ADVANCES IN CANCER RESEARCH VOLUME 30
Contributors to This Volume
Faye C. Austin
Alexei J. Likhachev
Joseph Ba16
...
10 downloads
786 Views
18MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
ADVANCES IN CANCER RESEARCH VOLUME 30
Contributors to This Volume
Faye C. Austin
Alexei J. Likhachev
Joseph Ba16
Willie Mark
Charles W. Boone
Alpha Peled
E. H. Cooper
Kazyrnir M. Pozharisski
Necharna Haran-Ghera
Jacob D. Shaposhnikov
Christopher R. Kintner
V. S. Shapot
Valeri F. Klimashevski
Joan Stone
Hiroshi Kobayashi
Bill Sugden
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 30-7979 ACADEMIC PRESS
New York
San Francisco London
A Subsidiary of Harcourt Brace Jovanovich, Publishers
COPYRIGHT @ 1979, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM O R BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING F R O M THE PUBLISHER.
ACADEMIC PRESS, INC.
111 Fifth Avenue, New York, New York 10003
United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. 24/28 Oval Road, London N W l 7DX
LIBRARY OF CONGRESS CATALOG CARD NUMBER:5 2 - 13360 ISBN 0- 12-006630-0 PRINTED IN THE UNITED STATES OF AMERICA
79808182
9 8 7 6 5 4 3 2 1
CONTENTS CONTRIBUTORS TO VOLUME 30
...............................................
ix
Acute Phase Reactant Proteins in Cancer E . H . COOPERA N D JOANSTONE I. I1 . 111 IV . V. VI . VII . VIII . IX . X XI . XI1. XI11. XIV
.
. .
Introduction .......................................................... Production. Half.Lives. and Destruction of Acute Phase Reactant Proteins . Prostate and Breast Cancers ........................................... Gastrointestinal Tract Cancers ......................................... Bladder and Gynecological Cancers ..................................... Lung Cancer ......................................................... Liver Cancer ......................................................... Lymphoma and Leukemia ............................................. Peritoneal and Pleural Effusions ........................................ Extrahepatic Synthesis and Concentration of Acute Phase Reactant Proteins Animal Tumor Systems ............................................... Haptoglobin Phenotypic Variation ...................................... The Biological Effects of Acute Phase Reactant Proteins in Cancer . . . . . . . . Mathematical Addendum .............................................. References ...........................................................
1 4
8 13 15
17 19 20 23 25 26 21 28 30 31
Induction of Leukemia in Mice by Irradiation and Radiation Leukemia Virus Variants NECHAMA HARAN-GHERA A N D ALPHAPELED I. I1 . I11. IV .
Introduction .......................................................... Radiation Leukemogenesis ............................................. Induction of Leukemia by the Radiation Leukemia Virus Variants ......... Concluding Remarks .................................................. References ...........................................................
45 48
62 81 83
On the Multiform Relationships between the Tumor and the Host V . S. SHAPOT
I . Introduction .......................................................... I1. Competitive Relationships between the Tumor and the Host . . . . . . . . . . . . . . . V
89 93
vi
CONTENTS
111. Effect of the Tumor on Biological Characteristics of the Host Tissue
.......
IV. Prospects for the Clinic ............................................... V. Conclusion ........................................................... References ...........................................................
115
137 139 143
Role of Hydrazine in Carcinogenesis JOSEPHB A L ~
.................................................
151
IV. Occurrence of Hydrazines in the Environment ........................... V. The Oncogenicity of Isonicotinylhydrazide .............................. VI . Experiments of H. Druckrey in the Production of Tumors with Hydrazine Compounds ........................................... VII. Production of Polyps and Tumors in the Intestinal VIII. Hydrazine-Caused Cancer ..................... IX. Does INH Produce Tumors in Humans? ........................... X. Methylhydrazine Derivatives, a New Class of Cyt XI. Hydrazine Therapy in Hodgkin’s Disease ................... ................................................ XII. Summary . . . References ........................................
153 I53
I. Introduction.
zine ........................................ 11. Toxic Effect 111. Hydrazine-Induced Alteration in Rat Liver .........................
155
161
Experimental Intestinal Cancer Research with Special Reference to Human Pathology KAZYMIR M. POZHARISSKI, ALEXEIJ. LIKHACHEV, VALERIF. KLIMASHEVSKI, A N D JACOBD. S H A P ~ S H N I K O V I. 11. 111. IV. V. VI , VII. VIII. IX.
X.
Introduction .......................................................... Experimental Models of Intestinal Tumors .............................. Morphology and Morphogenesis of Experimental Intestinal Tumors . . . . . . . . Factors Modifying Intestinal Carcinogenesis ............................. The Kinetics of Intestinal Epithelial Populations in Tumors and during Carcinogenesis ............................................. Biochemistry ......................................................... Immunology ......................................................... Metabolism of 1,2-Dimethylhydrazine and Related Substances . . . . . . . . . . . . . Interaction of 1,2-Dimethylhydrazine and Related Compounds with Cell Components ..................................................... Conclusion ........................................................... References ...........................................................
166 166 169 184
196 206 211 216 222 226 227
vii
CONTENTS
The Molecular Biology of Lyrnphotropic Herpesviruses BILLSUGDEN. CHRISTOPHER R . KINTNER. A N D WILLIEMARK I . Introduction .........................
..
. . . . . . . . . . . . 239
... ... IV . Experimental Tumor Studies ........................................ . . . V . Studies of MDV. HVS. and EBV in Tissue Culture . . . . . . . . . . . . . . . . . . . . . . V1. Identification and Properties of Virus-Related Products for MDV. HVS. and EBV ............ ..................... ... VII . Conclusion ................................................ ... I1 . A Brief Survey of Lymphotropic Herpes .................... I11. MDV, HVS. and EBV in Their Natural Hosts ........................
Addendum References
........................... ....................... ................................................
240 244 248 249
258 266 . . . 268 . . . 268
Viral Xenogenization of Intact Tumor Cells HIROSHIKOBAYASHI I. I1. 111. IV . V.
Introduction .......................................................... Acquisition of a Virus-Specific Antigen ................................. Increase in the Antigenicity of a Tumor-Specific Antigen . . . . . . . . . . . . . . . . . . Immune Responses against Xenogenized Tumor Cells .................... Summary ........................................................... References ..........................................................
279 280 287 292 295 297
Virus Augmentation of the Antigenicity of Tumor Cell Extracts FAYEC . AUSTINA N D CHARLES W . BOONS Introduction ......................................................... Virus Therapy of Cancer ............................................. Augmented Immunogenicity of Virus-Infected Tumor Cell Extracts ....... Mechanisms of Virus Augmentation of TATA Activity . . . . . . . . . . . . . . . . . . . Prospects for the Application of Virus-Augmented Tumor Antigens in Immunodiagnosis and Immunotherapy ............................... VI . Summary ........................................................... References .......................................................... I. I1. 111. IV . V.
SUBJECT INDEX ............................................................ CONTENTSOF PREVIOUS VOLUMES...........................................
301 303 308 329 338 339 340 347 351
This Page Intentionally Left Blank
CONTRIBUTORS TO VOLUME 30 Numbers in parentheses indicate the pages on which the authors’ contributions begin.
FAYEC. AUSTIN,Cell Biology Section, Laboratory of Viral Carcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20014 (301) JOSEPH B A L ~Department , of Pathological Anatomy, Semmelweis Medical University, Budapest, Hungary (151) CHARLESW. BOONE,Cell Biology Section, Laboratory of Viral Carcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20014 (301) E . H. COOPER,The Unit for Cancer Research, University of Leeds, Leeds, England (1) NECHAMAHARAN-GHERA, Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel (45) CHRISTOPHER R. KINTNER,The McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, Wisconsin 53706 (239) VALERIF . KLIMASHEVSKI, Laboratory of Experimental Tumors, N.N. Petrov Research Institute of Oncology, USSR Ministry of Public Health, Pesochny-2, Leningrad 188646, USSR (165) HIROSHIKOBAYASHI, Laboratory of Pathology, Cancer Institute, Hokkaido University School of Medicine, Sapporo, Japan (279) ALEXEIJ . LIKHACHEV, Laboratory of Experimental Tumors, N . N . Petrov Research Institute of Oncology, U S S R Ministry of Public Health, Pesochny-2, Leningrad 188646, USSR (165) WILLIEMARK,The McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, Wisconsin 53706 (239) ALPHAPELED,Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel (45) KAZYMIRM . POZHARISSKI, Laboratory of Experimental Tumors, N.N. Petrov Research Institute of Oncology, USSR Ministry of Public Health, Pesochny-2, Leningrud 188646, USSR (165) JACOB D. SHAPOSHNIKOV, Laboratory of Experimental Tumors, N.N. Petrov Research Institute of Oncology, USSR Ministry of Public Health, Pesochny-2, Leningrad 188646, USSR (165) ix
X
CONTRIBUTORS TO VOLUME
30
V. S. S H A P o r , Cancer Research Center, USSR Academy of Medical Sciences, Moscow, USSR (89) JOAN STONE, Department of Mathematics, University of Bradford, Bradford, England ( 1 ) BILLSUGDEN, The McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, Wisconsin 53706 (239)
ADVANCESINCANCERRESEARCH VOLUME 30
This Page Intentionally Left Blank
ADVANCES IN CANCER RESEARCH,
VOL. 30
ACUTE PHASE REACTANT PROTEINS IN CANCER
E.
H.Cooper and Joan Stone
The Unit for Cancer Research, University of Leeds. Leeds, England; and Department of Mathematics, University of Bradford, Bradford. England
I. Introduction .......................................................... 11. Production, Half-Lives, and Destruction of Acute Phase Reactant Proteins . . 111. Prostate and Breast Cancers ............................................ A. Prostate Cancer ..... ................................... B. Breast Cancer ..................................................... IV. Gastrointestinal Tract Cancers .......................................... V. Bladder and Gynecolog VI. Lung Cancer.. ....... VII. Liver Cancer ............................................. VIII. Lymphoma and Leukemia.. ............................................ IX. Peritoneal and Pleural Effusions ........................................ X. Extrahepatic Synthesis and Concentration of Acute Phase Reactant Proteins XI. Animal Tumor Systems ................................................ XII. Haptoglobin Phenotypic Variation ....................................... XIII. Biological Effects of Acute Phase Reactant Proteins in Cancer . . . . . . . . . . . . . XIV. Mathematical Addendum ............................................... Multivariate Methods .............................................
..........................
1
4 8 8 10 13 15 17 19 20 23 25 26 27 28 30 33 37
I. Introduction
The acute phase reactant proteins (APRPs) are mainly glycoproteins that alter their plasma concentration in response to stimuli produced by many forms of tissue injury, acute and chronic inflammation, connective tissue disorders, and cancer. Clearly as they respond to such a wide variety of stimuli it is self-evident that any changes in these proteins must be regarded as nonspecific. (For general review see Owen, 1967: Koj, 1974: Fisher and Gill, 1975.) Nevertheless, the quantitative and temporal responses of individual members of the APRPs can differ according to the nature of the stimulus and therefore may have diagnostic implications (Laurell, 1974 Hiramatsu et a / . , 1976; Fisher and Gill, 1975). The interest of APRPs in cancer can be considered from both a fundamental and an applied point of view. The underlying fundamental question is the reason for the alteration of the APRPs in response to I Copyright @ 1979 by Academic Press. Inc. All rights of reproduction in any form reserved ISBN 0-12-006630-0
2
E. H. COOPER A N D JOAN STONE
chronic degeneration and cancer and whether this has an advantage to the host or is an aberrant modification of various protective mechanisms that are a vital part of wound healing and the response to infection. The fundamental aspects concern the function of proteins such as a,-acid glycoprotein (alAGP), C-reactive protein (C-RP), and the C3 component of complement in the chronic tissue damage of cancer, as well as the local function of the antiproteases, a,-antitrypsin (a,AT) and a,-antichymotrypsin (alACT), in relation to the progression of an invading neoplasm as it infiltrates the normal host tissues. The main features of this family of proteins are shown in Table I. On the other hand, renewed interest is being taken in APRPs as components of a battery of biological tests for the monitoring of cancer. The history of investigation of APRPs in cancer shows how most of the earlier authors were content to produce a catalog of changes of a particular protein in a wide variety of cancers with very little attention to age, tumor load, or the patients' performance status. They were also handicapped by various technical problems which resulted in the proteins being quantitated indirectly as a measurement of their enzyme inhibitor activity (a,AT), binding capacity as in haptoglobin (HP), as indirect enzyme assays as in ceruloplasmin (CPL), or as partially purified serum fractions, e.g., seromucoid. The general availability of commercial specific antisera and the introduction of the simple single radial immuodiffusion technique of Mancini rf a / . (1965) has greatly enhanced its use in laboratory medicine both for research and routine. In patients in whom there is established cancer or a high probability of cancer, the tests may be required to help assess the prognosis, in particular to increase the accuracy of prediction of probable recurrence or TABLE I TO CLINICAL THEACUTEPHASE REACTANTPROTEINS WITH PARTICULAR RELEVANCE ONCOLOGY Protein a,-Acid glycoprotein a,-Antitrypsin a,-Antichymotrypsin Ceruloplasmin C-Reactive protein Haptoglobin Type 1-1 Type 2- I Type 2-2 Fibrinogen
Symbol aIAGP @,AT aIACT c PL C-RP HP HP 1 - 1 HP 2-1 HP 2-2
Molecular Weight 40,000 54,000 68,000 5 1,000
100,000 -200,000 =400,000 340,000
Amount in normal serum (gdliter) 0.55-1.4 2.0-4.0 0.3-0.6 0.15-0.6 < I 0 (mg/liter) 1 .o-2.2 1.6-3.0 I .2-2.6 2.0-4.5
ACUTE PHASE REACTANT PROTEINS IN CANCER
3
metastases in an individual. The tests may be required to help monitor therapy, especially when the tumor is no longer clinically detectable. Finally in this context is the provision of long-term monitoring of the patient at risk, so that early warning of relapse or progression can be given. As it is quite improbable that APRPs could play any role in population screening for cancer, especially as none of the presently available tumorrelated markers can seriously be advocated for this purpose (Hobbs, 1974; Neville and Cooper, 1976; Schwartz, 1978), this topic will not be discussed further. A second general concept germane to this review is the variation in an individual's general reaction to cancer, as witnessed by such gross indicators as weight loss and a vast array of more subtle biochemical imbalances. In some tumors the metabolic disturbances in the host are gross and coupled to major defects in organ function or cachexia (Theologides, 1971; Bodansky, 1975); in others the abnormalities are often subclinical, but they can decline fairly rapidly from a metastable state to a severe life-threatening illness for what is a relatively small increase of tumor burden or additional burden on a defective system. If skin cancers are excluded, then only 5% of all cancer patients will survive (Seidman et af., 1976) and since the greater proportion of deaths are associated with the effects of metastatic disease, it can be seen that knowledge of the effect of locally recurrent or disseminated tumor on biochemical homeostasis is needed as part of the information for guiding therapy on patients with advanced disease. Within the context of identifying the cancer patient's biochemical status a matrix of levels of various serum proteins, especially those with short half-lives and small pool sizes including certain APRPs, is a promising area for study. The early studies confirmed that advanced cancer was usually accompanied by a rise of a-globulins (Winzler, 1953), and later this was attributed to APRPs but the lack of specificity tended to discourage clinical oncologists from the idea that APRPs could be of any practical value. This discouragement was strongly reinforced by the belief in the late sixties and earlier part of this decade that the age of specific tumor markers had arrived, the virtue of tests being cancer specific was then loudly proclaimed. Unfortunately, tests such as the measurement of plasma carcinoembryonic antigen (CEA), a-fetoprotein (AFP), and beta subunit of human chorionic gonadotropin (PHCG) are nowadays realized to either lack specificity or to have some-;Jhat limited circumstances in which they can be used as an optimal marker (see Neville and Cooper, 1976: Bagshawe and Searle, 1977; King, 1978: Coletta, 1978 Schwartz, 1978; Krebs et al., 1978). Indeed if the behavior of PHCG in
4
E. H . COOPER A N D JOAN STONE
choriocarcinoma is the paradigm (Bagshawe, 1969), few markers can attain such sensitivity, and certainly it leaves many of the common forms of cancer without reliable specific markers, especially at the earlier stages of disease. This experience has resulted in a reappraisal of the use of some of the available nonspecific markers and examination of the value of including them in a battery of markers to monitor cancer, especially in cancers such as the kidney, bladder, and lymphomas where tumorrelated products have little to offer as markers (Neville and Cooper, 1976). There are several studies in the literature that strongly reinforce the view that changes in certain fractions of the plasma proteins, especially the seromucoids (Harshman et ul., 1974); Randle et ul., 1974) and Hp (Jayle e t al., 1968, might be helpful in monitoring cancer, the choice of the particular protein or group being influenced by the techniques available. The advent of monospecific antisera to many human plasma proteins provided a new opportunity to examine the changes in the individual APRPs and their implications in cancer. However, before assessing in what way APRPs could contribute to such a battery, it is important to have a clear idea of the objectives of monitoring. Essentially, monitoring is the collection of biochemical intelligence about the patient which, when taken with the appropriate clinical information, can help the clinician in decision making. The roster of markers from which a battery may be chosen are given in Table 11: for a discussion of multiparametric screening in clinical chemistry see Wolf et al. (1973) and Galen (1975).
II. Production, Half-Lives, and Destruction of Acute Phase Reactant Proteins
The plasma half lives of several APRPs based on tracing iodinated proteins injected into healthy subjects were established in the 1960s. Thus the estimates were 5.2 days for (wlAGP(Weisman el a l . , 1961), 24 days for Hp (Freeman, 1964; Krauss et al., 1966), 4.2 days for CPL (Koskelo et al., 1967, and 3.2 days for fibrinogen (McFarlane et a / . , 1964): this method gives albumin a plasma half-life of 19 days (Peters, 1970). These data in conjunction with the blood levels cannot provide accurate information on the rates of synthesis, as the blood level is the resultant of synthesis, catabolism, and partition between the blood and tissue fluids. The intravenous pool is estimated to be 40% of the total pool of albumin (Peters, 19751, 40% of the total CPL pool (Koskelo et al., 1963, and about 80% of the total fibrinogen pool (McFarlane et al., 1964
ACUTE PHASE REACTANT PROTEINS IN CANCER
5
TABLE I1 LIST OF SOMEPOTENTIAL TUMOR MARKER SUBSTANCES" 1. Tumor-Associated Antigens
Oncofetal wfetoprotein CEA Fetal sulfo-glycoprotein antigen B. Other tumor-associated antigens DNA Binding Proteins Hormones ACTH Calcitonin hCG and PHCG Prolactin Enzymes Acid phosphatase Alkaline phosphatases Aldolase Aminopeptidases Amylase Aryl sulfatase Aspartate arninotransferase Clycos yltransferases y-Glutarnyl transpeptidase Glutathione reductase Histaminase Metabolites and Degradation Products P-Aminoisobutyric acid Fibrinogen degradation products H ydroxyproline K-casein u-lactalburnin Catechol amines and metabolites A.
11.
111.
IV.
V.
Isocitrate dehydrogenase Lactate dehydrogenase Malate dehydrogenase Muramidase 5'-Nucleotidase Pepsinogen isoenzymes Phosphohexose isomerase Ri bonuclease Terminal deoxynucleotidyl transferase Tyrosinase Minor nucleosides Myeloma proteins Polyamines Protein-bound fucose p,-Microglobulin Acute phase reactant proteins
Modified from the list by Dr. R. W. Ruddon, Frederick Cancer Research Center, Frederick, Maryland 21701. (I
Takeda, 1966). The phenotype of Hp profoundly influences the balance between the intra- and extravascular pools, with Hp 1- 1 (mol. wt. 85,000) having an almost equal partition between the pools and Hp 2-2 (mol. wt. 200,000) with only 25% in the extravascular pool (Krauss, 1969). Clearly any extensive inflammatory process altering the permeability of the capillaries might influence this partition between the vascular and extravascular pool as is seen in inflammatory and noninflammatory effusions (Agostoni and Marasini, 1977). Koj (1974) has reviewed the evidence on the metabolism of APRPs in animals and humans during the course of acute and chronic elevations of the serum levels of these proteins. The
6
E. H. COOPER A N D JOAN STONE
general consensus is that the fractional catabolic rate is not influenced by the level of the APRPs in the blood. In relation to cancer the fractional catabolic rate of Hp has been reported to be unchanged in Hodgkin’s disease, even when it was raised to a concentration of 7.0 gm/liter (Krauss, 1969). O’Hara et al., (1967) investigated the Hp half-life in three normal subjects and in three patients with cancer using radioiodinated human Hp, the t l l z was 1.9, 2.1, and 2.1 days in their controls and 2.4, 2.7, and 2.8 days in the cancer patients. But these results must be regarded with some reservation as there is an uncertainty as to what is the true range of normal plasma half-lives and the influence of phenotype on the distribution of Hp in disease (Putnam, 1975). A similar stability has been observed in fibrinogen catabolism in inflammation and cancer (see Koj, 1974). More recently Lyman et al., (1978) have investigated the half-lives of fibrinogen in 30 cancer patients; they found that f 1 / 2 was 3.89 L 1.38 days in controls and 3.01 1.09 in cancer patients, and the half-life was shortest in patients receiving chemotherapy. However, compared to the other APRPs, there is the added complication of various changes in local or generalized deposition of fibrin that may influence the rate of removal of this protein. A starting point for a new chapter in the biochemistry of glycoproteins was the observation that the activity of certain hormones appeared to be dependent on their sialic acid content. This was first demonstrated for follicle stimulating hormone (Gottschalk rt al., 1960) and HCG (Goverde c>t al., 1968: Mori, 1969). Later, the studies of Ashwell and his group (see Ashwell and Morell, 1974, for review) were to extend this observation and produce a general hypothesis that the partial or complete loss of sialic acid from a plasma glycoprotein causes it to be rapidly removed from the circulation (Fig. 1). Beginning with a study of CPL, they were able to show this also applied to alAGP, Hp, and fetuin. Radioisotopic labeling indicated that the liver was the organ responsible for removing the asialated glycoprotein where it undergoes enzymic hydrolysis. A typical result is the finding that the half-life of human a,AT in the circulation of the rat is 18 hours, while in asialic AT it is almost toally cleared in about 30 minutes (Yu and Gan, 1977). The effect of the desialation seems to depend on the exposure of galactose residues which then trigger the recognition processes in the liver: this effect can be inhibited by altering the galactose group (Van den Hamer rt al., 1970). The binding sites on the liver cells appear to involve sialic acid, as treatment of hepatic parenchymal cells with neuraminidase removes the binding activity for asialoglycoprotein, but sialic acid also blocks the binding of glycoproteins to these cells (Ashwell, 1974); this dual function of sialic acid is still an enigma.
*
ACUTE PHASE REACTANT PROTEINS IN CANCER
7
FIG.1. Plasma survival of labeled asialoglycoproteins (from Ashwell and Morrell, 1974).
This mechanism could provide a way of distinguishing recently synthesized molecules from those that are “aged” and ready for removal, that is assuming the aging would involve loss of sialic residues. It is of interest that, in those proteins where a biological activity can be measured, desialation is not associated with a loss of function, for example asialo-a,AT, partially desialated alAT, and partially desialated alAT with oxidized galactose, maintain their trypsin inhibitory and chymotrypsin inhibitory potential (Yu and Gan, 1977). Hence the loss of hormone activity as the result of desialation is due to removal from the circulation rather than interference with the functional moiety. The clearance of alAT in subjects with alAT deficiencies has shown that the protein has a half-life of 6 days (Makino and Reed, 1970), while
8
E. H. COOPER
A N D JOAN STONE
iodinated alAT in recipients of the MM phenotype indicated the half-life was 4 days, and 5.5-6.5 days in the MM and MZphenotypes (Keuppers and Fallat, 1969). This is a somewhat unexpected result as these authors treated their preparation with neuraminidase. More recently, the behavior of the alAT isolated from PiM and PiZ subjects has been investigated by the simultaneous injection of the two labeled proteins into PiM subjects (Laurell et al., 1977). The mean fraction catabolic rates for M protein and Z protein were 0.26 and 0.40,respectively, which is too small a difference to explain why the alAT content of blood in PiZ phenotypes is only 15% of normal. Following injury or estrogen stimulation in M Z and ZZ patients there is a similar proportional increase of a,AT, but the absolute response is less in the ZZ patients. This points to either a small mRNA pool or a slower translation of the Z gene (Laurell or al., 1977). Ill. Prostate and Breast Cancers
It is convenient to consider these cancers together, as they are both hormone dependent and estrogens are frequently used to control their growth. This treatment produces a profound alteration in the plasma protein by its direct effects on their synthesis independent of any effect on the cancer (Musa or af., 1965; Laurell er al., 1968).
A. PROSTATE CANCER
Untreated carcinoma of the prostate has to be associated with raised serum a-globulins which are related to tumor stage and the treatment (Ablin c>t ul., 1973). An elevated pretreatment plasma fibrinogen in prostatic cancer has been significantly correlated with an increased proportion of deaths from all causes and from carcinoma of the prostate (Seal rr ul., 1976). More recently Seal rf ul. (1978) have confirmed that serum Hp levels are raised in prostatic cancer stages 111 and IV, stage I V being significantly higher than state 111. The pretreatment values were significantly correlated with death rates. Ward c’t af. (1977) examined the serum levels of Hp, alAGP, and CPL in benign prostatic hypertrophy (BPH), and untreated carcinoma of the prostate with and without bone metastases. The levels of this protein profile were not statistically different in BPH and the carcinoma without metastases, while carcinoma with bone metastases was accompanied by a significant increase in the four APRPs. They demonstrated that a mean discriminant function using the levels of the proteins in combination with the level of tartrate resistant serum acid
9
ACUTE PHASE REACTANT PROTEINS IN CANCER
phosphatase (SAP) and prealbumin (PALB) could give a correct classification of cases with and without bone metastases 88.6% of the time. The evidence for the bone metastases was the uptake of technicium 99 into the bone as shown by scintigraphy. Figure 2 illustrates the separation achieved by this discriminant function. Subsequently, Houghton ( 1978) reexamined the discriminant on a further set of 25 prostatic cancer patients without bone metastases and 45 with bone metastases and examined the value of the levels of ACT and C-RP in addition to the variates used in the earlier study. A stepwise discrimination showed the importance of the variates in descending order was SAP, alACT, Hp, PALB, alAT, alAGP, and C-RP. He obtained a 90% overall correct assignment with his discriminant function. However, the recent advances in the isolation of specific prostatic acid phosphatase (Foti rt al., 1977) and the finding that serum ribonuclease levels may be a guide in monitoring prostatic cancer (Chu et al., 1977) may result in a more powerful discriminant which would be less liable to be influenced by the vagaries of the APRPs’ response to therapy (Chu, 1978). Several forms of estrogen therapy and estrogen-containing drugs such as estramustine phosphate, a nornitrogen mustard linked to 17p-estradiol phosphate (Jonsson et al., 1975), have major effects on the plasma protein profile in prostatic cancer. These effects include a rise of serum alAT and CPL levels (Ward rt al., 1978) and plasminogen (Seal et al., 1976) and a fall of aIAGP and Hp (Ward rt al., 1978) and fibrinogen (Seal rt af., 1976). Seal and his colleagues (1976) pointed out that the fall of fibrinogen was unexpected as it rises in pregnancy and a rise may be induced by several forms of estrogen-containing oral contraceptives in women. They were unable to account for the rise in patients with prostatic cancer, except to note its effect was maximal when a 1 mg/day dose of diethylstilbestrol was used, the effect being absent at 5 mg/day. Provera, another estrogen used for
I
I
I
I
FIG. 2. Discriminant analysis scores using prealbumin, a,-antitrypsin, a,-acid glycoprotein, haptoglobin, serum acid phosphatase with prior logarithmic transformation of the data. Closed triangles represent metastatic carcinoma of the prostate. Circles represent nonmetastatic carcinoma of the prostate. Discriminant (D)= - 0.638(log,.SAP) 0.767(log,, PALB) - 2.074(log, a,AT) + 0.605(log, AGP) - 0.91 I(log,. Hp) + 4.996. From Ward et (I/. (1977).
+
10
E. H. COOPER A N D JOAN STONE
the management of prostatic cancer, had no effect on the APRPs as judged by changes in plasminogen or fibrinogen. The high a,AT to a,AGP ratio can provide an indication that the patient is receiving estrogens: this is sometimes a useful parameter to measure in the elderly on oral estrogens, as they do not always comply with the physician’s instructions (Table 111). Various other estrogen and steroid binding proteins are increased in the plasma in response to estrogen therapy: these include transcortin (Sandberg r t al., 1975), sex hormone binding globulin (Houghton r t a / . , 1977; Bartch et ul., 1977), and beta steroid binding protein (SP2) (Ward e t al., 1978). The last two proteins are probably the same (Bohn and Krantz, 1973; Bohn, 1974) though the former is assayed by competitive binding of labeled dihydrotestosterone and SP2 by immunodiffusion. B. BREASTCANCER Minton and Bianco (1974) reported serum electrophoresis has demonstrated a raised level of a,-globulins in 32 out of 39 advanced breast cancers with spread beyond local nodes, while in only 1 out of 1 1 with local disease. Pettingale and Tee (1977) examined the plasma protein profiles before and a year after excising an early breast cancer in 30 women, and in 30 women with benign breast disease. Only the levels of p,-glycoprotein preoperatively and CPL after one year were significantly higher in patients with breast cancer than in those with benign disease. An evaluation of APRPs as potential members of an array of markers to monitor breast cancer has been made by Cowen et al., (1978). In their study serum Hp, alAGP, CPL, and a l A T levels were evaluated in association with blood levels of CEA, K-casein, PHCG, and alkaline phosphatase. The values of the proteins in 501 women attending a breast clinic are shown in Table IV. The alAT levels were shown to be of no value as potential markers. They also made a limited study of serum CRP levels to identify the frequency of an abnormal value (>lo mdliter). The following results were obtained: Primary tumors Overt metastases after mastectomy without treatment Follow-up mastectomy clinically tumorfree Benign breast disease Controls
stages 1-111 stage IV
11/81 11/26 1124 3/102 0155 0126
(13.5%) (42%) (29%) (2.9%)
(0%) (0%)
Eighteen out of the 32 positive results (56%) had evidence of at least one
TABLE 111 alAT TO alAGP RATIOI N EACHGROUPI N PROSTATIC CANCER^
No estrogens Groupb
n Mean and standard deviation
1v I1 111 I 22 24 23 27 2.54 f 0.69 2.79 f 0.63 2.86 2 0.70 2.85 ? 0.77
Estrogens V 20 5.79
?
v1
VII
VIII
24 32 27 1.60 7 . 6 4 k 1.98 6.13 ? 2.12 6.50 ? 2.16
From Ward et al. (1977). I = benign prostatic hypertrophy, I1 = cancer prostate (CP) nonmetastatic untreated, 111 = CP metastatic untreated, IV = CP metastatic orchidectomy, V = CP nonrnetastatic-stilbestrol, 1 mg daily, V1 = CP nonmetastatic-stilbestrol, 3 mg daily, VII = CP rnetastaticstilbestrol, I rng daily, and VlII = CP metastatic-stilbestrol, 3 rng daily. a
TABLE IV INCIDENCE OF WOMENWITH BREASTCANCER A N D CONTROLS WITH RAISEDLEVELSOF TUMORASSOCIATED (TA) PRODUCTS AND THOSEWITH RAISEDLEVELSOF TUMOR ASSOCIATED AND/OR TUMORRELATED(TR) PRODUCTS
Control Benign breast disease Primary breast cancer stages I, 11, and I11 Postmastectomy clinically tumor-free Primary breast cancer stage IV and untreated metastatic cancer Metastatic breast cancer receiving treatment
Hpt >5.2 gdliter
AGP >1.4 gdliter
3/42
3/42
6/100
5/100
3/46
Alkaline C P L Phosphatase More than one >0.55 233 T A product gdliter IU/liter raised
Any one T A or TR marker raised"
Any two TA and one TR or Any two TA or any two TR and TR markers one TA markers raised" raised"
1/46
2/41 5/98 1/44
214 1 4/99 13/44
214 1 2/99 2145
141156
191156
11150
441155
14115s
48.1
16.0
5.1
14/34
11/34
5/34
23132
16/32
93.8
73.5
20.6
6/49
11/50
9/41
23/46
14/46
82.0
44.0
14.0
26.2% 29.0 39.1
OF
4.8% 5.0 8.6
0%
0 0
Expressed a s the minimum percentage. Some patients excluded as all the markers were not estimated. From Cowen et a/. (1978).
ACUTE PHASE REACTANT PROTEINS IN CANCER
13
other acute phase reactant protein elevated above the discriminant level. Coombes et al., (1977, who examined a much wider panel of "markers" though in only 36 patients with breast lumps prior to diagnosis and 17 breast cancers with overt metastases, considered that serum alAGP and C-RP levels were potentially useful and included them with ferritin, CEA, sialyl-transferase, alkaline phosphatase, and urinary hydroxypro1ine:creatinine ratios as part of their final panel. However, as Mach et a / . (1977) have pointed out, there is a paucity of reliable markers in breast cancer. The APRPs tend to show a general pattern of elevation that depends on state and tumor load, but there is very considerable variation within this pattern. The long natural history of breast cancer, especially in younger women is well known, where there may be a disease-free interval of several years between the treatment of the primary tumor and the occurrence of metastases. This long time has deterred any longitudinal studies, and unfortunately the long-term adjuvant treatment programs in the United States (Bonadonna, 1977) and Italy (Fisher, 1977) have not included any studies of the evolution of markers in the patients at risk. IV. Gastrointestinal Tract Cancers
The confirmation that the levels of plasma CEA can be raised in many forms of gastrointestinal diseases other than cancer (see Neville and Cooper, 1976) has raised the question of whether the alterations of serum proteins in various neoplastic and inflammatory bowel disease might interfere with the assay of CEA. Crawley et af. (1974), examining malignant and benign bowel disease and liver disease, found that CEA and a,AGP, and CEA and alAT showed significant correlations (p = <0.01) with correlation coefficients of 0.240 and 0.246, respectively; but there was sufficient scatter in the data to suggest these proteins do not influence the assay of CEA. An increased a-globulin has been observed with a raised CEA in primary colon cancer (Hsu and LoGerfo, 1972). A similar rise of a-globulins has been observed in patients after resection of a large bowel cancer with a raised plasma CEA but no clinical evidence of recurrence (Cooper et a/., 1976). They also undertook preliminary studies of the evolution of serum Hp levels in large bowel cancer using hemoglobin binding capacity which suggested this may be an adjunct to CEA, especially as there was good stability of the Hp levels in individual patients who remained tumor free for 12-16 months after resection of a primary tumor. Ward et al. (1977) considered that a battery of tests including the levels of plasma CEA and y-glutamyl transpeptidase (yGT) as well as those of PALB, Hp, a,AGP, and a,AT could help in the
14
E. H . COOPER A N D JOAN STONE
monitoring of large bowel cancer and possibly have predictive value as an aid to identifying patients with apparently curable disease who subsequently recur. This approach was tested practically by Bullen et al. (1977) who used such a battery to monitor patients with gastrointestinal cancer receiving chemotherapy. The battery could provide the clinician with an earlier advanced warning of progression than CEA, especially in stomach cancer, but unfortunately it appears to have little practical advantage in a known case of residual cancer as the alternative chemotherapies currently available are few and of limited efficacity (Carter, 1976). Recently Milano and his colleagues (1978) have measured serum PALB and C-RP levels in advanced large bowel cancer and found that the fall of PALB and coincidental rise of C-RP, in the absence of infection, usually heralded the terminal stages of the disease when the patient was still ambulent, death often ensuing with 1-3 months of this event. The ratio of C-RP to PALB was also found to be a helpful guide to the severity of burns and ulcerative colitis (Kohn et al., 1978). The reciprocal fall of PALB when the APRPs are rising after surgical injury are well described: the magnitude of this response is less than that of alAGP and Hp (Aronsen rt al., 1972). Indeed, this finding of an often rapid loss of a metastable equilibrium, as reflected in the rise in levels of APRPs and fall in PALB, seems to be a common feature in several forms of cancer. In stomach cancer and some benign diseases of the stomach the concentration of arAGP may be raised in the gastric juice. Rapp et al. (1975) found no alAGP in the juice from 104 patients with normal stomach or mild gastritis, and detectable alAGP in 34 out of 38 (8%) stomach cancers and 19 out of 165 (12%) in various ulcers and gastritis. Barbotte rt al. (1976) found that after insulin stimulation the aIAGP concentration in the gastric juice was abnormal in 23 out of 28 (82%) stomach cancers and 9 out of 30 (3Wo) benign lesions. They found a raised alAGP was more reliably associated with stomach cancer than the levels of CEA in the gastric juice which were raised in 1 out of 28 normal controls, 1 out of 30 benign lesions, and 9 out of 28 cancers. Suga and Tamura (1972), observing the effects of chemotherapy on advanced cancer during a short period of 8 weeks, considered that the change in levels reflected response or resistance. However, our experience in monitoring chemotherapy of stomach cancer over a longer time period of up to 15 months suggests that the APRPs tend to be relatively insensitive to the early phases of tumor growth after resection of a stomach cancer, the upward trend of the APRPs tending to be an event during the last 3-4 months of life. A similar observation has been made using alAGP as a marker (T. Fielding, personal communication). A typical example of the evolution of APRPs
15
ACUTE PHASE REACTANT PROTEINS IN CANCER PATIENT DIED
ACT
all AT
011 65AT
1.841.4-
320.60.2
-
WRMAL RANGES
I-
FIG.3. Evolution of three acute phase reactants in a patient with an inoperable carcinoma of the stomach (from Kelly et ul., 1978) .
in a case of advanced stomach cancer is shown in Fig. 3. However, caution must be exercised in the interpretation of the patterns of change in the profiles of APRPs in gastrointestinal disease when faced with the patient for the first time, as it is well known that major changes are produced by benign disease of the bowel such as Crohn's disease and ulcerative colitis (Marner et al. 1975: Pepys et al., 1977). The stimulus for the elevation of C-RP appears to be more pronounced in Crohn's disease than in ulcerative colitis (Pepys et al., 1977, 1978), though this is very likely to be influenced by patient selection. Nevertheless, the patterns of change often show subtle differences that help in possible differential diagnosis in upper gastrointestinal tract disease, but they would appear to have no contribution to the diagnosis of large bowel cancer.
V. Bladder and Gynecological Cancers
Bladder cancer presents an interesting opportunity to study the relationship of the APRPs response to increasing local invasion of an organ. Hollinshead et al. (1977), using a gradient polyacrylamide gel electrophoretic (PAGE) system to separate plasma proteins, observed that the
16
E. H . COOPER A N D JOAN STONE
ratio of the densities alAGP to PALB, which are the first two bands on PAGE gels, provided an index that appeared to be influenced by the extent and stage of bladder cancer; the values obtained by radial immunodiffusion showed a similar trend but were not so sensitive an index (Table V). Bastable et al. (1978) surveyed a number of APRPs and found that the levels of serum C-RP, alAGP, and q A C T could provide the basis of a discriminant system that might be included in any long-term monitoring (Table VI). Essentially, by the choice of appropriate levels, they could separate benign disease and superficial bladder cancer (T 1) from the invasive forms, but while the proportion of patients with abnormal APRPs increased with stage, there are undoubtedly some who have active tumors whose levels were within the 95th percentile of normal. Longitudinal studies in T, and T, bladder cancer showed that there may be periods of several months in which the profile of APRPs is fairly stable. These tend to correspond with a satisfactory clinical state. On the other hand, slow or rapid persistent elevations of the APRPs were invariably associated with clinical deterioration. The recruitment of a rising C-RP is usually a late event and when it occurs it is often when the serum albumin has fallen below normal limits (Cooper, 1978b). Mueller et al. (1971) have shown the advantages of serial measurements of serum Hp levels as an indicator of active disease in malignant ovarian cancers. Primary tumors less than 6 cm in diameter excited no elevation of Hp; patients with tumors over 6 cm in diameter had raised Hps. The
TABLE V CANCER SENSITIVITY INDEX (CSI) OF STAGED BLADDER CANCER SERAB Y Two DIFFERENT METHODS I N Two CENTERS~ Sera Mancini (Leeds) Bladder cancer T, Bladder cancer T,-s Bladder cancer T, Bladder cancer tumor-free
Number of cases
CSI
2S.D.
49 24 30 65
3.6 4.9 8.9 3.5
4.1 5.3 2.7
1.1
Densitometry-gradient PAGE (George Washington University) I1 1.05 0.4 Bladder cancer T, Bladder cancer T, 9 2.40 1.6 Bladder cancer T, 5 5.57 4.2 Bladder cancer T, II 11.00 12.5 Bladder cancer tumor-free 10 1.73 0.5 (I
From Hollinshead e f a / . (1977).
ACUTE PHASE REACTANT PROTEINS IN CANCER
17
TABLE VI FREQUENCY O F ELEVATION OF THREESERUMACUTEPHASEREACTANT PROTEINS" IN BLADDERCANCERA N D UROLOGICAL CONTROLS
T free + controlsb TI + v e T2 +ve T3 +ve T4 +ve
n
Normal
141
43
131 (93%) 35 (81%)
35 78
7 (20%) I I (14%)
One abnormality
Two or three abnormalities 1 (0.7%)
9 (6%) 7(16%) 6 (17%) 14(18%)
I(2%) 22 (63%) 53 (68%)
a,AGP > 1.25 gm/liter; a,ACT > 0.75 gmlliter; C-RP > 12 mglliter.
