological and experimental conditions. Accordingly, nonspecific Ts have been described in hosts bearing various types of tumors (Berendt and North, 1980; Plater et al., 1981; Herberman, 1983) or infected with a variety of nononcogenic viruses (Bendinelli, 1984). However, while RV-specific Ts have repeatedly been detected in infected (Bluestone and Lopez, 1982; Yefenof and Ben-David, 1983) and also in normal uninfected animals (Tilkin et al., 1984), the

156

MAURO BENDINELLI ET AL.

nonspecific suppressive activities described in RV-infected hosts have not generally been categorized as T cell mediated. Although Ts may act directly on B lymphocytes and macrophages, the bulk of their action is believed to be directed against Th (Klein, 1982b). Thus, the lack of massive faulty functioning of Th in RV-infected mice (Section IV,A) is in keeping with the concept that Ts activation does not play a major role in RV-associated immunodepression. Indeed, recent findings in FLC-infected mice (Garaci et al., 1981) and FeLV-infected cats (Stiff and Olsen, 1982) suggest that OV might damage Ts or their precursors. Rhabdoviruses may also selectively destroy specific Ts subpopulations (Sy et al., 1983). In contrast, activationhecruitment of suppressor macrophages appears to be a frequent occurrence in OV-infected animals. In an early study, it was noticed that macrophages of FLC-infected leukemic mice inhibited the antigen-independent synthesis of antierythrocyte antibody that occurs when peritoneal lymphocytes are cultured in uitro (Bendinelli, 1968, 1970). Subsequently, macrophages and macrophage-like cells derived from RV-infected animals were repeatedly shown to suppress immune functions as diverse as antibody and lymphokine production, lectin- and antigen-driven proliferation, T celI cytotoxicity, and NK activity (Roder et al., 1978; Grinwich et al., 1979; Caulfield and Cerny, 1980; Tagliabue et al., 1980; Migliorati et al., 1983; Varesio, 1983; Moody et al., 1984; Pettey and Collins, 1984). Suppressor macrophages found in the tumorous mass and in the spleen of MuSV-induced sarcoma-bearing mice have provided most of our knowledge in this area, but even in this model many uncertainties remain on how the suppressive activity is triggered and expressed (Varesio, 1983). Similar cells are frequently found in animals bearing transplanted tumors and are regarded as a major component of the immunodepression observed in such animals, no matter whether the carried tumors were originally RV induced or not (Bluestone and Lopez, 1979).Also in the MuSV model the elicitation of suppressor macrophage activity has generally been attributed to the neoplastic process rather than to MuSV infection per se. Hitherto, suppressor macrophages have most often been described in overtly tumorous RVinfected animals. Moreover, suppressor macrophages peaked in MuSV-infected mice with the size of the carried tumor (Kirchner et al., 1975) and appeared at the site of DMBA-MuLV-induced tumor growth earlier than in the spleen, thus raising the possibility that they were elicited within the tumor and then migrated to the spleen (Roder et al., 1978). Thus, even though threshold effects cannot be excluded, we must

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

157

conclude that on the whole available data do not permit one to incriminate the activation of professional suppressor cell circuits and, by inference, deregulation, as a major tumor-independent mechanism of RV-induced immunodepression. In certain instances, the identity of suppressor cells found in RV-infected animals has remained unclear. This applies to the suppressor cells induced by viruses of the REV group, the only avian RV that have consistently been shown to activate suppressor cells (Carpenter et al., 1977, 1978; Scofield and Bose, 1978; Walker et al., 1983; Bose, 1984), and to those found in the spleen of MuMTV-infected mice and described either as B cells (Rudczynski and Mortensen, 1978) or as macrophages (Tagliabue et al., 1980). The nonneoplastic spleen cells of mice infected with FMuLV and FLC which suppressed antibody responses in vitro were identified as B cells that appeared to act by releasing immunodepressive viral products (Bendinelli et aZ., 1979).That suppression is mediated by viral products is a possibility that should always be considered while studying RV-induced suppressor cells.

2. Nonviral Immunosuppressive Products During RV infections, immunocompetent cells are bound to become exposed to a multitude of nonviral molecules, originated in cells and tissues as a reaction to infection and inflammation, some of which may possess substantial local or systemic immunomodulating properties. For example, RV have been shown to induce interferon production in infected hosts (Blank and Murasko, 1980). Since each interferon type has distinctive capabilities in altering a variety of immune responses (Preble and Friedman, 1983), the consequences on the immune system of the mixtures of interferons produced after viral infection in uiuo are presently impossible to predict. An altered production of growth factors and other cytokines within or outside the immune system as well as hormonal imbalances may also play a part in overall immunodeficiency (Greenberger et al., 1980; Tagliabue et al., 1980; Koury and Pragnell, 1982). At present, however, there is a minimum amount of information bearing on these aspects, possibly due to the overshadowing in current test systems of putative nonviral immunomodulatory substances by immunoreactive virion components. Virus-unrelated substances with lymphoid cell inhibitory potential were detected in spleen extracts of FLC-infected mice (Specter and Friedman, 1978) and REVT-infected chickens (Carpenter et al., 1977), but these observations were not further pursued. Spleen cells of normal mice exposed in vitro to UV-inactivated avian oncoviral particles released a soluble

158

MAURO BENDINELLI ET AL.

factor, possibly an unstable protein different from interferon, with the ability to inhibit proliferative responses of fresh lymphocytes (Israel et al., 1980; Margolese et al., 1980), and the cells responsible for its production and for serving as primary target were identified as macrophages (Israel and Wainberg, 1981). The suppression of normal NK activity by FLC-infected splenocytes was partially reversed by indomethacin (unpublished results), suggesting a mediation by prostaglandins, which have been implicated in several immunosuppressive situations mediated by macrophages (Stenson and Parker, 1980), including the Mo-MuSV model (Brunda et al., 1980).

D. TUMOR-DEPENDENT MECHANISMS Even though many of the immunodepressive changes observed in RV infections may occur independently or in advance of the tumorigenic activity, it is clear that tumor-related mechanisms may contribute to the shutoff of immunological responsiveness. A detailed discussion of these aspects is beyond the scope of this article but, especially in acute OV infections, the additional intervention of such mechanisms should always be considered. As mentioned in Section 11, many immunological alterations observed in AKR mice develop or become more prominent coincidentally with or subsequent to the onset of spontaneous leukemia (Metcalf and Moulds, 1967; Ram et al., 1974; Normann et al., 1981). In rats inoculated intracranially with ASV the immune status varied considerably depending on the type of tumor developed (Roszman et al., 1973).Also, the late depression of cell-mediated responses by several MuLV may partly reflect tumor-related factors (Section 111).On the other hand, a substantial restoration of virus-specific and general immune reactivity is frequently observed following the spontaneous or provoked regression of FLC leukemia (Cerny et al., 1975; Chesebro et al., 1979; Genovesi et al., 1982), despite the fact that regressors may remain virus positive. The progression of tumors induced by acute OV is partly sustained by recruitment of newly transformed cells and often leads to extensive neoplastic infiltration of lymphoid organs and bone marrow (Weiss et aZ., 1982). This may result in depletion of immunocompetent cells and/or perturb mechanically, metabolically, or otherwise the milieu in which immunocytes grow, mature, and perform their functions. Dilution of immunocytes with nonreactive neoplastic cells has repeatedly been invoked to explain late immunodepressive effects. As the spleen of the FLC-leukemic mouse enlarges, it progressively impedes the

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

159

egress, but not the ingress, of circulating lymphocytes, thus contributing to perturb lymphocyte traffic (Bainbridge and Bendinelli, 1972). In addition, viral and nonviral tumor cells are known to release a variety of as yet poorly defined substances that in vitro suppress a number of immune functions (Kamo and Friedman, 1977; Mizel et al., 1980). Some tumors also make considerable amounts of immunomodulatory prostaglandins. For example, prostaglandins of the E series were increased up to over 50-fold in MuSV-induced sarcomas (Humes et al., 1974) and in spontaneous AKR leukemias (Karmali et al., 1984). As already mentioned (Section V,C), tumor cells have been shown to participate in the suppression of normal lymphoid responses in vitro by spleen cells of mice infected with MuLV (Kamo et al., 1975; Roman and Golub, 1976; Bendinelli et al., 1979). In one such model the presence of DNA on the surface of the leukemic cells was instrumental for suppression to occur (Russell and Golub, 1980).The production of soluble suppressive factors by selected clones and subclones of FLC-transformed cells was also unrelated to the synthesis of infectious virus and viral molecules (Bendinelli et al., 1985). VI. Role of Retrovirus-Induced Immunodeficiency in Pathogenesis

Current thinking on virus-induced immunodepression in general is that it may have important bearings on viral spread within the infected tissues, on the duration of infection, and on the mechanisms whereby cell and tissue damage are generated, and that it may predispose to initiation of superinfections, to reactivation of latent infections, and to chronic diseases of autoimmune or other nature (Bendinelli, 1984). However, there are relatively few investigations that have specifically monitored these aspects in retroviral infections. A. INFLUENCES ON THE INDUCING INFECTION AND ENSUING DISEASE

ON THE

Its early development has led to the assumption that the compromise of host immune responsiveness produced by RV facilitates the full expression of the pathogenic potential of the infecting virus by preventing the immune elimination of both infected cells and free virus (Jarrett, 1975; Essex, 1977; Specter and Friedman, 1978; Cerny and Essex, 1979; Rup et al., 1982). However, owing to difficulties in designing appropriate experiments, most of the supportive data are indirect and far from conclusive. RV infection and tumorigenesis are strongly influenced by the im-

160

MAURO BENDINELLI ET AL.

munological state of the host. For example, it is known that RV-infected transformed and nontransformed cells express a complex mosaic of antigens (such as viral components, depressed cellular antigens, and other ill-defined determinants) that can serve as targets for T-dependent cytolytic, cytostatic, and graft reactions, antibody and complement, antibody-dependent cell cytotoxicity, etc. (Blasecki, 1981; Levine, 1982). Pharmacological and other treatments which abrogate immune reactivity without abating the target cells for the virus greatly enhanced RV expression and neoplastic consequences. Conversely, numerous aspecific immunoaugmenting agents have been shown to increase resistance to RV, as revealed by reduced incidence and severity of neoplasia and increased host survival (Herberman, 1982; Bendinelli, 1985). In certain OV-host combinations, spontaneous regression is the usual outcome of the induced tumors and the phenomenon is known to be basically immunologic in nature (Radzichovskaja, 1967; Levy and Leclerc, 1977; Furmanski et al., 1979; Collavo et d., 1980; Johnson et d., 1980). In this general framework, reports that preinfection with selected MuLV increased the pathogenic action of related OV such as MuSV (Chirigos et al., 1968; Hooks et al., 1969) and the growth of transplanted tumors (Deodhar and Chiang, 1970) were considered strong presumptive evidence that the immunodepressive activity is a critical determinant of RV pathogenesis. Consistent with this concept, a direct relationship has repeatedly been noticed between OV-induced immunodepression and the development of tumors both in unmanipulated animals (Ceglowski and Friedman, 1969; Meredith et al., 1978; Anand and Steeves, 1980) and upon nonspecific immunopotentiation (Friedman and Ceglowski, 1973; Marx and Wheelock, 1976; Kiyohashi, 1981). On the other hand, evidence that RV-induced immunodepression is directed not only against unrelated antigens, but also against the antigens of the inducing virus, is skimpy. Inoculation of cats with inactivated FeLV and its purified envelope protein p15E abolished the immune response to the feline oncornavirus-associated antigen as well as lymphocyte blastogenic response to Con A (Olsen et al., 1980). Moreover, in a recent study chickens injected with immunosuppressive numbers of REV-T-producing transformed cells failed to develop detectable levels of REV-T-specific cytotoxic lymphocytes, while animals given lower nonsuppressive numbers of cells did (Walker et al., 1983). Recently, however, direct evidence that MuLV-induced immunodepression is a prerequisite for the progression of the induced leuke-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

161

mia has begun to accumulate. Transfer of peritoneal macrophages from normal syngeneic donors to FLC-leukemic mice caused regression in a significant proportion of otherwise progressing animals. Since the number of macrophages needed to induce regression was small, it was concluded that the efficacy of the treatment was not due to a direct antileukemia effect of the macrophages but was related to alleviation of FLC-induced immunodepression (Marcelletti and Furmanski, 1978).This is in line with observations that the induction of T cytotoxic responses (Treves et al., 1976; Tanyiama and Holden, 1979b; Gomard et al., 1981; Wybier-Franqui et al., 1982; Biasi et at., 1983)and also the triggering of established immune responses (Landolfo et al., 1977; Sharma et d . , 1979; Koppi and Halliday, 1982) against retroviral antigens is dependent on the availability of functional macrophages. In another study, successful serum therapy of FLC-infected mice was correlated with the ability of such treatment to prevent the establishment of the immunosuppressed state, possibly by reducing the level of circulating virus below the immunosuppressive threshold (Schiifer et al., 1976; Genovesi et al., 1982). It is clear, however, that the paucity of relevant information still precludes any definite conclusion about the precise importance of RVassociated immunodeficiency as a determinant of pathogenicity (Britt and Chesebro, 1983) and about the precise steps (viral vs. neoplastic, for example) where it is operative. Further studies should concentrate on these aspects and also on the possible interplay of nonspecific immunodeficiency with more specific forms of unresponsiveness observed in RV-infected hosts (Myburgh and Mitchison, 1976; Wedderbum et al., 1976; Wood, 1976; Collavo et al., 1981; Biasi et al., 1983). B. REDUCEDRESISTANCETO SUPERINFECTION That RV infections can increase the risk of developing serious secondary illness of fungal, bacterial, or viral origin is well documented in the cat. Cats infected with FeLV under natural conditions exhibit a greatly enhanced incidence of intercurrent diseases such as septicemia, stomatitis, peritonitis, pneumonia, and systemic infections sustained by microorganisms that are easily controlled by uninfected animals (Cotter et al., 1975; Essex, 1977; Hardy, 1980,1982). Indeed, it has been upheld that in free-living cats FeLV causes a higher lethality by functioning as immunodepressant than by acting as an oncogenic agent (Essex and Worley, 1981). Opportunistic infections (Candida albicans, Cryptosporidium, cytomegalovirus, etc.) and failure to thrive were prominent symptoms

162

MAURO BENDINELLI ET AL.

also in Macaca mulatta monkeys inoculated with MPMV (Fine et al., 1975; Fine and Schochetman, 1978) and in M . cyclopsis monkeys infected with a related type D virus originally isolated from a spontaneous lymphoma and subsequently incriminated as the possible causative agent of an AIDS-like acquired immunodeficiency that has occurred in colonies of captive monkeys (Daniel et al., 1984; Marx et al., 1984). In mice, there are no data dealing with naturally acquired superinfections, but a number of experiments have clearly shown that susceptibility to microbial challenge is highly enhanced by RV infections. In early studies, FLC and Mo-MuLV increased the severity of subsequent murine hepatitis viral infection (Gledhill, 1961), and other MuLV markedly enhanced parasitemia and mortality by protozoa, such as Babesia microti and Plasmodia, that are usually easily controlled by normal mice (Cox and Wedderburn, 1972; Cox et al., 1974). More recently, FLC and F-MuLV were seen to greatly augment the susceptibility of mice to several bacterial and viral pathogens (Bendinelli et al., 1985). The doses of cytomegalovirus required to establish a progressing infection. in or to kill FLC-infected mice were significantly reduced as compared to those required in normal mice, thus paralleling the increased risk of leukemic patients to develop disease and die due to cytomegalovirus infection (Mayo and Rap, 1980).Thus, parenthetically, MuLV-infected mice ought to represent ideal models for evaluating immune response modifiers as possible therapeutic tools for boosting antimicrobial defenses in immunosuppressed subjects. VII. The Retroviral Etiology of AIDS

By stringent criteria, a case of AIDS is defined clinically by a laboratory-proven Kaposi’s sarcoma and/or a life-threatening opportunistic infection in a person under 60 years of age with no history of underlying immunosuppressive illness or therapy. Since 1981, when this “novel” nosological entity became manifest, more than 10,OOO such cases have been reported in the United States and several hundreds have been identified in other countries. Their fatality rate is appallingly high. It is, however, generally agreed that the above definition only covers the most severe and advanced manifestations of the disease, and a number of affections of varying severity have been tentatively related to AIDS. Among these, the best defined is known as “lymphadenopathy syndrome” ( U S ) and is characterized by an unexplained per-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

163

sistent generalized lymphadenopathy with or without systemic symptoms, but without recognized opportunistic infections or neoplasms. At present, to establish with certainty whether and to what extent LAS and other forms of the “AIDS-related complex” are prodromal or mild AIDS or rather independent entities is impossible in the absence of pathognomonic features that precisely delineate the various syndromes and before the results of well-designed prospective cohort studies become available. Nevertheless, the fact that a lymphadenopathy represents the first clinical sign in 20-40% of AIDS patients favors the former concept (Centers for Disease Control, 1982a,b, 1984a,b). The clinical presentation of full-blown AIDS is quite varied (Fauci, 1983). The list of opportunistic pathogens detected in patients ranges from protozoa (Pneumocystis carinii pneumonia is present in 60% of cases) to helminths, fungi, bacteria, and viruses. The infections are particularly severe and persistent, often multiple, occurring either sequentially or simultaneously, and tend to localize at unusual organs or to become disseminated. Kaposi’s sarcoma, that was previously very rare in North America and Europe, is present in 30% of AIDS patients and is usually more aggressive than the classical types previously known in such countries. Other neoplastic complications include Burkitt’s-like lymphomas, squamous cell carcinomas of the anus and oral cavity, and additional proliferative diseases which share with the formers a suspected viral origin. Such various clinical presentations are linked by an underlying disorder of the immune system, as evidenced by the alteration of a number of immunogical markers and functions. In the frank disease the most prominent changes are an absolute lymphopenia, a profound diminution of circulating T4 lymphocytes, inversion of the T4/T8 ratio, anergiahypoergia to cutaneous recall antigens, and diminished in vitro lymphocyte responsiveness to antigens and mitogens. Other frequent abnormalities include elevated serum levels of immunoglobulins, a-thymosin and acid-labile a-interferon, impaired production of lymphokines (Murray et al., 1984), increased secretion of polyclonal antibody by B lymphocytes in uitro, and failure to make antibody responses upon deliberate immunization. Often NK and T cytotoxic activities are also defective (Seligmann et al., 1984). The pattern of infectious complications is compatible with deficits in the cells of the monocyte-macrophage lineage, and deficient chemotaxis and other defects of such cells have been observed (Smith et al., 1984). Alterations of accessory cells that could be detrimental to antigen presentation have also been described (Belsito et al., 1984; Kirkpatrick et al.,

164

MAURO BENDINELLI ET AL.

1984). In addition, peripheral lymphoid organs and thymus exhibit a varied degree of T lymphocyte depletion and other profound structural changes (Millard, 1984). Individuals with prodromal AIDS and other syndromes of the complex are more heterogeneous in that the above immunological and histological abnormalities are present irregularly and in varied combinations and are generally less profound (Seligmann et al., 1984). AIDS is progressive and, as far as it is known, constantly lethal. In the attempt to control its devastating course, a variety of immunoreconstitutive treatments and other therapeutic protocols have been used. Though some success has been reported in the control of the neoplastic and infectious complications, no convincing evidence of return to normal immunological functions has yet been reported (Fischinger, 1985). A number of noninfectious causes or predisposing factors had been initially postulated, but very early its epidemiology (temporal and geographical clustering, evidence of case-to-case contact) suggested that AIDS was due to an infectious agent. Furthermore, in North America and Europe the disease has almost exclusively striken certain specific population groups : promiscuous homosexual and bisexual men, intravenous-drug users, Haitians, and hemophiliacs. This led to conclusions that the disease was transmitted sexually and through exposure to blood or blood products. Presently, new groups at risk seem to be emerging (heterosexual partners of people in the aforementioned groups, infants of women at risk, recipients of blood transfusions, health care workers), and a larger spread of the disease to the general population is widely feared (Francis et al., 1983; Curran et al., 1984; Scott et al., 1984; Centers for Disease Controls, 1985a,b,c). On the other hand, in certain areas (equatorial Africa, The Caribbean) the epidemiological pattern seems to be different, showing no identifiable risk factors for the majority of patients (World Health Organization, 1984). Nevertheless, in the epidemiological situation typical of the advanced countries, alternative routes of transmission do not appear to be of importance. The recognition that a blood-borne infectious agent was responsible for the disease started an intensive search for a causative virus. Herpes simplex, Epstein-Barr, hepatitis B, and cytomegalovirus were found to be ubiquitous in the patients as well as in the persons at risk, but were soon dismissed as etiologic agents on solid grounds (although they tend to cause more severe infections in AIDS patients and can contribute to aggravate the immunodeficiency). Attempts to incriminate other known viruses, their variants, and associations of viruses

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

165

also failed (Hirsch et al., 1984; Quinnan et al., 1984).The significance of the vast array of virus-like particles seen by electron microscopy in plasma and lymphocytes was obviously limited (Armstrong, 1984; Sincovics et al., 1984). Because RV were known to be immunodepressive in animals, the human lymphotrophic RV HTLV-1, which has been causally associated with specific T4 cell malignancies occurring in selected world regions, and HTLV-2, first isolated from a T variant of hairy-cell leukemia, were also considered as possible causative agents. Opportunistic infections and T cell imbalances are frequently observed in patients with HTLV-l-positive leukemias. Furthermore, infection with HTLV1or HTLV-2 i n oitro can alter T cell function and, in some cases, lead to T cell death (Mitsuya et al., 1984a; Sarngadharan et al., 1984). For example, infection led a T cell clone with helper functions to induce an indiscriminate polyclonal activation of B cells and other clones to lose their cytotoxic function (Popovic et al., 1984a). However, virus isolation and serological studies clearly showed that the proportion of AIDS patients with signs of exposure to these viruses is low (Gallo et al., 1983, 1984a; Gelmann et al., 1983; Tedder et al., 1984). Subsequently, the attention was pointed onto a new (group of) human RV, whose prototypes were isolated by Montagnier and co-workers (Barre-Sinoussi et al., 1983; Vilmer et al., 1984), by Gallo and coworkers (1984a,b) (Popovic et al., 1984b), and by Levy et al. (1984) from patients with LAS and prodromal or full-blown AIDS. These prototypes have been respectively named lymphadenopathy-associated virus (LAV),HTLV-3, and AIDS-related retrovirus (ARV),but are most probably the same virus or closely related viruses (Feorino et al., 1984; Weiss, 1984). They are distinguishable from HTLV-1 and HTLV-2 morphologically, biochemically, and antigenically. Most importantly, they are not immortalizing as the human RV previously known, but rather tend to be cytopathic. Their relatedness to other RV is presently under scrutiny (Montagnier et al., 1984; Popovic et al., 1984b; Muesing et al., 1985).A resemblance to members of the family Lentivirinae has been emphasized (Gonda et al., 1985). Isolation and serologic procedures are currently being developed and estensively employed to determine the prevalence of exposure to HTLV-3/LAV and to correlate infection with risk of developing clinical symptoms and prognosis (Saxinger and Gallo, 1983; Brun-Vezinet et al., 1984a; Kalyanaraman et al., 1984; Sarngadharan et al., 1984; Schupbach et al., 1984). By now, there is unequivocal evidence that the new virus is the real initiator of the disease and not just another adventitious opportunistic agent. It has been isolated from the blood

166

MAURO BENDINELLI ET AL.

and bone marrow aspirates of a significant proportion of patients (Gallo et al., 1984b; Levy et al., 1984)and in a few instances also from asymptomatic seronegative sexual partners of patients with AIDS or U S (Salahuddin et al., 1984). Furthermore, serological evidence of exposure to HTLV-3/LAV has been shown to be common in patients and in groups at increased risk. For example, in different studies antibodies to a lysate of infected cells were detected in 68-100% of AIDS patients, in 84-100% of subjects with related conditions, in 87% of intravenous-drug abusers living in New York City, in 56-72% of individuals with hemophilia A, and in 22-65% of homosexuals. In contrast, the proportion of positivities among healthy controls from the same areas was less than 1% (Centers for Disease Control, 198413, 198513; Cheisong-Popov et al., 1984; Goedert et al., 1984; Melbye et al., 1984; Ramsey et al., 1984; Tsoukas et al., 1984; Spira et al., 1984; Weiss et al., 1985). Interestingly, the titers of HTLV-3/LAV-specific antibody in AIDS patients, although widely varied, were generally lower in advanced than in newly diagnosed cases (Sarngadharan et al., 1984; Kalyanaraman et al., 1984). The intense scrutiny under way has already led to a tumultuous progress in the understanding of HTLV-3/LAV properties and epidemiology. Since a systematic review of the results appears premature, the following is simply a florilegium of some major results. The virus has been adapted to grow on several established human lymphoid cell lines (Montagnier et al., 1984),and its ability to infect T4 lymphocytes has been linked to the presence of T4 antigen-related receptors (Dalgleish et al., 1984; Klatzmann et al., 1984). The viral genome of various isolates has been cloned and sequenced (Gonda et al., 1985; Muesing et al., 1985; Starcich et al., 1985), and a molecular mechanism for the cytolytic activity has been suggested (Sodroski et al., 1985).The virus has been shown to be inhibited by substances such as ribavirin and suramin (McCornick et al., 1984; Mitsuya et al., 1984b). The presence of virus in the brain, probably responsible for the encephalopathies frequently observed in AIDS patients, as well as in various other tissues of patients has been documented (Shaw et al., 1985).The demonstration that the virus is eliminated with the semen (Ho et al., 1984; Zagury et al., 1984)and saliva (Groopman et al., 1984) has confirmed the transmission routes suspected on epidemiologic grounds. Nonhuman host systems in which the virus reproduces the disease are being actively investigated (Alter et al., 1984; Francis et al., 1984) and are expected to be of great help in developing a protective vaccine and other specific preventive measures. Meanwhile, serology has shown that exposure to HTLV-3/LAV is

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

167

much more common than AIDS itself (Sarngadharan et al., 1984) and that the proportion of apparently healthy seropositive individuals has been steadily increasing in persons at risk from 1978 to 1984. This is believed to reflect the fact that, in analogy with most other infections, host response ranges from subclinical to severe (Center for Disease Control, 1984b). Furthermore, immunoenzymatic assays for antiviral antibodies have been licensed and are extensively used to screen donated blood to be transfused or manufactured into blood derivatives. Although their evaluation is still incomplete, it is hoped that their application-in conjunction with heat treatment to inactivate the virus (Spire et al., 1985)-will reduce the risk of infection associated with the use of such products. Inevitably, such an astonishing rate of progress has raised hopes that the deadly disease will be conquered in the near future. Such hopes will probably be fulfilled. Nevertheless, much remains to be learned. For example, the long-term prognosis of the infected persons who have not developed AIDS is not known. The latency between exposure and the first clinical symptoms of AIDS may range from 1to more than 4 years (Curran et al., 1984). Given such a long incubation, it is to be expected that the outcome of infection is strongly influenced by host, life-style, and environmental factors. To appreciate the public health importance of this aspect, suffice it to mention h a t hundreds of thousands of people are believed to have already been exposed to the virus. Interestingly, serological investigations have established that the infection was already present in Africa years before AIDS became evident in developed countries (Brun-V6zinet et al., 1984a). In any case, the oncogenic potential of many RV remains like a Sword of Damocles. Concerning the immunopathology of the infection, further studies should point to clarify the full range of HTLV-3-susceptible cells. The spectrum of cells supporting HTLV-1 and HTLV-2 replication has proved wider than initially thought (Clapham et al., 1983), and electron microscopy has evidenced RV-like particles within lymph node macrophages (Gyorkey et al., 1985) and in close association with dendritic cells (Armstrong and Horne, 1984) of patients with the prodroma1 stages of AIDS. Detailed attempts should also be made to elucidate whether the functional deficits of B lymphocytes and macrophages are secondary to the defect in T helper cells or a direct consequence of infection and to clarify the mechanisms of immunocompetent cell function disruption. The nature of the soluble immunosuppressive factors present in the serum and/or secreted by lymphocytes and macrophages of AIDS patients (Hennig and Tomar,

168

MAURO BENDINELLI ET AL.

1984; Laurence and Mayer, 1984) should also be investigated. A better understanding of these aspects might suggest rational strategies for reconstituting the immune response of diseased individuals (Lifson et al., 1984). No doubt, the experience gained in the study of RV-induced immunodeficiencies of lower animals will prove to be of great value in this context. VIII. Summary and Perspectives

That RV can adversely affect the functioning of the immune system has now been known for a quarter of a century and the effect has been more or less intensively investigated for nearly as long. In a limited number of infections of lower animals a multiplicity of phenomena pertaining to the whole organism and to isolated immunocompetent cells has been described. Infected hosts have been shown to mount reduced responses against a variety of immunogens and to present a subnormal resistance to superinfecting agents. Functional tests have established that all classes of immunocytes exhibit potentially important alterations in their effector, inducer, or immunoregulatory activities, and several pieces of evidence have suggested the centrality of macrophage impairment. Attempts to clarify the underlying mechanisms have indicated that many factors may contribute but that the direct interaction of the infecting virus with immunocompetent cells is of great importance. Even though it seems likely that active infection is a prerequisite for such cells to become grossly or permanently malfunctioning, the mere physical contact with the virion or with virion components has been shown to perturb both lymphocytes and macrophages. Thus, at least some RV can be viewed as self-replicating immunodepressive agents. Despite such detailed studies, the recent demonstration that RV are the etiologic agents of clinically important immunodeficencies of humans has come as an absolute surprise. On the other hand, even the clinical entities associated with infection by such viruses were entirely unknown until 1981. Though knowledge of the structure and biology of the etiologic RV is growing at a tremendous pace, at this juncture it is difficult to predict to what extent the information obtained from the study of RV-induced immunodeficiencies of lower animals applies to AIDS and related immunological disorders of humans. However, many of the lessons and methodologies derived from such studies are currently being fruitfully applied to primate systems. The expectation is that concerted efforts will translate-hopefully, in the near future-into efficient means for preventing acquired im-

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

169

munodeficiencies of clinical importance and for dealing with opportunistic infections of immunocompromised individuals.

ACKNOWLEDGMENTS The skilled technical assistance of Giulietta Cerretini and Luciana Montagnani and the bibliographical help of Bruno Benvenuti are deeply appreciated. During the preparation of this work M. B.’s laboratory was supported by grants from the Italian Research Council, Special Project “Oncology,” and from the Italian Ministry of Public Education.

REFERENCES Aaronson, S. A., and Stephenson, J. R. (1976). Biochim. Biophys. Acta 458, 323-354. Alter, H. J., Eichberg, J. W., Masur, H., Saxinger, W. C., Gallo, R. C., Macher, A. M., Lane, H. C., and Fauci, A. S. (1984). Science 226,549-552. Anand, R., and Steeves, R. (1980). Nature (London) 286,615-617. Anderson, L. J., Jarrett, W. F. H., Jarrett, O., and Laird, H. M. (1971).J . Natl. Cancer Inst. 47, 807-817. Armstrong, D. (1984). Prog. Med. Virol. 30, 1-13. Armstrong, J. A., and Home, R. (1984). Lancet 2,370-372. Bainbridge, D. R., and Bendinelli, M. (1972).J. Natl. Cancer Inst. 49,773-781. Banes, A. J., and Smith, R. E., (1977). Infect. Immun. 16,876-884. Banks, K. L., and Henson, J. B. (1973). Infect. Immun. 8,679-684. Barre-Sinoussi, F., Chermann, J. C., Rey, F., Nugeyre, M. T., Chamaret, S., Gruest, J., Dauguet, C., Axler-Blin, C., V6zinet-Brun, F., Rouzioux, C., Rozenbaum, W., and Montagnier, L. (1983). Science 220,868-871. Beard, J. W. (1980). In “Viral Oncology” (G. Klein, ed.), pp. 55-87. Raven, New York. Bellanti, J. A., and Herscovitz, H. B., eds. (1984). “The Reticuloendothelial System. A Comprehensive Treatise. Vol. 6. immunology,’’ p. 347. Plenum, New York. Belsito, D. V.,Sanchez, M. R., Baer, R. L., Valentine, F., and Thorbecke, G. J. (1984).N. Engl.]. Med. 310,1279-1282. Bendinelli, M. (1968). Immunology 14, 837-850. Bendinelli, M. (1970). In “Immunity and Cancer in Oncogenesis” (L. Severi, ed.), pp. 623-635. Division of Cancer Research, Perugia. Bendinelli, M. (1973).In “Virus Tumorigenesis and Immunogenesis” (W. S. Ceglowski and H. Friedman, eds.), pp. 181-221. Academic Press, New York. Bendinelli, M. (1977). Experienta 33,455-456. Bendinelli, M. (1981). In “Virus Oncogeni ad RNA” (C. De Giuli Morghen, ed.), pp. 318-330. Piccin, Padova. BendineIIi, M. (1984). In “Immunomodulation: New Frontiers and Advances” (H. H. Fudenberg, H. D. Whitten, and F. Ambrogi, eds.), pp. 161-196. Plenum, New York. Bendinelli, M. (1985).In “The Reticuloendothelial System: A Comprehensive Treatise. Vol. 10. Infection” (M. Escobar and J. P. Utz, eds.). Plenum, New York, in press. Bendinelli, M., and Friedman, H. (1976). Infect. Immun. 14,613-617. Bendinelli, M., and Friedman, H. (1980).Ado. Erp. Med. Biol. 121B, 91-97, Bendinelli, M., and Nardini, L. (1973a).Infect. Immun. 7, 152-159. Bendinelli, M., and Nardini, L. (1973b). Infect. Immun. 7, 160-166. Bendinelli, M., Nardini, L., and Campa, M. (1974).J . Gen. Virol. 22,207-214.

170

MAURO BENDINELLI ET AL.

Bendinelli, M., Campa, M., and Toniolo, A. (1975a).Infect. Immun. 11, 1031-1037. Bendinelli, M., Kaplan, G. S., and Friedman, H. (1975b).J.Natl. Cancerlnst. 55,14251432. Bendinelli, M., Toniolo, A., and Campa, M. (1975~). Infect. Immun. 11, 1024-1030. Bendinelli, M., Toniolo, A., and Friedman, H. (1976a). Ann N.Y. Acad. Sci. 276,431441. Bendinelli, M., Toniolo, A., and Friedman, H. (197613).Ado. Erp. Med. Biol. 66, 507512. Bendinelli, M., Campa, M., Toniolo, A., and Garzelli, C. (1978a).J.Gen. Virol. 39,243251. Bendinelli, M., Matteucci, D., Toniolo, A., and Friedman, H. (1978b).Erperientia 34, 1161-1162. Bendinelli, M., Matteucci, D., Toniolo, A., and Friedman, H. (1979).Infect. Immun. 24, 1-6. Bendinelli, M., Matteucci, D., Toniolo, A., and Friedman, H. (1980).Ado. Erp. Biol. Med. 121B, 493-508. Bendinelli, M., Cox, W. I., Specter, S., and Friedman, H. (1981). Cum. Chemother. lmmunother. 1121-1123. Bendinelli, M., Matteucci, D., and Toniolo, A. (1985). In “Viruses, Immunity, and Immunodeficiency” (A. Szentivanyi, H. Friedman, and S. Specter, eds.). Plenum, New York, in press. Bennett, M., and Steeves, R. A. (1970).J.Natl. Cancer Inst. 44, 1107-1119. Berendt, M. J., and North, R. J. (1980).J . Erp. Med. 151,69-80. Biasi, G., Collavo, D., and Chieco-Bianchi, L. (1983).J. Immunol. 131, 16-18. Bishop, J. M. (1982).Ado. Cancer Res. 37, 1-32. Blank, K. J., and Murasko, D. M. (1980).Nature (London) 283,494-495. Blasecki, J. W., ed. (1981). “Mechanisms of Immunity to Virus-Induced Tumors,” p. 354. Dekker, New York. Bluestone, J. A., and Lopez, C. (1979).J . Nutl. Cancer Inst. 63,1215-1220. Bluestone, J. A., and Lopez, C. (1982).J . Natl. Cancer Inst. 69,953-959. Bose, H. R. (1984). Microbiol. Sci. 1, 107-112. Brahic, M., and Haase, A. T. (1981).In “Comparative Diagnosis of Viral Diseases. Vol. IV. Vertebrates and Related Viruses. Part B. RNA Viruses” (E. Kurstak and C. Kurstak, eds.), pp. 619-658. Academic Press, New York. Britt, W. J., and Chesebro, B. (1983).J . Erp. Med. 157, 1736-1745. Broxmeyer, H. E., Ralph, P., Gilbertson, S., and Margolis, V. B. (1980).Cancer Res. 40, 3928-3933. Brunda, M. J., Holden, H. T., and Herberman, R. B. (1980).J.Immunol. 124,2682-2687. Brun-VBzinet, F., Rouzioux, C., BarrB-Sinoussi, F., Klankmann, D., Saimot, A. G., Rozenbaum, w., Christol, D., Gluckmann, J. c., Montagnier, L., and Chermann, J. c. (1984a). Lancet 1;1253-1256. Brun-VBzinet, F., Ronzioux, C., Montagnier, L., Chamaret, S., Gruest, J., BarrB-Sinoussi, F., Giraldi, D., Chermann, J. C., McCornick, J., Mitchell, S., Piot, P., Taelman, H., Mirlangu, K. B., Odio, W., Mbendi, N., Mazebo, P., Kayembe, K., Bridts, C., Desmyter, J.. Feiusod, F. M., and Quinn, T. C. (1984b). Science 226, 453-456. Bubbers, J. E., Elder, J. H., and Dixon, F. J. (1980).J.Immunol. 124,388-394. Butler, R. C., and Friedman, H. (1979).Ann. N . Y . Acad. Sci. 332,446-450. Butler, R. C., Friedman, H., and Nowotny, A. (1980).Adu. Erp. Med. Biol. 1214 315322.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

171

Butler, R. C., Frier, J. M., Chapekar, M. S., Graham, M. O., and Friedman, H. (1983). Infect. Zmmun. 39, 1260-1264. Carpenter, C. R., Bose, H. R., and Rubin, A. S. (1977). Cell. Zmmunol. 33,392-401. Carpenter, C. R., Rubin, A. S., and Bose, H. R. (1978).]. Immunol. 120, 1313-1320. Carter, J. K., and Smith, R. E. (1984).J. Virol. 50, 301-308. Carter, J. K., Ow, C. L., and Smith, R. E. (1983). Infect. Immun. 39,410-422. Carter, R. L., Chesterman, F. C., Rowson, K. E. K., and Wedderbum, N. (1970).Znt. J. Cancer 5, 103-110. Casali, P., and Oldstone, M. B. A. (1983). Curr. Top. Microbiol. Immunol. 104, 7-48. Casali, P., and Trinchieri, G. (1984).In “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), pp. 11-19. Springer-Verlag, Berlin and New York. Casali, P., Rice, G. P. A., and Oldstone, M. B. A. (1984).J . Exp. Med. 159, 1322-1337. Caulfield, M. J., and Cerny, J. (1980).J. Zmmunol. 124,255-260. Ceglowski, W. S., and Friedman, H. (1969).Nature (London) 224, 1318-1319. Ceglowski, W. S., and Friedman, H. (1970).]. Immunol. 105, 1406-1415. Ceglowski, W. S., and Friedman, H. (1975).Proc. SOC. E x p . Biol. Med. 148, 808-812. Celada, F., and hisjo, B. (1973). In “Virus Tumorigenesis and Immunogenesis” (W. S. Ceglowski and H. Friedman, eds.), pp. 263-270. Academic Press, New York. Centers for Disease Control (1982a).Morb. Mortal. Weekly Rep. 31,249-251. Centers for Disease Control Task Force on Kaposi’s Sarcoma and Opportunistic Infections (1982b). N. Engl. J . Med. 306,248-252. Centers for Disease Control (1984a).Morb. Mortal. Weekly Rep. 33, 337-339. Centers for Disease Control (1984b).Morb. Mortal. Weekly Rep. 33, 378-379. Centers for Disease Control (1985a).Morb. Mortal. Weekly Rep. 34, 1-5. Centers for Disease Control (1985b).Morb. Mortal. Weekly Rep. 34,101-103. Centers for Disease Control (1985~). Morb. Mortal. Weekly Rep. 34, 147-156. Cerny, J., and Essex, M. (1979). In “Naturally Occurring Biological Immuno-Suppressive Factors and Their Relationship to Disease” (R. Neubeneur, ed.), pp. 233-256. CRC Press, Cleveland, Ohio. Cerny, J., and Isaak, D. D. (1979). Int. J. Cancer 23,260-268. Cerny, J., and Stiller, R. A. (1975).J. Irnmunol. 115,943-949. Cerny, J., Essex, M., Rich, M.A., and Hardy, W. D. (1975).Int. J . Cancer 15,351-365. Cheisong-Popov, R., Weiss, R. A., Dalgleish, A., Tedder, R. S., Shanson, D. C., Jeffries, D. J., Ferns, R. B., Briggs, E. M., Weller, I. V. D., Mitton, S., Adler, M. W., Farthing, C., Lawrence, A. G., Gazzard, B. G., Weber, J., Harris, J. R. W., Pinching, A. J., Craske, J., and Barbara, J. A. J. (1984).Lancet 2,477-480. Chesebro, B., Bloom, M., Wehrly, K., and Nishio, J. (1979).J . Virol. 32, 832-837. Chirigos, M. A., Perk, K., Turner, W., Burka, B., and Gomez, M. (1968).Cancer Res. 28, 1055-1063. Choppin, J., Schaffar-Deshayes,L., and Levy, J. P. (1981).J.Immunol. 126,2352-2354. Cianciolo, G. J., Matthews, T. J., Bolognesi, D. P., and Snyderman, R. (1980).J. Immunol. 124,2900-2905. Cianciolo, G . J., Hunter, J., Silva, J., Haskill, J. S., and Snyderman R. (1981).J. Clin. Invest. 68, 831-844. Cianciolo, G . J., Lostrom, M. E., Tam, M., and Snyderman, R. (1983).]. E x p . Med. 158, 885-900. Clapham, P., Nagy, K., Cheisong-Popov, R., Exley, M., and Weiss, R. A. (1983).Science 222,1125-1127. Clark, S. P., and Max, T. W. (1983).Proc. Natl. Acad. Sci. U S A . 80, 5037-5041. Cockerell, G. L., and Hoover, E. A. (1977). Cancer Res. 37,3985-3989.

172

MAURO BENDINELLI ET AL.

Cockerell, G. L., Hoover, E. A., Krakowka, S., Olsen, R. G., and Yohn, D. S. (1976). J . Natl. Cancer Inst. 57,1095-1099. Collavo, D., Biasi, G., Colombatti, A., and Chieco-Bianchi, L. (1975).Br. J . Cancer 32, 331-338. Collavo, D., Zanovello, P., Leuchars, E., Davies, A. J. S., Chieco-Bianchi, L., and Biasi, G. (1980). J . Natl. Cancer Inst. 64,97-104. Collavo, D., Zanovello, P., Biasi, G., and Chieco-Bianchi, L. (1981).J . Immunol. 126, 187-193. Copelan, E. A., Rinehart, J. J., Lewis, M.,Mathes, L., Olsen, R., and Sagone, A. (1983).J. Immunol. 131,2017-2020. Cotter, S. M., Hardy, W. D., and Essex, M. (1975).J.Am. Vet. Med. Assoc. 166,449-453. Cox, F. G. E., and Wedderburn, N. (1972).J. Gen. Microbiol. 72,79-85. Cox, F. G. E., Wedderburn, N., and Sdaman, M. H. (1974).J. Gen. Microbiol. 85,358364. Cox, K. O., and Keast, D. (1973).J . Natl. Cancer Inst. 50,941-946. Cremer, N. E., Taylor, D. 0. N., and Lennette, E. H. (1969).J. Natl. Cancer Inst. 42, 695-702. Crittenden, L. B., Fadly, A. M., and Smith, E. J. (1982). Auian Dis.26,279-294. Curran, J. W., Lawrence, D. N., Jaffe, H.,Kaplan, J. E., Zyla, L. D., Chamberland, M., Weinstein, R., Lui, K.-J., Schonberger, L. B., Spira, T. J., Alexander, W. J., Swinger, G., Ammann, A., Solomon, S., Auerbach, D., Mildvan, D., Stoneburner, R., Jason, J. M., Haverkos, H. W., and Evatt, B. L. (1984).N . EngZ.1. Med. 310,69-75. Dalgleish, A. G., Beverley, P. C. L., Clapham, P. R., Crawford, D. H., Greaves, M. F., y d Weiss R. A. (1984). Nature (London) 312,763-767. Daniel, M. D., King, N. W., Letvin, N. L., Hunt, R. D., Sehgal, P. K., and Desrosiers, R. C. (1984). Science 223,602-605. DeLamarter, J. F., Stoye, J., Schumann, G., and Moroni, C. (1979). Haemutol. Blood Transfus. 23,413-415. Delarco, J. E., Rapp, U. R., and Todaro, G. J. (1978). Int. J. Cancer 21,356-360. DeMaeyer-Guignard, J. (1972). Science 177,797-799. Denner, J., Wunderlich, V., and Bierwolf, D. (1980).Acta Biol. Med. Ger. 39, 19-26. Dent, P. B. (1972). Prog. Med. Virol. 14, 1-35. Deodhar, S. D., and Chiang, T. (1970). Fed. Proc. Fed. Am. SOC. Erp. Biol. 29,560. Dorb, J. F., Schneider, M., and Mathb, G. (1969).Reo. Franc. Etud. Clin. Biol. 14,10031007. Dracott, B. N., Wedderburn, N., and Doenhoff, M. J. (1977). Immunology 33,573-580. Dracott, B. N., Wedderburn, N., and Doenhoff, M. J. (1978). ImmunoZogy 34,679-687. Dunlap, J. E., Nichols, W. S., Hebebrand, L. C., Mathes, L. E., and Olsen, R. G. (1979). Cancer Res. 39,956-958. Durban, E. M., and Boettiger, D. (1981a). Proc. Natl. Acad. Sci. U S A . 78,3600-3604. Durban, E. M., and Boettiger, D. (1981b). Virology 37,488-492. Eckner, R. J. (1975). /. Exp. Med. 142,936-948. Enrietto, P. J., and Wyke, J. A. (1983). Adv. Cancer Res. 39,269-314. Essex, M. (1977). Contemp. Top. Immunobiol. 6,71-79. Essex, M., and Worley, M. (1981).I n “Comparative Diagnosis ofviral Diseases. Vol. IV. Vertebrate and Related Viruses. Part B. RNA Viruses” (E. Kurstak and C. Kurstak, eds.), pp. 553-597. Academic Press, New York. Ewert, D. L., Vainio, O., and Halpern, M. S. (1983).J. Immunol. 131,3036-3041. Fadly, A. M., Lee, L. F., and Bacon, L. D. (1982). Infect. Immun. 37, 1156-1161.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

173

Farber, P., Specter, S., and Friedman, H. (1978).Zniect. Zmmun. 20,816-826. Fauci, A. S. (1983).J . Am. Med. Assoc. 249,2375-2376. Feorino, D. M., Kalyanaraman, V. S., Haverkos, H. W., Cabradilla, C. D., Warfield, D. T., JaEe, H. W., Harrison, A. K., Gottlieb, M. S., Goldfinger, D., Chermann, J.-C., BamSinoussi, F., Sara, T. T., McDougal, J. S., Curran, J. W., Montagnier, L., Murphy, F. A., and Francis, D. P. (1984). Science 225,69-72. Fine, D. L., and Schochetman, G. (1978). Cancer. Res. 38,3123-3139. Fine, D. L., Landon, J. C., Pienta, R. J., Kubicek, M. T., Valerio, M. G., Loeb, W. F., and Chopra, H. C. (1975).J . Natl. Cancer Inst. 54, 651-658. Fischinger, P. J. (1985). Curr. Probl. Cancer 9, 1-39. Fowler, A. K., Twardzik, D. R., Reed, C. D., Weislow, 0. S., and Hellman, A. (1977). Cancer. Res. 37,4529-4531. Francis, D. P., Curran, J. W., and Essex, M. (1983).J . Natl. Cancer Inst. 71, 1-4. Francis, D. P., Feorino, P. M., Broderson, J. R., McClure, H. M., Getchell, J. P., McGrath, C. R., Swenson, B., McDougal, J. S., Palmer, E. L., Harrison, A. K., Barr6Sinoussi, F., Chermann, J.-C., Montagnier, L., Curran, J. W., Cabradilla, C. D., and Kalyanaraman, V. S. (1984). Lancet 2, 1276-1277. Fredrickson, T. N., Lobue, J., Alexander, P., Schultz, E. F., and Gordon, A. S. (1972).J . Natl. Cancer Inst. 48, 1597-1603. Fredrikson, T. N., Langdon, W. Y., H o h a n , P. M., Hartley, J. W., and Morse, H. C. (1984).J . Natl. Cancer Inst. 472,447-454. Friedman, H. (1974). Isr. J. Med. Sci. 10, 1052-1074. Friedman, H., and Ceglowski, W. S. (1968).Nature (London) 218, 1232-1234. Friedman, H., and Ceglowski, W. S. (1971).J . Immunol. 107, 1673-1681. Friedman, H., and Ceglowski, W. S. (1973).In “The Role of Immunological Factors in Viral and Oncogenic Processes” (R. F. Beers, R. C. Telghman, and E. G. Bassett, eds \, pp. 187-209. Johns Hopkins Univ. Press, Baltimore, Maryland. Friedman, H., and Kateley, J. R. (1975).Am. J . Clin. Pathol. 63, 735-747. Friedman, H., Specter, S., Farber, P., and Ceglowski, W. S. (1979). Proc. Leukocyte Cult. Conf. 12,783-802. Furmanski, P., Dietz, M., Fouchey, S., Hall, L., Clymer, R., and Rich, M. A. (1979).J . Natl. Cancer Inst. 63,449-454. Gallo, R. C., Sarin, P. S., Gelmann, E. P., Robert-Guroff, M., Richardson, E., Kalyanaraman, V. S., Mann, D., Sidhu, G. D., Stahl, R. E., Zolla-Paznzer, S., Leibowitch, J., and Popovic, M. (1983). Science 220, 865-867. Gallo, R. C., Essex, M. E., and Gross, L., eds. (1984a). “Human T Cell Leukemia/ Lymphoma Viruses,” p. 437. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Gallo, R. C., Salahuddin, S. Z., Popovic, M., Shearer, G. M., Kaplau, M., Haynes, B. F., Palker, T. J., Redfield, R., Oleske, J., Safai, B., White, G., Foster, P., and Markham, P. D. (1984b). Science 204,500-503. Garaci, E., Migliorati, G., Jezzi, T., Bartocci, A., Gioia, L., Rinaldi, C., and Bonmassar, E. (1981). Int. J . Cancer 28,367-373. Gardner, M. B. (1978). Cuw. Top. Microbiol. lmmunol. 79,215-259. Gazzolo, L., Moscovici, M. G., and Moscovici, C. (1974). Virology 58,514-525. Gazzolo, L., Moscovici, M. G., Moscovici, C., and Vogt, P. K. (1975).Virology 67,553565. Gazzolo, L., Moscovici, C., and Moscovici, M. G. (1979).Infect. Immun. 23, 294-297.

174

MAURO BENDINELLI ET AL.