* Patients who had a cancer resected and were tumor-free. controls = urological patients without bladder cancer. + v e = tumor present in the bladder. From Bastable et a / . (1979).
values returned to normal after successful treatment and rose again if the tumor recurred and spread. Serum Hp levels have also been suggested as an indicator of metastases in carcinoma of the kidney (Vickers, 1974). Serum seromucoid levels have been studied extensively in about 300 benign gynecological conditions and 90 malignant diseases of the female genital tract by Randle e t al. (1974). The benign disease did raise the level of serum seromucoids above that in healthy women. On the other hand, in malignant disease the levels were raised and tended to reflect the stage of the disease. Carcinoma in situ was not associated with a n y increase of seromucoid. Latner r t ul. (1976), based on a observation that a broad spectrum antiprotease inhibitor, Aprotinin, inhibits tumor invasion in experimental animals (Latner et al., 1974), thought that alAT might be concerned with the body's reaction to cancer. They put this idea to the test in a study of women with carcinoma of the cervix and found a progressive rise from 2.4 2 0.09 gm/liter in normal women, not taking oral contraceptives, to 3.57 + 0,24gm/liter in advanced cancer of the cervix. They observed that women who had positive cervical smear cytology showed a significant rise of a,AT, the highest level being reached when the positive smear was accompanied by an intraepithelial tumor that had invaded through the basement membrane. This suggests that in this form of cancer, alAT production may be more sensitive to the signals produced by microinvasion than seromucoid (mainly alAGP). VI. Lung Cancer
There has been intermittent interest in APRPs in lung cancer. Abraham and Schutt (1970) and Schutt and Hoffmeister (1971) using PAGE systems
18
E. H . COOPER A N D JOAN STONE
reported that increased Hp and transferrin were good indicators for an early diagnosis of lung cancer. However, this did not appear to stand up to more critical analysis as the intensity of change in the levels of Hp (Favre rr al., 1972) and Hp and transferrin (Douma and van Dalen, 1974) would appear to be as great in various benign lesions of the lung, such as pulmonary embolism and pneumonia, as in bronchial carcinoma. Seromucoid fraction has been proposed as a possible aid to the monitoring of lung cancer (Harshman rt al., 1974). While APRPs may be able to draw the physician’s attention to the patients’ deterioration, the cause will not necessarily be apparent, as the signals produced by both the tumor-host interaction and those from infected tissue will invariably be mixed in lung cancer. The other aspects of interest have focused on the role of antiproteases in lung cancer. It is now well established that homozygous alAT deficiency occurs in about 0.01% of the population who have serum alAT levels that are about 1%-15% of that found in normal individuals. The homozygous deficient ZZ phenotype is clearly associated with chronic obstructive lung disease (Eriksson, 1965; Ganrot r t al., 1967) and familiar infantile hepatic cirrhosis (Sharp et al., 1969: Johnson and Alper, 1970; Porter e t al., 1972). It has been suggested the heterozygous deficient 10% of the population and have serum a,AT subjects, who are about alevels about 55% of normal, may be predisposed to chronic obstructive lung disease caused by tobacco smoke (Mittman et al., 1971; Lieberman, 1973). Recently, Thompson rt al. (1978) have found that a,-macroglobin ( a : , M ) increased in patients with lung disease who had low levels of serum alAT. They suggest that as a 2 M is normally unchanged by infection and injury its rise in the serum in AT deficiency could be a compensation for the loss of protease activity. This association of a,AT deficiency and lung disease has prompted several studies in patients with lung cancer. However, no association between alAT deficiency and lung cancer have been found. The levels of a,AT are raised in unresectable lung cancer independent of its histological type (Harris et al., 1974; Mickshe and Kokron, 1977, while those of a , M were unchanged (Mickshe and Kokron, 1977). Enzymological measurement (Rees et al., 1975) and immunological assays of @,ATlevels (Elson et al., 1974) have indicated there is a rise in the levels of serum alAT levels associated with smoking. These studies of serum a,AT levels indicate they are influenced by environmental effects. It has been suggested such a rise could be a protective mechanism to limit the damaging effect of the tobacco smoke. Hollinshead e t al. (1977) found the alAGP to PALB ratio to be 0.71 2 0.31 in healthy
ACUTE PHASE REACTANT PROTEINS IN CANCER
19
adults, 1.24 k 0.70in smokers, 1.67 & 0.75 in benign respiratory disease, 4.17 +- 3.76 in early lung cancer, and 8.78 f 9.08 in advanced lung cancer. When these ratios were measured by immunodiffusion the first three groups could not be distinguished, all having a ratio of approximately 4.0; the values in lung cancers were higher but with very wide variances. In summary, there would seem to be little evidence to directly link homozygous or heterozygous alAT deficiency and lung cancer. The evidence, suggesting that a raised alAT level in smokers is an advantage to the subjects, is conjectural.
VII. Liver Cancer
In Western populations primary liver cancer is a rare tumor, while metastatic cancer from the bowel, bronchus, and breast is common. Metastatic cancer in the liver is eventually always accompanied by a rise in APRPs, but in colon cancer this can be preceded for many months by a rise in CEA and sensitive serum enzymes such as yGT (Milano et al., 1978; Cooper and Neville, 1978). The rare association of homozygous or heterozgous alAT deficiency and hepatocellular or cholangiocellular primary liver cancer has been reported (Eriksson and Hagerstrand, 1974; Aagenaes et al., 1974; Lieberman, 1974; Lieberman et al., 1975; Zwi et al., 1975). This has led to searches as to whether this is an underlying cause of primary hepatoma, but its uncommon nature in Europe and North America makes it a difficult study. On the other hand, the tumor is not infrequent among African Negroes. Kew et al. (1978) have reported a study of alAT in 77 South African blacks with hepatic cancer. These patients nearly all had raised serum alAT levels (2.7-4.5 gdliter), while the normal range was 1.8-3.2 gdliter in the controls. There was no case of ZZ homozygous phenotype, nor any abnormal distribution of phenotypes in these African patients. Furthermore, they were unable to detect the periodic acid Schiff (PAS) positive, diastase-resistant globules in the hepatocytes of tumorbearing livers which are thought to be asialyl alAT in the livers of patients with ZZ aIAT deficiency (Sharp, et al., 1969). Major changes in plasma proteins are produced in many forms of benign and malignant liver disease (Hallen and Laurel], 1972; MurrayLyon and Williams, 1974; Skrede et al., 1975; Hiramatsu et al., 1976). However, there can be marked differences in the plasma protein profiles that can help in the discrimination of different forms of liver disease.
20
E . H . COOPER A N D JOAN STONE
VIII. Lymphoma a n d Leukemia The monitoring of lymphoma and leukemia pose different problems to those encountered in many other forms of cancer. The outlook in these diseases differs widely according to their type, so that acute leukemia in adults has a poor prognosis, while it is highly probable that some children with acute lymphoblastic leukemia will be cured. The earlier stages of Hodgkin’s disease (HD) can be cured and the more advanced forms effectively controlled (Kaplan, 1976), but it is the non-Hodgkin’s lymphomas and the diffuse forms of these diseases that often present difficulties in management. The problem can be stated more simply: where there are effective remedies then monitoring to produce early warning of relapse or resistance to a particular line of treatment could be helpful, such is the case in children’s leukemias and the chemosensitive tumors of children. When the treatments are eventually fairly ineffective, the early warning of deterioration may be only of academic interest to the physician, as in acute myeloid leukemias in children and adults. On the other hand, in non-Hodgkin’s lymphoma (NHL), which is often a protracted illness, monitoring may well influence clinical decision making as there are several combinations of therapy that can be tried and different timing as well. Furthermore, this approach may help in stratification of patients in clinical trial whose end point may often depend on opinions about whether the patient is in remission or has active disease, which is the absence of invasive investigations often lacks proof. Several APRPs have been proposed for the monitoring of Hodgkin’s disease; CPL and plasma copper levels can be raised but there is much controversy about their interpretation (Mortazavi et al., 1972: Hrgovcic et a / . , 1973; Foster e t al., 1973, although in HD Ray et al. (1973) found that the erythrocyte sedimentation rate (ESR) and serum CPL levels showed the best correlation with disease activity, and they give some examples of sequential studies to show this. On the other hand, Williams c’t ul. (1978) found neither the ESR nor serum copper levels a useful guide to disease activity in children with Hodgkin’s disease. This earlier work and that on Hp (Jayle e t al., 1968) and seromucoid (mainly (rlAGP) was often discounted as being clinically useful because levels could be influenced by factors other than the direct activity of the disease, and the variability of the response from one patient to another meant that the translation of the data derived on groups of patients to a given individual was difficult as the individual’s own normal limits were unknown. A study of Burkitt tumor in Nigeria by McFarlane and his colleagues (1967) is a good example of how measuring an array of plasma proteins including APRPs can indicate the way to proceed. In their study they compared
ACUTE PHASE REACTANT PROTEINS IN CANCER
21
the serum levels of CPL Hp, C-RP, and apM.They came to the conclusion that although the three APRPs were all highly significantly raised in active tumors the C-RP was the most useful as it completely disappeared from the serum when the patient was in remission. Plasma a p M levels were also depressed in the active tumors. They found there was no correlation between the various concentrations of APRPs in their series. By measuring the serum protein levels in the patient’s relatives, they could identify the levels to be expected in persons residing in a holendemic zone with chronic malaria and probably carrying various other parasitic diseases which are commonplace in t h e Burkitt tumor belt in Africa. More recently, Pepys e t al. (1978) reported that C-RP levels tend to discriminate symptomatic from asymptomatic HD: on the other hand, they found that the levels of serum amyloid P component, another member of the C-RP “family” of pentagonal disklike molecules, was little changed in HD. In the chronic leukemias in adults, there is usually a sustained elevation of Hp, alAT, and alAGP tending to be at their greatest levels when the disease is active. However, there is no correlation between the alAT level and the cell counts in chronic myeloid leukemia, suggesting the alAT is not elevated as a mechanism to protect the body from the putative excess of proteases in the vastly expanded pool of polymorphonuclear granulocytes (Child et al., 1977). On the other hand, a longitudinal analysis of the plasma protein profile in HD and NHL can provide patterns that appear to be related to various events in the evolution of the disease. It is of interest to note that Jayle et al. (1968) had described the evolution of the levels of serum Hp and seromucoid in 43 patients with HD observed serially for up to 8 years, clearly indicating that a persistently high Hp was a sign of bad prognosis; and in patients of longer survival the final phase of the illness was accompanied by erratic by progressively rising levels of these proteins for the last 1-2 years of their life. These studies were overlooked in the search for new ideas, and for a while the levels of serum copper were the center of attention. Our studies in lymphoma commenced with the serial measurement of a wide spectrum of serum APRPs (Child e t ul., 1978), but it soon became apparent that usually the patterns of the changes in the blood levels were concordant and highly correlated, which differs from the experience in Burkitt tumor (McFarlane et al., 1967). This has led to modifying the variate to be included in the monitoring array. It was shown by trial and error that the combination of the serum levels of C-RP, &-microglobulin (P2m), and the ESR could resolve several patterns in the lymphomas and chronic lymphatic leukemia (Amlot and Adinolfi, 1979; Cooper et a/., 1978; Cooper, 1979; Child et al.,
22
E. H. COOPER A N D JOAN STONE
1979). These three variates differ and yet are complementary. C-reactive protein is synthesized in the liver and is normally in low amounts in the sera ( < l o mg/liter) but rises very rapidly in response to injury and falls rapidly when the stimulus has been withdrawn. As with other APRPs, CRP can also exhibit sustained elevations in response to chronic stimulation. The p2m appears to be made by all nucleated cells and is an intrinsic part of the HLA (Poulik and Reisfeld, 1975); it has a low molecular weight ( 1 1,800), normal levels (0.8-2.4 mg/liter), and rapid turnover, with a fractional catabolism of about 1.4 per hour in health and in cancer patients (Karlson et al., 1978). The kidney appears to play the central role in the catabolism of p2m; more than 95% of the protein in the glomerular filtrate is reabsorbed by the proximotubular cells and metabolized (Wibell P I al., 1973). Hence alterations of renal function producing a raised creatinine are accompanied by an elevation of the serum p2m: however, an impaired renal function is a rare complication of lymphoma (Gagliano et al., 1976; Lynch et al., 1977). Nevertheless, transient rises of serum creatinine are common in sick patients, so it is always essential to measure the creatinine before attempting to interpret the significance of a raised serum p2m. Elevation of p2m has been observed in several forms of cancer (Kindt and Van Vaerenberg, 1976; Shuster e f a / . , 1976: Teasdale c>t al., 1977: Kin et a / . , 1977, as well as in various forms of chronic benign disease (Kin rf a l . , 1977). These earlier surveys showed a high probability of the p2m being raised in various lymphoid malignancies, which was the basis for our including this marker in the monitoring array. Finally, the ESR is the resultant of many factors which alter the physicochemical properties in the blood that influence red cell rouleaux formation and their rate of sedimentation. The fibrinogen content and the (r,AGP are correlated with the ESR, but the ESR appears to depend on other factors so that it can change either concordantly with the level of C-RP or may be dissociated as has been observed in the rheumatoid diseases (Amos ez a / . , 1977: Lancet Leader, 1977). Abd-el-Fattah et al. (1976) found that adding C-RP to blood in v i m had no effect on the ESR. Whatever its fundamental mechanism the ESR, when measured serially, still can provide useful information in monitoring HD; Le Bourgeois and Tubiana (1977) found the ESR gave an average lead time of 4.5 months over clinical assessment in detecting recurrence of HD stages I and I1 after, what would today be considered inappropriate, radiotherapy. The longitudinal studies have demonstrated that in some forms of active HD and NHL an upward trend of the levels of p2m, C-RP, and ESR may be correlated, either in untreated disease or recurrence. Frequently, the variates are discordant and the time course of each one produces several
ACUTE PHASE REACTANT PROTEINS IN CANCER
23
distinct patterns. The simplest pattern is seen in chronic lymphatic leukemia in which the p2m level is almost invariably raised during the course of the disease if the tumor burden produces symptoms. In those cases that progress and necessitate therapy, the values tended to be in the region of >8 mg/liter, the nonprogressive forms plateau <6 mg/liter and can probably remain at this level for several years. The ESR and C-RP are usually normal in chronic lymphatic leukemia. In the NHL, some patients with diffuse disease can achieve a metastable state for up to 2 or more years with chronic elevation of p2m and ESR but a normal CRP; others will exhibit a concordant rise of all the variants with disease activity, or this can be dissociated. A sustained abnormal level of C-RP or P2m seems to be a signal of residual disease but the interval between the rise and clinical recurrence is unpredictable, except that a rapid rise of either indicator is soon followed by clinical relapse. The pattern of CRP levels may oscillate markedly in some cases of advanced HD, who have received heavy chemotherapy. We have observed three such patients, all of whom had rising titers of antibody against BK virus, which may be a reflection of a particular immunological suppression from the disease and its treatment that allows these papova viruses to proliferate (Gardner, 1977). Finally, a practical advantage of using C-RP as a member of the panel is that it is possible to screen the sera rapidly with latex agglutination tests to obtain the approximate C-RP content and measure the concentration accurately by a nephelometric system within 3-4 hours. In our experience of closely monitored patients seen every 6 weeks, a very rapid rise of C-RP (2150 mg/liter) in a few days is more likely to be due to infection than the relapse of the lymphoma alone, unless the lymphoma has produced a major complication such as the rapid compression of a vital structure. IX. Peritoneal and Pleural Effusions
There have been two conflicting reports about the behavior of alAGP in malignant effusions. Rudman and his colleagues (1974) measured the concentration of alAGP and several other plasma proteins in noninflammatory effusions (heart failure, cirrhosis, etc.), nonneoplastic inflammatory effusions, and neoplastic effusions. The mean S.E. of the alAGP concentrations were 35 & 4, 65 7, and 130 k 13 mg/ml, respectively. They emphasized that the alAGP in the malignant effusions often had structural abnormalities indicated by a reduced sialic acid and hexoses content and variable hexosamine content. This change was considered to be a source of underestimation of the alAGP level when using radial
*
*
24
E. H . COOPER AND JOAN STONE
immunodiffusion assays. Indeed, this group believed that the alAGP concentration could have a diagnostic significance in identifying malignant effusions. On the other hand, Agostoni and Marasini (1973, in a smaller series of patients using immunodiffusion techniques, did not consider that malignant and inflammatory effusions could be distinguished. Booth rt al. (1977) using rocket techniques examined a wide spectrum of plasma protein in effusions and did not find them to have any discriminant effect. On the other hand, they found in patients with malignant effusions or a raised plasma CEA that the CEA in the effusion was often 3-4 times higher than in the blood. Pregnancy associated macroglobin did not exhibit this effect. The authors consider the raised CEA was produced by local secretion by tumor cells in the effusion. They thought the concentration of proteins in the effusions was partially a function of their size and partly a reflection of the nature of the lesion. The concentration of lower molecular weight proteins (PALB and alAGP) in noninflammatory effusions was about 30% of that in the serum rising to 70% in malignant and inflammatory effusions. We have examined the relative concentration of C-RP (mol. wt. 132,000) in the serum and pleural fluid in malignant effusions and find its concentration in the fluid is <5C% of that in the blood. As these patients had advanced cancers, the serum C-RP was usually 2 100 mg/liter at the time of measuring the levels in the effusions. Variants of alAGP have been observed in patients with various malignant diseases: among them is a glycoprotein of molecular weight about 50,000 with a lower sialic acid and higher fucose content than alAGP in Hodgkin’s disease (Abel and Good, 1966). This protein cross-reacts with (wlAGP,but it is known that mild hydrolysis of alAGP will not change its antigenicity (Krotoski and Weimar, 1965). Rudman et a / . (1972) have observed abnormal alAGP in the sera of patients with cancer. Bacchus (1977) has suggested that electrophoresis of seromucoid on polyacrylamide gels can be used to distinguish raised alAGP due to cancer and due to inflammatory or degenerative conditions. In their experience of 42 samples with raised seromucoid values, 21 out of 25 (82%) of those from patients with malignant diseases showed an increase of fractions with medium mobility, while 17 with nonmalignant disease showed increased bands in the slow mobility range. These observations have attracted little attention, possibily because of the time-consuming techniques needed to identify these variants; nevertheless, they would seem to merit further investigation. The introduction of crossed lectin affinity electrophoresis would seem to be a very suitable technique of identifying these changes. This has the advantages of the separation of proteins by electrophoresis and of separation of the glycosylated proteins according to their binding to the lectins (Bog Hansen e t al., 1975).
ACUTE PHASE REACTANT PROTEINS IN CANCER
25
X. Extrahepatic Synthesis and Concentration of Acute Phase Reactant Proteins
Although there is overwhelming evidence that the APRPs are normally synthesized in the liver (Koj, 1974; Putnam, 1975), there are observations that under exceptional circumstances extrahepatic synthesis is possible and that some tumors appear to concentrate certain APRPs. Gitlin and Perricelli (1970) have demonstrated that the human yoke sac produces AFP albumin, PALB, alAT, and transferrin. Localization of AFP and alAT have been demonstrated by the immunoperoxidase method in endodermal sinus tumors, which are rare cancers in children arising from the yoke sac. These tumors contain PAS diastase-resistant granules similar to those seen in the livers of patients with congenital alAT deficiency (Palmer et af., 1973; Palmer et al., 1976). The localization of several plasma proteins in the cells of endodermal sinus tumors has been confirmed by Tsuchida et al., (1976), who also demonstrated the synthesis of human alAT by one of these tumors when it had been transplanted as a xenograft into nude mice. It has been suggested that leucocytes may be a source of a, acid glycoprotein (Gahmberg and Anderson, 1978), but as yet this has not been studied in clinical medicine. Extracts of several human tumors and adjacent tissue have been shown to contain higher amounts of the protease inhibitors, aIAT, a s M , and antifibrin 111 than can be accounted for by the amount of blood in the tissue (Twining and Brecher, 1977). The mechanism by which this increased concentration of antiproteases is produced is unknown, although it is highly probable that the vessels in the vicinity of the tumor have an increased permeability. The inference that these antiproteases take some part in the control and reaction to proteolytic activity of the invading tumor cells is immediately apparent, but the precise balance in the system is far from clear. The distribution of alAT in normal, granuloma, and tumor tissue of rats has been studied recently by Ishibashi and his colleagues (1978). Unlike human aIAT, levels do not increase significantly in response to infection (Turner and Liener, 1977). The half-lives of Iz5I-a,AT was 25 and 33 hours in rats bearing granuloma and transplanted sarcoma, respectively. There was a marked incorporation of labeled a,AT by the granuloma and sarcoma compared to normal tissues, and this was greater than the uptake of albumin (ALB). However, the mechanism by which the a,AT concentrates in the inflammatory and tumor tissues was not clear. Among the normal organs examined the lung showed a preferential incorporation of a,AT. It was interesting to find the a,AT remained in a
26
E. H. COOPER A N D JOAN STONE
trichloroacetic acid-precipitable form, indicating it was not rapidly degraded in the tissues. The a,AT reaction with proteases results in the formation of complexes, which may be slowly reversible as a slow transfer of lZ5I-trypsin has been observed from alAT to a 2 M (Ohlsson, 1971). On the other hand, protease-a2M complexes are removed rapidly from the circulation, and there is evidence that a z M may act as a trap for proteases (Laurel1 and Jeppson, 1975). If this is true, it raises the question as to whether extravascular a,M complexes are removed from tumor tissues, possibly by macrophages. XI. Animal Tumor Systems
Study of the behavior of APRPs in response to various transplanted tumors can provide fundamental information about some of the pathways that may be involved. However, caution is advised before making a direct extrapolation to man, as the spectrum of stimuli induced by a transplanted tumor and the sensitivity of the liver to these stimuli may be different in man bearing a spontaneous tumor and its associated local and distant effects. Acute phase reactant protein have been studied mainly in rabbits and mice. In rabbits, Hp, Cp (Voelkel rt al., 1978), and C,RP, an analog of C-RP (Hokama et a!., 19691, are convenient markers of the APRPs. In the rat considerable attention has been focused on a z M which is absent in the sera of normal adult rats and appears in the serum in response to injury, pregnancy, and is present in fetal and neonatal animals. Koj (1974) has reviewed the literature on this protein and concludes i t arises in injured animals as a modification to a fetal biosynthetic pathway rather than the synthesis of a new protein. The a,-globulins in the serum of rats contain acid glycoprotein and another a,-glycoprotein (alAP) originating in the liver, which has the typical behavior of an acute phase reactant: this protein has been studied extensively by Darcy (1970). The recent studies of Voelkel et al. (1978) on rabbits bearing the VX, carcinoma appear to provide a new aspect on the control of the APRPs’ response in cancer. The VX, carcinoma produces large amounts of prostaglandin E2 (PGE,) in the plasma, and within a week of transplantation the host’s serum develops a blue color that becomes intense by 3-4 weeks. These workers demonstrated that the color was due to a marked rise in the CPL levels and this was accompanied by a concomitant increase in Hp without any change in the plasma ALB levels. This rise in CPL and Hp ran in parallel with the rise in serum PGEz and serum calcium. Giving indo-
ACUTE PHASE REACTANT PROTEINS IN CANCER
27
methacin, an inhibitor of prostaglandin synthesis, at the time of implanting the tumor completely prevented the hypercalcemia and largely inhibited the rise of CPL. They suggest that arachidonic acid, a metabolite of PGEz, might be responsible for the rise in the APRPs. The rise of CPL in rabbits bearing VX, carcinomas has been confirmed by Ungar-Waron rf al. (1978), who demonstrated that the levels reached were about 2-5 times those in pregnant rabbits. This concept is supported by the observation that systemically administered PGE in rabbit produces an elevation of H I T (Shim, 1976). There is also a fibrosarcoma in mice that produces PGE2 and hypercalcemia (Tashjian c t al., 1974) which could supply another model to test these relationships. XII. Haptoglobin Phenotypic Variation
Haptoglobins are a family of ap-globulins.Smithes (1955) demonstrated that with starch gel electrophoresis there are three major haptoglobin types (phenotypes) called Hp 1-1, Hp 2-1, and Hp 2-2. The phenotypes which can readily be identified, especially using gradient acrylamide gel electrophoresis (Baxter and Rees, 1974), are genetic markers which have been the subject of a great many population and geographical studies. Peacock (1966) had suggested that the Hp 2-1 and Hp 2-2 phenotypes confer resistance to leukemia. Larkin (1967) reported high Hp 1 - 1 gene frequency in females with colon cancer and reticulosis. Sezaki et al. (1973), reporting on Japanese experience, observed in the normal population a type 1-1 percentage of Hp gene frequency to be 6.1 and 0.300, respectively; in a leukemic population of 127 patients it rose to 12.1 and 0.359, respectively; a similar result was found in lymphomas. Moe (1970), describing experience of Norwegian children with malignant disease, could not confirm Peacock’s results. Papiha er af. (1974) restricted their studies to 30 patients with chronic lymphatic leukemia (CLL) in an English population and found the Hp 11 frequency of 0.425 in a normal population and 0.483 in the CLL which is not a significant difference. They summarized several of the papers on this topic indicating that there is little convincing evidence of any association between haptoglobin phenotype and susceptibility to leukemia. However, they point out the local gene frequencies can vary considerably, and this can also have profound influence in population studies involving multiracial groups. The Japanese experience needs repeating, but when confronted with the wider American and European studies it may not be sustained. However, it must be remembered there is a major difference in the frequency of the various types of leukemia in Japan
28
E. H . COOPER A N D JOAN STONE
compared to those in the Western world (Wynder and Hirayama, 1977). In the broader field of solid tumors, repeated studies from various parts of the world with reasonably pure ethnic groups, such as women with breast cancer in Poland (Jaegermann et al., 1970), cancer patients in a German clinic (Pietrek and Kindler, 1971), American women with ovarian cancer (Mueller P t al., 1971), or lung cancer in Marseilles (Favre c>t al., 1972), have failed to show any significant relation of Hp phenotype and cancer. We have a similar experience in a survey of the Hp phenotypes of patients with gastrointestinal cancer and hematological malignancies in Yorkshire. XIII. The Biological Effects of Acute Phase Reactant Proteins in Cancer
This is largely a subject for conjecture with many items of information on which to form hypotheses but little in the way of sound evidence. The overriding question is whether the production of raised levels of plasma APRPs in cancer confers any benefit to the host. In clinical terms it would seem that the majority of patients at some time during the evolution of advanced cancer will have an elevation of APRPs. The association between a progressive rise of the APRPs and a rapid evolution of the cancer suggest that the levels of APRPs tend to be conrolled by the production of signals at the tumor-host interface, the signal strength being related to the depth of penetration of the tissue, as seen in the cervix and bladder. On the other hand, when progression is temporarily slowed by therapy the levels of APRPs may return to within normal limits, normally the upper quartile of the range, and can remain in this metastable state for many months prior to rising again as the tumor once more becomes clinically aggressive. There is plenty of evidence that the signals influencing the levels of APRPs in cancer are not exciting changes in all proteins. This is clearly demonstrated by the dissociation of changes in the antiproteases with the levels of alAT and alACT increasing while that of a 2 M remains almost unchanged. A similar dissociation of response is seen with C-RP which is known to be synthesized in the liver (Hurlimann rt ul., 1966) and plasma protein SAP (Wegelius and Pasternack, 1976) (amyloid P component) which has a similar pentagonal disc-like molecular structure to C-RP and may be synthesized in the liver (Pepys rt af., 1978; Lev0 et ul., 1977; Osmond et al., 1977). Hemopexin, a heme-binding serum protein, which is synthesized in the liver, is another example of a biosynthetic system that is independent of the influences that change the levels
ACUTE PHASE REACTANT PROTEINS IN CANCER
29
of the APRPs in the majority of cancers (Muller-Eberhard and Liem, 1975). The levels of PALB and transferrin often tend to fall as other APRPs in the majority of cancers (Muller-Eberhard and Liem, 1975). The levels of PALB and transferrin often tend to fall as other APRPs rise in cancer, as well as in other forms of injury such as wounding, burns, rnyocardiol infarction, or chronic infections, which suggests the control of the levels of these proteins is affected by the same family of signals as those producing increased production of the APRPs. The biological activity of some of the APRPs is well defined, such as hemoglobin binding by Hp, the transport of copper by CPL, and the antiprotease activity of aIAT and alACT. As far as we know the failure to elicit a full alAT response in cancers in patients with heterozygous MZ and M S alAT phenotypes (about 1% of the population) is not associated with an especially florid form of cancer (A. M. Ward, personal communication). Indeed the clinical diseases related to alAT are associated with the deficiency of this protein rather than its chronic excess. Furthermore, the lack of direct harm arising from the chronic elevation of the APRPs in physiological states such as pregnancy or lactation or by sustained treatment with estrogens indicates that this component in the response to cancer is not intrinsically harmful as compared to the steady fall in serum ALB that accompanies weight loss in the cachexia of cancer. The APRPs that can modify immunological responses and the function of other protein cascades, such as the complement system and blood clotting, pose the far more interesting question of whether their action modifies the behavior of the tumor or purely generates random noise in an organism that has lost some of the fine control in its homeostatic mechanisms. Some APRPs can inhibit certain lymphocyte responses, including lymphocyte blast cell transformation induced by mitogens and mixed lymphocyte reaction: these include C-RP (Mortensen et af., 1975; Mortensen and Gerwurz, 1976), aIAGP (Chiu rt al., 1977, as well as other proteins including immunoregulatory alpha-protein (Menzoian rt al., 1974) and AFP (Murgita and Tomasi, 1975; Yachnin, 1975). However, the significance of these in vitro inhibitory activities on lymphocyte function in vivo is still uncertain. C-reactive protein has a particularly wide spectrum of activity in vitro as it can react with platelets (Fiedel and Gerwurz, 1976) and also activate complement by the classical pathway (Kaplan and Volanakis, 1974 Siegel et al., 1974 Siegel et al., 1975). The abnormal levels of alAGP, by their ability to bind drugs, may have consequences on the pharmacokinetics of various drugs. It has been shown that the rise of q A G P concentration paralleled binding of quini-
30
E. H. COOPER AND JOAN STONE
dine in postoperative patients (Fremstad et al., 1976): and alprenolol, a beta-blocker, and imipramine, a tricyclic antidepressant, are avidly bound by a,AGP: and this drug-binding is not correlated to the serum albumin concentration (Piafsky and Borga, 1977). However, the plasticizer (tris 2-butoxyethyl phosphate) present in Vacutainer tubes, commonly used for blood sampling, interfere with this drug-binding by q A G P (Borga et af., 1977). Fisher and Shifrine (1978) have proposed a hypothesis for the raised serum copper in cancer patients. In their scheme they propose that it is the catabolism of CPL that controls the copper level, and if the catabolism of CPL was slowed, it would cause the copper to rise. Their hypothesis suggests that there is a shift in the balance between the desialylation produced by hepatic neuraminidase activity and the resialylation produced by sialyl transferase activity, and an associated concomitant rise in blood sialic acid levels. However, this would imply a net prolongation of the half-life of the protein which has yet to be established. There is a good evidence that increased sialic transferase activity can be detected in the plasma in several forms of cancer (Bossmann et al., 1973: Bossmann and Hall, 1974) as well as raised sialyltransferase in human, mammary, and colonic carcinomas. However, the way in which q A G P appears to be modified in some cancer patients (see above) suggests the structure of APRPs, especially the sialylation and glycosialylation, could be changed as a result of contact with the cells bearing abnormal enzyme complements on their surfaces. XIV. Mathematical Addendum
Any attempt at mathematical analysis of serum APRP levels either individually or the profile of a combination of them is beset with many difficulties. Ascribing number values to the weight of plasma proteins in biological fluids is notoriously difficult, as the quality control and reference standards are not uniform to a high degree of precision. The assay techniques used vary between laboratories and the results from any particular laboratory have to be assessed with a knowledge of the reliability of the methodology used there and of its comparability with that of other sources. Landaas et al. (1978) report on the repertoire of serum protein analyses, the methods and reagents that are used, and the analytical variation among Scandinavian clinical chemical laboratories. Thirteen serum proteins were analyzed by various methods and the analytical results varied considerably between the laboratories. For most proetins the interlaboratory coefficient of variation was between 20% and 30%
ACUTE PHASE REACTANT PROTEINS IN CANCER
31
and the ratio of the highest to the lowest value was about 3. There was also considerable variation in results between analytical methods. They recommend that all clinical chemical laboratories agree on the use of a common reference standard with defined concentrations of the major serum proteins, that reference material be a human serum pool collected from so many individuals that the composition will be essentially the same when a new pool is collected years later, and a better standardization of the titer and specificity of commercial antisera. A proper interpretation of APRP data from cancer-bearing patients can be made only with a detailed knowledge of data produced for matched subjects in normal health by exactly the same methodology. Many factors need to be taken into account in establishing the distribution of values of any APRP occurring in normal health: reproducibility of any measurement; possible variability of level in an individual patient when samples are taken at different times of day, on different days, and in different weeks; and a general variation in level due to sex, age, or race, for example. The Hp and alAT values depend on phenotype, for instance. Large-scale data collection for normal-health subjects is needed to establish the shape of the distribution. A normal distribution presents no problems, but a skew distribution calls for further consideration. C-reactive protein is particularly difficult, as values are so low as to be undetectable by routine techniques in about half the normal population, though they can be detected by radioimmunoassay to two orders of magnitude lower than routine immunodiffusion (Kindmark and Thorell, 1972). A transformation of the data to a different measuring scale may produce a normal distribution, and this is the usual approach. An account of frequently used transformation is given by Armitage (1977). A common change of scale is to take logs in order to produce the log-normal distribution. Aitchison and Brown (1969) give a full account of the assumptions underlying such a distribution and its properties. Having established the distribution to be used as the reference for normal health, estimates can be made of any of its parameters which are of interest, probably the mean or median, and the variance, and a “normal range” of values, within which, say, 95% of all values can be expected to lie is estimated. Reed er al. (1971) criticize this reliance on the assumption of a specific type of distribution for the establishment of normal-range values and show that nonparametric methods, which make no assumption about the shape of the distribution, produce normal-range values which are, for all practical purposes, as accurate as estimates that assume the distribution of values to be normal, and that when values are not distributed as assumed, nonparametric estimates, which rely on ranking the values, are more accurate. As an example they use nonparametric
32
E. H . COOPER A N D JOAN STONE
methods to establish normal-range values of Hp from a sample of 100 apparently healthy persons. A comprehensive account of nonparametric tests can be found in Siegel’s book (1956). The changes of level of APRPs appearing in the presence of cancer are dependent on the site and stage of the tumor. Examples of the way the levels of these proteins may alter have been given in the earlier part of the text. The distribution of values of an APRP in any cancer-bearing group can present difficulties of interpretation. It is not always possible to be precise about which group a patient should be allocated to, in terms of staging, and a wide-spread distribution can be due to wrongly allocated patients, to large inherent variation in the values of patients with the same condition, or to the wide variety of patient conditions coming under one clinical description. This is especially true in “metastatic cancer,” which is a term embracing states as widely different as a solitary metastasis and a terminal illness with generalized carcinomatosis. Studies involving large numbers of patients within each cancer type, and within each clinically defined subgroup in the cancer type, are needed if it is desirable to establish the pattern of values of APRPs for each group. Evidence from the small groups of patients in the literature indicates that very early stage cancer patients have values with distributions not too different in shape from those of normal health, although there is a wider dispersion and a general elevation in level. More advanced cancers have a big dispersion of values, with an ill-defined distribution, and in fact, are possibly a mixture of distributions of subgroups, which are not detectable clinically or have been deliberately amalgamated by investigators. Visual inspection of the pattern of values in various cancers should suffice to establish that changes do occur in the presence of cancer: the mathematical task is to quantify the changes and to investigate if they can be used for diagnostic and prognostic purposes. The possibility of being able to predict staging or cancer load or prognosis from a particular APRP value seems remote, although one should not underestimate the possible potential of the information that could be acquired by having data from a large number of patients, even if only within one subgroup of one cancer type, and cooperative studies between centers to produce such data would be rewarding. APRPs cannot be specific markers of cancer. However the simultaneous assessment of several APRPs and their interrelations with tumor-related antigens or other general biochemical changes in cancer may provide information which is not otherwise available to the clinician. Once more, as much knowledge as possible is needed about the data from normal-health subjects before embarking on a study of data obtained
ACUTE PHASE REACTANT PROTEINS IN CANCER
33
from cancer-bearing patients. If analysis is to be distribution based, the usual product-moment correlation coefficient can be computed for pairs of APRPs. The early studies have shown that the correlations between pairs of plasma proteins in health are not very high (Table VII). Nonparametric analysis would be based on rank-correlation methods. Kendall (1962) gives an account of these; and their application to the correlations of alACT, aIAT, and aIAGP in various cancers is described by Kelly et a / . (1978).
MULTIVARIATE METHODS Exploratory data analysis of single APRPs and their correlations with other APRPs is essential before launching into multivariate methods which will enable us to deal with simultaneous data relating to various APRPs. A word of caution is needed here. There are numerous computer "packages" available now which provide the facility for performing several types of multivariate analysis on one set of data, and there is a great danger of a researcher feeding in the data and hoping that among all the analyses something "useful" will appear. The interpretation of such analyses must be done with a thorough knowledge of the assumptions underlying them, and it is essential to have expert statistical advice for this. In order to compare the profiles of combinations of APRPs in different patients, the patients must be in corresponding "states," not only in terms of cancer type and stage, but also in terms of the point in time they have reached in the evolution of the disease and its therapy. The points at which patients could be compared are pretreatment, at a fixed time after specific therapy, and at other decision-making points in the history of the disease. Points close to death (that is within one month) are to be avoided due to the breakdown of the homeostatic control systems. There are many multivariate methods for analyzing this type of data, and Kendall (1975) gives an account of the main distribution-based techniques: nonparametric methods are described by Puri and Sen (1971). Three techniques that could be useful in context of analyzing plasma protein data are described briefly below. I . D i s c r i m i m n t Analysis
Discriminant analysis is widely used, and Solberg (1975) has reviewed its application in clinical chemistry. A linear discriminant function is a linear combination o f p variables xl,
TABLE VII CORRELATION OF PAIRS OF PLASMA PROTEINS PALB Prealburnin Albumin a,-Antitrypsin a-Lipoprotein Orosomucoid a2-Macroglobin Haptoglobin Ceruloplasmin Transfenin Hemopexin Plasminogen &-Globulin /3,E-Globulin Fibrinogen
0.394 0.012 0.263 0.082 -0.444 0.302 -0.228 0.253 0.180 0.325 0.137 0.165 -0.158
From Ganrot (1972).
ALB
a,AT
-0.098 0.037 0.078 0.150 0.047 -0.132 0.189 0.003 0.174 0.124 0.352 0.073 0.195 0.142 -0.005 -0.116 -0.114 0.062 -0.024 0.085 0.128 -0.152 0.336
a-Lip.
Oroso.
a2M
-0.262 0.146 -0.176 0.209 0.159 0.115 0.170 -0.064 -0.187 -0.066
0.124 0.634 0.258 0.026 0.470 0.214 0.572 0.464 0.431
-0.069 0.220 -0.179 0.036 0.183 -0.029 -0.090 0.252
IN
Hp
0.072 0.007 0.166
0.343 0.405 0.379 0.485
NORMALHEALTHYSUBJECTS" CPL
Tf.
Hemop.
PI.-g.
p[
0.255 0.169 0.118 0.333 0.093 0.419
0.095 0.315 0.165 -0.067 0.021
0.278 0.363 0.166 0.142
0.295 0.207 0.166
0.586 0.499
PIE Fibr.
0.351
ACUTE PHASE REACTANT PROTEINS IN CANCER
35
...,x , (in this context, APRPs) measured on each patient in order to obtain an “index,” whose value determines to which group the patient should be assigned. The groups are determined initially according to criteria independent of the x’s, for instance, response to a specific treatment, and that discriminant function of the x’s which best separates the groups is established. A new patient’s index then allocates him or her to the appropriate response group. The probability of misclassification can be estimated for any set of data. Werner et al. (1972) use stepwise discriminant analysis for ascertaining the diagnostic effectiveness of specific protein assays. Agostoni el al. (1974) use discriminant analysis on data obtained from serum protein assays in chronic hepatitis and postnecrotic cirrhosis. Milford Ward el al. (1977a, b) use a stepwise approach to set up a discriminant function using logistic discriminating for separating patients with large bowel cancer into those who remained tumorfree for at least one year after surgery and those who developed metastases, the discriminant variables bieng log,, CEA, alAT, and a,AGP. They also used discriminant analysis to separate patients with prostatic cancer into those with and without metastases (Fig. 2). Linear discriminant analysis relies on the assumption that the parent populations relating to the discriminant groups have a common dispersion matrix. If there are more than two discriminant groups, an additional assumption is needed, that each group is multivariate normal. Both assumptions are somewhat drastic and unlikely to be met in practice. The abandonment of the requirement of a common dispersion matrix leads to a quadratic discriminating boundary, which is difficult to deal with in more than two dimensions, and this situation can be avoided by reducing the discriminator to two dimensions or by using nonparametric methods. Also, Kendall (1975) suggests the following:
x,,
One of the most interesting approaches to the discrimination problem now made possible by the computer is the display on a television unit of arbitrary projections of a constellation of points on to arbitrarily chosen planes. With the use of a lightpen and the virtually instantaneous recalculation of the necessary statistics, it is possible to explore the effect of removing outliers, discarding variables and so forth in a way which gives valuable insight into the data and the consequences of various ways of handling them.