Gelmann, E. P., Popovic, M., Blayney, D., Masur, H., Sidhu, G., Stahl, R. E., and Gallo, R. C. (1983). Science 220,862-865. Genovesi, E. V., Livnat, D., and Collins, J. J. (1982).Int. J . Cancer 30,609-624. Gerson, J. M., Varesio, L., and Herberman, R. B. (1981).Int. J. Cancer 27,243-248. Gillette, R. W., and Fox, A. (1977).J. Natl. Cancer Inst. 58,1621-1628. Gisselbrecht, S., Blaineau, C., Hurot, M. A., Pozo, F., and Levy, J. P. (1978).Cancer Res. 38,939-941. Gledhill, A. W. (1961). Br. J . Cancer 15,531-538. Godard, C. M., Augery, Y., and Jasmin, C. (1983).J . Gen. Virol. 64,2051-2055. Goedert, J. J., Samgadharan, M. B., Biggar, R. J., Weiss, S. H., Winn, D. M., Grossmann, R. J., Greene, M. H., Bodner, A. J., Mann, D. L., Strong, D. M., Gallo, R. C., and Blamer, W. A. (1984).Lancet 2,711-716. Gomard, E., Wybier-Franqui, J., and Levy, J. P. (1981).J. Immunol. 126,891-896. Gonda, M. A., Wong-Staal, F., Gallo, R. C., Clements, J. E., Narayan, O., and Gilden, R. V. (1985). Science 227, 173-177. Grant, C. K., Emisse, B. J., and Pontefract, R. (1984).Cancer Res. 44,498-502. Greenberger, J. S., Wroble, L. M., and Sakakeeny, M. A. (1980).]. Natl. Cancer Inst. 65, 841-851. Greenspan, E., and Wainberg, M. A. (1981).J. Reticuloendothel. SOC. 29, 105-116. Grinwich, K. D., Alexander, T. S., and Cemy, J. (1979).J. Immunol. 122, 1108-1115. Groopman, J. E., Salahuddin, S. Z., Sarngadharan, M. G., Markham, P. D., Gonda, M., Sliski, A., and Gallo, R. C. (1984). Science 226,447-449. Gyorkey, F., Melnick, J. L., Sinkovics, J. G., and Gvorkey, P. (1985).Lancet 1, 106. Halasa, J.,Friedman, H., and Ceglowski, W. S. (1972).]. Reticuloendothel. SOC. 11,468475. Hanafusa, H. (1977). Compr. Virol. 10,401-483. Haran-Ghera, N., Ben-Yaakov, M., and Peled, A. (1977).J. Immunol. 118,600-606. Hardy, W. D. (1980). In “Feline Leukemia Virus” (W. D. Hardy, M. Essex, and A. J. McClelland, eds.), pp. 3-78. Elsevier, Amsterdam. Hardy, W. D. (1982). Springer Semin. Immunopathol. 5,75-106. Hartzfeld, A., Pinter, A., Koo, G. C., and Boyse, E. A. (1981).lmmunogenetics 12, 153159. Hatten, B. A., and Dunton, H. (1978). Immunology 35,707-713. Hayward, W. S., Neel, B. G., and Astrin, S. M. (1982).Ado. Viral Oncol. 1,207-233. Hebebrand, L. C., Mathes, L. E., and Olsen, R. G. (1977).Cancer Res. 37,4532-4533. Hebebrand, L. C., Olsen, R.G., Mathes, L. E., and Nichols, W. S. (1979).Cancer Res. 39,443-447. Heineger, H.-J., Taylor, B. A., Hards, E. J., and Meier, H. (1975).Cancer Res. 35,825831. Hellman, A., Fowler, A. K., Twardzik, D. R., Weislow, 0. S., and Reed, C. D. (1980). Adu. E x p . Med. Biol. 121B, 99-109. Hennig, A. K., and Tomar, R. H. (1984). Clin. Immunol. Immunopathol. 33,258-267. Henson, J. B., and McGuire, T. C. (1974). Prog. Med. Virol. 18, 143-159. Herberman, R. B. (1982). In “Immunology of Human Infection. Part 11. Viruses and Parasites; Immunodiagnosis and Prevention of Infectious Diseases” (A. J. Nahmias and R. J. OReilly, eds.), pp. 273-314. Plenum, New York. Herberman, R. B. (1983).Adv. Host Defense Mech. 2,241-273. Herberman, R. B., Holden, H. T., Varesio, L., Taniyama, T., Puccetti, P., Kirchner, H., Gerson, J., White, S., and Keisari, Y. (1980). Contemp. Top. Immunobiol. 10, 6178.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

175

Hilgers, J., and Bentvelzen, P. (1978). Adu. Cancer Res. 26, 143-195. Hirano, S., Friedman, H., and Ceglowski, W. S. (1971).J. Immunol. 107, 1400-1409. Hirota, Y., Martin, M. T., Viljanen, M., Toivanen, P., and Franklin, R. M. (1980). Eur.1. Immunol. 10,929-936. Hirsch, M. S., Schooley, R. T., Ho, D. D., and Kaplan, J. C. (1984). Reu. lnf. Dis.6,726731. Ho, D. D., Schooley, R. T., Rota, T. R., Kaplan, J. C., Flynn, T., Salahuddin, S. Z., Gonda, M. A., and Hirsh, M. S. (1984). Science 226, 451-453. Hooks, J. J., and Detrick-Hooks, B. (1981). In “Comparative Diagnosis ofviral Diseases. Vol. IV. Vertebrate and Related Viruses. Part B. RNA Viruses” (E. Kurstak and C. Kurstak, eds.), pp. 599-643. Academic Press, New York. Hooks, W. A., Chirigos, M. A., and Chan, S. P. (1969). Cancer Res. 29, 1008-1012. Hoover, E. A,, Penyman, L. E., and Kociba, G. J. (1973). Cancer Res. 33, 145-152. Hoover, E. A., Rojko, J. L., Wilson, P. L., and Olsen, R. G. (1981).J.Natl. Cancer Inst. 67,889-898. Horak, I., Enjuanes, L., Lee, J. C., and Ihle, J. N. (1981).J. Virol. 37,483-487. Humes, J. L., Cupo, J. J., and Strausser, H. R. (1974). Prostaglandins 6,463-473. Ihle, J. N., and Lee, J. C. (1982). Curr. Top. Microbiol. Immunol. 98, 85-101. Israel, E., and Wainberg, M. A. (198l).J.Reticuloendothel. SOC. 29, 105-116. Israel, E., Yu, M., and Wainberg, M. A. (1979). Immunology 38,41-50. Israel, E., Beiss, B., and Wainberg, M. A. (1980). Immunology 40,77-85. Isaak, D. D., Price, J. A., Reinisch, C. L., and Cerny, J. (1979).J.Immunol. 123,18221828. Jaenish, R. (1976). Proc. Natl. Acad. Sci. U S A . 73, 1260-1264. Jarrett, W. F. H. (1975). B r . J . Cancer31,147-151. Johnson, C. S., Fouchey, S. P., and Furmanski, P. (1980).J. Natl. Cancer Inst. 64,645653. Johnson, C. S., Marcelletti, J., and Furmanski, P. (1984).J.Natl. Cancer Inst. 72,491496. Kalyanaraman, V. S., Cabradilla, C. D., Getchell, J. P., Narayanan, R., Braff, E. H., Chermann, J.-C., Barre-Sinoussi, F., Montagnier, L., Spira, T. J., Kaplan, J., Fishbein, D., Jaffe, H. W., Curran, J. W., and Francis, D. P. (1984). Science 225, 321-323. Kamo, I., and Friedman, H. (1977). Ado. Cancer Res. 25, 271-321. Kamo, I., Kateley, J. R., Kaplan, G., and Friedman, H. (1975). Proc. SOC. E x p . Biol. Med. 148,383-386. Karmali, R. A., Wustrow, T., Thaler, H. T., and Good, R. A. (1984).Cancer Res. 44,467471. Kassan, S. S., and Chused, T. M. (1977). Cell. Immunol. 30, 135-146. Kateley, J. R., and Friedman, H. (1976). Ann. N . Y. Acad. Sci. 276,466-478. KateIey, J. R., Holderback, J., and Friedman, H. (1974).1.Natl. Cancer Inst. 53, 11351140. Kateley, J. R., Kamo, I., Kaplan, G., and Friedman, H. (1974).J . Natl. Cancer Inst. 53, 1371-1378. King, N. W., Hunt, R. D., and Letvin, N. L. (1983). Am. J. Pathol. 113,382-388. Kirchner, H., Muchmore, A. V., Chused, T. M., Holden, H. T., and Herberman, R. B. (1975). J . Immunol. 114,206-210. Kirkpatrick, C. H., Davis, K. C., and Horsburgh, C. R. (1984).N . Eng2.J. Med. 311,857858. Kiyohashi, A. (1981). Nippon Gun Chiryo Gakkai Shi. 16,641-646.

176

MAURO BENDINELLI ET AL.

Klatzmann, D., Champagne, E., Chamaret, S., Gruest, J., Guetard, D., Hercend, T., Gluckman, J.-C., and Montagnier, L. (1984).Nature(London) 312,767-771. Klein, B., Le Bousse-Kerdiles, C., Smadja-Joffe, F., Pragnell, I., Ostertag, W., and Jasmin, C. (1982).Erp. Hematol. 10,373-380. Klein, G., ed. (1980).“Viral Oncology,” p. 842.Raven, New York. Klein, G., ed. (1982a).“Advances in Viral Oncology. Vol. 1. Oncogene Studies,” p. 262. Raven, New York. Klein, J. (1982b).“Immunology. The Science of Self-Nonself Discrimination,” p. 687. Wiley, New York. Komitowski, D., and Kurka, J. (1974).Erperientia 30,648-649. Koo, G. C., Ceglowski, W. S., Higgins, M., and Friedman, H. (1971).J . Immunol. 106, 815-830. Koppi, T. A., and Halliday, W. J. (1982).Immunol. Lett. 4, 135. Koury, M. J., and Pragnell, I. B. (1982).Nature (London)299,638-640. Kumar, V., and Bennett, M. (1981).Curr. Top. Microbiol. Immunol. 92,65-82. Kumar, V., Goldschmidt. L., Eastcott, J. W., and Bennett, M. (1978).J. Erp. Med. 147, 422-433. Kuzumaki, N., Kodama, T., Takeichi, N., and Kobayashi, H. (1974).Int. J . Cancer 14, 483-492. Landolfo, S., Herberman, R. B., and Holden, H. T. (1977).]. Immunol. 118,1244-1248. Langdom, W. Y., and Shellam, G. R. (1981).Cell. Immunol. 64, 102-118. Langweiler, M., Cockerell, G. L., and deNoronha, F. (1983).Cancer. Res. 43, 19571960.

Laurence, J., and Mayer, L. (1984).Science 225,66-69. Lee, J. C., and Ihle, J. N. (1976).J. Immunol. 117,1033-1038. Levine, A. J. (1982).Ado. Cancer. Res. 37,75-109. Levy, P., and Leclerc, J. C. (1977).Ado. Cancer Res. 24, 1-66. Levy, J. A., Hofhan, A. D., Kramer, S. M., Landis, J. A., Shimabukuro, J. M., and Oshiro, L. S. (1984).Science 225,840-842. Lieberman, D., and Sachs, L. (1979).Proc. Natl. Acad. Sci. U S A . 76,3353-3357. Lifson, J. D., Benike, C. J., Mark, D. F., Koths, K., and Engleman, E. G. (1984).Lancet 1, 698-702. Linial, M. (1982).Virology 119,382-391. Lust, J. A., Bennett, M., and Kumar, V. (1984).Int. J. Cancer 33,107-113. McAlack, R. F.,Cerny, J., and Friedman, H. (1971).J . Zmmunol. 107, 1752-1765. McCombs, C., Hom, J., Talal, N., and Mishell, R. I. (1975).]. Immunol. 115,1695-1699. McCornick, J. B., Getchell, J. P., Mitchell, S. W., and Hicks, D. R. (1984).Lancet 2, 1367-1368. McGany, M. P., Styles, B. D., and Mirand, E. A. (1978).J . Natl. Cancer Inst. 60,803810. Marcelletti, J., and Furmanski, P. (1978).J. Immunol. 120, 1-8. Marcelletti, J., and Furmanski, P. (1979).Cell 16,649-659. Margolese, R. G., Israez E., and Wainberg, M. A. (1980).Clin Exp. Immunol. 41,243251. Marshall, G. D., and Foster, B. G. (1978).FEMS Microbiol. Lett. 3,43-46. Martig, R. J., and Tribble, J. L. (1974).Infect. Immun. 10, 1018-1022. Marx, P. A.,and Wheelock, E. F. (1976).Ann. N.Y. Acad. Sci. 276,502-512. Marx, P. A., Maul, D. H., Osborn, K. G., Lerche, N. W., Moody, P., Lowestine, L. J., Henrickson, R. V., Arthur, L. O., Gilden, R. V., Gravell, M., London, W. T., Sever, J. L., Levy, J. A., Munn, R. J., and Gardner, M. B. (1984).Science 223, 1083-1086.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

177

Mascio, A. A., and Ceglowski, W. S. (1978).Infect. Immun. 19,486-492. Mathes, L. E., Olsen, R. G., Hebebrand, L. C., Hoover, E. A., and Schaller, J. P. (1978). Nature (London)274,687-689. Mathes, L. E., Olsen, R. G., Hebebrand, L. C., Hoover, E. A., Schaller, J. P., Adams, P. W., and Nichols, W. S. (1979). Cancer Res. 39,950-955. Matthews, R. E. F. (1982).Znteruirology 17, 1-199. Mayo, D. R., and Rapp, F. (1980). Infect. Immun. 29,311-315. Melbye, M., Biggar, R. J., Ebbesen, P., Sarngadharan, M. G., Weiss, S. H., Gallo, R. C., and Blattner, W. A. (1984). Br. Med. J . 289,573-575. Meredith, R. F., Okunewick, J. P., Brozovich, B. J., and Weaver, E. V. (1978). Dew. Comp. Zmmunol. 2,133-146. Meruelo, D., and Bach, R. (1983).Adu. Cancer Res. 40, 107-188. Metcalf, D., and Moulds, R. (1967). I n t . J . Cancer 2, 53-58. Metcalf, D., Ishidata, M., and Brumby, M. (1967).J . Natl. Cancer Znst. 38, 527-539. Migliorati, G., Jezzi, T., Frati, L., Bonmassar, E., Rossi, G. B., Garaci, E., and Riccardi, C. (1983).Znt. J . Cancer 31,81-90. Millard, P. R. (1984).J.Pathol. 143,223-239. Mitzuya, H., Guo, H. G., Megson, M., Trainor, C., Reitz, M. S., and Broder, S. (1984a). Science 223,1293-1296. Mitsuya, H., Popovic, M., Yarchoan, R., Matsushita, S., Gallo, R. C., and Broder, S. (198413). Science 226, 172-174. Mizel, S. B., Delarco, J. E., Todaro, G. J., Farrar, W. L., and Hilkifer, M. L. (1980).Proc. Natl. Acad. Sci. U S A . 77,2205-2208. Montagnier, L., Gruest, J., Chamaret, S., Dauguet, C., Axler, C., GuBtard, D., Nugeyre, M. T., Barr6-Sinoussi, F., Chermann, J.-C., Brunet, J. B., Klatzmann, D., and Gluckman, J. C. (1984). Science 225, 63-66. Moody, D. J., Specter, S., Bendinelli, M., and Friedman, H. (1984).J.Natl. Cancer Znst. 72,1349-1356. Moroni, C., and Schumann, G. (1977).Nature (London)269,600-601. Moroni, C., Forni, L., Hunsmann, G., and Schumann, G. (1980).Proc. Natl. Acad. Sci. U S A . 77,1486-1490. Morse, H. C., and Hartley, J. W. (1981).Curr. Top. Microbiol. Zmmunol. 98, 17-26. Mortensen, R. F., Ceglowski, W. S., and Friedman, H. (1974).J. Zmmunol. 112,20772086. Moscovici, C., and Gazzolo, L. (1982).Adu. Viral. Oncol. 1, 83-106. Moscovici, C., Samarut, J., Gazzolo, L., and Moscovici, M. G. (1981).Virology 113,765768. Muesing, M. A., Smith, D. H., Cabradilla, C. D., Benton, C. V., Laski, L. A., and Capon, D. J. (1985). Nature (London)313,450-458. Murray, H. W., Rubin, B. Y., Masur, H., and Roberts, R. B. (1984).N . E n g l . ] . Med. 310, 883 889. Myburgh, J. A., and Mitchison, S. A. (1976).Transplantation 22,236-244. Narayan, O., Wolinsky, J. S., Clements, J. E., Strandberg, J. D., Griffin, D. E., and Cork, L. C. (1982).J . Gen. Virol. 59, 345-356. Narayan, O., Kennedy-Stoskopf, S., Sheffer, D., Griffin, D. E., and Clements, J. E. (1983). Infect. Zmmun. 41,67-73. Nichols, W. S., Dunlap, J. E., Mathes, L. E., and Olsen, R. G. (1979).Zmmunol. Lett. 1, 67-71. Normann, S . J., Schardt, M., and Sorkin, E. (1981).J. Natl. Cancer Znst. 66, 157-162. Odaka, T., and Kohler, K. (1965).2. Naturforsch. 20b, 473-476.

178

MAURO BENDINELLI ET AL.

Okunewick, J. P., Meredith, R. F., Brozovich, B., and Weaver, E. V. (1978).Proc. SOC. Exp. Biol. Med. 157,449-452. Old, L. J., Clarke, D. A., Benaceraff, B., and Goldsmith, M. (1960).Ann. N.Y. Acad. Sci. 88,264-288. Oldstone, M. B. A. (1975).Prog. Med. Virol. 19,84-119. Oldstone, M. B. A., Fujinami, S. R., and Lampert, P. W. (1980).Prog. Med. Virol.26,4593. Oldstone, M. B. A., Rodriguez, M., Daughaday, W. H., and Lampert, P. W. (1984). Nature (Lodun) 307,278-281. Olsen, R. G., ed. (1979).“Feline Leukemia,” pp. 289. CRC Press, Boca Raton, Florida. Olsen, R. G., Mathes, L. E., Hoover, E. A., Nichols, S. W., and Hebebrand, L. C. (1980). Am. J . Pathol. 98,857-860. Ostertag, W., and Pragnell, I. B. (1981).Curr. T o p . Microbiol. Immunol. 94, 143-208. Pack, F. D., and Chapman, W. L. (1980). Exp. Pathol. 18,96-110, Perkins, E. H., Makinodan, T., Upton, A. C., Seibert, C., and Satterfield, L. C. (1971).J. Natl. Cancer Znst. 46,845-859. Pessano, S., Gazzolo, L., and Moscovici, C. (1979).Microbiobgica 2,379-392. Peterson, R. D. A., Hendrickson, R., and Good, R. A. (1963).Proc. SOC. Exp. B i d . Med. 114,517-520. Pettey, C. L., and Collins, J. J. (1984).Int. J . Cancer 34,269-276. Phillips, M. G., Gleichmann, H., Hirsch, M. S., Black, P., Memll, J. P., Schwartz, R. S . , and Carpenter, C. B. (1975).Cell. Zmmunol. 33,392-405. Plata, F., Goodenow, M. M., and Lilly, F. (1980).I. Exp. Med. 151,726-742. Plater, C., Debr6, P., and Leclerc, J. C. (1981). Eur.J. Zmmunol. 11,39-44. Popovic, M., Flomenberg, N., Volkman, D. J., Mann, D., Fauci, A. S., Dupont, B., and Gallo, R. C. (1984a). Science 226,459-462. Popovic, M., Sargadharan, M. G., Read, E., and Gallo, R. (1984b).Science 224,497-500. Preble, 0. T., and Friedman, R. M. (1983).Lab. Inoest. 49,4-18. Price, J. A., and Smith, R. E. (1982).Cancer Res. 42,3617-3624. Puccetti, P., and Holden, H. T. (1979).Znt. J . Cancer 23, 123-133. Purchase, H. G., and Witter, R. L. (1975).Cum. T o p . Microbiol. Zmmunol. 71,103-104. Quinnan, G. V., Masur, H., Rook, A. H., Armstrong, G., Frederick, W. R., Epstein, J., Manischewitz, J. F., Macher, A. M., Jackson, L., Ames, J.. Smith, H. A., Parker, M., Pearson, G. R., Parrillo, J., Mitchell, C., and Straus, S. E. (1984).]. Am. Med. Assoc. 252,72-77. Radzichovskaja, R. (1967).Proc. SOC. E r p . Biol. Med. 126, 13-18. Ram, M. D., Kohn, R. R., and Novak, D. (1974).J . Natl. Cancer Inst. 52, 1505-1514. Ramsey, R. B., Palmer, E. L., McDougal, J. S., Kalyanaraman, V. S., Jackson, D. W., Chaba, T. L., Holman, R. C., and Evatt, B. L. (1984).Lancet 2,397-398. Rangan, S . R. S., and Bang, F. B. (1967). Proc. SOC. Exp. Biol. Med. 125,593-598. Raschke, W. C., Baird, S., Ralph, P., and Nakoinz, I. (1978).Cell 15,261-267. Robinson, M. K., Manly, K. F., and Evans, M. J. (1980).J . Immunol. 124, 1022-1027. Roder, J. C., Tyler, L., Ball, J. K., and Singhal, S. K. (1978).Cell. Zmmunol. 36,128-142. Rojko, J. L., Hoover, E. A., Mathes, L. E., Olsen, R. G., and Schaller, J. P. (1979).J.Natl. Cnncer Inst. 63,759-768. Rojko, J. L., Hoover, E. A., Finn, B. L., and Olsen, R. G. (198l).J.Natl. Cancer Znst. 67, 899-910. Rojko, J. L., Hoover, E. A., Quackenbush, S. L., and Olsen, R. G. (1982).Nature (London) 298,385-388. Roman, J. M., and Golub, E. S . (1976)./. Exp. Med. 143,482-496.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

179

Rosenberg, N. (1982). Cum. Top. Microbiol. Immunol. 101,95-126. Roszman, T. L., Brooks, W. H., Markesbery, W. R., and Bigner, D. D. (1973). Cancer Res. 38,74-77. Rowson, K. E. K., and Mahy, B. W. (1975). Virol. Monogr. 13, 1-121. Rudczynski, A. B., and Mortensen, R. F. (1978).J. Natl. Cancer Inst. 60,205-211. Ruddle, N. K., Armstrong, M. Y. K., and Richards, F. F. (1976).Proc. Natl. Acad. Sci. U.S.A.73,3714-3718. Rup, B. J., Spence, J. L., Hoelzer, J. D., Lewis, R. B., Carpenter, C. R., Rubin, A. S., and Bose, H. R. (1979).J . Zmmunol. 123, 1362-1370. Rup, B. J., Hoelzer, J. D., and Bose, H. R. (1982). Virology 116,61-71. Russell, J. L., and Golub, E. S. (1980).J . Zmmunol. 125,2328-2333. Russell, S . W., and McIntosh, A. T. (1977).Nature (London) 268,69-71. Russell, S . W., Doe, W. F., and Cochrane, C. G. (1976).J. Immunol. 116, 164-166. Ruta, M., Bestwick, R., Machida, C., and Kabat, D. (1983).Proc. Natl. Acad. Sci. U.S.A. 80,4704-4708. Salahuddin, S. Z., Groopman, J. E., Markham, P. D., Sarngadharan, M. G., Redfield, R. R., McLane, M. F., Essex, M., Sliski, A., and Gallo, R. C. (1984).Lancet 2, 14181420. Salaman, M. H. (1969).Antibiot. Chemother. 15,393-406. Sarngadharan, M. G., Popovic, M., Bruch, L., Schupbach, J., and Gallo, R. C. (1984). Science 224,506-508. Saxinger, C., and Gallo, R. C. (1983).Lab. Znuest. 49,371-377. Sch&fer, W., Schwartz, H., Thiel, H. J., Wecker, E., and Bolognesi, D. P. (1976).Virology 75,401-418. Schenk, P. J., and Howe, M. L. (1979).J . Immunol. 122, 1874-1880. Schupbach, J,, Popovic, M., Gilden, R. V., Gonda, M. A., Sarngadharan, M. G., and Gallo, R. C. (1984). Science 224, 503-505. Schumann, G., and Moroni, G. (1978).J. Immunol. 120, 1913-1916. Scofield, V. L., and Bose, H. R. (1978).J . Immunol. 120,1321-1325. Scott, G. B., Buck, B. E., Leterman, J. G., Bloom, F. L., and Parks, W. P. (1984).N . Engl. J . Med. 310, 76-81. Seidel, H. J., and Nothdurft, W. (1976).J. Reticuloendothel. SOC. 19, 173-182. Seligmann, M., Chess, L., Fahey, J. L., Fauci, A. S., Lachmann, P. J., L’Age-Stehr, J., Ngu, J., Pinching, A. J., Rosen, F. S., Spira, T. J., and Wybran, J. (1984).N . Engl.]. Med. 311,1286-1992. Senn, H.-P., and Papoian, R. (1983).Eur. J . Immunol. 13,824-830. Sharma, J. M., Herberman, R. B., Djeu, J. Y.,and Nunn, M. E. (1979).J.Zmmunol. 123, 222-231. Shaw, G. M., Harper, M. E., Hanh, B. H., Epstein, L. G., Gajdusek, D. C., Price, R. W., Navia, B. A., Petito, C. K., OHara, C. J., Groopman, J. E., Cho, E.-S., Oleske, J. M., Wong-Staal, F., and Gallo, R. C. (1985). Science 227, 177-182. Shearer, G. M., Mozes, E., Haran-Ghera, N., and Benhvich, Z. (1973).J.Immunol. 110, 736-741. Siegler, R. (1970).J . Natl. Cancer Inst. 45, 135-148. Sincovics, J. G., Gyorkey, F., Melnick, J. L., and Gyorkey, P. (1984).Reu. Infect. Dis. 6, 745-760. Smith, P. D., Ohura, K., Masur, H., Lane, H. C., Fauci, A. S., and Wahl, S. M. (1984).J. C h . Znuest. 74,2121-2128. Smith, R. E. (1982). Curr. Top. Microbiol. Imrnunol. 101, 75-94. Smith, R. E., and Ivanyi, J. (1980).J. Immunol. 125,523-530.

180

MAURO BENDINELLI ET AL.

Smith, R. E., and Schmidt, E. V. (1982).Virology 117,516518. Smith, R. E., and van Eldik, L. J. (1978).Infect. Immun. 22,452-461. Snydermann, R., and Cianciolo, G. J. (1984).Zmmunol. Toduy 5,240-244. Sodroski, J., Rosen, C., Wong-Staal, F., Salahuddin, S. Z., Popovic, M., Arya, S., Gallo, R. C., and Haseltine, W. A. (1985).Science 227, 171-173. Specter, S., and Friedman, H. (1978).P h a m c o l . Ther. A 2,595-622. Specter, S., Patel, N., and Friedman, H. (1976a). Proc. SOC. E r p . Biol. Med. 151, 163167. Specter, S., Patel, C., and Friedman, H. (1976b).J. Natl. Cancer Inst. 56, 143-147. Specter, S., Bendinelli, M.,Ceglowski, W. S., and Friedman, H. (1978).Fed. Proc. Fed. Am. SOC. Exp. Biol. 37,97-101. Specter, S., Bendinelli, M., Cox, W. I., and Friedman, H. (1983).Ado. Erp. Med. Biol. 162,497-501. Spira, T. J., Des Jarlais, D. C., Mannor, M.,Yancovitz, S., Friedman, S., Garber, J., Cohen, H., Cabradilla, C., and Kalianaraman, V. C. (1984). N. Engl. J . Med. 311, 467-468. Spire, B., Dormont, B., Bad-Sinoussi, F., Montagnier, L., and Chermann, J. C. (1985). Lancet 1, 188-189. Stanton, H. F., Law, L. W., and Ting, R. C. (1968).]. Nutl. Cancer Inst. 40,1113-1131. Starcich, B., Ratner, L., Joseph, S. F., Okanoto, T., Gallo, R. C., and Wong-Staal, F. (1985). Science 227,538-540. Stephenson, J. R., ed. (1980). “Molecular Biology of RNA Tumor Viruses,” p. 613. Academic Press, New York. Stenson, W. F., and Parker, C. W. (1980).J. Zmmunol. 125, 1-5. Stiff, M. I., and Olsen, R. G. (1982).J. Clin. Lab. Znoest. 7 , 133-138. Stiff, M.I., and Olsen, R. G. (1983).J. Gen. Virol. 64,957-959. Stromberg, K., Benveniste, R. E., Arthur, L. O., Rabin, H., Giddens, W. E., Ochs, H. D., Morton, W. R., and Tsa, C.-C. (1984).Science 224,289-292. Stutman, 0. (1975).Ado. Cancer Res. 22,261-422. Svennerholm, B., Strannegard, O., and Lycke, E. (1978). Infect. Zmmun. 20,412-417. Sy, M.-S., Tsurufuji, M.,Finberg, R., and Benacerraf, B. (1983).]. Immunol. 131,30-36. Tagliabue, A., Boraschi, D., and McCoy, J. L. (1980.J. Zmmunol. 124, 2203-2208. Tagliabue, A., Luini, W., De Vito, G., and Boraschi, D. (1981).Br. J . Cancer 44,460463.

Taniyama, T., and Holden, H. T. (1979a).Int. J. Cancer 24, 151-160. Taniyama, T., and Holden: H. T. (1979b)./. Immunol. 123,43-49. Tedder, R. S., Shanson, D. C., Jeffries, D. J., Cheinsong-Popov, R., Clapham, P., Dalgleish, A., Nagy, K.,and Weiss, R. A. (1984).Lancet 2, 125-127. Temin, H. M. (1967).J . Cell. Physiol. 69,53-63. Tilkin, A. F., Gomard, E., Begue, B., and Levy, J.-P. (1984).J . immunol. 132,520-526. Ting, C.-C., and Rodrigues, D. (1979).J . Zmmunol. 123,801-807. Todaro, G. J. (1980). In “Viral Oncology” (G. Klein, ed.), pp. 291-309. Raven, New York. Tonietti, G., Rossi, G. B., Del Gobbo, V., Accinni, L., Ranucci, A., Titti, F., Premrov, M. G., and Garaci, E. (1983).Cancer. Res. 43,4355-4363. Toniolo, A., Matteucci, D., Pistillo, M. P., Gori, Z., and Bendinelli, M. (1980).J . Gen. Virol. 49,203-208. Toniolo, A., Matteucci, D., Conaldi, P. G., and Bendinelli, M. (1984).Med. Microbiol. immunol. 173,197-206. Toy, S . T., and Wheelock, E. F. (1975).Zmmunology 17,57-79.

RETROVIRUS-INDUCED IMMUNODEFICIENCIES

181

Trainin, Z., Wernicke, D., Ungar-Waron, H., and Essex, M. (1983). Science 220, 858859. Treves, A. J., Schechter, B., Cohen, I. R., and Feldman, M. (1976).J. Immunol. 116, 1059- 1064. Tsoukas, C., Gervais, F., Shuster, J., Gold, P., OShaughnessy, M., and Robert-Guroff, M. (1984). N. Eng1.J. Med. 311, 1514-1515. Ulrich, K., and Nex& B. A. (1985).J . Virol. 53,273-278, Unanue, E. R. (1981).Adu. Immunol. 31, 1-136. Vande Woude, G. F., Levine, A. J., Topp, W. C., and Watson, J. D., eds, (1984).“Cancer Cells. Vol2. Oncogenes and Viral Genes,” p. 657. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Varesio, L. (1983). In “The Reticuloendothelial System. A Comprehensive Treatise. Vol. 5. Cancer” (R. B. Herberman and H. Friedman, eds.), pp. 217-252. Plenum, New York. Vilmer, E., Barr6-Sinoussi, F., Rouxioux, C., Gazengel, C., Vezinet-Brun, F., Dauguet, C., Fisher, R. A., Manigner, P., Chermann, J.-C., and Griscelli, C. (1984).Lancet 1, 753-757. Virgin, H. W., and Unanue, E. R. (1984)./. Immunol. 133, 104-109. Wainberg, M. A., and Israel, E. (1980).I . Immunol. 124, 64-70. Wainberg, M. A., and Margolese, R. G. (1982). Clin. Exp. Immunol. 48, 163-170. Wainberg, M. A., Beiss, B., Fong, H., Beaupr6, S., and Menezes, J. (1983a).Cancer Res. 43,1550-1557. Wainberg, M. A., Vydelingum, S., and Margolese, R. G. (1983b).J . Zmmunol. 130,23722378. Walker, M. H., Rup, B. J., Rubin, A. S.,and Bose, H. R. (1983).Infect. Zmmun. 40,225235. Wecker, E., and Horak, I. (1982). Curr. Top. Microbiol. Immunol. 98,27-36. Wedderburn, N., Carter, R. L., and Salaman, M. H. (1976).Br.J. Cancer 34,390-400. Weislow, 0. S., Fisher, 0. U., Twardzik, D. R., Hellman, A,, and Fowler, A. K. (1981). Proc. SOC. E x p . Biol. Med. 166, 522-527. Weiss, R. (1984).Nature (London)309, 12-13. Weiss, R., Teich, N., Varmus, H., and Coffin, J., eds. (1982). “Molecular Biology of Tumor Viruses: RNA Tumor Viruses,” 2nd Ed., p. 1396. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Weiss, S. H., Goedert, J. J., Sarngadharan, M. G., Bodner, A. J., Gallo, R. C., and Blattner, W. A. (1985).J. Am. Med. Assoc. 253,221-225. Wellman, M. L., Kociba, G. J., Lewis, M. G., Mathes, L. E., and Olsen, R. G. (1984). Cancer Res. 44, 1527-1529. Wood, G. W. (1976). Cancer Res. 36,4552-4557. World Health Organization (1984). Bull. W. H. 0. 62,419-432. Wybier-Franqui, J., Gomard, E., and Levy, J.-P. (1982). Cell. Zmmunol. 68,287-301. Yamanouchi, K., Yoshikawa, Y., Nayami, H., Nishiyama, H., and Boschek, C. B. (1979). Jpn.J . Med. Sci. Biol. 32, 19-28. Yefenof, E., and Ben-David, J. (1983). Cancer Immunol. Immunother. 16,48-52. Yefenof, E., and Zilcha, D. (1982). Int. J . Cancer 30,355-359. Zagury, D., Bernard, J., Leibowitch, J., Safai, B., Groopman, J. E., Feldman, M., Sarnagadharam, M. G., and Gallo, R. C. (1984). Science 226,449-451. Zatz, M. M. (1975).J , Immunol. 115,1168-1170.

This Page Intentionally Left Blank

THE MOLECULAR ACTION OF PLATELET-DERIVED GROWTH FACTOR Brent H. Cochran Center for Cancer Research end Department of Biology, Massachusetts Institute of Technology, Cambridge,Massachusetts

I. 11. 111. IV. V. VI. VII. VIII. IX. X. XI. XII.

Introduction .................................................... The Biology of Platelet-Derived Growth Factor. ...................... Biochemistry of PDGF ........................................... The SisIPDGF Gene ............................................. The Biology of the Sis Oncogene. .................................. The PDGF Receptor ............................................. Metabolic Effects ofPDGF ........................................ PDGF Modulation of the EGF Receptor ............................ Effect of PDGF on Ion Fluxes. .................................... PDGF-Stimulated Protein Phosphorylations. ......................... Regulation of Gene Expression by PDGF ........................... Conclusion ..................................................... References. .....................................................

183 184 188 192 194 195 197 200 201 202 204 209 211

I. Introduction

There are two observations about the biology of tumor cells which must be incorporated into a molecular analysis of cancer. The first is that cancer cells exhibit a reduced requirement for the cell growth factors which control the division of their normal counterparts. The second is that the transformation from a normal to a cancerous cell is a genetic change which usually involves some type of chromosomal aberration (Klein, 1981; Rowley, 1982). Recent findings that cellular oncogenes are sometimes rearranged or mutated in human tumors have greatly advanced our understanding of the genetic basis of cancer (Parada et al., 1982; Taub et al., 1982; Capon et al., 1983). However, the functions of cellular oncogenes in normal and transformed cells has until recently remained obscure. The hope of those who work on growth factors and cell proliferation has always been that an understanding of normal cell growth regulation will lead directly to an understanding of the aberrations of a cancer cell. This hope was partially realized with the discovery that 183 ADVANCES IN CANCER RESEARCH, VOL 45

Copyright 0 1985 by Academic Press, Inr All lights of reproduction in aiiv form rewrved

184

BRENT H. COCHRAN

the transforming gene of the simian sarcoma virus (v-sis)has an ancestral relationship to the structural gene encoding platelet-derived growth factor (PDGF). This finding provided the first direct link between growth factors and oncogenes. Recent results on the regulation of gene expression by PDGF further extend the linkages between growth factors and oncogenes and suggest a framework for understanding oncogenes within the context of growth factor action. This article will review the biology of PDGF with an emphasis on understanding the relationship between growth factors and cell transformation. II. The Biology of Platelet-DerivedGrowth Factor

The first indication that there was a potent mitogen in platelets came from the observations of Samuel Balk in 1971. Balk compared the mitogenic potential of clotted blood serum to that of platelet-poor plasma using primary chicken embryo fibroblasts as an indicator system. He found that cells grown in plasma did not grow nearly as well as cells grown in serum. He speculated that a mitogenic substance was released from platelets during the clotting reaction. Soon afterward Kohler and Lipton (1974) and Ross et al. (1974) confirmed these results for mouse fibroblasts and monkey arterial smooth muscle cells and extended them by showing directly that platelet extracts restored growth-promoting activity when added back to platelet-poor plasma. Subsequently, it has been shown that PDGF is a minor protein of platelet &-granules and is released during platelet adherence and aggregation reactions (D. Kaplan et al., 1979; K. Kaplan et al., 1979). Platelet-poor plasma, which is prepared by withdrawing blood in the presence of clotting inhibitors and centrifuging out the cellular material, contains only 1-2 ng/ml of PDGF as measured by radioreceptor assay (Singh et al., 1982). By contrast human clotted blood serum contains 15-20 ng/ml PDGF. PDGF has been shown to stimulate the growth of fibroblasts in culture from chickens through humans. The action spectrum of PDGF includes cells primarily of mesenchymal origin including glial cells, smooth muscle cells, and fibroblasts. There is one report that it also stimulates growth and differentiation of erythroid precursors (Dainiak et al., 1983). In uiuo, PDGF is thought to play a role in the maintenance of the vascular lining, though direct evidence is lacking. Ross and Glomset (1976) have proposed that PDGF plays a key role in the development of atherosclerotic plaques by promoting the migration of smooth mus-

THE MOLECULAR ACTION OF PDGF

185

cle cells into the intima of damaged arteries and causing them to proliferate. It is clear that PDGF can stimulate the migration of smooth muscle cells and that platelets are involved in the genesis of experimental lesions (Grotendorst et al., 1981). However, direct evidence that PDGF is essential for plaque formation is lacking, and there may well be other growth factors from platelets, endothelial cells, and macrophages involved in this complex process (Leibovitch and Ross, 1976; Gajdusek et al., 1980; Oka and Orth, 1983). Phylogenetic analysis indicates that PDGF has been functionally conserved since the first chordates. Clotted blood sera from monkeys through the cyclostomatous fish contain a connective tissue mitogen which effectively competes with human PDGF for receptor occupancy. Sera from animals below tunicates on the vertebrate line of development and all animals tested on the invertebrate line of development have no measurable PDGF-like activity (Singh et al., 1982). Thus, the appearance of PDGF activity coincides with the appearance of the pressurized vascular system and lends further credence to the idea that PDGF plays a role in vascular wound repair. An important concept in the biology of PDGF is that it functions synergistically with other growth factors (Pledger et al., 1977; Vogel et al., 1978; Stiles et al., 1979; Clemmons and Van Wyk, 1981; Bright and Gaffney, 1982). PDGF by itself is a rather poor mitogen. In combination with growth factors found in platelet-poor plasma, it acts at nanogram per milliliter concentrations to stimulate greater than 90% of cells in a quiescent, density-arrested monolayer of BALBIc-3T3 cells to divide. Concentrations of PDGF and platelet-poor plasma which, by themselves, stimulate less than 20% of a 3T3 cell population to divide, will together stimulate all of the cells to divide. Thus, together, their effects are more than additive. It should be kept in mind, however, that high concentrations of PDGF will stimulate cells to cycle even in the absence of plasma factors. Conversely, high concentrations of plasma factors can stimulate some cells to divide in the absence of PDGF. PDGF can be distinguished from the growth factors in plasma by “order of addition” experiments. The PDGF is not needed continually by fibroblasts and can exert its action in a short time pulse (Pledger et al., 1978). Quiescent BALB/c-3T3 cells that have been exposed briefly to PDGF are able to respond to platelet-poor plasma for up to 24 hr after the removal of PDGF from the medium. Cells that have been primed to respond in such a way have been termed “competent.” Such cells do not begin traversing the cell cycle until the addition of the growth factors found in platelet-poor plasma. These

186

BRENT H. COCHRAN

factors have been termed “progression” factors. Quiescent monolayers of BALBk-3T3 cells are arrested at least 12 hr from the onset of DNA synthesis (S phase). If they are treated for 3 hr with PDGF and then placed in serum-free media for another 6 hr, it will still take them 12 hr to reach the onset of S phase after the time of addition of plateletpoor plasma to the culture. This competent state is not merely due to a persistent association with PDGF to the substratum or culture medium as has been shown with radiolabeled PDGF and with antiPDGF antibodies (Singh et al., 1983).Progression factors, on the other hand, must be present continually throughout the GI phase of the cell cycle in order to exert their effect. The requirement for two sets of signals for cell growth is not limited to fibroblasts. T lymphocytes, for instance, have been shown to require concanavalin A and T cell growth factor to stimulate their growth (Larsson and Coutinho, 1979). Like PDGF, concanavalin A also works in a pulse. The functional differences between PDGF and platelet-poor plasma allow the separation of growth factors into complementation groups of competence factors and progression factors. Table I shows the categorization of several known growth factors into competence and progression functions for fibroblasts. Leof et al. (1982) have shown that epidermal growth factor and somatomedin C can substitute for platelet-poor plasma in supporting the progression of competent cells through GI. However, additional factors, such as transfemn, are usually necessary to support exponential growth of sparse cells (Bockus et al., 1983). Some factors do not neatly fit into this categorization. Infection with TABLE I COMPETENCE AND PROGRESSION FACTOFS~ Agent

Progression activity

Competence activity

+ + +

-

Somatomedin A Somatomedin C Insulin Hydrocortisone Growth EGF PDGF FGF TPA SV40 “Adapted from Scher et al. (1979).

+I-

+ -

+ +

+ + + +

THE MOLECULAR ACTION OF PDGF

187

SV40,for instance, can complement both competence and progression activities. The phorbol ester, 12-O-tetradecanoylphorbol-13-acetate (TPA), has the interesting property of complementing either competence or progression factors, but it cannot substitute for both factors at the same time (Frank et al., 1979). To fully understand cell growth, it may be useful to extend the idea of functional growth factor complementation groups beyond competence and progression. Three other categories that come to mind are substratum requirements (attachment factors or transforming growth factors), survival factors (PDGF and others), and nutritional requirements (transferrin) (Gospodarowitz et al., 1983; Scher et al., 1982; Rudland et al., 1977). The requirement of fibroblasts for multiple growth factors and their separation into complementation groups is strikingly paralleled by recent attempts to place different oncogenes into different complementation groups. It is too early in this type of analysis to state that oncogene complementation groups correspond to growth factor complementation groups, but the idea is intriguing. The formation of competence is not only relatively stable, but can actually be transferred from the cytoplasm of one cell to another, as has been shown in an elegant series of experiments by Smith and Stiles (1981). They were able to demonstrate this by fusing competent, PDGF-treated cells and cytoplasts to quiescent fibroblasts in the presence of platelet-poor plasma. The hybrids in such experiments became competent to respond to plasma. This result was not simply the action of one S-phase nucleus transactivating another one, since the fusions were done on cells in early GI of the cell cycle. Furthermore, they demonstrated that the formation of this competence signal was blocked by inhibitors of RNA synthesis. This finding indicated that PDGF modulation of RNA synthesis is necessary for the mitogenic response. The first clues that there was a relationship between growth factors and the transformed state came very early in the study of PDGF. In the initial papers showing a difference in the growth-promoting properties of serum and platelet-poor plasma, Balk et al. (1973) demonstrated that chicken cells transformed by Rous sarcoma virus grew equally as well in either medium. Kohler and Lipton (1974) showed that the same was true for mouse fibroblasts transformed by SV40. Scher et al. (1978) extended these results to cells transformed by Kirstein sarcoma virus and to Abelson murine leukemia virus. Furthermore, they demonstrated that SV40 revertants requiring high serum or which grew to normal cell density both grew in the absence of PDGF.

188

BRENT H. COCHRAN

Abortive infection of mouse 3T3 cells with SV40 stimulated entry into S phase in the absence of serum entirely. Transfection of 3T3 cells with simian sarcoma virus (SSV) DNA also confers on cells the ability to grow in plasma (Armelin et al., 1984). The mechanism by which these different transforming agents allow cells to escape the need for PDGF is just beginning to be elucidated and is discussed further below. However, it is becoming increasingly clear that one of the ways that cells do this is by producing their own growth factors. Thus, the idea of autocrine stimulation has been borne out in at least a few cases. Certainly SSV-transformed cells produce a PDGF-like growth factor (see later). Other examples are known (Heldin et al., 1980; Gudas et al., 1982). In fact, in a survey of transformed mouse cell lines, most seemed to be producing a growth factor that could compete with the PDGF receptor (Bowen-Pope et al., 1984).The mechanisms by which growth factor production is switched on in these cells are unknown. Ill. Biochemistry of PDGF

Platelet-derived growth factor is found in the a-granules of human platelets. The study of this protein has been made difficult by (1)the fact that the growth factor is present in small quantities in human serum and is only a minor protein in platelet a-granules, (2) the fact that it sticks tenaciously to plastic and glass (Smith et aZ., 1982), and (3)the limited availability of clinically outdated human platelets from blood banks. PDGF is released from the a-granules during the platelet adherence and aggregation reactions (D. Kaplan et d.,1979; K. Kaplan et al., 1979).Thus, blood that has been allowed to clot contains PDGF even after removal of cellular debris. Singh et al. (1982) have estimated by radioreceptor assay that human serum contains -15 ng/ ml of PDGF. Assuming 200,000 plateleWp1 of blood, each platelet would contain 0.075 fg of PDGF, or -1300 molecules per platelet. By contrast, other platelet a-granule proteins are present in quantities 100-1000 times greater than PDGF. PDGF was first purified to homogeneity from clinically outdated human platelets in 1979 by Antoniades et al. Subsequently, several other groups have reported the purification and characterization of the elusive growth factor (Heldin et al., 1979; Deuel et al., 1981; Raines and Ross, 1982). Purification schemes use as starting material either clinically outdated platelets or platelet-rich plasma. The former has the advantage of increased initial purity and probably yields a product that has had

THE MOLECULAR ACTION OF PDGF

189

less exposure to serum proteases, while the latter has more PDGF activity. PDGF activity is usually assayed by measuring [3H]thymidine uptake by quiescent, density-arrested mouse 3T3 cells after a 24hr incubation with the growth factor preparation plus DME containing 5% platelet-poor plasma. Several purification protocols have been published and most include some combination of cation exchange chromatography, molecular sieving by gel filtration or HPLC, and a hydrophobic affinity step. Reported yields vary from 1.5 to 20%. PDGF has a molecular weight of 28,000-35,000 and an isoelectric point of 9.8. It is stable to boiling and to a variety of denaturation agents such as SDS and urea. It is susceptible to attack by proteases and by reducing agents. Amino acid analysis indicates that PDGF has the potential to form between five and eight disulfide bonds. This would account for its heat stability and susceptibility to reduction. PDGF is a glycoprotein which may contain 7% of the mass as sugars (Deuel et al., 1981). On nonreducing SDS-polyacrylamide gels the mitogenic activity has been resolved into several bands of activity varying slightly in molecular weight. The predominate forms have been termed PDGF I and PDGF I1 (Antoniades, 1981; Deuel et al., 1981; Johnsson et al., 1982). Presumably, they are both products of the same molecule which has been subjected to different posttranslational processing or proteolysis during purification. Upon reduction of the disulfide bonds and electrophoresis through SDS-polyacrylamide gels, PDGF is resolved into a group of peptide fragments between 10,000 and 18,000 Da. The number of peptides seen after reduction is between two and five. The variation may be due to different amounts of proteolysis occurring in the purification starting material. One model of PDGF structure holds that the “native” molecule consists of two polypeptide chains in the 15,000- to 18,000-MW range-one being -2000 Da larger than the other-held tightly together by five to eight disulfide bonds (Johnsson et al., 1982). The inability to study the synthesis of PDGF under physiological conditions has made it impossible to confirm this view, but recent studies on the v-sis protein give credence to the idea. In 1983 the first partial amino acid sequences for PDGF were reported (Antoniades and Hunkapillar, 1983; Waterfield et al., 1983). Figure 1 shows the partial amino acid sequence of PDGF aligned with the predicted amino acid sequence of the v-sis gene of the SSV and its cellular counterpart (Devare et al., 1983; Doolittle et al., 1983; Josephs et al., 1984a; Chiu et al., 1984). The N-terminal sequences of reduced peptides I-V are shown. Peptides I and 11 are the longest

190

BRENT H. COCHRAN

v-sip: -esis: --

--v-sis: cs1s:

psis:

7

C-slS:

Feptde I: P e p t i d e 11:

psis: -psis: Fept

I: Peptide 11: P e p t i d e 111:

Peptide Iv: Ps is : -csls:

Fade

I:

Peptide 11: Peptide 111: Pepti& N: Peptide v:

--*

* *--_--_--

*-----

__c

* * * * *

v-sis: -es1s:

7

Peptlde I: P e p t i d e 11:

Peptide v:

v-sis:

& zG: --

Peptlde I: Peptide 11:

v-sis: -psis: --

Peptide 11:

FIG.1. Sequence comparisonof PDGF, v-sis, and C - ~ iPeptides s. I-V are fragments of PDGF obtained after reduction of the molecule. A (*) signifies a matched amino acid. A (- - - -) indicates the continuation of a PDGF fragment through a region which has not yet been sequenced. This figure was adopted from Waterfield et al. (1983)and Josephs et al. (1984a).