2. Principal Component Analysis If there are observations of the p variables xl, x 2 ..., x, on each of individuals, we can set up linear combinations of the x,’s,
+ a12x2 + = azlxl + a2,xz +
y1 = y,
ClllXl
a l,XD
a2,,xpand so on
17
36
E. H. COOPER A N D JOAN STONE
The a’s are chosen in such a way that the y’s are uncorrelated,
and y 1 has the highest possible variance, and so represents better than any other linear combination of the x’s the difference between individuals. y z will have the next highest variance and so on. Although there are in general p principal components, the variation of all but a few will be small and attention can be centered on those few. If the original variables are normally distributed, the principal components are independent, as well as being uncorrelated. Principal component analysis would appear to be a useful tool for handling combinations of protein values in cancer patients, but it has a serious drawback which is that the components are not independent of the scales on which the original variables are measured. In a situation where some protein levels, such as Hp, increase about Ifold in the presence of advanced cancer, while the level of C-RP protein may increase 100-fold, this presents a difficulty. It can be tackled by standardizing the variables to have unit variance. When the principal components have been obtained, individuals can be ranked according to the value of the first one, and it may be possible to associate the ranking with a known grouping of individuals according to some attribute, say, evolution of the disease in cancer patients.
3. Cluster Anulysis This is a technique that sets up criteria for separating initially ungrouped data into “clusters.” Different criteria produce different cluster groupings, and the interpretation of these groupings in terms of the individuals being measured is left to the subjective judgment of the investigator, in the light of his or her own specialized knowledge. Everitt (1974) gives an account of various cluster analysis techniques. Cluster analysis may prove to be a useful tool in identifying subgroups in a specific cancer patient group, from their APRP values. Having exploited multivariate techniques for analyzing simultaneous data collected on patients, attention has to be directed at longitudinal studies over time of individual measurements relating to one patient. Ray et ul. (1973) have illustrated longitudinal studies of CPL in lymphoma. Milford Ward et al. (1977) give examples of APRP profiles measured sequentially in colorectal cancer after surgical treatment of the primary tumor, Cooper et al. (1978) report on serial measurement of p2m, APRPs, and ESR in non-Hodgkin’s lymphomas and chronic lymphocytic leukemia.
ACUTE PHASE REACTANT PROTEINS IN CANCER
37
Interpretation of measurments of an individual APRP over time, let along several APRPs, is a formidable task. From the limited amount of data that has been available to date, there is no prospect of associating a particular pattern of change of an APRP over time with a corresponding pattern of change in the stage of disease, although it is true to say that the clinicians with whom we are working are beginning from experience to recognize that sequential measurements after the initial treatment of a neoplasm tend to fall into recognizable patterns: (a) A stable value, within normal limits, is a feature of a patient whose condition does not deteriorate and may have been cured (b) A period during which the APRP value is normal, followed by a rise is often seen in the slower growing recurrence or metastases (c) A period during which the APRP is metastable about a value at the top of the normal range followed by a rise tends to accompany obvious residual disease (d) A steady rise in APRP levels associated with a progressive deterioration of the patient's condition (e) A completely erratic pattern: this is often indicative of the influence of secondary events such as infection in the lymphomas Sequential data of this kind, with succeeding values being highly correlated in general might be tackled by some kind of time series analysis (see, for instance, Hannan, 1963, but this approach would need the study of a very large number of patients with the same cancer type and stage followed in detail over the same period of progression of the disease. A pattern in which the variates stay normal and then make a sudden change leads one to speculate on the possibility of applying catastrophe theory to the problems of analyzing the behavior of APRPs in cancer (see Poston and Stewart, 1976). In any multivariate analysis the mathematician is always conscious that unknown factors and consequently omitted factors may be of vital importance. And such a situation may well be the case in contemporary studies of plasma proteins in cancer, as relatively little is known about the role of some of these proteins in health and disease.
REFERENCES Aagenaes, O . , Fagerohol, M . , Elgio, K . , Munthe, E., and Hovig, T. (1974). Postgrud. Med. J . 50, 365-375. Abd-El-Flattah, M . , Scherer, R., and Ruhenstroth-Bauer, G . (1976). Kliri. Wschr. 54, 169171.
38
E. H . COOPER A N D JOAN STONE
Abet, C. A., and Good, T . A. (1966). Clin.Chiin. Acta 14, 802-806. Ablin, R. J . , Gonder, M. J., and Soames, W. A. (1973). J. Urol. 110, 238-241. Abraham, K., and Schiitt, K. (1970). Z. Klin. Chem. Klin. Biochetn. 8, 92-98. Agostoni, A., and Marasini, B. (1977). A m . J. Clin. Pathol. 67, 146-148. Agostini, A., Marasini, B., Stabilini, R., Del Ninno, E., and Pontello, M. (1974). Clin. C h e m . 20, 428-439. Aitchison, J., and Brown, J. A. C. (1969). It1 "The Lognormal Distribution," Cambridge Univ. Press. Amlot, P. L., and Adinolfi, M. (1979). Eur. J . Cancer (in press). Amos, R. S., Constable, T.J., Crockson, R. A., Crockson, A. P., and McConkey, B. (1977). B. Med. J . 1, 195-197. Armitage, P. (1977). "Statistical Methods in Medical Research," Blackwell, Oxford. Aronsen, K. F., Ekulund, G., Kindmark, C. O., and Laurell, C. B. (1972). Scund. J . Clin. Lab. Invest. 29, Suppl. 124, 127-136. Ashwell, G. (1974). Biochem. Soc. S v m p . , Vol. 40, 117- 124. Ashwell, G., and Morrell, A. G. (1974). A d v . Enzymol. 41, 99-128. Bacchus, H. (1977). Crit. R e v . Clin. Lab. Sci. 8, 333-362. Bagshawe, K. D. (1969). "Choriocarcinoma." Arnold, London. Bagshawe, K. D., and Searle, F. (1977). Essuys Meti. Biochem. 3, 25-73. Barbotte, P. J . , Bali, J. P., Heintz-Remy, N., and Balmes, J. L. (1976). B i d . Ga.stroentero/. (Pnris) 9, 233-241. Bartsch, W., Horst, H. J., Becker, H., and Nehse, G. (1977). Actu Endocrinol. 85, 650664. Bastable, J . R. G., Richards, B., Haworth, S . , and Cooper, E. H . (1979). Br. J. Urol. (in press). Baxter, S. J., and Rees, B. (1974). Med. Sci. Law 14, 231-236. Bodansky, 0. (1975). "Biochemistry of Human Cancer." Academic Press, New York. Bohn, H . (1974). ArX. Gynttkol. 217, 219-231. Bohn, H., and Kranz, T. (1973). Arch. Gynaekol. 215, 63-71. Bonadonna, G. (1977). I n t . J . Rucliat. Oncol. Biol. Ph.vs. 2, 237-240. Booth, S. N., Lakin, G., Dykes, P., Burnett, and Bradwell, A. R. (1977). J. Clin. Puth. 30, 537-540. Borga, O., Piafsky, K. M., and Nilsen, O.G. (1977). Clin.Pharm. und Therup. 22, 539544. Bossmann, H . B., and Hall, T. C. (1974). Proc. N u t / . Acud. Sci. U . S . A . 71, 1833-1837. Bossmann, H. B., Bieber, G. F., Brown, A. E., Case, K. R., Geisten, D. M., Kimmera, T. W., and Lione, A. (1973). Nutitre (London) 246, 487-489. Bullen, B., Cooper, E. H., Turner, R., Neville, A. M., Giles, G. R., and Hall, R. (1977). Med. Peti. Oncol. 3, 289-300. Carter, S. K. (1976). J. N u t / . Cancer Inst. 56, 3- 10. Child, J. A., Roberts, B. E., Illingworth, S., and Cooper, E. H. (1977). Biornerlicine 27, 188-190. Child, J. A., Cooper, E. H . , Illingworth, S., and Worthy, T. S. (1978). Recent Prog. Cancer R e s . 64, 180-189. Child, J . A., Spati, B., Illingworth, S., Barnard, D., Corbett, S., Simmons, A. G . , Stone, J . , Worthy, T . S. , and Cooper, E. H. (1979). Ctrncer (in press). Chiu, K. M., Mortensen, R. F., Osmond, A. P., and Gewurz, H. (1977). Immunology 32, 997- 1005. Chu, T. M. (1978). Antibiot. Chernother. 22, 98-104.
ACUTE PHASE REACTANT PROTEINS IN CANCER
39
Chu, T. M., Wang, M. C., Kuciel, R., Valenzueca, L., and Murphy, G. P. (1977). Cancer Treat. Rep. 61, 193-200. Colleta, D. M. (1978). Antibiot. Chernother. 22, 166-175. Coombes, R. L., Powels, T. J., and Neville, A. M. (1977). Proc. Roy. Soc. Med. 70, 843845. Cooper, E. H. (1979). Recent Results Can. Res. 67, 97-103. Cooper, E. H., Turner, R., Geekie, A., Neville, A. M., Goligher, J. C., Graham, N. G., Giles, G. R., Hall, R., and Macadam, W. A. F. (1976). Biomedicine 24, 171-178. Cooper, E. H. (1978b). I n “Clinical Application of Carcinoembryonic Antigen Assay” (B. P. Krebs, C. M. Lalanne and M. Schneider, eds.), pp. 397-407. Excerpta Medica, Int. Cong. Ser. 439, Cooper, E . H . , and Neville, A. M. (1978). In “Gastrointestinal Tract Cancer” (M. Lipkin and R. A. Good, eds.), pp. 437-455. Plenum Medical, New York. Cooper, E. H., Bunning, R., Illingworth, S., Spati, B., and Child, J . A. (1978). Biomedicine 29, 154-158. Cowen, D. M., Searle, F., Milford Ward, A., Benson, E. A., Smiddy, F. G., Eaves, G., and Cooper, E. H. (1978). Eur. J. Cancer 14, 885-894. Crawley, J . M., Northam, B. E., King, J. P. G., Leonard, J. C., Booth, S. N., and Dykes, P. W. (1974). J . Clin. Path. 27, 130-134. Darcy, D. A. (1970). Brt. J . Exp. Path. 51, 51-59. Douma, G. J . , and van Dalen, A. (1974). Z. Klin. Chern. Klin. Biochern. 12, 474-476. Elson, L. A., Betts, T. E., and Darcy, D. A. (1974). In “Advances in Tumor Prevention, Detection and Characterization” (W. Davies and C. Maltoni. eds.), pp. 151- 153. Vol. I , Excerpta Medica. Congress Series. Eriksson, S. (1965). Acta Med. Scand. (Suppl.) 432, 1-80. Eriksson, S., and Hagerstrand, I. (1974). Acta Med. Sctrnd. 195, 451-458. Everitt, B. (1974). “Cluster Analysis.” Heinemann, London. Favre, R., Poirier, R., Kleisbauer, J . P., Arnoux, A,, and Lava], P. (1972). Poutnon Coeur 28, 165-167. Fisher, B. (1977). Med. Clin. North. A m . 6 , 953-965. Fiedel, B. A., and Gewurz, H. (1976). J . Imrnunol. 116, 1289-1294. Fisher, C. L., and Gill, L. W. (1975). “Serum Protein Abnormalities” (S. E. Ritzmann and J. L. Daniel, eds.). Little, Brown, Boston. Fisher, L., and Shifrine, M. (1978). Oncology 35, 22-25. Foster, M., Dawson, A., Pockington, T., and Fell, L. (1977). C h . Radio/. 28, 23-28. Freeman, T. (1964). Prof. B i d . Fluids 12, 344-347. Fremstad, D., Bergerud, K., Haffner, J. F. W., and Lunde, P. K. M. (1976). Eur. J. Clin. Phurtnacol. 10, 441-444. Gagliano, R. G., Costanzi, J. J., Beathard, G. A., Searles, H. E., and Bell, J. D. (1976). Amer. J . Meti. 60, 1026-1031. Gahrnberg, C.C., and Anderson, L.C. (1978). J. Exp. Med. 148, 507-521. Galen, R. S. (1975). Prog. Clin.Pathol. 6 , 83- 110. Ganrot, P. 0. (1972). Scand. J . Clin.Lab. Invest. 29, Suppl. 124, 83-88. Ganrot, P. 0..Laurell, C. B., and Eriksson, S. (1967). Scrrnd. J . Clin. Lab. Invest. 19, 205-208. Gardner, S . D. (1977). R r c . Advances Clin. Virul. 1, 93- 115. Gitlin, D., and Perricelli, A. (1970). Nature (London) 28, 995-997. Gottschalk, A., Whitten, W. K., and Graham, E. R. B. (1960). Biochirn. Biophxs. Acta 38, 183-184.
40
E. H . COOPER AND JOAN STONE
Goverde, D. C., Veenkamp, F. J. N., and Human, J. D. H. (1968). Actu Endocrinol. (Copenhagen) 59, 105- 119. Hallen, J . and Laurell, C. B. (1972). Scand. J . Clin. Lab. Invest. 29, Supp. 124, 97- 103. Hannan, E. J . (1967). “Time Series, Analysis.” Methuen, London. H a m s , C., Primack, A., and Cohen, M. H. (1974). Carrcrr 34, 280-281. Harshman, S . , Reynolds, V. H., Neumaster, T., Patikas, T., and Worrall, T. (1974). Curicer 34, 291-299. Hiramatsu. S., Kojima, J., Okada, S., Inai, S., and Ohmori, K. (1976). Acta HrptrtoGastroenterol. 23, 177- 182. Hobbs, J. R. (1974). Lancet ii, 1305- 1307. Hokama, Y., Yamaka, K., Scheller, C., and Nishimura, E. T. (1970). Cancer Res. 30, 1319- 1326. Hollinshead, A. C., Chuang, C. Y., Cooper, E. H., and Catalona, W. J. (1977). Cuncer 40, 2993- 2998. Houghton, A. L. (1979). M. S. Thesis Univ. of London. Houghton, A. L., Turner, R., and Cooper, E. H. (1977). Br. J . Urol. 49, 227-232. Hrgovcic, M., Tessmer, C. F., Thomas, F. B., Ong, P. S., Gamble, J. F., and Shullenberger, C. C. (1973). Cancer 32, 1512-1524. Hrgovcic, M., Tessmer, C. F., Brown, B. W., Wilbur, 3. R., Mumford, D. M., Thomas, F. B., Shullenberger, C. C., and Taylor, G. (1973). f r o g . Clin. Caricrr 5, 121-153. Hurlimann, J . , Thorbecke, G., and Hochwald, G. (1966). J . Exp. M r d . 123, 365-378. Hsu, C. C. S., and LoGerfo, P. (1972). froc. Soc. Exp. B i d . Mrd. 139, 575-578. Ishibashi, H., Shibata, K., Okubo, H., Kawamura, Z.. and Yanase, T. (1978). J . Lcrb. Clin. Mrd. 91, 576-582. Jaegermann, K., Glinska, H., Turowska, B., Urban, A., and Pawlicki, M. (1970). Nowotor? 20, 261-267. Jayle, M. F., Ennuyer, A., Chemama, R., Bataini, P., and Detroux, L . (1968). Sern. Hop. f u i s 44, 1861-1867. Johson, A. M., and Alper, C. A. (1970). fedintries 46, 921-925. Jonsson, G., Olsson, A. M., Luttrop, W., Cekan, Z., Purvis, K., and Diczfalusy, E. (1975). Vitarn. Horrn. 33, 351-376. Kaplan, H. S. (1976). Crrrrcrr Rrs. 36, 3863-3878. Kaplan, M. H., and Volanakis, J. E. (1974). J. Irnmrmol. 112, 2135-2147. Karlsson, F. A., Sege, K., Beauduin, M., Pluygers, E., Wibell, L., Groth, T., and Peterson, P. A. (1978). Sprnp. Brtrr 2-Microglo. frolf. Disord. (Exc.), pp. 31-47. European Press, Gent, Belgium. Kelly, U. L., Cooper, E. H., Alexander, C., and Stone, J. (1978). Riornrdicirir 28, 209215. Kendall, M. G. (1962). “Rank Correlation Methods.” Griffin, London. Kendall, M. G. (1975). ”Multivariate Analysis.” Griffin, London. Keuppers, F., and Fallat, R. J. (1969). Clin. Chirn. Actu 24, 401-403. Kew, M . C., Turnbull, R., and Prinsloo, I. (1978). Br. J . Crrricer 37, 635-638. Kin, K., Sakmalayaski, I., and Kawai, T. (1977). Gcrnrr 68, 427-434. King, R. B. (1978). Aririhiot. Clrernorhcr. 22, 47-58. Kindmark, C. O . , and Thorell, J. I. (1972). Surrid. J . Cliri. Ltrh. Irivest. 29, (Suppl.) 124, 49-53. Kinot, R.,and Van Vaerenberg, P. M. (1976). Actu Clin.Brlg. 31, Supp. 8, 33-37. Kohn, J., Hernandez, M., and Riches, P.G. (1978). Ricerctr C l h . L d . 8, Supp. 1, 61-70. Koj, A. (1974). Iri ”Structure and Function of Plasma Proteins” (A.C. Allison, ed.), pp. 72-131. Plenum, London.
ACUTE PHASE REACTANT PROTEINS I N CANCER
41
Koskelo, P., Kekki, M., Nikkila, E . A., and Virkkunen, M. (1967). Scund. J. Clin. Lab. Invest. 19, 259-262. Krauss, S. (1969). Blood 33, 865-876. Krauss, S . , Schrott, M., and Sarcione, E. J. (1966). A m . J. Med. Sci. 252, 184-191. Krebs, B. P., Lalanne, C. M., and Schneider, M. (1978). “Clinical Application of Carcinoembryonic Antigen.” Excerpta Medica, Amsterdam. Krotoski, W. A., and Weimer, H. E. (1965). Fed. Proc. 24, 506. Landaas, S., Skrede, S . , and Eldjarn, L. (1978) Scand. J. Clin. Lab. Invest. 38, 295-299. Larkin, M. F. (1967). J. Nail. Cancer Inst. 39, 633-638. Latner, A . L., Turner, G. A., and Lamin, M. M. (1976). Oncology 33, 12-14. Latner, A. L., Longstaff, E., and Turner, G. A. (1974). Br. J. Cancer 30, 60-67. Laurell, C. B. (1974). J . Clin.Path. 28, Supp. 6, 22-26. Laurell, C. B., and Jeppesson, J . 0. (1975). In “The Plasma Proteins” (F. W. Puttnam, ed.), Vol. 1, pp 229-264. Academic Press, New York. Laurell, C. B., Kullander, S., and Thorell, J. (1968). J . Clin. Lab. Invest. 21, 337-343. Laurell, C. B., Kullander, S., and Thorell, J. (1969). Clin. Chirn. Actu 25, 294-296. Laurell, C. B., Nosslin, B., and Jeppsson, I. 0. (1977). Clin. Sci. Mol. Med. 52, 457-461. Leading Article (1977). Luncet ii, 116. Le Bourgeois, J. P., and l’ubiana, M. (1977). Int. J. Rndiaf. Oncol. B i d . Phys. 2, 241-247. Levo, Y. B., Frangione, B., and Franklin, E. C. (1977) Nature (Lonclon) 268, 56-57. Lieberman, J. (1973). Med. Clin. North. A m . 57, 691-767. Lieberman, J. (1974). Ann. I n t . M e d . 81, 850-852. Lieberman, J . , Silton, R. M., Agliozzo, C. M., and McMahon, J. (1975). A m . J . Clin. Poth. 64, 304-310. Lyman, G. H., Bettigole, R.E., Robson, E., Ambrus, J . L., and Urban, H. (1978). Cancer 41, 1113-1122. Lynch, R. E., Kjellstrand, C. M., and Coccia, P. F. (1977). S e w . Oncol. 4, 325-334. Mach, J . P., Fritsche, R., Phan, M., Ozzello, L., and Stjernsward, J. (1977). Proc. Cliri. B i d . Res. 12, 91-108. Makino, S., and Reed, C. E. (1970). Lab. Clin.Med. 75, 742-746. Mancini, G., Carbonara, A. O., and Heremans, J . F. (1965). Immunochern. 2, 235-254. Marner, H . , Friborg, S., and Simonsen, E. (1975). Scund. J . Gastroenterol. 10, 537-544. McFarlane, A. S . , Todd, D., and Cromwell, S. (1964). Clin. Sci. 26, 415-420. McFarlane, H . , Ngu, V. A., Udeozo, I. 0. K., Osunkoya, B. O., Luzzatto, L., and Mottram, F. C. (1967). Clin.Chitn. Actu 17, 325-329. Menzoian, J. A., Glasgow, A. H., Nimberg, R. D., Cooperband, S. R., Schmid, K., Saparoschetz, I . , and Mannick, J. (1974). J. ln7tnunol. 113, 266-273. Micksche, M., and Kokron, 0. (1977). Oster Ziet Onkol. 3, 116- 119. Milano, G., Cooper, E. H., Goligher, J. C., Giles, G. R., and Munro Neville, A. (1978). J . Nut/. Ctincer I n s t . 61, 687-691. Minton, J. P., and Bianco, M. A. (1974). Arch. Surg. 109, 238-240. Mittman, Z., Lieberman, J., Marasso, F., and Miranda, A. (1971). Chest 60, 214-221. Moe, P. J. (1970). Actu Puetliut. Scuntl. 59, 381-384. Mori, K. F. (1969). Endocritm/og.v 85, 330-336. Mortazavi, S . H . , Bani-Hashemi, A . , Mozafari, M., and Raffi, A . (1972). Crrncer 29, 11931198. Mortensen, R. F. Osmand, A. P., and Gewurz, H. (1975). J. E.rp. Med. 141, 821-839. Mortensen, R. F., and Gewurz, H. (1976). J . Irnmunol. 116, 1244-1250. Mueller, W. K., Handschumacher, R., and Wade, M. E. (1971). Ohstef. Gvnurc. 38, 427435.
42
E. H . COOPER A N D JOAN STONE
Muller-Eberhard, U., and Liem, H. H. (1975). Ricerca 5, 275-291. Murgita, A., and Tomasi, T . (1975). J. Exp. Med. 141, 276-286. Murray-Lyon, I. M., and Williams, R. (1974). Clin.Chim. Acta 51, 303-308. Musa, D. U . , Seal, U. S., and Doe, R. P. (1965). J . Clin.Endocrinol. Metub. 25, 11631166. Neville, A. M., and Cooper, E. H. (1976). Ann. Clin. Biochem. 13, 283-305. O’Hara, H., Watawabe, K., and Wada, T. (1968). Clin. Chim. Actu 19, 41-47. Ohlsson, K. (1971). Scand. J . C h i . Lub. Invest. 28, 219-223. Osmand, A. P., Friedenson, B., Gewurz, H., Painter, R. H. Hoffman, T . , and Shelton, E. (1977). Proc. Null. Acad. Sci. 74, 739-743. Owen, J. A. (1967). A h . Clin. Chem. 9, 1-41. Palmer, P. E., Wolf, H. J., and Gherandi, G. J. (1973). Gastroenterology. 65, 284-293. Palmer, P. E., and Wolf, H. J. (1976). Arch. Pathol. Luh. Med. 100, 232-236. Palmer, P. E., Safii, H., and Wolf, H. J. (1976). A m . J . Clin. Pathol. 65, 575-582. Papiha, S. S . , Jones, A. L., Thompson, B., and Carr, D. (1974). J . Mrd. Genet. 11, 349352. Peacock, A. C. (1966). J . Nutl. Cancer Inst. 36, 631-639. Pepys, M. B., Druguet, M., Klaas, H. J ., Dash, A. C., Mirjah, D. D., and Petrie, A. (1977). “Immunology of the Gut.” In Ciba Found. Symp. 46 (New Series) (J. Knight, ed.), pp. 283-297. Elsevier-Excerpta Medica, Amsterdam. Pepys, M. B., Dash, A . C., Markham, R. E., Thomas, H. C., Williams, B. D., and Petrie, A. (1978). Clin. E x p . Immrtnol. 32, 119-124. Peters, T., Jr. (1970). Adv. Clin. Chem. 13, 37- 1 1 1 . Peters, T., Jr. (1975). In “The Plasma Proteins” (F.W. Putnam, ed.), pp. 133-172. Academic Press, New York. Pettingale, K. W.,and Tee, D. E. H. (1977). .I. Clin. Puthol. 30, 1048-1052. Piafsky, K. M., and BorgS, 0. (1977). Clin. Phurm. Therup. 22, 545-549. Pietrek, G., and Kindler, U. (1971). Verh. Deutsch. Ges. Inn. Med. 77, 764-766. Porter, C. A., Mowat, A. P., Cook, P. J. L., Haynes, D. W. G., Shilkin, K. B., and Williams, R. (1972). Br. Med. J. 3, 435-439. Poston, T., and Stewart, I. N. (1976). “Taylor Expansions and Catastrophes.” Pitman, London. Poulik, M. D., and Reisfeld, R. A. (1975). “Contemporary Topics in Immunochemistry,” Vol. 4, p. 175. Plenum, New York. Puri, M. L.. and Sen, P. K. (1971). “Non-parametric Methods in Multivariate Analysis.” Wiley, New York. Putnam, F. W. (1975). In “The Plasma Proteins” (F. W. Putnam, ed.), pp. 2-46. Academic Press, New York. Randle, G. H., Good, W., and Cumerbatch, K. N. (1974). Ohst. Gvnuecol Br. 81,479-482. Rapp, W., Heim, M.. Mikulica-Radecki, J. G., and Ludwig, R. (1975). Klin. Wscht. 53, 139-141. Ray, G. R., Wolf, P. H., and Kaplan, H. S. (1973). N N f . Cancer Inst. Mono. 36, 315-323. Reed, A. H., Henry, R. J., and Mason, W. B. (1971). Clin. Chem. 17, 275-284. Rees, E. D., Hollingsworth, W., Hoffman, T. R., Black, H., and Hearn, T. L. (1975). Arch. Environ. Health 30, 402-408. Rudman, D., Treadwell, P. E., Vogler, W. R., Howard, C. H., and Hollins, B. (1972). Cancer Res. 32, 1951- 1959. Rudman, D., Chawla, R. K., Del Rio, A . E., Hollins, B. M., Hall, E. Z., and Conn. J . M. (1974). J . Clin. Invest. 54, 147- 155.
ACUTE PHASE REACTANT PROTEINS IN CANCER
43
Sandberg, A. A., Rosenthal, H. E., Mittelman, A., and Murphy, G. P. (1975). Urology 6, 17-21. Schutt von, K. H., and Hoffmeister, H. (1971). Z. Klin. Chem. Klin. Biochem. 9, 201-206. Schwartz, M. K. (1978). Antibiot. Chemofher. 22, 176-184. Seal, U. S., and Doe, R. P. (1962). Cancer Chemother. Rep. 16, 329-334. Seal, U. S ., Doe, R. P., Byar, D. P., Corle, D. K., and Vet. Admin. Coop. Urol. Res. Group. (1976). Cancer 38, 1108- 1 1 17. Seal, U. S., Doe, R. P., Byar, D. P., Corle, D. K., and Vet. Admin. Coop. Urol. Res. Group (1978). Cancer 42, 1720-1729. Seidman, R., Silverberg, A., and Holle, A. I. (1976). Cancer 26, 14-32. Sezaki, T., Tanaka, S., Irino, S., and Hiraki, K. (1973). Acfu Med. Okajwinu 27, 141-147. Sharp, H. L., Bridges, R. A., Krivit, W., and Freier, E. F. (1969). J. Lab. Clin. Med. 73, 943-949. Shim, B. S. (1976). Nature (London) 259, 326-327. Shuster, J . , Gold, P., and Poulik, M. D. (1976). Clin. Chim. Acfu 67, 307-313. Siegel, J . , Rent, R., and Gewurz, H . (1974). J. Exp. M ed. 140, 631-646. Siegel, J., Osmand, A. P., Wilson, M. F., and Gewurz, H. (1975). J . Exp. Med. 142, 709721. Siegel, S. (1956). “Non-parametric statistics.” McGraw-Hill, New York. Skrede, S., Blomhoff, J. P., Elgjo, K., and Gjone, E. (1975) Sccrnd. J. Clin. Lab. Invest. 35, 399-406. Smithies, 0. (1955). Biochem. J. 61, 629-641. Solberg, H. E. (1975). Scand. J . Clin. Lub. Iwest. 35, 705-712. Suga, S., and Tamura, A. (1972). Concer Res. 32, 426-429. Takeda, Y. (1966). J. Clin.Invest. 45, 103- 11 1. Tashjian, A. H., Voelkel, E . F., Goldhaber, P., and Levine, L. (1974). Fed. Proc. 33, 8186. Teasdale, C., Mander, A. M., Fifield, R., Keyser, J. W., Newcombe, R. G., and Hughes, L. E. (1977). Clin. Chirn. Actu 78, 135-143. Theologides, A. (1971). Cancer 29, 484-488. Thompson, W. D., Boyd, G., McClay, A. L . C., and Horne, C. H. W. (1978). Biomed. Express 29, 124-125. Tsuchida, Y., Kanedo, M., Yokomori, K., Saito, S., Urano, Y., Endo, Y., Asaka, T., and Takeuchi, T. (1976). J. Pecliat. Surg. 13, 25-29. Turner, R., and Liener, I. E. (1977). Fed. Proc. 36, 765. Twining, S. S., and Brecher, A. S. (1975). Proc. Soc. Exp. Biol. Med. 150, 98-103. Twining, S. S., and Brecher, A. S. (1977). Clin. Chim. Actu 75, 143-148. Ungar-Waron, H., Gluckrnan, A., Spira, E., Waron, M., and Trainin, Z., (1978). Cancer Res. 38, 1296-1299. Van Den Hamer, C. J. A., Morell, A. G., Scheinberg, I. H., Hickman, J., and Ashwell, G. (1970). J . Biol. Chem. 245, 4397-4402. Vickers, M. (1974). J. Urol. 112, 310-312. Voelkel, E. F., Levine, L., Alper, C. A,, and Tashjian, A. H. Jr. (1978). J. Exp. Med. 147, 1078-1088.
Ward, A. M., Cooper, E. H., and Houghton, A. L. (1977b). Br. J . Urol. 49, 411-418. Ward, A. M., Cooper, E. H., Turner, R., Anderson, J. A., and Neville, A. M. (1977a). Br. J . Cancer 35, 170-178. Wegelius, 0..and Pasternack, A. (1976). In “Amyloidosis,” p. 9. Academic Press, London. Werner, M., Brooks, S. H., and Cohnen, G. (1972). Clin.Chem., 18, 116-123.
44
E. H . COOPER A N D JOAN STONE
Wibell, L., Evrin, P. E., and Berggard, (1973). Nephron 10, 320-331. Winzler, R . J . (1953). Ac/v. Cuncer Res. I , 506-549. Wolf, P. L., Williams, and von der Muehlle, E. (1973). “Practical Clinical Enzymology and Biochemical Profiling.’’ Wiley, London. Weisman, S., Goldsmith, B., Winzler, R. J., and Leeper, M. H. (1961). J. Luh. Clin. Met/. 57, 7- 15. Wilimas, J . , Thompson, E., and Smith, K. L. (1978). Ctrncer 42, 1929- 1935. Wynder, E. L., and Hirayama, T. (1977). Prev. M e t / . 6, 567-594. Yachnin, S. (1975). J. Exp. Med. 141, 242-256. Yu, S. D . , and Gan, J. C. (1977). Arch. Biochcrn. Biophys. 179, 477-485. Zwi, S . , Hurwitz, S. S . , Cohen, C., Prinsloo, I., and Kagan, E. (1975). S.Afr. Met/. J . 49, 1887- 1890.
ADVANCES IN CANCER RESEARCH,
VOL. 30
INDUCTION OF LEUKEMIA IN MICE BY IRRADIATION AND RADIATION LEUKEMIA VIRUS VARIANTS
Necharna Haran-Ghera and Alpha Peled Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel
11. Radiation Leukemogenesis .............................................. A. The Role of the Thymus and Bone Marrow ................ B. Age-Related Susceptibility to Radiation Leu C. Phenotypic Characteristics of the Thymus during Radiation Leukemogenesis D. Activation of Endogenous Leukemia Virus by Radiation . . . . . . . . . . . . . . . . E. Prevention or Enhancement of Radiation-Induced Lymphoid Tumors . . . . . 111. Induction of Leukemia by the Radiation Leukemia Virus Variants . . . . . . . A. Characteristics of the Radiation Leukemia Virus Variants . . . . . . . . . . . . . . . B. Target Cells Affected by the RadLV Variants .......................... C. Immunization by RadLV Variants . . . . . . . . .............. D. Genetic Control of Susceptibility . . . . . . . . . . . . . . . .......................... E. Phases in RadLV Leukemogenesis IV. Concluding Remarks .................................. ................... References . . . . . . . . . . . . . . . . .
45 48 49 52 54 57 59 62 64 69 72 75 78 81 83
I. Introduction
Awareness of the radiation hazard in relation to neoplasia was initially derived from the experience with human victims exposed to different types of radiation. Possible factors involved included environmental disorganization and disturbance of physiological regulation, cellular destruction leading to selective enhancement of cell division, and induction of cell alterations including chromosomal aberrations. The long latent period observed in radiation leukemogenesis suggested that several phases were involved in tumor development, and those were influenced by local effects and systemic influences. The field of experimental murine radioleukemogenesis initiated by Furth and Furth (1936) and intensively investigated subsequently by many research teams, provided important contributions to our knowledge of radiation-induced leukemias. The most extensive studies concerned with fundamentals of murine radiation leukemogenesis were those of Kaplan and colleagues who used the C57BL strain of mice as the model system. This strain, which exhibits a very low overt spontaneous inci45 Copyright @ 1979 by Academic Press. Inc. All rights of reproduction in any form reserved ISBN 0-12-006630-0
46
NECHAMA HARAN-GHERA A N D ALPHA PELED
dence of leukemia, was found to be extremely susceptible to the induction of thymic lymphomas by exposure to X-rays. Several reviews (Kaplan, 1964, 1967, 1974) have summarized the basic information on optimal conditions for lymphoma induction and prevention, etiologic agents, target cells for neoplastic transformation, and pathogenesis involved in radiation leukemogenesis. Thus, susceptibility to radiation leukemogenesis was dependent on the age of the mouse and the degree of exposure of the whole body to radiation. The incidence of leukemia was higher when the same total dose of radiation was given in several fractionated doses separated by intervals of only a few days (i.e., four whole-body exposures of 150 to 170 rads at 5-7 day intervals) rather than as a single exposure (Kaplan 1948; Kaplan and Brown 1952). Reduction or prevention of radiation leukemogenesis was achieved by shielding the bone marrow (Kaplan and Brown, 1951), the exteriorized spleen (Lorenz et al., 1953), or lymph nodes (Ilbery, 1967) during exposure of the rest of the body to X rays, or by injecting isologous bone marrow cells (Kaplan et a / . , 1953), spleen cells (Wallis et al., 1966), or peripheral blood lymphocytes (Peled and Haran-Ghera, 1969) into irradiated hosts shortly after termination of the radiation treatment. Although most of the radiation-induced tumors involved the thymus, radiation of the thymic area did not cause leukemia (Kaplan, 1949). Irradiation of both the thymus and bone marrow were shown to be most effective in inducing leukemia in the mouse. The involvement of an indirect mechanism in leukemia induction by X rays was indicated in the following studies. Lymphoid leukemia development in irradiated C57BL mice after being prevented by prior adult thymectomy (Kaplan, 1950) could be largely restored by subcutaneous implantation of thymus grafts from unirradiated syngeneic young mice (Kaplan and Brown, 1954). To determine if these tumors arose from the cells of the nonirradiated thymus implant or from other cells derived from the irradiated host, genetic and cytologic markers were used to define the origin of the tumors. Variable results were obtained by several investigators (Table I). In some studies (Kaplan et al., 1956: Barnes et a/., 1959) the normal thymus graft supplied the leukemic progenitor cells, while in other studies the majority of tumors were of cells from the irradiated host (Law and Potter, 1956, 1958; HaranGhera, unpublished results). The variable results reported by different researchers could be due to the different experimental procedures used (e.g., age of thymus graft and the time interval between termination of the radiation treatment and thymic implantation). The fact that some of the tumors originated from the nonirradiated graft suggested the involvement of a transmissible agent. Since Gross (1951) discovered that a type-
TABLE I REPOTENTIATION OF RADIATION LEUKEMOGENESIS I N THYMECTOMIZED HYBRID MICEBY SUBCUTANEOUS THYMUS GRAFTING
Irradiated hosts” (strain)
References Kaplan et
(I/.
(1956)
(C57BLIKa x C3H)F,
Law and Potter (1956)
(C57BL x A)F,
Law and Potter (1958) Barnes et a/. (1959) Haran-Ghera (unpublished)
(C57BL x A)F, (C57BL X CBA T,T,)F, (C57BL/6x BALB/c)F,
a
Time interval for thymus grafting after exposure to X rays 1-3 hours
1-3 hours
7 days 14-28 days Not given 1-3 hours
Thymus graft
Origin F1 C57BL C3H C57BL C57BL C57BL C57BL C57BL/6
Age (days) 7-14 7-14 7-14
Leukemia incidence in irradiated host
Leukemia origin
Host
Graft 11/12 6I7
18/54
33%
11/12
14/54
26% 3% 37%
II7 517
25%
315 317
2/58
12/38
1-12 1-12 1-10 2
Not given
1
15/22
12/48 91113
8%
7/14 68%
Adult mice, 6 to 8 weeks old, were thymectomized and 7-10 days thereafter exposed to fractionated radiation.
9/12
217 215 417 4/14 3/12
48
NECHAMA HARAN-GHERA A N D ALPHA PELED
C virus was responsible for the high spontaneous incidence of leukemia in the AKR strain of mice, a search for similar viruses in radiationinduced leukemias was undertaken. Viruses similar to the Gross agent were indeed extracted from radiation-induced thymomas in mice of strains C3H (Gross, 1958) and C57BL/Ka (Lieberman and Kaplan, 1959) and later from radiation-induced myeloid leukemias of the RF mice (Jenkins and Upton, 1963). However, the leukemogenicity of such virus isolates from primary radiation-induced tumors was extremely low and variable. It was thus proposed by Kaplan (1967) that radiation induces a number of different effects in the mouse including: (a) thymus injury followed by its regeneration, thereby making available an abundant supply of target cells for viral transformation; (b) injury of the bone marrow affecting thymus repopulation; and (c) activation of an endogenous murine leukemia virus (transmitted vertically in the C57BL strain) which could infect and transform the thymus cells. Although leukemogenic viruses have been isolated from radiation-induced thymomas (Lieberman and Kaplan, 1959; Latarjet and Duplan, 1962: Laznicka and Smetanova, 1963: Ilbery and Winn, 1964) and from tissues of irradiated nonleukemic C57BL/6 mice (Haran-Ghera, 1966: Haran-Ghera and Peled, 1967), a direct etiologic role of the virus in Xray leukemogenesis is still not established. In this review we shall assess recent experimental data concerning pathways involved in the development of leukemias in the C57BL strain induced by exposure to fractionated whole-body irradiation or by inoculating one of the different radiation leukemia virus variants presently available. In the absence of cofactors these virus variants have been shown to differ in their leukemogenic potential (see Table IV). We wish to reassess by these comparative studies the function of virus(es) in radiation leukemogenesis. II. Radiation Leukemogenesis
Exposure of young adult C57BL mice to fractionated irradiation (four weekly exposures of 150- to 170-rad whole-body irradiation) induces a high incidence (70%-100%) of thymic lymphomas after an average latency of about 200 days. The relationship of tumor development (thymic lymphoma or reticulum cell sarcoma) to age and variable treatments is summarized in Table 11. Since the thymus is the target organ for the overt expression of the disease, it has been assumed that the neoplastic transformation, involving viral etiology, occurs in the thymus (Kaplan, 1967, 1974). We shall discuss in this section the site of the radiation-induced trans-
49
INDUCTION OF LEUKEMIA IN MICE
TABLE I1 TUMOR DEVELOPMENT IN C57BL/6 MICEEXPOSED TO FRACTIONATED IRRADIATION Reticulum cell neoplasms
Thymic lymphoma Age of exposure (months) I 1
1 1 1 1
3 6 9 'I
Leukemogenic treatment 170 rads x 4 weekly on thymus only Thymectomy 1 week 170 rads x 4 weeklyb 170 rads x 4 weekly 170rads x 4-2 weeks 170 rads X 4 weekly 2 hours BMc--I07 i.v. 170rads x 4 weekly 7 days BMc-107 i.v. 170 rads x 4 weekly 170 rads x 4 weekly 170 rads x 4 weekly
Incidence
ALP"
ALP"
(days) Incidence
(days)
1/50 0/20
-
370 -
6/50 1/20
12% 5%
670 620
0/20
-
-
9/20
45%
540
17/20 7/20 6/29
85%
190
35% 20%
210 185
11/20 12/29
55% 40%
580 550
6/23
26%
130
14/23
60%
580
11/20 9/20 6/20
55% 45% 3M
200 190 270
7/20 7/20 13/20
359 35% 65%
580 500 270
2%
ALP = average latent period. The time interval between each of the four whole-body radiation exposures. BM = bone marrow from syngeneic C57BL/6 8-week-old female mice.
formation, the target cells involved, and the phenotype characteristics of these cells. Moreover, we shall review studies on the age-related susceptibility to radiation leukemogenesis, the radiation-induced "activation" of the endogenous leukemic virus, and its possible involvement in enhancing or preventing radiogenic lymphoma development.