(MWs of 14,500 and 17,500, respectively) and show homology not only with v-sis, but with each other as well. The sequence data of the Antoniades group support a similar conclusion. Aside from peptide 11, the sequence of PDGF shows identity to v-sis in 94 of 97 positions, with the three substitutions all being conservative. Minor variations should be expected, since the v-sis gene is derived from a baboon gene while the PDGF sequence comes from human genes. The amino acid sequence predicted from the human cellular sis gene corresponds even more closely to the PDGF amino acid sequence (Josephs et al., 1984a). No homologies have yet been found between the first 66

THE MOLECULAR ACTION OF PDGF

191

amino acids of v-sis and PDGF or of the last 54. However, it can be seen from Fig. 1 that v-sis is homologous to c-sis over virtually its entire length. The similarities between PDGF and v-sis are almost certainly not fortuitous since functional and immunological studies of the two proteins confirm the sequence data. SSV-transformed cells secrete a nonglycosylated 20,000-Da protein designated p20 (Thiel et al., 1981). Cell lysates and conditioned medium from SSV-transformed cells stimulate thymidine uptake in mouse 3T3 cells in a dose-dependent fashion (Deuel et al., 1983; Owen et al., 1984). Antiserum raised against human PDGF can neutralize this growth-promoting activity. The mitogenic activity has a specific activity identical to that of PDGF as quantitated by a PDGF radioimmunoassay. Anti-PDGF antibody precipitates a 20,000-Da protein from SSV-transfonned cells-presumably p2OSis(Deuel et al., 1983). Because of the production of a PDGF-like mitogen by SSV-transformed cells, it should be possible in the near future for more investigators to have access to PDGF by purifying it from conditioned media of SSV-transformed cells. Platelets are enucleate cells and do not synthesize proteins. The platelet precursor cell, the megakaryocyte, is thought to be the site of PDGF biosynthesis (Chernoff et al., 1980). Because megakaryocytes are a minor component of the bone marrow and have not been successfully passaged in culture, it has been impossible to study the biosynthesis of PDGF. The biosynthesis of the sis gene translation product can be studied, however. Robblns et al. (1983) have used antibodies raised against predicted amino- and carboxy-terminal peptides of v-sis to investigate the biosynthesis and processing of the sis gene product by SSV-infected marmoset cells in tissue culture. Their studies indicate that v-sis is first synthesized as a 28,000-Da polypeptide which is quickly cleaved into an 11,000-MW N-terminal peptide and a 20,000-Da C-terminal peptide. Interestingly, under nonreducing conditions anti-sis-N- and C-terminal sera both recognized a 56,000-MW protein, which strongly suggests that ~ 2 8 ~forms ' " a disulfide-bonded dimer under physiological conditions. Pulse-chase experiments indicate that under nonreducing conditions, the ~ 5 6di~ ' ~ mer is further processed into forms of 46,000, 35,000, 30,000, and ' " ~24"'"were the most prominent and 24,000 MW. Of these, ~ 3 5 ~and ~24"'"was the most stable. Anti-PDGF antibody recognized all of these forms and anti-carboxy-terminal p 2 W recognized everything but ~ 2 4 However, ~ ~ . anti-amino-terminal ~ 2 precipitated 8 ~ ~ none ~ of the processed products. These results suggest that the N-terminus of ~28"'"which is not homologous to PDGF is cleaved early in process-

192

BRENT H. COCHRAN

ing and is probably not part of the mitogenically active p2OSh(reduced) structure. The fact that the sis gene product is dimeric and susceptible to proteolytic cleavage is consistent with what has been learned about the structure of human PDGF. It is of interest to note that the unreduced p35 and p28 forms of sis have molecular weights very similar to those of PDGF. Further studies on v-sis and cellular sis biosynthesis as well as the complete PDGF sequence will be needed to finally resolve the molecular structure of PDGF and the sis gene translation product. However, the striking structural and biological similarities between PDGF and the sis protein, as well as the genetic studies cited below, make it a virtual certainty that the SSV acquired the structural gene for one of the PDGF polypeptide chains during the course of its evolution. IV. The SisIPDGF Gene

The sis gene has been captured by at least two retroviruses-the SSV, as previously mentioned, and the less well-studied ParodiIrgens feline sarcoma virus (PI-FSV)(Besmer et al., 1983).The acquisition of the cellular sis oncogenes by these two retroviruses seems to have involved two independent transduction events. The SSV is derived from a Gibbon ape leukemia virus and the PI-FSV has gag and enu genes from feline leukemia virus (Gelman et aZ., 1981; Besmer et al., 1983). The cellular sis gene shows a high degree of evolutionary conservation from humans to chickens. The SSV sis gene is most closely related to the woolly monkey c-sis as shown by DNA-DNA thermal denaturation analysis (Robbins et al., 1982b). The PI-FSV sis shows restriction patterns similar to that of feline c-sis. The PI-FSV sis appears to be fused to the viral gag protein in a 76,000-Da polyprotein. As yet, it is not known whether this peptide has growth factor activity. Future studies comparing the two known viral sis proteins should provide much insight into the relationship between PDGF structure and function. Many cell lines produce proteins which have mitogenic activity and compete for the PDGF receptor (Heldin et al., 1980; Graves et aE., 1983; Gudas et al., 1982; Bowen-Pope et aZ., 1984). Not all of these molecules have the same biochemical properties. Thus, it is tempting to speculate about a family of genes which code for a related group of PDGF-like molecules. However, all the molecular hybridization evidence to date suggests that there is only a single cellular sis gene. Thus, if there are multiple proteins which compete for the PDGF

THE MOLECULAR ACTION OF PDGF

193

receptor, they are either modified posttranslationally from the sis polypetide or are only distantly related to c-sis. Sequence analysis of the c-sis cDNA suggests that this mRNA encodes only one of the PDGF polypeptide chains (Josephs et al., 1984b). Another source of heterogeneity of PDGF may come from the possibility that several genes may be able to encode the second chain. Southern blot analysis of genomic DNAs from chickens to humans under relaxed hybridization conditions shows clearly that there is only one cellular sis gene (Wong-Staal et al., 1981; Robbins et al., 1982b). There is one report of multiple allelic forms of the c-sis gene (Wong-Staal et al., 1981), but another analysis of DNA from 12 different human tissue samples by the same group revealed the presence of only one allele (Dalla-Favera et al., 1981; Chiu et al., 1984). Thus, if there are multiple sis alleles, one of them greatly predominates. These findings are consistent with the idea that the multiple forms of PDGF seen in protein purified from platelets are primarily due to posttranslational processing and proteolysis. The human genomic sis gene has been cloned and its restriction map matches that seen in Southern blots (Dalla-Favera et al., 1981). The structure of the human c-sis gene reveaIs that the regions homologous to v-sis extend over 12 kb of DNA. The regions of homology consist of at least five exons and four intervening sequences. Even this large segment, however, does not represent the entire sis transcriptional unit or even all the coding regions. Sequence analysis of the first exon with v-sis homology and its 5' flank has revealed no candidate promoters or ATG initiator codons (Josephs et d., 1983). The initiation codon lies on another exon 5' to the first exon with viral homology (Gazit et al., 1984). The sequences of the other c-sis exons show strong homology to the amino acid sequences of several of the human PDGF peptides as well as the presence of a stop codon (see Fig. 1).The coding sequence of the smaller PDGF peptide chain (peptide 11) is not present and does not appear to be present in the same mRNA (Josephs et al., 1984b). RNA transcripts, 4.2 kb in length, homologous to c-sis have been detected in certain tumor cell lines (Eva et al., 1982; Westin et al., 1982; Graves et al., 1984). This transcript is -3 kb longer than the v-sis coding region. The size of the polypeptide that the 4.2 kb sis transcript encodes is 27 kDa (Josephs et al., 198413). Thus, it appears that human PDGF could be encoded by two different mRNAs. As yet, the only nontumor cells which have been shown to express sis are endothelial cells (Barrett et al., 1984). This finding provides additional evidence that PDGF plays a role in the vascular system. It

194

BRENT H. COCHRAN

is not known whether sis is expressed in megakarocytes, the presumptive platelet precursor cell. V. The Biology of the Sis Oncogene

SSV was isolated from a multifocal spontaneous fibrosarcoma of a pet woolly monkey [see Deinhardt (1980) for review]. It is the only known isolate of this virus from a nonlaboratory animal. The virus is replication defective and requires a helper virus for reproductionsimian sarcoma-associated virus (SSAV). The helper virus is closely related to the Gibbon ape leukemia viruses. The virus can grow in a wide variety of mammalian cells in uitro and can cause fibrosarcomas and gliomas when inoculated into marmosets. SSV also confers a transformed phenotype on mammalian fibroblasts in culture (Robbins et al., 1982a). PI-FSV has a biology almost identical to that of SSV (Besmer et al., 1983). It was isolated from the fibrosarcoma of a pet cat and has a helper virus related to feline leukemia virus. It can transform cells in culture and causes fibrosarcoma when injected into kittens. Both SSV and PI-FSV have the sis-transforming sequence, but seem to have evolved independently, as mentioned previously. Their sis genes are closely related to their respective host’s c-sis and differ as well in the location of v-sis within the viral genome. It is noteworthy that the tissues which these viruses transform require PDGF for growth in uitro. This implies that autogenous stimulation by PDGF may be all that is required to generate tumors in these tissue types. The sis gene may also be involved in the generation of human tumors. In a screen of human tumor cell lines for oncogenes, Eva et al. (1982) found sis gene expression in 8 of 23 lines studied. Sis RNA was present in five of six sarcoma lines and three of five glioblastomas. Sis expression was only detected in cell types which are known to respond to PDGF. Although PDGF-like growth factor production has not yet been correlated with sis expression in any of these tumor cell lines, it is known that some osteosarcomas secrete PDGF-like growth factors (Heldin et al., 1980; Graves et al., 1983). In a search for oncogenes in human hematopoietic cells, Westin et al. (1982) found sis expression in HUT 102 T cell lymphoma cell line. This finding is exciting because this cell line is a producer of the human T cell leukemia virus (HTLV). It is the only known oncogene to be associated with HTLV. Further studies by Salahuddin et al. (1984) have shown that production of a PDGF-like growth factor is closely linked to HTLV infection of T lymphocytes in uitro. However,

THE MOLECULAR ACTION OF PDGF

195

not all HTLV-infected cells express c-sis (Clarke et al., 1984). Thus, the possible role of sis in the genesis of HTLV-induced disease remains obscure. It is unknown how the sis gene becomes activated in these human tumor cells. Chromosomal rearrangements are a possibility, but so far none has been found by Southern analysis (Dalla-Favera et al., 1981). The human c-sis gene has been assigned to chromosome 22 on region 22qll > qter (Dalla-Favera et al., 1982). Chromosomal translocations involving this region are common in chronic myelogenous leukemia (CML) and include the well-known 9;22 translocation which generates the Philadelphia chromosome (Dalla-Favera et al., 1982; Rowley, 1983). It is interesting to note that though much attention has been paid to the possible involvement of the c-abl oncogene on chromosome 9 in CML patients, translocations of chromosome 22 are more consistently associated with the disease (Rowley, 1980). However, the chromosome 22 breakpoints in CML cluster at a position relatively distant from the location of c-sis (Groffen et al., 1984). Some cases of acute myelogenous leukemias, acute lymphocytic leukemia, and variant cases of Burkitt’s lymphoma also have rearrangements of chromosome 22 (Dalla-Favera et al., 1982). Future studies of chromosome 22 rearrangements and sis gene expression will be useful. However, it should be kept in mind that the target cell in CML is a megakaryocyte precursor, and presumably the sis gene is normally activated at some stage of this differentiation series. It may well be that there are DNA rearrangements near the sis locus in normal development, though little is known on this point. Thus, any associations found between sis and leukemias should be cautiously interpreted. Furthermore, the possibility exists that other platelet growth factors, such as TGF-/3 and an EGF-like factor, may also be near the sis locus and become activated in the same developmental pathway (Oka and Orth, 1983; Assioan et aE., 1984). VI. The PDGF Receptor

Experiments using 1251-labeledPDGF have demonstrated that it binds to a specific receptor on the surface of target cells. PDGF receptors have been found on fibroblasts, smooth muscle cells, and glial cells, but not on epithelial cells, endothelial cells, or lymphocytes (Heldin et al., 1981a; Bowen-Pope and ROSS,1982; Huang et al., 1982; Williams et al., 1982; Singh et al., 1982). Thus, PDGF receptors have only been found on cells that show a mitogenic response to PDGF. The binding is specific for PDGF and in general is not competed away

196

BRENT H. COCHRAN

by other growth factors, such as epidermal growth factor (EGF), fibroblast growth factor (FGF), insulin, or transforming growth factors p or a. In general, basic proteins such as cytochrome c, lysozyme, or ribonuclease A do not compete for the PDGF receptor (Heldin et al., 1981a; Huang et al., 1982). However, polylysine, histone 2b, and protamine sulfate do compete for the receptor at 100 pg/ml concentrations (Huang et al., 1982). Protamine sulfate is the most effective of these, blocking greater than 90% of the binding at 60 pg/ml, and can also block some of the biological responses to PDGF (Huang et al., 1984). The number of receptors for PDGF on responsive cells varies from 40,000 to 400,000 per cell, with mouse 3T3 cells having the most (Bowen-Pope and Ross, 1982; Heldin et al., 1981b; Huang et al., 1982). Reported dissociation constants for the PDGF receptor vary M. This variation reflects some procedural differfrom to ences as well as the fact that the measurements were made on different cell types. It is difficult to closely correlate receptor occupancy with mitogenic activity, but the data of Singh et al. (1983) would indicate that less than 100%occupancy is needed for a maximal mitogenic response. Part of the difficulty in making these measurements lies in the fact that at 37°C PDGF is rapidly internalized and degraded. The metabolism of the PDGF receptor complex indicates that it is internalized via the classic receptor-mediated endocytosis pathway, similar to EGF (Huang et al., 1982; Bowen-Pope and Ross, 1982; Heldin et al., 1982; Singh et al., 1983; Nilsson et al., 1983; Williams et al., 1982).Minutes after binding to the the receptor, the PDGF receptor complex is found clustered in coated pits. Shortly thereafter, it is found in lysozomes and at later times can be found associated with the Golgi (Nilsson et al., 1983). Within the lysozomes PDGF is rapidly degraded and iodotyrosine is released into the culture medium. The half-life of internalized PDGF is -1 hr. Primary amines such as methylamine, chloroquine, and ammonium chloride substantially inhibit PDGF degradation. There is no definitive evidence on whether internalization and degradation are required for the mitogenic response to PDGF. As with EGF, the PDGF receptors are down-regulated shortly after exposure to ligand (Heldin et al., 1982).It is unclear whether the receptors are recycled, but the reappearance of unoccupied receptors on the cell surface is completely blocked by cycloheximide, an inhibitor of protein synthesis. The receptor for PDGF appears to be a membrane protein of -170,000 MW (Glen et al., 1982; Heldin et al., 1983). Reduction of

THE MOLECULAR ACTION OF PDGF

197

this protein with 2-mercaptoethanol does not affect its migration in SDS-polyacrylamide gels, which indicates that it is composed of a single polypeptide chain. When membrane extracts are incubated with PDGF and [32P]ATP,phosphoproteins appear-including one that is the same size as the PDGF receptor. The kinase activity and substrate both bind to a PDGF affinity column strongly, suggesting that the PDGF receptor is an autophosphorylating protein kinase. The specificity of this kinase is primarily for tyrosine residues, but some phosphoserine and threonine can be detected as well. This kinase activity is distinct from the EGF receptor kinase and from CAMPdependent protein kinase (Ek et al., 1982; Ek and Heldin, 1982; Glen et al., 1982; Heldin et al., 1983; Nishimura et al., 1982; Pike et al., 1983). PDGF and EGF can both stimulate the phosphorylation of a synthetic peptide, which is similar to the autophosphorylation site of pp6OSrcprotein. Thus, it appears that tyrosine kinases associated with growth factor receptors may have overlapping specificities with the tyrosine kinases of oncogenes. Since 6 of the 20 known oncogenes code for tyrosine kinases, this suggests that growth factor receptors and some oncogenes may function in the same way (Bishop, 1983). VII. Metabolic Effects of PDGF

The initial consequences of PDGF receptor binding can be divided into two categories, which may be designated “immediate events” and “early events.” Immediate events are those that can be monitored within 1-10 min of PDGF exposure and/or can be shown to be independent of new RNA synthesis. Early events can be monitored within 30 min to 3 hr and require transcription. Table I1 gives a summary of these events. The discussion to follow will focus on the importance of these events in the mitogenic response to PDGF and their relationship to oncogene function. Several membrane-associated effects of PDGF have been reported, some of which are interrelated and may provide clues to the mechanism of PDGF action. PDGF stimulates an increased rate of turnover of phosphatidylinositol within 5 min of addition to cells (Habenicht et al., 1981). Other membrane phospholipids are not affected. This turnover is presumably mediated by the activation of a phosphatidylinositol-specific phosphoIipase C (Schier, 1980; Habenicht et al., 1981; Shier and Durkin, 1982). At about the same time, there is a release of arachidonic acid, a transient increase in diacylglycerol levels, and an increase in E-type prostaglandin production (Habenicht et al., 1981; Shier, 1980; Shier and Durkin, 1982; Rosengurt et al., 198313). The

198

BRENT H. COCHRAN

TABLE I1 METABOLICRESPONSES TO PDGF Immediate events (1-10 min andlor transcription independent) Tyrosine-specific phosphorylations Inhibition of EGF binding Phosphorylation of ribosomal protein S6 Increased phosphatidylinositol turnover Increase in diacylglycerol Arachidonic acid and prostaglandin release Reorganization of actin filaments Stimulation of Na+-H+ exchange Increased Na+-K+ ATPase activity Release of Ca2+from intracellular stores Stimulation of polysome formation Early events (30-180min and/or transcription dependent) Acquisition of intracellular mitogenic signal Induction of specific gene transcription, including c-myc, c-fos, and r-fos Appearance of new proteins in cytoplasm Stimulation of somatomedin C binding Increase in LDL receptors Stimulation of amino acid “A” transport system Chemotaxis

prostaglandin release then stimulates a small increase in intracellular CAMP. Drugs that block prostaglandin production fail to block the mitogenic response to PDGF (Shier and Durkin, 1982). However, increases in intracellular CAMPpotentiate PDGF action by lowering the effective dose needed (Wharton et al., 1982a; Rosengurt et al., 1983b). The transient increase in diacylglycerol is interesting because it is an activator of Ca2+-dependent protein kinase C which is also activated by tumor-promoting phorbol esters such as TPA (Castagna et al., 1982). TPA, as previously mentioned, is a mitogen for 3T3 cells similar to PDGF in that it promotes competence formation. Further evidence that PDGF and TPA act on the same protein kinase comes from the observation that they both induce the phosphorylation of an 80kDa protein whose phosphorylation is stimulated by phospholipase C and not CAMP. This phosphorylation is dependent on the phorbol ester receptor and is not stimulated by insulin or EGF (Rosengurt et al., 1983a). Protein kinase C is a serinekhreonine specific kinase and is thought to be the major cellular receptor for phorbol esters. It is

THE MOLECULAR ACTION OF PDGF

199

interesting to note, however, that TPA also stimulates phosphorylations on tyrosine within minutes of addition to chicken fibroblasts (Bishop et al., 1983). Investigation of the relationship between PDGF and TPA and their effects on protein kinases and mitogenesis should be a key area of future research. Transit of cells through the GI phase of the cell cycle is very sensitive to the rate of total protein synthesis (Brooks, 1977; Rossow et al., 1979). Transition from the resting to growing state is usually accompanied by an increase in protein synthesis (Pardee et al., 1978). PDGF increases the rate of protein synthesis when added to quiescent cells (Cochran et al., 1981; Owen et al., 1982). However, this increase in protein synthesis is not by itself sufficient to initiate the mitogenic response because a similar effect is caused by components of plateletpoor plasma, This effect is independent of nuclear functions and has been demonstrated in cytoplasts (Cochran et ul., 1981). The increased rate of protein synthesis is very likely mediated through the phosphorylation of ribosomal protein S6 (a nontyrosine phosphorylation). Insulin and multiplication-stimulating factor (MSA) also stimulate S6 phosphorylation (Thomas et al., 1979; Chambard et al., 1983). PDGF also stimulates increased transport by the amino acid "A" transport system (Owen et al., 1982). The maximum effect is seen only after 2 hr and is blocked by inhibitors of protein synthesis. This effect may thus be dependent on PDGF-inducible proteins. PDGF treatment can increase the number of cellular receptors for somatomedin C and low-density lipoprotein (LDL) (Clemmons et al., 1980; Witte and Cornicelli, 1980; Witte et al., 1982). The increase in the number of somatomedin C receptors is dependent on the subsequent addition of platelet-poor plasma and may help to prime the cells for response to progression factors. In human fibroblasts, PDGF is capable of stimulating the production of a somatomedin C-like growth factor (Clemmons and Shaw, 1983). The increase in LDL receptors is probably unrelated to the mitogenic response because mutant cells lacking such receptors show the normal growth response to PDGF. This increased LDL receptor activity could have an important role in the development of atherosclerosis, however. Another major effect of PDGF is on cytoskeletal elements. Rearrangements of actin filaments are seen within minutes of addition of PDGF to cells (Westermark et al., 1983; Bockus and Stiles, 1984). Transient centriole deciliation is seen within 1 hr of PDGF addition (Tucker et al., 1979). These changes may be related to PDGF induction of cell motility and chemotaxis and/or Ca2+fluxes. The chemo-

200

BRENT H. COCHRAN

tactic response is dependent on new RNA and protein synthesis. It is also possible that cyt6skeletal alterations may mediate changes in gene expression though this is only a conjecture (Ciejek et al., 1983). PDGF treatment can temporarily give cells a morphology similar to that seen in virally transformed cells (Westermark et al., 1983).Thus, there is some basis for connecting PDGF-mediated cytoskeletal changes with those mediated by viral oncogenes.

VIII. PDGF Modulation of the EGF Receptor

PDGF modulates the receptor for EGF. Within minutes of addition of PDGF to cells, the affinity of EGF for its receptor is reduced (Bowen-Pope et d.,1983; Collins et d.,1983).At a PDGF concentration of 0.5 ng/ml, up to 90% of the EGF binding can be inhibited. The binding can be inhibited even at 4"C, though the inhibition is more extensive at 37°C. The modulation is specific and not reciprocal. PDGF does not inhibit the binding of insulin to its receptor and EGF does not affect the PDGF receptor. The modulation of the EGF receptor by PDGF is transitory and recovers by 8 hr (Wrann et al., 1980). These modulations of the EGF receptor by PDGF are strikingly similar to the effects of the tumor promoter TPA on EGF binding (Magun et al., 1980).Furthermore, both PDGF- and TPA-treated cells become more sensitive to the mitogenic effects of EGF (Wharton et al., 1983; Magun et al., 1980). This sensitization could be important in oioo, since the concentration of EGF in plasma is less than 1 ng/ml (BowenPope and Ross, 1983).The similarity of the TPA and PDGF effects on the EGF receptor suggests that this effect may be mediated through protein kinase C,which is directly activated by TPA. A possible mechanism for the effect of PDGF on the EGF receptor might be that PDGF stimulates the phosphorylation of the EGF receptor. This question has not been closely investigated, but it does seem clear from studies on isolated membranes that (1)PDGF-stimulated phosphorylation of the EGF receptor is much less than phosphorylation of its own receptor and that (2) EGF induces the phosphorylation of its own receptor more than PDGF does (Ek and Heldin, 1982; Nishimura et al,, 1982; Pike et al., 1983). Since these studies were not done on intact cells, these results must be treated with caution. In this regard, it is interesting to note that TPA stimulates the phosphorylation of the insulin and somatomedin C receptors in intact lymphocytes (Jacobs et al., 1983b).

THE MOLECULAR ACTION OF PDGF

201

IX. Effect of PDGF on Ion Fluxes

PDGF is known to have three effects on the ion fluxes of quiescent cells. It stimulates Na+-H+ exchange, Na+-K+ pump activity, and induces the release of Ca2+ from intracellular stores (Frantz et at., 1980; Frantz, 1982; Cassel et al., 1983; Moolenaar et al., 1984).PDGF, EGF, and TPA all simulate Na+-K+ pumping submaximally when used alone on quiescent cells. Combinations of growth factors that are effective in stimulating DNA synthesis elicit maximal pumping activity. The maximal increase in activity is about 2-fold and is sustained for less than 1 hr in serum-stimulated cells (Frantz, 1982). The result is a net increase in K+ influx. Blocking Na+-K+ ATPase activity with the drug ouabain does not prevent cells from becoming competent when treated with PDGF. Thus, this activity is not necessary for the mitogenic response to PDGF. PDGF promotes a 0.1-0.15 unit increase in cytoplasmic pH (Cassel et al., 1983; Schuldiner and Rozengurt, 1982).This cytoplasmic alkalization is electroneutral and blocked by amiloride-mediated inhibition of Na+ influx. The pH change begins after a lag of 2 min, is maximal by 10 min, and is sustained for at least 1 hr in the presence of PDGF alone. Serum-induced cytoplasmic alkalization is transient. EGF also induces cytoplasmic alkalization via Na+-H+ exchange (Moolenaar et al., 1983). Thus, this effect does not explain the differences in mitogenic action of these two growth factors. Na+ influx per se does not stimulate cell growth (Frantz, 1982). However, alkaline shock treatment of at least one cell type will make them competent to respond to plasma factors (Zetterberg et al., 1982). PDGF induces a net loss of intracellular Ca2+which is coupled to a release of Ca2+found in intracellular membranous storage sites. The resulting transient increase in intracellular Ca2+may be mediated by the formation of inositol triphosphate (Berridge et al., 1984). Inositol triphosphate is formed as a result of phosphatidylinositol-4,5-bisphosphate breakdown and is thought to be the second messenger for Ca2+ release. This ionic effect of PDGF is unusual in that it is specific to PDGF as EGF, TPA, insulin, and platelet-poor plasma have either no or much smaller effects on calcium fluxes. The calcium release causes a transient increase in cytoplasmic Ca2+concentration and could be responsible for centriole deciliation (Tucker et al., 1983). It is interesting to note that some virally transformed cell lines have increased Na+-K+ pumping and/or increased Na+-H+ exchange relative to normal cells (see Frantz, 1982, for review). Future studies on

202

BRENT H. COCHRAN

ion fluxes are needed to determine whether (1)they are part of the mitogenic response and (2) whether they are a result of receptor phosphorylation in growth factor treated and transformed cells or are independent events. X. PDGF-Stimulated Protein Phosphorylations

One of the most tantalizing relationships between growth factors and oncogenes is that of the tyrosine kinase activity. Six of the twenty or so viral oncogenes encode or are tightly associated with a tyrosine kinase activity (Bishop, 1983). Many of the polypeptide growth factor receptors are associated with tyrosine kinase activity including PDGF, EGF, insulin, and somatomedin C (Ushiro and Cohen, 1980; Kasuga et al., 1982; Jacobs et al., 1983a). However, little is known about the nature and function of proteins that are phosphorylated on tyrosine in response to growth factors and oncogenes. Some data are known, however, about whether growth factors and oncogenes phosphorylate the same proteins in a given cell type. The results of the most extensive study to date of this question are summarized in Table I11 (Cooper et al., 1982).As can be seen from the table, tyrosine phosphorylations caused by EGF and PDGF are more similar to each other than they are to those induced by transforming viruses. In fact, the only protein which seems to be phosphorylated by both a growth factor and a transforming virus is the 39-kDa protein which was not detectably phosphorylated in all strains of Swiss-3T3 cells used in this study. The cytoskeletal protein vinculin, which is phosphorylated in RSV-infected cells, is not phosphorylated in response to growth factors. In chick cells, PDGF, EGF, TPA, and Rous sarcoma virus all phosphorylate on tyrosine a 42-kDa protein similar to the 43-kDa proteins shown in Table I1 (Nakamura et al., 1983; Cooper et aZ., 1984). The PDGF-stimulated phosphorylations were sustained for longer times than those of EGF in these studies. The identity of the tyrosine kinase which catalyzes these phosphorylations is unknown. It may be the growth factor receptors, but this is not necessarily the case. The fact that the TPA receptor is a serinel threonine kinase raises the possibility of a kinase cascade. Consistent with this idea, Cooper et al., (1984) report that PDGF treatment can enhance the kinase activity of pp6WC.Two caveats should be kept in mind when assessing a study such as this. First, the sensitivity of detection of phosphorylations in an in viuo labeling experiment does not allow the detection of minor phosphoproteins. For instance, in

203

THE MOLECULAR ACTION OF PDGF

TABLE 111 CORRELATION OF PROTEINPHOSPHORYLATIONS INDUCED BY GROWTHFACTORS AND TRANSFORMING VIRUSES~ Protein (kDa) 29 39 42 43, pl 7.0 43, p l 7.3 45, pl6.7 45, p l 6.9 45, p l 7.3 46 81

PDGF

EGF

+/-

+ + + + +

+ +

EGF(A431)

RSV

ST-FeSV

A-MuLV

+

+ +

+ +

+ +

+ +

+ + +

+ +

+

+

This table correlates the effect of different mitogenic agents and retroviruses with the phosphorylation on tyrosine of specific proteins in mouse cells. A (+) indicates the protein is phosphorylated. A (+/-) indicates that phosphorylation was not consistently observed in all the mouse cell lines tested. EGF (A431) denotes phosphorylations observed in the A431 cell line which has an elevated number of EGF receptors, but no PDGF receptors. The viruses used are Rous sarcoma virus (RSV), Sneider-Thayer feline sarcoma virus (ST-FSV), and Abelson murine leukemia virus (A-MuLV). This table was adopted from Cooper et al. (1982).

this study the phosphorylation of the EGF and PDGF receptors was not detected. There could be many such minor proteins which were not detected. It is the minor proteins which would be likely candidates for a regulatory role. Second, these phosphorylations are presumably cell cycle dependent, at least in the case of the growth factors. Thus, it may be unfair to compare phosphorylations in quiescent, growth factor-treated cells with those in cycling tumor virus-transformed cells. Much more work will have to be done on tyrosine phosphorylations to understand their role in cell growth control and transformation. With all the excitement generated by the fact that some oncogenes have tyrosine kinase activity in common with growth factor receptors, other phosphorylations on serine and threonine residues which are stimulated by PDGF and other growth factors have not received as much attention. It should be noted, however, that in the study of Cooper et al., (1982),it was found that PDGF stimulated the phosphorylation of 11proteins on residues other than tyrosine. Furthermore, they found that every PDGF-stimulated tyrosine-phosphorylated protein also contained either phosphoserine or phosphothreonine, or

204

BRENT H. COCHRAN

both. Perhaps the residue of phosphorylation is not as critical as the protein substrate. At least one nontyrosine phosphorylation stimulated by PDGF is well characterized. PDGF along with several other growth factors stimulates the phosphorylation of ribosomal protein S6. The phosphorylation of this protein most likely has a role in the increase in protein synthesis and polysome formation seen after the addition of growth to quiescent cells (Chambard et al., 1983).Phosphorylation of S6 seems to be CAMPdependent and dependent on the regulation of intracellular pH via the Na+-H+ exchange, as mentioned above. It is also clear that S6 phosphorylation is not sufficient to initiate the mitogenic response of growth factor-treated quiescent cells (Chambard et al., 1983; Thomas et al., 1979; Cochran et al., 1981). The role of other nontyrosine phosphorylations in PDGF-treated cells is not known. Before leaving the subject of phosphorylations, it should be noted that phorbol esters and PDGF both stimulate the phosphorylation of an 80-kDa protein in 3T3 cells. This is further evidence that these two growth factors may share a common mechanism of action, at least in part (Rozengurt et al., 1983a). Though it is reasonable to suppose that growth factor action is mediated through protein phosphorylations, it is important to remember that there is no direct evidence on this point. Genetic studies on cloned receptor genes will be of great importance in resolving the role of kinase activity in growth factor-mediated mitogenesis. XI. Regulation of Gene Expression by PDGF

PDGF regulates not only the cytoplasmic effects previously noted, but also gene expression in the nucleus. Furthermore, Smith and Stiles (1981) have shown that the acquisition of the stable mitogenic signal in response to PDGF is dependent on de novo RNA synthesis. Specific regulation of RNA and protein synthesis has been found in response to PDGF. Pledger et al. (1981) have described the appearance of five new proteins in response to PDGF treatment of quiescent, density-arrested BALB/c-3T3 cells. Their molecular weights were 29,35,45,60, and 70 kDa. The induction of the 29-, 35-, and 70-kDa proteins was blocked by inhibitors of RNA synthesis. Synthesis of the other two proteins is presumably controlled by a translational mechanism. Insulin and EGF had little effect on the synthesis of these proteins, but FGF, another competence factor, induced the 29- and 35-kDa proteins. The regulation of these proteins is fairly specific, as only about

THE MOLECULAR ACTION OF PDGF

205

1% of the spots seen in a two-dimensional protein gel change upon addition of PDGF. The 35-kDa protein has been identified by antisera to a major excreted protein family (MEP) of retrovirus-transformed cells (Scher et al., 1983).This is a major protein representing -0.5% of the total protein in PDGF-treated cells and almost 2% of the total nonnuclear protein in a spontaneously transformed 3T3 cell line. Five cell lines that were either spontaneously or chemically transformed constitutively synthesized this 35-kDa MEP. TPA also induced the synthesis of this MEP protein family. Though MEP secretion and constitutive production is correlated with cell transformation, no function has been shown for them. The 29-kDa PDGF-inducible protein is associated with the cell nucleus. Its synthesis in cycling cells is confined to the first half of GI and the protein is stable for 9-12 hr (Olashaw and Pledger, 1983). Its function is unknown, but it has been proposed as a mediator of competence. In an attempt to further understand gene regulation by PDGF, we have recently isolated cDNA clones of PDGF-inducible gene sequences (Cochran et al., 1983).These clones were isolated by differential colony hybridization screening of a cDNA library made from poly(A)+RNA from PDGF-treated cells. Out of an 8000 clone library, five independently inducible gene sequences were found. Two of these genes, termed]E and KC, illustrate the general pattern of gene regulation by PDGF. The KC gene has a 1200 nucleotide mRNA and encodes a 10,000-Da polypeptide, as shown by in vitro translation. The J E gene has a 990-nt mRNA and encodes a 19,000-Da polypeptide. A 10-to 20-fold induction of these genes can be detected within 1 hr of PDGF treatment, and the induction ofJE is 60-fold with a maximum PDGF dose. At maximum induction, they are moderately abundant RNAs.]E is found in -3000 copies/cell and is 0.9% of the total poly(A)+RNA, and KC is represented in about 700 copies/cell and is 0.2% of the total poly(A)+ RNA. The induction of these genes is blocked by inhibitors of RNA synthesis, but not by inhibitors of protein synthesis. Thus, their induction is a primary response to PDGF and not a consequence of cell growth. We have recently shown by in vitro transcription assays that they are regulated at the transcriptional level (unpublished results). Inhibitors of protein synthesis actually slightly induce these genes and superinduce them when used in combination with PDGF. The effect is particularly striking for KC, as its induction is 20-fold stronger in the absence of protein synthesis. This gene may be subject to feedback inhibition of transcription.

206

BRENT H. COCHRAN

This gene regulation is specific to PDGF. EGF and FGF only slightly induceJE and KC, and insulin and platelet-poor plasma not at all. TPA gives an intermediate induction (unpublished results). A wide variety of non-growth factor drugs and proteins do not induce these genes. The induction ofJE and KC is transient and their levels have fallen to near baseline by the time cells begin S phase. As yet, no functions have been assigned to these genes, and it is not known how PDGF mediates their induction. However, all of the genes induced by PDGF as yet described show the same pattern of regulation as do theJE and KC genes. They are induced at the transcriptional level in a transient manner by competence factors. Growth factors with progression activity have little effect on these genes. Inhibitors of protein synthesis tend to superinduce their expression. The most direct demonstration of a functional link between two oncogenes has been the observation that PDGF induces the accumulation of c-myc mRNA (Kelly et al., 1983). PDGF induces c-myc 40fold within 3 hr of addition to quiescent BALBk-3T3 cells. Two other competence factors, FGF and TPA, induce the synthesis of c-myc mRNA, but not as strongly as PDGF. EGF, insulin, and platelet-poor plasma have less than a 2-fold effect in a 3-hr treatment. The levels of c-myc in the induced state are far lower than those of the other PDGFinducible mRNAs JE and KC. C-myc is present at -5-10 copies/cell when fully induced. The induction of c-myc is interesting not only because it is a known oncogene, but also because it is induced in response to mitogen stimulation of other cell types. Treatment of T lymphocytes with concanavalin A or B lymphocytes with lipopolysaccharide (LPS) both induce c-myc (Kelly et al., 1983). These two mitogens are similar to PDGF in that they both are first signals in a multisignal pathway. C-myc is also induced early in regenerating liver (Makino et al., 1984). Thus, c-myc induction is correlated with the early response to mitogens in a variety of cell types. This observation, along with the results discussed below, gives credence to the idea that the normal role of c-myc is to regulate the proliferative state of the cell. It has been hypothesized that c-myc is regulated by a labile repressor protein (Leder et al., 1983). The finding that cycloheximide slightly induces c-myc in lymphocytes and mouse fibroblasts and superinduces in the presence of mitogens is consistent with this idea. This behavior is also consistent with that seen for theJE and KC genes induced by PDGF, suggesting that these genes may be subject to similar regulation.

THE MOLECULAR ACTION OF PDGF

207

To fully appreciate the significance of the regulation of c-myc by PDGF, it is necessary to understand something about the biology of the myc gene. Myc is the transforming gene of several avian RNA tumor viruses. These viruses induce an unusually large array of neoplasms, including sarcomas, carcinomas, and leukemias. The nonacute transforming virus, avian leukosis virus, appears to function by inserting its long terminal repeat (LTR) enhancer sequence into the genome at regions near the cellular myc gene, thus activating its transcription (Hayward et al., 1981). Further interest in the myc gene has come from findings that the myc gene is involved in rearrangements with the immunoglobulin locus in human Burkitt’s lymphoma and in mouse plasmacytomas (Taub et al., 1982).Amplifications of the c-myc gene have been found in a human promyelocytic leukemia and a human colon carcinoma (Alitalo et al., 1983; Collins and Groudine, 1982). Most of the known rearrangements of c-myc in tumor cells appear to involve normal regulatory sequences rather than mutations within the colding regions of the gene. The model of c-myc regulation by a labile repressor protein accounts for these observations by proposing that any disruption of the repressor binding site will lead to constitutive production of c-myc. Thus, the finding that mitogens induce c-myc in normal quiescent cells in a transient manner suggests that constitutive production of c-myc might lead to constitutive cell growth and give the cell independence of normal growth regulatory signals. Consistent with this hypothesis is that in some cell systems vmyc transformation leads to enhanced cellular proliferation (Palmieri et al., 1983). Armelin et al. (1984) have assessed the growth properties of BALB/ c-3T3 which have been transfected with the mouse c-myc gene under the control of exogenous promoters. The conclusion of their experiments is that the expression of c-myc is responsible for part of the mitogenic response to PDGF. By putting the c-myc under control of the mouse mammary tumor virus LTR, it is possible to induce the expression of c-myc by adding hydrocortisone to the culture medium. The addition of hydrocortisone to confluent quiescent BALBk-3T3 cells containing these myc constructs increased the proportion of cells entering S phase over a 24-hr period from 0.5 to 28% in the absence of PDGF. While this is a substantial mitogenic response, it is not as strong as the response to PDGF usually seen. Another indication that myc forms a component of the mitogenic response to PDGF is that constitutive expression of c-myc confers on cells a hypersensitivity to EGF. Density-arrested BALB/c-3T3 cells show almost no mitogenic response to EGF. Constitutive expression

208

BRENT H. COCHRAN

of c-myc allows cells to respond in a strong mitogenic fashion to subnanogram/milliliter concentrations of EGF. Thus, c-myc expression alone enhances the cellular growth response to EGF, just as PDGF treatment does. A third criterion by which myc-expressing cells show a degree of independence from the PDGF requirement is by the ability of myc transfectants to grow clonally in medium lacking PDGF. Myc-transfected BALBh3T3 cells will form colonies on fibronectin-coated tissue culture dishes in medium containing platelet-poor plasma 60% efficiently as cells plated in PDGF-containing medium. This is substantially better than the less than 10%efficiency of colony formation of normal BALBb3T3. However, it is not as good as the greater than 90% colony formation efficiency obtained with cells that have been transfected with the v-sis gene in PDGF-free medium. Thus, the general conclusion of these transfection experiments is that myc forms a component of the mitogenic response to PDGF, but is probably not the only gene or event involved. A further and more direct demonstration that c-myc mediates a part of the competence promoting activity of PDGF has been the microinjection experiments of Kaczmarek et al. (1985).They have shown that microinjection of the myc protein into quiescent BALBb3T3 cells makes them competent to respond to the growth factors in plateletpoor plasma. This effect is seen even in the presence of anti-PDGF antibody. A puzzling aspect of myc biology has been that despite the obvious involvement of c-myc with a variety of cancers, it is unable to induce focus formation in fibroblasts or render them able to grow in soft agar. However, the ability to induce cell growth and the ability to grow in soft agar are independent functions. PDGF does not cause cells to grow in soft agar and transforming growth factors do not induce cells to proliferate in the absence of other growth factors. These two functions may have their analogs in independent oncogenes. Thus, it is not surprising to find that not every oncogene induces anchorage-independent growth or functions entirely in the absence of other oncogenes. This notion is entirely consistent with the finding of Land et al. (1983) that myc functions in a synergistic way with ras and T antigen in order to fully transform rat embryo fibroblasts. The finding of the BZym gene in Burkitt’s lymphoma cell lines (Diamond et al., 1983) is a further indication that oncogenes, like growth factors, function synergistically in distinct complementation groups. It seems likely that some of the other genes induced by PDGF may also have a role to play in the mitogenic response. Two strong candidates for such a role are the PDGF-inducible genes related to thefos

THE MOLECULAR ACTION OF PDGF

209

oncogene. Sequence analysis of theJB gene isolated from the cDNA library has demonstrated that this gene bears a strong homology to the third exon of the cellularfos oncogene (Cochran et al., 1984). In this exon there is a 12 of 14 amino acid homology stretch and a 51 of 59 nucleotide match. The homology is great enough to permit cross-hybridization at low stringency. The sequence homology is not nearly as strong as the predicted match between c-sis and PDGF, but is comparable to that of N-mycand c-myc (Schwab et al., 1983).The sequences of the relatedfos gene (r-fos) and c-fos diverge greatly to either side of the third c-fos exon. In addition to the r-fos gene, it has recently been shown by several investigators that the c-fos gene itself is inducible by PDGF (Cochran et al., 1984; Greenberg and Ziff, 1984; Kruijer et al., 1984; Muller et al., 1984).The c-fos gene shows a pattern of regulation similar to that of the other PDGF-inducible genes. The induction is rapid, is not blocked by inhibitors of protein synthesis, and is superinduced under these conditions. The induction is transient and peaks early in GI. Other competence factors such as FGF and TPA induce the gene, whereas progression factors have little or no effect. The induction can be seen at the transcriptional level within 15 min of PDGF addition (Greenberg and Ziff, 1984). There is evidence that c-fos induction occurs 15-30 min prior to c-myc induction, but the significance of this is unclear. It seems doubtful that thefos gene would activate the myc gene, since both are induced in the absence of protein synthesis. In addition, it appears that fos expression is shut off more rapidly than some of the other PDGF-inducible genes. Thefos protein product has disappeared by 3 hr of PDGF treatment (Kruijer et al., 1984; Muller et al., 1984). XII. Conclusion

Through the study of the molecular biology of PDGF, the role of oncogenes in cell transformation and their relationship to growth factors is beginning to be understood. This relationship is illustrated schematically in Fig. 2. It appears that the deregulation of molecular events which are normally controlled by growth factors can lead to ceII transformation. Several known oncogenes can be associated with specific molecular events regulated by growth factors. Some oncogenes, such as the sis gene, may code for a growth factor which is capable of stimulating cell growth through an autocrine-type mechanism. Other oncogenes may code for functions similar to events inside the cell that are regulated by growth factors. It seems quite likely that some oncogenes will either turn out to be growth factor

2 10

BRENT H. COCHRAN

ONCOGENES IN SERIES

--.

4 :

FIG.2. Oncogenes in series.

receptors or mimic the functions of them. This may be a likely explanation for the coincidence that a large number of oncogenes as well as growth factor receptors are tyrosine kinases. The recent finding by Downward et al. (1984) that the erb-B gene is closely related to the EGF receptor supports this idea. A third connection between oncogenes and growth factors is that genes which are regulated by growth factors and are important in the response to them may be oncogenes. The regulation of c-myc and cfos by PDGF and lymphocyte mitogens is a strong example of this. It is possible that other genes regulated by PDGF or other mitogens also have roles in cell transformation. Several key questions remain about the mechanism of growth factor action and its involvement in cancer. How does PDGF transmit a signal from a receptor on the cell surface to the nucleus in order to regulate a few specific genes? It seems likely that PDGF-induced phosphorylations are involved, but there is no direct evidence to support this idea. Other mechanisms can be envisioned. Do oncogenes other than sislPDGF exert their action by regulating gene expression? SV40 large T antigen has been shown to regulate gene expression (Schutzbank et al., 1982), but is this a general phenomenon for oncogenes? It is also tempting to speculate that growth factor expression is regulated by oncogenes. The large number of tumor cells which secrete growth factors give credence to this idea (Delarco and Todaro, 1978; Nickel1 et al., 1983; Ozanne et al., 1982; Tarsio et al., 1983). But

THE MOLECULAR ACTION OF PDGF

211

just how important is autocrine stimulation of cell growth in tumorigenesis? The simian sarcoma virus appears to work in such a manner, but is this indicative of a widespread cancer mechanism? How does the myc gene function? Is it solely responsible for making cells “competent”? Are other genes involved? Questions such as these will be important in future research on both growth factors and oncogenes.

REFERENCES Alitalo, K., Schwab, M., Lin, C. C., Varmus, H. E., and Bishop, J. M. (1983).Proc. Natl. Acad. Sci. U . S A . 80, 1707-1711. Antoniades, H. N. (1981). Proc. Natl. Acad. Sci. U.S.A. 78,7314-7317. Antoniades, H. N., and Hunkapiller, M. W. (1983).Science 220, 963-965. Antoniades, H. N., Scher, C. D., and Stiles, C. D. (1979).Proc. Natl. Acad. Sci. U S A . 76, 1809-1813. Armelin, H. A., Armelin, M.C.S., Kelly, K., Stewart, T., Leder, P., Cochran, B. H., and Stiles, C. D. (1984).Nature (London)310,655-660. Assoian, R. K., Frolik, C. A., Roberts, A. B., Miller, D. M., and Sporn, M. B. (1984).Cell 36,35-41. Balk, S . D. (1971).Proc. Natl. Acad. Sci. U.SA. 68,271-275. Balk, S., Whitfield, J. F., Youdale, T., and Braun, A. C. (1973).Proc. Natl. Acad. Sci. U S A . 70,675-679. Barrett, T . B., Gajdusek, C. M., Schwartz, S. M., McDougall, J. K., and Benditt, E. P. (1984). P ~ o c Natl. . Acad. Sci. U . S A . 81,6772-6774. Berridge, M. J., Heslop, J. P., Irvine, R. F., and Brown K. D. (1984).Biochem. J . 222, 195-201. Besmer, P., Snyder, H. W., Murphy, J. E., Hardy, W. D., and Parodi, A. (1983).J.Virol. 46,606-613. Bishop, J. M. (1983).Annu. Rev. Biochem. 52,301-354. Bishop, R., Martinez, R., Nakamura, K. D., and Weber, M. J. (1983).Biochem. Biophys. Res. Commun. 115,536-543. Bockus, B. J., and Stiles, C. D. (1984).Exp. Cell Res. 153, 186-197. Bockus, B. J., Scher, C. D., and Stiles, C. D. (1983).In “Myelofibrosisand the Biology of Connective Tissue” (P. D. Berk, ed.). Thieme, Stuttgart. Bowen-Pope, D. F., and Ross, R. (1982).J.B i d . Chem. 257,5161-5171. Bowen-Pope, D. F., Dicorleto, P. E., and Ross, R. (1983).J . Cell Biol. 96,679-683. Bowen-Pope, D. F., Vogel, A., and Ross, R. (1984). Proc. Natl. Acad. Sci. U . S A . 81, 2396-2400. Bright, M. D., and GaEney, E. V. (1982).E x p . Cell Res. 137,309-316. Brooks, R. F. (1977).Cell 12, 311-317. Capon, D. J., Chen, E. Y.,Levinson, A. D., Seeburg, P. H., and Goeddel, D. V. (1983). Nature (London) 302,33-37. Cassel, D., Rothenberg, P., Zhuang, Y. X., Deuel, T. F., and Glaser, L. (1983).Proc. Natl. Acad. Sci. U S A . 80,6224-6228. Castagna, M., Takai, Y.,Kaibuchi, K., Sano, K., Kikkawa, U., and Nishizuka, Y. (1982).J. Biol. Chem. 257,7847-7851. Chambard, J. C., Franchi, A., LeCam, A., and Pouyssegur, J. (1983).J . Biol. Chem. 258, 1706-1713. Chernoff, A., Levine, R. F., and Goodman, D. S. (1980).J . Clin. Inuest. 65, 926-930. Chiu, I.-M., Reddy, E. P., Givol. D., Robbins, K. C., Tronick, S. R., and Aaronson, S. A. (1984).Cell 37, 123-129.

212

BRENT H. COCHRAN

Ciejek, E. M., Tsai, M. J., and O’Malley, B. W. (1983).Nature (London)306,607-609. Clarke, M. F., Westin, E., Schmidt, D., Josephs, S. F., Ratner, L., Wong-Staal, F., Gallo, R. C., and Reitz, M. S., Jr. (1984).Nature (London) 308,464-467. Clemmons, D. R., and Van Wyk, J. J. (1981).J. Cell Physiol. 106,361-367. Clemmons, D. R., and Shaw, D. S. (1983).J. Cell Physiol. 115, 137-142. Clemmons, D. R., Van Wyk, J. J., and Pledger, W. J. (1980).Proc. Natl. Acad. Sci. U S A . 77,6644-6648. Cochran, B. H., Lillquist, J. S., and Stiles, C. D. (1981).J . Cell Physiol. 109,429-438. Cochran, B. H., Reffel, A. C., and Stiles, C. D. (1983).Cell 33,939-947. Cochran, B. H., Zullo, J., Verma, I. M., and Stiles, C. D. (1984).Science 226,1080-1082. Cohen, S., Carpenter, G., and King, L. J. (1980).J . B i d . Chem. 255,4834-4842. Collins, S., and Groudine, M. (1982).Nature (London)298,679-681. Collins, M. K. L., Sinnett-Smith, J. W., and Rozengurt, E. (1983).J. B i d . Chem. 258, 11689-11693. Cooper, J. A., Bowen-Pope, D. F., Raines, E., Ross, R., and Hunter, T. (1982).Cell 31, 263-273. Cooper, J. A., Seflon, B. M., and Hunter, T.(1984).Mol. Cell. B i d . 4,30-37. Crossin, K. L., and Carney, D. H. (1981).Cell 23,61-71. Dainiak, N., Davies, G., Kalmanti, M., Lawler, J., and Kulkami, V. (1983).J . Clin. Inuest. 71, 1206-1214. Dalla-Favera, R., Gelmann, E. P., Gallo, R. C., and Wong-Staal, F. (1981). Nature (London) 292,31-35. Dalla-Favera, R., Gallo, R. C., Giallongo, A., and Croce, C. M. (1982).Science 218,686688. Deinhardt, F. (1980).In “Viral Oncology” (G. Klein, ed.). Raven, New York. DeLarco, J. E., and Todaro, G. J. (1978).Nature (London)272,356-358. Deuel, T.,Huang, J. S., Proffitt, R. T., Baenziger, J. U., Chang, D., and Kennedy, B. B. (1981).J. B i d . Chem. 256,8896-8899. Deuel, T.F., Huang, J. S., Huang, S. S., Stroobant, P., and Waterfield, M. D. (1983). Science 221,1348-1350. Devare, S. G., Reddy, E. P., Law, J. D., Robbins, K. C., and Aaronson, S. A. (1983).Proc. Natl. Acad. Sci. U S A . 80,731-735. Diamond, A., Cooper, G. A., Ritz, J., and Lane, M. A. (1983).Nature (London)305,112116. Doolittle, R. F., Hunkapiller, M. W., Hood, L. E., Devare, S. G., Robbins, K. C., Aaronson, S. A., and Antoniades, H. N. (1983). Science 221,275-276. Downward, J., Yarden, Y., Mayes, E., Scrace, G., Totty, N., Stockwell, P., Ullrich, A., Schlessinger, J., and Waterfield, M. D. (1984).Nature (London) 307, 521-527. Ek, B., and Heldin, C. H. (1982).J. Biol. Chem. 257, 10486-10492. Ek, B., Westermark, B., Wasteson, A., and Heldin, C. H. (1982).Nature (London) 295, 419-420. Eva, A., Robbins, K. C., and Andersen, P. R., Srinivasan,A., Tronick, S. R., Reddu, E. P., Ellmore, N. W., Galen, A. T., Lautenberger, J. A., Papas, T. S., Westin, E. H., WongStaal, F., Gallo, R. C., and Aaronson, S. A. (1982).Nature (London)295,116119. Franb, C. N. (1982).In “Genetic Expression in the Cell Cycle” (G. M. Padilla and K. S. McCarthy, Sr., eds.), pp. 411-445. Academic Press, New York. Frank, C. N., Stiles, C. D., and Scher, C. D. (1979).J. Cell. Physiol. 100,413-424. Frank, C. N., Stiles, C. D., Pledger, W. J., and Scher, C. D. (1980).J . Cell Physiol. 105, 439-449. Gajdusek, C., DiCorleto, P., Ross, R., and Schwartz, S. M. (1980).J. Cell Biol. 85,467472.

THE MOLECULAR ACTION OF PDGF

213

Gazit, A., Igarashi, H., Chiu, I., Srinivasan, A., Yaniv, A., Tronick, S. R., Robbins, K. C., and Aaronson, S. A. (1984).Cell 39,89-97. Gelman, E. P., Wong-Staal, F., Kramer, A., and Gallo, R. C. (1981).Proc. Natl. Acad. Sci.

U S A . 78,3373-3376.

Glenn, K., Bowen-Pope, D. F., and Ross, R. (1982).J . B i d . Chem. 257,5172-5176. Gospodarowicz, D., Gonzalez, R., and Fujii, D. K. (1983).J. Cell. Physiol. 114,191-202. Graves, D. T.,Owen, A. J., and Antoniades, H. N. (1983).Cancer Res. 43,83-87. Graves, D.T.,Owen, A. J., Barth, R. K., Tempst, P., Winoto, A., Fors, L., and Hood, L. E. (1984).Science 226,972-974. Greenberg, M. E., and Ziff, E. B. (1984).Nature (London)311,433-438. Groffen, J., Stephenson, J. R., Heisterkamp, N., deKlein, A., Bartram, C. R., and Grosveld, G. (1984).Cell 36,93-99. Grotendorst, G. R., Seppa, H. E. J., Kleinman, H. K., and Martin, G. R. (1981).Proc. Natl. Acad. Sci. U . S A . 78,3669-3672. Gudas, L.J., Singh, J. P., and Stiles, C. D. (1982).Cold Spring Harbor ConJ Cell ProliJ

9.

Habenicht, A. J. R., Glomset, J. A., King, W. C., Nist, C., Mitchell, C. D., and Ross, R. (1981).J. Biol. Chem. 256, 12329-12335. Hayward, W.S., Neel, B. G., and Astrin, S. M. (1981).Nature (London)290,475-480. Heldin, C. H., Westermark, B., and Wasteson, A. (1979).Proc. Natl. Acad. Sci. U S A . 76,

3722-3726.

Heldin, C. H., Westermark, B., and Wasteson, A. (1980).J. Cell. Physiol. 105,235-246. Heldin, C. H., Westermark, B., and Wasteson, A. (1981a).Proc. Natl. Acad. Sci. U S A .

78,3664-3668.

Heldin, C. H., Westermark, B., and Wasteson, A. (1981b).Biochem. J . 193,907-913. Heldin, C.H., Wasteson, A., and Westermark, B. (1982).J. Biol. Chem. 257,4216-4221. Heldin, C . H., Ek, B., and Ronnstrand, L. (1983).J. Biol. Chem. 258, 10054-10061. Huang, J. S., Huang, S. S., Kennedy, B., Deuel, T. F. (1982).J.Biol. Chem. 257,8130-

8136.

Huang, J. S., Nishimura, J., Huang, S. S., and Deuel, T. F. (1984).]. Cell. Biochem. 25,

205-220.

Hunter, T., and Sefion, B. M. (1980).Proc. Natl. Acad. Sci. U S A . 77, 1311-1315. Jacobs, S., Kull, F. C. J., Earp, H. S., Svoboda, M. E., Van Wyk, J. J., and Cuatrecasas, P. (1983a).J . BioZ. Chem. 258,9581-9584. Jacobs, S., Sahyoun, N. E., Saltiel, A. R., and Cuatrecasas, P. (1983).Proc. Natl. Acad. Sci. U S A . 80,6211-6213. Johnsson, A., Heldin, C. H., Westermark, B., and Wasteson, A. (1982).Biophys. Res. Commun. 104,66-74. Josephs, S . F., Favera, R. D., Gelmann, E. P., Gallo, R. C., and Wong-Staal, F. (1983). Science 219,503-505. Josephs, S. F., Guo, C., Ratner, L., and Wong-Staal, F. (1984a).Science 223,487-490. Josephs, S. F., Ratner, L., Clarke, M. F., Westin, E. H., Reik, M. S., and Wong-Staal, F. (1984b). Science 225,636-639. Kaczmarek, L., Hyland J. K., Watt, R., Rosenberg, M., and Baserga, R. (1985).Science

228,1313-1315.

Kaplan, D. R., Chao, F. C., Stiles, C. D., Antoniades, H. N., and Scher, C. D. (1979). Blood 53, 1043-1052. Kaplan, K. L.,Broekman, J. M.,Chernoff, A., Lesznick, G. R., and Drillings, M. (1979). Blood 53,604-615. Kasuga, M., Zick, Y., Blithe, D. L., Crettaz, M., and Kahn, C. R. (1982).Nature (London)

298,667-669.