A. THEROLE OF
THE
THYMUS A N D BONEMARROW
The target organ for the overt development of most spontaneous or induced murine lymphoid leukemias is the thymus. Thymectomy was shown to have a most profound inhibitory effect on the incidence of spontaneous leukemia (Furth, 19461, as well as on virus- and radiationinduced tumors (Kaplan, 1950: Law and Miller, 1950). By contrast, in some instances lymphatic leukemia induction by chemical carcinogens was shown to occur both in intact and thymectomized mice (Kirschbaum and Liebelt, 1945: Haran-Ghera et ul., 1967: Haran-Ghera and Peled, 1973: Armuth, 1976).
50
NECHAMA HARAN-GHERA A N D ALPHA PELED
The role of the thymus in murine leukemogenesis is complex and not yet clearly elucidated. Thymic susceptibility to leukemia induction is apparently maximal at birth. Thus, exposure of adult mice to leukemogenic agents may restore the neonatal level of susceptibility by producing a selective state of maturation arrest in the lymphoid cell population of the regenerating injured thymus (Kaplan, 1967). In general, depletion of thymic lymphocytes, through focal thymocyte atrophy, induces an abundance of reticular cells and leads eventually to the development of thymic lymphomas (Dunn et NI., 1961: Rappaport and Baroni, 1962: Haran-Ghera and Kaplan, 1964). Large immature lymphoblastic thymocytes, appearing in the thymus following whole-body exposure to X rays, have been considered as target cells for neoplastic transformation (Kaplan, 1967). Shielding either the femoral bone marrow or the exteriorized spleen of C57BL mice during radiation and injecting the syngeneic normal bone marrow cells after irradiation have both been shown to enhance the progressive regeneration of the radiation-injured thymus. These methods also effectively prevent development of lymphomas. Thus, nonirradiated marrow replenishes the thymus more efficiently than irradiated bone marrow. Protection of bone marrow from irradiation may therefore influence rates of migration of bone marrow-derived thymocyte precursor cells to the thymus. A special blast-type population characterized by its scLrce dense chromatin (forming irregular blocks scattered throughout the whole nucleus) and its compact nucleolus has been described as x cells (Haot rt cd/., 1973: Boniver et ( I / . , 1978). These cells were shown to occur i n abundance after radiation disturbs the pool of lymphoid precursors for marrow and thymus lymphopoiesis. Thus, x cells were suggested to be target cells for lymphoid tumor development. The prevention of leukemia by normal bone marrow transplantation was shown to reduce drastically the amount of these x cells in the thymus. Another thymic component, the epithelial-reticular stroma has also been suggested many years ago as an essential factor for the progression to autonomous frank neoplasia (Law, 1957). Radiation-induced thymic lymphomas appear only after a long latent period. Studies were carried out to define the site and time at which transformed cells, having the potential to develop eventually into overt leukemia, appear. The transplantation-bioassay method, originally introduced by Furth and Boon (1944), was used to establish the presence of potential leukemic cells in mice at a stage when they do not yet show clinical disease. This method involved intravenous injection of bone marrow cells from C57BL/6 mice treated with fractionated irradiation into irradiated (400 rads-a nonleukemogenic dose) syngeneic or hybrid (BALBlc x C57BL/6) F, mice. The bone marrow cells were collected
INDUCTION OF LEUKEMIA IN MICE
51
one to several weeks following termination of the leukemogenic treatment. The use of hybrid mice in this analysis enabled to identify the origin of the preleukemic cells (by transplanting the tumor cells developing in F1 hybrid mice into F, and parental strains). The origin of leukemic cells from parental C57BL/6 donor strain would indicate the actual transfer of established potentially leukemic cells among the transferred bone marrow cells, whereas leukemias of F, host origin would indicate the presence of a noncellular leukemogenic agent among the transferred cells that caused neoplastic transformation in the thymus or other lymphoid organs of the host. Potential leukemic cells were observed in the bone marrow (and not in thymus or spleen cells tested similarly) of C57BW6 mice exposed to fractionated irradiation within 30 days following termination of the radiation treatment. The methods known to reduce (by syngeneic bone marrow transplantation shortly after irradiation) or prevent (by thymectomy before exposure to fractionated irradiation) radiation leukemogenesis in the mouse did not prevent or reduce the occurrence of these preleukemic cells in the bone marrow. They occurred at a similar time, incidence, and site as those induced in the irradiated intact mice. These results indicated the transforming effect of X rays on a certain bone marrow population irrespective of the ultimate susceptibility of the treated mice to the development of overt leukemia (Haran-Ghera, 1976). An important conclusion from these experiments was that the radiation-induced cell transformation occurred in the bone marrow rather than in the thymus. This observation provided an explanation for the fact that irradiation of the target organ (e.g., the thymus) did not cause tumor development in mice (Kaplan, 1967). Comparative studies on the phenotype characteristics of the early occurring transformed cells and overt leukemic cells indicated that in contrast to the radiation-induced leukemic cells that express the cell surface component Thy- 1 (Chazan and Haran-Ghera, 1976), the radiation-induced preleukemic cells in the marrow lack this antigen (Haran-Ghera, 1977) (see Section 11, C for further discussion on phenotype characteristics of thymocytes during X-ray leukemogenesis). These observations would suggest that X rays act initially on stem cells or prothymocytes present among the bone marrow populations. These preleukemic cells would need the thymic microenvironment to further differentiate into overt T leukemia cells. Micklem (1960) has shown that thymic lymphocyte precursors that reside in the bone marrow may repopulate the thymus during 2 to 4 weeks after termination of irradiation. The injured repopulated thymus might thus provide the microenvironment for the proliferation of the preleukemic cells present in the bone marrow. In this case, enhanced thymus repopulation (by shielding procedures of the hemopoietic organs
52
NECHAMA HARAN-GHERA A N D ALPHA PELED
or by injection of syngeneic bone marrow or spleen cells after radiation treatment) might prevent the migration of the preleukemic cells to the microenvironment favorable to their further proliferation. Two fundamental phases occurring during radiation leukemogenesis could thus be suggested: ( I ) the initial X-ray-induced neoplastic transformation occurring in the bone marrow (a non-thymic-dependent phase) and (2) migration of the preleukemic cells from the bone marrow to the thymic microenvironment which induces a proliferative phase enabling the ultimate growth of leukemic cells. Mutagenic agents act directly on somatic cells to alter their genetic material in some specific way and thereby lead to neoplastic transformation. Since radiation is a mutagenic agent, its initial effect may be a direct mutagenic alteration of stem cells or prothymocytes present in the bone marrow. Kaplan (1974) believes that somatic mutation is an untenable mechanism for the leukemogenic action of radiation for the following reasons: nonlinearity of the dose response relationship in radiation leukemogenesis, an increased yield of tumors with fractionated irradiation, the requirement for whole-body rather than local X-irradiation of the thymus, the indirect induction mechanism, and the striking protection by shielding or injection of bone marrow cells. But if one considers the possibility that radiation leukemogenesis is not necessarily a one-stage effect, but rather involves several phases for its development, the somatic mutation theory remains plausible. Thus, the transformation of somatic target cells in the bone marrow, by a direct mutagenic alteration of the genetic material by irradiation, could induce preleukemic cells. The additional X-ray-induced environmental changes (including thymus injury and transient host immune impairment) could contribute to the proliferation of these transformed cells in the appropriate thymic microenvironment.
B. AGE-RELATEDSUSCEPTIBILITY T O RADIATIONLEUKEMOGENESIS The age at the time of the first exposure to radiation is an important determinant in radiation leukemogenesis. Susceptibility of C57BL mice to fractionated whole-body irradiation was found to be maximal at 2-5 weeks of age and decreased rapidly thereafter (Kaplan, 1948: see also Table 11). The target organ for overt leukemia development is the thymus, and it could be assumed that the alteration of susceptibility with age might be related to thymus involution or lack of target cells in the thymus of older mice. Experimental evidence (Haran-Ghera, 1975) indicated that thymus architecture did not appear to be responsible for the
INDUCTION OF LEUKEMIA IN MICE
53
reduced susceptibility to radiation leukemogenesis observed with increased age. Thymectomized mice of variable ages carrying a young thymus graft (the thymus graft developed well in the different age groups) were exposed to fractionated irradiation. An age-dependent decrease in leukemia incidence was observed, although the thymus in these tested mice was of the same age, and the variable factor was the host’s age. The capacity of fractionated irradiation to induce potential leukemic cell transformation among bone marrow cells of older C57BL/6 mice (6and 9-month-old ones) was indicated by using the previously described bone marrow transfer method. A high incidence of potential leukemic cells were demonstrated among the irradiated bone marrow cells in older mice shown to be resistant to overt leukemia development (Haran-Ghera, 1976). These findings clearly indicated that the age increase did not affect the susceptibility of bone marrow cells to undergo neoplastic transformation. It is interesting to stress the unique occurrence of reticulum cell neoplasm type A, following transplantation of bone marrow cells taken from 9-month-old irradiated mice, in contrast to thymic lymphoma following transfer of bone marrow from 1- to 6-month-old irradiated mice. These findings might suggest that in young mice, transformed stem cells eventually migrate into the thymus and there become T leukemic cells: whereas in older mice the altered thymus microenvironment would prevent further differentiation of the X-ray transformed stem cells into T cells. These transformed cells would remain in a more primitive form of undifferentiated reticulum cell and ultimately develop into reticulum cell neoplasms (Table 11). The occurrence of reticulum cell neoplasms as a “late” radiation effect has also been described by Van Bekkum e t ul. ( 1977). He proposed dividing the occurrence of lymphoreticular tumors in C57BL mice after X-ray treatment into three distinct periods: (I) tumors appearing 4-9 months after radiation treatment comprising of thymic lymphosarcomas, their development depending on exposure to radiation and presence of a thymus: (11) tumors developing between 918 months, their occurrence being radiation dependent but thymus independent; and (111) tumors occurring from 18 months onward, these being both thymus and radiation independent. Thus, the presence of functional thymus tissue has two effects: (1) accelerating the appearance of lymphosarcomas and (2) favoring the development of lymphosarcomas over reticulum cell sarcomas in irradiated C57BL mice. The observation that older mice are less susceptible to radiation leukemogenesis than young mice might perhaps be related to the occurrence of natural immunization of older mice against overt thymoma development by the endogenous leukemogenic virus itself (Haran-Ghera, 1975: Ihle, 1978), by spontaneously occurring preleukemic cells in several-
54
NECHAMA HARAN-GHERA A N D ALPHA PELED
month-old C57BL/6 mice (Haran-Ghera, 1978a), or by an age-dependent increase of autogenous immune respone to endogenous RNA tumor viruses normally observed in several strains of mice (Ihle er a l . , 1973). Indeed, an age-dependent resistance of C57BL/6 mice (5- to 8-month-old) to syngeneic leukemic cells induced by the radiation leukemia virus in contrast to the proliferation of carcinogen-induced firbosarcoma cells was demonstrated (Haran-Ghera, 1975). The infusion of serum from 8-monthold C57BL mice into 6-week-old C57BL/6 mice concurrently with radiation treatment reduced leukemia incidence and extended the latent period (Friedman and Haran-Ghera, unpublished data). In conclusion, the age-dependent resistance to radiation leukemogenesis operates on the proliferation phase level rather than on the initial transformation phase (Haran-Ghera, 1976). Age-dependent host factors might interfere with the proliferation of preleukemic cells. The lack of suitable thymus microenvironment in several-month-old irradiated C57BL mice might prevent overt T lymphoma development (though the irradiated mice are carriers of preleukemic cells). These mice might eventually, after a prolonged latency, develop reticulum cell neoplasms (see Table 11). C. PHENOTYPIC CHARACTERISTICS OF THE THYMUS DURING RADIATIONLEUKEMOGENESIS The target organ for the overt development of most spontaneous or induced murine lymphoid leukemias is the thymus. Thymic lymphocytes can be divided into two categories on the basis of surface markers and functional properties. The major thymus subpopulation, considered to be steroid and radiation sensitive, consists of immunologically inactive cells and its antigenic characteristics include high levels of 8 antigen and low levels of H-2 alloantigen, and the TL antigen (when present at all). The minor subpopulation, which is relatively radiation and cortisone resistant, has a high H-2 antigen content and lower levels of 8, lacks the T L antigen, and possesses the capacity to respond to phytohemagglutinin and mount graft versus host reactions as well as reacting in mixed lymphocyte cultures (Cerottini and Brunner, 1967: Raff, 1971). Based on antigenic properties of the cell surface of mouse thymocytes, radiationinduced T leukemias in C57BL/6 mice were shown to have characteristics of the minor thymus subpopulation. Therefore, it seemed plausible to assume that a transient change in thymus subpopulations following radiation injury might be providing a favorable proliferative *‘milieu” for the migration of preleukemic cells. Indeed, the correlation between the
INDUCTION OF LEUKEMIA IN MICE
55
availability of an “elevated” thymus minor subpopulation for several weeks and leukemogenic treatment in terms of induction of a high or low overt leukemia incidence was quite remarkable. Thus, the effective leukemia induction system of four weekly exposures to 170 rads whole-body irradiation changed the thymus subpopulation pattern. Radiation injury to the thymus was reflected in drastic reduction in thymus weight due to elimination of the radiation-sensitive thymocytes, and the cells surviving showed high levels of H-2” alloantigen, as expected. An abundance of high H-2, low 0 thymus-derived lymphocytes were actually present in the thymus for several weeks, in spite of the fact that the thymus had already regenerated and almost regained its initial normal weight. It is important to note that there was a reduced incidence (similar to normal thymus levels) of high H-2” bearing thymocytes following four exposures to 170 rads at f4-day intervals, a treatment shown to yield a low leukemia incidence (Kaplan and Brown, 1952; see also Table 11). Bone marrow shielding during the last radiation exposure or bone marrow administration within 1-2 hours after the last radiation treatment also reduced drastically the radiation-induced elevated high H-2 minor thymus subpopulation. The fact that one single marrow shielding or administration of bone marrow cells during or after the last radiation exposure is as effective in preventing leukemia development as repeated preventive treatments, after each radiation exposure, coincided with the findings that the susceptible minor high H-2” thymus subpopulation was present in abundance for an extended period of time only after the last radiation exposure. In contrast to the specific thymus subpopulation pattern occurring following fractionated irradiation, exposure to a single nonleukemogenic dose of 400 rads caused elimination of the minor high H-2, low 0 population for many weeks. The transient induced changes in thymus subpopulations were shown to be age dependent. Evaluation of the thymus population pattern in 5- and 8-month-old irradiated mice indicated a marked decrease in the availability of thymocytes bearing high levels of H-2” alloantigens and low theta level (although thymus regeneration weight curves were similar, irrespective of age differences). These observations suggest that older mice, after treatment, lack the preferable thymus environment for the migration and proliferation of preleukemic cells from the bone marrow into the thymus (Chazan and HaranGhera, 1976). Cell surface antigens belonging to the T L system are frequently found on radiation-induced leukemias in C57BL mice, although TL is not detected in the thymus of normal C57BL mice (Old et u l . , 1963). Studies carried out (Stockert and Old, 1977) to test whether that anomalous TL expression coincides with the transformation event has indicated the
56
NECHAMA HARAN-GHERA A N D ALPHA PELED
occurrence of TL+ in the thymus shortly after the radiation treatment and before overt leukemia occurs. The results do suggest that activation on the TLa locus occurs during the preleukemic phase of radiation leukemogenesis. Nevertheless, it should be ascertained whether mice expressing TL+ shortly after radiation treatment will indeed be the ones to develop TL+ leukemias. The T L conversion might be due to some specific precursor bone marrow population that repopulates the thymus following specific injury induced by fractionated irradiation. Similar to the studies previously described, it would be interesting to see the specificity of this TL conversion: whether exposure to doses that are not leukemogenic or treatments that prevent leukemia development (injection of bone marrow cells following last exposure to X rays) would provide similar TL- to TL+ conversion. Such studies would ascertain that TL is indeed a marker for preleukemic changes occurring in the thymus early during radiation leukemogenesis. Thle and co-workers ( 1978) have supported the concept that leukemia development involves lymphoid precursors blocked in their ability to terminally differentiate. Thus, a correlation of altered distribution and regulation of the enzyme terminal deoxynucleotidyl transferase (TdT) to induction of leukemia has been proposed by these investigators. This enzyme TdT has been purified and characterized in association with thymocyte differentiation (Bollum, 1974). TdT activity in thymocytes of mice has been shown to be associated with two enzyme fractions defined as peak I and peak I1 after separation by phosphocellulose chromatography. Peak I was found to be more closely associated with the cortisoneresistant thymus subpopulation, whereas the major thymus population was shown to express mainly peak I1 (Kung e t al., 1975). TdT was also shown to be present in a subpopulation of the bone marrow. Since its presence in the bone marrow was demonstrated in a Thy-1- cell population that could be induced by thymopoietin to become a Thy-l+ cell population, Silverstone et al. (1976) proposed that TdT was associated with prothymocytes. Variable TdT peaks were shown to occur in different strains of mice and a certain age relationship was also described. Thus, the expression of TdT peak I1 was almost lacking from AKR and NIH Swiss mice (in normal thymus and in thymomas) and decreased in other strains with increase in age. Leukemogenic doses of irradiation were shown to alter the normal differentiation of TdT positive cells in the thymus and bone marrow of C57BL/6 mice. The thymus of young normal C57BL/6 mice was shown to display TdT activity of both peaks until the age of 6-8 months when it decreased to low levels, whereas in the irradiated thymus TdT activity remained at low levels (less than 20% of that found in controls). Both TdT peaks were also demonstrated in
INDUCTION OF LEUKEMIA I N MICE
57
the normal bone marrow of C57BL/6 mice, particularly evident in one of the cell fractions (A) obtained following separation by discontinuous BSA density gradient. This bone marrow fraction A previously shown to include prothymocytes (Silverstone et a l . , 1976) showed a five- to tenfold increase in TdT activity following radiation treatment. Radiation-induced thymic localized tumor cells were shown to have TdT peaks I and 11, whereas only peak I was found in leukemic cells infiltrating peripheral lymphoid tissue. In \?iitu passage lines, as well as in titro cell lines of the radiation-induced thymomas displayed only TdT peak I (Pazmino and Ihle, 1976: Pazmino rt ul., 1977; Ihle, 1978). Thus, a phenotypic loss of TdT positive cells have been correlated with the development of thymic lymphomas. An alternative suggestion could correlate the findings that the majority of radiation-T-induced leukemias have the phenotypic characteristics of the minor thymus subpopulation (Chazan and Haran-Ghera, 1976), and that this population expresses mainly TdT peak I (Kung ct d.,1975). Thus, the leukemias that originate from the minor thymus subpopulation would display mainly TdT peak I (peak I1 in the localized thymoma could represent normal thymocytes left over that would be eliminated by transplantation).
D. ACTIVATION OF ENDOGENOUS LEUKEMIA VIRUSBY RADIATION Evidence that murine leukemia virus replication may be activated by radiation was initially provided by demonstrating leukemogenic activity of cell-free extracts prepared from radiation-induced leukemias (Gross, 1958; Lieberman and Kaplan, 1959). Cell-free extracts prepared from nonleukemic bone marrow of irradiated mice collected shortly following X-ray exposure were also shown to have leukemogenic activity (HaranGhera, 1966: Haran-Ghera and Peled, 1967). These cell-free extracts possessed a low leukemogenic activity when inoculated into newborn mice of the strain of extract origin, but this leukemogenic activity could be amplified by serial in i ~ i ipassaging. ~) Transient virus activation within several days following radiation treatment was also demonstrated in irradiated bone marrow transplantations. Hybrid mice reconstituted with bone marrow cells from C57BL/6 mice within several days following exposure to fractionated irradiation developed leukemias of F, origin, whereas no leukemias occurred in hybrid recipients reconstituted with normal C57BL/6 bone marrow cells of matching age and sex. It should be stressed that methods known to reduce or prevent radiation leukemogenesis did not affect the release of the leukemogenic agent present in irradiated bone marrow (Haran-Ghera, 1976; 1978a). Electron micro-
58
NECHAMA HARAN-GHERA A N D ALPHA PELED
scopic examination of several organs from mice exposed to fractionated irradiation or to a single nonleukemogenic X-ray dose revealed the presence of virus particles budding from cell membranes within a few days following exposure to radiation (Gross and Feldman, 1968). The search for viral antigen expression, using the immunofluorescence method, in organs of mice exposed to fractionated irradiation, revealed that thymocytes were devoid of antigen expression throughout the long latent period, although leukemic cells were detected one month after X-ray treatment (Lieberman ef ( I / . , 1976). Most radiation-induced leukemias in the C57BL strain of mice have also been shown to lack demonstrable murine leukemia virus (MuLV) or MuLV-related antigens. Some antigen expression though was indicated among the bone marrow cells and, to a lesser extent, in the spleen (Lieberman et a / . , 1976). Similar observations described by Haas (1977a) indicated the presence of MuLV gsa positive cells in the bone marrow at about 8 days after termination of the radiation treatment and persisted for about 3 weeks. Kaplan and associates have postulated that exposure of C57BL mice to X rays may lead to a redepression or activation of a latent viral genome and that thereafter only partial virus expression is required to initiate transformation in susceptible cell types. The distribution pattern and the infectivity titers of the endogenous virus in irradiated tissues (quantitated by using the XC plaque assay) did not appear to correlate with the leukemogenic process. The expression of infectious MuLV in the thymus of irradiated mice was negative during leukemogenesis, and the infectivity patterns and titers in the other tissues were inconsistent with thymoma induction. The MuLV infectivity patterns in various tissues of partial-body irradiated mice (inducing a low overt leukemia incidence) or in thymectomized irradiated mice (thereby abolishing overt lymphoma development) were essentially similar to those of whole-body irradiated mice (Nagao, 1977; Yokoro ef d.,1977). The accelerated appearance of antibody against MuLV demonstrated in C57BL/6 mice exposed to fractionated irradiation supported the hypothesis that irradiation activates the expression of endogenous ecotropic virus (Ihle et a / . , 1976a). This immune response was characterized by high titers of antibodies against viral envelope proteins gp71 and p15 (E). However, no correlation between the development and persistence of this humoral immune response against MuLV and the subsequent development of thymomas was demonstrated. In extensive studies of radiation-induced thymomas in C57BL/6 mice, using both serological and biochemical assays, no detection of overt virus expression or any seroepidemiological association between the endogenous ecotropic C-type viruses and leukemia development was observed. Reconstitution of irradiated mice by syngeneic normal bone marrow suppressed
INDUCTION OF LEUKEMIA I N MICE
59
the occurrence of thymic lymphoma, without any detectable effect on the rate of appearance of a humoral immune response against the virus. It was therefore concluded that bone marrow suppression of leukemia development was not related to its potential influence on the humoral immune response against the virus. Cultures of tumor cells established from radiation-induced tumors did not express viral antigens in their cytoplasm for several passages. These conclusions were based on a variety of assays which included competition radioimmune precipitation assays for AKR MuLV gp71, p30, and p12: cytotoxicity assays with antisera against gp71 or p12: and assays for reverse transcriptase and hybridization of AKR MuLV cDNA with cellular RNA from radiationinduced thymomas (Ihle et al., 1976b). Since the MuLV genome is present in C57BL/6 mice, the possibility of its transient expression during radiation leukemogenesis cannot be excluded. Experimental evidence has been provided indicating that the presence of infectious ecotropic MuLV is not essential for induction of lymphomas by irradiation. Thus, C57L mice developed thymomas following exposure to fractionated irradiation, although no ecotropic typeC virus has been isolated from normal C57L mice, from X-ray induced tumors, or their transplants (Arnstein et a / ., 1976). Comparable studies were also carried out in NIH Swiss mice shown to genetically lack at least a portion of the AKR MuLV genome (Chattopadhyay et a / ., 1974). These mice were shown to be susceptible to radiation leukemogenesis, developing a high incidence of thymic lymphomas following exposure to fractionated irradiation. Studies concerned with virus expression in irradiated NIH Swiss mice, following serological and competition assays, have indicated a lack of detectable antibodies against MuLV and absence of ecotropic-specific virion component during the process of radiation leukemogenesis (Ihle, 1977). These results demonstrated that radiationinduced thymomas could occur in strains of mice that were genetically deficient in at least a portion of the MuLV viral genome presumed responsible for thymomas in C57BL/6 mice. In summary, although endogenous virus replication in C57BL/6 was shown to be activated by irradiation, its possible role as the causative leukemogenic agent in radiation-induced leukemias still seems unresolved. E. PREVENTION OR ENHANCEMENT OF RADIATION-INDUCED LYMPHOID TUMORS The proposition that the endogenous leukemia virus, activated by X rays, might serve as a causative agent in radiation leukemogenesis initi-
60
NECHAMA HARAN-GHERA A N D ALPHA PELED
ated studies on the prevention of radiation-induced tumors, using methods that affect viral activity during tumor development. An attempt to prevent radiation-induced thymomas in C57BL/6 mice was made by inducing active immunity (repeated treatments during 8 weeks since the start of exposure to X rays) to endogenous type-C virus with inactivated Rauscher or Gross MuLV or by transferring passive immunity to endogenous type-C virus with goat anti-Gross MuLV IgG (Peters r t al., 1977). A significant reduction in tumor incidence was observed following immunization with Rauscher MuLV or by passive transfer of goat anti-Gross MuLV IgG. Thus, circulating antibodies to endogenous type-C virus induced by active immunization or by passive treatment with anti-Gross MuLV IgG diminished development of radiation-induced thymoma. Concurrent treatment of mice with fractionated irradiation and interferon was also shown to decrease leukemia incidence in C57BL/6 mice (Lieberman Pf nl., 1971). In both studies the preventive treatment was given during several weeks or months after the radiation treatment. Thus, the inhibitory effects may have been on tumor growth and not necessarily on virus replication. Antilymphocyte serum (ALS) administration was also evaluated in radiation leukemogenesis. It has been shown that ALS enhances leukemia incidence in susceptible strains following MuLV inoculation, probably by suppressing immunological surveillance against transformed cells (Allison er NI., 1967; Hirsch and Murphy, 1968). Balner (1971) studied the influence of two types of ALS antisera (horse and rabbit) on the development of radiation-induced leukemia in mice. Horse ALS did not affect the incidence and latency periods of lymphomas, but mice treated with rabbit ALS failed to develop lymphomas. The inhibitory effect of rabbit ALS could be attributed to the presence of antibodies to MuLV in the rabbit ALS or to a more pronounced cytotoxic effect afforded by rabbit's ALS on developing preleukemic or leukemic tumor cells. It should be stressed though that in these studies the ALS treatment was started 1 week after the last X-ray treatment and was continued for 3 months thereafter. Thus, this procedure also probably permitted the cytotoxic effect of ALS on developing leukemic cells. Attempts were also made to inhibit radiation leukemogenesis by immunization with cells infected with radiation leukemia virus (RadLV) [B l-S(RadLV)] (Lieberman and Kaplan, 1977). Lymphoma induction by whole-body X-irradiation of C57BL/Ka mice was significantly inhibited (relative to the incidence in PBS-treated controls) by immunization with virus-producing B 15 (RadLV) cells, but not with nonvirus producer Bl-5 cells. Mice injected once with BL/Ka (B) virus alone were not protectd against lymphoma development. Thus, a vaccinationlike protection was conferred by B 1-5
INDUCTION OF LEUKEMIA IN MICE
61
(RadLV) cells but not by the free virus. These experiments suggested that virus-induced cell surface membrane antigenic determinants might have played a role in this immunization procedure. It was suggested (Haran-Ghera and Peled, 1967, 1968) that radiationinduced immune impairment might permit incipient tumor cells to persist and progress to frank neoplasia by interfering with normal immune surveillance. Tests were carried out to evaluate the immune suppression induced by fractionated irradiation and its possible contribution to leukemia development. A marked depression in direct (IgM) plaque forming capacity to sheep red blood cells (SRBC) was observed in irradiated mice for several months. This impairment did not seem to contribute to leukemia development since the disease appeared in mice exhibiting either a low or a high response to SRBC, and the latency was also found to be unrelated to the degree of immunosuppression induced by radiation. A significant reduction in graft versus host (GVH) response was observed up to 1 month following fractionated irradiation. Skin graft survival was not prolonged in irradiated mice, nor could such mice accept an allogeneic tumor graft (Haran-Ghera, 1976). Increased immune impairment by the concomitant treatment of mice with fractionated irradiation and antithymocyte serum (ATS) enhanced radiation leukemogenesis (Haran-Ghera and Peled, unpublished results). C57BL/6 mice were exposed to four weekly doses of 170-rad whole-body irradiation and four weekly injections of ATS (0.5 ml subcutaneous each time) given 3 days after each exposure to 170 rads. The additional ATS treatment enhanced tumor development as expressed in a shorter latent period for overt tumor development (100 days in contrast to 180 days in the controls). An immune response reflected in the occurrence of antibodies against MuLV following radiation was considered as an indication that radiationinduced immunosuppression was not associated with radiation leukemogenesis (Ihle, 1978). One should perhaps though consider that the contribution of host immune impairment should not necessarily be related to viral expression but perhaps rather to preleukemic cells present in the bone marrow shortly after the radiation treatment. Recent studies concerned with virus-associated malignancies could support this suggestion. Thus, antigens acquired by the malignant transformation were shown to differ from virus-associated antigens. A difference on the molecular levels between viral antigens and some antigenic moiety present on transformed cells has been demonstrated. (Okazaki et al., 1976: Fenyo and Klein, 1976; Siegert et al., 1977; Fenyo et al., 1977). In conclusion, from the available present data, the possible contribution of host immune impairment to radiation leukemogenesis affecting viral etiology and/or transformed preleukemic cells is still unresolved.
62
NECHAMA HARAN-GHERA A N D ALPHA PELED
Ill. Induction of Leukemia by the Radiation Leukemia Virus Variants
A role for viruses in murine thymomas was first demonstrated by Gross ( 1951) when he transferred leukemia from the high leukemia AKR strain
to the low leukemia C3H/Bi strain with cell-free extracts of a spontaneous tumor. Eight years later leukemogenic cell-free extracts were obtained from radiation-induced thymomas (Gross, 1958; Lieberman and Kaplan, 1959). A virus associated with radiation leukemogenesis, called RadLV, was first isolated by Lieberman and Kaplan in 1959 from radiation-induced thymomas in C57BL/Ka mice exposed to fractionated irradiation. Cell-free extracts prepared from these tumors, when inoculated into syngenetic newborn mice, induced a low incidence of leukemia after a long latent period. However, the leukemogenic activity increased and the latent period decreased substantially with in vivo serial passages. This virus preparation will be referred to in this review as RadLV. In contrast to RadLV isolates obtained from thymomas, a leukemogenic agent was also isolated from nonleukemic bone marrow of young adult C57BL/6 mice within several days after having exposed them to fractionated irradiation (Haran-Ghera, 1966). This virus also had low leukemogenic activity when tested in newborn mice and serial passages in v i ~ wincreased its leukemogenic potential in newborns. Injection of this isolate into the thymus of adult C57BL/6 mice induced a low leukemia incidence. This low incidence could be increased provided the inoculated mice were exposed shortly before or after intrathymic virus administration to a low nonleukemogenic dose of X rays (Haran-Ghera, 1971). This virus variant will be referred to as D-RadLV. The “D” expresses its dependence on additional coleukemogenic treatment for high leukemogenic activity in adult C57BL/6 mice. A selected serial passage line of DRadLV was found to induce a high leukemia incidence following DRadLV injection into the thymus of adult normal C57BL/6 mice. This virus variant will be cited as A-RadLV (“A” standing for its autonomous high leukemogenic activity in adult C57BL/6 mice). Latarjet and Duplan ( 1962) isolated a thymoma-inducing leukemogenic agent from thymomas induced in C57BL/6 mice by fractionated irradiation. After its in v i w passaging for several years a virus variant was obtained that exhibited a different pathological activity. This virus variant, which will be referred to in this review as RS-RadLV, was shown to affect spleen and lymph nodes of adult C57BL/6 mice, thus inducing lymphoreticular sarcomas (Mistry and Duplan, 1973). The different RadLV variants (see Table 111) that have been isolated from C57BL mice and propagated in vivo for many years were shown to
TABLE 111 ORIGINA N D LEUKEMOGENIC ACTIVITYOF RadLV VARIANTS Virus origin
Leukemogenic activity in: Newborn
virus designation
Strain
Donor treatment
Adult
Strain
Leuk.
RadLV
C57BL/Ka Thymoma induced by 170 rads x 4
C57BL/Ka
High
D-RadLV
C57BL/6J
Bone marrow collected 7 days post 170 rads x 4
C57BL16J
High
A-RadLV
C57BLl6J
Thymoma induced by intrathymic D-RadLV injected in adults
C57BL/6J
High
RS-RadLV
C57BL/6J
Thymoma induced by 170 rads x 4
C57BL/6J
High
Strain BL/Ka B 10 B6/J BlO.BR BIO.RIII BI0.S BIO.T(6R) B6N BIO.D, BIO.BR BIO.RIII BI0.S B lO.T(6R) B6/J B 10 BIO.BR BI0.M BI0.S BIO.T(6R) B6/J
Leuk. High Low Low High High High High Low Low High High Intermediate Intermediate High High High High Low High High
References Lieberman and Kaplan (1959) Meruelo r t ul. (1977a) Lonai and HaranGhera (in preparation) Haran-Ghera (1966: 1971) Lonai and HaranGhera (1977: in preparation) Haran-Ghera et al. ( 1977) Lonai and HaranGhera (1977)
Mistry and Duplan (1973)
64
N E C H A M A HARAN-GHERA A N D A L P H A PELED
affect different target cells and thereby express different leukemogenic activity. In the previous section we discussed events occurring during radiation leukemogenesis. In this section we would like to discuss leukemia induction by the RadLV variants, their characteristics, the target cells affected, their immunizing capacity, genetic control of susceptibility, and pathways during tumor development.
A. CHARACTERISTICS OF THE RADIATION LEUKEMIA VIRUSVARIANTS
1. RatlLV (Lirhrmiun and Kuplnn, 1959) Cell-free extracts prepared from radiation-induced thymoma in C57BL/ Ka when inoculated intraperitoneally into syngeneic newborn mice induced low incidence of leukemia after a long latent period. The sensitivity of the bioassay to RadLV was increased to a great extent when the virus was inoculated directly into the thymus of newborn or adult syngeneic mice or into a thymus implanted under the renal capsule of adult mice (Haran-Ghera et NI., 1966). Subsequently, t h e leukemogenic activity of cell-free extracts was amplified by serial intrathymic in LVivo passages so that it became highly leukemogenic in several-week-old C57BL/Ka mice. The in vi\v replication of RadLV in the thymus, following its administration to adult C57BL/Ka mice, and lack of virus expression in other organs (except at a low incidence in bone marrow cells) introduced the designation “thymotropic” to this leukemogenic agent (Declbve et d., 1974: 1975a). This selective localization of RadLV in the thymus was quite different from the widespread distribution of the Gross AKR virus (N-tropic) in many tissues (Rowe and Pincus, 1972). Extrathymic RadLV replication using IF assays for virus indentification was demonstrated following its administration to newborn mice. A high percentage of IF positive cells was indicated in the bone marrow and spleen cells. This virus expression decreased with increasing age. Transient extrathymic virus replication was partially restored by whole-body irradiation (this observation could be related to lymphocyte migration following X-ray treatment). Neonatal thymectomy did not diminish the capacity of bone marrow cells to maintain RadLV replication. Infectious RadLV persisted in the bone marrow since cell-free extracts prepared from these cells gave rise to high levels of viral antigen when injected into the thymus of secondary hosts, despite the lack of virus-specific antigens (detected by IF) in these bone marrow cells when extracts were prepared from them (Lieberman and Kaplan, 1976). RadLV isolated from virus-induced lymphomas in C57BL/Ka was
INDUCTION OF LEUKEMIA IN MICE
65
shown to be thymotropic (T+) and leukemogenic (L+) in vivo, though it replicated very poorly on mouse fibroblasts in vitro F,- (Decleve et a l . , 1976). Decleve et al. (1977) pointed out that the naturally occurring RadLV was distinct from the previously recognized classes of endogenous murine ecotropic and xenotropic C-type viruses since the latter were found to be fibrotropic. The T+L+ RadLV was shown to be a B-tropic ecotropic virus, replicating with one-hit kinetics in thymocytes of Fv- 1" strain and poorly in thymocytes of Fv-1" strains. Three other types of endogenous viruses isolated from C57BL/Ka mice were shown to be devoid of thymotropism or leukemogenic activity (T-L-) but did replicate on fibroblasts. (1) BUKa (B) replicated with one-hit kinetics on Fv-I"" fibroblasts and therefore was designed as a B-tropic ecotropic MuLV (Decleve et al., 1975b); (2) the RadLV", a B-tropic RadLV passaged in vitro on (BALB/c x NIH/Swiss)F,-Fv-l"" and on (C57BL/Ka x NIH)F,Fv-1"" was shown to have lost its leukemogenic activity but maintained infectivity for fibroblasts (Decleve et al. 1975b): and (3) Bl/Ka (N) replicated preferentially on Fv- I"" fibroblasts (Lieberman r t NI., 1976). Normal spleens of C57BL mice were also shown to harbor the N-tropic virus (Odaka, 1973), and it was demonstrated transiently in radiogenic lymphomas in C57BL/Ka mice (Lieberman et al., 1976). By cocultivation methods Lieberman et al. (1977a) found that normal spleen and bone marrow cells from 11- to 12-month-old C57BL/Ka mice carried the Ntropic virus besides the B-tropic ecotropic and a xenotropic virus. They recovered the xenotropic virus, designated Bl/Ka (x), by cocultivation techniques, from all lymphoid and hemopoietic organs of C57BL young and old mice. The four types of endogenous viruses isolated from C57BL/ Ka, the B-tropic, RadLV", N-tropic, and the xenotropic (x) virus were found to have differences in their envelope glycoproteins. A monospecific antiserum prepared against gp71 of an N-tropic virus readily neutralized Bl/Ka (B) and BI/Ka (N) but failed to neutralize RadLV. Antixenotropic gp71 antiserum neutralized Bl/Ka (x) as well as the xenotropic viruses present in most RadLV preparations, but failed to neutralize either the thymotropic or the leukemogenic activity of RadLV. These results suggested that the thymotropic leukemogenic RadLV (T+L+)particles differed antigenically from the B-, N-, and X-tropic endogenous viruses (Declbve et al., 1976). The designation "thymotropism" might perhaps be premature since the RadLV preparation tested consists of a mixture of viruses. The fact that infectious RadLV persists both in thymus and in the bone marrow should also be considered. The high expresseon of viral antigens in the thymus might be due to the fact that this organ consists of T lymphocyte subpopulations, whereas in other lymphoid organs the T or T precursor
66
NECHAMA HARAN-GHERA A N D ALPHA PELED
cells involve only one moiety of a variable size among different cell populations. 2. The D-RudLV Voriant (Hliran-Ghero, 1066) This virus variant was isolated from the bone marrow of nonleukemic C57BL/6 mice several days after they had been exposed to fractionated irradiation. Cell-free extracts, after several serial passages, induced a high leukemia incidence when injected intraperitoneally or into the thymus of newborn C57BL/6 mice. Virus injection into the thymus of adult mice induced a low incidence of leukemia. This incidence could be increased provided the inoculated mice were exposed shortly before or after intrathymic virus administration to a low nonleukemogenic dose of X rays (Haran-Ghera, 1971). Bone marrow shielding during exposure to X rays or bone marrow cell transfusion (treatment known to repair radiation effects) after irradiation did not change the efficient coleukemogenic effect of X rays. This virus isolate was originally marked as "passage 127" and later designated as D-RadLV (Haran-Ghera ef ul., 1977). The induction rate of leukemia in adult C57BL/6 mice was also increased using ATS or urethane as coleukemogenic agents (Haran-Ghera and Peled, 1968: Haran-Ghera, 1971: also see Table IV). It is interesting that delaying the coleukemogenic treatment for several weeks after D-RadLV intrathymic inoculation reduced the incidence of leukemia (Haran-Ghera, 1975). In contrast to the age-dependent susceptibility to radiation leukemogenesis (previously described), increased age did not affect the susceptibility of C57BL/6 mice to D-RadLV. A high incidence of leukemia was observed even when 1-year-old mice were exposed to 400 rads following intrathymic D-RadLV administration. D-RadLV has been propagated in our laboratory i n \ i \ w in thymic tissue of C57BL/6 mice. The characteristics of this variant have been confirmed in each passage by injecting DRadLV-induced tumor extracts directly into the thymus of both young adult normal and irradiated C57BL/6 mice. In this manner the low leukemogenic activity of the virus in normal unirradiated adult C57BL/6 mice is ascertained. Extrathymic D-RadLV replication, defined as leukemogenic activity in bioassays, was demonstrated following virus intraperitoneal injection into newborn C57BW6 mice. A transient leukemogenic activity (several weeks following virus injection) was demonstrated in cell-free extracts prepared from the bone marrow and spleen besides the permanent activity in preparations from the thymus (Haran-Ghera, 1972). The intrathymic infection by D-RadLV, although low or nonleukemogenic in normal adult
67
lNDUCTlON OF LEUKEMlA I N MICE TABLE IV OF LEUKEMIA I N C57BL/6 MICE BY RadLV COFACTORS AFFECTINGINDUCTION VARIANTS D-RadLV Coleukemogenic treatment”
Leukemia incidence
A-RadLV
ALP (days)
Leukemia incidence
ALP (days)
Age 1 month 6 months 12 months Virus injected intrathymically in 6week-old C57BL/6 mice plus: 400-rad WB 400-rad WB-30 days after virus 400-rad WB 1 hr lo7 BM i.v. 400-rad WB + bone marrow shielding ATS x 2b Urethane x 2c
10% -
140 -
10/10
0/10 0/10
10/10
100%
95 120
10/10
2/20
1/10 10% 8/10 80% 20/20 100% 31/40 15/20
77% 75%
120 I16 116
140
8/10 9/17
1OWc 80% 60%
120 90 190
100%~ 85 7/10 70% 80 14/16 85% 105 N.D.