214

BRENT H. COCHRAN

Kelly, K., Cochran, B. H., Stiles, C. D., and Leder, P. (1983).Cell 35,603-610. Klein, G. (1981).Nature (London)294,313-318. Kohler, N., and Lipton, A. (1974).Exp. Cell Res. 87,297-301. Kozbor, D., and Croce, C. M. (1984).Cancer Res. 44,438-441. Kruijer, W., Cooper, J. A., Hunter, T., and Verma, I. M. (1984).Nature (London) 312, 711-716. Land, H., Parada, L. F.,and Weinberg, R. A. (1983).Nature (London) 304,596606. Larsson, E. L., and Coutinho, A. (1979).Nature (London) 280,239-241. Leder, P., Battey, J., Lenoir, G., Moulding, C., Murphy, W., Potter, H., Stewart, T., and Taub, R. (1983). Science 222, 765-771. Leibovich, S. J., and Ross, R. (1976).Am. J . Pathol. 84,501-513. Leof, E. B., Wharton, W., Van Wyk, J. J., and Pledger, W. J. (1982).E x p . Cell. Res. 141, 107-1 15. Magun, B. E., Matrisian, L. M., and Bowden, G. T. (1980).J. B i d . Chem. 255,63736381. Makino, R., Hayashi, K., and Sigamura, T. (1984).Nature (London) 310,697-700. Moolenaar, W. H., Mummery, C. L., van der Saag, P. T., and de Last, S . W. (1981).Cell 23,789-798. Moolenaar, W. H., Tsien, R. Y.,van der Saag, P. T., de Laat, S. W. (1983). Nature (London)304,645-648. Moolenaar, W. H., Tertoolen, L. G. J., and de Laat, S. W. (1984).J.B i d . Chem. 13,80668069. Muller, R., Bravo, R., Burckhardt, J., and Lurran, T. (1984).Nature (London) 312,716720. Nakamura, K. D., Martinez, R., and Weber, M: J. (1983).Mol. Cell Biol. 3,380-390. Nickell, K. A., Halper, J., and Moses, H. L. (1983).Cancer Res. 43, 1966-1971. Nilsson, J., Thyberg, J., Heldin, C. H., Westermark, B., and Wasteson, A. (1983).Proc. Natl. Acad. Sci. U S A . 80,5592-5596. Nishimura, J., Huang, J. S., and Deuel, T. F. (1982).Proc. Natl. Acad. Sci. U S A . 79, 4303-4307. Oka, Y., and Orth, D. N. (1983).J. Clin. Inuest. 72, 249-259. Olashaw, N. E., and Pledger, W. J. (1983).Nature (London) 306,272-274. Owen, A. J., Geyer, R. P., and Antoniades, H. N. (1982).Proc. Natl. Acad. Sci. U S A . 79, 3203-3207. Owen, A. J., Pantazis, P., and Antoniades, H. N. (1984).Science 225, 54-56. Ozanne, B., Wheeler, T.,and Kaplan, P. L. (1982).Fed. Proc. Fed. Am. SOC.E x p . Biol. 41,3004-3007. Pdmieri, S., Kahn, P., and Graf, T. (1983). EMBO J . 2,2385-2389. Parada, L. F.,Tabin, C. J., Shih, C., and Weinberg, R. A. (1982).Nature (London) 297, 474-478. Pardee, A. B., Dubrow, R., Hamlin, J. L., and Kletzien, R. F. (1978).Annu. Reu. Biochem. 47,715-750. Pike, L. J., Bowen-Pope, D. F., Ross, R., and Krebs, E. G. (1983).J . Biol. Chem. 258, 9383-9390. Pledger, W. J., Stiles, C. D., Antoniades, H. N., and Scher, C. D. (1977). Proc. Natl. Acad. Sci. U S A . 74,4481-4485. Pledger, W. J., Stiles, C. D., Antoniades, H. N., and Scher, C. D. (1978). Proc. Natl. Acad. Sci. U S A . 75,2839-2843. Pledger, W. J., Hart, C. A., Locatell, K. L., and Scher, C. D. (1981).Proc. Natl. Acad. Sci. U.SA. 78,4358-4362.

THE MOLECULAR ACTION OF PDGF

2 15

Raines, E. W., and Ross, R. (1982).J.Biol. Chem. 257,5154-5160. Robbins, K. C., Devare, S. G., Reddy, E. P., and Aaronson, S. A. (1982a).Science 218, 1131-1133. Robbins, K. C., Hill, R. L., and Aaronson, S. A. (1982b).J. Virol. 41, 721-725. Robbins, K. C., Antoniades, H. N., Devare, S. G., Hunkapiller, M. W., and Aaronson, S. A. (1983). Nature (London) 305,605-608. Ross, R., and Glomset, J. A. (1976).N . Engl. J . Med. 295, 639-377. Ross, R., Glomset, J. A., Kariya, B., and Harker, L. (1974).Proc. Natl. Acad. Sci. U S A . 71,1207-1210. Rossow, P., Riddle, V. G. H., Pardee, A. B. (1979). Proc. Natl. Acad. Sci. U S A . 76, 4446-4450. Rowley, J. D. (1980).Annu. Rew. Genet. 14, 17-39. Rowley, J. D. (1982). Science 216, 749-751. Rowley, J. D. (1983). Nature (London)301,290-291. Rozengurt, E., Legg, A., and Pettican, P. (1979).Proc. Natl. Acad. Sci. U.S.A.76,12841287. Rozengurt, E., Rodriguez-Pena, M., and Smith, K. A. (1983a).Proc. Natl. Acad. Sci. U S A . 80,7244-7248. Rozengurt, E., Stroobant, P., Waterfield, M. D., Deuel, T. F., and Keehan, M. (1983b). Cell 34,265-272. Rudland, P. S., Durbin, H., Clingan, D., and de Asua, L. J. (1977).Biochem. Biophys. Res. Commun. 75,556-561. Salahuddin, S . Z., Markham, P. D., Lindner, S. G., Gootenberg, J., Popovic, M., Hemmi, H., Sarin, P. S., and Gallo, R. C. (1984). Science 223, 703-707. Scher, C. D., Pledger, W. J., Martin, P., Antoniades, H. N., and Stiles, C. D. (1978).J. Cell. Physiol. 97, 371-380. Scher, C. D., Shepard, R. C., Antoniades, H. N., and Stiles, C. D. (1979). Biochim. Biophys. Acta 560,217-241. Scher, C. D., Young, S. A., and Locatell, K. L. (1982).J.Cell. Physiol. 113,211-218. Scher, C. D., Dick, R. L., Whipple, A. P., and Locatell, K. L. (1983).Mol. Cell Biol. 3, 70-8 1. Schuldiner, S., and Rozengurt, E. (1982). Proc. Natl. Acad. Sci. U.S.A. 79,7778-7782. Schutzbank, T., Robinson, R., Oren, M., and Levine, A. J. (1982).Cell 30,481-490. Schwab, M., Alitalo, K., Klempnauer, K. H., Varmus, H. E., Bishop, J. M., Gilbert, F., Brodeur, G., Goldstein, M., and Trent, J. (1983).Nature (London)305, 245-248. Shier, W. T. (1980). Proc. Natl. Acad. Sci. U S A . 77, 137-141. Shier, W. T., and Durkin, J. P. (1982).J. Cell. Physiol. 112, 171-181. Singh, J. P., Chaikin, M. A., and Stiles, C. D. (1982).J. Cell BioE. 95,667-671. Singh, J. P., Chaikin, M. A., PIedger, W. J., Scher, C. D., and Stiles, C. D. (1983).J . Cell. Biol. 96, 1497-1502. Smith, J. B., and Rozengurt, E. (1978).Proc. Natl. Acad. Sci. U S A . 75,5560-5564. Smith, J. C., and Stiles, C. D. (1981). Proc. Natl. Acad. Sci. U . S A . 78,4363-4367. Smith, J. C., Singh, J. P., Lillquist, J. S., Goon, D. S., and Stiles, C. D. (1982).Nature (London)296,154-156. Stiles, C. D., Capone, G. T., Scher, C. D., Antoniades, H. N., Van Wyk, J. J., and Pledger, W. J. (1979).Proc. Natl. Acad. Sci. U.S.A. 76, 1279-1283. Tarsio, J. F., Rubin, N. A., Russel, P., Gregerson, D. S., and Reid, T. W. (1983).Exp. Cell Res. 146, 71-78. Taub, R., Kirsch, I., Morton, C., Lenoir, G., Swan, D., Tronick, S., Aaronson, S., and Leder, P. (1982). Proc. Natl. Acad. Sci. U S A . 79, 7837-7841.

216

BRENT H. COCHRAN

Thiel, H. J., Matthews, T. J., Weinhold, K. J., and Broughton, E.M. (1981).Virology 115, 401-405. Thomas, G., Siegman, M., and Gordon, J. (1979).Proc. Natl. Acad. Sci. U S A . 76,39523956. T h o , G. J., DeLarco, J. E., Marquardt, H., Bryant, M. L., Sherwin, S. A., and Sliski, A. H. (1979).Cold Spring Harbor ConJ Cell Prolif. 6,113-131. Tucker, R. W., Scher, C. D., and Stiles, C. D. (1979).CeZE 18, 1065-1072. Tucker, R. W., Snowdowne, K. W., and Borle, A. B. (1983).J . Cell Biol. 97,343. Ushiro, H., and Cohen, S. (1980).J . Biol. Chern. 255,8363-$365. Vogel, A., Raines, E., Kariya, B., Rivest, M. J., and Ross, R. (1978).Proc. Natl. Acud. Sci. U S A . 75,2810-2814. Waterfield, M. D., Scrace, G. T., Whittle, N., Stroobant, P., Jobnsson, A., Wasteson, A., Westennark, B., Heldin, C. H., Huang, J. S., and Deuel, T. F. (1983).Nature (London) 304,35-39. Westermark, B., Heldin, C. H.,Ek, B., Johnsson, A., Mellstrom, K., Nister, M., and Wasteson, A. (1983).In “Growth and Maturation Factors” (G. Guroff, ed.), pp. 75114. Wiley, New York. Westin, E. H., Wong-Staal, F., Gelmann, E. P., Favera, R. D , Papas, T. S., Lautenberger, J. A., Eva, A., Reddy, E.P., Tronick, S. R., Aaronso ,S. A,, and Gallo, R. C. (1982). Proc. Natl. Acad. Sci. U S A . 79,2490-2494. Wharton, W., Leof, E.B., Olashaw, N., Earp, H. S., and Pledgel, W. J. (1982a).]. Cell. Physiol. 111,201-206. Wharton, W., Leof, E.,Pledger, W. J., and OKeefe, E.J. (1982b1. Proc. Natl. Acad. Scd. U S A . 79,5567-5570. Wharton, W., Leof, E., Olashaw, N., O’Keefe, E. J., and Pledger, W. J. (1983).Exp. Cell Res. 147,443-448. Williams, L. T., Tremble, P., and Antoniades, H. N. (1982).Procj Natl. Acad. Sci. U S A . 79,5867-5871. I Witte, L. D., and Cornicelli, J. A. (1980). Proc. Natl. Acad. Sci./U.SA.77,5962-5966. Witte, L. D., Cornicelli, J. A., Miller, R. W., and Goodman, D. S. (1982).1.Biol. Chem. 257,5392-5401. Wong-Staal, F., Dalla-Favera, R., Franchini, G., Gelmann, E.P., and Gallo, R. C. (1981). Science 213,226-228. Wrann, M., Fox, C. F.,and Ross, R. (1980). Science 210, 1363-1365. Zetterberg, A., Engstrom. W., and Larsson, 0. (1982).Ann. N.Y. Acad. Sci. 397, 130147.

4

TRICHOTHECENES. ZEARALENONE. AND OTHER CARCINOGENIC METABOLITES OF FUSARlUM AND RELATED MICROFUNGI R . Schoental Department of Pathology. The Royal Veterinary College. University of London. London. England

. .

I Introduction ................................................... I1 Secondary Metabolites of Fusaria ................................. I11 Epidemiological Considerations .................................. A Alimentary Tract Tumors ..................................... B Geochemistry. Selenium. and Cancer ........................... C . Selenium and Esophageal Cancer in South Africa . . . . . . . . . . . . . . . . . D . Esophageal Cancer in China and Elsewhere ..................... IV. Occurrence and Pathological Effects of T-2 Toxin . . . . . . . . . . . . . . . . . . . V Diacetoxyscirpenol (Anguidine. NSC-141537) ....................... VI The Antileukemic Baccharinoids and Other Macrocyclic Trichothecenes A Baccharinoids ............................................... B. Miotoxins .................................................. C. Satratoxins.................................................. VII Metabolism of T-2 Toxin ........................................ VIII Effects of T-2 Toxin and Related Trichothecenes on the Immune System................................................ IX Detection and Estimation of Trichothecenes ........................ X Chemistry and Biological Activity of Zearalenone . . . . . . . . . . . . . . . . . . . XI Teratogenic Effects of Zearalenone and Bone Lesions . . . . . . . . . . . . . . . . XI1. Metabolism of Zearalenone ...................................... XI11 Estrogenic Agents and Zeranol ................................... A . Anabolic Action of Estrogenic Agents ........................... B . Metabolism of Zeranol ....................................... XIV. Estrogenic Agents and the Development of Sex Organ Abnormalities and Tumors ....................................... A . In Human Beings ............................................ B . In Experimental Animals ..................................... XV Carcinogenic Effects of Zearalenone ............................... A. In F344N Rats and B6C3F1 Mice .............................. B. In Wistar Rats ............................................... XVI. Occurrence of Zearalenone and Distribution of Mycotoxins . . . . . . . . . . . A. Natural Occurrence of Zearalenone in Agricultural Products . . . . . . . . B. Distribution of Mycotoxins in Milling Fractions of Cereals . . . . . . . . . XVII Methods of Detection and Estimation of Zearalenone and Its Estrogenic Derivatives ....................................... XVIII . Attempts at Detoxication of Fusarial Mycotoxins ..................... XIX. Conclusions ................................................... References ....................................................

.

. .

. .

.

.

.

. . . .

.

.

218 220 222 223 225 227 228 231 234 236 236 237 239 240 243 244 245 248 249 251 253 252 253 253 256 257 257 258

268 268 269 270 271 272 274

217 ADVANCES IN CANCER RESEARCH. VOL. 45

Copyright B 1985 by Academic Press. Inc. All rights of reproduction in any form reserved .

218

R. SCHOENTAL

I. Introduction

Significant advances in any scientific discipline often come about by the application of ideas based on new observations in other fields or by serendipitous findings in the course of experimentation. This applies also to advances in cancer research. A breakthrough in the understanding of the etiology of certain tumors followed the recognition that moldy feeds, which cause losses in animal husbandry from reproductive and other health disorders, contain relatively simple chemical substances produced as secondary metabolites under specific conditions by the toxigenic strains of microfungi. Many of the toxic or otherwise deleterious secondary metabolites, known as mycotoxins, when isolated in pure form have been found to reproduce experimentally similar disorders in both livestock and laboratory animals. Some of the mycotoxins induced tumors in rodents and in certain other animals when tested appropriately. The interest in the economically and otherwise important mycotoxins has been rapidly growing during the last quarter of a century and is illustrated by the proliferation of publications and books dealing with the natural distribution, production, isolation, and chemistry of many of these interesting substances and with their biochemical, biological, and pathological effects (e.g., Kadis et al., 1971; Turner, 1971; Purchase, 1974; Christensen, 1975; Rodricks, 1976; Rodricks et al., 1977; Wyllie and Morehouse, 1977, 1978; Moreau, 1979; Uraguchi and Yamazaki, 1978; Schoental and Connors, 1981; Shank, 1981; Turner and Aldridge, 1983; Ueno, 1983a; IARC, 1983a,b; Rechcigl, 1983; Searle, 1984). Many synthetic substances, some potentially carcinogenic, have been introduced into the environment during the last 50 years. Cancer in earlier times must have been caused by carcinogenic agents present in the “natural” environment, such as ultraviolet and ionizing radiation, radioactive minerals and certain microelements, asbestos, certain alkaloids and other secondary metabolites of plants and microorganisms, as well as viruses. Such agents continue to be health risks to animals and man and might have been involved in the etiologies of cancers which are still considered as “spontaneous.” Some of these factors are beyond the possibility of control, but elimination of even a single agent would be of practical value in reducing the incidence of cancers by delaying their appearance beyond the lifespan of the individual. Identification of sources of carcinogens is therefore of primary importance with the view to eliminate or reduce human exposure to those that are tumorigenic.

CARCINOGENIC METABOLITES OF FUSARIUM

219

The latent period between the exposure to a carcinogen and the time of the appearance of recognizable tumors can be very long. Under conditions of “optimal” dosage in experimental animals, it corresponds to more than a quarter of their respective life-spans. The latent period for most human cancers is estimated to be 20-30 years. This explains the difficulties in tracing the agents responsible for the tumors in a cancer patient. In this review, the term “tumor” will be used in preference or alternately to “cancer.” The distinction between various degress of neoplasias is often not sharp or impossible. It is becoming increasingly evident that nutrition does not deal with idealized situations in which food would consist of pure proteins, carbohydrates, lipids, salts, and vitamins, but that, in reality, food contains, in addition, a variety of impurities, some of which, although present in minute trace quantities, may have striking effects on the metabolism, the cellular integrity, and the survival of the individual. Mycotoxins (the secondary metabolites of fungi) including those produced by Fusal-ia, the common field microfungi, are among the “naturally” occurring food contaminants, which often confound epidemiological and experimental evidence on the role of nutritional factors in relation to human and animal health. The studies of mycotoxins as potential carcinogenic agents are still in their infancy, except for the aflatoxins, which for various reasons including economic and political (and also because of their striking fluorescence that makes them readily identifiable) have been a favorite subject during the last 20 years or so. The aflatoxins (B1, GI, and MI) proved to be effective liver carcinogens in rats and in some other experimental animal species and may be involved in cancers of this organ in man, though definite evidence on the extent to which aflatoxins contribute among other hepatotoxic agents to human liver disorders and tumors has still to be obtained (Busby and Wogan, 1984). Extrapolation of data obtained experimentally in certain animals to other animal species (or man) is not always justified in view of differences in the enzymatic endowment of different species and variations in their ability to metabolize, activate, and detoxify xenobiotics. Epidemiological evidence of exposure to some specific environmental factors and its correlation with the incidence of particular types of cancer in man is more directly relevant and, under specific conditions (e.g., massive exposure in industry), gave significant leads, which resulted in the recognition that radiation, polycyclic aromatic hydrocarbons, aromatic amines, asbestos, and polychlorinated hydrocarbons are carcinogenic hazards for man (Hiatt et al., 1977).

220

R. SCHOENTAL

However, carcinogenic agents in food to which large numbers of the general population may be exposed are more difficult to identify, and this is illustrated by the fact that, till the present, the etiologies of the most common cancers have remained unrecognized. In this review, epidemiological and experimental evidence is presented which suggests that some secondary metabolites of Fusarium microfungi deserve serious consideration as playing a part in the etiologies of human cancers of the sex organs and of the digestive tract and of the brain. As some carcinogenic agents can have also acute, subacute, and teratogenic action when their dosage is excessive or when acting during the perinatal period or on the very susceptible young, the identification of the causes of such more acute disorders also can give useful leads to cancer. Many of the known “natural” carcinogens have been identified as a result of their ill effects on livestock, when economic pressures stimulated detailed investigations, as in the case of the pyrrolizidine alkaloid-containing plants (Schoental, 1976c),bracken (Dobereiner et al., 1967; Tokarnia and Dobereiner, 1975; Evans, 1976; Hirono et at., 1984) and the mold products, aflatoxins (Busby and Wogan, 1984),T-2 toxin (Bamburg et al., 1968), zearalenone (Mirocha et al., 1968; Christensen et al., 1972),and others (Shank, 1981; Martin, 1981; Schoental, 1981a,b,e, 1984e). The relevance of the past observations made on livestock disorders to the etiology of human diseases and cancer point to the need to pay more attention to similar comparative studies in the future. II. Secondary Metabolites of Fusaria

The Fusariurn microfungi are a group of widely distributed soil microorganisms. Many species are phytotoxic (Marasas et al., 1971; Joffe and Palti, 1974; Palti, 1978) and can cause great losses of various agricultural and horticultural crops. Their taxonomy is complicated and controversial (Nelson et al., 1981; Booths, 1971). Among the hsarial secondary metabolites are the cytotoxic trichothecenes, biosynthesized by the terpene route (Tamm and Breitenstein, 1980), and the estrogenic zearalenone, biosynthesized by the polyketide route via the acetate-malonate-CoA pathway (Steele et al., 1974; Vleggaar and Steyn, 1980). Secondary metabolism usually becomes significantwhen the conditions for the basal cellular functioning are not favorable or when the proportion of nutrients is not conducive to the biosynthesis of normal

CARCINOGENIC METABOLITES OF FUSARIUM

22 1

cell constituents needed for life and growth. Under such conditions, fungi are able to go into “suspended animation,” they form spores, and some species can transform the surplus substrate materials into secondary metabolites, possibly as internal reserves. The conditions that determine the production of deleterious secondary metabolites by Fusaria have yet to be more fully elucidated. Under apparently similar culture conditions, a particular strain or isolate can unpredictably produce greatly variable yields of the toxic metabolites. Micronutrients may be involved, especially those having pronounced catalytic action, e.g., selenium. Certain epidemiological data seem to point to the importance of geochemical factors in the incidence of certain disorders, including tumors in humans and animals. Fusarial secondary metabolites are sporadically present in livestock feeds and have been recognized as the cause of a variety of disorders, “ mycotoxicoses,” some of which are fatal. A recent example was reported from Israel where imported barley that caused diarrhea and reduced milk yield in dairy cows was found to contain T-2 toxin (Shlosberg et al., 1984). Among livestock that survived such mycotoxicoses, long-term sequelae have not been recorded (Miessner and Schoop, 1929; Petrie et al., 1977; Schoental, 1981e). However, the lack of publications should not be confused with the lack of carcinogenic potentialities of these compounds for livestock. Tumors of the sex organs and of the alimentary tract are often encountered in veterinary oncology (Cotchin, 1977, 1984). The possibility that commercial laboratory animal diets may sporadically be contaminated with fusarial mycotoxins and may confound the evaluation of carcinogenicity tests also has to be considered (Schoental, 1974b, 1979a-c). Usually moldy grain is not used for human consumption. However, under exceptional circumstances due to harvest failure or war, people may be forced to consume moldy grain to alleviate famine. Pellagra that used to occur among the poor in the Mediterranean countries and elsewhere, though considered an “avitaminosis,” was related to the consumption of corn (Zea mays) that became moldy either in the course of transport or in storage and was mainly fusarial mycotoxicosis (Schoental, 1980e). The sensitivity of pellagrins to sunshine is related to the disturbances in porphyrin metabolism. When rats are given T-2 toxin, they excrete increased concentrations of urinary coproporphyrin (Schoental and Gibbard, 1978). It would be of interest to know the incidence and the types of tumors that occur among the “cured” pellagrins.

222

R. SCHOENTAL

111. Epidemiological Considerations

The appearance of cancer in humans may take 20-30 years from the time of exposure to a causative agent. The most recent epidemiological data collected in the 1980s would include the effects due to the various synthetics, such as pesticides, food additives, and oral contraceptives, some of which are known to be carcinogenic. These synthetics came into widespread use since the 1950s after the last world war. For the evaluation of “natural” carcinogens, more relevant are earlier data (e.g., on cancer mortality in 24 countries for the year 1964-1965) (reported by Segi et al., 1969). Cancers, which caused death in the 196Os, are likely to have been initiated before World War I1 and would have been due mainly to “natural” carcinogenic agents, endogenous and/or exogenous. Carcinogenic mycotoxins, including those produced by Fusarium species, may have contributed to their etiol-

om. Segi’s data indicate that, in countries situated in the temperate zones with high rainfall and frequent damp and cool weather, the predominant neoplasias were those of the gastrointestinal tract (which represented about one-third of all the cancers in men and in women) and, in women, tumors of the breast and sex organs, which represented another one-third of all cancers (Fig. 1). It seems significant that the data for mortality from breast cancers in women in most of the western countries were very similar, but were about six times higher than among Japanese women. Yet, among the Japanese immigrants in the United States the incidence of tumors of the sex organs and of the gastrointestinal tract becomes within one or

FIG.1. Mortality ratio from cancer of the sex organs (about 1 :3) and of the gastrointestinal tract (about 1 :3) to all the cancers among women in England and Wales.

CARCINOGENIC METABOLITES OF FUSAZUUM

223

two generations similar to that prevailing in the United States (Haenszel and Kurihara, 1968; Wynder and Hirayama, 1977). Environmental factors evidently are mainly responsible for the differences in the incidence of such tumors. Recently, neoplasms of the alimentary tract, of the breast, and of certain other organs have been correlated with diet, especially diets high in fats and low in fiber with low intake of fruits and vegetables and with high intake of alcoholic drinks (Bjelke, 1974; Proc. Maribou Symp., 1979). Alcoholic beverages and high fat diets have also been correlated with cardiovascular disorders, though controversies remain unresolved as to whether the deleterious effects are due to animal or vegetable fats, to cholesterol, to peroxides, or to something else. The particular substances in the various items of diet which have the deleterious effects have remained mainly unidentified. Fusarial mycotoxins which are lipophilic compounds may sporadically contaminate fats (Schoental, 1981c), as well as certain alcoholic beverages, notably beers (Schoental, 1980a, 1981b; Haikara, 1983; Payen et al., 1983). Supplementation of alcoholic drinks with thiamine, advocated by Centerwall and Criqui (1978), might indeed counteract some effects of mycotoxins. The effects of fiber (Burkitt, 1982) proved to be not invariably beneficial. When different preparations of bran and of other forms of fiber were tested experimentally, the results were variable and inconclusive (Kelsey, 1978; Asp et al., 1979; Jacobs, 1983). This is not surprising for mycotoxins, when present in cereals, usually concentrate in the outer parts of the grain.

A. ALIMENTARYTRACT TUMORS Though dietary factors are suspected to be involved in the etiology of tumors of the alimentary tract, the actual agents have not yet been identified. Many natural and synthetic substances that occasionally contaminate food can induce tumors of the digestive tract in experimental animals, including certain nitroso compounds and mycotoxins, especially the irritant trichothecene, T-2 toxin (Schoental, 1981b). It is of particular interest that the incidence of cancer of the various parts of the gastrointestinal tract (such as the buccal cavity, esophagus, stomach, duodenum, small and large intestine, colon, and rectum) varies greatly in different geographical areas and also can vary at different times in the same area. Thus, the incidence of stomach cancer in the United States has been decreasing in recent years, while that of col-

224

R. SCHOENTAL

orectal tumors has been increasing. Similar trends also appear elsewhere. The most striking regional differences, up to more than 100-fold, have been reported in the incidence of cancer of the esophagus. In the high-risk areas, esophageal cancer can account for about 20% of all cancers (Warwick and Harington, 1973; Cook-Mozaffari, 1979; Kinlen, 1981; Pfeiffer, 1982). Pockets of very high incidence of esophageal cancer occur in Iran (the northeastern parts of the Caspian littoral, the Gonbad and Gorgiin districts) (Kmet and Machboubi, 1972), in the USSR, especially Turkmen Republic, Kazakhstan, and Uzbekistan (Serenko and Romenski, 1970; Napalkov et d.,1982),in the Henan province of China (Lin and Tang, 1980; Li et al., 1980), in Kenya (the northeastern shores of Lake Victoria), and in Central Kenya. Significantly high incidence of ruminal cancer has also been reported among Masai cattle, which graze in the Nasampolai valley of the Narok district (Ahmed and Cook, 1969; Plowright et aZ., 1971)and in South Africa (Bantustan and Transkei) (Burrell, 1962; Burrell et al., 1966; Oettle, 1964). Less striking, elevated incidence of esophageal cancer occurs in some other geographical areas, e.g., in Curacao [possibly related to the use of certain plants (Dunham et aZ., 1974)], in Japan [among those who eat bracken (Hirayama, 1979)],in Puerto Rico (Martinez, 1969),in the United States [among blacks (Fraumeni and Blot, 1977)], in northern France (Tuyns, 1970,1979), in northern Sweden (Larsson et al., 1973, in Finland (Appelqvist, 1972), in Trieste (Giarelli et al., 1980), and elsewhere. Usually, the incidence in men is higher than in women, yet in some areas the incidence of esophageal cancer in women exceeds that in men [e.g., in Iran (Kmet and Mahboubi, 1972) and in Alaska (Hurst, 1964)l. More than a single factor is likely to be responsible for tumors that occur in animals and man. “In oesophageal cancer, we have an example of a cancer that is surely due to environmental factors, but is just as surely not caused everywhere by the same carcinogen” (Berg, 1977). The factors that have epidemiologically been correlated with high risk of esophageal cancer in specific areas include alcoholic drinks, the custom of drinking very hot tea, the ingestion of residues from smoked opium containing polycyclic compounds (Hewer et al., 1978),trauma, e.g., caused by crystals of silica in millet flour (O’Neil et al., 1980), etc. As for the alcoholic drinks, direct evidence has recently been obtained that sporadically T-2 toxin is present in small concentration in some samples of beer (Payen et aZ., 1983) and that zearalenone, when present in barley at a low level, increased during

CARCINOGENIC METABOLITES OF FUSARZUM

225

malting up to 50-fold (Haikara, 1983), supporting the misgivings that, during periods of particularly severe weather conditions, mycotoxins could signficantly contribute to the ill effects attributed to alcoholic beverages (Schoental, 1980a, 1981b). B. GEOCHEMISTRY, SELENIUM, AND CANCER Alcoholic drinks, however, could not be of significance among the Muslim Turkman population of Iran in the Gonbad and Gorgiin districts of the Mazandaran province, where women have the highest recorded incidence of esophageal cancer (women 174 per 100,000, and men 109 per 100,000). b e t and Mahboubi (1972) stress that the soil in this area consists of alluvial sediment which requires irrigation and yields poor crops of barley, wheat, and cotton. The population subsists mainly on bread, milk, and tea. The nutritional status is very low, especially among women who experience very many pregnancies. The “natural” vegetation of this area consists mainly of the halophilic Artemisia and Astragalus plants, certain species of which are known to be “indicators” of seleniferous soils and selenium accumulators (Rosenfeld and Beath, 1964; Kingsbury, 1964). Astragalus plants are being used in China (and elsewhere) as herbal medicines (Dr. Li, personal communication, 1983). Seleniferous soil can impart toxic properties to plants grown in such areas. Selenium may become incorporated into the proteins, replacing sulfur in the respective amino acids, or into other sulphur compounds, transforming these into toxic antimetabolites. Rosenfeld and Beath (1964) indicate that the beasts of burden of Marco Polo and of other travelers in Turkmenistan and Western China became ill and lost hoofs due to eating the local selenium-rich plants, while locally bred animals did not suffer ill effects. This has been considered the earliest report on selenium toxicity. In the republics of the USSR, Turkmen Republic, Kazakhstan, and Uzbekistan, adjacent to Iran and the Caspian sea, the incidence of esophageal cancer is also very high (Serenko and Romanski, 1970; Napalkov et al., 1983).It is noteworthy that, in these republics of the USSR (as in certain other parts of the Russian grain belt), outbreaks used to occur of disorders known as “alimentary toxic aleukia” ( A M ) (Nesterov, 1948; Mayer, 1953; Joffe, 1960, 1978). According to Mayer (1953), “The disease (ATA) is reported in several republics and sections along the Volga River, in the Kirov and Molotov Regions, in Uljanovsk, in the Udmurt, in Kujbishev, the Tartar and the Bashkir Republics, in the Saratov and Chkalov Districts, in Kazakhstan, in the

226

R. SCHOENTAL

Sverdlovsk and Cheljabinsk sections of the Ural Mountains, also in the neighbourhood of the Siberian cities of Tobolsk, Omsk, Tomsk, Novosibirsk and at some parts of the Central Asian District, as far south as Tashkent in Uzbekistan.” This syndrome [known also as “septic angina” (necrotizing tonsillitis), “panmyelotoxicoses,” and by a variety of other names] followed the consumption of bread made from flour derived from overwintered, moldy grain contaminated by the toxigenic species of Fusarium poae and Fusarium sporotrichioides and of other microfungi and by their mycotoxins. ATA was particularly severe in undernourished people. Its effects were comparable to those following radiation. The outbreaks usually began in April or May and ended in July. ATA consisted of several stages. A period of 3-9 days of acute gastrointestinal tract disorder and of irritation of the mucous membranes of the oral cavity was followed by a period of about 0.5-2 months of apparent improvement, during which leukopenia, thrombocytopenia, and hemorrhagic rash may have become apparent. In severe cases, the third stage followed consisting of acute, often fatal infections of the throat, the lungs, necrotizing pharyngitis, and hemorrhagic diathesis. Recovery was slow even in the less serious cases (Nesterov, 1948; Mayer, 1953; Joffe, 1978). The mycotoxin mainly responsible for ATA has been identified as the trichothecene, T-2 toxin, which can reproduce many features of ATA in experimental animals (Joffe, 1978; Yagen et al., 1977; Lutsky et al., 1978). Rats given crude extracts from cultures of the species of Fusaria which were incriminated in the 1940s outbreak of ATA in the Orenburg District showed necrotizing lesions in the esophagus (Schoental and Joffe, 1974). Among rats given several intragastric (ig) doses of pure T-2 toxin, acute lesions of the esophagus were not seen, but among the rats that survived more than a year after the first dose, some developed tumors in various parts of the gastrointestinal tract (GT),the pancreas, and the brain (Schoental et al., 1979). The question, why tumors localize in particular areas of the GI tract cannot as yet be satisfactorily answered. The pH varies greatly along the GI tract, and so do various enzyme systems (Zedeck, 1980; Soullier et al., 1981), some of which activate, while others detoxify certain intermediate metabolites of the administered parent carcinogen. The degree of the stomach’s fullness, the particulars of the diet, the availability of specific vitamins, coenzymes, micronutrients, etc., all play some role. The interrelationships between all these factors and the administered substance(s) determine the end effect.

CARCINOGENIC METABOLITES OF FUSARZUM

227

One can only speculate why it is that, in the USSR, where outbreaks of ATA are known to have occurred at various times in the grain belt with black and loamy soils (extending from the European parts to Eastern Siberia and Birobidzhan), there appears to be, to some extent, an inverse relationship between the incidence of cancer of the stomach and that of the esophagus (Napalkov et al., 1982).Could the different levels of soil selenium explain this situation (Sindeeva, 1964)? General trends of the cancer rates per 100,000 population of the whole USSR from 1965 to 1980 have shown a decrease of stomach cancer (about 25%), a decrease of esophageal cancer (more than 25%), but an increase, almost doubling, of cancer of the rectum (in both males and females) and almost doubling of breast cancer in women (from 13.7 to 22.2 per 100,000). This is reminiscent of the trends that occurred also in the United States and elsewhere. However, for the identification of the etiological factors that contributed to the development of cancer in any individual, overall data are not helpful. Case control studies that would match exactly the date and the place of birth, the exact location, the diet, the style of life, the use of medicines (herbal or synthetic), etc. not only of the patient, but also of the parents, may give more significant leads. It is not yet known whether individuals who recovered from ATA in the USSR developed, as late sequelae of this syndrome, tumors of the esophagus or of the digestive tract, and also whether trichothecene mycotoxins may have been involved in such tumors in the USSR and elsewhere (Schoental, 1979c, 1980c, 1983~).

C. SELENIUM AND ESOPHAGEAL CANCER IN SOUTHAFRICA In South Africa, the Beaufort and Dwyka series of the geological Karoo system of sedimentary rocks have been reported to be seleniferous (Brown and de Wet, 1962). The selenium content of wheat samples from South Africa was reported to be higher than from various other parts of the world (Robinson, 1936). It may be significant that high levels of selenium (10-30 mgkg) have been found in the livers of poultry affected by aflatoxin mycotoxicoses in South Africa (Abrams, 1966). Moreover, toxigenic Fusarium species (especially F. moniliforme) have been isolated more frequently from maize grown in areas of high incidence of esophageal cancer in the Transkei than from maize grown in other areas (Marasas et al., 1979,1984; Marasas, 1982). The incidence of cancer of the esophagus in Transkei increased apparently in an epidemic-like fashion since the 1940s and attained its peak about 1955. Burrell (1962) considered that this increase was re-

228

R. SCHOENTAL

lated to “the disastrous period during 1931-1933; after 2 years long of the worst drought in recorded history, 1933 brought gale-force winds. The sky was thick with dust. Torrential rains followed. The Beaufort sedimental soil, because of leaching, could no longer sustain healthy plant growth.” Such severe weather conditions are conducive to outbreaks of mycotoxicoses. The spores of soil microfungi are dispersed by the wind. When rains come, microfungi grow and produce their mycotoxins on the decomposing plant debris. Exposure of the population of Transkei to trichothecene mycotoxins in 1933 may have contributed to the tumors which became apparent in the course of the subsequent two decades or so. Abundance of fusarial mycotoxins which are phytotoxic could also cause poor crops, while a low nutritional status would sensitize the people to the action of mycotoxins. A high rate of esophageal cancer has recently been reported also among the Zulu men in Natal, South Africa. The main factors which have been considered as contributing to this type of neoplasma are bought maize [relative risk (RR):5.731 and commercial cigarette smoking (RR: 2.64) (van Rensburg et al., 1985). Surveillance of maize bought by poor blacks in South Africa for fusarial mycotoxins is obviously indicated. Certain Fusariurn species have long been known to be phytotoxic to a variety of crops including the palm tree, barley, wheat, oats, maize, rice, beans, peas, soya beans, tomatoes, potatoes, onions, asparagus, cotton, flowers, etc. Plants differ widely in their susceptibility to microfungi and to their mycotoxins (Nelson et d.,1981). Plants that are resistant can absorb trichothecene mycotoxins from the soil and sporadically may become toxic (Jarvis et al., 1981; Habermehl et aZ., 1984; Schoental, 1984c,e). D. ESOPHAGEAL CANCER IN CHINA AND ELSEWHERE

In China in the Linxian area of the northeastern Henan province, a high incidence of esophageal cancer has been reported not only in people, but also among chicken and other livestock (Lin and Tang, 1980; Li et aZ., 1980). The Chinese isolated a number of microfungi from local agricultural products, including Fusarium species, which, when inoculated on corn bread to which some sodium nitrite was added, produced nitroso compounds. Among the nitroso compounds, N-methylbenzylnitrosamine was detected which is known to induce esophageal tumors in experimental rodents. The Chinese found the

CARCINOGENIC METABOLITES OF FUSARIUM

229

soil of this region deficient in molybdenum, but did not report its selenium content. The types of weeds and of other vegetation that may be used locally for medicinal and other purposes may also be relevant. In Great Britain, Stocks pioneered the idea that the incidence of cancers may be related to geology and geochemistry, especially to the content of microelements, but did not take into consideration selenium (Stocks and Davies, 1960). In England and Eire pronounced local differences exist in the content and distribution of selenium in soils and in plants. Elevated levels have been reported in the soil of Counties Limerick, Meath, and Tipperary in Eire and in Staffordshire and Devon in England. In some localities, the soil contained up to 90 ppm of selenium and some plants contained up to 108 ppm (Webb and Atkinson, 1965; Nye and Peterson, 1975). In some areas, pockets of elevated selenium levels in top soils occur even in countries considered to have selenium deficiencies, such as New Zealand (Wells, 1967)or Finland (Koljonen, 1978).The occurrence of Fusarium mycotoxins in Finland has been reviewed (Hintikka, 1983). Selenium is an essential micronutrient. It is a constituent of glutathione peroxidase (Sunde and Hoekstra, 1980) and also may play other roles in the maintenance of the cellular redox potential and in the metabolism of xenobiotics (Diplock, 1976; Spallholz et al., 1981; Burk, 1983) (including the trichothecenes). Its deficiency can cause white muscle and other disorders in animals (Levander, 1982, 1983), and supplementation of livestock diets with sodium selenite has been advocated (Frost and Lish, 1975). However, high levels of selenium are toxic. Its role as a carcinogenic or as an anticarcinogenic agent is still controversial (Shapiro, 1973; Schrauzer, 1979; Jacobs and Griffin, 1981;Whanger et al., 1982). Selenium compounds are photosensitive. Some have the tendency to decompose and to yield the inert elemental selenium in uitro, in bacterial culture (Falcone and Nickerson, 1963; Gerrard et al., 1974),and also in the animal body (Schoental and van Dorst, 1983) (Fig. 2). This may be a factor responsible for the conflicting results often obtained and could confound their interpretation. The toxicity of excessive intake of selenium is well documented (Rosenfeld and Beath, 1964; Glover, 1970). Ingestion of diets containing more than 3 ppm of selenium can be deleterious and higher levels can be fatal. The main features of selenium poisoning in man include abdominal pain, nausea, vomiting, diarrhea, garlic odor of the breath and sweat (due to the volatile dimethylselenide), metallic taste, phar-

230

R. SCHOENTAL

FIG.2. Rat stomach showing streaks of deposited elemental red selenium. x 1.7.

yngitis, irritation of the mucous membranes simulating bronchopneumonia, chronic dermatitis, hair loss, fatigue, nervousness, and depression. Congestion of the liver and glomerulonephritis may occur. Acute sore throat among laboratory workers, who were associated with selenium studies in dogs, was attributed to exposure to dimethylselenide (recognized by its garlic smell) exhaled by the animals in a confined space (Motley et al., 1937). Resistance to selenium can be induced in experimental animals. Rats, for whom the LDm of selenium is 3.5 mgkg body weight when given as a single sc injection of selenite, have been found able to withstand about three times as high a dosage, when starting with low doses of 1.6 mgkg body weight and then gradually increasing the twice weekly injections over 140 days to 12 mgkg body weight (Cameron, 1947). The mechanism of this resistance remained unexplained and might have been due to the induction in specific tissues of reductive enzymes that have the ability to reduce selenium salts to its inert elemental form (Schoental and van Dorst, 1983). Certain of the selenium compounds belong to the ever growing group of agents which may be able to modify the development andlor the localization of tumors (Schoental, 198Oc; Zedeck and Lipkin, 1981)and possibly might also affect cardiovascular disorders (Salonen et al., 1982).These agents include vitamins A and E, ascorbic acid, the B group of vitamins, and their respective coenzymes, as well as the micronutrients which are involved in the regulation of respective enzymatic activities and of the cellular redox potential (Schoental,

CARCINOGENIC METABOLITES OF FUSARZUM

231

1976a). The appropriate balance of micronutrients is essential for the physiological functioning of the mammalian body and its organs. Inadequate, as well as excessive, intake of any of these factors can have deleterious consequences. The striking effects of diet on resistance to toxic agents include both the macronutrients and micronutrients. The question, why one particular person develops a tumor when so many apparently might have been similarly exposed resembles the question why does a tumor arise at a particular site and in a particular cell, though the carcinogen presumably had circulated through the body? What are the additional local factors involved? The degree of exposure and the biochemical “fitness” at the time of exposure determine the absorption, metabolism, and elimination. The eventual level of the carcinogenic entity that reached the cell at a time of a critical stage in its mitotic division may be relevant to the initiation of neoplasia. The probability that enough of the carcinogenic entity will be available to “hit, fit, and stick” at the critical site in a cell and allow it to survive is evidently very small. IV. Occurrence and Pathological Effects of T-2 Toxin

T-2 toxin [3or-hydroxy-4~,15-diacetoxy-8a-(3-methylbuty~loxy)12,13-epoxy-A%icothecene] belongs to a group of more than four dozen tetracyclic sesquiterpenoid compounds, known as trichothecenes (Fig. 3) (Godtfredsen et al., 1967), in recognition of the fact that the first member of this group was trichothecin, isolated from cultures of Trichothecium roseum in the course of a search for antibiotics and antifungal agents (Freeman and Morrison, 1948). The features in the trichothecenes responsible for their toxicity are the isolated double bond at C-9 and C-10, the epoxy ring on C-12 and C-13, and the presence of at least some esterified alcoholic hydroxyls. In T-2 toxin three of its four hydroxyls are esterified, two at C-4 and C-15 by acetic acid moieties and one at C-8 by a 3-methylbutyric acid group. T-2 toxin was first isolated from cultures of Fusarium tricinctum, a

FIG.3. The structure of T-2toxin [3c~-hydroxy-4~,l5-diacetoxy-8a-(3-methylb~ty1y1oxy)-12,13-epoxy-Ag-tricothecene]. (R’ = H; R2, R3 = CH3 CO; R4 = (CH& CH CH2 CO.)

232

R. SCHOENTAL

toxigenic contaminant of moldy corn, which caused cattle losses in the United States from a hemorrhagic syndrome (Bamburg et al., 1968).At a subsequent, similar outbreak of hemorrhagic syndrome among cattle in Wisconsin, T-2 toxin was found to occur naturally in the moldy corn at a level of approximately 2 ppm (Hsu et al., 1972).T-2 toxin was also detected in moldy fodder that caused outbreaks of similar mycotoxicoses in livestock in several countries including Canada (Greenway and Puls, 1976), eastern Europe (Szathmary et al., 1976; Szathmary, 1983), Japan (Yoshizawa and Morooka, 1981; Ueno, 1983), Scotland (Petrie et al., 1977), and elsewhere (see Martin, 1981). It has been detected in sorghum and in various other foods in India (Rukmini and Bhat, 1978; Ghosal et al., 1976, 1977,1978), in potatoes (Lafont et al., 1983), and in other agricultural products. T-2 toxin has been identified as a constituent of “poaefusarin,” obtained from cultures of F. poae in the USSR (Mirocha and Pathre, 1973). It has also been isolated together with some of its hydrolysis products, HT-2 toxin and neosolaniol (Table I), from cultures of F. sporotrichioides (Yagen et al., 1977), which, together with F. poae, had been involved in an outbreak of disorders among the human population of the Orenburg district in the USSR during World War I1 when, due to famine, people had to consume bread made from moldy millet. These often fatal disorders came to be known as ATA or “septic angina” (Nesterov, 1948; Mayer, 1953; Sarkisov, 1954; Joffe, 1960, 1978). Originally, steroidal constituents obtained from cultures of these Fusarium species were suspected to be the causative agents of ATA (Olifson, 1957). However, recently, workers in the USSR have confirmed that the toxicity of F . sporotrichiella cultures was due to their content of trichothecenes, consisting mainly of T-2 toxin, HT-2 toxin, and neosolaniol (Kotik et al., 1979). TABLE I T-2 TOXINAND ITS DEACYLATED METABOLITES

R Groups Substance

R’

R2

R3

R4

T-2 toxin HT-2 toxin 15-Deacetyl-HT-2 toxin (T-2 toxin triol) Neosolaniol 4Deacetylneosolanio1 T-2 tetra01

H H H

COCH3 H H

COCH3 COCH3 H

COCHeCH(CH3)g COCH2CH(CH3)2 COCHgCH(CH3)2

H H H

COCH3 H H

COCH3 COCHB H

H H H

~~~

~

CARCINOGENIC METABOLITES OF FUSARIUM

233

Steriods isolated by careful fractionation of extracts from cultures of F. sporotrichioides were not toxic (Yagen et al., 1977, 1980). Pure T-2 toxin reproduced in cats some of the features of ATA, including the characteristic leukopenia (Sato et al., 1975; Lutsky et al., 1978). T-2 toxin is one of the most toxic among the trichothecenes (Bamburg and Strong, 1971). The trichothecenes’ chemistry, relative acute and chronic toxicities, and other aspects of their effects on animals in uiuo and on isolated cells in uitro have been extensively reviewed (Bamburg and Strong, 1971; Smalley and Strong, 1974; Ohtsubo and Saito, 1977; Sato and Ueno, 1977; Ueno, 1977a,b, 1980, 1983a; Palti, 1978). A bibliography published in 1980 contained 2978 references (Meyer and Frank, 1980). In long-term experiments T-2 toxin induced in the digestive tract of white Wistar rats lesions and tumors (benign and malignant). These included ulcers and squamous hyperplasias of the forestomach, adenocarcinomas of the glandular stomach and of the duodenum, multiple pancreatic tumors of the acinar and of the islet cells, and also brain tumors and cardiovascular lesions (Schoental et al., 1979). The vascular effects caused by T-2 toxin included hypertension (Wilson et al., 1982). The neoplastic lesions were found in rats, which survived more than a year after the first of a few, relatively large, ig doses of T-2 toxin. In experiments in which rats were killed 8 months after having received in the diet 5-15 ppm of T-2 toxin, no neoplastic lesions have been seen (Marasas et al., 1969). However, papillomas with hyperkeratoses in the squamous part of the forestomach have already been reported previously in the USSR in rats that were fed grain contaminated with F. poae (Rubinstein et al., 1967). Similar lesions were seen in the esophagus and in the squamous part of the stomach of rats given crude alcoholic extracts from F. poae and F . sporotrichioides (Schoental and Joffe, 1974). It is not yet known whether the carcinogenic effects are due to T-2 toxin per se or to some of its metabolic products. Using fusarenon-X, a trichothecolone derivative in which the eighth-position is oxidized to a keto group, Japanese workers found only low incidence of tumors among their experimental rodents. Their controls also had some tumors (Ohtsubo and Saito, 1977; Saito et al., 1980). T-2 toxin does not appear to require metabolic activation by specific liver enzymes and is not hepatotoxic. It exerts direct cytotoxic action on several types of cells, including cardiac myocytes (Yarom et al., 1983)and the endothelial cells lining the arteries, as evidenced by the

234

R. SCHOENTAL

thickening of the arterial walls in some of the rats surviving large doses of T-2 toxin (Schoental et al., 1979; Wilson et al., 1982).In mice, datoxin B1 exerted synergistic action on the acute toxicity of T-2 toxin (Lindenfelsen et al., 1974). Like most of the trichothecenes, T-2 toxin is not mutagenic in tests in uitro which use Escherichia coli or Salmonella typhimurium, whether without or in the presence of hepatic microsomal fraction S-9 (Ueno and Kubota, 1976; Ueno et al., 1978; Wehner et al., 1978). The action of T-2 toxin has been described as radiomimetic (Saito et al., 1969; Ueno, 197%). T-2 toxin inhibits the synthesis of proteins (Ueno et al., 1969; Ueno and Shimada, 1974; Wei and McLaughlin, 1974; Cundliffe et al., 1974; Cundliffe and Davies, 1977)and of DNA (Ueno and Shimada, 1974; Ueno, 1983a). Actively proliferating cells are particularly sensitive to T-2 toxin, such as the myelocytes in the bone marrow (Yarom et al., 1984a),the lining of the gastrointentinal tract, etc. I n uitro T-2 toxin was shown to inhibit the phagocytosis of opsonized streptococci by human leukocytes and their chemotaxis, which explains the lack of cellular infiltration in response to experimental infections of rats with staphylococci (Yarom et al., 1984) or of mice with herpes simplex virus (Friend et al., 1983). In uitro, T-2 toxin inhibits the aggregation of platelets from human blood (Yarom et al., 1984b) and, after a lag period, causes hemolysis of rat erythrocytes (Segal et al., 1983). T-2 toxins (and certain other trichothecenes) cause similar acute and subacute effects in a number of animal species including cats, mice, rats, guinea pigs, cattle, pigs, dogs, monkeys, chicken, turkeys, geese, trout, as well as in human beings (Marasas et al., 1969; Bamburg and Strong, 1971; Wyatt et al., 1973; Smalley and Strong, 1974; Ohtsubo and Saito, 1977; Sat0 and Ueno, 1977; Ueno, 1977a,b, 1980, 1983a; Yagen et al., 1977). T-2 toxin was teratogenic when tested in mice (Stanford et al., 1975). The effects on humans of the trichothecene, diacetoxyscirpenol (Anguidine, NSC-141537), have been observed in the course of its evaluation for the treatment of cancer patients. V. Diacetoxyscirpenol(Anguidine, NSC-141537)

The observations of decreased numbers of leukocytes in the blood of patients suffering from ATA led to the testing of fusarial preparations in mouse leukemias (Lyass, 1955). More recently, the effects of diacetoxyscirpenol have been evaluated in uitro and in vivo with the view of using it as an antitumor agent.

CARCINOGENIC METABOLITES OF FUSARIUM

.

O R ~

235

-

FIG.4. The structure of diacetoxyscirpenol(Anguidine, NSC-141537). It differs from

T-2toxin by the absence of the C-8 substituent.

Diacetoxyscirpenol (DAS) (Fig. 4) was the first toxic trichothecene isolated from Fusarium scirpi (Brian et al., 1961). Its total synthesis has recently been accomplished (Brooks et al., 1983).Due to its cytostatic and cytotoxic action, DAS has been suggested as an anticancer agent (Stiihelin et al., 1968). DAS, under the name of anguidine (NSC141537), appeared to be active in uiuo against transplantable colon carcinoma induced by N-methyI-N-nitrosourethane in C57BU6 mice or by l$-dimethylhydrazine in BALBlc I mice (Corbett et al., 1977). When tested in uitro on human colon cancer cell lines, DAS blocked the transition stages from GI to S and from G2 to mitosis (Dosic et al., 1978). Anguidine was evaluated as an antitumor agent in patients with various tumors in several United States hospitals (Murphy et al., 1978; Goodwin et al., 1978; Belt et al., 1979; DeSimone et al., 1979, etc.), but the results were disappointing. Phase I1 evaluation of anguidine in another 42 patients also proved disappointing. Minimal regression was seen in only 2 patients (of which one died from sepsis 11 weeks after the start of therapy), in 13 patients, the disease remained stable, and the therapy failed completely in 24 patients. The toxicity symptoms were described as “nausea, vomiting, phlebitis, hypotension, erythema of the skin, confusion, chills, fever, ataxia, and weakness.” Myelosuppression was less severe when injections were given over 4 or 8 hr, instead of a quick “push,” but the other toxicity symptoms of the gastrointestinal, the cardiovascular, and the CNS remained prominent. Patients with liver damage or cardiac instability were more severely affected, and the dosage of anguinine had to be lower than 5 mg/m2 in such cases. “The poor patient tolerance associated with anguidine treatment makes this drug less than desirable for the management of patients with advanced malignancies.” “The relative lack of antitumor activity and the excessive degree of toxicity . . . are somewhat discouraging” (Yap et al., 1979). Deoxyniualenol (DON), known also as vomitoxin, is a trichothecene frequently present in moldy cereals often together with zearalenone

236

R. SCHOENTAL

(Vesonder et al., 1981; Vesonder, 1983; Trenholm et al., 1980, 1981, 1983). This compound has no substituents at position C-4, and its hydroxyls are not esterified. It is therefore much less toxic than DAS or T-2 toxin. Its long-term effects have not yet been reported. Fusarium moniliforme recently received much attention. It has been found in areas of high incidence of esophageal tumors, such as the Linxian county in China (Li et al., 1980; Lin and Tang, 1980) and in Transkei, South Africa (Marasas et al., 1979; Marasas, 1982). The South African workers found liver tumors and hyperplasia of the basal cells of the esophageal epithelium in rats that were fed corn (maize) inoculated with F. moniliforme (MRC strain 826). The chemical nature of the carcinogenic agent(s) in this case is not known (Marasas et al., 1984). Rats given in diet DAS (10 ppm for 10 weeks) developed “hyperplasia and thickening of the esophageal mucosa,” while rats given Nmethylbenzylnitrosamine had “macroscopically visible papillomas and invasion of submucosa” (Craddock and Sparrow, 1985). VI. The Antileukemic Baccharinoids and Other Macrocyclic Trichothecenes

A. BACCHARINOIDS

In the course of a search for “natural” antileukemic agents, the plant, Baccharis megapotamica Spreng (Astraceae) was investigated. This plant is known to cause serious toxicoses in cattle that graze in the marshy areas of Brazil where it grows. Dried Baccharis plants from Brazil, collected in May, 1975near Curitiba, approximately 1000 miles south of Rio de Janeiro were found to contain 200-300 ppm of macrocyclic trichothecenes, named baccharinoids, which were very active in vivo against mouse leukemia P-388 and in vitro against human nasopharyngeal KB cells (Kupchan et al., 1976). Baccharinoids comprise a group of macrocyclic trichothecenes which are related to roridin A. Some have an additional epoxide ring (Baccharin, Fig. 5) on the C-9 and the C-10 positions; others have a phydroxyl on the C-8 position in the trichothecene moiety. The acyl moiety in the macrocycle of roridin is also oxidized in some of the baccharinoids (Jarvis et al., 1983). Subsequent studies have shown that baccharinoids can be formed in the plant in the course of translocation of mycotoxins, such as verrucarin A or roridin A, produced by the soil microfungi, e.g., Myrothe-

CARCINOGENIC METABOLITES OF FUSARIUM

237

FIG.5. Baccharin, one of the macrocyclic trichothecenes, isolated from the Brazilian (toxic to livestock) plant, B . megapotamica. It has additional epoxy rings (1) at C-9 and (2-10 and (2) at C-2' and C-3' (due to metabolic oxidation of roridin A in the plant). Baccharin shows higher antileukemic activity in mice than roridin A.

cium uerrucaria. When seedlings of Baccharis plants were kept in hydroponic cultures to which roridin A was added (e.g., 50 mg roridin A in 5 ml of ethanol was added to 50 ml of distilled water, 10 p M in calcium sulfate), similar baccharinoids were found in the aerial parts of the plants. The roots contained only traces of the trichothecenes. Control seedlings, kept in aqueous solutions with or without added ethanol, showed no trace of mycotoxins (Jarvis et al., 1981). It is of interest that the high antileukemic activity was shown by the /3-epimers of chemically prepared oxidation products of roridin A. However, the P-epoxide of anguidine shows apparently no activity, though the parent compound is active (Jarvis et al., 1980, quoting Doyle). A review of the macrocyclic trichothecenes, their biosynthetic precursors, the trichoverrins, and the relation of structure to biological activity of these interesting natural products has been published (Jarvis et al., 1983).It would be of interest to know their respective carcinogenic potentialities. B. MIOTOXINS Another Brazilian shrub, Baccharis coridifolia (Compositae), is known to be very toxic to livestock in Brazil and in the neighboring countries. It can cause gastrointestinal and nervous disorders also in

238

R. SCHOENTAL

experimental animals (Tokamia and Dobereiner, 1975).This plant has been found to contain macrocyclic trichothecenes, roridins A and E, absorbed from the soil, and also a new trichothecene, miotoxin A (Fig. 6). Its structure has been determined to be 4’-hydroxyroridin E (using NMR spectroscopy and X-ray analysis) (Habermehl et al., 1984). Two additional macrocyclic trichothecenes, miotoxin B and C, isolated from B. coridifolia, are also derivatives of roridin A. Their respective acyl moieties differ in the oxidation states of C-2’, C-3’, or C-3‘ (Habennehl and Busam, 1984). These workers report that the microfungus, M. uerrucaria, was isolated from the soil surrounding the growing plants. Esophageal and rumenal tumors encountered in cattle have usually been attributed to bracken, Pteridium aquilinum, that also grows abundantly in certain parts of Brazil (Dobereiner et al., 1967; Evans, 1976; Cotchin, 1984).However, trichothecenes absorbed from the soil may contribute also to the carcinogenic effects of bracken, which is resistant to certain Fusaria (Hutchinson, 1976; Hirono et al., 1984; Schoental, 1984c,e). Translocation of mycotoxins from soil into growing plants appears to be more frequent than previously realized. Another example, recently demonstrated, is the case of the tremorgenic mycotoxin, verruculogen, causing ryegrass staggers (Day and Mantle, 1981). Another Brazilian plant, Dqfenbachia seguinae Schott (Araceae) (Caladiumsequinum), which is known locally as able to induce sterility (Kingsbury, 1964),has been investigated with the aim of using it as oral contraceptive (Madaus and Koch, 1941). During the war, extracts from Diffenbachia plants grown in hot houses in Germany, prepared

OH

FIG.6. Miotoxin A (4’-hydroxyroridin E) isolated with other macrocyclic trichothecenes from the Brazilian plant, B . cordifolia DC (mio-mio)toxic to livestock.

CARCINOGENIC METABOLITES OF FUSARIUM

239

by I. G. Farben under the name of “Salistrum,” were added to the soup of Jewish inmates of the Auschwitz concentration camp, other inmates serving as controls, and its effects on the reproductive organs were examined. Some of the inmates were later rescued and brought to the United States. Dr. Jean Jofen, a psychologist, working in New York discovered that children born in the United States to the survivors of Nazi concentration camps showed surprisingly low IQ (Jofen, 1972).These children of low intelligence were the offspring mainly of women, who spent some time at Auschwitz concentration camp. No such effects were noticed among the offspring of inmates of other concentration camps. Such effects on the intelligence of offspring of the women exposed to extracts from Diffenbachia plants indicate the need for identification of the active agent(s), which the inhumane experiments of the Nazis on humans did not succeed to do (Arditti and Rodriguez, 1982). In view of the evidence described above that the toxicity of B . megapotamica was traced to trichothecene mycotoxins translocated from the soil and metabolized in the growing plant (see also Schoental, 1 9 8 4 ~it) might ~ be justified to investigate Brazilian D. sequinae for the presence of trichothecenes and also of the estrogenic mycotoxins, such as zearalenone and its derivatives. The trivial name of Diffenbachia plants is “dumbcane.” It resembles sugar cane, but it causes irritation and numbing of the buccal cavity, edema of the tongue, and interferes with swallowing and breathing when ingested (Kingsbury, 1964). Such effects would be compatible with the known action of the cytotoxic and irritant trichothecenes, while the contraceptive effects (inhibition of menstruation that swiftly followed the consumption of “Salistrum,” as described by Jofen, 1972) might have been due to zearalenone and its derivatives. Both types of mycotoxins can be produced by Fusarium species. Follow-up of the children with low IQ would appear indicated.

C. SATRATOXINS Macrocyclic trichothecenes are also produced by the microfungus, Stachybotrys atra (syn. Stachybotrys alternans and Stachybotrys chartarum) hyphomycetes, which have been the cause of poisoning and disorders, known as stachybotryotoxicoses, among farm animals, especially horses in eastern Europe, and have affected man also (Drobotko, 1945; Forgacs and Carll, 1962).These soil microfungi are major decomposers of cellulose. Their black spores carried by the air are

240

R. SCHOENTAL

153

OH Me

FIG.7. Satratoxin H, one of the macrocyclic trichothecenes produced by Stachybotrys atra, the hay microfungi, known to cause stachybotryotoxicoses in man and livestock.

often present in hay and straw. Like certain Fusarium species, S. atra can produce, under appropriate environmental conditions and at relatively low temperatures, very toxic trichothecenes, the satratoxins, characterized by a macrocyclic diester ring between C-4 and C-15 of the trichothecene moiety. The first identified was satratoxin H (Fig. 7.) (Eppley and Bailey, 1973; Eppley et al., 1977)which contains a tetrahydropyran ring between C-6’ and C-12’. Satratoxins F and G differ fiom it by having an epoxy ring on the C-2’ and the C-3’ (Eppley et al., 1980; Jarvis et al., 1983). In satratoxin F, an acetyl group is present instead of a methyl carbinol at C-6’. Satratoxin F represents an intermediate oxidation stage between the verrucarins and roridins, but whether it is a biosynthetic intermediate is not known (Jarvis et al., 1983). The carcinogenic potentialities of the satratoxins remain to be investigated, as well as of the related cytotoxic macrocyclic vertisporin, produced by Verticimonosporium diffructum, in which a dihydroxylated tetrahydrofuran ring is fused to the tetrahydropyran ring of satratoxins and the C-7’ and C-8’ double bond is saturated (Minato et al., 1975). VII. Metabolism of T-2 Toxin

A number of deacylated metabolites of T-2 toxin was detected in animal excreta: HT-2 toxin, 15-deacetyl-HT-2 toxin (T-2 toxin triol), neosolaniol, 4-deacetylneosolanio1, and T-2 tetra01 (Table 11). The amounts of some of these metabolites corresponded to about 1% of the administered dose or less. The major metabolites included 3’-hydroxy-T-2 toxin and 3’-hydroxy-HT toxin (Yoshizawa et al., 1982),

CARCINOGENIC METABOLITES OF FUSARTUM

24 1

TABLE I1 T-2 TOXIN,ITSMETABOLITES, AND THEIR RELATIVETOXICITIES

Substance T-2 toxin 3'-Hydroxy-T-2 toxin HT toxin (4-deacetyl-T-2toxin) 3'-Hydroxy-HT-2 toxin Neosolaniol (8-deacyl-T-2toxin) T-2 toxin trio1 4-Deacetylneosolanio1 T-2 tetra01

Toxicity in rodents

++++ ++++ +++ +++ ++ . + + +

which are also produced in vitro by liver homogenates from mice and monkeys (Yoshizawa et al., 1984). Radiolabeled [3-3H]T-2 toxin of high specific activity was synthesized by oxidation of the hydroxyl of T-2 toxin at the C-3 position with dimethylsulfide-N-chlorosuccinimideto a carbonyl, followed by reduction of this carbonyl with sodium[3Hlbor~hydride(Wallace et al., 1977), and was used for metabolic studies in mice and rats (Matsumoto et al., 1978), in chicken and laying hens (Chi et al., 1978a,b; Yoshizawa et al., 1980a),in pigs (Robison et al., 1979),and in a lactating cow (Yoshizawa et al., 1981). The radioactivity of [3H]T-2 toxin disappears rapidly from the animal tissues and is excreted almost completely in the urine and feces within 72 hr. A small proportion was secreted into the pig's and cow's milk (Yoshizawa et al., 1981).As [3-3H1T-2 toxin, in the course of its metabolism, has been reported to yield some 3H20 this indicates that exchange of tritium can occur in the body. Thus, the results based on estimation of tritium radioactivity have to be interpreted with caution (Mirocha, 1983). The enzymatic pathways leading to the various T-2 toxin metabolites depend on the availability of the respective coenzymes. Hence the nutritional state, especially in respect to B and other vitamins, can critically affect the toxic manifestations. During the outbreaks of fusarial mycotoxicoses, malnourished people and animals were indeed more severely affected (Schoental, 1983).As yet, no evidence is available whether the opening of the epoxide ring, and molecular rearrangement, or the oxidation at the C-9 and C-10 double bond takes place in cells at the site of T-2 toxin action. The enzyme systems

242

R. SCHOENTAL

which are involved in the carcinogenic action of T-2 toxin are unknown. Deacylation of T-2 toxin has also been observed in vitro (Ellison and Kotsonis, 1974; Ohta et al., 1977; Yoshizawa et al., 1980b).The S9 microsomal fractions from human, bovine, mouse, monkey, or rat liver homogenates proved very effective in the removal of the acetyl group at the C-4 position to form HT-2 toxin. Gastric and intestinal strips from rats, although less effective than the liver fractions, were also able to deacylate T-2 toxin. The removal of the acidic residues and the formation of T-2 tetraol proceed in a stepwise manner which appears to depend on the type of the tissue and on the pH (Yoshizawa et al., 1980b, 1981).The products of deacylation are less toxic than T-2 toxin. Among the major metabolites, 3’-hydroxy-T-2 toxin and 3’-hydroxy-HT toxin, which are also formed in uitro, appear mainly to retain the toxicity of the parent compounds (Yoshizawa et al., 1984). Although T-2 toxin is mainly eliminated in the excreta of broiler chickens (Chi et al., 1978b) some of it is carried over into the eggs when given to laying hens (Chi et al., 1978a). The eggshells may become thin and fragile, their production and hatchability decreased (Wyatt et al., 1975). Eggs of hens, especially the egg yolk and its lipids, have previously been reported to be carcinogenic to mice, but the carcinogenic agent(s) have not been identified (Szepsenwol, 1963, 1964). Infertile eggs that contain red blood droplets should not be consumed by man according to Jewish tradition, which paid much attention to preventive medicine (Schoental, 1980b). It would be of interest to know whether fusarial metabolites play a role in Marek‘s disease or in other lymphoproliferative disorders of chicken (Nazerian, 1973) and of other animal species. It is not unlikely that nivalenol, deoxynivalenol, etc., which frequently are found in stored cereal grains, may represent products of metabolic activity of certain microorganisms, which are able enzymatically to hydrolyze ester groups and to bring about the respective oxidations andfor reductions, that transform T-2 toxin and related trichothecenes to the much less toxic deoxynivalenol, etc. Partial detoxification by hydrolysis of the ester trichothecenes has been suggested as an explanation of the occurrence of pellagra among people who consumed bread made from (moldy) maize, but not among those who made it into tortillas (Schoental, 1980e).The preparation of corn flour for tortillas includes soaking the maize grain in warm slaked lime. This treatment would obviously hydrolyze ester groupings and to some extent detoxify mycotoxins. The incidence of

CARCINOGENIC METABOLITES OF FUSARZUM

243

tumors of the digestive tract among “recovered’ pellagrins needs to be investigated. VIII. Effects of T-2 Toxin and Related Trichothecenes on the Immune System

Many carcinogenic agents (including radiation) are immunosuppressive (Berenbaum, 1964; Stjernsward, 1969) and so is T-2 toxin, DAS, and the trichothecene-producing microfungi, Fusariurn niuale (Saito et al., 1969) and S. atra (Danko and Krasznai, 1976). Stiihelin et al. (1968) described lesions in the thymus and lymphoid tissues caused by DAS. T-2 toxin causes similar lesions and immunosuppression in several animal species (including man) when exposed to subtoxic, large levels (Schoental and Joffe, 1974; Boonchuvit et al., 1975; Schoental et at., 1979; Rosenstein et d.,1979; Lafarge-Frayssinet et al., 1979,1981; Jagadeesan et al., 1982). The immunosuppression appears to be transitory and may be followed by recovery; the depletion of lymphoid cells may be followed by regeneration, hyperplastic or neoplastic changes critically depending on the dose. The ester trichotehcenes are cytotoxic to actively proliferating cells. They suppress lymphocytic functions, especially of the T cells, inhibit antibody response in uiuo and in uitro, and inhibit allograft rejection in mice (Rosenstein et al., 1979). Their effects on the T-independent immune response is controversial (Otokawa et al., 1979; Rosenstein et al., 1981).The mechanisms by which the time-related, immunological changes are induced are complex (Faith et al., 1980), and their relations to the metabolic processes involving the trichothecenes are little understood. Yet, immunosuppression by the trichothecenes, when it occurs, may have important consequences in relation to bacterial and viral disorders (Richard et al., 1978; Schoental, 1981d; Friend et al., 1983)and to lymphoreticular and certain other tumors in animals and man.

Although liver microsomal fractions are very effective in deacylation of T-2 toxin, no effects on the hepatic DNA have been observed in uiuo or in uitro. It is the DNA of lymphocytes which is affected by T-2 toxin and which develops breaks both in uiuo and in uitro, but the lesions appear to be reversible (Lafarge-Frayssinet et al., 1981; Jagadeesan et al., 1982).Studies of the incorporation of [3H]thymidine into human fibroblasts in uitro indicate that unscheduled DNA synthesis may occur after treatment with T-2 toxin, HT-2 toxin, and T-2 toxin tetra01 (Agrelo and Schoental, 1980; Oldham et al., 1980).

244

R. SCHOENTAL

IX. Detection and Estimation of Trichothecenes

A number of biological and physicochemical methods have been developed for the detection and the estimation of the toxic trichothecenes. The recent improvements have increased their sensitivity and specificity (Ueno et al., 1973; Pathre and Mirocha, 1977; Ueno, 1980, 1983; Takitani and Asabe, 1983). A simple skin test was used in the isolation of the active principles from cultures of toxigenic Fusarium species, which depends on the grossly recognizable, irritant and necrotizing action of ester trichothecenes at the site of their application to the shaved skin of various rodents (Joffe, 1960; Ueno et al., 1970; Chung et al., 1974; Hayes and Schiefer, 1979). The sensitivity of the skin varies depending on the site of application, on the species, and on the age of the animal; the skin of rabbits and guinea pigs is more sensitive than that of mice and rats. This test is semiquantitative and allows the detection of about 0.2 pg of T-2 toxin per assay. Other biological methods are based on the toxicity of trichothecenes to plants (Marasas et al., 1971; Joffe and Palti, 1974; Siriwardana and Lafont, 1978), protozoa (Ueno and Yamakawa, 1970), brine shrimps (Artemia salina) (Harwig and Scott, 1971; Eppley, 1974; Prior, 1979), chick embryos (Ichinoe et aZ., 1972; Polzhoefer and Niehuss, 1980), and to various mammalian cells. Lymphocytes are particularly sensitive to T-2 toxin and to DAS (Lafarge Frayssinet et al., 1981; RobbanaBarnat et al., 1982). A sensitive biological method for the estimation of trichothecenes is based on their inhibition of protein synthesis evaluated by the incorporation of labeled leucine into rabbit reticulocytes. This test is able to detect T-2 toxin at 0.03 puglml of the medium (Ueno et al., 1969; Ueno and Shimada, 1974). The mechanism of inhibition of protein synthesis is related to the structure of the trichothecenes: T-2 toxin and DAS inhibit the initiation stage, while venucarol, for example, inhibits the elongation-termination stage (Wei and McLaughlin, 1974; Cundliffe et al., 1974; Cundliffe and Davies, 1977). Physicochemical methods are more suitable for the quantitation and for the identification of individual constituents present in mixtures of mycotoxins. Unless present in very high concentration, extracts from biological materials have to be cleaned up, however, before use either by partition between various solvents (Tatsuno et al., 1973; Eppley, 1975), by dialysis (Roberts and Patterson, 1975; Patterson and Roberts, 1979), or by adsorbent columns (Mirocha et d., 1976b; Naoi, 1983).

CARCINOGENIC METABOLITES OF FUSARIUM

245

Where reference substances are available, individual trichothecenes can be detected by chromatographic methods, such as mono- or two-directional thin-layer chromatography (TLC), or high-resolution TLC (using two-way developments in several solvent systems). The positions of the trichothecene on the developed plates are visualized by spraying with sulfuric acid, p-anisaldehyde, 4-(p-nitrobenzyl) pyridine, or nicotinamid0-2-acetylpyridine~ etc. (Takitani et al., 1979; Sano et al., 1982; Takitani and Asabe, 1983). More sensitive are methods in which, after appropriate derivatization, the mixtures are resolved by gas-liquid chromatography (Ikediobi et al., 1971; Romer et al., 1978) or by high-performance liquid chromatography (Schmidt et al., 1981) with electron capture or mass spectrometric detection (Pareles et al., 1976; Collins and Rosen, 1979; Pathre and Mirocha, 1978; Szathmary et al., 1980; Scott et al., 1981). In order to reliably identify specific trichothecenes, the use of more than one procedure is desirable. The methodology of estimation of trichothecenes has been assessed (Eppley, 1979; Scott et al., 1981; Scott, 1982), and recommendations for further studies have been proposed by a panel of experts (Eppley, 1982). Radioimmunoassay of T-2 toxin in biological materials has been described which allows the detection of a few ppb of T-2 toxin in cereals and also in blood, milk, and urine (Lee and Chu, 1981a,b). A recently simplified procedure should facilitate the search for this carcinogenic mycotoxin in biological materials (Fontelo et al., 1983). The main problem involved is how to obtain representative samples from great amounts-shiploads-of nonhomogeneous agricultural products that might contain “hot spots” of fungal activity with high levels of mycotoxins in an otherwise wholesome bulk (Shlosberg et al., 1984). The possibility has also to be considered that during storage the known mycotoxins may undergo further metabolic changes to, as yet, not identified products, some of which may be nontoxic (Grove and Mortimer, 1969). X. Chemistry and Biological Activity of Zearalenone

Zearalenone (Z) [6-(l0-hydroxy-6-oxo-trans-l-undecenyl)-~-resorcylic acid lactone] (Fig. 8) [Chem. Abstr. designation [ S - ( E ) ] 3,4,5,6,9,10-hexahydro- 14,16-dihydroxy-3-methyI1H-2-benzoxacyclotetradecin-177(8H)-dione]is a secondary metabolite of Fusarium graminearum (Gibberella zeae) and of a number of other Fusarium

246

R. SCHOENTAL

FIG. 8. The structure of zearalenone [6-(lO'-hydroxy-6'-oxo-trans-l'-undecenyI)-Presorcylic acid lactonel.

species, which are common contaminants of cereals, particularly of corn (maize, 2. mays). Z was isolated as a result of a search for estrogenic agent(s) which were responsible for sporadic outbreaks of hyperestrogenism among pigs consuming moldy feed. Such pigs showed " vulvar hypertrophy and occasional vaginal eversion in females, preputial enlargement in castrated males and prominent mammary glands in both sexes" (Stob et al., 1962). Z appeared also effective in improving the growth rate and feed utilization in sheep and was suggested for use as anabolic agent in animal husbandry (Stob et aZ., 1962). The structure of Z has been established by Urry et al. (1966).Z and a number of its derivatives and analogs have been synthesized by several groups of workers (Girotra and Wendler, 1967, 1969; Taub et al., 1968;Mattas et al., 1968; Hurd and Shah, 1973; Masamune et al., 1975) and have been discussed in relation to some of their biological effects (Shipchandler, 1975). Various aspects of chemical, biochemical, and biological activities of Z and of some of its derivatives have been extensively reviewed (Mirocha et aZ., 1971, 1977; Mirocha and Christensen, 1974; Pathre and Mirocha, 1976, 1977; Hidy et aZ., 1977; Hurd, 1977). F. graminearurn Schwabe when grown on maize (45% moisture) for about 10 weeks at 12°C will produce, besides 2, four additional metabolites: 5-formylzearalenone, 7'-dehydrozearalenone, and two epimeric 8'-hydroxyzearalenones (Bolliger and Tamm, 1972). Fusarium roseum ( F . graminearum) can form also other metabolites, including 6',8'-dihydroxyzearalene and two epimeric 3'-hydroxyzearalenones, which accumulate as the cultures age. These are products of second phase oxidation of Z. Additional products have been detected in cultures of F. rosetcm (F. graminearum), in which the macroring is open, and in which the side chain is degraded to various compounds, including 3,5-dihydroxyphenylaceticacid (Steele et aZ., 1976; Pathre et al., 1980). In Z, the lactone ring is relatively stable and resistant to hydrolysis,

CARCINOGENIC METABOLITES OF FUSARIUM

247

being sterically hindered by the presence of the methyl group at C-10‘ and by hydrogen bonding with the hydroxyl on C-2 of the benzene ring; hence, the simpler products might be either intermediates in the biosynthesis of Z or of other ramification pathways. The production of Z as well as of its various derivatives varies greatly, not only depending on the strain and the particulate isolate of Fusarium, but also on the environmental conditions and the age of the cultures (Hidy et at., 1977; Naik et al., 1978).An isolate of F. roseurn “Gibbosum” was found to produce high yields of Z (up to 5% of dry weight of a rice culture) when grown at temperatures 20-25°C on parboiled rice of 53-60% moisture for more than 9 days. This isolate was successfully used for the preparation from [l-I4C]acetate of [ W ] Z with relatively high specific activities (1.63-46.5 pCi/mmol) for metabolic studies (Hagler and Mirocha, 1980).F. roseum “Gibbosum” also produces on rice cultures small amounts of a-zearalenol (Hagler et al., 1979).A convenient method has been devised for the detection of the various mycotoxins produced by Fusarium isolates, by growing them in vermiculite cultures containing a liquid medium in an inert mineral matrix (Richardson et al., 1984). In cereals, small amounts of zearalenol(s) accompany Z. In two samples of oats from Finland, containing 25 and 135mgkg of Z, the levels of zearalenol were 1.5and 4.0 mgkg, respectively. A sample of American corn contained 18 mgkg of Z and 0.15 mgkg zearalenol (Mirocha et al., 1979).

b

a

C

FIG.9. The structure of estrogenic compounds zearalenol (a), estradiol (b), and diethylstilbestrol (c)showing the two, almost equidistal hydroxyl groups at the ends of their molecular structures. These hydroxyl groups are essential for estrogenic activity.

248

R. SCHOENTAL

I’

a

b

FIG.10. The structures of zearalenol (a) and zeranol (b) (Ralgro)devoid of the C-1’ and C-2’ double bond.

As in the case of estradiol-17p which is more active than estrone, so is zearalenol more active than zearalenone (about three times), but the relative value depends on the method used for evaluation. The estrogenic activity of estrogens depends critically on the two hydroxyls at the end of their molecular structures, which are present in zearalenol, estradiol, and diethylstilbestrol (Fig. 9), as well as in zeranol (Fig. 10). The estrogenic activity greatly depends also on the stereochemical configuration of the hydroxyl in the nonaromatic moiety of their structures (Hurd, 1977; Duax and Weeks, 1980). XI. Teratogenic Effects of Zearalenone and Bone Lesions

High levels of Z, like those of other estrogenic agents (whether the “natural” hormones, the synthetic steroids, diethylstilbestrol, and other stilbene derivatives, etc.), will affect fertility, reproduction, and lactation in many animal species including humans, and in the susceptible fetus, may cause death or developmental defects, which could manifest themselves as structural, functional, or behavioral abnormalities (IARC, 1979; Nora and Nora, 1975; Schardein, 1980) or neoplasias (Rice, 1979; Herbst and Bern, 1981; Schoental, 1974a, 1981d, 1982a,b, 1983a,d). In teratogenicity studies of zearalenone in rats, decrease of maternal growth and incidence of bone lesions in the fetuses was dose dependent and at 10 mg/kg/day, highly significant (Table 111) (Ruddick et aZ., 1976). Excessive medullary trabeculation of long bones has been described in rats given zearalenone in their diet (Becci et aZ., 1982; Report NTR-81-54, 1982). Splay-leg syndrome occurs in piglets from sows consuming moldy diet and has been reproduced using Z, which decreased the number and the weights of some of the live-born piglets (Christensen et al., 1972; Miller et al., 1973; Chang et al., 1979). A piglet born to a sow given Z-contaminated rations (40ppm) had bone deformations (Chang et al., 1979).

CARCINOGENIC METABOLITES OF FUSARIUM

249

TABLE 111 MATERNALAND FETALEFFECTS OF ZEARALENONE ADMINISTERED OMLLYDURING DAYS6-15 OF GESTATION TO THE RAT' Dose (mgkg)

0 1

5 10

Litters

Maternal weight gain (g) (mean)

Litter size (mean)

Fetal weight (g) (mean)

Fetal skeletal defects (%)

Deciduomas (%)

10 10 9 10

53.3 56.2 50.0 38.7

12.9 12.7 11.6 11.3

4.9 4.5 4.9 4.3

2.4 12.8 26.1 36.8

6.5 5.2 5.4 10.3

Ruddick et al. (1976).

The potential of estrogenic agents to induce lesions of the bones and possibly bone tumors (osteosarcomas) needs more study (Silberberg and Silberberg, 1971; Little, 1973; Highman et al., 1981). In the USSR, a bone disease, which used to occur among school children (mainly in northern Siberia, in the area of Lake Baikal) and was known as Kashin-Beck or as Urov disease or as osteoarthrosis deformans, has been correlated with the consumption of bread made from grain contaminated with Fusarium sporotrichiella. The incidence of KashinBeck disease has been reported to decrease when the cereals were imported from unaffected areas (Schwarzman, 1937; Rubinstein, 1949; Nesterov, 1961). XII. Metabolism of Zearalenone

In metabolic studies in vivo using mainly females of several species of animals, Z has been found to undergo reduction to zearalenols and conjugation (Mirocha et al., 1981). The quantitative aspect of the metabolic transformation of Z depends on the species, the age and the sex of the animal, its endowment of enzymes and their stereospecificity, etc. Using liver homogenates, or their fractions, the reduction in vitro of Z to the two epimeric a- and p-zearalenols has been found to involve several enzyme systems (reductases and/or 3a-hydroxysteroid dehydrogenases) which require coenzymes, either NADH or NADPH, and show species differences (Kiessling and Pettersson, 1978; Ueno and Tashiro, 1981; Olsen and Kiessling, 1983), but no conjugation has been observed in uitro. In attempts to isolate some of the enzymes involved in the reduction of Z by rat liver microsomal fraction, a

250

R. SCHOENTAL

hydrogenase has recently been obtained of high-molecular-weight, 2,000,000, which may be an aggregate (Tashiro et al., 1983). Z, which in uiuo affects the genital organs, the hypothalamus and the pituitary, and the production of gonadotropins luteinizing hormone (LH) and follicle-stimulating hormone (FSH) (Hobson et al., 1977), affects also the metabolic pathways of the male hormones in uitro (shown using human hyperplastic prostate) (Thouvenot and Morfin, 1980). The essential role of NAD-coenzymes in the metabolism of Z indicates that the nutritional statws of the individual animal, the availability of B vitamins, etc., as well as the levels of Z may influence the metabolic pathways and the biological effect. It may be expected that, at very high levels, the phenolic hydroxyls of the resorcylic moiety of Z may also exert some less-specific toxic effect. When pigs were given 40 ppm of crystalline Z added to their fodder for 9 days, starting 8 days after parturition, they secreted into the milk Z and its metabolites which caused, in the piglets, swollen and reddened vulvae. Z was present only in trace quantities (0.5-1.0%). The main compounds were p-zearalenol (82-86%) and a-zearalenol(1318%), and these corresponded in the milk at 42-44 hr after feeding to about 1ppm (Hagler et al., 1979). A cow, given 25 ppm of zearalenone in the diet for 7 days, produced milk containing zearalenone, a-zearalenol, and 8-zearalenol which were mostly conjugated and in tutu corresponded to about 1.2 ppm (Mirocha et al., 1981). The proportion of the a-isomer was higher in the cow than in the pig. The urine of man, cow, pig, and rabbit given Z contains mostly conjugated metabolites (Mirocha et al., 1981). Though the estrogenic activity of the conjugated resorcylic acid macrolactones has not yet been reported, it is likely that they would be active and estrogenic when administered orally or ingested, and in the case of infants, which live mainly on milk, such conjugated estrogens may not be quite harmless. When ['*C]Z was given to laying hens (10 mgkg; 1.54 &fig), the radioactivity (94% of the administered dose) was eliminated in the excreta within 72 hr. One-third of the eliminated radioactivity represented unchanged Z, another one-third, polar metabolite(s). Edible muscle retained little activity, but lipophilic metabolite(s) persisted in the fat and in the egg yolk beyond 72 hr after dosing, corresponding to about 2 ppm in the egg yolk (Dailey et al., 1980). The resistance of chicken to Z and its derivatives may be explained by the rapid excretion of these substances as such or as glucuronides. Egg yolk and its lipids have been reported to be carcinogenic in

CARCINOGENIC METABOLITES OF FUSARIUM

251

some strains of mice and to induce lymphosarcomas and tumors of the lung, of the mammary glands, of the brain, of the stomach, etc. (Szepsenwol, 1963,1964).The carcinogenic substances have not been identified, but it has been concluded that more than one carcinogen may be involved (Szepsenwol, 1964). The distribution and the effects of the carcinogenic agents in these hen eggs would be not incompatible with the characteristics of the mycotoxins from Fusarium. The screening of foodstuffs for Z, its derivatives, and other fusarial toxins should become an essential routine so as to give a better indication on the quantitative distribution of these mycotoxins and on their health hazards to livestock and also to human beings. XIII. Estrogenic Agents and Zeranol

A. ANABOLIC ACTIONOF ESTROGENIC AGENTS

Estrogenic agents, notably diethylstilbestrol (DES),have been used in livestock husbandry as anabolics, which are able to stimulate the growth of young animals and to improve food utilization (compare Lu and Rendel, 1976). Z already has been noted to have anabolic action in sheep by Stob et al. (1962).Its reduction products, zearalanone and the two epimeric zearalanols [known as zeranol (the a-epimer) and taleranol (the P-epimer)] which have no double bond at C-1’ and C-2’, are not fluorescent. a-Zeranol has been found to be more effective as an anabolic agent. Under the name of “Ralgro,” it has been extensively used in animal husbandry and has replaced DES, the use of which was prohibited in 1979 by the Food and Drug Administration (FDA) following the recognition of its carcinogenic potentialities to humans when medicinally given during pregnancy (Herbst et al., 1971; see also Herbst and Bern, 1981; Proc. Cong. 1981; Kaufman et al., 1984). However, the claims that the use of estrogens increases meat production and therefore that their benefits are worth the risk are not always supported by experience. In the case when the anabolics are superimposed on an already large intake of estrogenic agents in the diet (e.g., coumestrol from soya meal, genistein from subterranean clover, etc., and/or Z produced by the Fusarium molds) so that their total amount exceeds the level that is optimal for anabolic action, decreased growth follows. This already has been demonstrated 36 years ago by Meites (1949).White rats which were injected daily for 30-75 days with large doses of estrogens (estrone, estradiol, or DES at

252

R. SCHOENTAL

0.1 mg/day in 0.1 ml of corn oil) had lower weight gains and ate less than controls. In the case of DES, a range of doses (from 0.001 to 0.1 mg/day) was used and showed clearly that the higher the dose, the greater was the depression of weight which paralleled the decrease of food consumption. High levels of the estrogenic Z (of the order of 10 mgkg body weight) and of its reduced derivatives are similarly growth inhibiting (Kiessling, 1982), besides having deleterious effects on fertility (and otherwise) in livestock (Reynolds, 1980; Cooper, 1981; Long et al. 1982) and in experimental animals (Christensen et al. 1972; Miller et al. 1973; Report NTP-81-54, 1982). The anabolic DES was reported to induce significantly increased incidence of cryptorchid testes in Sprague-Dawley rats exposed during the 6-18 days of gestation to 80 pgkg body weighdday. The biologically equivalent dosage of zeranol (which is more than 100 times less active) should correspond to at least 8 mgkg body weighdday. When tested at the level of 4 mgkg body weighvday, no significant effects were reported (Wardell et al., 1983). In mice, subcutaneous injections of zeranol (1-10 mgkg body weighdday) during days 10 through 16 of gestation decreased the number of live births and increased medullary trabeculae in bones (Davis et al., 1982). Though only small quantities of anabolics, such as Ralgro, are retained in animal tissues and in their products, the concern they cause is not unjustified. Human beings are exposed to a variety of estrogenic agents from diverse sources (Table IV). The traces of anabolics may sometimes be the proverbial “straw that broke the camel’s back.”

B. METABOLISM OF ZERANOL The distributionmetabolism, and excretion of zeranol (Fig. 10) (Ralg o ) has been studied in females of human and several animal species (rat, rabbit, dog, and monkey), using [ 1‘,2’-3H~lzeranolobtained by catalytic reduction of Z (Migdalof et al., 1983).The radioactivity in blood had a half-life of about 1 day, and the major part of a single oral dose was excreted in the urine and feces within 120 hr. While the urine of rabbits and man contained the larger part of excreted radioactivity in the form of conjugated zeranol and some unidentified polar metabolite(s), in the other species, the feces were the main route of excretion of the unchanged zeranol. Small quantities of radioactivity were detected between 1,2, and 4 hr after a single oral dose in all the rat tissues, the liver showing the highest values.

CARCINOGENIC METABOLITES OF FUSARIUM

253

TABLE IV

EXAMPLES OF THE VARIOUS SOURCES OF ESTROGENIC AGENTSTO WHICHPEOPLEMAY BE EXPOSED ~

~

~~

Natural The physiological steroidal h o p o n e s produced endogenously by the human being (estrone, estradiol-17/3, etc.) Similar steroidal hormones present as constituents of animal products Zearalenone and related resorcylic acid lactones (metabolites of Fusariurn molds) present as sporadic contaminants in foods and drinks made from contaminated agricultural products or derived from livestock that consumed moldy fodder Plant estrogens (e.g., coumestrol in soya beans) Synthetic (of steroidal, stilbene, or other chemical structures) use In oral contraceptives For suppression of lactation As replacement therapy during menopause In cosmetics As anabolics For the treatment of breast cancer For the treatment of prostate cancer For the prevention of threatened abortion and for pregnancy tests has been mostly discontinued

Zeranol appears to be reversibly oxidized to zearalanone; this, in

the course of reduction, may form both the a- and p-epimers. The pepimer, known as taleranol, is less active. In the rabbit, conjugation of these compounds also occurs. Some polar unidentified metabolite(s) detected in human urine may be due to second phase hydroxylation and/or to opening of the macrolactone ring (Migdalof et al., 1983). It is of interest that zeranol was tested in women as replacement estrogen for the alleviation of menopausal disorders (Utian, 1973) as compound P. 1496 or Frideron (compare the Extra Martindale Pharmacopoeia, 1982). Its carcinogenic potentialities, which are likely to resemble those of Z and of other estrogenic agents, have not yet been evaluated. XIV. Estrogenic Agents and the Development of Sex Organ Abnormalities and Tumors

A. IN HUMANBEINGS It has been shown that regardless of their chemically diverse structures and greatly varied specific estrogenic activities, various estrogenic agents, including Z, bind in vivo and in vitro to the same recep-

254

R. SCHOENTAL

tors in the target organs of animal and man (Korenman, 1969; Shutt and Cox, 1972; Greenman et al., 1977,1979; Boyd and Wittliff, 1978; Kiang et at., 1978; Martin et al., 1978; Katzenellenbogen et at., 1979; Blankenship et al., 1982). In primates, the levels of gonadotropins are also affected by Z as they are by estradiol-17P (Hobson et al., 1977). The total load of estrogenic agents to which an individual is exposed determines the effects; the most striking and of greatest importance is exposure to excessive total levels of estrogenic agents when it occurs during the perinatal period when the sex and other target organs are undergoing active development and differentiation. Precocious sexual development, breast enlargement, ovarian cysts, and abnormal bone age have been reported in young children, some only 8 months old, from Puerto Rico during 1978-1982 (Comas, 1982; S a h z de Rodriguez and Toro-Sola, 1982; Bongiovanni, 1983).Anabolics, such as DES or zeranol, were suggested to be possibly involved (SAenz de Rodriguez and Toro-Sola, 1982). Breast enlargement has also been observed in Haitian men, who spend some time at Fort Allen in Puerto Rico, one of the United States Naturalization Services Centers, before their immigration into the United States. Anabolics in their meat supplies have again come under suspicion (Anon, 1982). Yet, the amounts of anabolics carried over into animal products are usually very small, but in these cases, they might have acted concomitantly with Z which might have been present in imported cereals (corn does not grow in Puerto Rico) and/or other imported food (Schoental, 1983a). In the blood of some of the children showing precocious sexual development, Z has been detected by radioimmunoassay (SAenz de Rodriguez, 1984). Precocious breast enlargement among school children, male and female, has occurred previously in Italy (Scaglioni et al., 1978; Fara et al., 1979) and Bahrain (Kimball et al., 1981), but the types of estrogenic agents involved have not been identified. In Chile, precocious puberty in infants has been traced to their treatment for a skin disorder with an ointment containing estrogens (Beas et al., 1969). Though these effects in children appear mostly to subside, the long-term fate of such sexually precocious children remains to be investigated. In the female offspring of women who, during pregnancy, have been treated with large doses of estrogenic agent (mainly DES) in the belief that such treatment will prevent abortion, vaginal abnormalities, adenomatosis, and clear cell adenocarcinoma have been detected after the girls had reached the age of 8 years or more (Herbst et al., 1971; Bibbo et al., 1977; Antunes et al., 1979; Lingeman, 1979; Hoover, 1980; Herbst, 1981; Scully et al., 1981). Among some of the

CARCINOGENIC METABOLITES OF FUSARIUM

255