16/19 15/20
84% 75%
118 107
” The coleukemogenic treatment was started 2 days after virus inoculation. N.D. = not done: WB = whole-body radiation. * ATS = antithymocyte rabbit serum: 0.5 ml injected subcutaneously twice at a five-day interval. Urethane 1 mg/gm body weight in saline, injected intrapentoneally once weekly. C57BL/6 mice, does induce gsa positivity in most thymocytes. Thus, virus expression does not necessarily indicate thymocyte transformation. The possibility (Haas, 1977a) that the D-RadLV variant converts thymocytes to MuLV gsa positive cells but is not a transforming virus seems unlikely for the following reasons: D-RadLV was shown to induce preleukemic cells in the bone marrow of C57BL/6 mice (Haran-Ghera, 1978a) and it is highly leukemogenic in several strains of mice (Lonai and HaranGhera, in preparation). 3 . The A-RadLV Variant (Peled and Haran-Ghera, 1971; Haran-Ghera et al., 1977)
Occasionally, a serial passage line of D-RadLV when injected intrathymically into normal C57BL/6 mice induced a high leukemia incidence without additional X-ray treatment. This high leukemogenic potential of the virus could be further maintained by serial passage lines of this variant. Such a variant originally selected in 1968 as “passage 136” was later designated as A-RadLV. The “A” stands for its “autonomous”
68
NECHAMA HARAN-GHERA A N D ALPHA PELED
high leukemogenic activity in its strain of origin C57BL/6 (no age restriction noted) without additional coleukemogenic treatment (Haran-Ghera et al., 1977). A-RadLV has been maintained all these years as an in i 9 i v o passage line in the thymus of normal C57BL/6 mice. Tissue culture lines grown from A-RadLV thymomas induced in C57BU6 mice have been shown to produce high titers of leukemogenic virus which act without irradiation (Haas, 1974). Murine lymphoid cells have been infected in v i m with purified virus derived from an A-RadLV tissue culture line (136.4). I n l’itro infection of thymocytes was evaluated by expression of viral antigens as detected by immunofluorescence. Thymocytes were shown to be highly susceptible to infection by this virus, whereas murine fibroblasts were refractory to it (Haas and Hilgers, 1975). For this reason the virus was labeled thymotropic. However, the extracts of A-RadLVinduced thymomas also contained ecotropic and xenotropic type-C viruses which could be propagated in iyitro on fibroblast cell lines. When these different viruses were inoculated alone and in various combinations into C57BL/6 mice, only the thymotropic virus had leukemogenic activity i n vivo (Haas, 1978a). Increased leukemogenic activity was observed after simultaneous inoculation of the thymotropic virus with either the B-tropic ecotropic virus or the xenotropic virus thus indicating a helper function of nonleukemogenic virus (Haas, 1978a). Recently Haas ( 1977b) reported that the A-RadLV-derived thymotropic virus is cytolytic for mink lung cells and that its gp70 envelope is a recombinant of xenotropic and ecotropic envelope genes [similar to the mink cell focus forming virus (MCF) demonstrated by Hartley et 01. (1977) in preleukemic and leukemic thymuses of AKR]. This recombination was suggested to occur in the thymus epithelium reticulum or after passage of the ecotropic virus through thymocytes, and was a prerequisite for transformation. Confirmation of these observations on MCF-like agent has not been made. 4. The RS-RudLV Variant (Mistry and Diiplan, 1973) Latarjet and Duplan ( 1962) isolated a thymoma-inducing leukemogenic agent from thymomas induced in C57BL mice by fractionated irradiation. After several years of serial passages in iiiv this virus exhibited a very different pathological activity. It affected spleen and lymph nodes and gave rise to lymphoreticular sarcomas (Mistry and Duplan 1973; Ricciardi-Castagnoli rt ul., 1973). This variant is designated in the present review as RS-RadLV. Two viral isolates were derived from RS-RadLVinduced reticular sarcomas (Guillemain et NI., 1977). The first is a leukemogenic virion released in high titers in the plasma of rats inoculated with a rat adapted virus (F-13). The second, 3C, has been produced from
I NDUC TI ON OF LEUKEMIA I N MICE
69
a cell line (13-3C), established from spleen cells of C57BL/6 mice infected with the RS-RadLV. This isolate is highly leukemogenic for adult C57BLl 6 mice (irrespective of t h e route of virus administration). The 3C virus isolated from tissue culture was found to be positive in XC tests and contains reverse transcriptase. The fluid of the 13-3C culture line was shown to be a mixture of ecotropic and xenotropic viruses. RNA-DNA hybridization experiments have indicated a close relationship of the RSRadLV-3C virus to Gross leukemia virus, and to a lesser degree to the Rauscher leukemia virus (Astier et a / . , 1977). Interestingly, the cells grown in v i m (originating from tumors induced by both virus variants) when injected into young adult C57BL/6 mice, induced tumors of host origin. These results suggested that cells transformed by RS-RadLV are not proliferating upon transplantation, but rather release virus which induces the disease LIP MOW (Guillemain et ul., 1977). B. TARGET CELLSAFFECTEDBY
THE
RadLV VARIANTS
Direct intrathymic inoculation of RadLV revealed that the first foci of transformed cells appeared in the outer corticoid zone of the thymus (Haran-Ghera ef a/., 1966). The first detectable sign of cytoplasmic viral antigens following RadLV infection appeared in the outer thymus cortex. It was therefore suggested that the large mitotically active cells of the thymic outer cortex serve as target cells for infection and subsequently for leukemia development (Declkve e r a / ., 1975a). The rapid thymocyte transformation was further illustrated using an experimental model comprised of thymectomized hybrid mice grafted with a parental C57BL thymus which was later repopulated by host lymphoid cells. The time of RadLV inoculation, following thymus grafting, was an important factor in determining the cellular origin of the leukemia, whether of parental or hybrid origin. The early RadLV-induced tumors were of donor origin and the rest of recipient origin. Thus, the time course for the transition of tumor origin from donor to recipient genotype paralleled the replacement of donor by recipient F, lymphocytes in thymus grafts (Kaplan, 1974). Some early experiments by Lieberman and Kaplan (1966) suggested that only thymocytes and peritoneal exudate cells acted as target cells for transformation and leukemia development. Recent experiments (Kaplan and Lieberman, 1976) have indicated that lymphoid cells removed from thymus, spleen, lymph nodes, bone marrow, and fetal liver could be infected and undergo neoplastic transformation in recipient mice. Weanling bone marrow and fetal liver cells infected with RadLV yielded a very high incidence of leukemias of donor origin in contrast to thymocytes that exhibited a reduced transformation level.
70
NECHAMA HARAN-GHERA A N D ALPHA PELED
The tropism of RadLV to specific lymphoid populations was evaluated by determing the immunological competent of T and B cells in RadLVtreated mice (Lieberman et al., 1977b). A selective impairment of several immune functions induced by RadLV inoculation suggested that RadLV interfered with normal functions of specific T-lymphocyte subpopulations: This observation reflected the thymotropic nature of RadLV. The virus, for instance, impaired or eliminated the function of the T-helper component involved in the response to SRBC. The secondary response reflected in establishment or maintenance of immunological memory was not affected by RadLV. The influence of RadLV on the T-cell function, involved in cell-mediated immune responses, was contradictory: MLC was augmented in RadLV-infected mice whereas CML activity, reflected in the ability to destroy specific target cells, was deficient. Further analysis based on Ly antigen expression indicated that the short-lived thymus-dependent lymphocytes bearing the Ly 1.2.3 phenotype were impaired following RadLV inoculation. I n contrast, functions mediated by bone marrow-derived lymphocytes were not affected by RadLV administration: for example, formation of IgM-secreting immunocytes to the thymus-independent immunogen polyvinylpyrrolidone (PVP) was not reduced by RadLV infection. The RadLV variants, A-RadLV and D-RadLV, were shown to affect different target cells of the immune system. Studies on the cellular basis of immunosuppression caused by these two variants indicated that the immunocompetent function of thymocytes, at least at the helper T-cell level, was impaired by A-RadLV, whereas D-RadLV affected the marrow cell population. The thymus-independent response to PVP or SIII was not affected by A-RadLV but was reduced following D-RadLV infection (Peled and Haran-Ghera, 1974: Haran-Ghera r t al., 1977). Both D-RadLV and A-RadLV variants did not affect the T-cell function involved in graft rejection or in graft versus host response (Peled and Haran-Ghera, 1971). The variable tropism of the RadLV variants to either thymocytes or bone marrow cells was further expressed in transformation experiments (Peled and Haran-Ghera, in preparation). A high leukemia incidence of parental C57BL/6 origin was obtained mostly when bone marrow cells from C57BL/6 origin were incubated in i'itro with D-RadLV and thereafter injected into (BALB/c X C57BL/6)F, intact irradiated recipients. The cells involved were probably prothymocytes present in the bone marrow (Haran-Ghera o r NI., 1978). The susceptible cells to A-RadLV i l l rqitro transformation, in similar experiments, were shown to be thymocytes rather than bone marrow cells. It could thus be suggested that the initial virus-thymus cell interaction after A-RadLV inoculation in C57BL/ 6 mice would lead to high leukemia induction: whereas virus-bone mar-
INDUCTION OF LEUKEMIA IN MICE
71
row interaction occurring after D-RadLV inoculation would yield a low leukemia incidence. Both D-RadLV and A-RadLV “transformed cells” required the presence of an intact thymus in order to develop into overt leukemia and remained “refractory” when injected into thymectomized mice (Kaplan and Lieberman, 1976: Peled and Haran-Ghera, in preparation). The thymic microenvironment was required for about 8 weeks following injection of the RadLV-infected target cells into the appropriate recipients. The murine leukemia viruses, although associated with thymoma development, have been shown to replicate in other organs besides the thymus (Rowe and Pincus, 1972: Lieberman et al., 1976: Haas 1977a). Moreover, other lymphoid organs besides the thymus contain cells that have the potential to act as targets for the neoplastic transformation. For example, intraperitoneal administration of A-RadLV into C57BL/6 mice following adult thymectomy resulted in the development of reticulum cell neoplasms (RCN). A high incidence occurred after a long latent period. However, cell-free centrifugates from these tumors (RCN), upon injection in the thymus of young C57BW6 mice, induced thymomas (HaranGhera and Peled, unpublished results). Law (1957) suggested that the thymus epithelial reticulum rather than thymocytes were essential for the development of the virus-infected target cells into overt T lymphomas. The thymus epithelium reticulum could provide t h e suitable microenvironment for the differentiation, maturation, and proliferation of the transformed prothymocytes into T cells: Indeed, it was recently demonstrated (Haran-Ghera, 1977; 1978a) that the proliferation of preleukemic cells to overt leukemia expression was dependent on the presence of an intact thymus at least for several weeks after preleukemic cell induction. Recently Waksal e t a / . (1976) have reported that i/i ritro monolayers of thymic epithelial reticulum supported the signal for neoplastic transformation of young normal AKR thymoctes overlaid on them. Similar observations were reported by Haas (Haas et a/., 1977: Haas 1978b), concluding that C57BL/6 normal thymocytes were transformed to leukemic thymocytes when cocultured i n v i m on thymus epithelial reticulum monolayers (TER) derived from A-RadLV-induced leukemic thymuses. Contradictory results were obtained by Peled and Haran-Ghera (1978) who demonstrated by the use of parental-F, combinations that the developing leukemias were of TER origin. The existence of hidden leukemic cells among the TER was further proven by the injection of TER cells alone into the appropriate recipients. This procedure resulted in the development of T lymphatic leukemias of the TER genotype origin. It was suggested (Peled and Haran-Ghera, 1978) that thymic epithelial reticulum might be considered as a favorable site for the maturation of
72
NECHAMA HARAN-GHERA AND ALPHA PELED
transformed prothymocytes to mature T lymphocytes, or for the autonomous proliferation of transformed T lymphocytes. C. IMMUNIZATION BY RadLV VARIANTS Murine leukemia viruses are vertically transmitted (Gross, 195I ) and have been found to be integrated into the genome of mice (Gelb (11. 1973: Chattopadhyay et nl., 1974: Rowe, 1973). Thus, it was originally thought that mice were tolerant to the antigens of their endogenous viruses (Kaplan, 1967: 1974). The fact that little resistance to the isotransplantation of virus-induced leukemias has been reported supported the suggestion that C57BL/6 mice were naturally tolerant to viral associated antigens. The availability of different RadLV variants with different leukemogenic potential introduced a means to study immunization by the RadLVs. It was found that D-RadLV inoculated in C57BL/6 mice or ARadLV in (BALBk x C57BL/6)F, hybrids led to the appearance of antitumor immunity (Haran-Ghera, 1969 and Haran-Ghera and Rubio, 1977). C57BL/6 mice inoculated with D-RadLV developed immunity to isotransplantation of leukemic cells provided D-RadLV was administered intrathymically, since any other route of virus administration (intraperitoneal, intravenous, subcutaneous, intrarenal, or intrasplenic) did not elicit transplantation resistance. The onset of resistance was noted 15 days after D-RadLV intrathymic administration and was retained for at least 90 days. Removal of the thymus from 15 days onward, after DRadLV administration, did not diminish host resistance. Direct interaction of D-RadLV thymocytes was therefore a prerequisite for the induction of an immune response. Only those virus dilutions that retained leukemogenic activity in mice (evaluated by further exposing such DRadLV-treated mice to 400 rads) were found to have an immunizing effect (Haran-Ghera, 1970). Moreover, newborn C57BL/6 mice, susceptible to all RadLV variants, when inoculated intraperitoneally with DRadLV were resistant for several months to isotransplantation of RadLVinduced leukemic cells (Haran-Ghera, 1969). Injecting D-RadLV into the thymus of young adult C57BL/6 mice prevented further leukemia induction by D-RadLV or A-RadLV as well as the proliferation of leukemic cells induced by these virus variants. This effect was specific since the mice were susceptible to X-ray-induced leukemic cells. The indication that C57BW6 mice could be immunized with D-RadLV intrathymic inoculation suggested that the low leukemia incidence obtained by delaying the coleukemogenic treatment following D-RadLV administration might
INDUCTION OF LEUKEMIA IN MICE
73
involve an immune reaction. Indeed, reinoculation of such treated mice with D-RadLV (either into the thymus in sitir or into a several-week-old intrarenal thymus graft) plus coleukemogenic treatment resulted in a reduced incidence of leukemias (Haran-Ghera, 1970). In contrast to DRadLV, which is low leukemogenic in its strain of origin (C57BL/6) and which induced immunity following intrathymic infection, the high leukemogenic variant A-RadLV failed to confer immunity in this strain. However, it was found that similar to D-RadLV i n C57BL/6 mice, the low leukemogenic potential of A-RadLV in (BALBk x C57BL/6)F1 hybrid mice resulted from immunological activation following its intrathymic administration (Haran-Ghera and Rubio, 1977). Immune lymphoid cells present in various hematopoietic organs, from day I 1 onward after D-RadLV intrathymic inoculation, were shown to mediate adoptive transfer of immunity in vivo and could also inhibit the growth of leukemic cells in vitro as tested by the cytostasis assay (Peled, 1977). Of particular interest was the finding of highly efficient antitumor neutralizing cells in the thymus of D-RadLV-injected mice (Peled and Haran-Ghera, 1971: Peled, 1977). Since it has been shown (Haran-Ghera, 1975) that D-RadLV persisted for many months in the thymus, it could be assumed that “immune” thymocytes served as a vector for a virus which immunizes the new host. This possibility was excluded by showing that adoptive transfer of immunity was restricted to viable thymocytes. However, t h e possibility that a transfer of preleukemic thymocytes confers immunity of the second host cannot be ruled out. Analysis of the cell type involved in the cell-mediated immunity to D-RadLV-induced leukemic cells in and in vifro demonstrated that cells adherent to nylon-wool column, probably of the B-cell type, were involved (Peled and Berke, 1975: Peled unpublished). The use of preferential agglutination by soybean agglutinin (SBA) lectin (Reisner ef d.,1976) permitted the recovery of an enriched population of B cells which was found to have highly inhibitory effects on leukemic cells in i T i v o and in vifro (Peled, unpublished). The specificity of the immunization conferred by D-RadLV was demonstrated by both i17 r*ivo and it? 1,irr.o tests. The transplantation resistance induced by intrathymic D-RadLV injection prevented proliferation of both D-RadLV- and A-RadLV-induced leukemias (HaranGhera and Rubio, 1977), whereas different dilutions of the high leukemogenic A-RadLV variant failed to induce transplantation resistance. The capacity of D-RadLV-induced immunization to prevent induction and proliferation of A-RadLV-induced leukemias could be due to the fact that A-RadLV originated from a D-RadLV-induced thymoma and the immune lymphoid cells were triggered by a RadLV-associated antigen common to both these variants. No cross-reactivity was found with other ~ i 1 7 0
74
NECHAMA HARAN-GHERA A N D ALPHA PELED
tumors originating in C57BL/6 mice like a syngeneic benzpyrene-induced fibrosarcoma, 3LL lung carcinoma, or the EL, leukemia (Peled, 1977). Of particular interest were the findings that this immune resistance was specific for RadLV-induced neoplasms. In parallel to the above mentioned data concerning D-RadLV-induced cell-mediated immunity, the nature and function of humoral immunity following immunization and leukemia induction by the RadLV variants has been studied. The serum of 6- to 8-week-old C57BL/6 mice receiving one intrathymic injection of D-RadLV or hyperimmune serum (obtained by two intrathymic D-RadLV inoculation at a 2-week interval followed by lo6 D-RadLV-induced leukemic cells challenged subcutaneously) neutralize the leukemogenic capacity of both D-RadLV and A-RadLV variants. These sera also contain antibodies which bind to cells bearing radiation leukemia virus antigens (e .g., to D-RadLV- or A-RadLV-induced leukemic thymocytes but not to normal thymocytes). Elevated levels of antibodies in the sera were associated with the presence of preleukemic cells in the bone marrow (Haran-Ghera, 1969: Peled and Haran-Ghera, 1971: Haran-Ghera and Rubio, 1977). Serum IgG and IgM titers were found to be low, at most up to 1:64 dilution in immunized mice and 1:64-1: 132 in hyperimmunized mice versus 1: 16 dilution in sera of 6-month-old normal C57BL/6 mice. D-RadLV and A-RadLV leukemogenicity could be neutralized with immune serum or immune y-globulin fraction (C. M. Friedman, unpublished data). Intraperitoneal injection of D-RadLV into newborn C57BL/6 mice yields a long latent period of 5 months to I year before development of thymomas in up to 60% of the injected mice. Sera taken from these animals at different ages exhibited antiviral neutralizing activity up until several weeks before overt leukemia appeared: then this activity abruptly disappeared (Haran-Ghera, 1972). These findings are suggestive of a humoral immune role in the dependent preleukemic stage and its abrogation in the critical autonomous proliferation stage. The reason for this decrease in humoral response is not known. It could be related to viral replication in specific immune cells (see above). Normal C57BL/6 mice harbor endogenous C-type MuLV of ecotropic and xenotropic varieties (Levy, 1978). Natural antiviral binding antibodies to endogenous MuLV as demonstrated by RIA have been reported in adult C57BL mice (Nowinski, 1974; Martin and Martin, 1975: Hanna et d.,1976). Neutralization of fibroblastic ecotropic MuLV has not been regularly observed (J . A. Levy, personal communication). Sera collected from normal C57BL/6 mice at different ages, however, have been found to have neutralizing antibodies to D-RadLV and A-RadLV variants (Haran-Ghera, 1975). Conceivably these naturally occurring antibodies
INDUCTION OF LEUKEMIA IN MICE
75
against RadLV-induced leukemic cells might also be due to increased titers of the endogenous RadLV in C57BL/6 mice (Haran-Ghera and Peled, 1967) and to the fact that aged C57BL/6 mice are carriers of preleukemic cells (Haran-Ghera, 1978a). The "nonspecific" reactions of the above mentioned immune sera, due to endogenous viral antigen expression, was absorbed out with normal spleen, thymus, and bone marrow cells, and thereafter specific antipreleukemic and leukemic antibodies have been demonstrated. Various in v im tests of antibody functions were carried out to determine whether a simple leukemic-cell damage function could be attributed to serum antibodies. Complement-dependent antibody-mediated cytotoxicity experiments indicated that circulating antitumor antibodies raised by D-RadLV intrathymic inoculation in C57BL/6 mice or A-RadLV in F1 hybrids lysed the target D-RadLV-induced leukemic cells. Conversely, no cytotoxic activity was found in sera of C57BL/6 mice treated with ARadLV (Haran-Ghera and Rubio, 1977). Recent studies have described a method to immunize C57BL/6 mice against RadLV by vaccination (Lieberman and Kaplan, 1977). The injection of C57BL/Ka mice with virus-producing B 1-5 (RadLV) cells induced immunization against both transplantation of leukemic cells and actual induction of leukemia by RadLV. The authors believe that the immune response was directed toward antigenic determinants of RadLV. Furthermore, Treves et al. (1977) demonstrated generation of anti-B1-6 (RadLV) T-cytotoxic lymphocytes which were sensitized via RadLV-fed macrophages. The fact that Kaplan's RadLV was not able to serve as an immunizing agent like D-RadLV might be due to actual differences between the different RadLV variants. His agent appears to act like ARadLV in C57BL/6 mice. Genetic studies of resistance and susceptibility to RadLV leukemogenesis (Meurelo ef a / . , 1977a,b; Lonai and HaranGhera, 1977) indicate differences among the RadLV variants which might be related to the varied origins of these virus isolates (see Table 111).
D. GENETICCONTROL OF SUSCEPTIBILITY It has already been shown many years ago (MacDowell and Richter, 1935) that susceptibility to the development of leukemia is genetically inherited in certain strains of mice. Host immune response has been considered as one possible factor of this host resistance to leukemia development. Immune responsiveness to many antigens has been shown to be under genetic control, regulated by I r genes linked to the major histocompatibility complex within the H-2 complex (Benacerraf and
76
NECHAMA HARAN-GHERA A N D ALPHA PELED
McDevitt, 1972; Dorf et a l . , 1975). Thus, the variable susceptibility of C57BL mice to leukemogenesis by the different RadLV variants led to a survey of the relationship and function of H-2 linked resistance and sensitivity genes to the virus variants. In assessing this genetic control of resistance, the leukemogenic activity of A-RadLV was evaluated in different congenic strains with the same independent H-2 haplotype on at least two different backgrounds. The Fv-1” type congenic series on the B6, BlO, BALB, and A background were found to be sensitive to ARadLV leukemia induction, while the Fv-1” type on the C3H and DBA were resistant to leukemogenesis by A-RadLV. These results were in agreement with the it7 i i t r o studies which suggested that RadLV is a Btropic virus (Declkve et al., 1975b). The resistant locus to A-RadLV designated as Rrv-1 was mapped to subregions I-A and I-B of the H-2 complex. Lonai and Haran-Ghera (1977) suggested that Rrv-1 may be in complementation with a second locus to the right of it between Rrv-1 and H-2D. This localization and complementation of the two loci for resistance were characteristic of I r genes (Schreffler and David, 1975) and indicated a possible relationship between the genetic regulation of immune responsiveness and susceptibility to leukemia. The resistant versus sensitive H-2 haplotype for A-RadLV leukemia (Lonai and Haran-Ghera, 1977) were found to be different from those described for the Gross leukemia virus (Lilly and Pincus, 19731, as well as for the RadLV preparation used by Meruelo er al. (1977a). When comparing the sensitivity of different Fv- l h congenic mouse strains to leukemia induction by A-RadLV, D-RadLV, and RadLV (provided by Kaplan in 1967 and subsequently passaged in C57BL/6 mice in Haran-Ghera’s laboratory), it was found that the H-2 linked pattern of resistance to these RadLV variants was different (Lonai and HaranGhera, in preparation). All three variants, including D-RadLV, which was previously shown to be low leukemogenic in C57BL/6 mice [strain from which D-RadLV was originally isolated and induced high leukemia incidence following coleukemogenic treatment (Haran-Ghera, I97 111, were highly leukemogenic in different strains of mice. The most apparant dissimilarity between A-RadLV and the other two variants was that ARadLV was highly oncogenic in strains carrying H-2”, whereas D-RadLV and RadLV were weakly oncogenic in this haplotype. RadLV which, in the original studies of Kaplan and associates, was highly leukemogenic in the strain or origin (C57BL/Ka H-2”) was found to be weakly leukemogenic in C57BL/lO(H-2”)(Meruelo e t al., 1977a). The data of Meruelo rt al. (1977a) on RadLV are basically similar to that of Lonai and HaranGhera (in preparation) with RadLV and suggest that the sensitivity of
INDUCTION OF LEUKEMIA IN MICE
77
C57BL/Ka mice to RadLV could be explained by non-H-2 linked resistance genes in this strain. Meruelo and associates ( 1977a,b) studied recently the genetic control mechanism involved in leukemogenesis by Kaplan’s RadLV isolate. AIthough RadLV was originally isolated from the C57BWKa strain, other strains were shown to be susceptible or resistant to RadLV leukemogenesis. The resistance was shown to be associated with a gene(s) in the D region of the H-2 complex (Meruelo et ul., 1977a). Thus, the strains BlO.D, and B1O.S showed marked susceptibility to RadLV, whereas BIO. T(6R) and BlO.S(7R) mice were resistant. Among a variety of C57BL/IO derived H-2 congenic and recombinant strains of mice, all strains having the H-2D“ allele were markedly resistant to the disease, whereas those strains having the H-2Dg and H-2D alleles were susceptible. These observations suggested that the resistance to RadLV-induced leukemia is inherited as a dominant gene. Surprising, the BIO with the H-2D” allele (like the C57BL/6) was found to be relatively resistant (Meruelo c’r crl., 1977a). Besides the linkage of leukemogenic resistance to H-2, Meruelo rt 01. (1977b) found an additional gene not linked to H-2 with an influence on the RadLV leukemogenesis designated Srlv- 1. This gene confers susceptibility to RadLV-induced leukemogenesis and is dominantly expressed. The mechanism of action of Srlv-1 is presently unknown, although preliminary data indicate that it might affect virus replication. Furthermore, studying the mechanism by which H-2D linked genes confer resistance to RadLV-induced leukemias, Meruelo rr crl. (1978) showed that H-2 linked genetic control apparently does not influence initial virus replication, but instead alters virus spread. Expression of H-2K determinants was significantly increased on the cell surface of the majority of thymocytes of infected mice whether they were susceptible or resistant H-2 types. In contrast, an increase in H-2D expression was detected on thymocytes of resistant H-2 haplotypes, and not on thymocytes of susceptible H-2 types. The authors, therefore, suggested that altered expression of H-2 antigen plays a very significant role in the mechanism of host defense to virus infection. Chazan and Haran-Ghera (1976) had previously reported that cells bearing high levels of H-2 antigens appeared in abundance in the thymus of RadLV-injected mice. These cells were present in greater amount than that usually found in the normal young adult thymus of C57BL/6 mice. Their data suggested that most of the spontaneous leukemias of the AKR strain or the T leukemias induced by RadLV originated from a minor cell subpopulation of the thymus, which consisted of thymocytes bearing high levels of H-2 antigen and low levels of Thy-I. Although it is suggested by Meruelo rr nl. (1978) that the
78
NECHAMA HARAN-GHERA A N D ALPHA PELED
increase in H-2 expression after RadLV infection occurs on all cells of the thymus, the possible involvement of a subpopulation expansion cannot be ruled out. Such a phenomenon could occur by an extremely rapid proliferative rate of a small percentage of thymocytes that populates a thymus which is devoid of cells susceptible to a selective thymolytic effect of D-RadLV (Haran-Ghera, 1968). The fact that D-RadLV (originally isolated from irradiated nonleukemic bone marrow of C57BL/6 mice) and RadLV (originating from a radiogenic thymoma in a C57BL/Ka mouse) were shown to be weakly leukemogenic in their haplotype of origin (Meruelo et al., 1977a; Lonai and HaranGhera, in preparation) suggests that a selection mechanism under H-2 control may be responsible for the preleukemic stage described by HaranGhera (1978a). Lonai and Haran-Ghera raised the possibility that ARadLV which is highly leukemogenic in its haplotype of origin may be an example of an oncogenic virus that “slipped through” this selection mechanism. One likely mechanism for resistance to RadLV-induced leukemogenesis would be tumor surveillance via immunologically competent cells and/or regulation or production of antibodies associated with a locus with similar localization. Evidence for involvement of a locus (X-1) with similar localization to Rgv-1 (Lilly, 1970) which directs immune responsiveness to transplantable radiation-induced leukemias has been described by Sat0 et al. (1973). Support for this assumption comes also from the findings of Aoki et a / . (1966) that mice of the resistant H-2 haplotype showed detectable anti-Gross virus antibodies. All these findings suggest that H-2 linked Ir genes are involved in the host’s immune response to leukemogenesis.
E. PHASESI N RadLV LEUKEMOGENESIS A prerequisite for studying the occurrence of phases during tumor development is the demonstration of tumor cells shortly after the tumorigenic treatment or during the long latent period. Using a transplantation bioassay method, the presence of preleukemic cells in mice at a stage when they do not show clinical manifestations of the disease has recently been described. Potential leukemic cells designated as preleukemic cells could be detected in almost each mouse treated with different leukemogenic agents (X rays, chemical carcinogens, and RadLV variants) irrespective of the ultimate occurrence of the disease due to the leukemogenic treatment (Haran-Ghera, 1973; 1976; 1977; 1978a,b). For example, a persistent presence of preleukemic cells was demonstrated among the bone marrow cell population of mice injected intrathymically with D-
INDUCTION OF LEUKEMIA IN MICE
79
RadLV (a treatment which ultimately yielded only a very low incidence of overt leukemia). Thus, bone marrow was the main site for D-RadLVinduced potential leukemia cells detected as early as 10 days after DRadLV infection (Haran-Ghera, 1978a). The proliferation of transferred preleukemic cells was shown to depend on exposure of the recipients to a low nonleukemogenic dose of X rays. These findings indicated that the proliferation of these preleukemic cells were dependent on specific host environment, whereas leukemic cells from overt neoplasms could proliferate equally well in normal and irradiated recipients. Experimental evidence, provided recently, has suggested that qualitative rather than quantitative differences existed between the early occurring preleukemic cells and leukemic cells from overt diseased mice which were detected only after a prolonged latent period (Haran-Ghera et a / ., 1978). The potential leukemia-incuding cells detected in the bone marrow shortly after D-RadLV inoculation differed in several respects from the leukemic cells present in the clinically expressed disease (Haran-Ghera et ul., 1978: Haran-Ghera, 1978b). (1) The leukemic cells were shown to have the phenotype of the minor thymus subpopulation, expressing high levels of H-2 alloantigen and relatively low levels of Thy-I. In contrast, preleukemia cells were found to be refractory to lysis by anti-Thy- 1 serum and complement but were inducible to become Thy-I+ following in vitro incubation with thymopoietin or ubiquitin. These results suggested that the preleukemia cells present in the bone marrow shortly after leukemogenic treatment were committed precursor T cells, prothymocytes. (2) The preleukemic cells were found to differ in their immunogenicity; thus D-RadLV-induced preleukemic cells, in contrast to leukemic cells, induced resistance to subsequent challenge with leukemia cells. The presence of these immunogenic preleukemic cells, following intrathymic inoculation of D-RadLV, could perhaps be partly responsible for the buildup of immunity in C57BL/6 mice following D-RadLV injection (discussed in Section 111,C). (3) The proliferation of preleukemic cells was shown to be dependent on specific environmental requirements, in contrast to leukemic cells that proliferated equally well in normal or irradiated syngeneic recipients. The specific factors required for the proliferation of preleukemic cells varied depending on whether they were induced by low or high potential leukemogenic agents. These requirements were evaluated by transplanting preleukemic cells at various stages following leukemogenic treatment into syngeneic intact or thymectomized irradiated or normal hosts (Haran-Ghera, 1977: 1978b). Tissues taken from D-RadLV-treated mice developed into overt leukemias only when transplanted into intact irradiated recipients, but never in normal recipients. In mice treated with D-RadLV followed by radiation (therby induc-
80
NECHAMA HARAN-GHERA A N D ALPHA PELED
ing high leukemia incidence) the preleukemic cells would proliferate both in irradiated and normal recipients at about 80 days following virus injection, several weeks before clinical expression of the disease. In mice treated with A-RadLV the preleukemic cells could develop into overt leukemia when transferred into normal recipients already 30 days after their induction. The above mentioned differences between the early occurring potential leukemia-inducing cells justified their designation as preleukemic cells, dependent on host factors and thymus microenvironment for their further differentiation into T leukemias. The occurrence of several fundamental phases preceding the clinical tumor expression are therefore suggested: ( 1 ) the transformation of normal cells, probably prothymocytes into preleukemic cells and (2) further differentiation and/or proliferation of the transformed cells into autonomously growing cells, probably controlled by host factors. Proliferation of preleukemic cells may involve a ”dependent” and an “autonomous” phase. The autonomous stage in which cells proliferate irrespective of host conditions might be an essential phase preceding the final overt expression of the disease. The high or low potential of the leukemogenic agents might be related to their initial capacity to induce “dependent” or “autonomous” immortalized transformed cells. The capacity of D-RadLV to induce a high incidence of preleukemic cell, (irrespective of its low overt leukemia induction capacity), might indicate that a leukemogenic agent could exert its transformation effect without necessarily providing proliferative capacity that would lead to overt tumor development. Thus, D-RadLV preleukemic cells induced in C57BL/6 mice remained always in a “dependent” stagetheir further proliferation depending on the presence of specific environments (radiation, thymus subpopulations, time sequence between virus inoculation, and coleukemogenic treatment). Irradiation of D-RadLVinoculated mice shortly after virus inoculation introduced the required proliferative environment for the D-RadLV-induced preleukemic cells, when residing in such treated hosts for longer periods, and changed ultimately their characteristics to autonomous variants. The requirement of 30 days for the early expression of autonomy demonstrated in ARadLV-induced preleukemic cells might express quantitative rather than qualitative changes during this short period. The transition from dependency to autonomy might involve a single differentiation course, from the stem-cell-prothymocyte phase to subsequent maturation afforded by thymus microenvironment. The recent findings concerning the “phagocytic” activity of thymus epithelial reticulum suggest that these cells might be considered as a favorable site for the autonomous transformation of leukemic cells. Thus, an autonomous
INDUCTION OF LEUKEMIA IN MICE
81
clonal selection of leukemic cells might be related to their prolonged existence within the “phagocytic” cells of the thymus epithelial reticulum (Peled and Haran-Ghera, 1978). One could also suggest the possibility that during sustained proliferation, the preleukemic cells may undergo somatic mutation or chromosomal rearrangement and thereby become autonomous. In recent cytogenetic studies (Wiener et al., 1978) most of the A-RadLV-induced leukemic cells where shown to possess a tumor stem line with a chromosome constitution of 41 chromosomes (much less in the dependent D-RadLV-induced preleukemic cells). The G-banding pattern analyzed in all these leukemias with the hyperdiploid stem line of 41 chromosomes revealed the presence of trisomy in chromosome 15. Thus, the trisomy 15 in the majority of A-RadLV-induced leukemias might be the end result of selection of transformed cells with an unbalanced gene load, endowing the preleukemic cells with the selective growth advantage necessary to achieve full autonomy. Recently Dofoku et a / . (1975) and Chang et al. (1977) have shown that chromosome 15 is frequently present in a trisomic state in spontaneous lymphomas in AKR mice and in X-ray-induced mouse leukemia in the C57BL strain. It could thus be suggested that these specific chromosomal aberrations might be a prerequisite for full autonomous development of malignancy in i s i v o . The specificity of this chromosomal aberration seems to be related with the prothymocyte-thymocyte differentiation pathway. IV. Concluding Remarks
The target organ for the overt development of most spontaneous or induced murine lymphoid leukemias is the thymus. The thymus has, therefore, been considered by many investigators as the site of neoplastic transformation and proliferation. In recent studies the bone marrow has been shown to be the primary site of transformation by different leukemogenic agents, like chemical carcinogens, fractionated irradition, or leukemogenic viruses. The induction of transformation by different leukemogenic agents was established both in adult intact and thymectomized mice; these studies stressed the occurrence of transformation in the absence of the thymus. Potential leukemic cells have been identified in the bone marrow of most mice treated with leukemogenic agents irrespective of the ultimate overt leukemia development expressed mostly in the thymus. Experimental data has suggested that the initial transformation by leukemogenic agents involves prothymocytes present among the bone marrow subpopulations. These transformed preleukemic prothymocytes require the thymus microenviranment to further proliferate
82
NECHAMA HARAN-GHERA A N D ALPHA PELED
into overt mature T leukemia cells. Thymus injury and its regenerative process, induced by the different leukemogenic agents, is probably an important facet in leukemogenesis. The ability to abrogate radiogenic leukemogenesis by syngeneic normal bone marrow transfer or by marrow shielding during irradiation may involve this second phase of leukemogenesis. These procedures may enhance thymus repopulation and thereby prevent differentiation and/or proliferation of the preleukemic cells originating in the bone marrow. The development path ways during leukemogenesis seem to involve at least two fundamental phases preceding the overt expression of a tumor: (1) transformation of normal stem cells or prothymocytes into preleukemic cells and (2) differentiation and proliferation of the transformed cells into autonomously growing malignant cells. The leukemogenic potential (high or low) of the different leukemogenic agents is determined by different tumor cell-host relationships. A low leukemogenic agent leads to the formation of a preleukemic “dependent” clone of cells. This transformation in itself is insufficient to cause tumor development and has sometimes the capacity of contributing to a state of immunization. A latent period ensues during which the transformed cells are dependent on their environment, their proliferation is controlled and a host-tumor balance is maintained. At this stage, host immunity may have a significant role. The presence of various cofactors (not necessarily leukemogenic by themselves) is necessary before the “dependent” clone can become established into an “autonomous” clone. Host irradiation, the availability of specific lymphoid subpopulations, intact thymus, and immune impairment are but a few of these potential cofactors that affect this proliferation phase. The ultimate autonomous transformation (affected by uncontrolled differentiation factors, chromosomal aberrations, or chromosomal rearrangements) may permit their escaping from host control and environmental influences. Comparative studies on the similarity and diversity of initial developmental pathways in leukemogenesis by X rays or by RadLV variants question the relationship between these two types of leukemogenic agents and suggest that one or the other alone can be directly involved in the transformation event. The presence of an intact thymus was shown to be a prerequisite for the establishment and proliferation of preleukemic cells induced by the RadLV variants but not for those cells induced by fractionated wholebody irradiation. Moreover, viral antigen expression was always demonstrated in RadLV-induced tumors but quite rarely in radiation-induced tumors, and immunological cross-reactivity was demonstrated among RadLV-induced
INDUCTION OF LEUKEMIA IN MICE
83
tumors but not in radiation-induced tumors. This absence of viral expression in radiogenic lymphomas raises doubts as to whether the passive immunotherapy approach using antisera against RadLV-coded cell surface antigens involves a specific antiviral effect. Such antisera may induce a nonspecific cytotoxic effect on available preleukemic cells or may affect the availability of specific lymphoid subpopulations or other host factors that are essential for the autonomous proliferation of preleukemic cells. In conclusion one must consider the possibility that initial transformation of target cells to preleukemic cells might involve a viral etiology in some cases and a direct X-ray or a chemical carcinogen effect in others. Further differentiation and/or proliferation of these transformed cells to autonomous leukemic cells might be controlled by numerous other factors including physiological and/or genetic host factors and even perhaps viruses (e.g., xenotropic virus) not necessarily having leukemogenic potential. The transition from the “dependent” phase to autonomy is a prerequisite for overt leukemia development, and toward this end viruses may not always be required. In essence both viral and nonviral etiology should be considered in murine leukemogenesis.
ACKNOWLEDGMENTS We are indebted and most grateful to Dr. Jay A. Levy for valuable discussions and criticism of this manuscript. We also wish to thank Dr. P. Lonai for helpful criticism and Mrs. Esther Majerowich for patient assistance in the preparation of the manuscript. Work cited from the authors’ laboratory was supported by Public Health Service Contract N01CB-43930 and N01-CB-74151 from the Division of Cancer Biology and Diagnosis, National Cancer Institute.
REFERENCES Allison, A. C., Berman, L. D., and Levey, R. H. (1967). Naturr (Londun) 215, 185-187. Aoki, T., Boyse, E. A., and Old, L. J. (1966). Cancer Res. 26, 1415-1419. Armuth, V. (1976). Br. J. Cancer 34, 516-522. Amstein, P., Riggs, J. L., Oshiro, L. S., Huebner, R. J., and Lennette, E. H. (1976). J. Natl. Cancer Inst. 57, 1085-1090. Astier, T., Janowski, M., Guillemain, B., Baugnet-Mahiev, L., Maisin, J. R., and Duplan, J. F. (1977). In “Radiation’-Induced Leukemogenesis and Related Viruses ”( J . F. Duplan, ed.), INSERM Symp. 4. Elsevier/North-Holland Biomedical Press, Amsterdam. Balner, H . (1971). Rev. Eur. Etud. Clin. Biol. 16, 981-986. Barnes, D. W. H., Ford, C. E., Ilbery, P. L. T., Jones, K. W., and Loutit, J. F. (1959). Acta Unio Int. Contra Cancrum 15, 544-548. Benacerraf, B., and McDevitt, H. 0. (1972). Science 175, 273-274.
84
NECHAMA HARAN-GHERA A N D ALPHA PELED
Bollum. F. J . (1974). In "The Enzymes.'' (P. D. Boyer, ed.), Vol. 10, pp. 145-171. Academic Press, New York. Boniver, J., Simar, L. J., Courtoy, R., and Betz, E. H. (1978). Canccr Res. 38, 52-57. Cerottini, J . C., and Brunner, K. T. (1967). Itnmrrnology 13, 395-403. Chang, T. D., Biedler, J . L., Stockert, E. et al. (1977). AACR Abst.. 225. Chattopadhyay, S . K., Lowry, D. R. Teich, N. M., Levine, A . S . , and Rowe, W. P. (1974). Proc. Ncrtl. Acud. Sci. U.S.A. 71, 167-171. Chazan, R., and Haran-Ghera, N. (1976). Cell. Immrrnol. 23, 356-375. Decltve, A., Lieberman, M., and Kaplan, H. S. (1977). Virology 81, 270-283. Decltve, A., Sato, C. Lieberman, M., and Kaplan, H. S. (1974). Proc. Natl. Acud. Sci. U.S.A. 71, 3124-3128. Declkve, A , , Travis, M., Weissman. I . L., Lieberman, M., and Kaplan, H. S. (1975a). Cancer Rcs. 35, 3585-3595. Declkve, A , Niva, 0.. Gelman, E. P., and Kaplan, H. S. (l975b). Virologv 65, 320-332. Decltve, A., Lieberman, M., Ihle, J . M., and Kaplan, H. S. (1976). Proc. Natl. Accrd. Sci. U.S.A. 73, 4675-4679. Dofoku, R.. Biedler, J. L., Spengler, B. A., et al. (1975). Proc. Nurl. Acud. Sci. U . S . A . 72, 1515-1517. Dorf, M. J., Stimpfling, J. H., and Benacerraf, B. (1975). J . Exp. Med. 141, 1459-1463. Dunn, T . B., Moloney, J. B., Green, A. W., and Arnold, B. (1961). J . Nutl. Cancr,:fnsr. 26,189-22 I . Fenjo, E. M., Klein, G. (1976). Natrrre (London) 260, 355-356. Fenjo, E. M., Yefenof, E., Klein, E . , and Klein, G . (1977). 1.Exp. Mrd. 146, 1521-1533. Furth, J. (1946). J . Geront. 1, 46-57. Furth, J . , and Boon, M. C . (1944). AAAS Res. Conf. Cancer 128-138. Furth, J., and Furth. 0. (1936).Am. J . Cancer 28, 54-65. Gelb, L. D., Milstein, J. B.. Martin, M. A., and Aaronson, S. A. (1973). Nature, New Biol. 244, 76-79. Gross, L. (1951). Proc. Soc. Exp. Biol. 76, 27-32. Gross, L. (1958). Acta Haemutol. 19, 353-361. Gross, L., and Feldman, D. G. (1968). Cancer Res. 28, 1677-1685. Guillernain, B.. Mamoun, R., Astier, T., Portail, J . P . , Legrand, E., and Duplan, J . F. (1977). In "Radiation-Induced Leukemogenesis and Related Viruses" (J. F. Duplan, ed.) INSERM Symp. 4, pp. 297-309. Elsevier, Amsterdam. Haas, M. (1974). Cell 1, 79-83. Haas, M. (1977a). J . Nail. Cancer Inst. 58, 251-257. Haas, M. (1977b).In "Advances in Comparative Leukemia Research" (Bentvelzen. Hilgers and Yohn, eds.), pp. 115-126. ElseviedNorth-Holland Biomedical Press, Amsterdam. Haas, M. ( 1978a).J . Virol. 25, 705-709. Haas, M. (1978b). Int. J. Cancer 21, 115-120. Haas, M., and Hilgers, J. (1975). Proc. Nail. Acad. Sci. U.S.A. 72, 3546-3550. Haas, M.,Sher, T., and Smolinsky, S. (1977). Cancer Res. 37, 1800-1807. Hanna, M. G., Ihle, J . N., Lee, J. C. (1976). Cancer Res. 36, 608-614. Haot, J., Betz, E. H.. and Revesz, L. (1973). Natrrre, NPWBiol. 244, 21 1-212. Haran-Ghera, N. (1966). In:. J . Cancer 1, 81-87. Haran-Ghera, N. (1968). Isr. J . Med. Sci. 4, 1169-1 180. Haran-Ghera, N. (1969). Nature (London) 222, 992-993. Haran-Ghera, N. (1970). Bibliogr. Haem. 36, 261-266. Haran-Ghera, N. (1971). Isr. J. Med. Sci. 7, 17-25. Haran-Ghera, N. (1972). Nature, New Biol. 238, 21-23.
INDUCTION OF LEUKEMIA IN MICE
85
Haran-Ghera, N. (1973). Nature, New B i d . 246, 84-86. Haran-Ghera, N. ( 1975). In “Dynanmic Aspects of Host-Parasite Relationships” (A. Zuckerman, ed.), pp. 107-131. Halsted Press, New York. Haran-Ghera, N. (1976). In ”Biology of Radiation Carcinogenesis” (J. Yuhas, R. Tennant and J. Regen, eds.), pp. 245-260. Raven Press, New York. Haran-Ghera, N. (1977). In “Radiation-Induced Leukemogenesis and Related Viruses” ( J . F. Duplan, ed.), INSERM Symp. 4, pp. 79-89. Elsevier, Amsterdam. Haran-Ghera, N. (1978a). J . Natl. Cancer Inst. 60,707-710. Haran-Ghera, N. (1978b). Leitk. Res. 2, 11-22. Haran-Ghera, N., and Peled, A. (1967). Br. J . Cancer 21, 730-738. Haran-Ghera, N. and Peled, A. (1968). Isr. J. Med. Sci. 4, 1181-1 187. Haran-Ghera, N., and Peled, A. (1973). Nature (London) 241, 396-398. Haran-Ghera, N., and Rubio, N. (1977). J . Immrtnol. 118, 607-61 1 . Haran-Ghera, N., and Kaplan, H. S. (1964). Cancer Res. 24, 1926-1931. Haran-Ghera, N., Ben-Yaakov, M., and Peled, A. (1977). J . Immunol. 118, 600-606. Haran-Ghera, N., Kotler, M., and Meshorer, A. (1967). J . Natl. Cancer Inst. 39, 653-661. Haran-Ghera, N., Lieberman, M., and Kaplan, H. S. (1966). Cancer Res. 26,438-441. Haran-Ghera, N., Rubio, N., Leef, F., and Goldstein, G. (1978). Cell. Immrtnol. 37, 308314. Hartley, J. W., Wolford, N. K., Old, L. J . , and Row, W. P. (1977). Proc.. N u t / . Acad. Sci. U.S.A. 74, 789-792. Hirsch, M. S., and Murphy, 1. A. (1968). Nature (London) 218, 478-479. Ihle, J. N. (1977). I n “Radiation-Induced Leukemogenesis and Related Viruses” ( J . F. Duplan, ed.), INSERM Symp. 4, pp. 265-274. Elsevier, Amsterdam. Ihle, J. N. (1978). Semin. Hemato/. 15, 95-1 15. Ihle, J. N., Joseph, D. R., and Pazmino, N . H . (l976b).J. Exp. Med. 144, 1406-1423. Ihle, J. N., McEwan, R., and Bengali, K. (1976a).J . Exp. Med. 144, 1391-1405. Ihle, J. N., Yurconic, M., and Hanna, M. G. (1973). J. Exp. Med. 138, 194-208. Ilbery, P. L. T . (1967). Nature (London) 215, 655-656. Ilbery, P. L. T., and Winn, S. M. (1964). Aust. J . Exp. Biol. Med. Sci. 42, 133-148. Jenkins, V. K., and Upton, A. C. (1963). Cancer Res. 23, 1748-1755. Kaplan, H. S. (1948). J. N u t / . Cancer Inst. 9, 55-66. Kaplan, H. S. (1949). J. Natl. Cancer Inst. 10, 267-270. Kaplan, H . S. (1950). J. Natl. Cancer Inst. 11, 83-86. Kaplan, H. S. (1964). J. Natl. Cancer Inst. Monogr. 14, 207-217. Kaplan, H. S. (1967). Cancer Res. 27, 1325-1340. Kaplan, H . S. (1974). Series Haematol. 7, 94-167. Kaplan, H. S., and Brown, M. B. (1951). J. Nut/. Cancer Inst. 12, 427-436. Kaplan, H. S., and Brown, M. B. (1952). J . Nut/. Cancer Inst. 13, 185-192. Kaplan, H . S., and Brown, M. B. (1954). Science 119, 439-441. Kaplan, H. S., and Lieberman, M. (1976). Blood Cells 2, 301-317. Kaplan, H. S., Brown, M. B., and Paull, J. (1953). J . Natl. Cancer Inst. 14, 303-316. Kaplan, H . S., Carnes, W. H., Brown, M. B., and Hirsch, B. B. (1956). Cancer Res. 16, 422-425. Kirschbaum, A., and Liebelt, A. (1945). Cancer Res. 15,689-691. Kung, P. C., Silverstone, A. E., McCaffrey, R. P., and Baltimore, D. (1975).J. Exp. Med. 141, 855-865. Latarjet, R., and Duplan, J. F. (1962). I n t . J. Radiat. Biol. 5, 339-344. Law, L. W. (1957). Ann. N. Y. Acad. Sci. 68, 616-635. Law, L. W., and Miller, J. H . (1950). J. Natl. Cancer Inst. 11, 253-256.
86
NECHAMA HARAN-GHERA AND ALPHA PELED
Law, L. W., and Potter, M. (1956). Proc. Natl. Acad. Sci. U . S . A . 42, 160-167. Law, L. W., and Potter, M. (1958). J. Nail. Cancer Inst. 20,489-493. Laznicka, M., and Smetanova, K. (1963). Bull. Assoc. Fr. Erud. Cancer 50, 651-657. Levy, J. A. (1978). In "Current Topics in Microbiology and Immunology," Vol. 79, pp. 113-213. Springer-Verlag, Berlin, New York. Lieberman, M., and Kaplan, H. S. (1959). Science 130, 387-388. Lieberman, M., and Kaplan, H. S. (1966). Nail. Cancer Inst. Monogr. 22,549-557. Lieberman, M., and Kaplan, H. S. (1976). Blood Cells 2, 291-299. Lieberman, M., and Kaplan, H. S . ( 1977). 1n"Biology of Radiation-Induced Leukemogenosis and Related Viruses" (J. F. Duplan, ed.) INSERM Symp. 4, pp. 127-132. Elsevier, Amsterdam. Lieberman. M., Kaplan, H. S., and Decleve, A. (1976). In "Biology of Radiation Carcinogenesis" (J. M. Yuhas, R. W. Tennant and J. D. Regan, eds.), pp. 237-244. Raven Press, New York. Lieberman, M., Merigan, J. C., and Kaplan. H . S. (1971). Proc. Soc. Exp. Biol. 138, 575578. Lieberman, M., Decleve, A., Gelmann, E. P., and Kaplan, H. S. (1977a). I n "RadiationInduced Leukemogenesis and Related Viruses" (J. F. Duplan, ed.) INSERM Symp. 4, pp. 23 1-246. Elsevier, Amsterdam. Lieberman, M., Segal, S., Finn, O., Zan-Bar, I . , and Kaplan. H. S. (l977b). In "RadiationInduced Leukemogenesis and Related Viruses" (J. F. Duplan, ed.) INSERM Symp. 4, pp. I 15- 125. Elsevier, Amsterdam. Lilly, F. (1970). Bihliogr. Hacvnatol. 36, 213. Lilly, F., and Pincus, T. (1973). Adv. C'trnc~~r Res. 17, 231-277. Lonai, P., and Haran-Ghera. N. (1977). J. E x p . Med. 146, 1164-1 168. Lorenz, E., Congdon, C. C., and Uphoff, D. (1953). J . Natl. Cuncw I n s t . 14, 291-299. MacDowell, E., and Richter, M. N . (1935). Arch. Path. 20, 709-71 I . Martin, S. E., and Martin, W. J. (1975). Int. 1. Cancer 15, 658-664. Meruelo, D., Lieberman, M., Deak, B., and McDevitt, H. 0. (1977b). J. Exp. Meil. 146, 1079- 1087. Meruelo, D., Lieberman, M., Ginzton, N., Deak, B., and McDevitt, H. 0. (1977a). J . E.rp. Mrd. 146, 1079- 1087. Meruelo, D., Nimelstein, S. H., Jones, P. P., Lieberman, M., and McDevitt, H. 0. (1978). J . E i p . M d . 147, 470-487. Micklem, H. S . (1960). Transpl. B i t / / . 26, 436-441. Mistry, P. B., and Duplan, J. F. (1973). B d l . Ctrncer 60, 287-300. Nagao, K. (1977). Hiros/ii/nu J. Med. Sci. 26, 177-188. Nowinski, R. C., and Kaehler, S. L. (1974). Science 185, 869-870. Odaka, T. (1973). J. Virol. 15, 332-337. Okazaki, E., Aoki, T., and Kozlowski, E. (1976). Irrt. J . Cancer 17, 640-616. Old, L. J., Boyse, E. A., and Stockert, E. (1963). J. Nntl. Cirncrr Inst. 31, 977-986. Pazmino, N . H., and Ihle, J . N. (1976). J . Inimrrnol. 117, 620-625. Pazmino, N. H., McEwan, R. N., and Ihle, J . N. (1977).J . ln~/nuno/.119, 494-499. Peled, A. (1977). L e d . Ri;s. I , 333-343. Peled, A., and Berke, G. (1975). I s r . J . Mi~tl.Sci. I I , 1396. Peled, A., and Haran-Ghera, N. (1969). Isr. J. Med. Sci. 5, 12-13. Peled, A., and Haran-Ghera, N . (1971). Natrwe, Neiv Biol. 232, 244-245. Peled, A., and Haran-Ghera, N. (197.4). I/nni/tno/ogy 26, 323-329. Peled, A., and Haran-Ghera, N. (1978). Nulure (London) 274, 266-268. Peters, R. L., Sass, B., Stephenson, J. R., Al-Ghazzouli, I. K., Hino, S., Donahoe, R. M.,
INDUCTION OF LEUKEMIA IN MICE
87
Kende, M., Aaronson, S. A., and Kelloff, G. J. (1977). Proc. N a t l . A c ~SLY. . U.S.A. 74, 1697-1701.
Raff, M. C. (1971). Nature, New B i d . 229, 182-184. Rappaport, H . , and Baroni, C. (1962). Cancer Res. 22, 1067-1074. Reisner, Y . , Ravid, A., and Sharon, N. (1976). Biochem. Biophys. Res. Commrrn. 72, 15851591. Ricciardi-Castagnoli, P., Jadin, J. M., and Maisin, J . R. (1973). Cuircw Res. 33, 2476-2488. Rowe, W. P. (1973). "G.H.A. Clowes Memorial Lecture.'' Cancer Res. 33, 3061-3068. Rowe, W. P., and Pincus, T. (1972). J . Exp. Med. 135, 429-436. Sato, H., Boyse, E. A., Aoki, T., Iritani, C., and Old, L. J . (1973). J . Exp. M d . 138,593606. Schreffler, D. C., and David, C. S. (1975). A d v . Irnniiinol. 20, 125-195. Siegert, W., Fenyo, E. M., and Klein, G. (1977). Int. J. Cancer 20, 75-82. Silverstone, A. E., Cantor, H., Goldstein, G., and Baltimore, D. (1976). J. Exp. Med. 144, 543-548. Stockert, E., and Old, L. J. (1977). J. Exp. Med. 146, 271-276. Treves, A. J., Feldman, M., and Kaplan, H. S. (1977). J. N u t / . C m m r I m t . 58, 15271529. Van Bekkum, D. W., Knaan, S., and Zucker, C. (1977). I n "Radiation-Induced Leukemogenesis and Related Viruses" (J. F. Duplan, ed.) INSERM Symp. 4, pp. 51-63. Elsevier, Amsterdam. Waksal, S . D., Smolinsky, S., Cohen, I. R., and Feldman, M. (1976). Nature (London) 263, 512-514. Wallis, V., Davies, A. J . S., and Koller, P. C. (1966). Nature (London) 210, 500-504. Wiener, F., Ohno, S. , Spira, J., Haran-Ghera, N., and Klein, G. (1978). J. Narl. Cancer Inst. 61, 227-238. Yokoro, K., Nagoa, K., Ito, K., Kawamura, Y., and Imamura, N. (1977). I n "RadiationInduced Leukemogenesis and Related Viruses" (J. F. Duplan, ed.) INSERM Symp. 4, pp. 133-148. Elsevier, Amsterdam.
This Page Intentionally Left Blank
ADVANCES IN CANCER RESEARCH, VOL. 30
ON THE MULTIFORM RELATIONSHIPS BETWEEN THE TUMOR AND THE HOST
V. S. Shapot Cancer Research Center. USSR Academy of Medical Sciences. Moscow, USSR
I. Introduction ........................................................... 89 11. Competitive Relationships between the Tumor and the Host . . . . . A. Carbohydrate Metabolism ................................. B . Nitrogen Metabolism ....................................... C. Lipid Metabolism ........................................ 111. Effects o f the Tumor on Biological Characteristics of the Host Tissue ....... I15 ......................................... 115 B. Cytoplasmic Informational RNAs and RNP Complexes . . . . . . . . . . . . . . . . . 125 C. Disorders of Endocrine Regulation .................................... 127 D. Immunodepression .................................................. 133 IV. Prospects for the Clinic ................................ V. Conclusion ............................................................ 139 References ............................................................ 143
I. Introduction
It is well known that cancer patients often die exhibiting dysfunction of vitally important organs, and disorders of regulation of metabolism and endocrine functions. In these cases death is caused directly by infections, disturbances in the balance between the coagulative and anticoagulative systems of the blood (thrombosis, thrombophlebitis, hemorrhage), and not from metastases that hinder the vital activities of the host. All the above phenomena are often called complications that allegedly coincide with but have no direct relation to the main disease, that is, the progression of the malignant tumor, whereas actually they are a reflection and result of the pernicious action of the neoplasm on the homeostasis of the host. The highest percentage of mortality is due to various infections. For example, according to the findings of Klastersky et al. (1972), who analyzed the results of the autopsies of 157 patients who succumbed to malignant neoplasms, in about 32% death was due directly to infection caused by gram-negative microorganisms resistant to accepted antibiotics. These authors note that the infection resulting in lethality was most 89 Copyright
0 1979 by
Academic Press, Inc.
All rights of reproduction in any form reserved
ISBN 0-12-00663&0
90
V. S. SHAPOT
commonly encountered not only in patients with acute leukemia, but also in patients with cancer of the upper respiratory passages, the gastrointestinal tract, and the urogenital organs without any signs of leukopenia. According to the autopsies made over a period of four years (19711974) in the Cancer Research Center of the U.S.S.R. Academy of Medical Sciences, as many patients died of infectious complications (peritonitis, mediastinitis, pulmonary abscess, pericarditis, pneumonia, etc.) as of the progression of the main disease, staphylococci being the causative agents in 75% of the cases (Smolyanskaya and Grinenko, 1976). According to the author of the “Annals of Internal Medicine” editorial (1973), “infection and cancer are old friends.” From the data cited in the article it follows that the most common causes of death of oncologic patients are pneumonia and sepsis usually induced by nonvirulent white staphylococci and fungi. The mortality caused by different kinds of infection in oncologic patients sometimes reaches 65% (Sickless et al., 1973). The elevated sensitivity of oncologic patients to infection indicates a sharply marked imrnunodepression engendered by development of the tumor. Immunodepression in oncologic patients and animals with tumors was also repeatedly demonstrated by direct determination of the state of their cellular and humoral immunity (see, for example, Harris and Copeland, 1974). The possible causes of this surprising phenomenon deserve special attention and will be considered below (Section 111,D). Numerous cases are also encountered in which cardiovascular cornplications in patients with internal cancer are the direct causes of their death. The development of malignant tumors is often accompanied by peripheral neuropathies, myelopathies, polymyositides, and myasthenic syndromes, syndromes which sometimes precede the diagnosis of the main disease (Tyler, 1974). Twenty percent of all patients with polymyositis had malignant tumors (Pearson, 1966). In a group of dermatomyositis patients past 40 years of age, cancer was diagnosed in 50% of the cases (Niebauer, 1974). Acanthosis nigricans, a rare dermatosis, in persons past 40 years of age denotes internal cancer in 60%-100% of the cases (Barriere, 1975; Storck, 1976), acanthosis nigricans being diagnosed in most of the patients before the appearance of tumor symptoms (Dedkova and Raben, 1977). More than 150 cases of severe hypoglycemia have been reported by world literature in patients with large malignant tumors without any signs of hyperinsulinemia (see Papaioannou, 1966; Shapot, 1972). The basis for suspecting the “paraneoplastic” origin of the above pathologies is the impossibility of eliminating these syndromes by the
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
91
accepted and usually effective therapeutic measures (diagnosisex juvantibus). A proper evaluation of the above complications offers prospects for revealing the malignant tumor during the preclinical period by the different forms of its pernicious action on the host. On the other hand, their “paraneoplastic” nature is in many cases demonstrated by the disappearance of the corresponding symptoms after removal of the discovered tumor. We are still far from understanding the ways and means that connect the development of neoplasm with “paraneoplastic” phenomena. Only the first steps are being made in this direction. Elucidation of the mechanisms underlying these phenomena will apparently make it possible to enhance the resistance of the host to the tumor process by a scientifically substantiated correction of the disorders of homeostasis caused by the tumor and thereby in some measure to reduce mortality from cancer. The possibility of selectively sensitizing the tumor to the damaging factors employed in the clinic, viz., radiotherapy, chemotherapy, and hyperthermia, proved to be one of the helpful side effects of such studies. The achievements of selective “vulnerability” of the neoplasm makes it already possible to reduce the doses of these influences and safeguard against injury to the host tissues unaffected by the neoplasm (see Section IV) . In this article we are able to analyze only an insignificant part of the aforesaid “paraneoplastic” syndromes, but we deem it necessary to mention them in order to illustrate the multiformity and complexity of the effects exerted by the tumor on the host. Moving on to the main subject of the article, we wish to note that we shall deal only with the most common of all malignant tumor effects on the host’s homeostasis, omitting specific disorders caused by tumors of endocrine origin or by ectopic hormones produced by other tumors. It is possible to distinguish at least two principal and interrelated groups of manifestations of systemic effect. One of them consists in changes in the metabolism and hormonal balance of the host caused by successful competition of the neoplasm with the normal tissues for vitally important metabolites and trophic factors. The other form may be defined as influence on distant tissues manifested by decreasing differentiation, changes in enzymic characteristics, diminution of the sensitivity of target tissues to the hormones that regulate their functions, and disturbance in the negative feedback systems which coordinate the activities of central and peripheral endocrine glands. To designate “nontumor” syndromes encountered in oncologic patients, or the so-called complications of the main disease, the literature
92
V . S. SHAPOT
makes use of most diverse terms, which hampers the analysis of corresponding phenomena, the determination of their relation to the tumor growth, and their rational classification. A very popular term is “cancer intoxication.” The idea of the general poisoning of the organism by products of tumor disintegration has no reasonable substantiation. Rapid disintegration of tumor tissue is a rare phenomenon, and cancer fever is now considered the result of autoimmune pathology or secondary infection, and nephritis is considered the manifestation of hyperuricemia. No more felicitous is the term “cancer cachexia” since it has no concrete content. It does not indicate, for example, the specificity of cachexia for cancer. Moreover, the term produces an unjustified impression of an extreme degree of the oncologic patient’s exhaustion. Actually, however, emaciation at the terminal stages of tumor development is not a frequent phenomenon at all. Shabad (1936) analyzed records of 932 autopsies performed in Leningrad on persons who had died of cancer of various sites and noted extreme general emaciation as the cause of death only in 11.8% of the cases. The emaciation was associated mainly with such tumor localizations as hindered the passage of food (cancer of the gullet, stomach, intestines). At the same time in some cancer loci the patients exhibited adiposis. In the already mentioned study by Shabad (1936), 75% of the patients died not of a generalized tumor process, but of complications in the form of purulent infections. The term “paraneoplasia” would seem to be more precise, but it, too, cannot be considered entirely adequate. The prefix “para” qualifies the phenomena as concomitant to the tumor process, whereas actually they are caused by it and it is for them the etiologic factor. Moreover, the term paraneoplastic is usually applied (see the Symposium “Paraneoplastic Syndromes,” 1974) without any reason also to syndromes engendered not by all malignant neoplasms, but by only one of their specific groups, viz., tumors of endocrine origin uncontrollably producing corresponding hormones (insulinomas, pheochromocytomas, carcinomas of the adrenal cortex, etc.). We suggest that this category of phenomena should be designated as hormonal tumor syndrome, which is a narrower term than paraneoplasia. If the tumor elaborates a product that is not characteristic of the homologous tissue, for example, adrenocorticotropic or parathyroid hormones in bronchogenic cancer, or erythropoietin in kidney cancer, it would be logical to add the word ectopically to the above term, the designation now reading ectopically hormonal tumor syndrome. Of an entirely different nature is the feature common to all malignant
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
93
neoplasms, and unconnected with their elaboration of hormones, namely the ability to affect distant tissues and physiologic systems of the host, disordering its homeostasis, which in itself often leads to lethal results. In our opinion the most appropriate term for this property will be “systemic effect,” which we are now going to use. Cachexia may be recognized as a topical and not obligatory manifestation of the tumor’s systemic effect on the host. Since cachexia may also be caused by other factors, it would be expedient to call it not “cancer cachexia,” but “cachexia of the cancer patient.” There is an opposite aspect of the problem in question we are not able to touch upon in this paper. By this we mean induced alterations in the organism’s homeostasis that either predispose or hamper tumor development (“anticarcinogenesis”). This topic has been competently covered by Kavetzky (1977) in his comprehensive monograph to which the reader is referred. II. Competitive Relationships between the Tumor and the Host
A. CARBOHYDRATE METABOLISM A notion of the tumor as a trap for glucose in the host put forward some years ago (Shapot, 1968, 1970, 1972, 1975) now seems to be reasonably well substantiated. A tumor in vivo finds itself in a state of “glucose hunger” and is able additionally to consume and metabolize large quantities of glucose administered. An extremely high rate of glucose uptake by cancer cells as compared with a relatively slow glucose influx from the host underlies the above phenomenon. As a result, low, sometimes undetectable levels of glucose are being maintained in the tumor itself and in surrounding medium (Gorozhanskaya and Shapot, 1964; Gullino ej al., 1964, 1967; Nakamura and Hosoda, 1968: Shapot 1972, 1975). Recently additional data supporting this view have been obtained. Using a highly sensitive modification of the glucose oxidase method, permitting the determination of as low a glucose concentration as 0.0005 mghample, Shapot et al. (1976) could not detect even traces of glucose in highly malignant mouse Guelshtein 22a, 61, and 60 hepatomas. As to the slow growing Guelshtein 48 hepatoma, in some instances traces of glucose were found. The above regularity does not seem to be confined to transplantable tumors but holds for human malignancy as well. In all malignant tumors
94
V . S. SHAPOT
(surgically removed and thereafter immediately placed in 0.02 M NaF to prevent glycolysis) of stomach (n = 19), mammary (n = 1 3 , uterus (n = 18), kidney (n = 8), ovary (n = 3) examined, no glucose was found. In benign human blastomas, glucose levels ranged from 0.08 to 0.22 mg/g. In tissues of healthy subjects (victims of accident) [lung, stomach, kidney (n = 10, each), mammary, gland, heart, uterus, ovary (n = 3, each)], the content of glucose ranged from 0.17 to 1.24 mg/gm' (Shapot et al., 1977). It is worth mentioning here that in human malignant tumors of stomach, lung, uterus, and ovary the hexokinase, a key enzyme of glycolysis, was shown to have K , , about 10 times lower M ) than that of the homologous human normal tissues (Monakhov et d., 1978). It is likely that a high affinity of hexokinase for glucose endows the malignant neoplasm with the advantage to metabolize glucose at such negligible concentrations. Undetectable glucose levels in the tumor tissue creates an enormous glucose concentration gradient between the arterial blood and tumor80 mg/100 ml versus zero! Such a gradient favors successful competitior of the tumor with the host tissues for glucose and ensures a permanent hypoglycemic strain on the host. Severe hypoglycemia found in animals carrying ascites Ehrlich carcinoma and Brown-Pierce carcinoma was already described earlier (Blinov et a / . , 1974; Shapot and Blinov, 1974; Shapot and Blinov, 1975; Shapot, 1975). Numerous data from the literature as to profound hypoglycemia in cancer patients with large (from 900 gm to 18 kg) tumors of nonendocrine origin were also listed (Shapot, 1972). Analysis of the above observations led us to the conclusion that the hypoglycemia in question was induced by the tumor avidly consuming glucose in case compensatory mechanisms of the host failed to counterbalance this tumor activity. Some other authors are of the same opinion (Papaioannou, 1966; Carey et al., 1966; Nissau et al., 1968; Jacob et al., 1969; Chowdhury and Bleicher, 1973; Marks et al., 1974). Direct measurements attest to an extremely high rate of glucose uptake by the tumor. Utilization of glucose by a woman with intensively metastasizing leiomyosarcoma was found to be 12 mg/kg/min, i.e., 4 to 12 times as high as normal (Carey et al., 1966). However, many clinicians still maintain that hypoglycemia in cancer patients with extrapancreatic tumors is due to hyperinsulinemia or to an excess of insulinlike peptides (NSILA) in the blood, presumably pro]
The above values are of course, underestimated, since corpses lay several hours prior
to the time samples were taken, and intensive glycolysis could proceed during that time.
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
95
duced by the neoplasm. We believe that it is time to give up such a view. Sometimes hypoglycemia in tumor-bearing animals and cancer patients develops concomitantly with normal or even decreased blood insulin levels. As for animals, we have shown (Shapot and Blinov, 1975) that hypofunction of insulin-producing apparatus was provoked by a hypoglycemic status of the host (feedback effect). Rabbits carrying Brown-Pierce carcinoma characterized by a marked hypoglycemia (40 mg/100 ml) display after glucose load a flattened glycemic curve typical of latent diabetes, A similar phenomenon in cancer patients was noted by many authors. More spectacularly, functional insulin deficiency manifests itself when a repeated (double) glucose load was applied to rabbits with Brown-Pierce carcinoma whose level of immunoreactive insulin in the blood was found to be close to normal (15 t 2.0 versus 16 & 3 pU/ml in controls). The second injection of glucose to control animals, 30 minutes after the first one, was accompanied by a decline of the hyperglycemic curve, whereas in rabbits carrying Brown-Pierce carcinoma the second glucose load provoked a sharp jump of the glycemic curve. When, however, the rabbits were “saturated” with glucose by infusion 30 days running from the very moment of tumor implantation, a relative hyperglycemia could be maintained, and the second glucose load was followed by a more drastic decline of blood sugar levels than that observed with control groups. Thus, hypoglycemia in tumor-bearing animals being prevented, sufficient insulin was released to effectively eliminate an excess of the blood glucose. In mice with Ehrlich carcinoma who also developed severe hypoglycemia (50 mg/100 ml versus 90 mg/100 ml in intact mice), serum insulin levels sharply decreased (from 16 & 2.0 down to 11 f 1.5 pU/ml) as the mass of ascites cancer cells enlarged (Blinov, 1974). In rats carrying Zajdela hepatoma with only a slight hypoglycemia, the serum insulin concentrations was found somewhat diminished. Thus the above observations do not support the idea that hypoglycemia in tumor hosts is provoked by the action of an excess of insulin circulating in the blood. Marks et al. (1974) examined three patient with (1) Hodgkins disease, (2) an anaplastic large metastasizing melanoma, and (3) fibrosarcoma of enormous size. All patients mentioned suffered from severe hypoglycemia. A continued infusion of 10% glucose was necessary to prevent hypoglycemia in patient 2, the fasting serum insulin levels in all three patients being barely detectable. Two patients, however, had elevated serum NSILA levels. Two weeks after surgical removal of the tumor, the blood sugar in female patient 3 rose from 44 mg/100 ml to 96 mg/100 ml, but the serum
96
V . S . SHAPOT
NSILA dropped from 850 to 250 pUlml. However, 17 months later her serum NSILA was found elevated again with no sign of hypoglycemia. On incubation of the melanoma removed from patient 3 the tumor released neither immunoreactive insulin nor NSILA into the medium. The above authors, as we do, believe that hypoglycemia induced by the growing tumor is responsible for hypofunction of the host's insulin-producing apparatus. Why don't the majority of tumor-bearing animals and cancer patients develop hypoglycemia? It is obvious that glycogenolysis and gluconeogenesis as compensatory mechanisms must be involved in overcoming the hypoglycemic pressure exerted by the tumor on the host. There is a growing body of evidence to indicate that the liver and the muscle of tumor-bearing animals lose their glycogen as the tumor is progressing. This was the case with the liver of rats carrying Walker 256 carcinoma and of mice with Ehrlich ascites carcinoma (Granzov and Beheim, 1972). The content of liver glycogen in mice with Guelshtein 22a hepatoma and in rats with Zajdela hepatoma was found to be as low as 13% to 32% of that of controls (Mishineva et a l . , 1973). Only 5% and 10% of the original content of glycogen in rabbit liver and muscle, respectively, was left by day 20 after implantation of Brown-Pierce carcinoma, the dissemination of tumor nodes in the host being in contrast drastically increased: by day 10, 20, and 30 up to 25.1 2.4; 58 f 8.8, and 93.8 k 4.0 (in arbitary units), respectively (Shapot and Blinov, 1975). Potential capacity of mouse, rat, and rabbit liver to synthesize glycogen during tumor growth was not found to be impaired (Pattillo, 1971; Mishineva et nl., 1973; Shapot, 1975). Hence the only reason for liver glycogen depletion must be its intensive mobilization to meet the glucose requirements of both the tumor and the host. However, depletion of liver and muscle glycogen in rabbits with Brown-Pierce carcinoma did not prevent hypoglycemia attaining 38 mgl 100 ml by day 30 after tumor implantation (Shapot and Blinov, 1974). Therefore mobilization of glycogen storage may play only a minor role in maintaining normoglycemia in the tumor host. But it is possible that depletion of both liver and muscle glycogen serves as a trigger to intensify the main compensatory mechanism, namely, glyconeogenesis from noncarbohydrate compounds (see below). There are some observations that seem to support such a proposition (Marks et nl., 1974). A patient with melanoma developed severe hypoglycemia. On autopsy, his liver with no sign of metastases was rich in glycogen. Thus in the above case there was a correlation between insufficient gluconeogenesis and the loss by the liver of the capacity to mobilize glycogen.
*
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
97
However, in some other instances the above correlation did not exist. Gluconeogenesis was stimulated in rats carrying Zajdela hepatoma, although in their liver a large amount of glycogen was found (our observations). Similar findings were reported by Nisselbaum (1972). No depletion of liver glycogen occurred in rats with Morris 5123 hepatoma even after 24 hours starvation, and their blood sugar levels remained close to normal values. Probably phosphorylase of these animals could not be converted to the active “a” form, as was the case with certain hepatomas (Sato and Tsuiki, 1972). Gluconeogenesis in the tumor host had not been studied systematically and directly as a process. It was assessed sometimes in cancer patients indirectly by determining the activation of the Cori cycle (see e.g., Holroyde et d.,1975). We undertook a systematic study of gluconeogenesis from noncarbohydrate compounds in rabbits, rats, and mice with various types of tumors. Labeled amino acids 14C-tyrosine, 14C-glycine,14-glutamicacid, and 14Calanine were used as glucose precursors. We focused on glucogenic amino acids since extensive protein catabolism occurs in the tumor host favoring gluconeogenesis from these sources. Newly formed radioactive glucose was isolated either by paper chromatography or by a highly specific method-chromatography on an ionexchange column using glucose oxidase (Friedman et al., 1967). Gluconeogenesis was studied in liver and kidney of tumor-bearing and intact animals. The liver is known to maintain normoglycemia altering the rate of gluconeogenesis in correspondence with fluctuations in the composition of the body’s inner medium (Exton, 1972; Eisenstein, 1973). The kidney is also involved in this process for the inhibition of gluconeogenesis in both kidneys results in hypoglycemia within an hour (Cahill, 1970). According to Felig (1975), alanine comprises about half of all amino acids taken up by human liver and serves as the main precursor of glucose formed from products of muscle protein catabolism. A similar phenomenon was observed in animals (Felig, 1975). Preferential conversion of the carbon skeleton of alanine to glucose when compared with gluconeogenesis from other glucogenic amino acids was fully confirmed in our experiments on animals. It turned out that all animals studied fell into two distinct groups, depending on the type of tumor they had. In mice with Crocker carcinoma, Ca-755, Guelshtein solid and ascites hepatomas and in rats with Zajdela hepatoma, normoglycemia was maintained until death, and only slight hypoglycemia in Zajdela hepatoma-bearing rats was sometimes observed.
98
V . S . SHAPOT
Animals of the second group (mice with Ehrlich ascites carcinoma and rabbits with Brown- Pierce carcinoma) developed severe hypoglycemia. A direct correlation between the normoglycemia and stimulated gluconeogenesis in tumor-bearing animals was established (Shapot and Blinov, 1974; Blinov, 1974; Shapot, 1975; Blinov ef a l . , 1975). In the first group endogenous formation of glucose from labeled amino acids was found to be elevated manyfold, especially in the case of large tumors. For instance, in rats with Zajdela hepatoma by day 5 after implantation the content of the radioactive glucose in the blood, liver, and kidney increased 2-fold, in mice with Guelshtein solid hepatoma by day 20 after implantation-3.5-, 5 5 , and 6-fold, respectively. As for mice with Ehrlich ascites carcinoma characterized by growing hypoglycemia, their gluconeogenesis was markedly inhibited by day 4 after implantation (Table I). In rabbits with Brown-Pierce carcinoma, who also developed hypoglycemia, gluconeogenesis was inhibited only in the kidney cortex; in liver it was stimulated six-fold. However, the concentration of radioactive glucose in the blood was not elevated. Hence in this instance a compensatory intensification of liver gluconeogenesis was not sufficient to counterbalance the loss of glucose trapped by the tumor disseminated throughout the host tissues. The next question to be answered was whether it was possible to unmask the tendency toward hypoglycemia even in cases where normal sugar levels were maintained, as in mice (CBA x C57BL) F, with Ca-
TABLE I GLUCONEOGENESIS" I N MICEW I T H EHRLICH ASCITESCARCINOMA Tissues Mice From "C-tyrosine: Control Carrying Ehrlich carcinoma 4 days 8 days From ''C-glycine: Control Carrying Ehrlich carcinoma 4 days 8 days
Liver
Blood
Kidney 130
1,350 t
133
1,070 f
290 ? 1,320 -+
32 200
1,150-+ 135 1,330 ? 205
15,600 -+ 1.600 7,680 t 640 8,540 -+ 1,020
22,140
730 t
86
1,230 f
270
2,210
12,550 t 2,920
14,820 f 1,600 16,010 t 2,600
10,400 -+ 1.260 10,050 t 990
f
~~
"
Gluconeogenesis expressed as 14C-glucose(cprn/ml or cpm/mg tissue) formed
TUMOR AND HOST MULTIFORM RELATIONSHIPS
99
755. Here we made use of the approach that included a continuous "saturation" of the host with glucose from the very moment of tumor implantation and subsequent cessation of glucose injection after a certain period. Both intact and tumor-bearing (CBA x C57BL) F, mice were injected with 10 mg/gm glucose for 10 days, and then the extra glucose supply was discontinued and the animals were left on the standard diet. Intact and tumor-bearing mice of the same line, which did not receive extra glucose, were under observation as well. As could be expected, the rate of liver and especially kidney gluconeogenesis declined as a result of glucose injections in both control and tumor-bearing animals. However, the difference between them was revealed after the extra glucose supply ceased. In healthy mice the rate of glucose formation from 14C-tyrosine increased somewhat, but the initial level was not obtained within the next 10 days. In contrast, gluconeogenesis in tumor-bearing mice under the same conditions was sharply accelerated, especially in the kidney cortex, and highly exceeded the normal level. As to the tumor-bearing mice that did not receive glucose, gluconeogenesis in their kidney increased continuously after implantation (Fig. 1). These data support the idea that normoglycemia in mice carrying Ca-755 is maintained owing to elevated gluconeogenesis. In spite of two-, four-, and sevenfold stimulation of liver and kidney gluconeogenesis, the concentration of 14C-glucose in the blood is in-
L
I
I
I
I
I
U
5
10
15
20
Days from LmpLantation FIG. 1 . Gluconeogenesis in the adrenal cortex of control (1, 3) and tumor-bearing (2, 4) mice. 3 and 4 were injected with glucose (10 mg/gm weight) twice a day for the first 10 days.