older age groups, uterine abnormalities, ectopic pregnancy, and frequent abortions have also been reported (Kaufinan et al., 1984). Among the male offspring exposed in a similar way in utero to DES, some developed genital abnormalities, including epididymal cysts, testicular hypoplasia and cryptorchidism, low sperm counts, and abnormal sperm morphology (Gill et al., 1976, 1979, 1981; Bibbo et al., 1977; Henderson et al., 1979, 1982; Stenchever et al., 1981; Whitehead and Leiter, 1981). Cryptorchidism has often been correlated with increased risk of testicular neoplasms (Bolande, 1977). Indeed recently, instances of testicular seminoma in two young men, 28 and 27 years old, have been reported (Conley et al., 1983). It is of particular interest that among the five siblings of one of these young men, two sisters developed vaginal adenosis and adenocarcinoma, respectively, and one brother had oligospermia. Their mother received DES treatment during all her pregnancies (Conley et al., 1983).Increased, more than doubled, incidence of congenital abnormalities in male organs have been reported to have occurred since the 1960s in Great Britain (Chilvers et al., 1984; Matlai and Beral, 1985), in Sweden (Aarskog, 1979; Kallen and Winberg, 1982), in Hungary (Czeizel, 1985), and elsewhere. The incidence of testicular cancer in young men also has been increasing (Clemmesen, 1968; Scottenfeld et al., 1980; Nethersell et al., 1984), especially among the higher social classes (Davies, 1981; Waterhouse, 1985). The rise of the various abnormalities of male genital organs and of male homosexuality and feminization is likely to be related to the presence of a variety of estrogenic agents in food and medicines to which oral contraceptives have been added since the 1960s. As yet, no attention has been paid to a possibility that patients with similar genital abnormalities, but who were not exposed to DES in utero, might have been exposed instead to environmental Z. A lead to such a possibility might be obtained from their birth dates, whether these did correlate with unusually severe weather conditions in the localities of their parents domiciles at the perinatal age. Weather conditions prevailing in any locality could be traced from the Meteorological Offices. The contribution of exogenous estrogenic agents to the incidence of idiopathic breast tumors in women proved difficult to trace (Moore et al., 1983; Vessey, 1984; Grant, 1985). To quote Lipsett (1979), “In contrast to uterine cancer where the base line incidence is low, and small absolute increases are therefore perceptible, the high rate of occurrence of breast cancer will always make detection of small increases difficult.” The difficulties and in-

256

R. SCHOENTAL

consistencies that have been encountered in epidemiological surveys, designed to evaluate the effects of estrogenic agents used for specific purposes (such as for the prevention of abortion, for menopausal disorders, for oral contraceptives, etc.), might have been aggravated because the existence of mycoestrogens has not even been considered. Z would be expected to induce similar effects to those due to DES in males and in females who may have been exposed to its appropriate levels and at the appropriate stages of development, whether in utero or via milk and other foods. Instances of such exposures must sporadically have occurred from times immemorial and might have been responsible for some of the idiopathic abnormalities and neoplasias of the genital tract and of other target organs, as well as for abnormal sexual behavior, in animals and man. The conclusion that among the various estrogenic agents “there are differences in potency, metabolism, and excretion, but there is no difference in the fundamental mode of action” and that “DES is not uniquely carcinogenic, and that any estrogen reaching the responsive tissues in equivalent amounts would exert the same effect” (Lipsett, 1979) applies equally to equiestrogenic amounts of Z. The controversy whether the use of oral contraceptives (OC) may increase the risk of breast cancer (as it does of endometrial, cervical, and possibly of pituitary tumors) is approaching resolution. Careful selection of controls in their case-control studies of the effects of OC (especially those with high content of progestogen compounds) showed an association with increased risk of breast cancer in young women (the RR was approximately 4 after 5 years of OC use) (Pike et al., 1983; Anonymous, 1983; McPherson et al., 1983) (see also Vessey et aZ., 1983; Grant, 1985, in relation to cervical tumors). Some of the causes of the controversies regarding the use of OC have been reviewed by Grant (1985). The problem no doubt would become clearer, if exposure to zearalenone and to the other fusarial mycotoxins could be included in epidemiological considerations of cancer incidence.

B. IN EXPERIMENTAL ANIMALS

Male rodents experimentally exposed to estrogens (steroidal or DES) have long been known to develop various lesions and tumors of the target organs, especially if treated early in life (Lacassagne, 1932; Greene et al., 1939; Dunn and Green, 1963; Dunn, 1979; Arai, 1968; Arai et al., 1978, 1983; McLachlan et al., 1975; McLachlan, 1980, 1981; Rustia and Shubik, 1976; Nomura and Kanzaki, 1977). Z and its estrogenic congeners are no exception. The effects of Z on the male

CARCINOGENIC METABOLITES OF FUSARIUM

257

have been reported already in the first publication of Stob et al. (1962), and since then, testicular abnormalities also have been observed in experimental animals and in livestock treated with the anabolic zeranol. The treated male animals, including cattle, sheep, pigs, chickens, guinea fowl, etc., develop dose-dependent anatomical and behavioral abnormalities (Davis et al., 1977; Ralston, 1978; Wiggins et al., 1979; VBnyi and Szeky, 1980a,b; Mess et al., 1979).Neoplastic and other effects of estrogenic agents on the female sex organs are more widely known (see e.g., Marchand, 1977; Forsberg and Kalland, 1981; Jull, 1976; Schoental, 1974a, 1976~). XV. Carcinogenic Effects of Zearalenone

A. IN F344/N RATS AND B6C3F1 MICE

The results of long-term testing of zearalenone have only appeared in the form of a detailed report from the United States Department of Health and Human Services, NTP-81-54, NIH Publ. No. 83-1791,1982 entitled the “Carcinogenesis Bioassay of Zearalenone (Case No. 17924-92-4) in F344/N rats and B6C3F1 mice.” The mice were given 50 and 100 ppm Z, the rats 25 and 50 ppm Z in the diet for 103 weeks to groups of 50 animals of each sex. The surviving animals were killed when approximately 27 months old. The panel of experts concluded that Z is carcinogenic in mice in which it increased the incidence of pituitary tumors in both sexes and of hepatocellular adenomas in the females, but “under the conditions of this bioassay, Z was not carcinogenic for F344/N rats of either sex.” The F344/N rats developed interstitial cell tumors of the testes in more than 90% of the controls while the two groups of rats treated with the lower and higher dosages of Z developed 86% and 78%, respectively. Testicular atrophy was present in 2% controls, 52% at low and 34% at high levels of Z. The incidence of all primary types of tumors in the rats was in males, 102/50 in the controls, 93/50 for low Z and 87/48 for high Z; in females, 53/32,67/40, and 49/33, respectively. No interpretation of these results seems possible. The F344 rats have been known to develop greatly variable incidences of interstitial cell testicular and other tumors (Sass et al., 1975; Goodman et al., 1979). Such animals are unsuitable for carcinogenicity tests, especially of estrogenic agents. The incidence of tumors of various types among untreated laboratory rodents can vary greatly, even in the same laboratory at various times (Gilbert et al., 1958; Burek, 1976; Goodman et al., 1979; Roe,

258

R. SCHOENTAL

1981, Deerberg et al., 1981). The “spontaneous” tumor incidence, e.g., in mice, was reported to have changed from 10 to 80% within the 10 years, 1961-1971 (Roe, 1981).The striking changes were suggested to have been due to longer survival of the specific pathogenfree (SPF) animals; yet, germ-free F344 rats have very low incidence of “spontaneous” tumors (Sacksteder, 1976). More satisfactory explanation appears to be that various batches of the laboratory animal diet contain variable amounts of carcinogenic contaminants, such as insecticides, chlorinated hydrocarbons, etc., and especially of mycotoxins including the estrogenic zearalenone (Schoental, 1974b, 1979a,b; Coleman and Tardiff, 1979).Exposure to high levels of Z of the parent animals and of the offspring (in utero and/or during the perinatal period) could lead to situations mimicking inherited susceptibility to tumors of the genital and of the other organs, which are targets for estrogenic and toxic agents (Schoental, 1983d). Perinatal exposure to estrogenic agents can also have significant effects upon the developing immune system, as shown by Blair (1981).

B. INWISTARRATS In long-term experiments in which Wistar rats were given in the diet levels of Z (not exceeding 3 mgkg), the incidence of tumors was reported to be similar to that in the controls (historical controls!), but the types and the numbers of the respective tumors have not been specified (Becci et al., 1982; see also Becci and Stevens, 1984). In my own experiments in which white, SPF, Wistar-derived rats (obtained from the M.R.C. Laboratory Animal Centre, Carshalton, Surrey) were exposed perinatally to Z alone or in conjunction with nontoxic doses of T-2 toxin, some of the animals, males and females, that survived for more than a year developed various lesions and tumors, benign and malignant, mainly of the pituitary and of the sex organs. Of particular interest appears an experiment in which young (about 5 weeks old) females were treated daily for 3 days with applications to the skin (from which the fur hair was removed with scissors)of solutions of (A) Z (20 mg/ml in ethyl acetate), (B) the same solution of Z and a solution of T-2 toxin (1mg/ml in aqueous 10% ethanol), and (C) T-2 toxin only (1mg/ml, the same solution as in B). Four weeks after the first application, the rats were mated with untreated males. While the young of the first litter in C survived well, all the young from the first litters of A and B died within a few days after birth and so did most from the second litters. The survival of the young from the third litters in A and in B was good. The father rats

CARCINOGENIC METABOLITES OF FUSAlUUM

259

were then removed and the young rats, when 1 week old, were injected intraperitoneally in A with 0.05 ml of Z in aqueous ethanol corresponding approximately to 10 mgkg body weight, in B with 0.05 ml of Z followed by 0.05 ml T-2 toxin in 10% aqueous ethanol corresponding approximately to 0.5 mgkg body weight, and in C with T-2 only (at 0.5 mgkg body weight). Similar ip injections were repeated at 2 weeks of age, and the animals were left without further treatment. The rats that survived were killed when they were 20-23.5 months old. No significant lesions were found among the rats treated with only T-2 toxin at these low-dose levels. However, most of the rats in groups A and B had significant lesions, more pronounced in group B. In group B, tumors of the pituitaries and of the sex organs were particularly striking. Examples of the various neoplastic changes in the genital organs of the females are illustrated in Figs. 11-23. These examples include various sized endometrial adenomas and adenocarcinomas, most of which were unilateral; some metastasized, spreading among the viscera in the peritoneal cavity and forming small nodules on the lining membranae. Endometritis was common. The ovaries were often cystic. Among the many, often large, mammary tumors, fibroadenomas predominated. Among the males, degenerative and proliferative changes of the genital organs were common. Atrophy of the testis and the epididymis and of the oligospermia is illustrated in Figs. 24 and 25. Adenomatous prostatic glands are illustrated in Fig. 26, and interstitial Leydig cell tumors are illustrated in Figs. 27 and 28. Both males and females often had greatly enlarged pituitaries (Fig. 29) of which many were hemorrhagic, but some were adenomas with areas indicating malignancy (Fig. 30). The experiments indicate the following. 1. Z applied to the skin is absorbed and can affect the outcome of pregnancy from matings that take place after the Z as such must have been eliminated. 2. Low levels of T-2 toxin that were sufficient to cause the characteristic local skin lesions per se do not seem to affect pregnancies from matings that followed the applications of T-2 toxin, nor did they induce significant long-term effects. 3. Two intraperitoneal doses of Z (about 10 mgkg body weight) given within the first two weeks after birth caused significant uterine lesions, mainly endometriosis and endometrial, adenoma among the rats that survived more than 20 months. 4. Two similar intraperitoneal doses of Z (about 10 mgkg body

FIG.11. Genital organs of a female rat, killed when 20.5 months old, showing a large tumor, endometrial adenoma, at the lower part of the left uterine horn. x 1.35. FIG.12.Longitudinal section through the right uterine horn and ovary (a)and through the large endometrial adenocarcinoma on the left horn (b). Hematoxylin and eosin (H and E); x4. FIG.13.Higher magnification of the tumor from Fig. 12,showing glandular papillary proliferations of the endometrial adenocarcinoma. H and E; ~ 5 0 . 260

FIG.14. Cross section of a uterine horn with endometrial adenocarcinoma. H and E; x5. FIG.15. Higher magnification of the adenocarcinoma from Fig. 14, showing endometrial glandular proliferation. H and E; ~ 6 4 . FIG.16.Another area ofthe tumor in Fig. 14, showing the deeply stained proliferating glandular cells embedded in the stroma. H and E; XU. FIG. 17. Higher magnification from the same tumor (Fig. 14), showing variation in size of proliferating cells and mitotic figures. H ,and E; ~640. 261

262

R. SCHOENTAL

FIG.18. Squamous metaplasia with some irregularities, possibly early squamous cell carcinoma of the uterine cervix. R and E; x64.

FIG.19. Another area from the same rat, showing proliferation of endometria1glands and an area of infection. H and E; ~ 6 4 . FIG.20. Genital organs from a rat, showing large endometrial adenocarcinoma. Metastatic nodules are on the peritoneal lining. x2. FIG.21. Section through the liver of this rat, shawing the endometrial adenocarcinomaspreading between the liver Iobes. H andE; ~ 3 .

FIG.22. Microscopic appearance of a very malignant undifferentiateduterine tumor with bone metaplasia that metastasized in the peritoneal cavity, the lung, etc. H and E; X64. FIG.23. Higher magnification of this undifferentiatedendometrial tumor from Fig. 22, showing the variations in cell type and size. H and E; x640. 264

FIG.24. Deformed, atrophic testis and epididymis. H and E; x5. FIG.25. Atrophic tubulus with very few spermatozoa and interstitial hyperplasia (higher magnification fiom Fig. 24). H and E; X64. 265

FIG.26. Adenomatous prostatic glands and intraglandular fibrosis. H and E; ~ 6 4 . FIG.27. Interstitial Leydig cell tumor of the testis. H and E; x64. FIG.28. Higher magnification showing the interstitial cells of the tumor in Fig. 27. H and E; x640. 266

FIG.29. Rat brain showing a large depression at the site from which the pituitary was removed, and the pituitary is shown at the side of the brain. x2.0. FIG.30. Adenoma of a pituitary with areas of malignant transformation. H and E; x 7.5. 267

268

R. SCHOENTAL

weight) had more striking effects when the rats were given, in addition, two intraperitoneal doses of T-2 toxin (0.5 mgkg body weight, following each injection of Z), and this also induced endometrial carcinomas in females. 5. The male rats treated similarly with Z + T-2 toxin develop lesions of genital organs, mostly unilateral testicular atrophy, and Leydig cell tumors. Some also develop adenomatous lesions of the prostatic glands. 6. Both males and females develop enlarged pituitaries, mostly hemorrhagic, but in some of the adenomas, areas of malignancy were present. 7. Some of the male rats exposed early in life to Z + T-2 toxin that developed testicular and pituitary lesions became greatly belligerent (Schoental, 1982a) suggesting that their behavior may be an expression of sexual abnormalities. XVI. Occurrence of Zearalenone and Distribution of Mycotoxins

A.

NATURAL OCCURRENCE OF ZEAEULENONE IN AGRICULTURAL PRODUCTS

The levels of Z in cereals and in other agricultural products can vary greatly, depending on the environmental conditions during their harvesting and during their storage and transport. The weather, rainfall and temperature, and the type of substrate and its moisture content play critical roles (Stoloff, 1976; Bennett and Shotwell, 1979; Sutton et al., 1980). As a result of wet harvest, levels of Z in corn have been reported to be sporadically very high, e.g., up to 141 mgkg in Ontario, Canada (Funnell, 1979) or up to 200 mgkg in Yugoslavia (Muzic et al., 1976). Information regarding the occurrence of Z in grain and in other foodstuffs is fragmentary. Usually animal feeds have been tested when visibly moldy or as a result of outbreaks of adverse effects in livestock, such as sterility or hyperestrogenism in pigs, which are very susceptible to the estrogenic action of Z (Miller et al., 1973; Mirocha et al., 1977). Similar outbreaks elsewhere (see Martin, 1981) and screening studies indicate a worldwide occurrence of this fusarial estrogen (including Africa, Australia, Canada, Finland, France, Hungary, Italy, India, Japan, Poland, Rumania, Serbia, Sweden, Yugoslavia, United Kingdom, United States, and USSR; see also Jemmali et al., 1978; Andrew et al., 1981, and others). From England, only a few data have been published. Z-producing

CARCINOGENIC METABOLITES OF FUSARZUM

269

Fusaria have been detected in barley and other grains after wet harvest (Hacking et al., 1976; Gross and Robb, 1975) and in imported maize, distillers maize, maize gluten, etc. (Buckle, 1983).On one occasion, hay has been found to contain 16 ppm of Z (Mirocha et al., 1968).Z has been detected in stored grain (Stoloff, 1976; Stoloff et al., 1976; Bennett and Shotwell, 1979; Shotwell et al., 1977; Stahr et al., 1980; Enari et al., 198l),in cornflakes (Scott et al., 1978),tomato paste (Bottalico, 1981), etc. The moisture content in freshly harvested grain is usually 15-20% of its weight and would allow the growth of many toxigenic microfungi, such as the Alternariae, Aspergilli, Penicillia, and others. The organisms which are more heat resistant will continue to grow and to produce their toxic metabolites during storage, in the “hot spots.” However, Fusaria, such as F . roseum (F. culmorum) and F . poae, are less thermostable and more hygrophilic. They require a moisture content of about 30-50% for optimal growth. Their survival has been reported not to exceed 3-4 months on grain, especially on dry grain (Pelhgte, 1969). The Fusarium microfungi which are not thermostable may disappear in the course of grain storage or silage (Pelhate, 1977) and/or during the preparation of diets and foods (Buckle, 1983). However, their secondary metabolites (including Z) are more thermostable and remain long after the microfungi can no longer be isolated, e.g., in imported grains (Gilbert et al., 1983; Buckle, 1983). In silage, levels of Z up to 87.3 ppm have been found by Mirocha et al. (1974). Z often occurs in conjunction with DON (Trenholm et al., 1983), also known as vomitoxin or food refusal factor, because pigs refuse to consume feeds containing this trichothecene, which induces vomiting (Vesonder et al., 1981; Vesonder, 1983). Concentrations of zearalenone ranging from 0.1 to 8.0 ppm and of DON from 0.1 to 41.6 ppm have recently been reported in feed samples that contained corn harvested during wet weather in 1981 in Illinois (Cot6 et al., 1984). OF MYCOTOXINS IN MILLING B. DISTRIBUTION FRACTIONS OF CEREALS

Molds usually grow on the surface of grains and, on (dry or wet) milling, would be retained together with their lipophilic mycotoxins, including zearalenone, in the outer fraction, such as the germ and the bran, while the starch itself retains only very small parts of the mycotoxins originally present in the grain. At experimental dry milling of contaminated corn, most of its Z was found in the husks and the oily fractions of corn (Bennett et al., 1976). Similar distribution of Z and of

270

R. SCHOENTAL

T-2 toxin between the various fractions of contaminated corn was found on wet milling (Bennett et aZ., 1978; Bennett and Anderson, 1978; Collins and Rasen, 1981). Bran and other by-products of milling, as well as oilseed meals after extraction of the vegetable oils are used for livestock feeds. If these are derived from moldy starting materials, the mycotoxins and their lipophilic metabolites would be carried over into the livestock's fat and other products, and though usually the amounts may be very small, sporadically their levels could be significant. The contribution of mycotoxins to cardiovascular disorders and to certain tumors that have been correlated with high-fat diets (Kinlen, 1981, 1983) (e.g., of the breast, colorectum, etc.) warrants investigation. Similarly, though increased consumption of roughage may be desirable (Burkitt, 1982), attention must be paid to screening the bran, gluten, and similar preparations sold for human consumption to ensure that they are free from mycotoxins. While in some cases highfiber diets did to some extent protect from experimental colon tumors (Watanabe et al., 1979; Eastwood and Passmore, 1983),other workers sometimes obtained just the opposite effects (Wilson et al., 1977; Cruse et al., 1978). Recently, Jacobs (1983) reported that the addition of 20% wheat bran (soft white, certified food grade by the American Association of Cereal Chemists, St. Paul, Minnesota) to a defined fiber-free diet increased the incidence of colon tumors in Sprague-Dawley rats (given 13weekly sc injections of dimethylhydrazine, 20 mgkg body weight) several times above that found in similarly treated rats maintained on the fiber-free diet. The incidence of tumors in small intestines was not affected (Jacobs, 1983, and references therein). Rats which were given such wheat-bran diets developed hyperplasias of the epithelium in the large intestines (Jacobs and White, 1983). Bran derived from moldy wheat might contain mycotoxins. This raises the question, what to do with cereals and other agricultural products found to contain deleterious mycotoxins? The idea that contaminated products could be used after appropriate dilution with sound ones is definitely not sound, though such mixing may prevent fatal, acute effects. XVII. Methods of Detection and Estimation of Zearalenone and Its Estrogenic Derivatives

The detection and estimation of zearalenone among the many constituents and various mycotoxins, which are usually present in moldy

CARCINOGENIC METABOLITES OF FUSARZUM

27 1

cereals and in other agricultural products and foodstuffs, are difficult problems. Several procedures have been devised for the “cleanup” of crude extracts to make them suitable for TLC and for other physical methods (Shotwell, 1977; Hagan and Tientjen, 1979; Gimeno, 1983; Patterson and Roberts, 1979; Haladay, 1980). The possibility of mistaking the methyl ether of alternariol for the fluorescent Z on TLC has been recognized (Seitz et al., 1975). Gas chromatography as well as high-pressure liquid chromatography have also been used for the estimation of Z. Its dimethoxy- or diand trimethylsilyl derivatives show good sensitivity and specificity, especially when combined with high-resolution mass spectrometry (Mirocha et al., 1974; Holder et al., 1977; Moller and Josefsson, 1978; Scott et al., 1978; Cohen and Lapointe, 1980; Cooper et al., 1981; Trenholm et al., 1980, 1981, 1983).The limit for detection (among six mycotoxins) of Z is 1-2 pgkg (Howell and Taylor, 1981). A biological method that is based on the uterotrophic activity of estrogens (Dorfhan and Dorfman, 1954) and measures the increase of the weight of the uterus of immature rodents is simple and does not require expensive apparatuses, but is not specific for Z and its estrogenic derivatives. Moreover, in the concomitant presence of cytotoxic agents, such as T-2 toxin (Mirocha et al., 1976b), the increase of the uterine weight by estrogenic agents may be abolished, and under such conditions, the biological test can be misleading (Mirocha et al., 1978; Palyusik et al., 1981). Radioimmunoassay has been developed for the estimation of zearalenone and zearalenol in biological fluids such as human blood serum (Thouvenot and Morfin, 1983).The residues of the anabolic zeranol in animaI tissues can also be estimated by radioimmunoassay (Dixon, 1980; Carter et al., 1984).These methods will no doubt facilitate more extensive screening of food for safety from these estrogenic agents than was practicable in the past when the methods were more complicated and required expensive apparatuses. XVIII. Attempts at Detoxication of Fusarial Mycotoxins

Neither washing nor treatment with propionic acid, formaldehyde, or ammonia yield safe products or wholesome silage from mycotoxincontaminated products. Even pretreatment with “Gasol,” which consists of formaldehyde and propionic or other acids and apparently might prevent fungal growth, is not safe; it could actually stimulate Fusaria to greater production of Z if the concentration of the inhibitor should prove insufficient (Bennett et al., 1980; Kallela and Saasta-

272

R. SCHOENTAL

moinen, 1982). Similarly, the production of T-2 toxin by Fusarium acuminatum has been reported to be activated when the fungal growth was only partly inhibited by an insufficient Concentration of sorbic acid, 0.025% (Gareis et al., 1984). Mycotoxin-contaminated products can be used for the making of ethanol, methanol, or gas. The Russians already have recognized 50 years ago that moldy grain contaminated with Fusaria can cause ATA and is unsuitable for human consumption, but could be used for the preparation of alcohol by fermentation (Mayer, 1954, quoting Nazarov, 1934; Joffe, 1960).Yeast (species of Saccharomyces) are relatively resistant to T-2 toxin and affected only by its very high concentration (Schappert and Khachatourians, 1983). T-2 toxin and related trichothecenes are not volatile, nor is Z. Volatile substances, which are responsible for the characteristic “musty” smell of molds, would warrant detailed study. Bennett et aE. (1981)have shown that when moldy corn, which cwtained high levels of Z (8.0 or 33.5 ppm) was fermented, the steam-distilled alcohol was free from Z which remained in the fermentation medium and in the solid residue obtained by filtration. During fermentation starch disappears; hence, the concentrations of Z in the residual fractions increased to almost double of those present originally in the oorn. However, fermented foods or alcoholic beverages that are consumed without distillation retain Z (Lovelace and Nyathi, 1977; Martin and Keen, 1978). The concentration of Z actually may greatly increase during soaking and malting as shown in the case of barley (Haikara, 1983) and sporadically could present health hazards, e.g., to heavy beer drinkers (Enstrom, 1977; Schoental, 1980a, 198lb). It has been suggested that feminization of adult alcoholics may be due to the inhibition by alcohol of testosterone synthesis, as during oxidation of alcohol, the body becomes depleted of NAD coenzymes (Mendelson and Mello, 1979). The possibility that the feminization may be due to Z and the mycoestrogens that sporadically may be present in alcoholic drinks should also seriously be considered (Schoental, 1980a, 1981b; Martin, 1981). XIX. Conclusions

The carcinogenic potentialities of fusarial mycotoxins have yet to be more fully evaluated. Though the available experimental data are scanty, epidemiological and circumstantial evidence strongly indicates that prevention of common human cancers may largely depend

CARCINOGENIC METABOLITES OF FUSARZUM

273

on the decrease or elimination of carcinogenic mycotoxins from the environment and food. In the past, man was aware of the deleterious effects of molds which were included in the prophecies of doom by the Hebrew Seers. Thus, e.g., Amos 4:6-11: I have given you . . . want of bread in all your places. I have witheld the rain from you when there were yet three months to the harvest and I caused it to rain upon one city, and caused it not to rain upon another city: one piece was rained upon, and the piece whereupon it rained not withered. So two or three cities wandered into one city to drink water.. . .I have smitten you with BLASTZNG and MILDEW; the multitude of your gardens and your vineyards and your fig-trees and the olive trees has the palmerworm devoured. I have sent among you the pestilence after the manner of Egypt: your young men have I slain with the sword, and have carried away your horses, and I have made the stink of your camps to come up even into your nostrils . . . I have overthrown some among you, as when God overthrew Sodom and Gomorah. Amos 5:3:The city that went forth a thousand shall have an hundred left, and that which went forth an hundred shall have ten left.

It is remarkable how deep was the prophet’s understanding of the interrelationships between meteorological events and their consequences that may lead to the decimation of the population. Food shortage and famine due to unusual weather conditions and prolonged drought, when followed by rains, cause proliferation of molds and of other pests that destroy vegetation, spread pestilence, cause loss of livestock, and death among young men, who become aggressive, violent, and die fighting. The contaminated environment can lead eventually to the decimation of the human population. In the Bible, the ill effects of molds have been often stressed, and drastic procedures prescribed for their elimination, e.g., from human dwellings (Lev. 14:34-46) and food (Schoental, 1980b, 1984a, 198513). Nearer to our times, large scale disasters caused by mycotoxins occurred in people who were usually undernourished (Schoental, 1983e) in the form of ergotism among populations consuming rye bread contaminated with ergot alkaloids produced by CZaviceps purpurea (Bove, 1970; Matossian, 1983).The fatalities in the USSR during the early 1940s due to “alimentary toxic aleukia” were caused by fusarial mycotoxins (Nesterov, 1948; Mayer, 1953; Joffe, 1978), and those in Japan after World War I1 were caused by the hepatotoxins from Pencillian islandicum in imported “yellow rice” (Uraguchi and Yamazaki, 1978).The effects of such “epidemics” on the incidence of cancer in the surviving population has yet to be evaluated.

274

R. SCHOENTAL

The present generations living in air-conditioned homes and consuming mass-produced food treated with pesticides, antioxidants, and other additives only seldom encounter molds growing on food. They are unaware that mycotoxins which have been present in the agricultural products may persist during their processing in factories and may remain in the ready to eat foods. The evidence that suppression of immunity can follow exposure to large doses of T-2 toxin and/or estrogenic agents (Blair, 1981) may explain the greater efficacy of few large doses of such “natural” carcinogens over that of continuous low dosage, which tends to induce tolerance. The main hazards of mycotoxins appear related to the occasional exposure to very high levels as a result of unusual weather conditions. ACKNOWLEDGMENTS I am indebted to Professor E. Cotchin for hospitality in his department and for the evaluations of some of the lesions. I thank MI. M. Robins and his colleagues for the sections, the MRC Laboratory Animal Centre for the rats, and Ms. S. Spencer for the excellent care of the animals. I also thank Mr. R. F. Legg for the photomicrographs and MIS. D. Brewster for her constant, invaluable help. Partial support of the animal experiments by the MRC is gratefully acknowledged.

REFERENCES

Aarskog, D. (1979).N . Engl. J. Med. 300,75-78. Abrams, L. (1965).J. South Afr. Vet. Med. Assoc. 36,5-13. Agrelo, C. E.,and Schoental, R. (1980).Toricol. Lett. 5, 155-160. b e d , N., and Cook, P. (1969).Br. J. Cancer 23,302-312. Allen, N. K.,Peguri, A., Mirocha, C. J., and Newman, J. A. (1983).Poultry Sci. 62,282289. Andrews, R. I., Thompson, B. K., and Trenholm, H. L. (1981).J.A.O.C.S. 989-991A. Anonymous (1982). Lancet 2,284. Anonymous (1983). Lancet 2,947-948. Antunes, C . M. F.,Stolley, P. O., Rosenheim, N. B., Davies, J. L., Tamasera, J. A., Brown, C., Bumett, L., Rutledge, A., Pokempuer, M., and Garcia, R. (1979). N . Engl. J. Med. 300,9-13. Appelqvist, P. (1972).Acta Chirurg. Scand. (Suppl.)430, 1-62. Arai, Y. (1968).Erperientia 24, 180-181. Arai, Y.,Chen, C.-Y., and Nishizuka, Y. (1978). Gann 69,861-862. Arai, Y., Mori, T., Suzuki, Y.,and Bern, H. A. (1983).Znt. Reu. Cytol. 84,235-268. Arditti, J., and Rodriguez, E. (1982).1. Ethnophannacol. 5,293-302. Asp, N.-G.,Bauer, H., Dahlquist, A., Fredlund, P., and Oste, P. (1979).Nutr. Cancer 1, 70-73. Bailey, D., Cox, G., Morgaraidge, K., and Taylor, J. (1976). Toricol. Appl. Pharmacol. 37,144. Bamburg, J. R., and Strong, F. M. (1971). In “Microbial Toxins, A Comprehensive

CARCINOGENIC METABOLITES OF FUSARZUM

27.5

Treatise, Algal and Fungal Toxins” (S. Kadis, A. Ciegler, and S. J. Ayl, eds.), Vol. 7, pp. 207-292. Academic Press, New York. Bamburg, J. R., Riggs, N. V., and Strong, F. M. (1968).Tetrahedron 24,3329-3336. Beas, F., Vargas, L., Spada, R. P., and Merchak, N. L. (1969).J . Pediatr. 75, 127-130. Becci, P. J., and Stevens, K. R. (1984).J. Appl. Toxicol. 4,222-223. Becci, P. J., Voss, K. A., Hess, F. G., Gallo, M. A., Parent, R. A., and Stevens, K. R. (1982).J . A p p l . Toxicol. 2,247-254. Belt, R. J., Haas, C. D., Joseph, V., Goodwin, W., Moore, D., and Hoogstraten, B. (1979). Cancer Treat. Rep. 63, 1993-1995. Bennett, G. A,, and Anderson, R. A. (1978).J.Agric. Food Chem. 26,1055-1060. Bennett, G. A., and Shotwell, 0. L. (1979).J.Am. Oil Chem. SOC. 56,812-819. Bennett, G. A., Peplinski, A. J., Brekke, 0.L., Jackson, L. K., and Wickser, W. R. (1976). Cereal Chem. 53,299-307. Bennett, G. A., Vandegraft, E. E., Shotwell, 0.L., Watson, S. A., and Bacan, B. J. (1978). Cereal Chem. 55,455-461. Bennett, G. A., Shotwell, 0. L., and Hesseltine, C. W. (1980).]. Am. Oil Chem. SOC. 57, 245-247. Bennett, G. A., Lagoda, A. A., Shotwell, 0. L., and Hesseltine, C. W. (1981).J.Am. Oil Chem. SOC. 58,974-976. Berenbaum, M. C. (1964). Br. Med. Bull. 20,159-164. Berg, J. W. (1977).In “Origins of Human Cancer (H. H. Hiatt, J. D. Watson, and J. A. Winsten, eds.), pp. 15-19. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Bibbo, M., Gill, W. B., Azizi, F., Blough, R., Fang, V. S., Rosenfield, R. L., Schumacher, G. F. B., Sleeper, K.,Sonek, M. G., and Wied, G. L. (1977).Obstet. Gynecol. 49,l8. Bjelke, E. (1974). Scand. J . Gastroenterol. Suppl. 9, 1-235. Blair, P. B. (1981).In “Developmental Effects of Diethylsilbestrol” (A. L. Herbst and H. A. Bern, eds.), pp. 167-178. Thieme, Stuttgart. Blankenship, L. T., Dickey, J. F., and Bodine, A. B. (1982). Theriogenology 17,325331. Bolande, R. P. (1977).In “Genetics of Human Cancer” (J. J. Mulvihill, R. W. Miller, and J. F. Fraumeni, Jr., eds.), pp. 43-75. Raven, New York. Bongiovanni, A. (1983).J. Pediatr. 103,245-246. Boonchuvit, B., Hamilton, P. B., and Burmeister, H. R. (1975). Poultry Sci. 54, 16931696. Booth, C. (1971).“The Genus Fusarium.” Commonwealth Mycolog. Inst. Kew, Surrey. Bottalico, A. (1981).Phytopathol. Mediterr. 20, 129. Bove, F. J. (1970). “The Story of Ergot.” Karger, Basel. Boyd, P. A., and Wittliff, J. L. (1978). Toxicot. Enuiron. Health 4, 1-8. Brian, P. W., Grove, A. W., Hemming, J. F., Lows, D., and Norris, G. L. F. (196l).J.Exp. Bot. 12, 1-12. Brooks, D. W.,Grothaus, P. G., and Mazdiyasiu, H. (1983).J.Am.Chem. SOC. 105,44724473. Brown, J. M. M., and de Wet, P. J. (1962). Onderstepoort J . Vet. Res. 29, 111-135. Buckle, A. E. (1983).Bull. Br. Mycol. SOC. 17, 14. Burek, J. D. (1978). In “Pathology of Aging Rats.” CRC Press, Boca Raton, Florida. Burk, R. F. (1983).Annu. Reo. Nutr. 3,53-70. Burkitt, D. P. (1982).In “Nutrition and Killer Diseases” (J. Rose, ed.), pp. 1-7. Noyes, Park Ridge, New Jersey.

276

R. SCHOENTAL

Burrell, R. J. W. (1962).J. Natl. Cancer Inst. 28,495-514. Burrell, R. J. W.,Roach, W. A., and Shadwell, A. (1966).J. Natl. Cancer Inst. 36,201214. Busby, W. F., Jr., and Wogan G. N. (1984). ACS Monogr. 182,945-1136. Cameron, G. R. (1947).J. Pathol. Bacterial. 59,530-545. Carter, A. P., Dixon, S. M., and Becci, M. H. (1984).J. Vet. P h a m c o l . Therp. 7,17-21. Centerwall, B. S., and Criqui, M. H. (1978). N . Engl. J . Med. 299,285-289. Chang, K.,Kurtz, H. J., and Mirocha, C. J. (1979).A m . ] . Vet. Res. 40, 1260-1267. Chi, M. S., Robison, T. S., Mirocha, C. J., Behrens, J. C., and Shimoda, W. (1978a). Poultry sci. 57, 1234-1238. Chi, M. S., Robison, T. S., Mirocha, C. J., Swanson, S. P., and Shimoda, W. (1978b). Toricol. Appl. Pharmacol. 45,391-402. Chi, M. S., Mirocha, C. I. Weaver, , G. A., and Kurtz, H. J. (1980). Appl. Enuiron. Microbiol. 39, 1026-1030. Chilvers, C., Pike, M. C., Forman, D., Fogelman, K., and Wadsworth, M. E. J. (1984). Loncet 2,330-332. Christensen, C. M. (1975). “Molds, Mushrooms and Mycotoxins.” Univ. of Minnesota Press, Minneapolis. Christensen, C. M., Mirocha, C. J., Nelson, G. H., and Quast, J. F. (1972).Appl. Microbiol. 23,202. Chung, C. W., Trucksess, M. W., Giles, A. L., Jr., and Friedman, L. (1974)~. Assoc. Off. Anal. Chem. 57,1121-1127. Claridge, C. A., and Schmitz, H. (1979).Appl. Enuiron. Microbiol. 37,693-6. Clemmesen, J. A. (1968).Acta Pathol. Microbiol. Scand. 72,348-349. Cohen, H., and Lapointe, M. R. (1980).J . Assoc. Of-Anal. Chem. 63, 642-646. Coleman, W. E., and Tardiff, R.G. (1979).Arch. Enuiron. Contam. Toricol. 8,693-702. Collins, G. J., and Rosen, J. D. (1979).J . Assoc. Off. Anal. Chem. 62, 1274-1280. Collins, G. J., and Rosen, J. D. (1981).J . Food Sci. 46,77-79. Comas, A. P. (1982). Lancet 1, 1299-1300. Conley, G. R.,Sant, G. R.,Veci, A. A., and Mitcheson, D. (1983).J.Am. Med. Assoc. 249, 1325-6. Cook-Modari, P. (1979). Nutr. Cancer 1,51-60. Cooper, R. A. (1981). Brt. Vet. J . 137,513-519. Corbett, T. H., Griswold, D. P., Roberts, B. J., Peckham, J. C., and Schabel, F. M., Jr. (1977). Cancer 40,2660-2680. Cotchin, E. (1977). In “Pathology of the Female Genital Tract” (A. Blaustein, ed.), pp. 26-65. Springer-Verlag, Berlin and New York. Cotchin, E. (1984).J. P a t b l . 142, 101-127. C8te, L.M., Reynolds, J. D., Vesonder, R. F., Buck, W. B., Swanson, S. P., Coffey, R.T., and Brown, D. C. (1984).J. Am. Vet. Med. Assoc. 184, 189-192. Cruse, J. P., Lewin, M. R., and Clarke, G. C. (1978). Lancet 2, 1278-1280. Cundliffe, E., and Davies, J. E. (1977).Antimicrob. Agents Chemother. 11,491-3. CundlifTe, E., Cannon, M., and Davies, J. E. (1974).Proc. Natl. Acad. Sci. U S A . 71,3034. Czeizel, A. (1985). Lancet 1,462-463. Dailey, R. E., Reese, R. E., and Brouwer, E. A. (1980).J. Agric. Food Chem. 28,286291. Danko, G., and Krasznai, G. (1976).Mag. Allatrou. h p j a 31,233-238. Davies, J. M. (1981). Lancet 1,928-932.

CARCINOGENIC METABOLITES OF FUSARIUM

277

Davis, G. J., McLachlan, J. A., and Lucier, G. W. (1977).Toxicol. Appl. Pharmacol. 41, 138 (Abstr.). Davis, G. R. F., Westcott, N. D., Smith, J. D., Neish, G. A., and Schiefer, H. B. (1982). Can. J. Microbiol. 28,259-261. Dawkins, A. W., Grove, J. F., and Tidd, D. K. (1965).Chem. Commun. 2, 27-28. Day, J. B., and Mantle, P. G. (1981).Vet. Rec. 106,463-464. Deerberg, F., Rehm, S., and Pittermann, W. (1981).Vet. Pathol. 18,707-713. De Simone, P. A., Greco, F. A., and Lessner, H. I. (1979).Cancer Treat. Rep. 63,20152017. Diplock, A. T. (1976).Crit. Reo. Toxicol. 4,271-329. Dixon, S . N. (1980).J.Vet. Phamucol. Ther. 3, 177-181. Dobereiner, J., Tokarnia, C. H.,and Canella, C. F. C. (1967).Pesqui. Agropecu. Bras. 2, 489-504. Dorfinan, R. T., and Dorfinan, A. S. (1954).Endocrinology 55,65-69. Dosik, G . M., Barlogie, B., Johnson, D. A., Murphy, W. K., and Drewinko, B. (1978). Cancer Res. 38,3304-3309. Drobotko, V. G . (1945).Annu. Reu. Sou. Med. 2,238-242. Duax, W. L., and Weeks, C. M. (1980). In “Estrogens in the Environment” (J. A. McLachlan, ed.), pp. 11-22. Elsevier, Amsterdam. Dunham, L. J., Sheets, R. H., and Morton, J. F. (1974).J . Nutl. Cancer Znst. 53, 12591269. Dunn, T. B. (1979).Recent Results Cancer Res. 66, 175-192. Dunn, T. B., and Green, A. W. (1963).J . Natl. Cancer Znst. 31,425-455. Eastwood, M. A., and Passmore, R. (1983).Lancet 2,202-206. Ellison, R. A,, and Kotsonis, F. N. (1974).Appl. Microbiol. 27,423-424. Enari, T.-M., Ilus, T., Niku-Paavola, M.-L., Nummi, M., Ylimaki, A., and Koponen, H. (1981).Eur. J . Appl. Microbiol. 11,241-243. Enstrom, J. E. (1977).Br. J . Cancer 35,674-683. Eppley, R. M. (1974).J . Assoc. Ofi.Anal. Chem. 57,618-620. Eppley, R. M. (1975).J. Assoc. Off Anal. Chem. 58,906-908. Eppley, R. M. (1979).J . Am. Oil Chem. SOC. 56,824-829. Eppley, R. M. (1982).J . Assoc. Off Anal. C h m . 65,892-893. Eppley, R. M., and Bailey, W. J. (1973).Science 181,758-760. Eppley, R. M., Mazzola, E. P., Highet, R. J., and Bailey, W. J. (1977).J.Org. Chem. 42, 240-243. Eppley, R.M., Mazzola, E. P., Stack, M. E., and Dreifuss, P. A. (1980).f.Org. Chem. 45, 2522-2523. Eppley, R. M., Trucksess, M. W., Nesheim, S., Thorpe, C. W., Wood, G. E., and Pohland, A. E. (1984).J . Assoc. Off Anal. Chem. 67, 43-45. Evans, I. A. (1976).ACS Monogr. 173,690-700. Faith, R. E., Luster, M. I., and Vos, J. G. (1980).Reo. Biochem. Toricol. 2, 173-211. Falcone, G., and Nickerson, W. J. (1963).J . Bacteriol. 85,754-762. Fara, G . M., Del Corvo, G., Bernuzzi, S., Bigatello, A., DePietro, C., Scaglioni, S., and Chiumello, G. (1979).Lancet 2,295-297. Flury, E., Mauli, R., and Sigg, H. P. (1965).Chem. Commun. 2,26-28. Fontelo, P. A., Beheler, J., Bunner, D. L., and Chu, F. S. (1983).Appl. Enuiron. Microbiol. 45,640-643. Forgacs, J., and Carll, W. T. (1962).Adu. Vet. Sci. 7,273-382. Forsberg, J . G., and Kalland, T. (1981).Cancer Res. 41,721-734.

278

R. SCHOENTAL

Fraumeni, J. F., Jr., and Blot, W. J. (1977).J. Chron. Dis. 30,759-767. Freeman, G. C., and Momson, R. I. (1948).Nature (London) 162,30. Friend, S . C. E., Schiefer, H. B., and Babiuk, L. A. (1983). Vet. Pathol. 20,737-360. Frost, D. V., and Lish, P. M. (1975).Annu. Reu. P h a m c o l . 15,259-284. Funnell, H. S. (1979).Can.J . Comp. Med. 43,243-246. Gareis, M., Bauer, J., and von Montgelas, A. (1984).Appl. Enuiron. Microbiol. 47,416418. Gedek, B., and Bauer, J. (1983). Deu. Food Sci. 4,301-307. Gerrard, T. L., Telford, J. N., and Williams, H. H. (1974).J. Bacteriol. 119, 1057-1060. Ghosal, S., Chakrabarti, D. K., and Basu Chaudhury, K. C. (1976).J . Phann. Sci. 65, 160-161. Ghosal, S., Chakrabarti, D. K., and Chaudhury, K. C. B. (1977).ErperientM 33,574575. Ghosal, S., Bigwas, K., and Chattopadhyay, B. K. (1978).J. P h a m . Sci. 67,1768-1769. Giarelli, L., Silvestri, F., Ferlito, A., Brollo, A., and Clocchiata, L. (1980).Clin. Otohryng. 5,13-22. Gilbert, C., Gillman, J., Loustalot, P., and Lutz, W. (1958).Br. J . Cancer 12,565-593. Gill, W. B., Schumacher, G. F. B., and Bibbo, M. (1976).J.Reprod. Med. 16,147-153. Gill, W. B., Schumacher, G. F. B., Bibbo, M., Straus, F. H., and Schoenberg, H. W. (1979).J . Urol. 122,36-39. Gill, W. B., Schumacher, G. F. B., Hubby, M. M., and Blough, R. R. (1981). Zn “Developmental Effects of DES in Pregnancy” (A. L. Herbst and H. A. Bern, eds.), pp. 103-119. Thieme, Stuttgart. Gimeno, A. (1983).J. Assoc. Off. Anal. Chem. 66,565-569. Girotra, N. N., and Wendler, N. L. (1967).Chem. Znd. 1,493. Girotra, N. N., and Wendler, N. L. (1969).1. Org. Chem. 34,3192-3194. Glover, J. R. (1970).Znt. Med. Surg. 39,5044. Godtfredsen, W. O., Grove, J. F., and Tamm, C. (1967).Helu. Chim. Acta 50,1666. Goodman, D. G., Ward, J. M., Squire, R. A., Chu, K. C., and Linhart, M. S. (1979). Tox~COZ. Appl. Pha-col. 48,237-248. Goodwin, W., Haas, C. D., Fabian, C., Heller-Bettinger, I., and Hoogstraten, B. (1978). Cancer 42,23-26. Grant, E. (1985).“The Bitter Pill.” Elm Tree Books, Hamish Hamilton, London. Greeder, G. A., and Milner, J. A. (1980).Science 209,825-827. Greene, R. R., Bumll, M. W., and Ivy, A. C. (1938).Science 88,130-131. Greenman, D. L., Boyd, P. A., Mehta, R. G., and Wittliff, J. L. (1977).J . Toricol. Enuiron. Health 3,348. Greenman, D. L., Mehta, R. G., and Wittliff, J. L. (1979).J . Toxicol. Enoiron. Health 5, 593-598. Greenway, J. A., and Puls, R. (1976).C a n . / . Comp. Med. 40, 12-15. Gross, J., and Robb, J. (1975).Ann. Appl. Biol. 80,211-216. Grove, J. F., and Mortimer, P. H. (1969).Biochem. P h a m c o l . 18,1473-78. Grunicke, H., Bock, K. W., Becher, H., G h g , V., Schnierds, J., and Puschendorf, B. (1973). Cancer Res. 33,1048-1053. Habermehl, G. G., and Busam, L. (1984).Liebigs Ann. Chem. 1746-1754. Habermehl, G. G., Busam, L., and Stegemann, J. (1984).J.Naturforsch. (C) 39,212-16. Hacking, A., Rosser, W. R., and Dervish, M. T. (1976).Ann. Appl. Biol. 84,7-11. Haenszel, W., and Kurihara, M. (1968).J.Natl. Cancer Znst. 40, 43-68. Hagan, S. N., and Tientjen, W. H. (1975).J . Assoc. Off. Anal. Chem. 58,620-621. Hagler, W. M., and Mirocha, C. J. (1980).Appl. Enuiron. Microbiol. 39,668-669.

CARCINOGENIC METABOLITES OF FUSARlUM

279

Hagler, W. M., Mirocha, C. J., and Pathre, S. V. (1981).App. Enuiron. MicrobioZ. 41, 1049- 1051. Hagler, W. M., Mirocha, C. J., Pathre, S. V., and Behrens, J. C. (1979). AppZ. Enuiron. Microbiol. 37,849-853. Haikara, A. (1983).Proc. Eur. Brew. Conoen. Congr., 19th, pp. 401-408. Hanvig, J., and Scott, P. M. (1971).Appl. Microbiol. 21, 1011-1016. Hayes, M. A., and Schiefer, H. B. (1979). Food Cosmet. Toxicol. 17,614-621. Henderson, B. E., Benton, B., Jing, J., Yu, M. C., and Pike, M. C. (1979). Int.]. Cancer 23,598-601. Henderson, B. E., Ross, R. K., Pike, M. C., and Casagrande, J. T. (1982).Cancer Res. 42, 3232-3239. Herbst, A. L. (1981). I n “Developmental Effects of DES in Pregnancy” (A. L. Herbst and H. A. Bern, eds.), pp. 63-70. Thieme, Stuttgart. Herbst, A. L., and Bern, H. A., eds. (1981). “Developmental Effects of Diethylsilbestrol (DES) in Pregnancy.” Thieme, Stuttgart. Herbst, A. L., and Scully, R. E. (1970). Cancer 25,745-757. Herbst, A. L., Ulfelder, H., and Poskanzer, D. C. (1971).N . Engl. J . Med. 284,878-881. Hewer, T., Rose, E., and Ghadirian, P. (1978).Lancet 2,494-496. Hiatt, H. H., Watson, J. C., and Winsten, J. A., eds. (1977). “Origin of Human Cancer.” Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Hidy, P. H., Baldwin, R. S., Greasham, R. L., Keith, C. L., and McMullen, J. R. (1977). Ado. Appl. Microbiol. 22,59-82. Highman, B.,Roth, S. I., and Greenman, D. L. (1981).J.Natl. Cancer Inst. 67,653-662. Hintikka, E.-L. (1983). Ado. Food Sci. 4,221-228. Hirayama, T. (1979). Nutr. Cancer 1(3), 67-81. Hirono, I., Yamada, K., Niwa, H., Shizuri, Y., Ojika, M., Hosaka, S., Yamaji, T., Wakamatsu, K., Kigoshi, H., Niiyama, K., and Uosaki, Y. (1984). Cancer Lett. 21, 239246. Hobson, W.,Bailey, J., and Fuller, G. B. (1977).I . Toxicol. Enoiron. Health 3,43-57. Holaday, C. E. (1980).J . Am. Oil Chem. Soc. 57,491A-492A. Holder, C. M., Nony, C. L., and Bowman, M. C. (1977).J. Assoc. Of- Anal. Chem. 60, 272-278. Hoover, R. (1980).In “Estrogens in the Environment” (J,McLachlan, ed.), pp. 353-376. Elsevier, Amsterdam. Hopkins, G. L., and West, C. E. (1976). Life Sci. 19, 1103-1116. Howell, M. V., and Taylor, P. W. (1981).J. Assoc. Off. Anal. Chem. 64,1356-1363. Hsu, I. C., Smalley, E. B., Strong, F. M., and Ribelin, W. E. (1972). Appl. Enoiron. Microbiol. 24, 684-690. Hurd, R. N. (1977).In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 379-391. Pathotox Publ., Park Forest South, Illinois. Hurd, R. N., and Shah, D. H. (1973).J . Med. Chem. 16,543-545. Hurst, E. E., Jr. (1964). Cancer 17, 1187-1196. Hutchinson, A. (1976). Bot. J. Linn. SOC.73, 145-150. lchinoe, M., Kurata, H., and Ueno, Y. (1972).J. Food Hyg. Soc. Jpn. 13,481. Ikediobi, C. O., Hsu, I. C., Bamburg, J. R., and Strong, F. M. (1971).Anal. Biochem. 43, 327-340. IARC (International Agency for Research on Cancer) (1979). IARC Monogr. 21. IARC (International Agency for Research on Cancer) (1983a).ZARC Monogr. 31,265278.

280

’R: ‘SCHOEN+AL

IARC (International Agency for Research on Cancer) (1983b). IARC Monogr. 31,279291. Jacobi, J., Lloyd, H., and Meares, J. (1975).H o w . Metab. Res. 7,228-230. Jacobs, L. R. (1983). Cancer Res. 43,4057-4061. Jacobs, L. R., and White, F. A. (1983).Am. J. Clin. Nutr. 37, 945-953. Jacobs, M. M., and Griffin, A. C. (1981). In “Inhibition of Tumor Induction and Development” (M. S. Zedeck and M. Lipkin, eds.), pp. 169-188. Plenum, New York. Jagadeesan, V., Rukmini, C., Vijayaraghovan, M., and Tulpule, P. G. (1982).Food Cosmet. TOXicOl. 20, 83-87. Jansson, B., Seibert, G. B., and Speer, J. F. (1975).Cancer 36,2373-2384. Jarvis, B. B., Stahly, G. P., Pavanasasivam, G., and Mazzola, E. P. (1980).J. Med. Chem. 23,1054-1058. Jarvis, B. B., Midiwo, J. O., Tuthill, D., and Bean, G. A. (1981). Science 214,460-462. Jarvis, B. B., Eppley, R. M., and Mazzola, E. P. (1983).Ado. Food Sci. 4,20-38. Jemmali, M., Ueno, Y.,Ishii, K., Frayssinet, C., and Etienne, M. (1978).ExperientM 34, 1333-1334. Jofen, J. (1972).Proc. World Congr.Jewish Studies 5th (Jerusalem), Val. 2, pp. 55-71. Joffe, A. Z. (1978).In “Mycotoxic Fungi, Mycotoxins, Mycotoxicoses. An Encyclopedic Handbook” (T. D. Wyllie and L. G. Morehouse, eds.), Vol. 3, pp. 21-86. Dekker, New York. Joffe, A. Z. (1960). Bull. Res. Connun. Isr. 80,81-95. Joffe, A. Z., and Palti, J. (1974).Mycoputhol. Mycol. Appl. 52,209-218. Jull, J. W. (1976).ACS Monogr. 173,52-82. Kadis, S., Ciegler, A., and Ajl, S. J., eds. (1971).“Microbial Toxins, A Comprehensive Treatise,” Vol. 7. Academic Press, New York. Kallela, K., and Saastamoinen, I. (1982).Nord. Vet. Med. 34, 124-129. Kallen, B., and Winberg, J. (1982).Acta Paediatr. Scand. Suppl. 293, 1-21. Kamimura, H., Nishijima, M.,Yasuda, K., Saito, K., Ibe, A., Nagayama, T., Ushiyama, H., and Naoi, Y. (1981).J. Assoc. Ofl.Anal. Chem. 64, 1067-1073. Katzenellenbogen, B. S., Katzenellenbogen, J. A., and Mordecai, D. (1979).Endocrinology 105,33-42. Kaufinan, R. H., Noller, K., Adam, E., Irwin, J., Gray, M., Jefferies, J. A,, and Hilton, J. (1984). Am. J . Obstet. Gynecol. 148,973-982. Kelsey, J. L. (1978).Am. J . Epiderniol. 107,236-244. Kiang, D. T., Kennedy, B. J., Pathre, S. V., and Mirocha, C. J. (1978).Cancer Res. 38, 3611-3615. Kiessling, K.-H. (1982).Acta Pharmacol. Toxicol. 51, 154-158. Kiessling, K. H., and Pettersson, H. (1978). Actu Pharmucol. Toxicol. 43,285-290. Kimball, A. M., Hamadeh, R., Mahmood, R.A. H., Khalfan, S., Muhsin, A., Ghabrial, F., and Armenian, H. K. (1981).Lancet 1,671-672. Kingsbury, J. M. (1964). “The Poisonous Plants of the United States and Canada.” Prentice-Hall, New York. Kinlen, L. (1981). In ‘‘Dietary Influences on Cancer: Traditional and Modern” (R. Schoental and T. A. Connors, eds.), pp. 91-107. CRC Press, Boca Raton, Florida. Kinlen, L. J. (1983). Br. Med. J . 286, 1081-1082. b e t , S., and Mahboubi, E. (1972).Science 175,846-853. Koljonen, T. (1978).Ambio 7, 169-171. Korenman, S. G. (1969). Steroids 13, 163. Kotik, A. N., Chernoba, J. V. T., Komissarenko, N. F., and Trufanova, V. A. (1979). Mikrobiol. Zh. 41,636.

CARCINOGENIC METABOLITES OF FUSARIUM