100
V. S. SHAPOT
creased no more than by 50%, this fact presumably reflecting an intensive utilization of glucose by the tumor. In the other experiments the dynamics of glycogen accumulation in liver as well as the rate of its synthesis from ''C-tyrosine in this organ were studied. From Fig. 2 one sees depletion of liver glycogen in nonstarving mice by day 20 after tumor implantation; its content decreased more than 90%-fold. As a result of a continuous injection of glucose, liver glycogen in both intact and tumor-bearing mice increased manyfold, but after extra glucose supply ceased the content of liver glycogen in control mice gradually declined, never reaching the initial level; whereas in tumor-bearing animals depletion of liver glycogen proceeds very rapidly to the level characteristic of tumor-bearing mice that had no glucose injections. The most plausible explanation for the above observations is an intensive mobilization of liver glycogen to counterbalance the tendency toward hypoglycemia induced by the tumor. Figure 3 shows a stimulation of the synthesis of I4C-liver glycogen in tumor-bearing mice under usual conditions. Continuous injections of glucose inhibited liver gluconeogenesis, but cessation of extra glucose supply from day 11 onward was followed by a sharp stimulation of this process in the case of tumor-bearing mice only, whereas in intact animals the specific radioactivity of the newly formed glycogen from 14C-tyrosine increased slowly and within 10 days remained lower than the initial values. Thus the results described reveal a direct correlation between the
Oags from
LmpLantution
FIG.2. Accumulation and depletion of liver glycogen in control (1, 3) and tumor-bearing (2, 4, 5) mice. 3 and 4 were injected with glucose twice a day for the first 10 days, 5 , for the last 5 days.
TUMOR AND HOST MULTIFORM RELATIONSHIPS
101
Days from impLontation FIG. 3. '*C-glycogen formation from U-'4C-tyrosine in the liver of control (1, 3) and tumor-bearing (3, 4, 5 ) mice. 3 and 4 were injected with glucose twice a day for the first 10 days: 5 , for the last 5 days.
depletion of host liver glycogen and a drastic acceleration of liver glyconeogenesis induced by the avid uptake of glucose by the tumor (Blinov and Shapot, 1974b; Shapot, 1976). What are the factors responsible for the stimulation of gluconeogenesis in the host? Gluconeogenesis is under multihormonal control. Insulin acts as an inhibitor; glucagon, epinephrine, growth hormone, and glucocorticords accelerate endogenous formation of glucose. We do not know of any data indicating elevated levels of the blood glucagon. The same holds for epinephrine, although the possibility of its hypersecretion is consistent with the phenomenon of a stimulated lipolysis in adipose tissues and muscle described for the tumor's host. As for glucocorticoids, there are many observations as to their elevated levels both in cancer patients (Saez, 1971, 1974), and tumor-bearing animals (Samundjan, 1973). Glucocorticoids induce hyperglycemia in rats (Herrmann and Staib, 1969), cows (Heitzman et a / . , 1971), and man (Kelly, 1959; Egorova, 1965) and raise the activity of key enzymes of gluconeogenesis (Weber, 1968; Knox and Sharma, 1968; Mertvetzov, 1969). We believe that hyperfunction of the adrenal cortex may play an important role in the intensification of glyconeogenesis as a response to hypoglycemic strain of the tumor on the host. In our experiments one injection of cortisol into intact mice starved for 24 hours induced a gradual elevation of the blood sugar especially pronounced by the 24th hour. A similar phenomenon has been noted
102
V. S . SHAPOT
with respect to liver glycogen. Its content increased sevenfold within 24 hours after cortisol injection. Formation of glycogen from labeled amino acids was also accelerated five- to sixfold (Blinov et al., 1975). Here we would like to draw attention to the fact that cortisol injected in the same doses to mice with Ehrlich carcinoma, which developed severe hypoglycemia, failed to stimulate gluconeogenesis, unlike its normal effect exerted on mice with Crocker sarcoma, which were able to maintain normoglycemia. Daily injection of large doses of cortisol (5 mg) for 8 days were needed to raise the blood sugar level in mice with Ehrlich carcinoma, this observation indicating an elevated threshold of sensitivity to the hormone. It is appropriate to mention a report on two cases by Nissau er a l . ( 1968). Cancer patients with large tumors, reticulosarcoma and lymphosarcoma, developed pronounced hypoglycemia. Injection of glucocorticoids or epinephrine failed to raise their blood sugar levels. The authors suspected impairment of gluconeogenesis in both patients which was not able to counterbalance an excessive uptake of glucose by the tumor (studied in vitro after its surgical removal). All findings described above support our view that normoglycemia in the tumor’s host is ensured by stimulated gluconeogenesis from noncarbohydrate sources, its impairment resulting in hypoglycemia.
METABOLISM B. NITROGEN The tumor is a trap for nitrogen, a concise and expressive term coined by Mider (1951, 1953) as a result of his excellent studies. However, Mischenko (1940) seemed to be the first to discover the capacity of the tumor to take up nitrogen of the host’s tissue proteins. The experiments were performed on transplantable rat and chicken tumors, but his observations applied to cancer patients as well. Mischenko inferred that the tumor is growing at the expense of assimilation of products of host’s muscle protein breakdown. He wrote that the tumor is able to consume both exogenous and endogenous nitrogen compounds and ensures its endogenous nutrition inducing an “intertissue exchange.” Mischenko’s experiments demonstrated that even during starvation the host supplies the tumor with building blocks and that cachexia may be caused by an intensified hydrolysis of host tissues. Mider’s and Mischenko’s idea was confirmed by LePage et al. (1952). Starving rats carrying a rapidly growing Flexner carcinoma lost within 5 days 31% of their weight and 3% of liver protein. In spite of this the weight of the carcinoma markedly increased; its growth rate being only
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
103
half as high as in similar animals kept on a protein-rich diet. Rat proteins and nucleic acids, prelabeled with l4C-glycine, after tumor implantation lost their label very rapidly, whereas the specific radioactivity of tumor proteins and nucleic acids sharply increased. It is quite obvious that the tumor somehow enhances protein catabolism of the host tissues and may at the same time hinder protein synthesis in them. The data of Lundholm (1975) and Lundholm et al. (1976) support this view. In 43 patients with various sites of tumor (esophageal, gastric, hepatic, pancreatic, colon, rectal, renal carcinomas), the above authors found a significant elevation of the cathepsin D activity in muscle lysosomes, enhancement of muscle protein breakdown, and a slower than normal protein synthesis as revealed by measuring incorporation of labeled leucine performed in vitro (biopsy). It was possible to stimulate the incorporation rate of leucine into soluble protein in cancer patient’s muscle fibers by increasing the normal human plasma amino acid level 10-fold in the incubation medium. The degree of stimulation was equal to that observed in control specimens. According to Lundholm (1973, the above effect distinguished cancer disease from malnutrition. In Blinov’s experiments (1974) in rabbits with Brown-Pierce carcinoma the total nitrogen content sharply decreased in the heart by day 10, and in skeletal muscle by day 20 after implantation. The host finds itself under subcaloric conditions, therefore preferential breakdown of muscle proteins to meet the requirements of visceral organs and the tumor might be expected. Numerous nitrogen balance studies both with tumor-bearing animals and cancer patients failed to unambiguously answer the question. Hence warranted doubts were expressed as to the validity of such an approach (Costa, 1977; Blackburn et a l . , 1977). Indeed, in our experiments on rabbits with Brown-Pierce carcinoma a pronounced negative nitrogen balance within 13 to 18 days after implantation was observed. In the terminal stage quantities of nitrogen excreted within 2 days were as high as about 1.5 gm. In this case the dissemination of the tumor throughout the host tissue attained such a degree that it was not able to trap all the nitrogen released by the host. In contrast, in rat 45 sarcoma, which grew locally, during the whole period after implantation, an apparent positive nitrogen balance was found to be maintained in the host. The weight of the animal carcass (less 15 to 49 gm neoplasm) was reduced on the average by 8 gm. Hence exogenous, dietary nitrogen as well as endogenous nitrogen seemed to be consumed by the growing tumor (Shapot and Blinov, 1975). Similar observations on cancer patients were reported by Blackburn et af. (1977).
104
V . S. SHAPOT
The most physiologically significant tissue losses were from the viscera, indicated by the low albumin, transferrin, and total lymphocyte count accompanied by the marked impairment of cell-mediated immunity. The reason for stimulated catabolism of nitrogen biopolymers in the tumor host remains unclear. One may assume that both starvation and certain hormonal imbalance, in part hypersecretion of glucocorticoids, are involved. Hypertrophy of the adrenal cortex of the tumor host with the eventual dystrophy of the gland by the terminal stage of tumor growth drew the attention of many authors long ago (Begg, 1958). A hypersecretion of glucocorticoids in cancer patients (Saez, 1971; 1974) and tumor-bearing animals (Samundjan, 1973) was also reported. According to our observations made with Bunatyan (Shapot and Blinov, 1975), the concentration of the total 1 I-oxycorticosteroids as well as of biologically active corticosterone in the blood of rabbits with Brown-Pierce carcinoma by days 20 and 30 after implantation was significantly elevated. There are reasons to believe that hypersecretion of glucocorticoids may cause stimulation of protein breakdown, especially in the tumor’s host. Blinov (1974) has shown that cortisol administered to rabbits with Brown-Pierce carcinoma additionally stimulated catabolism of nitrogen polymers. The content of the total nitrogen in urine increased 16%, NH, 10096, and creatin 110960 as compared with tumor-bearing rabbits that received no hormone. Moreover, cortisol exerted on tumor-bearing rabbits a more pronounced effect than on healthy ones. Excretion of NH3 and creatine by the former was three times higher than that by the latter. Jewel1 and Hunter (1971) reported that adrenalectomy prevented the stimulation of albumin catabolism in rats with Walker 256 carcinoma. Thus there is an apparent similarity in the ability of glucocorticoids and the tumor to enhance protein catabolism in the body. The following question arises: In what way does the tumor induce a hypersecretion of glucocorticoids in the host? We have already described the role played by stimulated gluconeogenesis in counterbalancing hypoglycemic pressure exerted by the tumor. Glucocorticoids are known to enhance gluconeogenesis, and it is likely that their hyperproduction in the host is just a response to the tendency toward hypoglycemia induced by the tumor. If true, “saturation” of the host with glucose would reduce catabolic processes. Indeed, administration of glucose was shown to reduce the losses of nitrogen by surgical noncancer patients (Abbott et al., 1959; Holden et a l . , 1957) and starving animals (Strautmans and Schmidt, 1966). A highcarbohydrate diet tended to normalize previously elevated levels of free
TUMOR AND HOST MULTIFORM RELATIONSHIPS
105
amino acids in the host’s blood to the same degree as did removal of the tumor (Mustea, 1971). Food enriched in carbohydrates prevented the loss of nitrogen in the brain of mice carrying MFS fibrosarcoma (Mallick et al., 1968). Blinov (1974) (see also Shapot and Blinov, 1975) administered 2 gm/kg 40% glucose daily to rabbits with Brown-Pierce carcinoma. As a result their blood total nitrogen, urea, glutamine, nitrogen of free amino acids, creatine, and creatinine sharply decreased as compared with carcinomabearing rabbits that did not receive glucose. The content of the total nitrogen in the spleen and myocardium increased as did creatine in the latter tissue; the level of creatinine was reduced. Thus, a prolonged “saturation” of the host with glucose obviously alleviates catabolic effects of the tumor. In this connection it is worth mentioning again that excess of administered glucose inhibits gluconeogenesis (see Section 11,A). Factors that favor the uptake of amino acids as products of the host’s tissue protein catabolism by the tumor remain unknown. Some speculations, however, may be made concerning glutamine. Glutamine is known to serve as one of the most important precursors of the synthesis of tumor protein and purine nucleotides. The content of glutamine in tumor cells, however, was found to be negligible (Roberts et al., 1971) owing to its extremely rapid assimilation and a low capacity of the tumor to synthesize glutamine (Schreck et al., 1973). In the host tissues, glutamine levels gradually decreased and sometimes, as in the case of rapidly growing Walker 256 carcinoma and Novikoff hepatoma, dropped to barely detectable values (Wu and Morris, 1970). Similar results were obtained (Shapot and Blinov, 1975) with rabbits with Brown-Pierce carcinoma. From day 10 after implantation onward the content of glutamine in rabbit liver, heart, and skeletal muscle (but not in kidneys) decreased, while that in the blood increased. The latter fact suggested the release of glutamine from host tissues. Proceeding from the observations described above, one may propose that a preferential transfer of blood glutamine to the tumor is conditioned by a high gradient between its concentration in the arterial blood and that in the tumor, as it occurs with glucose (see Section 11,A). Dystrophy of vital organs may to a certain degree be conditioned not only by enhanced catabolism of their proteins, but by a reduced rate of protein synthesis also. A fraction of glucogenic amino acids would inevitably be channeled to the carbohydrate metabolic pathway when gluconeogenesis was stimulated in the tumor host. Thereby this fraction would not be available any longer for the participation in polypeptide chain formation.
106
V. S . SHAPOT
Another reason for a hindered protein synthesis in the host may be impairment or rearrangement of the protein-synthesizing apparatus. Clark and Goodlad (1975) have described a certain deficiency in ribosomes of skeletal muscle as a result of the pernicious effect of Walker 256 carcinoma growing in rats. It is widely known that both in cancer patients and tumor-bearing animals dystrophy of skeletal muscle occurs very often. Clark and Goodlad (1975) studied ribosomes of gastrocnemius very thoroughly. Their protein-synthesizing capacity was only half as high as that of corresponding muscle of intact rats. Hybrid ribosomes were prepared composed of the large ribosomal subunit of gastrocnemius from the tumor host and the small unit from the normal muscle and vice versa. The defect was found to be localized in the small subunit after the initiation of polypeptide chain synthesis. There are other data indicating a rearrangement of the protein-synthesizing apparatus in the liver of the tumor host, for example, a shift in the proportion of membrane-bound and free ribosomes in favor of the latter. In normal rat liver, according to our observations, the membranebound to free ribosomes ratio is 3.0 5 0.25. This ratio is a reflection of a high specialization of hepatocytes since membrane-bound ribosomes, unlike free, synthesize preferentially extracellular proteins subsequently transferred to the circulation. According to Yap et al. (1977), 98% of mRNA coding for serum albumin is associated with membrane-bound polyribosomes. In hepatomas the opposite picture can be seen: at the early stages of growth the relative proportion of membrane-bound ribosomes is reduced and the membrane-bound to free ribosomes ratio eventually drops to 0.5 5 0.14. From day 3 after implantation of Zajdela hepatoma onward the relative proportion of membrane-bound ribosomes in rat liver starts reducing and by day 6 the above coefficient reaches the value of 0.63 -t 0.26, very close to that characteristic of the tumor itself. Special measures are taken to collect all cytoplasmic membrane-bound ribosomes including those sedimenting with mitochondria and nuclei (Pushkina et al., 1976). One of the possible interpretations of the phenomenon described would be that it is a defense reaction of the host against the action of the tumor as a trap for nitrogen-the protein-synthesizing apparatus is reformed to ensure the synthesis of intracellular, structural proteins. Indeed, the pool of free ribosomes was found to be enlarged on malnutrition (Ekren and Vatrin, 1972). We made attempts to verify the above observation and let intact rats starve for 48 hours with no limitation in water supply. The proportion of free ribosomes in their liver was reduced from 3 2 0.25 to 0.78-1.3.
TUMOR AND HOST MULTIFORM RELATIONSHIPS
107
However, forced nutrition of rats with Zajdela hepatoma within the 2 last days when anorexia sets on (infusion of glucose, interlipid and protein hydrolyzate) prevented the reduction of the membrane-bound to free ribosomes ratio only partially; it did not rise higher than 1. Hence, the effect of the tumor on the host liver ribosomes cannot be explained only by starvation. As for rat regenerating liver (48 hours after partial hepatectomy) the opposite picture could be seen-the proportion of membrane-bound ribosomes rose up to 83%-85% of total ribosome pool. In this connection one more example of the functional reformation of protein-synthesizing machinery in the host liver can be mentioned (Shapot and Berdinskich, 1975; Shapot, 1975). It is well known that tumor growth both in humans and animals is often accompanied with hypoalbunemia. The reason for this may be at least twofold, a result of stimulated albumin catabolism (Jewel1 and Hunter, 1971) or hampered synthesis of this protein, or both. Specific immunoprecipitation of polyribosomes with antibodies against serum albumin followed by ultracentrifugation of the precipitate on sucrose concentration gradient allowed us to demonstrate in the case of animals with Schwetz leukemia and Zajdela hepatoma that in their livers the fraction of polyribosomes involved in the formation of albumin polypeptide chains was sharply reduced (Fig. 4), the total content of polyribosomes remaining unchanged. The Zajdela hepatoma completely lost the capacity to synthesize serum albumin. The phenomena described above are regarded as manifestations of dysdifferentiation of the host’s vital organs which synthesize proteins
959
700.0
FIG.4. Proportion (in percent of total) of rat liver and hepatoma polyribosomes synthesizing serum albumin: 1, control: 2, animals with Schwetz leukemia: with Zajdela hepatoma: 3, 4, Zajdela hepatoma cells.
I08
V . S . SHAPOT
needed for the whole body at a reduced rate. Both in trauma and tumor growth, the synthesis of secreted proteins such as transferrin, lipoproteins, and serum albumin decreased (Blackburn et d., 1977). Our knowledge of nucleic acid metabolism in the tumor host is scarce. Here we report some recent data on the successful competition of the tumor with the host lymphoid tissues for the precursors of nucleic acid, pyrimidine nucleotides. The common pyrimidine precursor of RNA-uridine monophosphate (UMP)-is synthesized through two different pathways: (a) de n o w , from COP, NH,, and aspartic acid, with the formation of an intermediate compound, orotic acid; and (b) a “salvage” route, including the reaction of uracil with ribosylpyrophosphate, leading to uridine. Proportions of these two pathways vary depending on the type of tissue, but the “salvage” one prevails in rapidly growing malignant tumors. Comparative study of the proportions of the above pathways as a reflection of the effect of the tumor on the host proved very informative. According to our previous observations with Vornovitskaya (see Shapot, 1975), rat liver utilizes mainly orotic acid as a precursor of RNA pyrimidine nucleotides, the incorporation of uridine into RNA being 22-fold lower. In regenerating liver this difference in favor of orotic acid rises to about 60. As for rat Zajdela hepatoma, the opposite picture is seen. Uridine serves as a predominant precursor of RNA pyrimidine nucleotides, its incorporation being 10 times that of orotic acid. The most spectacular results were obtained (Vornovitskaya et id., 1979) with C3HA mice carrying rapidly growing Guelshtein 22a hepatoma, which like Zajdela hepatoma utilizes mainly uridine. In mouse liver, unlike rat liver, both orotic acid and uridine were involved in RNA synthesis equally. In contrast, in the mouse spleen and thymus, the preferential precursor of RNA pyrimidine nucleotides turned out to be uridine. The incorporation ratio of 14C-orotic acid to 14C-uridine into RNA in the above lymphoid organs was 0.21-0.22, whereas in 22a hepatoma it was only 0.05. We have studied the changes in this ratio in the process of tumor growth (Fig. 5). In the liver and spleen orotic acid incorporation remained practically unchanged within day 1 to day 8 after implantation, in thymus being even somewhat higher: whereas the incorporation of uridine sharply decreased, particularly in the spleen. It dropped after 24 hours after implantation to as low as 13%, and on the fifth and eighth days to 4% of the initial value (Fig. 6). Hence the tumor intercepts for its own needs the precursor that is required first, omitting orotic acid which is utilized poorly. It is obvious that such a pumping of uridine would affect lymphoid organs first, while the liver can effectively utilize orotic acid.
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
0
2
4
109
fi
6
nuys from LmpLantation
FIG.5. Incorporation of 14C-oroticacid ( 1 ) and 14C-uridine(2) into liver RNA of hepatomabearing mice.
The phenomenon of successful competition of the tumor with the host for thymidine as a precursor of DNA was demonstrated by us on C3HA mice with Guelshtein hepatornas as well (Gerstein et a l . , 1978). In the acid-soluble fraction of 22a hepatoma, on the fifth day after implantation, as early as 5 minutes after 14C-thymidineadministration, the label was found to be very high and then gradually declined within 55 minutes, obviously being utilized in DNA synthesis since the label in DNA rose in parallel (Fig. 7). An intensive incorporation of the labeled exogenous
7
,
0
1
1
2
1
1
4
1
1
6
1
1
8
Daus from ImpLantation
FIG.6. Incorporation of 14C-oroticacid (1) and I4C-uridine in spleen and thymus RNA of hepatoma-bearing mice.
110
V. S . SHAPOT
FIG. 7. Incorporation of I4C-thymidine into mouse hepatoma DNA ( I , 2, 3) and acidsoluble fraction ( I ' , 2', 3'): 1 and 1', 22a hepatoma on the fifth day of the implantation: 2 and 2', 22a hepatoma on the eighth day, 3 and 3', 48 hepatoma (1-1.5 month growth).
thymidine into DNA of rapidly growing Morris hepatomas was described earlier by the Weber group (Ferdinandis et al., 1971). On the eighth day after implantation close to the terminal period, the initial ( 5 minuntes) level of the 14C-thymidinelabel in the acid-soluble fraction was markedly reduced as well as the rate of its incorporation into DNA. In slow growing Guelshtein 48 hepatoma both of the above indices were low. The kinetics of the incorporation of 14C-thymidineinto DNA and the I4C-label in the acid-soluble fraction of the spleen of intact and 22a hepatoma-bearing mice was also studied. On the second day from implantation a reduced radioactivity, calculated on the basis of mg DNA, could be observed within 5 minutes. On the fifth day, the period of rapid tumor growth, both the initial value of the label in the acid-soluble fraction and the rate of I4C-thymidine incorporation into DNA were decreased about 60% (Fig. 8). It is likely that rapidly growing 22a hepatoma intercepts exogenous thymidine for its own DNA synthesis, interfering with the influx of thymidine to even rapidly proliferating host tissues.
111
TUMOR AND HOST MULTIFORM RELATIONSHIPS
Slow growing 48 hepatoma did not alter significantly thymidine incorporation into DNA and its influx into the acid-soluble fraction in any of the host tissues studied. The same holds for rat Zajdela hepatoma. Suppression of DNA synthesis in the spleen of 22a hepatoma-bearing mice seems to be a result not just of a competition of the tumor for the precursor, but of other factors as well. We noted certain alterations in the distribution of the 14C-labelin the spleen acid-soluble fraction between thymidine (along with products of its catabolism), TMP, and TDP TTP (Table 11). From the second day onward a specific distribution of the label in the spleen was shifted in favor of nonphosphorylated compounds; the radioactivity of TMP being the most reduced, particularly low on the fifth and eighth days after implantation (Gerstein el al., 1978). The above alterations may be caused by a hindered thymidine phosphorylation and/ or the stimulation of its catabolism. Indeed, the activity of thymidine kinase from spleen of 22a hepatoma-bearing mice was found to be sharply reduced-from 1.32 f 0.37 to 0.074 f 0.03 nmoles of TMP formed/mg protein. Suppression of DNA synthesis in the spleen of mice with 22a hepatoma correlates with a two- to threefold reduction of the organ’s size. In the thymus and liver of mice with 22a hepatoma, unlike spleen of the same
+
FIG.8. Incorporation of 14C-thymidineinto the spleen DNA (A) and acid-soluble fraction (B) of tumor-bearing mice: 1, control: 2, on the second day after implantation of 22a hepatoma; 3, on the fifth day: 4, on the eighth day: 5, 48-hepatoma-bearing mice.
I12
V . S. SHAPOT
TABLE I1 DISTRIBUTION OF THE RADIOACTIVITY I N T H E ACID-SOLUBLE FRACTION I N PERCENT ~~
14C-thymidine label
Tissue 22a hepatoma 48 hepatoma Liver
Spleen Spleen of 22a hepatoma's host Spleen of 48 hepatoma's host
Term after tumor implantation
Thymidine + products of its catabolism
TMP
TTP
5 days 8 days 2 months Control
30.3 t 44.0 t 75.0 69.0 t
*
3.2 3.4 2.7 1.5
3.8 2 3 . 3 6.6 t 0.9 20.0 t 2.6 19.0 t 1.3
65.9 t 5.7 49.4 t 2.8 5.0 t 0.4 12.0 t 0.7
Control 2 days 5 days 8 days 2 months
64.0 t 71.0 t 81.5 t 80.0 t 60.0 t
5.5 5.8 4.5 3.5 10.0
25.0 15.5 5.5 5.3 27.0
t 3.5 -c 3.9 t 2.5 t 2.4 k 7.3
11.0 t 2.0 13.5 c 3.2 13.0 c 2.0 14.7 t 3.4 13.0 5 3.2
animal, no changes in the incorporation of thymidine into DNA and distribution of the label in the acid-soluble fraction occurred. Other authors, however, reported a significant stimulation of DNA synthesis in spleen of mice and rats with tumors of various loci (Morgan and Cameron, 1973). One can infer that reactions of the host to the growing tumor are multiform and nonstereotypic depending on the type of neoplasm, its growth rate, and the animal species.
C. LIPIDMETABOLISM As we have demonstrated above, the functioning of a malignant neoplasm as a glucose trap leaves in certain measure an imprint also on the state of other types of metabolism, nitrogen metabolism in particular. We shall now consider the alterations in lipid metabolism that occur in the host in the process of tumor growth. Hyperlipidemia and the depletion of tissue lipids have long been noted in animals with tumors (Haven and Bloor 1956): moreover, hyperlipidemia begins to show particularly clearly at the moment the mass of tumor (Walker, 256) equals 1% of the host weight, diminishing, however, during the terminal period, when the mass of the neoplasm is doubled. At the same time Costa (1973) noted that Krebs mouse carcinoma-2 induces a 50% diminution of the lipid content of the host tissues as early as 7 days after transplantation, still not having grown large. There is a statistically reliable (607 meq/liter versus 358 meq/liter) or
TUMOR AND HOST MULTIFORM RELATIONSHIPS
113
an almost twofold increase in concentration of unsaturated free fatty acids in the plasma of 40 patients with malignant neoplasms of various loci (Mays, 1971). However, the author does not consider this phenomenon specific for cancer because he observed still higher hyperlipemia in noncancer patients, who suffered from other chronic diseases that led to a catastrophic loss of their weight. Lipidemia, owing to the elevated level of trigliceride-rich, very low density lipoproteins (CVLDL) in tumor-bearing animals and cancer pa1977) to be a result tients, was reported (for references see Kralovic et d., of the depletion of fat deposits. The above authors demonstrated that growth of rat Walker-256 carcinoma is accompanied by mobilization of free fatty acids, which precedes or parallels the depletion of carcass neutral lipid at the early stages of tumor development when its mass is yet small. Kralovic et al. (1977) noted an increase in the size of the adrenal gland of the host at this time, suggesting some stress. From our point of view these phenomena may be a result of hypersecretion of glucocorticoids enhancing gluconeogenesis in the host. We shall now dwell on studies (Lankin, 1971, 1973 a,b; Lankin and Neyfakh, 1973; Polyakov ef al., 1977; Lankin et al., 1976, 1977) in which the changes in the lipid content of host tissues were investigated in four models of mouse tumors, viz., ascites form of Ehrlich mouse carcinoma, sarcoma 37, sarcoma 180, and Walker carcinoma 256. A rapid depletion of the fat deposits was observed, with a steady decrease of total lipids in the omentum and skeletal muscles up to the terminal period when, in Ehrlich carcinoma, it fell to 11% and 4% of its initial value, respectively. N o changes were recorded in the brain and kidney lipids, while in the liver and blood plasma the increases proved to be extreme, reaching their maximum on the fifth day mainly through enrichment of the liver with triglycerides, and of blood plasma on the seventh day after transplantation. The liver of animals carrying various tumors displayed highly elevated levels of fatty acids as constituent of total lipids (Table III), by day 7 or 9 particularly these alterations were extreme. An enormous increase in the relative content of higher unsaturated fatty acids, oleic and linoleic, could be observed in the blood plasma during the period of intensive tumor growth, i.e., between the fifth and eighth days, and a progressive diminution of these acids in the omentum. The depletion of the fat deposits and increase in the transport form of lipids (triglycerides) in the plasma and liver, as well as in free unsaturated fatty acids, attest to intensive mobilization of lipids caused by the tumor growing in the host. The tumor apparently utilizes for its needs mainly free fatty acids
114
V. S. SHAPOT
TABLE 111 CONTENT OF FATTYACIDS IN TUMORHOSTLIPIDS“ Fatty acids in percent Liver of
16:O
Control animals Mice with Ehrlich ascites carcinoma (the seventh day) Mice with 37 ascites sarcoma (the seventh day) Mice with Crocker ascites sarcoma (the seventh day) Rats with Walker 256 carcinoma (the ninth day)
18:l
18:2
100
I00
141
322
I00 138
I63
424
I62
20Y
233
I33
I67
294
187
Lankin (1973a).
(Spector, 1967) supplied by the host in complex with serum albumin or as a constituent of VLDL. At the same time the tumor itself does not exhibit any appreciable accumulation of lipids: nor is there any ketonemia in the host. This suggested to the researchers that the tumor itself used the mobilized lipids as the main source of energy, oxidizing them to CO, and H,O (Mays, 1971; Lankin, 1973a,b; Spector, 1975). We, however, consider such interpretation unlikely. The ability of the tumor to oxidize lipids was in its time demonstrated in experiments in vitru with an excess of oxygen in the medium. As a matter of fact, in vivo the tumor is unable vigorously to oxidize large amounts of lipids because it is under conditions of progressing hypoxia (Vaupel, 1974; Tannok, 1976; Shapot, 1976). The main source of energy for it is anaerobic glycolysis (Shapot, 1975). There are reasons to assume that the mobilization of lipids, hyperlipidemia, is a result of hypoglycemic “pressure” exerted by the tumor on the host. The mobilized lipids are assimilated by the tumor (as precursors of its membrane lipids), but mainly by the host tissues which, owing to a deficiency of glucose intercepted by the tumor, switch over in large measure to oxidation of fatty acids: especially since hyperlipidemia is of itself capable of limiting utilization of glucose by muscular tissue in conformity with Randle cycle but inducing at the same time the synthesis of enzymes involved in gluconeogenesis (Weber et al., 1966). Hyperproduction of glucocorticoids, which is characteristic of the tumor-affected organism, acts in the same direction, i.e., favoring utilization of fatty acids (FA) as a source of energy and reducing assimilation of glucose by muscle tissue. The concentration of unsaturated FA in the blood,
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
115
mobilized from the deposited fat and other tissues, depends on the level of glucose in the blood. Hyperglycemia induced by administration of glucose into the organism completely blocks the mobilization of nonesterified unsaturated FA (Gordon et al., 1957; Dole, 1956, see also Waterhouse er a f . , 1969). A recent study on mice with Ehrlich ascites carcinoma (Baker et a f . , 1978) confirms the above observations. Tumor-bearing animals had elevated FA fasting levels and on being refed glucose-rich food displayed inhibition of FA mobilization. Even in the case where the fasting plasma-free FA concentration in patients with cancer was normal, induction of high hyperglycemia was required to suppress plasma FA levels (Edmondson, 1966). On the other hand, any influence that hampers metabolism of glucose, for example, administration of 2-deoxyglucose (Laszlo et a f . , 1960), favors mobilization of FA. Thus hyperlipidemia and depletion of the fat deposits observed in the tumor-affected organism, although nonspecific of cancer, since these phenomena may be induced by other factors, including starvation, are conditioned by the parasitic character of tumor growth. Mobilization of lipids and elevated amounts of higher polyunsaturated fatty acids circulated in the tumor host may entail an increase in the risk of blood coagulation (Gjesdal, 1976). Prostacyclin (PGI,) is known to prevent platelet aggregation. Accumulation of lipid peroxydes, e.g., 15-hydroperoxide of arachidonic acid, a potent inhibitor of PG1,-synthesizing system (Moncada et al., 1976) of endothelium and other tissues (Dembinska-Kiec et al., 1977), would favor blood coagulation through enhancement of the formation of thromboxane (TXA,) since both PGI, and TXA, derive from the common precursorcyclic peroxide of PGGz (Prostglandins, 1977). TXAz, synthesized in platelets, lung, and stomach, causes aggregation of platelets, vasoconstriction (Zmuda et a f . , 1977), and eventually thrombosis. Thus, impairment of lipid metabolism in the host may provoke imbalance in the blood coagulative-anticoagulative system.
Ill. Effect of the Tumor on Biological Characteristics of the Host Tissue
A. ENZYMES Since Greenstein’s classic experiments (1947) with spontaneous and transplantable mouse and rat tumors, it has been known that the activity of catalase in the host liver is suppressed and that it becomes normal
1 I6
V . S. SHAPOT
again after surgical removal of the tumor. Later it was revealed that the tumor inhibits the synthesis of this enzyme and not its activity. According to Kushiwagi et al. (1972), the liver of the mouse with ascites hepatoma AH-49 H exhibits a sharp decrease in incorporation of labeled leucine into the polypeptide catalase chains that form on polyribosomes precipitated by antibodies against the pure enzyme preparation. At one time Shapot et ul. (1963) found, besides water-soluble catalase, another form of this enzyme firmly associated with the lipoprotein of liver cells. It is extracted by n-butanol after preliminary removal of the soluble form and accounts for more than two-thirds of the total activity of the liver catalase. In the experiments with highly malignant Guelshtein ascites hepatoma 22a transplanted to C3HA mice we showed that the activity of only the water-soluble catalase diminishes in the host liver, while that of the lipid-bound enzyme does not change. The ratio of the second form of catalase to the first therefore increased from 2.5 to 4 (Davidova et at., 1970). Interesting phenomena pertaining to the changes in the activity of 23 liver enzymes in animals with slowly and rapidly growing Morris hepatomas, as well as Walker 256 carcinoma, were noted by Herzfeld and Greengard (1972). The growth of Walker carcinoma in 10 to 14-day-old sucklings prevented the appearance in their livers of the enzymes corresponding to the given stage of differentiation, for example, ornithine aminotransferase, glucokinase, glutamine synthetase, and malate-NADPdehydrogenase. At the same time, premature induction of ornithine aminotransferase by glucocorticoids was observed in the liver of these animals. After transplantation of tumors to adult animals, the content of those enzymes increases in their livers, the activity of which is relatively high in rapidly growing hepatomas and in embryonic liver. Some of these enzymes (tryptophan oxygenase, tyrosine aminotransferase), as the authors point out, are particularly intensively induced by glucocorticoids, and the increase in their activity must therefore be regarded as rather a nonspecific phenomenon associated with stress. Those enzymes that are barely active or are entirely absent in hepatomas and embryonal liver such as ornithine aminotransferase and glutamine synthetase diminish. Hexokinase isozyme I11 proved to be an exception to this rule. It is very low in embryonic liver and is very active in hepatoma. It is precisely because of the increase in its content in the host liver that the total hexokinase becomes more active. Our findings (Shapot et af., 1976) agree with those of Herzfeld and Greengard (1972) and Farron (1972). In the livers of mice with a rapidly growing Guelshtein hepatoma 22a, the total activity of hexokinase and
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
117
its isozyme 111 sharply increased and the activity of glucokinase, the marker enzyme of the liver, diminished 66%-75%. We took notice of one more feature. In tumors, as was pointed out above (Section II,A), glucose is maintained at an extraordinarily low level. In our endeavor to come as close as possible to the actual conditions existing in vivo, we discovered that in the medium containing 0.1 mM glucose, i.e., 2% of physiologic, the activities of hepatoma 22a hexokinase 1 and I1 are low, and only one isozyme 111, characterized, unlike the other two, by a particularly high affinity for glucose, is the most active. In the liver of mice with hepatoma 22a, we observed the same tendency, viz., at a low concentration of glucose in the .medium the highest activity was that of isozyme hexokinase 111 (Fig. 9), the activity of all three hexokinase isozymes at a physiologic glucose concentration significantly exceeding that of the normal liver. According to Teras and Isok (1974), the activity of total hexokinase in the liver of mice with Guelshtein hepatoma 22a increased 2.5-fold on the 28th day of tumor growth and that of glucose-6-phosphate dehydrogenase 2-fold, whereas the activity of glucose-6-phosphatase diminished 50%. In the liver of rats with Morris hepatomas the activity of lysosomal enzymes-8-galactosidase, arylsulfatase, P-glucuronidase, cathepsin, and acid ribonuclease-increases (Shamberger et al., 1971). These findings agree with those of Schersten et al. (1971) concerning the increase in the activity of the same lysosomal enzymes in the liver of 12 patients
FIG. 9. Chromatography of mouse liver hexokinase isozymes on the DEAE-cellulose column on KCI concentration gradient. (A) Normal liver, (B) liver of mice with Guelshtein 22a hepatoma, (C) 22a hepatoma cells. Solid line shows enzyme activity in 5.5 mmolesll glucose: broken line shows enzyme activity in 0.1 mmoIes/l glucose.
118
V . S. SHAPOT
with kidney cancer (see also Lundholm er al., 1976). Jaroszewicz ef al. (1976) examined 13 arninotransferases of the liver of rats with Guerin carcinoma. Of these the activity of valine, isoleucine, and methionine aminotransferases were increased 50%. Earlier we (Krechetova er al., 1972) found that latent endoribonuclease of membrane-bound liver ribosomes belongs to their structural proteins being localized in the small subunit. These ribosomes labeled in vivo with RNA 14C-oroticacid, after 48 hours of incubation under conditions optimal for RNase, split their own RNA 80%-85%. On the other hand, endogenous 14C uridine-labeled RNA of membrane-bound ribosomes of Zajdela hepatoma, rat hepatoma 27, and mouse hepatoma 22a under the same incubation conditions remained intact, which indicated the absence of endogenous ribonuclease activity. As a matter of fact, the enzyme could not be isolated from membrane-bound ribosomes of the above tumors, the possibility of inactivation of RNase by an inhibitor being excluded. Endogenous RNA breakdown in membrane-bound liver ribosomes of hepatoma-bearing rats on incubation was barely detectable. A partially purified preparation of ribonuclease from membrane-bound ribosomes of the liver of tumor carriers proved to be far less active than that from the corresponding ribosomes of normal liver (Table IV), whereas the activity of the enzyme preparation from free ribosomes of the tumor host did not change. Further purification of RNase from membrane-bound ribosomes of normal liver by chromatography on a DEAE-sephadex-A-50 column showed that the main activity was eluted with 0.15-0.17 M NaCl (Fig. 9) the specific activity of the enzyme increasing 200- to 250-fold. In this fraction obtained from Zajdela hepatoma membrane-bound ribosomes, no activity was detected at all, whereas in the case of membrane-bound ribosomes from the liver of animals with Zajdela hepatoma the activity was very low (Fig. lo), making no more than 3% the normal value (note the difference between the DPM scales in Figs. 1OA and B). At the same time we observed no essential difference in the activity of RNase of free ribosomes eluted from the column with NaCl 0.17-0.22 M solution (main activity peak) among the normal liver, the liver of a tumor carrier, and hepatoma. It follows that the activity of RNase of membranebound ribosomes from the liver of the tumor’s host changes in the same direction as in the hepatomas themselves (Sukhovaet a / . , 1978), although these changes are not so strongly pronounced. In contrast to that, RNase activity of membrane-bound ribosomes from rat regenerating liver (48 hours after partial hepatectomy) turned out to be even 25% higher than that of control. A malignant human tumor very appreciably affects the activity of
119
TUMOR AND HOST MULTIFORM RELATIONSHIPS
TABLE IV ACTIVITY OF ENZYMEPREPARATIONS FROM RIBOSOMES I N ARBITARY UNITS(EU) Membrane-bound
Source of enzyme Rats: Liver (normal) Animals with Zajdela hepatoma Animals with 27 solid hepatoma Ascites Zajdela hepatoma 27 hepatoma Mice: Liver (normal) Animals with 22 hepatoma Animals with 48 solid hepatoma Ascites 22a hepatoma Solid 48 hepatoma
Free
EU
Percentage as compared with the liver
EU
Percentage as compared with the liver
25-33 1.4-4.9
Io w0 9.3
12-18 8.8-14
1 00% 76
6.0-8.0
24
14-18
I00
0.0-0.6
1
8.6-12
70
0.0-1.5
2
16.0-18.0 6.0-8.5 5.3
0.0-0.54 1.1-3.0
I00 25
18 1
7
12
19.1-21 10-14
80
100 70
15.5
77
11.6-17.0 12-16
71 70
pyrimidine nucleoside kinases of the nuclear sap of liver cells uneffected by metastases (Borzenko et al., 1977). Normally the activity of these enzymes in the liver is negligible, but in the tissue of hepatocellular cancer the activity of thymidine kinase increases 40- to 50-fold and that of uridine kinase 120- to 150-fold. The activity of these enzymes in the liver of patients with gastrointestinal cancer changes in the same direction, but less strongly (Table V). Herzfeld and Greengard’s findings (1977) fully agree with the above data concerning thymidine kinase of the liver of cancer patients. The increase in thymidine kinase activity in the liver of rats with transplanted Morris hepatoma 7777 and lymphoma RVC 290 was enormous, in the first case exceeding normal activity 5-fold and in the second case 100fold. It is interesting that with the growth of other transplantable tumors, for example adenomasarcoma CCCS and Walker carcinoma, the activity of thymidine kinase in the host liver increased only negligibly. We were surprised to discover, specifically for lymphoid tissue of rats, a sharp decrease in adenosine deaminase activity after Zajdela hepatoma implantation (Gerstein et af., 1978); in the liver of these animals the
120
V. S. SHAPOT
I.0
0.6
0.2
FrmJian
number
FIG. 10. Fractionation of the RNase preparation from rat liver and hepatoma membranebound ribosomes on the DEAE-column. (A) Enzyme preparation from normal liver ribosomes: ( X - - x ) enzyme activity, (0-0) A,,,. (B) Enzyme preparation from the liver of hepatoma-bearing rats: (&--A) enzyme activity, (-0) from Zajdela hepatoma cells: ( x x X ) enzyme activity (zero). (0-0) AZR,).Enzyme activity was measured as DPM of the degraded t4C-labeled ribosomal RNA preparation used as substrate.
activity of adenosine deaminase hardly changes (Fig. 11). Diminished deamination of deoxyadenosine may lead in the lymphoid organs, spleen and thymus, to selective accumulation and subsequent phosphorylation of the latter to deoxyadenosine triphosphate (Carson rt al., 1977), which is capable of suppressing DNA synthesis through the inhibition of ribonucleotide reductases (Reichard, 1968) and TMP kinase (Vornovitskaya e / al., 1968, 1972). It is natural to assume that such impairment of metabolism in lymphoid organs must affect their normal functioning as a part of the immune defense system of the host. Lawson e/ ul. (1977) studied the synthesis of urea and the activity of carbamylphosphate synthetase in Morris hepatomas of various rates of growth and differentiation, as well as in the liver of tumor carriers. The activity of carbamylphosphate synthetase proved to be sharply diminished in the liver of rats with large, slowly growing hepatomas characterized by intensive synthesis of urea. Experiments with parabionts (healthy rats and tumor-carrying rats with a low activity of the liver enzyme) yielded no indication as to the presence in the blood of the latter of any factors affecting its activity.