28 1

Kupchan, S. M., Jarvis, B. B., Dailey, R. G., Jr., Bright, W., Bryan, R. F., and Shizuri, Y. (1976).J. Am. Chem. SOC. 98,7092-7093. Lacassagne, M. A. (1932).C. R. Acad. Sci. 195,630-632. Lafarge-Frayssinet, C., Lespinats, G., Loisillier, F., Mousset, S., Lafont, P., Rosenstein, Y., and Frayssinet, C. (1979).Proc. SOC. E r p . Biol. Med. 160,302-311. Lafarge-Frayssinet, C., Decloitre, F., Mousset, S., Martin, M., and Frayssinet, C. (1981). Mutat. Res. 88, 115-123. Lafont, P., Girard, T., Payen, J., Sarfati, J., and Gaillardine, M. (1983).Microbiol. Alim. Nutr. 1, 147-152. Larsson, L., Sandstrom, A., and Westling, P. (1975).Cancer Res. 35,3308-3316. Lee, K. Y.,Poole, C. F., and Zlatkis, A. (1980).Anal. Chem. 52,837-842. Lee, S., and Chu, F. S. (1981a).J. Assoc. Off- Anal. Chem. 64, 156161. Lee, S., and Chu, F. S. (1981b).J. Assoc. Of$ Anal. Chem. 64,684-688. Levander, 0.A. (1982).Ann. N. Y.Acad. Sci. 393,70-82. Levander, 0.A. (1983).Fed. Proc. Fed. Am. SOC. E r p . Biol. 42,1721-1725. Li Mingxin (Li Min-Hsin), Li Ping, and Li Baorong (Li Pao-Jung) (1980).Adu. Cancer Res. 33,173-249. Liao, L. L., Grollman, A. P., and Horowitz, S. B. (1976).Biochim. Biophys Actn 454,

273-284.

Lin, P., and Tang, W. (1980).J. Cancer Res. Clin. Oncol. 96, 121-130. Lindenfelsen, L. A., Lillehoj, E. B., and Burmeister, H. R. (1974).J. Natl. Cancer Inst.

52,113-116.

Lingeman, C. H. (1979).Recent Results Cancer Res. 66,l-196. Lipsett, M. B. (1979).Cancer 43, 1967-1981. Little, K. (1973).“Bone Behaviour.” Academic Press, New York. Long, G. G., Diekman, M.,Tuite, J. F., Shannon, G. M., and Vesonder, R. F. (1982).Am. J . Vet. Res. 43, 1599-1603. Lovelace, C. E., and Nyathi, C. B. (1977).J. Sci. Food Agric. 28,288-292. Lu, F. C., and Rendel, Y., eds. (1976).“Anabolic Agents in Animal Production.” Thieme, Stuttgart. Lutsky, I., Mor, N., Yagen, B., and Joffe, A. Z. (1978).Toricol. Appl. Pharmacol. 43,

111-124.

Lyass, L. S. (1955).Vopr. Oncol. 6,79. McKeown, T., and Record, R. G. (1951).Lancet 1, 192-196. McLachlan, J. A., ed. (1980).“Estrogens in the Environment.” Elsevier, Amsterdam. McLachlan, J. A. (1981).In “Developmental Effects of Diethylstilbestrol (DES) in Pregnancy” (A. L. Herbst and H. A. Bern, eds.), pp. 148-157. Thieme, Stuttgart. McLachlan, J. A., Newbold, R. R., and Bullock, B. (1975).Science 190,991-992. McPherson, K.,Neil, A., Vessey, M. P., and Doll, R. (1983).Lancet 2, 1414-1415. Madaus, G., and Koch, F. E. (1941).Z. E r p . Med. 109,68-87. Mann, D.D., Buening, G. M., Hook, B., and Osweiler, G. D. (1982).Vet. Immunol. Immunopathol. 3,411-417. Mann, D. D., Buening, G. M., Hook, B., and Osweiler, G. D. (1983).Am. J . Vet. Res. 44,

1757-59.

Marasas, W. F. 0. (1982).In “Cancer ofthe Oesophagus” (C. J. Pfeiffer, ed.), pp. 29-40. CRC Press, Boca Raton, Florida. Marasas, W.F. O., Bamburg, J. R., Smalley, E. B., Strong, F. M., Ragland, W. L., and Degurse, P. E. (1969).Toricol. Appl. Phannacol. 15,471-482. Marasas, W. F. O., Smalley, E. B., Bamburg, J. R., and Strong, P. M. (1971).Phytopatho b g y 61, 1488-1491.

282

R. SCHOENTAL

Marasas, W. F. O., van Rensburg, S. J.. and Mirocha, C. J. (1979).Agric. Food Chem. 27, 1108-11 12. Marasas, W. F. O., Kriek, N. P. J., Fincham, J. E.,and van Rensburg, S. J. (1984).Int. J . Cancer 34,383-387. Marchant, J. (1977).In “Pathology of the Female Genital Tract.” (A. Blaustein, ed.), pp. 1-25. Springer-Verlag, Berlin and New York. Martin, P. M. D. (1981).In “Dietary Influences on Cancer: Traditional and Modern” (R. Schoental and T. A. Connors, eds.), pp. 175-200. CRC Press, Boca Raton, Florida. Martin, P. M. D., and Keen, P. (1978).Subouruudia 16,15-22. Martin, P. M., Horwik, K. B., Ryan, D. S., and McGuire, W. L. (1978).Endocrinology 103,1860-1867. Martindale Pharmacopoeia, The Extra (1982).28th Ed., p. 1770. Pharmaceutical Press, London. Martinez, I. (1969).J. Natl. Cancer Inst. 42,1069-1094. Masamune, S., Kamata, S., and Schilling. (1975).J . Am. Chem. SOC. 27,3515. Matlai, P., and B e d , V. (1985).Lancet 1,108. Matsumoto, H., Ito, T., and Ueno, Y. (1978).Jpn.J . E r p . Med. 48,393-399. Matossian, M. K. (1983).Med. Hist. J . 18,33-42. Mayer, C. F. (1953).Mil. Surg. 113, 173-189,295-315. Meites, Y. (1949).Am. J. Physiol. 159,281-286. Mendelson, J. H., and Mello, N. K. (1979).N. Engl. J. Med. 301,912-921. Mess, B., Ruzsas, C., Woller, L., and Biro-Gosztonyi, M. (1979).Neuroendocrinol. Lett. 1, 1-5. Meyer, H., and Frank, H. K. (1980).Chem.Abstr. 92,209649 (Bibliographyof Fusarium toxins: A review with 2978 refs.) Miessner, H., and Schoop, G. (1929).Dtsch. Tieriirzt. Wochenschr. 37, 167-170. Migdalof, B. H., Dugger, H. A., Heider, J. G., Cooms, R. A., and Terry, M. K. (1983). Xenobiotics 13,209-222. Miller, J. K., Hacking, A., Harrison, J., and Gross, V. J. (1973).Vet. Rec. 93,555-559. Mills, J. L., Stolley, P. V., Davies, J., and Moshang, M. (1981).Am. J . Dis. Child. 135, 743-745. Minato, H., Katayama, T., and Tori, K. (1975).Tetruhydron Lett. 2579-2582. Mirocha, C. J. (1983).Deu. Food Sci. 4, 177-194. Mirocha, C. J., and Christensen, C. M. (1974).In “Mycotoxins” (I. F. H. Purchase, ed.), pp. 129-48. Elsevier, Amsterdam. Mirocha, C. J., and Pathre, S. (1973).Appl. Microbiol. 26, 719-724. Mirocha, C. J., Harrison, J., Nichols, A. A., and McClintock, M. (1968).Appl. Microbiol. 16,797-798. Mirocha, C. J., Christensen, C. M., and Nelson, G. H. (1971).In “Microbial Toxins, A. Comprehensive Treatise” (S. Kadis, A. Ciegler, and S. J. Ajl, eds.), Vol. 7, pp. 107138. Academic Press, New York. Mirocha, C. J., Schauerhamer, B., and Pathre, S. V. (1974)J.Assoc. Off. Anal. Chem. 57, 1104-1110. Mirocha, C. J., Pathre, S. V., Schauerhamer, B., and Christensen, C. M. (1976a).Appl. Environ. Microbiol. 32, 553-556. Mirocha, C. J., Pathre, S. V.,and Behrens, J. (1976b).J.Assoc. Of$ Anal. Chem. 59,221. Mirocha, C. J., Pathre, S. V., and Christensen, C. M. (1977).In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 345-364. Pathotox Publ. Park Forest South, Illinois.

CARCINOGENIC METABOLITES OF FUSARIUM

283

Mirocha, C. J., Pathre, S. V., Behrens, J., and Schauerhamer, B. (1978).Appl. Enoiron. Microbiol. 35,986-987. Mirocha, C. J., Schauerhamer, B., Christensen, C. M., Niku-Paavola, M. L., and Nummi, M. (1979).Appl. Enoiron. Microbiol. 38,749-750. Mirocha, C. J., Pathre, S. V . ,and Robison, T. S. (1981).Food Cosmet. Toricol. 19,25-30. Mdler, T. E., and Josefsson, E. (1978).J . Assoc. Off. Anal. Chem. 61,789-792. Moreau, C . (1979). “Moulds, Toxins and F o o d (M. 0. Moss, ed. and trans.). Wiley, New York. Moore, D. H., Moore, D. H., 11, and Moore, C. T. (1983).Ado. Cancer Res. 40,189-253. Motley, H. L., Ellis, M. M., and Ellis, M. D. (1937).J . Am. Med. Assoc. 109, 1718. Murphy, W. K., Burgess, M. A., Valdivieso, M., and Bodey, G. P. (1978). Proc. Am. Assoc. Cancer Res. 19,411 (Abstr. C-419). Muzic, S., Bogdanic, C., and Balzer, I. (1976). Chem. Abstr. 85 (76529). Naik, D. M., Bush, L. V., and Barron, G . L. (1978). Can.]. Plant Sci. 58, 1095-1097. Naoi, Y. (1983). Deu. Food Sci. 4, 121-124. Napalkov, N. P., Tserkovny, G. F., and Merabishvili, V. M. (1982). Cancer Incidence U.S.S.R. IARC Sci. Publ. No. 48,2nd Ed. Nazerian, K. (1973).Ado. Cancer Res. 17,279-315. Nelson, P. E., Toussoun, T. A., and Cook, R. J., eds. (1981). “Fusarium, Diseases, Biology and Taxonomy.” Pennsylvania State Univ. Press, University Park, Pennsylvania. Nesterov, V . C. (1948).KZin. Med. 26(7),34-41. Nesterov, A. I. (1961).Arthritis Rheum. 7,29-40. Nethersell, A. B. W., Drake, L. K., and Sikora, K. (1984).Br. J . Cancer 50,377-380. Nomura, T., and Kanzaki, T. (1977). Cancer Res. 37, 1099-1104. Nora, H., and Nora, J. J. (1975).Arch. Enoiron. Health 30, 17-21. Nye, S. M., and Paterson, P. J. (1975).In “Trace Substances in Environmental Health IX.A Symposium” (D. D. Hemphill, ed.), pp. 113-121. Univ. of Missouri Press, Columbia. Oettle, A. G. (1964).J . Natl. Cancer Inst. 33,383-439. Ohta, M., Ishii, K.,and Ueno, Y. (1977).J . Biochem. (Tokyo) 82, 1591-1598. Ohtsubo, K., and Saito, M. (1977).In “Mycotoxins in Human and Animal Health” (J. V . Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 255-262. Pathotox, Park Forest South, Illinois. Oldham, J. W., Allred, L. E., Milo, G. E., Kindig, O., and Capen C. C. (1980). Toxicol. Appl. Pharnacol. 52, 159-168. Olifson, L. E. (1957).D. J. Mendeleyeu Chem. SOC. 7,37-46. Olsen, M., and Kiessling, K.-H. (1983).Acta Pharnacol. Toricol. 52,287-291. O’Neill, C . H., Hodges, G. M., Riddle, P. N., Jordan, P. W., Newman, R. H., and Flood, R.J. (1980). Int. J . Cancer 26,617-628. Otokawa, M.,Shibahara, Y., and Egashira, Y. (1979).Jpn.J . Med. Sci. Biol. 32, 37. Palti, J. (1978).Acta Phytomed. J . Phytopathol. Suppl. 6. Palyusik, M., Harrach, B., Mirocha, C. J., and Pathre, S. V. (1980).Acta Vet. Acad. Sci. Hung. 28,217-222. Palyusik, M., Harrach, B., Horvath, G., Gaal, T., and Mirocha, C. J. (1981).Mag. Allatom. Lapja 36,799-803. Pareles, S. R., Collins, G. J., and Rosen, J. D. (1976).J.Agric. Food Chem. 24,872-875. Pathre, S . V., and Mirocha, C . J. (1976).Ado. Chem. Ser. 149, 178-227. Pathre, S. V., and Mirocha, C. J. (1977).In “Mycotoxins in Human and Animal Health”

284

R. SCHOENTAL

(J. V.Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 229-253. Pathotox Publ. Park Forest South, lllinois. Pathre, S. V.,and Mirocha, C. J. (1978).Appl. Environ, Microbiol. 35,992-994. Pathre, S. V., Fenton, S. W., and Mirocha, C. J. (1980).]. Agric. Food Chem. 28,421. Patterson, D. S., and Roberts, B. A. (1979).J. Assoc. Off.Anal. Chem. 62, 1265-1267. Payen, J., Girard, T., Gaillardin, M., and Lafont, P. (1983). Microbiol. Alim. Nuts. 1, 143-146. Pelhgte, J. (1969). Phytopathol. Z. 64,7-20. Pelhgte, J. (1977).Folk Vet. Lat. 7 , 1-16. Pepeljnjak, S. (1983).Ado. Food Sci. 4,265-272. Peters, C. A., and Hurd, R. N. (1975).J.Med. Chem. 18,215-217. Petrie, L., Robb, J., and Stewart, A. F. (1977). Vet. Rec. 101,326. Pfeiffer, C. J., ed. (1982).“Cancer of the Esophagus.” CRC Press, Boca Raton, Florida. Pike, M. C., Henderson, B. E., Krailo, M. D., Duke, A., and Roy, S. (1983).Lancet 2, 926-930. Plowright, W., Linsell, C. A., and Peers, F. 6 .(1971).Br. J . Cancer 25,72-80. Polzhoefer, K., and Niehuss, M. (1980).2. Lebensmittelforsch. 170, 124-128. Prior, M. G. (1979).Can.J . Comp. Med. 43,352-355. Proc. Marabou Symp. Food Cancer (1979). Nutr. Cancer 1(2),23-106. Proc. Nat. Conf. Cynecol. Cancer (1981).Cancer Suppl. 48,425. Puls, R.,and Greenway, J. A. (1976).Can.J. Comp. Med. 40, 16-19. Purchase, I. F. H., ed. (1974).“Mycotoxins.” Elsevier, Amsterdam. Ralston, A. T. (1978).J. Anim. Sci. 47,1203-1206. Rechcigl, M., Jr., ed. (1983). “CRC Handbook of Foodborne Diseases of Biological Origin.” CRC Press, Boca Raton, Florida. Reddy, B. S., and Ohmari, T. (1981).Cancer Res. 41,1363-1367. Report NTP-81-54 (1982). In “Carcinogenesis Bioassay of Zearalenone (Case No. 17924-924).F.344/N Rats and B6C3F1 Mice”(N1H Publ. No. 83-1791). U.S.Dept. Health & Human Services, Research Triangle Park, N.C. 27709. Reynolds, 1. P. (1980).Vet. Rec. 107,367-369. Rice, J. M., ed. (1979).Natl. Cancer Inst. Monogr. 51. Richard, J. L., Cyzewski, S. J., Pier, A. C., and Booth, G. D. (1978).A m . ] . Vet. Res. 39, 1674- 1679. Richardson, K. E., Hagler, W. M. J., and Hamilton, P. B. (1984).Appl. Enuiron. Microbid. 47,643-646. Robbana-Barnat, S., Lafarge-Frayssinet,C., and Frayssinet, C. (1982).Proc. Int. IUPAC Symp. Mycotorins Phycotoxins, 5th, l a r d Sept. pp. 269-272. Roberts, B. A., and Patterson, D. S. P. (1975).J. Assoc. Off. Anal. Chem. SS, 11781181. Robinson, W. 0. (1936).Ind. Eng. Chem. 28,736-738. Robison, T. S., Mirocha, C. J., Kurtz, H. J., Behrens, J. C., Weaver, G. A., and Chi! M. S. (1979).J . Agric. Food Chem. 27,1411-1413. Rodricks, J . V., ed. (1976). Adu. Chem. Ser. 149. Ridricks, J. V., Hesseltine, C. W., and Mehlman, M. A., eds. (1977). “Mycotoxins in Human and Animal Health.” Pathotox Publ. Park Forest South, Illinois. Roe, F. J. (1976).In ”Anabolic Agents in Animal Production” (F. C. Lu and J. Rendel, eds.), pp. 227-237. Thieme, Stuttgart. Roe, F. J. C. (1981).Proc. Nutr. SOC. 40,57-65. Romer, T. R., Boling, T.M., and MacDonald, J. L. (1978).]. Assoc. Off Anal. Chem. 61, 801-808.

CARCINOGENIC METABOLITES OF FUSARZUM

285

Rosenfeld, I., and Beath, 0. A. (1964). “Selenium, Geobotany, Biochemistry, Toxicity and Nutrition.” Academic Press, New York. Rosenstein, Y., and Lafarge-Frayssinet, C. (1983). Toxicol. Appl. Pharmacol. 70, 283288. Rosenstein, Y.,Lafarge-Frayssinet, C., Lespinats, G., Loisillier, F., Lafont, P., and Frayssinet, C. (1979). Zmmunology 36, 111-117. Rosenstein, Y., Kretschmer, R. R., and Lafarge-Frayssinet, C. (1981). Immunology 44,

555-560. Rothenbacher, H., Wiggins, J. P., and Wilson, L. L. (1975).Am. J . Vet. Res. 36, 13131316. Rubinstein, Y. I., Kukel, Yu P., and Kudinova, G. P. (1967). Vopr. Pitan 26, 57-61. Rubinstein, Y. I. (1949). Hyg. Sanitation 12,35-39. Ruddick, J. A., Scott, P. M., and Harwig, J. (1976). Bull. Enuiron. Contam. Toricol. 15, 678-681. Rukmini, C., and Bhat, R. V . (1978).J . Agric. Food Chem. 26,647-649. Rustia, M.,and Shubik, P. (1976). Cancer Lett. 1, 139-146. Sacksteder, M. R. (1976).J. Natl. Cancer inst. 57, 1371-1373. Shenz de Rodriguez, C. A. (1984).N. Engl. J. Med. 310,1741-1742. Shenz de Rodriguez, C. A., and Toro-Sola, M. A. (1982). Lancet 1, 1300. Saito, M., Enomoto, M., and Tatsuno, T. (1969). Gann 60,599-603. Saito, M., Horiuchi, T., Ohtsubo, K., Hatanaka, Y., and Ueno, Y. (1980).Jpn.J. E x p . Med. 50,293-302. Salonen, J. T., Alfthan, G., Huttunen, J. K., Pikkarainen, J., and Puska, P. (1982).Lancet 2,175-179. Sano, A., Asabe, Y., Takitani, S., and Ueno, Y. (1982).J . Chromatogr. 235,257-265. Sarkisov, A. K. (1954).J . Microbiol. Epidemiol. Immunobiol. 1,43-47. Sass, B., Rabstein, L. S., Madison, R., Nims, R. M., Peters, R. L., and Kelloff, G. J. (1975). J . Natl. Cancer Inst. 54,1449-1456. Sato, N., and Ueno, Y. (1977). In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 295-307. Pathotox Publ., Park Forest South, Illinois. Sato, N., Ueno, Y., and Enomoto, M. (1975).Jpn.J. Pharmacol. 25,263-270. Scalioni, S., DiPietro, C., Bigatello, A., and Chinmeell, G. (1978). Lancet 1, 551-552. Schappert, K.T., and Khachatourians, G. G. (1983).Appl. Enuiron. Microbiol. 45,862867. Schardein, J. L. (1980). Teratology 22,251-270. Schmidt, R., Ziegenhagen, E., and Dose, K. (1981).J . Chromatogr. 212,370-373. Schoental, R. (1974a). In “Multiple Primary Malignant Tumours” (L. Severi, ed.), pp. 1003-1011. Div. Cancer Res., Instit. Anat. Pathol. Histol., Perugia. Schoental, R. (1974b). Cancer Res. 34,2419-2420. Schoental, R. (1976a). Br. J. Cancer 33,668-669. Schoental, R. (1976b). FEBS Lett. 61,111-114. Schoental, R. (1976~).ACS Munugr. 173,626-689. Schoental, R. (1979a).Nutr. Cancer 1, 13-14. Schoental, R. (1979b). Br. J . Cancer 39,101. Schoental, R. (1979~).Front. Gastrointest. Res. 4, 17-24. Schoental, R. (1980a).Nutr. Cancer 2,88-92. Schoental, R. (1980b). Preuent. Med. 9,159-161. Schoental, R. (1980~).Lancet 2,593. Schoental, R. (1980d). Znt. J . Enuiron. Studies 15,65.

286

R. SCHOENTAL

Schoental, R. (198Oe). Biochem. SOC. Trans. 8,147-150. Schoental, R. (1981a). In “Dietary Influences on Cancer, Traditional and Modem” (R. Schoental and T. A. Connors, eds.), pp. 109-148. CRC Press, Boca Raton, Florida. Schoental, R. (1981b).In “Dietary Influences on Cancer, Traditional and Modem (R. Schoental and T. A. Connors, eds.), pp. 149-173. CRC Press, Boca Raton, Florida. Schoental, R. (1981~).Nutr. Cancer 3,57-62. Schoental, R. (1981d). Int. J. Enuiron. Studies 18,67. Schoental, R. (1981e).J. Vet. Pharmucol. Ther. 4, 1-6. Schoental, R. (1982a). In ‘‘Behavioral Models and the Analysis of Drug Action” (M. Y. Spiegelstein and A. Levy, eds.), pp. 475-478. Elsevier, Amsterdam. Schoental, R. (1982b). Proc. Meet. Mycotoxins Anim. Dis., 4th, MAFF, London, pp. 3-4. Schoental, R. (1983a).Lancet 1,537. Schoental, R. (1983b). Microbiol. Aliment. Nutr. 1, 101-107. Schoental, R. (1983~).Int. J. Enuiron. Studies 21, 159-164. Schoental, R. (1983d). Br. J. Cancer 47,569-570. Schoental, R. (1983e).Nutr. Health 2,147-152. Schoental, R. (1984a). Perspect. Biol. Med. 28, 154-157. Schoental, R. (1984b).J . Appl. Toxicol. 4,222. Schoental, R. (1984c).Chem. Br. 20,1108, 1113. Schoental, R. (1984d).ACS Monogr. 182,1137-1169. Schoental, R. (1984e).Vet. Rec. 115,500. Schoental, R. (1985a). Br. J. Psychiat. 146,115-119. Schoental, R. (1985b).Astronel Korot, in press. Schoental, R., and Connors, T. A., eds. (1981). “Dietary Influences on Cancer, Traditional and Modem” CRC Press, Bota Raton, Florida. Schoental, R., and Gibbard, S. (1978). Biochem. SOC. Trans. 7,127-129. Schoental, R., and Joffe, A. Z. (1974).J. Pathol. 112,34-42. Schoental, R., and van Dorst, S. (1983). Biochem. SOC. Trans. 11,782. Schoental, R., Joffe, A. Z., and Yagen, B. (1979).Cancer Res. 39,2179-2189. Schottenfeld, D., Warshaver, M. E., Sherlock, S., Zauber, A. G., Leder, M., and Payne, R. (1980).Am. J. Epidemiol. 112,232. Schrauzer, G. N. (1979). Adu. Nutr. Res. 2,219-244. Schutte, K. H. (1968). 1.Natl. Cancer Znst. 41,821-824. Schwartzman, M. L. (1937). Sou. Med.]. 3,181-184. Scott, P. M. (1982).I . Assoc. Off.Anal. Chem. 65, 876. Scott, P. M., Panalaks, T., Kanhere, S., and Miles, W. F. (1978). J. Assoc. Off. Anal. Chem. 61,593-600. Scott, P. M., Lau, P.-Y., and Kanhere, S. R. (1981).]. Assoc. Off. Anal. Chem. 64,13641371. Scully, R. E., and Welch, W. R. (1981). In “Developmental Effects of DES in Pregnancy” (A. L. Herbst and H. A. Bern, eds.), pp. 26-45. Thieme, Stuttgart. Searle, C. E., ed. (1984). ACS Monogr. 182. Segal, R., Milo-Goldzweig, I., Joffe, A. Z., and Yagen, B. (1983). Toricol. Appl. PharmaC O ~70,343-349. . Segi, M., Kurihara, M., and Matsuyama, T. (1969).“Cancer Mortality for Selected Sites in 24 Countries; No. 5 (1964-1965).” Dept. of Public Health, Tohoku University School of Medicine, Sendai, Japan. Seitz, L. M., Sauer, D. B., Mohr, H. E.,Burroughs, R., and Paukstelis, J. V . (1975).J. Agric. Food Chenr. 23,l-4. Serenko, A. F., and Romenski, A. A., eds. (1970).“Cancer Morbidity and Mortality Data

CARCINOGENIC METABOLITES OF FUSAlUUM

287

in USSR.” Moscow, Medicine, IARC Internal Technical Report 70/003, prepared by A. Tuyns. Shank, R. C., ed. (1981). “Mycotoxins and N-Nitroso-compounds: Environmental Risks,” Vols. 1-2. CRC Press, Boca Raton, Florida. Shapiro, R. J. (1973).In “Organic Selenium Compounds, Their Chemistry and Biology” (D. L. Klayman and W. H. H. Gunther, eds.), pp. 693-726. Wiley (Interscience), New York. Sheehan, D. M., Frederick, C. B., Branham, W. S., and Heath, J. E. (1982).Carcinogenesis 3, 957-999. Shipchandler, M. T. (1975). Heterocycles 3,471-520. Shlosberg, A., Klinger, Y.,and Malkinson, M. (1984).Vet. Rec. 114, 387. Shotwell, 0. L. (1977). In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 403-413. Pathotox Publ., Park Forest South, Illinois. Shotwell, 0. L., Goulden, M. L., Bennett, G. A., Plattner, R. D., and Hesseltine, C. W. (1977).J . Assoc. Off. Anal. Chem. 60,778-783. Shutt, D. A., and Cox, R. I. (1972).J . Endocrinol. 52,299-310. Sigg, H. P., Mauli, R., Flury, E., and Hauser, D. (1965).HeEu. Chim. Acta 48,962-988. Silberberg, M., and Silberberg, R. (1971). In “The Biochemistry and Physiology of Bone” (G. H. Bourne, ed.), Vol. 3, 2nd Ed., pp. 401-484. Academic Press, New York. Sindeeva, N. D. (1964).“Mineralogy and Types of Deposits of Selenium and Tellurium. Wiley (Interscience), New York. Siriwardana, T.M. G., and Lafont, P. (1978).Appl. Environ. Microbiol. 35, 206-207. Slavik, M. (1978).Recent Results Cancer Res. 63,282-287. Smalley, E. B., and Strong, F. M. (1974). In “Mycotoxins” (I. H. F. Purchase, ed.), pp. 199-228. Elsevier, Amsterdam. Soullier, B. K., Wilson, P. S., and Nigro, N. D. (1981).Toxicol. Lett. 12,343-348. Spallholz, J. E.,Martin, J. L., and Ganther, H. E., eds. (1981).“Selenium in Biology and Medicine.” Avi Publ., Westport, Connecticut. Stjihelin, H., Kalberer-Rusch, M. E., Signer, E., and L a z e , S. (1968).J. Anneimittelforsch. 18, 989-994. Stahr, H. M.,Ross, P. F., and Obioha, W. (1980).J . Agric. Food Chem. 28,286. Stanford, G. K.,Hood, R. D., and Hayes, A. W. (1975).Res. Commun. Puthol. Pharma~01.10,743-746. Steele, J. A., Lieberman, J. R., and Mirocha, C. J. (1974).Can. J . Microbiol. 20,531. Steele, J. A., Mirocha, C. J., and Pathre, S . V. (1976).J . Agric. Food Chem. 24,89-97. Stenchever, M. A., Williamson, R. A., Leonard, J., Karp, L. E., Ley, B., Shy, K., and Smith, D. (1981).Am.J. Obstet. Gynecol. 140, 186-191. Steyn, P. S., Vleggaar, R., Rabie, J., Kriek, P. J., and Harington, J. S. (1978). Phytochemistry 17,949-951. Stjernsward, J. (1969).Antibiot. Chemother. 15,213-233. Stob, M., Baldwin, R. S., Tuite, J., Andrews, F. N., and Gillette, K. G. (1962).Nature (London) 196,1318. Stocks, P., and Davies, R. I. (1960).Br. J . Cancer 14,8-22. Stoloff, L. (1976).Adu. Chem. Ser. 149,23-50. Stoloff, L., Henry, S., and Fransis, 0. T., Jr. (1976).J.Assoc. Off. Anal. Chem. 59, 118121. Sunde, R. A., and Hoekstra, W. G. (1980).Nuts. Rev. 38,265-273. Sutton, J. C., Baliko, W., and Funnel], H. S. (1980).Can.J . Plant Sci. 60, 149-155.

288

R. SCHOENTAL

Szathmary, C. I. (1983). Deu. Food Sci. 4,229-250. Szathmary, C. I., Mirocha, C. J., Palyusik, M., and Pathre, S. V. (1976).Appl. Enuiron. Microbiol. 32,579-584. Szathmary, C. I., Galacz, J., Vida, L., and Alexander, G. (1980). J . Chromatogr. 191, 327-331. Szepsenwol, J. (1963).Proc. SOC. E x p . B i d . Med. 112, 1073-1076. Szepsenwol, f. (1964).Proc. SOC. E x p . Biol. Med. 116, 1136-1139. Takitani, S., and Asabe, Y. (1983).Deu. Food Sci. 4, 113-120. Takitani, S., Asabe, Y., Kato, T., Suzuki, M., and Ueno, Y. (1979).J . Chromatogr. 172, 335-342. Tamm, Ch. (1974).Prog. Chem. Org. Nat. Prod. 31,64-117. Tamm, Ch., and Breitenstein, W. (1980). In “Biosynthesis of Mycotoxin; a Study in Secondary Metabolism” (P. S. Steyn, ed.), pp. 69-104. Academic Press, New York. Tashiro, F., Shibata, A., Nishimura, N., and Ueno, Y. (1983).]. Biochem. 93,1557-1566. Tatsuno, T., Ohtsubo, K., and Saito, M. (1973).Pure Appl. Chem. 35,309-313. Taub, D., Girotra, N. N., Hoffsommer, R. D., Kuo, C. H., Slates, H. L., Weber, S., and Wendler, N. L. (1968). Tetrahedron 24,2443-2461. Thouvenot, D., and Moffin, R. F. (1980).]. Steroid Biochem. 13,1337-1345. Thouvenot, D., and Moffin, R. F. (1983). Appl. Environ. Microbiol. 45, 16-23. Thouvenot, D., Moffin, R., Di Stefano, S., and Picart, D. (1981).Eur. J. Biochern. 121, 139-145. Tiel, P. G., Meyer, C. J., and Marasas, W. F. 0. (1982).J . Agric. Food Chem. 30,308312. Tokamia, C. H., and Dobereiner, J. (1975).Pesqui. Agropecu. Bras. Ser. Vet. 10,79-97. Treip, C . S. (1983).J . Puthol. 141,29-40. Trenholm, H. L., Locksley, W. R., and Farnworth, E. R. (1980). J . Assoc. Off. Anal. Chem. 63,604-611. Trenholm, H. L., Warner, R. M.,and Farnworth, E. R. (1981).J.Assoc. Off. Anal. Chem. 64,302-310. Trenholm, H. L., Cochrane, W. P., Cohen, H., Elliot, J. I., Farnworth, E. R., Friend, D. W., Hamilton, R. M. G., Standish, J. F., and Thompson, B. K. (1983).J. Assoc. Off. Anal. Chern.66,92-97. Turner, W. B. (1971). “Fungal Metabolites.” Academic Press, New York. Turner, W. B., and Aldridge, D. C. (1983). “Fungal Metabolites 11.” Academic Press, New York. Tuyns, A. J. (1970).Int. J . Cancer 5,152-156. Tuyns, A. J. (1979).Cancer Res. 39,2840-2843. Ueno, Y. (1977a).In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 189-207. Pathotox PubI., Park Forest South, Illinois. Ueno, Y. (1977b).Pure Appl. Chern. 49,1737-1745. Ueno, Y. (1980).Ado. Nutr. Res. 3, 301-353. Ueno, Y., ed. (1983a).“Trichothecenes Chemical, Biological, Toxicological Aspects.” Kodasha Ltd., Tokyo and Elsevier, Amsterdam. Ueno, Y. (1983b).Deu. Food Sci. 4,125-135. Ueno, Y., and Kubota, K. (1976).Cancer Res. 36,445-451. Ueno, Y., and Shimada, N. (1974).Chern. Pharm. BUZZ. 22,2744-2746. Ueno, Y., and Tashiro, F. (1981).J. Biochern. 89,563-571. Ueno, Y., Tashiro, F., and Kobayashi, T. (1983).Food Cosmet. Tozicol. 21, 167-173. Ueno, Y., and Yamakawa, H. (1970).Jpn.J. E z p . Med. 40,385-390.

CARCINOGENIC METABOLITES OF FUSARZUM

289

Ueno, Y., Hosoya, M., and Ishikawa, Y. (1969).J . Biochem. (Tokyo) 66,419-422. Ueno, Y., Ishikawa, Y., Amakai, K., Nakajima, M., Saito, M., Enomoto, M., and Ohtsubo, K. (1970).Jpn.J . Erp. Med. 40, 33-38. Ueno, Y., Sato, N., Ishii, K., Sakai, K., Tsunoda, H., and Enomoto, M. (1973).Appl. Microbiol. 25,699-704. Ueno, Y., Shimada, N., Yagasaki, S., and Enomoto, M. (1974). Chem. Phann. Bull. 22, 2830-2835. Ueno, Y., Kubota, K., Ito, T.,and Nakamura, Y. (1978). Cancer Res. 38, 536-542. Ueno, Y., Nakayama, K., Ishii, K.,Tashiro, F., Minoda, Y., Omori, T., and Komagata, K. (1983).Appl. Enuiron. Microbiol. 46, 120-127. Uraguchi, K.,and Yamazaki, M., eds. (1978).“Toxicology, Biochemistry and Pathology of Mycotoxins.” Wiley, New York. Urry, W. H., Wehrmeister, H. L., Hodge, E. B., and Hidy, P. H. (1966). Tetrahedron Lett. 27,3109-3114. Utian, W. H. (1973).Br. Med. J . 1,579-581. Vannier, B., and Raynaud, J. P. (1980).J. Reprod. Fertil. 59,43-49. van Rensburg, S. J., Benad6, A. S., Rose, E. F., and du Plessis, J. P. (1983).Nutr. Cancer 4,206-216. van Rensburg, S . J., Bradshaw, E. S., Bradshaw, D., and Rose, E. F. (1985).Br.J.Cancer 51,399-405. V h y i , A., and Szeky, A. (1980a).Mag. Allatoru. Lapja 35,242-246. V h y i , A., and Szeky, A. (1980b).Mag. Allatoru. Lapja 35,247-252. Vesonder, R. F. (1983).Deu. Food Sci. 4,210-217. Vesonder, R. F., Ellis, J. J., and Rohwedder, W. K. (1981).Appl. Enuiron. Microbiol. 42, 1132-1 134. Vessey, M. P., Lawless, M., McPherson, K. M., and Yeates, D. (1983).Lancet 2,930934. Vessey, M. P. (1984).J. R. SOC. Med. 77,542-549. Visconti, A., and Bottalico, A. (1983).J. Agric. Food Chem. 31, 1122-1123. Vleggaar, R., and Steyn, P. S. (1980).In “Biosynthesis of Mycotoxins. A Study in Secondary Metabolism” (P. S. Steyn, ed.), pp. 395-422. Academic Press, New York. Vlattas, I., Harrison, I. T., TokBs, L., Fried, J. H., and Cross, A. D. (1968).J.Org. Chem. 33,4176-4179. Von Bolliger, G., and Tamm, Ch. (1972). Helu. Chim. Acta 55,3030-3048. Wallace, E. M., Pathre, S. V., Mirocha, C. J., Robison, T. S., and Fenton, S. W. (1977).J. Agric. Food Chem. 25,836-838. Wardell, R. E., Seegmiller, R. E., and Bradshaw, W. S. (1982). Teratology 26, 229237. Warwick, G. P., and Harington, J. S . (1973).Ado. Cancer Res. 17,81-229. Watanabe, K., Reddy, B. S., Weisburger, J. H.,and Kritchevsky, D. (1979a).J . Natl. Cancer Inst. 63, 141-145. Waterhouse, J. A. H. (1985).1. R. SOC. 78 Suppl. 6, 3-7. Webb, J. S., and Atkinson, W. J. (1965).Nature (London) 208,1056-1059. Wehner, F. C., Marasas, W. F. O., and Thiel, P. G. (1978).Appl. Enuiron. Microbiol. 35, 659-662. Wei, C. M., and McLaughlin, C. S . (1974).Biochem. Biophys Res. Commun. 57, 838844. Wells, N. (1967).N. Z.J. Sci. 10, 142-179. Whanger, P. D., Schmitz, J. A., and Exon, J. H. (1982).Nutr. Cancer 3,240-248. Whitehead, E. D., and Leiter, E. (1981).J . Urol. 125,47-50.

290

R. SCHOENTAL

Wiggins, J. P., Rothenbacher, H., Wilson, L. L., Martin, R. J., Wangsness, P. J., and Ziegler, J. H. (1979).J. Anim. Sci. 49,291-297. Wilson, C. A., Everard, D. M., and Schoental, R. (1982). Toxicol. Lett. 10,35-40. Wilson, R. B., Hutcheson, D. P., and Wideman, L. (1977). Am. J. Clin. Nutr. 30, 176181. Workshop on Fat and Cancer (1981). Cancer Res. 41,2678-2827. Wyatt, R. D., Colwell, W. M., Hamilton, P. B., and Burmeister, H. R. (1973). Appl. Microbial. 26, 757-761. Wyatt, R. D., Doerr, J. A., Hamilton, P. B., and Burmeister, H. R. (1975). Appl. Microbiol. 29,641-645. Wyllie, T. D., and Morehouse, L. G., eds. (1977). “Mycotoxic Fungi, Mycotoxins, Mycotoxicoses: An Encyclopedic Handbook,” Vol. 1. Dekker, New York. Wyllie, T. D., and Morehouse, L.G., eds. (1978). “Mycotoxic Fungi, Mycotoxins, Mycotoxicoses: An Encyclopedic Handbook,” Vols. 2 and 3. Dekker, New York. Wynder, E. L., and Hirayama, T. (1977). Preuent. Med. 6,567-594. Yagen, B., Joffe, A. Z., Horn, P., Mor, N., and Lutsky, I. I. (1977). In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, G. W. Hesseltine, and M. A. Mehlman, eds.), pp. 329-336. Pathotox Publ., Park Forest South, Illinois. Yagen, B., Horn, P., Joffe, A. Z., and Cox, R. H. (1980).J. Chem. SOC. Perkin Trans. I, 2914-2917. Yap, H. Y., Murphy, W. K., Distefano, A., Blumenschein, G. R., and Body, G. P. (1979). Cancer Treat. Rep. 63,789-791. Yarom, R.,More, R., Sherman, Y., and Yagen, B. (1983).Br.]. Exp. Pathol. 64,570-577. Yarom, R., More, R., Eldor, A., and Yagen, B. (1984a). Toxicol. Appl. Pharmucol. 73, 210-217. Yarom, R., Sherman, Y., More, R., Ginsburg, I., Borinski, R., and Yagen, B. (198413). Toxicol. Appl. Pharmucol. 75,6048. Yoshizawa, T., and Morooka, N. (1981). In “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 309-321. Pathotox Publ., Park Forest South, Illinois. Yoshizawa, T., Swanson, S.P., and Mirocha, C. J. (1980a).Appl. Enuiron. Microbiol. 39, 1172- 1177. Yoshizawa, T., Swanson, S.P., and Mirocha, C. J. (1980b).Appl. Enuiron. Microbiol. 40, 901-906. Yoshizawa, T . , Mirocha, C. J., Behrens, J. C., and Swanson, S. P. (1981).Food C o m e t . Toxicol. 19, 31-39. Yoshizawa, T., Sakamoto, T., Ayano, Y., and Mirocha, C. J. (1982).Agric. B i d . Chem. 46, 2613-2615. Yoshizawa, T., Sakamoto, T., and Okamoto, K . (1984).Appl. Enuiron. Microbiol. 47, 130-134. Zedeck, M. S. (1980). Preuent. Med. 9,346-351. Zedeck, M. S.,and Lipkin, M. (1981). “Inhibition of Tumor Induction and Development.” Plenum, New York.

COOPERATION BETWEEN MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS: AN EXPERIMENTAL MODEL OF TUMOR PROGRESSION? Nicolas Glaichenhaus, Evelyne Mougneau, Gis&le Connan, Minoo Rassoulzadegan, and FranCois Cuzin Unit6 273 de I’lnstitut National de la Sant6 et de la Recherche MBdicale, Centre de Biochimie, Universitb de Nice, Nice, France

I. Introduction ...................................................... 11. The Multiple Oncogenes of DNA Tumor Viruses ...................... A. Polyoma Virus ................................................. B. Adenoviruses .................................................. 111. Cooperation between Cellular Oncogenes ............................ A. Qualitative Distinction between Two “Complementation Groups” ..... B. Quantitative Dependence on the Level of Oncogene Products.. ....... C. Multistep Pathways and Multigenic Determination .................. IV. “Immortalization” by Genes of Group I: A Complex Phenotype . . . . . . . . . . A. Immortalization, in Vitro Assays, and Biological Significance. . . . . . . . . . B. Changes in Serum Factor Requirements ........................... C. Sensitivity to Tumor Promoters.. ................................. D. Expression in 3T3 Cells Provides Biological Assays and Selective Procedures for Oncogenes of Group I ..................... V. Changes in the Expression of Cellular Genes Inducedby Group I Oncogenes ..................................... VI. Early Stages of Transformation. ..................................... References .......................................................

291 292 292 293 294 294 294 295 296 296 298 298 299

300 301 303

I. Introduction

Stepwise progression of tumoral diseases toward increased invasiveness, hormonal independence, and resistance to chemical and physical treatment is an all-too-common phenomenon which has been analyzed in various experimental systems since the early work of Rous and co-workers (1935, 1941). The stepwise character and apparent irreversibility of these phenotypic changes wouId be most easily accounted for by a sequential accumulation of genetic alterations. Alterations of cellular genes associated with malignancy (oncogenes) were recently identified first as a result of the natural capacity of retroviruses for gene transduction, and subsequently by transfection of genomic DNA from human tumors into mouse cells (see Muller and Verma, 1984, for review). In clear contrast with clinical and biological 291 CoDvrieht 0 1985 bv Academic Press. Inc.

292

NICOLAS GLAICHENHAUS ET AL.

observations, however, these genes appeared to induce an advanced tumoral state with no detectable transition through intermediate stages. Although, as discussed in Section III,C, a multistep pathway does not necessarily require multiple genes, one might, in a pessimistic mood, have questioned the relevance to the cancer problem of these one-step experimental models. Recent studies by Land et al. (1983) and Ruley (1983) suggested, however, a more positive approach. These authors extended to cellular oncogenes some of the conclusions established for DNA tumor viruses (see Cuzin, 1984, for review). On this basis, a second generation of experimental models can now be contemplated which might allow us to identify separate gene functions, cellular targets, and eventually, successive steps in the transformation process. It is reasonable to hope that, still imperfect and highly oversimplified as they are, these multigenic models will help molecular genetics to get one step closer to human pathology, even if it is a modest one. A first positive indication is the number of new questions that can be submitted to experimentation. One possible outcome, discussed in the present review, is the definition on a genetic basis of a very early transition from the fully normal state to a “high-risk state,” where the cells, still apparently normal in many respects, can be further acted upon to produce malignant progeny. II. The Multiple Oncogenes of DNA Tumor Viruses

Unlike most retroviruses, oncogenic DNA viruses which have been analyzed so far appear to require more than one gene to establish and maintain neoplastic transformation. This was extensively documented for polyoma viruses and adenoviruses, and more recent observations suggest a similar situation for two herpesviruses of clinical importance, herpes simplex type 2 and Epstein-Barr virus (Jariwalla et al., 1983; Griffin and Karran, 1984).

A. POLYOMA VIRUS Three proteins (T antigens) are expressed from integrated viral sequences in transformed rodent cell lines: the chromatin-associated large T protein, the membrane-associated middle T protein, and the cytosolic small T protein (see Ito, 1980, for review). Analyzing the effects in rat embryo fibroblasts (REF) and in established 3T3 cells of modified viral genes (plt, pmt, and pst) which separately encode each protein (Treisman et al., 1981; Zhu et al., 1984)led us to identify their

MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS

293

complementary functions in a stepwise transformation pathway leading from the truly normal REF cell to a fully transformed and highly tumorigenic state (Rassoulzadegan et al., 1982, 1983; Cuzin, 1984; Cuzin et al., 1984). Completion of a first transformation step, dependent on the expression of the p l t gene (large T), is required for further transformation of REF cells by the pmt gene, encoding the middle T protein. As described in Section IV, cells expressing only large T exhibit a series of new phenotypic characters, the most conspicuous of which is the ability for long-term growth in cell culture without the concomitant acquisition of tumorigenic properties (“immortalization”) as originally observed for the 3T3 cell lines by Todaro and Green (1963). Transformation of REF cells can be achieved in two steps by the successive introduction of the p l t and p m t genes. However, simultaneous transfer of these two genes did not lead to focus formation unless the third early viral gene, p s t (small T protein), was also present. Further analysis of this apparently paradoxical situation suggested that transformation by p m t requires a cellular function which is constitutively expressed in established cell lines and is induced by the small T protein in REF cells.

B. ADENOVIRUSES Cell lines transformed by oncogenic adenoviruses do not usually maintain a complete copy of the viral genome, but only a subgenomic fragment corresponding to part of the early genes. Transfection experiments indicated that this fragment (region E l ) is indeed sufficient for complete transformation. It contains two transcription units, E1A and ElB, each of them encoding two proteins made from distinct mRNAs produced by alternate splicing of their primary transcripts (see Logan et al., 1984, for review). Neither E1A alone nor E1B alone is capable of inducing a fully transformed state in primary cultures of rat or hamster fibroblasts. Transfection of E 1A promotes the establishment of nontumorigenic clones (Houweling et al., 1980; Van den Elsen et al., 1982). Transfection of E1B produces no detectable effect. An ElA gene product is required for transcription of E 1B (transactivation, see below), but this is not a sufficient explanation for the lack of transforming ability of the isolated E1B genes, since mutants expressing constitutively E1B in the absence of E1A are also transformation negative (Van den Elsen et al., 1983; Logan et al., 1984). Taken together, these results indicate a two-step transformation

294

NICOLAS GWCHENHAUS ET AL.

pathway with an intermediate stage corresponding to immortalized nontumorigenic fibroblast clones. Whether these two steps can be related to those defined by the polyoma plt and pmt genes will be discussed in Section IV.

111. Cooperation between Cellular Oncogenes

A. QUALITATIVE DISTINCTION BETWEEN Two “COMPLEMENTATION GROUPS” Striking similarities were noticed by Land et al. (1983) and by Ruley (1983) between the transforming potential of polyoma virus and adenovirus genes and that of cellular oncogenes in the same rodent fibroblast cells. As polyoma pmt, the ras genes transform established 3T3 cells, but not primary REF cells. In addition, transforming ras alleles were observed to coexist with a rearranged form of the c-myc proto-oncogene in human tumor cell lines (Murrayet al., 1983);hence the hypotheses (1)that ras genes might need other active oncogenes to transform a normal embryonic fibroblast and (2) that, in human cancer cells also, the mutated ras genes might cooperate with other genes to establish and/or maintain the malignant state. Cooperation between mutated ras alleles and other genes was demonstrated in two different types of primary rat cells, embryo fibroblasts and baby kidney cells (BRK) (Land et al., 1983; Ruley, 1983). These other genes, hereafter referred to as Group I oncogenes, include plt, ElA, and two rearranged forms of myc [one found in the avian retrovirus MC29, as a gag-myc fused gene (v-myc) (see Graf and Stehelin, 1982, for review), and the other representing a translocated derivative of the murine myc gene (c-myc), cloned from a mouse plasmacytoma line and expressed under the control of the simian virus 40 (SV40) early promotor]. None of these genes induces focus formation by itself when transfected into either 3T3 or REF cells. B. QUANTITATIVE DEPENDENCE ON THE LEVEL OF

ONCOGENE PRODUCTS Including the v-myc gene in Group I is in a way paradoxical. Indeed, transfection of the gene under control of its original avian long terminal repeat (LTR) sequences produces foci neither on established cell lines nor on REF cells and it immortalizes the latter (Land et al.,

MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS

295

1983; Mougneau et al., 1984; Ruley et al., 1984). However, expression of the gene after infection with MC29 virus leads to efficient transformation of chicken myeloid and fibroblast cells (see Graf and Stkhelin, 1978, for review). Furthermore, Vennstrom et al. (1984)reported that a reconstructed murine retrovirus genome comprising the myc gene of OK10 virus transformed both REF and established mouse cells. These different phenotypic effects are likely to correspond to different levels of expression of the gene and are reminiscent of the fact that, as other potential oncogenes, the normal c-myc gene can be activated by the integration of LTR enhancer and promotor sequences next to the normal gene (Hayward et al., 1981; see Miiller and Verma, 1984, for review). The importance of such quantitative effects was also stressed by the recent work of Spandidos and Wilkie (1984).These authors used DNA constructs in which a mutated TUS allele was associated with several enhancers, so that its expression was increased more than 10-fold as compared with the original tumor cells (or with REF cells under the conditions where cooperation with Group I genes was demonstrated). Expression at these high levels induced both immortalization and transformation of REF cells, The effect of a mutated ras gene product in the regulation of cell growth appears therefore to depend on its intracellular concentration. Here again, a dose dependence has been observed for the wild-type allele, normal c-ras genes being expressed at high levels in various tumor cells (DeFeo et al., 1981; McCoy et al., 1983; Thor et al., 1984) and overexpression under LTR control leading to cell transformation in vitro (Chang et al., 1982). These findings can hardly be considered unexpected, considering these critical and probably pleiotropic elements in the complex cellular machinery. Furthermore, these results do not question the validity of the two-step pathway described for REF cells under defined experimental conditions, but it warns us that the crude genetic data are not to be taken on an all-or-none qualitative basis. It may in fact eventually appear that simple dose effects will still not be sufficient to account for complex cellular responses. Other regulations might be critical, such as a change in the intracellular localization of the gene product or in the timing of its expression with respect to the cell cycle.

C . MULTISTEPPATHWAYS AND MULTIGENIC DETERMINATION The determination of successive steps by distinct genes was operationally important, as it leads to a search for the critical cellular tar-

296

NICOLAS GWCHENHAUS ET AL.

gets. However, one must not forget that there is nothing against the possibility of a unique gene determining more than one transformation step. One clear example is that of the transformation process induced by SV40 virus (see Tooze, 1980, for review). In spite of a noticeable homology between the proteins encoded by polyoma and SV40 viruses, the large T protein of the latter appears to exert the functions of both the large T and middle T proteins of polyoma: one gene, encoding an obviously complex multifunctional protein, is sufficient both for immortalization and for transformation of REF cells. Such a model of one pleiotropic gene product acting at different stages might also apply to the apparently one-step transformation induced by most retroviruses (see Section I). Viral oncogenes have undergone multiple changes as compared with their cellular homologues (see Miiller and Verma, 1984, for review). In addition to changes in the cisacting sequences which regulate gene expression (see Section III,B), multiple alterations of the protein structure and/or deletion of a functional domain might be responsible for a series of different modifications in cellular physiology and growth control mechanisms. IV. “Immortalization” by Genes of Group I: A Complex Phenotype

It was suggested by Land et ul. (1983)that the cooperation observed between myc and M S genes reflected the ability of the former to immortalize REF cells. The rearranged c-myc and the v-myc genes confer in fact to REF and BRK cells two properties exhibited by 3T3 lines, ability to grow as isolated colonies at low cell density and unlimited growth in culture (Mougneau et al., 1984; Ruley et al., 1984). Two other properties, growth at low concentrations of serum and reactiveness to chemical promotors, were conferred on either REF or 3T3 cells by transfer of p l t , c-myc, v-myc, or E1A (Rassoulzadegan et al., 1982,1983; Mougneau et al., 1984; Connan et al., 1985). Such phenotypic characteristics are not apparent in 3T3 cells and thus appear to be properties imparted by the genes of Group I. Moreover, they provide biological assays and possible selection procedures for the genes of this class. A.

IMMORTALIZATION, in Vitro ASSAYS,AND BIOLOGICAL SIGNIFICANCE

Since the pioneer work of Todaro and Green (1963), immortalization of the rodent fibroblast was defined by the expression of two properties, permanent growth in culture and growth at low cell den-

MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS

297

sity, under a precisely defined set of conditions which are obviously highly artificial for a mouse or rat fibroblast [plastic substratum, Dulbecco modified Eagle’s medium (DMEM), COZ buffered and supplemented with 10%fetal or newborn calf serum]. Although these two properties are acquired during the tumoral evolution induced by DNA viruses (Vogt and Dulbecco, 1962, 1963),their expression is not by itself sufficient for malignant growth in the animal, and the cells maintain a normal growth control in culture (low saturation density and anchorage dependency). It is difficult to assess whether the acquisition of immortality in vitro corresponds in uivo to a significant step of tumorigenesis. Although obviously required for the isolation of transformed clones, it is not likely to be only a cell culture artifact: tumor development did not occur following injection in nude mice of REF cells transfected with ras genes (Land et al., 1983), nor after injection in newborn rats of polyoma pmt DNA, under conditions where wild-type pol yoma virus DNA was tumorigenic (Asselin et al., 1983). On the other hand, it is often pointed out that, in species other than rodents, even highly malignant tumor cells could not be propagated in culture. This is clearly the case for most human cancers and, among the reference animal systems, for chicken cells transformed by viruses, such as Rous sarcoma virus (RSV) (STC gene) and even MC29 (myc gene) (D. St6helin, personal communication). It may just be that human and chicken cells require something more than calf serum and D M E M medium to grow, as it may be that these cancer cells evolve through pathways different from that followed in vitro by the rat fibroblasts. One way of approaching this question would be to ask how the introduction of rearranged myc, plt, or E1A genes would affect the growth properties of a normal human cell. Available experimental data are still limited. From our unpublished observations, it appears that the p l t gene confers on human primary cells in culture the ability to grow for a significant number of additional generations, but that these cells will eventually reach crisis and die (our unpublished results). A similar difference between the in vitro behavior of transformed cells of two closely related species was observed for retroviral oncogenes in avian cells. Whereas no chicken cell line could be established following infection with any of these viruses, clonal cell lines could be derived from quail embryonic cells infected with RSV (neuroretina and neural crest) (Pessac et al., 1983; B. Pessac, personal communication or with MC29 virus ( D . Stkhelin, personal communication). Although more experimental data are obviously needed, our present suggestion would be that expression of Group I oncogenes is

298

NICOLAS GLAICHENHAUS ET AL.

likely to be important at an early stage of in vivo tumorigenesis and in oitru transformation, both in rodent and other species’ fibroblasts. At this stage, however, only the rodent cell would express a potential for unlimited growth in DMEM medium supplemented with calf serum, since human and chicken cells lack some genetic or environmental “factor(s).” Immortalization would thus represent an operational assay for this early stage of transformation only in some particular species and cell types. On the other hand, the activity of oncogenes of Group I can be revealed by a series of independent phenotypic properties.

B. CHANGES IN SERUM FACTOR REQUIREMENTS A convenient assay for serum-independent growth properties conferred by the p l t gene was provided by the observation that cells of the established FR3T3 rat line do not grow and eventually die in lowserum medium (0.5% calf serum). Clones expressing plt could be readily selected by colony formation in this medium, where they would thereafter happily multiply (Rassoulzadeganet al., 1982,1983). On the other hand, established FR3T3 cells transformed in high-serum medium by the pmt gene of polyoma virus [middle T transforlines] were shown to lack the independence from serum mants (MTT) factors for growth in culture which is characteristic of polyoma transformants: in low-serum medium, they did multiply, but reverted to a normal phenotype. They could be complemented by transfer of the p l t gene for anchorage-independent growth or focus formation (Rassoulzadegan et al., 1982). The p l t , ElA, v-myc, and rearranged c-myc genes were equally efficient in these two assays (Mougneau et al., 1984; our unpublished results). Serum-independent growth after transfer of myc genes was independently observed by Armelin et al. (1984), who compared the plating efficiencies of BALB/c-3T3 mouse cells in serum and in plasma [platelet-derived growth factor (PDGF) depleted]. It is not yet possible to ascertain whether these different assays reflect the same physiological event(s).

C. SENSITIVITY TO TUMOR PROMOTERS

Cells carrying genes of Group I are not only sensitive to the transforming effect of other oncogenes (Group 11),but they also become susceptible to chemical promoters such as 12-O-tetradecanoylphorbol-1Sacetate (TPA) (Connan et al., 1985). In both REF and FR3T3 cells, TPA treatment after transfer of the p l t , v-myc, or c-myc genes

MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS

299

led to focus formation and to colony formation in suspension. Neither TPA alone nor, obviously, genes of Group I in the absence of TPA had any effect, whereas their combination was as efficient as the transfection of a complete polyoma genome ( p l t + pmt) or of a combination of myc and M S genes. Cell lines established from TPA-induced foci maintained a transformed phenotype in the absence of the drug. D. EXPRESSION IN 3T3 CELLSPROVIDES BIOLOGICAL ASSAYSAND SELECTIVE PROCEDURES FOR ONCOGENES OF GROUP I Although the changes of growth control induced in REF cells by plt, myc, and E1A include clonal immortality in culture, the properties of these cells are distinct from those of 3T3 cells, initially selected for permanent growth in culture after unknown genetic change(s). This is especially clear in the case of FR3T3 cells (Seif and Cuzin, 1977), which are neither serum independent nor reactive to TPA. These properties are likely to be related with a low number of generations in cell culture (less than 50 in these experiments). It is a trivial observation that the behavior of established lines changes with increasing number of passages in culture due to a strong selective pressure in favor of subclones with increased serum independency and faster growth. High-passage FR3T3 cells actually reacted to TPA treatment by an increase in growth rate and saturation density (our unpublished results). As commonly observed with other mouse (BALBh-3T3 and NIH-3T3) and rat lines (Rat-1)at high passage numbers, they also produced spontaneous transformants at a significant rate. The phenotype of cells carrying activated Group I genes is different from that of 3T3 cells which have been maintained for a comparable number of generations in culture and more comparable to what these cells will become at high passage numbers, i.e., in all likelihood, after the accumulation of a number of additional genetic events. These observations provide a series of specific assays for the genes of this group: immortalization of REF cells, complementation of MTT cells for focus formation or colony formation in suspension in lowserum medium, colony formation by 3T3 cells in low-serum medium or in plasma (Rassoulzadegan et al., 1982; Armelin et al., 1984; Mougneau et al., 1984),and focus formation after TPA treatment (Connan et al., 1985). Immortalization of REF cells is often made technically difficult by relatively high backgrounds of spontaneous immortalization events and a limited efficiency of DNA transfection (our unpublished observations), but this is not the case for assays which can be

300

NICOLAS G W C H E N H A U S ET AL.

performed on mouse or rat 3T3 cells taken after a reasonable number of generations in uitro. Combining these assays with transfection of genomic DNA from cancer cells at various stages of their progression is a possible way of searching for new oncogenes, which would not be detected by focus formation on NIH-3T3. V. Changes in the Expression of Cellular Genes Induced by Group I Oncogenes

It is of interest to note that at least two of the proteins encoded by Group I oncogenes, polyoma large T and adenovirus ElA, act in the regulation of viral transcription. Polyoma large T, as its SV40 equivalent, represses the early promotor and activates late gene expression (see Hand, 1981, for review; Brady et al., 1984; Keller and Alwine, 1984; G. Khoury and M. Yaniv, personal communication). E1A proteins are required for transcription of the other early regions of the adenovirus genome (Berk et al., 1979; Jones and Shenk, 1979; Nevins, 1981) and activate expression of various cellular genes, but repress enhancer-dependent transcription (Borrelli et al., 1984, and references therein). The rnyc protein exhibits definite similarities with both large T and E1A: presence in the cell nucleus (Donner et al., 1982; Abrams et al., 1982; Alitalo et al., 1983),binding to DNA (Bunte et al., 1983,1984; Persson and Leder, 1984),and amino acid sequence homology with E1A (Ralston and Bishop, 1983). One may therefore consider the hypothesis that oncogenes of Group I act by modulating, either positively or negatively, the expression of a series of cellular genes. The close similarity of the phenotypes induced by p l t , rnyc, or E1A in rodent fibroblasts suggests that all three proteins might act on common targets, most likely on genes critical for the regulation of cell growth. Identification of such genes was undertaken by searching cDNA libraries, representative of the polyadenylated RNAs of cells expressing the polyoma large T protein, for cellular transcripts present in these cells in elevated amounts (Glaichenhaus et al., 1985). The libraries, in vector pBR322, were constructed from FR3T3 cells expressing large T. In order to compare gene expression between cells grown under comparable conditions, differential screening was performed between these cells (FR3T3-LTl) and cells which express only the middle T protein (MTT4) cultivated in medium containing only 0.5% serum where they both exhibit a normal phenotype (Rassoulzadegan et al., 1982; and Section IV,B). It was first observed, in agreement with previous results on fully

MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS

301

transformed cells (Scott et al., 1983), that the overall pattern of gene expression was not dramatically changed in the presence of the large T protein, a significant increase being observed for only 9 out of 5000 cDNA clones. Increases were of a rather modest amplitude (3- to 5fold), but they were reproducible between different cell lines expressing large T, which had been derived either from FR3T3 or from REF cells, and were either fully transformed ( p l t + pmt) or only immortalized (plt only). None of these nine probes gave a positive signal when hybridized under stringent conditions with plasmid DNA carrying sequences from known oncogenes (abl, erbA, erbB,fes, fms, fos, myb, myc). Two of these mRNAs were present in increased amounts not only in polyoma-immortalized or -transformed cells, but also in cells immortalized by myc and by E lA, and interestingly, in mouse plasmacytoma cell lines with rearrangements of the c-myc gene (Mushinski et aZ., 1983; Stanton et al., 1984), whereas its expression was low in REF cultures and in normal spleen cells. Another observation of possible interest was made for a group of four other mRNAs whose patterns of expression were modified in pltimmortalized cells. Their intracellular levels were changing with the phase of the cell cycle during exponential growth, and this dependence was modified after transformation by polyoma virus. This was deduced fiom the analysis of Cytoplasmic and nuclear RNAs of GI, S, or Gz cell populations prepared by automated sorting of exponentially growing cells on the basis of their DNA content. As shown by Imbert et aZ. (1984),this method of analysis of gene expression during the cell cycle provides unambiguous results without resorting to the use of inhibitory drugs. It is in addition not limited to the analysis of the Go to GI transition, as is the case upon stimulation of serum-deprived or otherwise arrested cultures. Analysis for the presence of these four particular RNAs revealed that they were mostly present during the Gz phase in FR3T3 cells. In polyoma-transformed cells, by contrast, amounts of the specific RNAs were greater in GI than in Gz.These results suggest that the analysis of the transition induced by the p l t or myc genes is a possible approach to the critical controls of the cell cycle. VI. Early Stages of Transformation

Experimental evidence summarized in this review defines an early stage of the oncogenic transformation process induced by the p l t and E 1A oncogenes, and, at least under some defined conditions of expres-

302

NICOLAS GLAICHENHAUS ET AL.

sion, by the rearranged forms of the myc gene. Expression of these genes not only confers on normal rodent fibroblasts a set of negrowth properties, such as immortalization or serum independence but it also creates a high-risk state where the cells become responsive to environmental stimuli (TPA)which do not induce permanent trans formation of normal cells. It is clear that improving our knowledge o these early stages will be critical in order to understand how the disease starts and to devise future diagnostic and eventually therapeu tic strategies. What is still a matter of speculation is the possible generality of tht early stage of the tumoral process as observed in rodent embryo fibro blasts. One view, probably overoptimistic, would be that all malignan diseases start with an early stage strictly comparable to that inducec by myc or p l t , with only some phenotypic variations dependent on the cell type and species. The opposite view would be that the transfor mation stage defined by immortalized rodent cells corresponds to ont particular pathway followed by this particular cell type under a de fined and highly artificial set of experimental conditions. Truth migh lie between these two extremes: for instance, it is likely that studie: on the expression of rearranged myc genes in REF cells will tell u: something about their function in clinically important cancers, such a: B cell neoplasias. It is also likely, however, that the high-risk state induced by m y c and “myc-like” oncogenes is not the only possible starting point of tumor progression. An interesting case of a stepwise transformation process which may involve different steps is that i n duced by papillomaviruses. Attention was recently focused on viruse: of this family, which may be important in the development of humar cancers (see Orth et al., 1980; Boshart et al., 1984, for reviews). The most extensively studied is the bovine papillomavirus type 1(BPV1) which transforms rodent fibroblasts in culture. A series of relativelj well-defined transformation stages were described (Grisoni et al. 1984; Meneguzzi et al., 1984; Cuzin et al., 1985), leading to a highly malignant state. Here again, cells at the first stages maintained struc tures and growth control close to normal, but exhibited both increased frequencies of spontaneous transformation and enhanced responsive ness to tumor promotors. Properties of the BPV1-transformed lines a{ this early stage, however, were significantly different from those i n duced by the genes of Group I. Under the experimental condition: where these genes immortalized REF cells, BPVl DNA did not pro. mote the establishment of cell lines and BPV1-transformed mouse and rat 3T3 lines remained highly serum dependent for growth in culture The papillomavirus oncogene(s) might thus provide us with a clue to 2

MULTIPLE ONCOGENES LN RODENT EMBRYO FIBROBLASTS

303

distinct pathway of tumoral progression. The future will tell us how many more early stages and transformation routes remain to be discovered in addition to the polyomdadenoviruslmyc and to the papilloma early steps.

ACKNOWLEDGMENTS We thank Kenneth B. Marcu and Dominique StBhelin for helpful discussions and critical reading of the manuscript.

REFERENCES Abrams, H. D., Rohrschneider, L. R., and Eisenman, R. N. (1982). Cell 29,427-439. Alitalo, K.,Ramsay, G., Bishop, J. M., Ohlsson Pfeifer, S., Colby, W. W., and Levinson, A. D. (1983).Nature (London) 306,274-277. Armelin, H. A., Armelin, M. C. S., Kelly, K., Stewart, T.,Leder, P., Cochran, B. H., and Stiles, C. D. (1984).Nature (London) 311,655-660. Asselin, C., Gklinas, C., and Bastin, M. (1983).Mol. Cell. Biol. 3, 1451-1457. Berk, A. J., Lee, F., Harrison, T., Williams, J., and Sharp, P. A. (1979).Cell 17,935-944. Borrelli, E., Hen, R., and Chambon, P. (1984).Nature (London) 312,608-612. Boshart, M., Gissmann, L., Ikenberg, H., Kleinheinz, A., Scheurlen, W., and zur Hausen, H. (1984).EMBO]. 3,1151-1157. Brady, J., Bolen, J. B., Radonovich, M., Salzman, N., and Khoury, G. (1984).Proc. Natl. Acad. Sci. U S A . 81,2040-2044. Bunte, T., Greiser-Wilke, I., and Moelling, K. (1983).EMBOJ. 2, 1087-1092. Bunte, T., Donner, P., P f d , E., Reis, B., Greiser-Wilke, I., Schaller, H., and Moelling, K. (1984).EMBOJ. 3, 1919-1924. Chang, E. H., Furth, M. E., Scolnick, E. M., and Lowy, D. R. (1982).Nature (London) 297,479-483. Connan, G., Rassoulzadegan, M., and Cuzin, F. (1985).Nature (London)314,277-279. Cuzin, F. (1984).Biochim. Biophys. Acta 781, 193-204. Cuzin, F., Rassoulzadegan, M., and Lemieux, L. (1984). Cancer Cells (Cold Spring Harbor Lub.) 2, 109-116. Cuzin, F., Meneguzzi, G., BinBtruy, B., Cerni, C., Connan, G., Grisoni, M., and de Lapeyriere, 0. (1985).In “Papilloma Viruses: Molecular and Clinical Aspects” (P. M., Howley and T. R. Broker, eds.), UCLA Symposia on Molecular and Cellular Biology, New Series, Vol. 32. Liss, New York. In press. DeFeo, D., Conda, M. A., Young, H. A., Chang, E. H., Lowy, D. R., Scolnick, E. M., and Ellis, R. W. (1981). Proc. Natl. Acad. Sci. U.SA. 78, 3328-3332. Donner, P., Greiser-Wilke, I., and Moelling, K. (1982).Nature (London) 296,262-266. Glaichenhaus, N., LBopold, P., Masiakowski, P., Zerlin, M., Julius, M., Marcu, K. B., Vaigot,P., and Cuzin, F. (1985).Cancer Cells (Cold Spring Harbor Lab.),in press. Graf, T.,and Stbhelin, D. (1982).Biochim. Biophys. Acta 651,245-271. Griffin, B. E., and Karran, L. (1984).Nature (London) 309,78-82. Grisoni, M., Meneguzzi, G., de LapeyrSre, O., Binktruy, B., Rassoulzadegan, M., and Cuzin, F. (1984).Virology 135, 406-416. Hand, R. (1981).Biochim. Biophys. Acta 651,l-24.

304

NICOLAS GWCHENHAUS ET AL.

Hayward, W. S., Neel, B. G., and Astrin, S. M. (1981). Nature (London) 290,475-480. Houweling, A., van den Elsen, P. J., and van der Eb, A. J. (1980). Virology 105,537-550. Imbert, J., Lawrence, J. J., Coulier, F., Jeunet, E., Billotey, V., and Birg, F. (1984). EMBOJ. 3,2587-2591. Ito, Y. (1980). In “Viral Oncology” (G. Klein, ed.), pp. 447-480. Raven, New York. Jariwalla, R. J., Aurelian, L., and Ts’o, P. 0. (1983). Proc. Natl. Acad. Sci. U S A . 80, 5902-5906. Jones, N., and Shenk, T. (1979). Proc. Natl. Acad. Sci. U S A . 76,3665-3669. Keller, J. M., and Alwine, J. C. (1984). Cell 36,381-389. Land, H., Parada, L. F., and Weinberg, R. A. (1983). Nature (London) 304,596-602. Logan, J., Pilder, S.,and Shenk, T. (1984). Cancer Cells (Cold Spring Harbor Lab.) 2, 527-532. McCoy, M. S., Tool, J. J., Cunningham, J. M., Chang, E. H., Lowry, D. R., and Weinberg, R. A. (1983). Nature (London)302,79-81. Meneguzzi, G., BinBtruy, B., Grisoni, M., and Cuzin, F. (1984).EMBOJ. 3,365-371. Mougneau, E., Lemieux, L., Rassoulzadegan, M., and Cuzin, F. (1984). Proc. Natl. Acad. Sci. U S A . 81,5758-5762. Miiller, R., and Verma, I. M. (1984). Curr. Top. Microbiol. Zrnmunol. 112, 73-115. Murray, M. J., Cunningham, J. M., Parada, L. F., Dautry, F., Lebowitz, P., and Weinberg, R. A. (1983).Cell 33, 749-757. Mushinski, J. F., Bauer, S.R., Potter, M., and Reddy, E. P. (1983). Proc. Natl. Acad. Sci. U S A . 80,1073-1077. Nevin, J. R. (1981). CelE 26,213-220. Orth, G., Favre, M., Breitburd, F., Croissant, O., Jablonska, S., Obalek, S., JanabekChonelskia, M., and Rzesa, G. (1980). Cold Spring Harbor Con$ Cell Prolif. 7, 259-282. Persson, H., and Leder, P. (1984). Science 225,718-721. Pessac, B., Girard, A., Romey, G., Crisanti, P., Lorinet, A. M., and Calothy, G. (1983). Nature (London)302,616-618. Poste, G., and Fidler, I. J. (1980). Nature (London) 283, 139-145. Ralston, R., and Bishop, J. M. (1983). Nature (London) 306,803-806. Rassoulzadegan, M., Cowie, A., Can,A., Glaichenhaus, N., Kamen, R., and Cuzin, F. (1982). Nature (London)300,713-718. Rassoulzadegan, M., Naghashfar, Z., Cowie, A., Cam, A., Grisoni, M., Kamen, R., and Cuzin, F. (1983).Proc. Natl. Acad. Sci. U S A . SO, 4354-4358. Rous, P., and Beard, J. W. (1935).J . Exp. Med. 62,523. Rous, P., and Kidd, J. G. (1941).J. E r p . Med. 73,365-389. Ruley, H. E. (1983). Nature (London)304,602-606. Ruley, H. E., Moomaw, J. F., and Maruyama, K. (1984). Cancer Cells (Cold Spring Harbor Lab.)2,481-486. Scott, M. R. D., Westphal, K. H., and Rigby, P. W. J. (1983). Cell 34,557-567. Seif, R., and Cuzin, F. (1977).J. Virol. 24,721-728. Spandidos, D . A., and Wilkie, N. M. (1984). Nature (London) 310,469-475. Stanton, L. W., Yang, J. Q., Eckhardt, L. A., Birshtein, B., Harris, L., and Marcu, K. B. (1984). Proc. Natl. Acad. Sci. U S A . 81,829-833. Thor, A., Horan Hand, P., Wunderlich, D., Caruso, A., Muraro, R., and Schlom, J. (1984). Nature (London)311,562-564. Todaro, G. J., and Green, H. (1963).J . Cell Biol. 17,299-313. Tooze, J. (1980). “DNA Tumor Viruses,” 2nd Ed. Cold Spring Harbor Laboratories, Cold Springs Harbor, New York.

MULTIPLE ONCOGENES IN RODENT EMBRYO FIBROBLASTS

305

Treisman, R., Novak, U., Favaloro, J., and h e n , R. (1981).Nature (London)292,595-

600.

Van den Elsen, P. J., de Pater, S., Houweling, A., van der Veer, J., and van de Eb, A. (1982).Gene 18, 175-185. Van den Elsen, P., Houweling, A., and van der Eb, A. (1983). Virology 128,377-390. Vennstrom, B., Kahn, P., Adkins, B., Enrietto, P., Hayrnan, M. J., Graf, T., and Luciw, P. (1984). EMBO J . 3,3223-3229. Vogt, M., and Dulbecco, R. (1962). Virology 16,41-51. Vogt, M., and Dulbecco, R. (1963). Proc. Natl. Acad. Sci. U S A . 49, 171-179. Zhu, Z., Veldman, G. M., Cowie, A., Carr, A., Schaffhausen,B., and Karnen, R. K. (1984). J . Virol. 51, 170-180.

This Page Intentionally Left Blank

INDEX

A N-Acetoxyacetylaminofluorene DNA damage, 47 in a sequences, 69 Acquired immunodeficiency syndrome (AIDS) clinical symptoms, 162-163 human retroviruses and, 165, 167-168 immune system disorders, 163-164 new retrovirus HTLV-3LAV as initiator of, 165-167 transmission, 164 Adaptive response to ionizing radiation, human, murine cells, 87 to MNNG in E. coli, 78 in mammalian cells, 78-80 Adenoviruses, oncogenic rodent fibroblast transformation, 293 ElA and ElB gene cooperation, 293 AIDS, see Acquired immunodeficiency syndrome Alcoholic beverages mycotoxins, 224-225 zearalenone persistence, 272 Alimentary toxic aleukemia (ATA) T-2 toxin-induced in cats, 233 in USSR grain belt, human, 225-227 microfungal mycotoxins and, 226227 outbreaks and development, 226 symptoms, 226 Alimentary tract, tumors etiology, 223-225 geographical areas and, 223-224 mycotoxins in alcoholic beverages and, 224-225 T-2 toxin-induced in rats, 233 Amino acid “A” transport system, PDGF-stimulated, 198-199 307

3-Aminobenzamide, poly(ADP-ribose) synthetase inhibition, 71 Anguidine, see Diacetoxyscirpenol Antibody response by splenocytes, FLC and, 152-153 Aphidicolin, DNA repair and, 60-61, 64-65 Arachidonic acid, PDGF-induced release, 197-198 ATA, see Alimentary toxic aleukemia ATP, intracellular poly(ADP4bose) synthesis and, 72-74 Avian myelocytomatosis virus HB1 (HBI virus) in embryonic bursal lymphocytes, 112-114 v-myc function, 112-1 14

6 Baccharinoids antileukemic action, 236-237 from Baccharis megapotamica plants, 236-237 macrocyclic structure, 237 Baccharis cordifolia miotoxins A, B, and C, isolation from, 237-238 roridins A and E, absorption from soil, 238 B. megapotamica baccharinoid formation effect of roridin A in ethanol, 237 mycotoxin translocation and, 236237 Bacillus Calmette-Guerin (BCG) intralesional, tumor regression and, 36-38 B. subtilis W-methylguanine methyltransferase, constitutive, 89

308

INDEX

B cells retrovirus effects on antibody responses, 141-142 mitogen-induced blastogenesis, 142 suppressor activity induction, 157 threshold for activation, 142 BCG, see Bacillus Calmette-Guerin Benzamide, poly(ADP-ribose) synthetase inhibition, 71 Benzo[a]pyrene dioperoxide, DNA damage, 47 Benzoylated napthoilated DEAE-cellulose (BND-cellulose) in DNA repair assay, 57-58 Blym oncogenes ChBlym-1 in bursal lymphomas, chicken absence in preneoplastic cells, 113 acting via transfemn pathway, 112 cloning, 110-111 proteins encoded by, 110, 112 stage-specific activation, 113-114 detection technique, 108 discovery, 107 HuBlym-1 in Burkitt’s lymphoma, human activation in B cell neoplasms, 116 cloning, 115-116 homology with ChBlym-1, 114-117 proteins encoded by, 117-1 18 NIH/3T3 cell-transforming activity, 107-111,113-116, 118-120 tumorigenesis in B cells and, 107, 118 BND-cellulose, see Benzoylated napthoylated DEAE-cellulose Breast cancer oral contraceptives and, 256 in USA and Japan, -223,227 in USSR, 227 Burkitt’s lymphoma, human oncogenes omyc, 115 HuBlym-1, see Blym oncogenes Bursa1 lymphomas, chicken induction by avian leukosis virus, 108 oncogenes ChBlym-1, see Blym oncogenes omyc, 110 V ~ Y C 113-114 , pathogenesis, 108-109

C Calcium ion, intracellular PDGF-induced release, 201 Cancer, human diet and, 223 environmental factors and, 222-223 synthetics and, 222 Cell culture, human plt effect on growth, 297 Cell cycle DNA excision repair during, 59-60 fibroblast, effects of PDGF, 185-187 progression factors, 185-187 mRNA for polyoma large T protein during, 301 Cell lines, human malignant, W-methylguanine acceptor activity, 76-77 virus-transformed, @-methylguanine acceptor activity, 76-78 Cell lines, quail from embryonic cells, induction by Rous sarcoma virus, 297 Cell lines, rodent BALB/o3T3 (mouse) cell cycle, growth factor effects, 185 serum-independent growth, mycinduced, 298 FR3T3 (rat) serum factor replacement by p l t , 298 TPA effect on growth, 299 NIH/3T3 (mouse), transformation by Blym oncogenes, 107-111, 113116, 118-120 sensitive to UV, DNA excision repair deficiency, 91 Cereals Fusarium-infected, zearalenone production, 245247 mycotoxin contamination of grain methods of detection, 270-271 in milling fractions and bran, 269270 during storage, 268-269 Chromosome 22, human translocations in chronic myelogenous leukemia, 195

INDEX

Concomitant immunity, tumor-induced, mouse augmentation by cyclophosphamide, 27-30 endotoxins, 33-36 intralesional adjuvants, 36-38 ionizing radiation, 30-33 decay during tumor growth, 6-7, 12, 15,36 effect on tumor growth, 5 passive transfer, 12-14 by Ly 1-,2+ cytolytic T cells, 13-19, see also Cytolytic T cells by Ly 1+,2- helper T cells, 18 specificity during tumor growth, 6 suppressor T cell effects, 22-27,36, 38-40; see also Suppressor T cells Contraceptives, oral, breast cancer and, 256 Corynebacterium pamum intralesional, tumor regression and, 36-38 Cyclophosphamide combined with adoptive immunotherapy concomitant immunity and, 29-30 suppressor T cell elimination and, 29 tumor regression and, 29-30 immunoprophylactic effects, 27-28 immunotherapeutic action on MOPOC-315 plasmacytoma, 28-29 Cytolytic T cells Ly 1-,2+ phenotype, concomitant immunity passive transfer, 13 generation inhibition by suppressor T cells, 15-19 kinetics, 14-15 Cytosine arabinoside DNA repair and, 63-65 poly(ADP-ribose) synthetase and, 72 Cytoskeleton, PDGF-induced changes, 198-200

D Deoxynivalenol long-term effect unknown, 236 in moldy cereals, 235

309

Deoxynucleoside hiphosphate error-prone DNA synthesis and, 89 Diacetoxyscirpenol antitumor activity tests, human, 234235 isolation fiom Fusarium scirpi, 235 structure and synthesis, 235 Diaglycerol, PDGF-induced transient increase, 197-198 Diet, human, neoplasms and, 223 DiethylstiIbestroI as anabolic in animal husbandry, 251 replacement by zeranol, 251 side effects, 252, 256-257 effects on human sexual development, 254-256 structure, 247 Dijfenbachia sequinae, extract given to women, offspring intelligence and, 238-239 sterility induction, Brazil, 238 DNA 0-alkylated sites, see also 06-Methylguanine methyl nitrosourea-induced, 74 repair, 74-78 complementary (cDNA), to polyoma large protein mRNA, 300-301 damage in uiuo, recognition by cells, 50 heterogeneity in protein association, 67 sequence-imposed, 67 linker, repair, 67-68 in lymphocytes, T-2 toxin-induced breaks, 243 mutagenic lesions bypass by replication, 81-82 instructive, 80 noninshuctive, 80-81 nucleosomal, repair, 67-68 single-stranded (ssDNA) lesions apyrimidinic sites, 47, 49, 51 bypass by DNA polymerase, 47- 50 induction, 47 viral, reaction with mutagen, scheme, 48 DNA excision repair absence in xeroderma cells, 61-63

310

INDEX

restoration in revertants, 92 adduct loss detection, see also Nucleotides analytical methods, 55 immunological assays, 53, 54 (table) in a sequences, induction by various agents, 69 in bacteria, 51-52 during cell cycle, 59-60 deficiency in UV-sensitive rodent lines, 91 discovery, 50-51 DNA replication and, 82-84 DNA synthesis inhibitors and, 63-65 enzymes involved in, 51 in linker regions, 67-69 in mammalian cells, 52-53 mechanisms of, 65-66 MEDLINE data base, 1985,93-95 new nucleotide insertion patch size assays, 55-56 with radioactive thymidine, 5 5 , s pool size effects, 58-59 separation on BND-cellulose, 5758 nuclease sensitivity, 68-69 in nucleosomes, 67-68 in penneabilized skin cells, 61 poly(ADP-ribose) synthetase and, 7172 rate of, 66 DNA N-glycosylases, excision repair and, 51, 52 DNA ligase, excision repair and, 51 DNA polymerases, in excision repair, 47-48,50-51 I1 and I11 in bacteria, 60 a- and p-, in mammalian cells, 60-61, 66 during cell cycle, 59 DNA synthesis error-prone translesion deoxynucleoside triphosphate effect, 89 in E. coli mutations, 84-85 “SOS” cascade of reactions, 85, 88; see also SOS system in virus-infected cells, 85-87 halt in uiuo during repair, 50 inhibition by T-2 toxin, 234

lesion bypass and, 82-84 stops and pauses, detection in uitro, 46-50 unscheduled, deficiency in murine cells, 92-93 herpes simplex virus low survival and, 93 DNA synthesis inhibitors in DNA repair assays, 63-65

E E. coli, see Escherichia coli EGF, see Epidermal growth factor EGF receptor phosphorylation by PDGF, 200 ECF binding inhibition, 200 Endotoxin combined with immunotherapy, large tumor regression, 34-36 small tumor regression, 33-34 Epidermal growth factor (EGF) mitogenic response induction, c-myc role, 207-208 as progression factor, 186 tyrosine phosphorylation, 202-203 Escherichia coli (E. coli) DNA excision repair mechanism of, 51-52 polymerases I1 and 111, 60 mutations error-prone DNA translesion synthesis, 84-85 MNNC-induced, 74 adaptive response to, 78 DNA 0-alkylated site repair, 7475, 78 06-methylguanine acceptor protein formation, 74 from noninstructive lesions, control by umuC gene products, 84-85 umuD gene products, 84-85 Esophagus, cancer in China, microfungi in food and, 228229 Fusarium monil~ormeand, 236 in Iran, selenium-rich plants and, 225

311

INDEX

in South Africa cigarette smoking and, 228 fusarial mycotoxins in maize and, 227-228 selenium-rich wheat and, 227 severe weather, mycotoxicoses and, 228 in USSR grain belt, moldy grain consumption and, 225-227; see also Alimentary toxic leukemia Estrogenic agents including zearalenone and diethylstilbestrol, 253-257 genital abnormalities and, 255 precocious sexual development and, 254 testicular cancer and, 255 natural and synthetic sources for humans, 252-253

F Fibroblasts adenovirus-transformed, hamster, rat two-step pathway, 293-294 viral E l A and E I B gene cooperation, 293 PDGF-induced competence, 185-187 progression factor-induced GI phase, 186-187 simian sarcoma virus-transformed growth in PDGF absence, 188 v-sis-coded protein, 189-193 virus-transformed, PDGF not required, 185-186 Fibrosarcoma Meth A induction of concomitant immunity, 12-15,2224 Ly 1+,2- suppressor T cells, 9-10, 18-22 transplantation tolerance, 26 regression by cyclophosphamide, 29 gamma radiation, 33-34 Friend leukemia complex (FLC) immune response depression, mouse, 135 (table), 143 compartmentalization, 151-152 time course, 137

inhibition of antibody response by splenocytes, 152-153 thymic factor production, 151 Fucocoumarins, DNA damage in a sequences, 69 Fusariurn secondary metabolites, 220-223 mycotoxicoses in livestock and, 221 T-2 toxin isolation from, 232 F . graminearurn, see F. roseurn F. rnonilijorme esophageal tumors and, 236 F. roseurn production of zearalenone and its metabolites 245-246 culture condition effects, 247 F . scripti diacetoxyscirpenol isolation from, 235 F. sporotrichiella grain contamination, USSR, 249 Kashin-Bek disease in children and, 249

G Gamma radiation suppressor T cell elimination, 31-33 tumor regression and, 31-33 Genes JB, PDGF-induced, 208 homology to fos oncogenes, 208-209 JE, PDGF-induced, 205-206 KC,PDGF-induced, 205-206 oncogenes, see Oncogenes PGGF-encoding, 184

H HB1 virus, see Avian myelocytomatosis virus HeLa cells DNA repair rate, 66 Helper T cells Ly 1+,2-phenotype concomitant immunity transfer, 18 tumor priming for endotoxin-induced regression, 35

312

INDEX

Herpes simplex virus survival in murine and human cells, 92-93 Histone 2b, binding to PDGF receptor, 196 Human T cell leukemia virus PDGF-like factor in infected cells, 194 sis gene in infected cells, 194-195 Hydroxyaminoquinone 1-oxide, DNA damage, 47 Hydroxyurea DNA repair and, 61,63-65 reaction with mutagens, 63 poly(ADP-ribose) synthetase and, 72

I Immunoadjuvants, intralesional concomitant immunity increase, 37 cytotoxic T cell generation, 37 tumor regression and, 36-38 Immunoassays for altered nucleotides in DNA repair, 5 3 , s (table) T-2 toxin detection, 245 Immunodeficiencies, retrovirus-induced endogenous retroviruses and, 129-132 exogenous retroviruses and, 130-131 (table), 132-134 history, 125-127 lymphoid cell alterations, 137-148 accessory cells, 143-145 B cells, 141-145 nonspecific defense mechanisms, 145-148 T cells, 139-141 multiple cellular pathways of, 148 in pathogenesis, effects on induced leukemia, 160-161 neoplasia, 158-160 pathogen action of related retroviruses, 159-160 susceptibility to microbial infections, 161-162 tumor-related mechanisms, 158-159 virion component effects, 152-154 Immunodepression, retrovirus-induced antigenic challenge and, 135-136 compartmentalization, 136- 137 host’s genotype and, 135

time course, 136, 137 virus-host combination and, 134-135 Immunodepression, trichotheceneinduced, 243 Immunosuppression, see Immunodepression Ion fluxes, PDGF stimulation of Cae+release, 201 Na+-H+ exchange, 201 Na+-K+ pumping, 201

K Kashin-Bek disease bone deformation in children, USSR, 249 Fusarium sporotrichiella-contaminated grain and, 249

L Leukemia chronic myelogenous chromosome 22 translocations, 195 megakaryocytes as target cells, 195 induced, activation by retrovirus infection, 160-161 Leukocytes, T-2 toxin inhibitory effects, 234 Low-density lipoprotein receptor PDGF-induced stimulation, 198-199 Lymphocytes DNA breaks, induction by T-2 toxin, 243 retrovirus replication, 149-153 T-2 cytotoxicity detection, 244

Macrophages immunosuppression in tumor bearers, mouse, 10 retrovirus effects on functions, 146 (table), 148 as accessory cells, 143-145 as effector cells, 145, 147 envelope protein role, 153 lymphocyte-inhibiting substance production, 158 as suppressor cells, 156-157

313

INDEX

tumoricidity, 35 Mastocytoma P815 induction of concomitant immunity, 17,22-25, 37 Ly 1+,2- suppressor T cells, 9-10, 18-22 regression by Corynebacterium parvum, 36-38 Megakaryocytes PDGF biosynthesis site, 191 target cells in chronic myelogenous leukemia, 195 06-Methylguanine induction in DNA by alkylating agents, 74 removal by acceptor protein, 74-79 06-Methylguanine acceptor protein in mammalian cells, 75-76 effects of viral and malignant transformation, 76-78 W-methylguanine removal &om DNA, 74-75 mechanism of, 75-76 restricted specificity, 76 O6-Methy1guaninemethyltransferase constitutive in Bacillus subtilis, 89 Methyl methanesulfonate DNA damage induction, 56, 63 effect on MNNG-treated E. coli, 75 UV-induced excision repair and, 7374 N-Methyl-N-nitro-N’-nitrosoguanidine (MNNG) adaptive response induction in E. coli, 78 mammalian cells, 78-80 06-methylguanidine removal from DNA, 78-80 E. coli mutation induction, 74 DNA 0-alkylation, 74-75 W-methylguanine acceptor protein formation, 74 nucleotide pool size and, hamster cells, 58 Milk, cow, pig zearalenone metabolites, 250 Miotoxins from Baccharis cordifolia plants, 237-

238

macrocyclic structure, 238 toxicity to livestock, Brazil, 237-238 MMS, see Methyl methanesulfonate MNNG, see N-Methyl-N-nitro-N’-nitrosoguanidine Monocyte-macrophage lineage retrovirus replication, 149-151 Mycotoxicoses, livestock secondary metabolites of Fusarium and, 221 Mycotoxins in alcoholic beverages, 224-225, 272 cancerogenic potentialities, 218-223, 272-274 fermentation-resistant, 272 in milling fractions of grain localization in husks, oily fraction, and bran, 269-270 testing by animal feeding, 270 mycotoxicoses in livestock and, 221

N NAD, intracellular poly(ADP-ribose) synthesis and, 71-72 Natural cytotoxic cells, retrovirus effects, 147 Natural killer cells, retrovirus effects, 147 Neutropenia, retrovirus-induced, kitten, 147 Nuclease, see UV endonuclease Nucleosomes, DNA repair, 67-68 Nucleotides, in DNA repair immunoassays, 53, 54 (table) patch size assay, 55-56 in bacteria, 56 in cell cultures, 57 flash photolysis and, 56 pool size changes, 58-59 effects on mutations, 88-89 Nucleotide sequence oncogenes ChBlym-1 and HuBlym-1, 116-117

Oncogenes Blym, see Blym oncogenes c-fos, PDGF-induced, 209

314

INDEX

c-myc in bursal lymphomas, chicken, 109110 PDGF-induced cell growth independence from PDGF and, 208 cell response to EGF and, 207-208 cellular proliferation and, 206-207 integration of enhancer and promoter sequences, 295 c-sis evolutionary conservation, 192 in human tumors, 194-195 translation product comparison with PDGF, 191-193 E I A of adenovirus, in rodent fibroblasts, 293 E I B of adenovirus, in rodent fibroblasts, 293 in early stages of viral transformation, 301-303 Group I (plt, EIA, v - ~ ~ YC-TTI~C) c, cellular gene expression and, 300 in combination with TPA, polyoma genome transfection, 298-299 at early step of in vitro transfonnation, 298 REF immortalization, 296-297,299 myc in BALB/c3T3 mouse cells, serumindependent growth induction, 298 -PDGF relationship, 209-211 p l t , polyoma virus-induced effect on human cell growth, 297 in FR3T3 rat cells, required for serum-independent growth, 298 in REF large T protein expression, 292293 serum-independent growth induction, 298 pmt, polyoma virus-induced in REF, 292-293 pst, polyoma virus-induced in REF, 292-293 ras, cooperation with c-myc in tumor cell lines, human, 294,296 Group I oncogenes in REF, 294-296

tyrosine kinase activity, 202-203 v-myc of HBl virus, in embryonic bursal lymphocytes, 113-114 v-sis of simian sarcoma virus, 184

P Parodi-Irgens feline sarcoma virus isolation and properties, 192,194 sis gene, 192-195, see also Oncogenes, c-sis PDGF, see Plateletderived growth factor PDGF receptor characterization, 196-197 competitive binding of histone 2b, 196 polysine, 196 protamine sulfate, 196 number per cell, 196 -PDGF compIex, metabolism, 196 protein kinase activity, 197 on target cell surface, 195 Pellagra, moldy corn consumption and, 221,242-243 pH, cytoplasmic, PDGF-induced increase, 201 Plants,hsarial mycotoxin susceptibility, 228 Plasma, platelet-poor progression hctors, 185-187 Plasmacytoma MOPOC-315 cyclophosphamide-inducedregression, 28-29 Platelet-derived growth factor (PDGF) amino acid sequence, 189-191 biosynthesis in megakaryocytes, 191 characterization, 189 conservation since first chordates, 185 discovery, 184 EGF receptor modulation, 200 encoding by two mRNAs, 193 functions in vivo, 184-185 gene expression regulation c-myc mRNA accumulation, 206-209 five new protein induction, 204-205 in a-granules of platelets, human, 188 metabolic effects early, transcription-dependent, 198200

INDEX

immediate, translation-independent, 197-198 -oncogene relationship, 209-21 1 purification, 188-189 stimulation of fibroblast growth, 184 progression factor required for, 185-187 short-time pulse action, 185 ion flux, 201-202 protein phosphorylation ribosomal protein S6 and, 204 tyrosine and, 202-204 in T cells infected with human T cell leukemia virus, 194 Poly(ADP4bose) synthetase ATP level and, 72-74 DNA excision repair and, 71-72 induction in cells by UV, 72 cytosine arabinoside and, 72 hydroxyurea and, 72 NAD level and, 71-72 properties, 70-71 Poly(ADP4bose) synthetase inhibitors DNA excision repair and, 71-72 sister chromatid exchanges and, 71 Polylysine, binding to PDGF receptor, 196 Polyoma virus REF transformation, 292-293 induction of three new T proteins, 292-293 Progression factors, from platelet-poor plasma fibroblast cell cycle induction, 185-187 substitution by EGF, 186 somatomedin C, 186 Prostaglandin E, PDGF-induced release, 197-198 Protamine sulfate, binding to PDGF receptor, 196 Protein kinase, PDGF receptor activity of, 196 Proteins Blym oncogene-encoded in chicken bursa1 lymphomas, 110, 112 in human Burkitt’s lymphoma, 117118

315

c-sis-encoded, structure, 190-193 HA-encoded, viral transcription and, 300-301 large T polyoma plt-encoded, 292-293 viral transcription and, 300-301 SV40-encoded, with middle Tprotein functions, 296 middle T, polyoma pmt-encoded, 292293 ribosomal S6, phosphorylation by PDGF, 204 small T, polyoma pst-encoded, 292293 v-sis-encoded in fibroblasts amino acid sequence, 189-190 biosynthesis, 191-192 comparison with PDGF, 191-193 Protein synthesis inhibition by T-2 toxin, 234 PDGF-induced, 198-199,204-205 UV-induced in human fibroblasts, 87 Psoralen, DNA damage, 47 Pyrimidines removal from DNA, 47,49 W-induced dimers in DNA nucleosomal and linker, 68 removal during repair, 68-69 nuclease sensitivity and, 69

Ralgro, see Zeranol Rat embryo fibroblasts (REF) immortalization in oitro, 296-297, 299-300,302 serum-independent growth, induction by plt transfer, 298 transformation by polyoma virus, 292-293 viral gene-encoded T proteins, 292-293 rus allele cooperation with Group I oncogenes, 294 retrovirus genome, 295 SV40, viral gene-encoded T proteins, 296 REF, see Rat embryo fibroblasts Retroviruses -antibody complexes, circulating, 155

3 16

INDEX

endogenous immunodeficiency induction in uitro, 129, 130-131 (table) immunoresponsiveness of preneoplastic animals and, 132 onmgenicity, 120 exogenous immunodepressive properties, 130131 (table), 133-134 tumor-related mechanisms, 158-159 transmission, 132 genome comprising myc gene of OK10 virus, REF transformation, 295 human KTLV-1 and -2 in AIDS patients, 165, 167 induction of interferons, 157 macrophage-produced substances, 158 Lentivirinae, “slow” disease in animals and, 128 new HTLV-3LAV, as AIDS initiator, 165-167 properties, 166-167 Oncovirinae, neoplasm induction, 128 replication in immunocompetent cells functional imparirnent and, 151-153 lymphoid, 149 metabolism changes and, 151 monocyte-macrophage lineage, 149-151 new surface antigen induction, 151, 160 Spumavirinae, syncitial, 128 suppressor activity induction in B cells, 157 macrophages, 156-157 virion components display onto plasma membrane, 151 envelope protein role, 153 immunodepression by, 152-154 RNA, messenger (mRNA) for PDGF, 193 for polyoma large T protein, 301 Roridin A absorbed from soil by Baccharis cordifolia, 238 Baccharis megapotarnica, 236-237 in ethanol, baccharinoid formation and, 237

Roridin E in Baccharis coridifblia, absorbed from soil, 238 Rous sarcoma virus quail embryonic cell infection, cell line production, 297 tyrosine phosphorylation, 202-203

S Sarcoma S1509 A Ly 1-,2+ suppressor T cells and, 1921 Sarcoma SAl concomitant immunity, transfer by Ly 1-,2+ T cells, 34 regression by endotoxin combined with immunotherapy, 34-35 Ly 1+,2- T cell role, 35 gamma radiation, 32 Satratoxins macrocyclic structure, 240 from microfungus Stachybotrys atra, 239-240 Selenium esophageal cancer and, 225,227 as essential micronutient, 229 resistance to, induction in rats, 230 toxicity, symptoms, human, 229-230 Serum factors cell growth independent from after myc gene transfer, mouse, 298 after p l t gene transfer, rat, 298 Simian sarcoma virus fibroblast transformation, see Fibroblasts isolation and properties, 194 sis gene, see Oncogenes, c-sis, v-sis Simian virus 40 (SV40) in mutagen-pretreated cells mutability, 86-87 survival, 86 REF transformation, 296 Skin test, trichothecene detection, rodents, 244 Somatomedin C, as progression factor, 186 Somatomedin C receptor, PDGF-stimulated, 198-199

INDEX

SOS system, mutagen-induced

317

Transplantation tolerance antitumor immunity suppression and, 25-26 Trichothecenes baccharinoids, 236-237 deoxynivalenol, 235-236 detection methods chromatographic separations, 245 toxicity to skin, mammalian cells, and plants, 244 diacetoxyscirpenol, 234-235 immunosuppression induction, 243 miotoxins, 237-239 satratoxins, 239-240 T-2 toxin, 231-234,240-243 toxicity and structure, 231 Trichothecium roseum trichothecin isolation from, 231 T-2 toxin cytotoxicity, inhibition of DNA synthesis, 234 leukocyte phagocytosis and chemotaxis, 234 protein synthesis, 234 deacetylation by hepatic homogenates, 242 detection by protein synthesis inhibition in reticulocytes, rabbit, 244 radioimmunoassay, 245 T skin test, 244 toxicity to lymphocytes, 244 T cells effects on animals, acute and subacute, retrovirus effects on 234 cell-mediated reactions in vivo, induction of 139-140 ATA in cats, 233 helper activity and, 140 digestive tract tumors, rat, 233 functions in uitro, 139 transitory immunosuppression, 243 lymphokine secretion and, 139-141 12-0-Tetradecanoylphorbol-13-acetate DNA breaks in lymphocytes, 243 isolation from Fusarium spp., 232 @PA) metabolites of, excretion and toxicity, in combination with Group I onco240-241 genes, polyoma genome transfecpartial detoxification, pellagra and, tion, 298-299 242-243 effect on FR3T3 rat cell growth, 299 structure, 231 Thymic factors, FLC effect, 151 synthesized, tritiated, metabolism in Thymidine, radioactive, in DNA repair rodents, 241 assay, 55,58-59 TPA, see 12-O-Tetradecanoylphorbol-13- zearalenone carcinogenic effect increase, Wistar rats, 258-259, 268 acetate Tumor immunogenicity, mouse Transferrin, Blym oncogene functions discovery, 2-5 and, 112, 118

absence in Bacillus subtilis, 89 in E . coli, 85, 88, 89 in hamster cells, 88 in human fibroblasts, 87 Stachybotrys atra satratoxins, isolation from, 240 stachyboiqotoxicosis induction, human, horse, 239-240 Suppressor T cells, retrovirus-induced, 155 Suppressor T cells, tumor-induced, mouse concomitant immunity down regulation, 7, 14, 36 inhibition of cytolytic T cell generation, 15-19 tumor escape from immunity and, 22-25,38-40 elimination by cyclophosphamide, 28-29 gamma radiation, 31-33 X-radiation, 30-32 generation, kinetics, 11-12 Ly 1-,2+ phenotype, 19-21 Ly 1+,2- phenotype, 9-10,19-22 passive transfer of suppression, 8-9 transplantation tolerance and, 25-26

318

INDEX

escape from immune destruction and,

X

38-40

suppression as transplantation tolerance, 25-26 Tumors retrovirus-induced, immunodeficiencies and, 158-159 zearalenone-induced, Wistar rats in female genital organs, 258-264 in male genital organs, 258-259, 265-266 in pituitaries, 259,267-268 stimulation by T-2 toxin, 258-259, 268 Tyrosine, phosphorylation by EGF, 202-203 oncogenes, 202-203 PDGF, 202-204 transforming viruses, 202-203

Ultraviolet (UV) DNA damage, 47,52-56 in a sequences, 69 immediate repair, 66 repair in nucleosomes and linker regions, 67-69 induction of poly(ADP4bose) synthetase, 72 cytosine arabinoside and, 72 hydroxyurea and, 72 protein synthesis, human fibroblasts, 87 UV, see Ultraviolet UV endonuclease, Micrococcus luteus in DNA excision repair glycosylase activity, 51 pyrimidine dimer recognition, 55, 68,69

V Viruses, see also specific viruses in cells pretreated with mutagens mutability increase, 86-87 survival increase, 85-86 transformation induction, early stages, 301-303

Xeroderma pi gmentosum herpes simplex virus low survival in, 92-93 revertant cells, restoration of excision DNA repair, 92 UV resistance, 92 skin tumors and, 91 light-induced lesions, 91 UV-induced DNA damage, lack of repair, 61-63 X-radiation adaptive response induction, murine cells, 87 suppressor T cell elimination, 30-32 tumor regression and, 30-32

z Zearalanols a-epimer, see Zeranol estrogenic activity, 248 isolation from oats and corn, 247 structure, comparison with estrogens, 247-248 as zearalenone metabolites, 249-250 Zearalenone in alcoholic beverages, 272 carcinogenic effects in B6C3F1 mice, 257 in rats F344/N strain, unaffected, 257 Wistar, benign and malignant lesions, 258-268 conversion to wand p-zearalanols, 249-250 defect induction in fetal bones, 248-249 fetal growth,248-249 maternal growth, 248-249 estrogenic activity, side effects, 251257 in animals, 251-252,256-257 in humans, 253-256 from Fusarium-contaminated corn, 245-246 in grain detection methods

INDEX

biological, uterotrophic activity, 27 1 chromatographic, 271 radioimmunoassay, 271 in milling fractions and bran, 269270 moisture effect, 269 during storage, 269 testing by animal feeding, 268-269 variable content, 268-269 worldwide occurrence, 268

319

hyperestrogenism in pigs and, 246 metabolism in hepatic homogenates, 249-250 metabolites, secretion with milk, pig, cow, 250 properties, 246-247 structure, 246 Zeranol as anabolic in animal husbandry, 251 metabolism in women, 252-253 oxidation to zearalanone, 253

This Page Intentionally Left Blank

CONTENTS OF RECENT VOLUMES

Volume 34 The Transformation of Cell Growth and Transmogrification of DNA Synthesis b y Simian Virus 40 Robert G. Martin Immunologic Mechanisms in UV Radiation Carcinogenesis Margaret L. Kripke The Tumor Dormant State E. Federick Wheelock, Kent J . Weinhold, and Judith Levich Marker Chromosome 14q- in Human Cancer and Leukemia Felix Mitelman Structural Diversity among Retroviral Gene Products: A Molecular Approach to the Study of Biological Function through Structural Variability James W.Gautsch, John H. Elder, Fred C. Jensen, and Richard A. Lemer Teratocarcinomas and Other Neoplasms as Developmental Defects in Gene Expression Beatrice Mintz and Roger A. Fleischman Immune Deficiency Predisposing to Epstein-Barr Virus-Induced Lymphoproliferative Diseases: The X-Linked Lymphoproliferative Syndrome as a Model David T. Purtilo INDEX

Volume 35 Polyoma T Antigens Walter Eckhart

Transformation Induced by Herpes Simplex Virus: A Potentially Novel Type of Virus-Cell Interaction Berge Hampar Arachidonic Acid Transformation and Tumor Production Lawrence Levine The Shope Papilloma-Carcinoma Complex of Rabbits: A Model System of Neoplastic Progression and Spontaneous Regression John W . Kreider and Gerald L. Bartktt Regulation of SV40 Gene Expression Adolf Graessman, Monika Gruessman, and Christian Mueller Polyamines in Mammalian Tumors, Part I Giuseppe Sculabrino and Maria E. Feriolo Criteria for Analyzing Interactions between Biologically Active Agents M o m s C. Berenbaum INDEX

Volume 36 Polyamines in Mammalian Tumors, Part I1 Giuseppe Scalabrino and Maria E . Ferioli Chromosome Abnormalities in Malignant Hematologic Diseases Janet D. Rowley and Joseph R. Testa Oncogenes of Spontaneous and Chemically Induced Tumors Robert A. Weinberg Relationship of DNA Tertiary and Quarternary Structure of Carcinogenic Processes 321

322

CONTENTS OF RECENT VOLUMES

Philip D. Lipetz, Alan G. Galsky, and Ralph E. Stephens Human B-Cell Neoplasms in Relation to Normal B-Cell Differentiation and Maturation Processes Tore Godul and Steinar Funderud Evolution in the Treatment Strategy of Hodgkin’s Disease Gianni Bonadonna and Armando Santoro Epstein-Barr Virus Antigens-A Challenge to Modem Biochemistry David A. Thorley-Lawson, Clark M . Edson, and Kathi Geilinger INDEX

Volume 37 Retroviruses and Cancer Genes J . Michael Bishop Cancer, Genes, and Development: The Drosophila Case Elisabeth Gateff Transformation-Associated Tumor Antigens Arnold I . Leuine Pericellular Matrix in Malignant Transformation Kari Alitalo and Antti Vaheri Radiation Oncogenesis in Cell Culture Carmia Borek Mhc Restriction and Zr Genes Jan Klein and Zoltan A. Nagy Phenotypic and Cytogenetic Characteristics of Human B-Lymphoid Cell Lines and Their Relevance for the Etiology of Burkitt’s Lymphoma Kenneth Nilsson and George Klein Translocations Involving zg Locus-Carrying Chromosomes: A Model for Genetic Transposition in Carcinogenesis George Klein and Gilbert Lenoir INDEX

Volume 38 The SJVJ Spontaneous Reticulum Cell

Sarcoma: New Insights in the Fields of Neoantigens, Host-Tumor Interactions, and Regulation of Tumor Growth Benjamin Bonavida The Initiation of DNA Excision-Repair George W . Teebor and Ktystyna Frenkel Steroid Hormone Receptors in Human Breast Cancer George W . Sledge, Jr. and William L. McGuire Relation between Steroid Metabolism of the Host and Genesis of Cancers of the Breast, Uterine Cervix, and Endometrium Mitsuo Kodumu and Toshiko Koduma Fundamentals of Chemotherapy of Myeloid Leukemia by Induction of Leukemia Cell Differentiation Motoo Hozumi The in Vitro Generation of Effector Lymphocytes and Their Employment in Tumor Immunotherapy Eli Kedur and Dauid W. Weiss Cell Surface Glycolipids and Glycoproteins in Malignant Transformation G. Yogeeswaran INDEX

Volume 39 Neoplastic Development in Airway Epithelium P. Nettesheim and A. Marchok Concomitant Tumor Immunity and the Resistance to a Second Tumor Challenge E. Gorelik Antigenic Tumor Cell Variants Obtained with Mutagens Thierry Boon Chromosomes and Cancer in the Mouse: Studies in Tumors, Established Cell Lines, and Cell Hybrids Dorothy A. Miller and Orlando J . Miller

CONTENTS OF RECENT VOLUMES

Polyomarvirus: An Overview of Its Unique Properties Beverly E. Grifjn and Stephen M . Dilworth The Pathogenesis of Oncogenic Avian Retroviruses Paula]. Enrietto andJohn A. Wyke Adjuvant Chemotherapy for Common Solid Tumors Dauid A. Berstock and Michael Baum INDEX

Volume 40 SMethylcytosine, Gene Regulation, and Cancer Arthur D. Riggs and Peter A. Jones Immunobiology of Infection with Human Cytomegalovirus H . Kirchner Genetics of Resistance to Virus-Induced Leukemias Daniel Meruelo and Richard Bach Breast Carcinoma Etiological Factors Cathleen T. Moore Treatment of Acute Leukemia-Advances in Chemotherapy, Immunotherapy, and Bone Marrow Transplantation Gosta Gahrton The Forty-Year-Old Mutation Theory of Luria and Delbriick and Its Pertinence to Cancer Chemotherapy Howard E . Skipper Carcinogenesis and Aging Vladimir N.Anisimov INDEX

Volume 41 The Epidemiology of Diet and Cancer Tim Byers and Saxon Graham Molecular Aspects of Immunoglobulin Expression by Human B Cell Leukemias and Lymphomas John Gordon Mouse Mammary Tumor ViNS: Transcriptional Control and Involvement in Tumorigenesis

323

Nancy E. Hynes, Bernd Groner, and Rob Michalides Dominant Susceptibility to Cancer in Man David Harnden,John Molten, and Terry Featherstone Multiple Myeloma, Waldenstrom's Macroglobulinemia. and Benign Monoclonal Gammopathy: Characteristics of the B Cell Clone, Immunoregulatory Cell Populations and Clinical Implications Hiikan Mellsfedt, Gijran Holm, and Magnus Bjijrkholm Idiotype Network Interactions in Tumor Immunity Hans Schreiber Chromosomal Location of Immunoglobulin Genes: Partial Mapping of These Genes in the Rabbit and Comparison with Ig Genes Carrying Chromosomes of Man and Mouse Leandro Medrano and Bernard Dutrillaux INDEX

Volume 42 Immunological Surveillance of Tumors in the Context of Major Histocompatibility Complex Restriction of T Cell Function Peter C.Doherty, Barbara B. Knowles, and Peter]. Wettstein Immunohistological Analysis of Human Lymphoma: Correlation of Histological and Immunological Categories Harald Stein, and Kad Lennert, Alfred C. Feller, and David Y. Mason Induced Differentiation of Murine Erythroleukemia Cells: Cellular and Molecular Mechanisms Richard A. Rififkind,Michael Sheffeery, and Paul A. Marks Protoneoplasia: The Molecular Biology of Murine Mammary Hyperplasia Robert D. Cardijf

324

CONTENTS OF RECENT VOLUMES

Xiphophorus as an in Vivo Mociel for Studies on Normal and Defective Control of Oncogenes Fritz Anders, Manfred Schartl, Angelika Barnekow, and Annerose Anders Contrasuppression: The Second Law of Thymodynamics, Revisited Douglas R. Green and Richard K. Gershon

Tumor Angiogenesis Judah Folkmon Fusion Proteins in Retroviral Transformation Karin Moelling Application of Migration Inhibition Techniques in Tumor Immunology Robert Szigeti INDEX

INDEX

Volume 44

Volume 43 Cancer Metastasis: Experimental Approaches, Theoretical Concepts, and Impacts for Treatment Strategies Volker Schirwnacher The Canine Transmissible Venereal Tumor: A Unique Result of Tumor Progression D. Cohen Biological and Molecular Analysis of p53 Cellular-Encoded Tumor Antigen Vardu Rotter and David Wolf Monoclonal Antibodies Reactive with Breast Tumor-Associated Antigens Jeffrey Schlom, David Colcher, Patricia Horan Hand, John Greiner, David Wunderlich, Maureen Weeks, Paul B. Fisher, Philip Noguchi, Sidney Petska, and Donald Kufe

Human Cancer-Associated Antigens: Present Status and Implications for Immunodiagnosis Do0 Sulitzeanu Genetic Suppression of Tumor Formation Ruth Sager The Nude Mouse in Cancer Research Beppino C. Giovanella and Jargen Fogh Seroepidemiological Studies on Nasopharyngeal Carcinoma in China Y.Zeng Plasminogen Activators, Tissue Degradation, and Cancer K . Dan@,P. A. Andreasen, J . Grandahl-Hansen, P. Kristensen, L. S. Nielsen, and L. Skriver The Mechanism of Action of mAMSA B. Marshall and R. K. Ralph Retroviruses as Chromosomal Genes in the Mouse Christine A. Kozak INDEX