TUMOR AND HOST MULTIFORM RELATIONSHIPS
121
Here we shall emphasize that not all the tumors examined by Lawson et al. (1977) suppress the activity of host liver carbamylphosphate synthetase. It was clearly marked in the case of heptomas 5123D, 21, and 47C in which urea formation is retarded, whereas in the liver of rats with hepatomas 20 and 9618A the activity of the enzyme was close to normal. It follows that in these studies we encounter again a phenomenon of nonstereotypic effect of malignant tumors on enzyme characteristics of the host tissues. It is, furthermore, noteworthy that tumor carriers in certain instances exhibit diminished inducibility by xenobiotics, zoxazolamine and pentobarbital, of microsomal enzymes that catalyze the primary reactions of their metabolism; this observation undoubtedly reflects a tendency toward reduction of rough endoplasmic reticulum, i.e., the membranebound ribosomes we noted previously with respect to the liver of rats TABLE V NUCLEOSIDE KINASES I N THE LIVERNUCLEAR SAP:I4C-TMP A N D I4C-UMP FORMED" Source of tissue Healthy persons (rr = 10) Hepatomas
Liver metastases Cancer of stomach Stage 11 Stage 111
Stage IV
Cancer of small intestine Stage IV Cancer of cecum Stage 11 Cancer of duodenum Stage I Stage IV a
In cpdrng protein.
Thymidine kinase 4.192 150,400 212,950 187,312 142.217
* 992.1
Uridine kinase 1,654 2 723.4 240.800 374,500 287,300 192.405
5,638 4,874 12,997 8,070 10,000 8,950 7,480 39,714 63,801 45,759 59.714
889 1,002 15,850 93 7 4 13.21 I 13,200 I1,100 14,840 48.849 54.875 68.2 I I
2 1,570
I0.hSI
18,244
15 ,424
1.975 75.424
3.489 79.500
I22
V. S . SHAPOT
nmole/inin mgprot ein
Liver hepatoma Act L V L by of odenosme deammase cn 7Leen
thymus
r o t t~ssues
FIG.11. Adenosine deaminase activity in tissues of normal and Zajdela hepatoma-bearing rats. Activity of the enzyme was measured in the cytoplasmic fraction, obtained by centrifugation of the nuclei-free tissue homogenate in a buffer containing 0.05 M tris-HCI (pH 8), 0.25 M sucrose, and 0.15 M KCI, spectrophotometrically by determination of the initial velocity of the reduction in absorption at 265 nM in the Unicam AR 25 Linear Recorder. Reaction proceeded at 37°C in the 3 ml solution containing 0. I N sodium phosphate (pH 7.0), 4 x 10-5M adenosine, 10 ml protein fraction (30 to 60 pg protein). The 0.019 difference in the absorption value was assumed to correspond to 1 pg adenosine deaminated. (0) Control, (a) tumor-bearing rats.
with Zajdela hepatoma (Section I1,B). Diminished inducibility of microsome enzymes was discovered by Rosso rt al. (1971) in their studies of animals with ten models of rapidly and slowly growing tumors, viz., Walker carcinoma, solid and ascitic forms of Flexner-Jobling carcinoma, sarcoma 45, Guerin carcinoma of uterus, Yoshida rhabdomyosarcoma AN-130 and 602, sarcoma 180, Ehrlich solid and ascites carcinoma. The microsomes of tumor carriers’ liver metabolized zoxazolamine, aniline, amidopyrine, and p-nitroanisol much more poorly than did the microsomes of control animals. This warranted the conclusion that impaired in the tumor host are the reactions of aryl- and alkyl hydroxylation, as well as N - and O-demethylation. As a result, animals with the above tumors, especially females, turned out to be much more sensitive to the paralyzing action of zoxazolamine and the hypnotic action of pentobarbital. For example, after administration of a standard dose of pentobar-
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
123
bital, females with sarcoma 180 slept ten times longer than did healthy animals. After surgical removal of the tumor the metabolism of xenobiotics in the host liver was normalized. As a result of the imbalance between the systems of generation of lipid peroxides (NADPH-dependent microsomal dioxygenase), on the one hand, and its inhibition (0,-dismutase) and the utilization of the above peroxides (GSH-peroxidase, GSSG-reductase), on the other hand, the microsomes and mitochondria of the liver of animals with Ehrlich ascites carcinoma and primary sarcoma induced by 3,Cbenzpyrene accumulate excessive amounts of lipid peroxides (Lankin et al., 1977). The excess of lipid peroxides produces a toxic effect on the microsomes, i.e., the activity of the inducible enzyme, 3,4-benzpyrenehydroxylase is sharply suppressed and the level of cytochrome P-450diminishes. In the mitochondria the activity of monoaminoxidases decreases, but the ability to deaminate adenosine monophosphate, and coenzymes containing this nucleotide, NAD, CoA, etc., increases (Khuzhamberdyiev et al., 1973). It is probably precisely the accumulation of peroxides of unsaturated fatty acids that accounts for the partial uncoupling of respiration and phosphorylation in the mitochondria of the tumor carriers’ liver (Shapot et a l . , 1976; Morozkina and Shapot, 1976). The effect of the slowly growing sarcoma 45 and rapidly growing Walker carcinoma on the energy metabolism of mitochondria of tumor carrier’s liver, isolated under maximum sparing conditions, and then incubated with succinate or glutamate, was studied by the polarographic method. The extent of dissociation of respiration and phosphorylation increased with the stage of tumor growth and in accordance with the degree of its malignancy. A decrease in the ADP/O coefficient and the respiratory control-diminished stimulation of oxygen consumption after addition of ADP-was observed. The NAD-dependent link of the respiratory chain proved to be the most vulnerable. The uncoupling of respiration and phosphorylation could to a certain extent be normalized if the mitochondria of the tumor carrier’s liver were incubated with serum albumin which, as is well known, easily binds lipid peroxides and unsaturated fatty acids. The above interpretation seems to be supported by the data of Chan and Higgins (1978). The in vitro aging of rat liver mitochondria was accompanied by an increase in the fatty acids levels with a concomitant uncoupling of oxidative phosphorylation. Replacement of the medium by 0.25 M sucrose containing 1% defatted bovine serum albumin reduced the free fatty acids levels and restored oxidative phosphorylation (raised
124
V . S. SHAPOT
respiratory control index and ADP: 0 ratio). A substantial fraction of '2SI-iodinated serum albumin was shown to bind to mitochondria1 membranes under these conditions. Last, we shall mention a characteristic shift in the spectrum of lactic dehydrogenase (LDH) isozymes in all the tumor-unaffected tissues of cancer patients we studied (Shapot, 1975) jointly with Gorozhanskaya. These organs clearly show a diminished share of isozyme I, a change in the direction of the cathode fractions, and an increase in the coefficient (Gorozhanskaya and Shapot, 1971) which reflects the quantitative ratio of isozyme V to isozyme I activities (Table VI). We noted an enormous increase in the coefficient V:I for all the malignant tissues, whereas in benign neoplasms this coefficient does not differ from normal (Gorozhanskaya and Shapot, 1971). Similar results were obtained by American authors (Hawrylewiez rt ol., 1974) concerning the influence of malignant mammary tumor on portions of tissue of the same organ not affected by the neoplasm. Recalculation of these authors' data for the purpose of deriving the aforesaid coefficient showed that normal tissue of the mammary gland is charac-
TISSUELDH-1
AND
TABLE VI V ISOENZYMES I N PERCENTTO LDH ACTIVITY
THE
TOTAL
Tissue Kidney Control Cancer patient Carcinoma Stomach Control Cancer patient Carcinoma Lung Control Cancer patient Carcinoma Mammary gland Control Cancer patient Carcinoma Uterus Control Cancer patient Carcinoma
43.2 t 1.6 34.7 f 1.9 14.4 -c 2.0
2.9 t 0.6 5.5 t 0.9 22.7 t 2.9
0.06 0.16 1.57
30.1 t 1.9 22.8 2 1.4 3.7 2 0.4
7.0 k 0.5 10.2 t 0.7 23.1 ? 1.0
0.23 0.44 6.20
28 25
20.5 2 1.0 9.8 t 1.1 6.3 t 1.3
12.3 2 1.7 12.8 2 1.0 26.7 t 1.8
0.60 1.30 4.30
6 30 30
14.1 t 1.0 5.7 0.9 3.7 2 0.6
*
9.3 & 0.5 5.7 t 4.6 32.1 & 4.6
0.66 1.00 9.10
6 30 30
11.0 -c 2.0 9.0 t 1.2 5.4 2 0.6
7.0 2 0.6 7.8 2 1.0 18.2 2 1.7
0.63 0.79 3.70
15
16 17 15
40 45 15
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
125
terized by a coefficient that is close to 0, malignant tissue by a coefficient of 5.2, and the portion of the gland distant from the tumor has a coefficient of 0.36. In the latter case appear cathode isozymes IV and V which are usually absent in the normal gland, while isozyme I diminishes about 66%. The above phenomena denote a tendency toward dysdifferentiation of specialized tissues manifested in changes in enzyme characteristics inherent in normal specialized tissue.
B. CYTOPLASMIC INFORMATIONAL RNAs
AND
RNP COMPLEXES
There are reasons to believe that abnormalities in many biological characteristics of distant host tissues induced by the growing tumor may be to a certain degree conditioned by certain features in the multistep process of gene expression. Such deviations are obvious in transplantable hepatomas but could be revealed in the liver of tumor-bearing animals as well, although they are less pronounced. No appreciable differences in DNA-like RNA (D-RNA) between normal rat liver, Zajdela hepatoma, and liver of hepatoma-bearing rats could be detected in fractionating them by ultracentrifugation on a sucrose gradient and by chromatography on a methylated albumin Kiselgur column (MAK) using a salt and temperature gradient which permits separation of RNAs according to their molecular weight and nucleotide composition (Ellem and Sheridan, 1964: Lichtenstein and Shapot, 1971). Corresponding sedimentograms and chromatograms proved nearly identical. However, drastic differences were revealed on studying cytoplasmic nonribosomal RNAs of the three tissues examined. To selectively label both nuclear and cytoplasmic D-RNAs, a partial actinomycin D block (low doses) was used arresting the synthesis of ribosomal RNAs but not affecting D-RNA synthesis. It turned out that liver cytoplasmic D-RNAs were far more heterogeneous in hepatoma liver and in tumor-bearing animals than in normal rat liver. On MAK chromatography all the cytoplasmic D-RNA of normal liver, unlike nuclear D-RNA of the same tissue (eluted at an elevated temperature), was released in the salt fraction, i.e., was represented by a population of molecules loosely bound to the adsorbent and of relatively low molecular weight. In contrast to that in the two other tissues a significant proportion of cytoplasmic D-RNA was eluted both in the salt and “temperature” fraction, the latter representing RNA tightly bound to the adsorbent (Shapot
126
V. S . SHAPOT
et ul., 1977). A similar heterogeneity in cytoplasmic D-RNA of hepatoma and liver of tumor-bearing rats was revealed by sedimentation characteristics (Lichtenstein et al., 1978). The above findings indicate a reduced selectivity in the nucleus to cytoplasm transport of D-RNA in these tissues as compared with normal liver. What kind of factors may be involved in producing such an effect? Schumm and Webb (1975) perfected the model system operating in vitro, which permits study of alterations in the transport of mRNA from nucleus to cytoplasm depending on the nature of the cytosol which serves as incubation medium. These authors using competitive DNA : RNA hybridization techniques demonstrated that the mRNA transferred from normal liver nuclei to the homologous cytosol differs significantly from that transferred from the same nuclei to the hepatoma cytosol. Particularly impressive are the findings of Schumm and Webb concerning a sixfold stimulation of the transport of mRNAs from normal rat liver nuclei and broadening of their spectrum in the presence of a partially purified protein fraction obtained from the blood plasma of tumor carriers. Twenty-five percent of the mRNA sequences transported in the medium containing the plasma protein fraction from normal animals turned out to be different from that transported in the presence of the corresponding fraction from the plasma of tumor-bearing animals. These experiments carried out on animals bearing Morris 9618A and 5123D hepatomas and Novikoff hepatoma yielded identical results. Let us mention one more observation (Shapot et ul., 1977) concerning the transport of D-RNA from nucleus to cytoplasm. The dynamics of the cytoplasmic D-RNA specific radioactivity under a complete actinomycin D block of nuclear RNA synthesis (“chase”) was studied. In normal liver this block leads to a rapid decline in the specific radioactivity of cytoplasmic D-RNA, whereas in liver of tumor-bearing animals it remains unchanged in hepatoma, being even somewhat elevated. Under the same conditions the radioactivity of nuclear RNAs dropped in all three tissues studied. The same regularity was established for poly(A)-containing cytoplasmic D-RNAs. A complete arrest of RNA synthesis induced a rapid reduction of their content in normal liver but not in the liver of hepatomabearing rats and in the hepatoma itself (Lichtenstein et al., 1978). The most reasonable explanation for the phenomena described would be that nuclei of hepatoma and of liver of tumor-bearing animals contain a large pool of cytoplasmic mRNA precursors which is able to maintain within at least 30 minutes the nucleus-to-cytoplasm transport at a constant level, unlike normal liver in which a nuclear pool of mRNA precursors becomes exhausted very rapidly once transcription is blocked.
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
127
Indeed, kinetic curves of the specific radioactivity of cytoplasmic DRNAs under a partial actinomycin D block demonstrate that in hepatoma and liver of tumor-bearing animals these RNAs appear from the nucleus very soon and within 60 minutes reach the plateau, while in normal liver D-RNA transfer from the nucleus proceeds slowly, its content in the cytoplasm increases gradually and does not reach plateau even within 120 minutes (Shapot et al., 1977). There is one additional feature which is characteristic of hepatoma and liver of tumor-bearing animals. In hepatoma the proportion of rapidly labeled informosomes (postribosomal ribonucleoprotein complexes) was found significantly elevated at the expense of the reduction of polyribosoma1 mRNPs. This tendency is less pronounced in the liver of tumorbearing animals. The above phenomena with selectively labeled D-RNA were revealed using three independent methods: by fractionation on sucrose concentration gradient, isopicnic ultracentrifugation in CsCl gradient, and finally by the technique used in our laboratory (Lichtenstein et al., 1975), namely, nucleoprotein-celite chromatography permitting the separation of nucleoproteins on the basis of the tightness of the RNAprotein bonds. The latter method allowed us to obtain polyribosomal mRNPs, a very firmly built complex, and informosomes characterized by loose RNA-protein bonds in the opposite poles of the chromotogram. This feature, predominant proportion of informosomes in the tumor and liver of tumor-bearing animals, seems to be specific for them and is not found in regenerating liver. In addition, a shift in the polysomes/monosomes ratio toward the latter was observed in hepatomas and liver of tumor bearing animals (Shapot et al., 1977). All the deviations listed turned out to be typical not only of Zajdela hepatoma and the liver of its host but were verified in this laboratory on mouse Guelshtein hepatomas (Zborovskaya et a/., 1978).
c. DISORDERS O F ENDOCRINE REGULATION We have already mentioned the absence of gluconeogenesis stimulation in mice with Ehrlich carcinoma despite the profound hypoglycemia caused by the tumor. Special experiments have shown (Blinov and Shapot, 1974a; Blinov et al. 1975) that transplantation of this tumor sharply elevates the threshold of liver tissue sensitivity to the action of glucocorticoids which stimulate glycogenesis, gluconeogenesis and glycogen neogenesis. For example, whereas in healthy mice one day after administration of 5 mg of hydrocortisone the content of 14C-glucoseand I4Cglycogen in the liver (radioactive alanine served as precursor) increased
128
V. S. SHAPOT
6-fold and 1 lj-fold, respectively, in tumor-bearing animals they increased only 3- and 2-fold. Moreover, under conditions that cause hyperproduction of glucocorticoids and other hormones that stimulate gluconeogenesis (various forms of stress) in the liver of these mice, gluconeogenesis remained at the initial level or even diminished 50% (paradoxical reaction), whereas in healthy animals the rate of gluconeogenesis doubled. The elevation of the sensitivity threshold of gluconeogenesis to hydrocortisone was noted in the liver of rats with Zajdela hepatoma as well (Blinov er al. 1975). This phenomenon is very likely the result of a decrease in the content and the change in the properties of glucocorticoid cytosol receptors. As a matter of fact, K,,, to dexamethasone of the liver of rats 5 days after transplantation of Zajdela hepatoma diminishes from (108M-l) x 3.8 2 0.23 to (108M-l) x 1.74 2 0.29 0, < 0.001), while the number of binding sites decreases from 4.8 2 0.15 X mole/mg mole/mg protein p < 0.05 (Dmitriyeva et protein to 3.76 ? 0.5 x a l . , 1976). It is curious that in some cases it is possible to reveal diminished reactivity of the tissues of the tumor host not only to hormones but also to other regulatory factors. It is well known that stimulation of gluconeogenesis in the liver may be induced by an excess of glucogenic amino acids of corresponding enzymes (substrate induction). Control mice and animals with Ehrlich ascites carcinoma were injected with considerable amounts (0.6 gm/kg) of serine, as an inductor and with small quantities of 14C-alanine,as a radioactive glucose precursor. In this case the rate of gluconeogenesis increased sharply, about threefold, in the liver of healthy mice, whereas in the liver of animals with the tumor gluconeogenesis it not only failed to be stimulated, but was even slightly depressed (Shapot, 1975a). We also studied the changes in reactivity of cancer patients' lymphocytes to mitogens and hormones. On blast-transformation of peripheral lymphocytes induced by mitogens (phytohemagglutinin), three forms of RNA-polymerase are known to be stimulated. RNA synthesis catalyzed by A and B RNA-polymerases was studied under these conditions in isolated lymphocyte nuclei. The activity of these two RNA-polymerases in melanoma patients (n = 18) was higher than in donors (n = 32) 6 and 2 times, respectively; in lung cancer patients (n = 18) 5 and 4 times as high; and in sarcoma patients (n = 20) 7 and 8 times as high as in donors. Unlike the RNA-polymerases of the donors' lymphocytes the activity of these enzymes in cancer patients was stimulated by phytohemmagglutinin to a lesser degree, on the average of 50% for form A and 65% for form B. Dexamethasone inhibited the stimulated lymphocyte RNA-polymerase
TUMOR AND HOST MULTIFORM RELATIONSHIPS
I29
A in donors and cancer patients about equally, while the sensitivity of RNA-polymerase B to the hormone was lowered 50%. It was also shown that the reduced sensitivity of the RNA-synthesizing system of lymphocytes to glucocorticoids develops in parallel with the dissemination of the melanoma, and it was proposed that the lowered reactivity of cancer patients' lymphocytes to glucocorticoids is conditioned by a decrease in representation of the TI-cell subpopulation and its replacement by nonsensitive T,-cells (Ioannesyants et af., 1977, 1978). It is probable that resistance of certain human leukemias and lymphomas to glucocorticoid treatment is due to the T,-cell or B-cell origin of these neoplasms. Further we (with Shelepov and Davidova) demonstrated functional disturbances in the insular apparatus in rats with two forms of hepatomas (see Shapot, 1978). Figure 12 shows that in normal rats (1.5 months old) and in rats with Zajdela hepatoma the maximum rise in the glycemic curve occurs in response to a glucose load, as was to be expected, on the 30th minute. In healthy animals a synchronous increase in immunoreactive insulin (IRI) in the blood was observed, its maximum also being reached toward the 30th minute. As for rats with Zajdela hepatoma, they
Hours
afhef gLucose LoaU
FIG. 12. Effect of the glucose load on the serum sugar and immunoreactive insulin (IRI) levels in rats. Glycemic curves: 1, control: 2, rats carrying Zajdela hepatoma. IRI curves: 3, control: 4, rats with Zajdela hepatoma.
130
V . S . SHAPOT
are characterized by a relatively low initial IRI level in the blood serum and its delayed increase after a glucose load with also a delayed decrease, the initial IRI level failing to be reached even with 3 hours of observation. The amplitude of reactive hyperglycemia at its maximum (in the 30th minute) in rats with Zajdela hepatoma turned out to be higher than in normal animals, but the assimilation of glucose by peripheral tissues was retarded. An analogous study conducted on rats with solid hepatoma 27 found no initial hypoglycemia in these animals despite the elevated level of IRI. The assimilation of glucose after the load was also prolonged in these animals, and there was no synchronism in the peak of the glycemic curve and the IRI level: but, unlike the rats with Zajdela hepatoma, the deviations from the norm were, in this case, not stereotypic (Fig. 13). From the data obtained, the mean rates of glucose assimilation were calculated (glucose tolerance test). In healthy rats this rate was 3.1 mg 96 glucose/min: in rats with hepatoma 27, 0.56 mg/lOOmYmin; and in rats with Zajdela hepatoma, 0.9 mg/100 m l h i n . In the latter case the return of the glucose level from the reactive maximum to the initial one was prolonged to 138 minutes. t
T
//
r-.--.
Houm from the gLuco5-e Load FIG. 13. Effect of the glucose load on the serum sugar and IRI levels in rats. Rats carrying solid hepatoma 27. Glycemic curve: 1: IRI curves: 2, 3, 4; different animals.
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
13 1
TABLE VII SERUM INSULIN/GLUCOSE INDEX IN
RATS AFTER
THE
GLUCOSELOAD Time after the administration of glucose (minutes)
Rats carrying Control rats
Zajdela hepatoma
27 hepatoma
0.280 0.420 0.450 0.134 0.170
0.135 0.117 0.224 0.290 0.300
0.510 0.220 0.245 0.220 0.098
~~~~~
0 30 60 120 180
Table VII shows the changes in the insulin-glucose index (IRI in pU/ ml to glucose in mg/100 ml) in tumor-bearing animals compared with healthy ones in the glucose tolerance test. The increase in the index in rats with Zajdela hepatoma, beginning from the 60th minute after the load, indicates that to assimilate an excess of glucose they require more insulin than normal: in other words, the sensitivity threshold of their tissues is elevated. In rats with hepatoma 27 attention is called to the extraordinarily high initial index. Interesting observations were made by Teras (1979) on the changes in the inducibility of hexokinase, glucose-6-phosphate hydrogenase, and glucose-6-phosphatase by triiodothyronine (T,) in the liver of mice with Guelshtein hepatoma 22a. On the seventh day of tumor growth the hormone caused a much greater increase in the activity of the above enzymes-8.5-, 3-, and 2.7-fold, respectively-than their induction in normal liver (supersensitivity!). However, a different picture was observed on the 28th day after transplantation of the tumor, viz., the liver nearly lost its sensitivity to T,. The above phenomena obviously reflect a profound distortion of the reactivity of the host target tissue to T,. According to Horvath et al. (1975) induction of tyrosine aminotransferase by cortisol in the liver of chickens with transplanted hepatoma MC 29 is diminished. In the thymus of healthy chickens the activity of thymidine kinase decreases 80% after two injections of cortisol or dexamethasone in the course of 48 hours, whereas thymidine kinase of the thymus of chickens with hepatoma did not react to the hormone at all (Naray et af., 1977). Patients with stomach cancer (n = 75) exhibited a progressive (with the spread of the tumor process), statistically reliable decrease in the level of immunoreactive insulin in the blood serum from 18.5 2 4.2. p U /
I32
V . S . SHAPOT
ml in early stages (versus 23.3 k 4.0 in the control group with ulcer and gastritis, n = 28) to 11.8 r 1.6 pU/ml in the 111-IV stages (Vereschagina et al., 1977). The same authors noted as the most characteristic disorder in these patients, impairment of the inverse correlation between the production of adenohypophyseal hormone and the peripheral hormones. The content of the somatotropic hormone in the blood is sharply increased to 1.45 & 0.43 mg/ml in the late stages versus 0.30 & 0.01 in the control group. The coefficient of thyrotropic hormone (TTH) to triiodothyronine in the blood increases from 0.85 (control group) to 1.5 with a tendency toward a decrease in the level of T3 and an increase in the ratio of the total thyroxin to T3. The authors interpret these phenomena to result from an overstrain of thyroid function and its eventual exhaustion. Similar phenomena were discovered by Afrikyan (1977) in patients with lung cancer (n = 102), stomach cancer (n = 92), mammary cancer (n = 26), acute leukemia, lymphosarcoma, and myeloma (n = 56). Regardless of the localization and histogenesis of the malignant tumor the patients exhibited before treatment, an increase in the level of the thyrotropic hormone and a decrease in the concentration of T, in the blood with a coefficient of TTH/T, = 2 and higher. There were no such deviations in cases of benign neoplasms. Low T, syndrome was recognized also by other authors (Dessaint et al., 1978) as characteristic of cancer patients. In cancer patients, thyroid hypofunction apparently causes hyperplasia by the mechanism of negative feedback, overstrain of the adenohypophysis, which leads to depletion of basophils and subsequent dystrophy and even atrophy of the gland (Monastyrskaya, 1963). The results of histochemical studies of the thyroid and adenohypophysis of rats with transplanted tumors corroborate these conclusions (Morozkina er a / ., 1975). The reduced responsiveness of target tissues to corresponding hormones is likely to be characteristic of the tumor’s host, although this feature may manifest itself differently in each given instance. In women with mammary cancer, unlike healthy women, the elevated levels of estrogens during the menstrual (ovarian) cycle did not suppress the production of follicle stimulating hormone (FSH). No reduction in the blood FSH after administration of sinestrol into patients with mammary cancer in menopause could be observed. However, on remission induced by respective therapy an excess of sinestrol did suppress FSH production in such patients and failed to do it on recurrence (Sharoukhova, 1974; see Fig. 14). According to Dilman (1974), in patients with cancer of endometrium, mammary gland, and colon, a glucose load did
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
133
not reduce the levels of growth hormone in the blood. A paradoxical reaction to glucose load-hyperproduction of growth hormone-in patients with cancer of the endometrium was reported (Benjamin, 1974). In patients (n = 37) with the same localization of tumor the release of growth hormone in response to hypoglycemia induced by insulin proved very poor (Madajewicz et al., 1977). Patients with mammary cancer display elevated levels of blood cortisol which could not be suppressed by the administration of dexamethasone. The same phenomena were observed in patients with other forms of cancer. The author (Saez, 1974) draws attention to the abnormally high threshold in the hypothalamo-hypophyseal mechanism regulating the secretion of the adrenal steroids characteristic of patients with various forms of cancer. The feedback regulation of ACTH production by adenohypophysis is preserved, but higher amounts of cortisol are needed for this mechanism to be put in effect, particularly in patients with advanced cancer. All the observations described leave no doubt as to the profound disorders in endocrine regulation provoked by the malignant tumor in the host as one of the manifestations of its systemic effects.
D. IMMUNODEPRESSION Immunodepression doubtless is characteristic of both tumor-bearing animals and cancer patients, the degree of immune deficiency increasing with the neoplasm growth and dissemination. Nonspecific immune defense (a particular vulnerability to infection) as well as antitumor immunity seem frequently to be reduced (Harris and Copeland, 1974). There are some indications that this phenomenon is a reflection of impairment of certain links of immunogenesis. The proliferation and differentiation of T- and B-lymphocytes are hindered. The growing tumor affects the cooperation of T and B cells. In addition, the activation of T suppressors which weaken both humoral and cell-mediated immunity was demonstrated in mice with Ca-755 (Petrov and Khaitov, 1977). Analogous observations were made in other laboratories as well. Lung macrophages of rats with Walker 256 carcinoma growing in their lungs were shown to exhibit a sharply reduced capacity for phagocytosis, particularly at the advanced stages of tumor development. Phagocytosis was suppressed to the highest degree when the macrophages were incubated in the presence of the host serum. One to 3 weeks after tumor implantation, glucose oxidation by the host lung macrophages was strongly inhibited. In other words, a serious impairment of energy me-
134
V. S. SHAPOT
9
Befure
treutment RemLssLon
Recurrence
FIG. 14. Excretion of follicle-stimulating hormone by patients with mammary carcinoma after sinestrol treatment (Sharoukhova, 1974).
tabolism supporting macrophage functional activity was provoked by the tumor (Gudewicz and Saba, 1977). It is of interest that very low density ( p = 1.006) lipoproteins and low density (1.006 < p < 1.063) lipoproteins, isolated from cell-free ascites fluid of MM46 tumor, growing in syngeneic C3HIHE mice, suppressed lysis of this tumor by activated macrophages in vitvo (Yamazaki et ul., 1977). The exact reasons for the phenomenon of immune depression in the tumor host are not yet known: many factors seem to be involved. It is likely that certain substances are released in the circulatory system that interfere with immunogenesis. Normal human blood serum contains “immunoregulatory a-globulin” and related polypeptides capable of inhibiting various T-cell-mediated reactions: proliferations of lymphocytes in response to contact with specific antigens or mitogens as well as imrnunity of mice to syngeneic tumors, effector cells, or target cells being unaffected. “Immunoregulatory a-globulin” was found in particularly high concentrations in the serum of patients with metastasizing cancer and ascitic fluids (Wang e f ul., 1977). Hence, an excessive synthesis of this a-globulin and related peptides, thus flooding the blood of cancer patients, may be one of the factors inducing imrnunodepression. The above authors believe that a reversible block of the process of “recognition” in the presence of “irnrnunoregulatory substances” underlies the
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
135
loss of cytotoxicity to the syngeneic tumor by sensitized mouse lymphocytes. The data on the immunosuppressive effect of the products of fibrinogen proteolytic degradation were reported by Girman et al. (1976). Dialyzable low molecular weight peptides (below 5,000 daltons) obtained by treatment of fibrinogen by thrombin were shown to suppress 98% blast-transformation of peripheral lymphocytes and sharply reduce the rosette-formation capacity of mouse lymphocytes immunized with sheep erythrocytes. Girman et af. examined 19 cancer patients and 3 patients with hepatocirrhosis. In most cases a correlation between the presence of fibrinogen degradation products in the serum and its immunodepressive activity was found, but in four patients no correlation could be revealed. The serum of rats carrying Morris 5123D hepatoma exerted an immunodepressive effect inhibiting T-cell function first and subsequently that of B cells, this effect increasing with the enlargement of the tumor. The active factor, presumably an antigen-antibody complex, was thermolabile, nondialyzable, and could be removed from the serum by adsorption on lymphocytes (Schumm et al., 1976). With primary hepatoma and teratocarcinoma the part of an immunodepressive factor may be played by a-fetoprotein (Murgita and Tomassi, 1 97 5a,b) . Let us turn to the possible effects of endocrine disorders and deviations in metabolism on the immune status of the tumor host. Hyperproduction of adrenal steroids in the tumor hosts occurs frequently as was mentioned above (Section 11,B).According to Kendysh (1972), glucocoritcoids induce protein degradation not only in skeletal muscle but in the thymus and spleen as well, and thereby supply stimulated gluconeogenesis with amino acids as glucose precursors. Involution of thymus manifested by a shrinkage of the gland and increasing destruction of thymus lymphocytes is a constant phenomenon in tumor-bearing animals, observed during the growth of Walker 256 carcinoma, Yoshida sarcoma, Zajdela hepatoma, benzpyrene-induced rat sarcoma, plasmacytoma of the Syrian hamster, spontaneous lung carcinoma, S-180 sarcoma, and mouse leukemia (Ertle, 1973). Before metastases occurred (Walker carcinoma), the involution of the thymus was reversible provided the tumor and adjacent lymph nodes were surgically removed (Toma and Simu, 1973). Thymus involution was regarded as a result of hyperfunction of the adrenal (Ertle, 1973) which may represent, from our point of view, a compensatory response of the host to the tendency toward hypoglycemia (see Sections II,A and B) induced by the tumor.
136
V . S . SHAPOT
Kirikuta rr d.(1970) point to the apparent analogy between the involution of thymus provoked by cancer and stress thymolytic atrophy, leading to an increase in the glucocorticoid levels in the blood. Bilateral adrenalectomy was shown to retard the involution of the thymus in rats carrying Walker 256 carcinoma. A correlation between elevated 17-oxycorticosteroid concentrations in the blood and depressed immune reactions was found among patients with the most advanced mammary cancer (Saez, 1974). The interception of nucleic acid pyrimidine nucleotide precursors, thymidine and uridine, from the thymus and spleen by the tumor (Section II,B) may have a direct bearing on immunodepression in the host. In addition, we recall our observation of a sharp decrease in the activity of spleen and thymus adenosine deaminase in rats with Zajdela hepatoma while its activity in their livers remains unchanged (Section III,A). Many hereditary combined forms of immune deficiency are known to be a result of the absence or a very low activity of the enzymes catalyzing the conversion of adenosine to inosinic acid, and particularly of the first one-adenosine deaminase (Giblett et d.,1972). This defect leads to the accumulation of the deoxyadenosine (to be more exact, dATP) exerting a toxic effect on lymphocytes (Allison et a / . , 1977; Carson et ( I / . , 1977) and predominantly on T-cell population. It is quite possible that a reduced activity of adenosine deaminase has its share in the phenomenon, namely, causing T-cell dysfunction. Indeed a reduced (50% to 25%) activity of adenosine deaminase in the blood of 321 cancer patients as compared with that of patients with benign tumors and other diseases was reported by Dinescu-Romalo c’t a / . (1977). However, the above findings would have been more convincing were the measurements made directly on peripheral lymphocytes. Precisely this has been done by Uberti L’t a / . (1976) on individuals with various types of solid malignant tumors. On the average the activity of adenosine deaminase in their lymphocytes was found to be substantially lower than that in lymphocytes of healthy subjects. Similar observations concering lymphocytes of patients with lung cancer were reported by Ogawa cr a / . (1977). Finally, we will dwell on the data of Dudrick’s group (Copeland o r d . , 1977) as to the restoration of the impaired cell-mediated and humoral immunity in subjects with malnutrition by means of parenteral hyperalimentation, i.e., infusion of the solution containing glucose, amino acids, electrolytes and vitamins ( A , D, E, C, B,, B,, B,, niacin, and panthotenate). It may be expected N priori a benefic effect of the adequate parenteral nutrition on cancer patients who because of a parasytic character of tumor growth suffer from malnutrition besides impairment of intestinal
TUMOR AND HOST MULTIFORM RELATIONSHIPS
137
adsorption of enterally supplied food. Proceeding from the above reasons, Copeland rt al. (1977) have studied rats with Morris 5123 hepatoma divided into three groups, kept on: (a) a carbohydrate-rich protein-free diet, (b) a standard diet, and (c) adequate parenteral nutrition. In animals of group (c) no stimulation of tumor growth occurred and, what was the most important, immunocompetency was fully restored in contrast to the other two groups. Dudrick’s group (Souchon et al., 1975) showed in addition that adequate parenteral nutrition alleviated gastrointestinal toxicity of 5-F-uracil both in healthy rats and in cachexic cancer patients subjected to this chemotherapy. The effect of adequate parenteral nutrition was then studied i n emaciated patients with malignant neoplasms of various localizations. In 17 out of 23 cancer patients subjected to chemotherapy, who displayed impaired cell-mediated immunity, within 11 days of a systematically applied parenteral nutrition the skin test became positive and subsequent chemotherapy proved effective. Copeland et al. (1977) inferred that reduced immunity in the tumor host may be partly a result of malnutrition.
IV. Prospects for the Clinic
With due regard to the characteristics of carbohydrate metabolism of cancer cells in the host described above (Section II,A), it is possible to achieve a selective increase in their sensitivity to nonspecific antitumor agents, and thus be able to use them in smaller amounts and thereby relatively harmless to normal tissues by increasing the gap between toxic and therapeutic doses. Numerous attempts made in recent decades to block the growth of tumors by suppressing glycolysis have, as is well known, failed. We (Shapot, 1968, 1970) already suggested the possibility of selectively sensitizing tumors to the action of damaging factors-chemical, radiation, and thermal-without inhibiting glycolysis, but by stimulating it to the utmost, creating through hyperglycemia conditions for the maximum saturation of the glycolytic enzymes. This idea also occurred to Ardenne and his associates (1972) who are still developing it in their studies. The gist of the approach is that by involving an enormous amount of glucose in metabolism of cancer cells (satisfying their high potential capacity for glycolysis that is not realized under usual conditions), the tumor accumulates an excess of lactic acid, which leads to sharp self-acidification of the tissue whose pH may diminish to 6.2 or lower, whereas the pH of normal tissues persists at the former level. In such a condition cancer
138
V. S. SHAPOT
cells become particularly vulnerable to damage, probably because of an increase in the permeability of lysosomes membranes and the release from them of hydrolytic enzymes which are conducive to cell autolysis (Ardenne and Rieger, 1972). According to Ardenne’s findings, chemical agents that affect membranes accelerate the above process and, consequently, intensify the antitumor effect of the complex therapy. Hyperthermia also plays a role in complex tumor therapy, since many authors have shown cancer cells to be more sensitive to overheating than normal cells are. Grievous experience has shown that humans do not tolerate extreme hyperthermia, i.e., a temperature of 44”C, for which reason only moderate hyperthermia, not above 40”C, is used in treatment. The attempts of the group headed by Aleksandrov and Fradkin at the Minsk Oncologic Institute (Savchenko rt al., 1977) to put into practice the above principles of selective sensitization of tumors in terminal and inoperable patients with far advanced and generalized forms of malignant tumors, melanomas, chondro- and osteosarcomas, and mammary cancer, in particular, seem to us very promising. Maintenance of prolonged hyperglycemia for 6 to 22 hours at a level somewhat higher than 4 gdliter, by administration of concentrated glucose solutions through a catheter into the superior vena cava proved to be optimum. Hyperglycemia is combined with (a) 4-5 hours of hyperthermia (41°C) with simultaneous cooling of the head and neck; (b) administration of various antitumor drugs; (c) administration of a thermosena lysosome sitizer of cancer cells-naphtidon and vitamin A-as sensitizer: and (d) irradiation. Endotracheal anesthesia with artificial ventilation of the lungs is necessary during the period of hyperthermia. The second “shock,” i.e., chemotherapy or combined chemo- and radiotherapy in conjunction with hyperglycemia and brief hyperthermia, is administered 72 hours after the first session. Today the procedures are so perfected that they are absolutely safe for the patient. It is possible to reduce the doses without diminishing the therapeutic effect: for example, those of antitumor drugs have been reduced 50%-92% and irradiation 90%-93%. Fifteen days after the procedures, and with local hyperthermia immediately afterward, the tumors diminish, necrosis occurs, and “shrinkage and drying-in” set in even when, under usual conditions, the tumor is resistant to chemo- or radiotherapy. This treatment makes possible surgical removal of the remaining tumor nodes, which is necessary in order to prevent recurrences (Savchenko et al., 1977). Application of local hyperthermia (in cases of superficial tumors) by high frequency currents is particularly effective in the above complex of
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
139
treatment since it makes possible targeted heating of neoplasms to 42”43°C. By the end of 1976, combined treatment in its general and local variants had been administered to 80 patients with advanced forms of cancer considered “hopeless.” The results can be judged at least by the following figures: 34 are still alive, 25 from 1 to 6 years. Thus it is obvious that the hyperglycemic background essentially enhances the effectiveness and selectivity of the usual therapy, especially when chemical, radiation, and thermal factors are combined. It would be appropriate to devote some attention to one more clinically important aspect of the systemic action of tumors. A novel method of quantitatively evaluating the functional state of the liver by the kinetics of absorption and excretion of radioactive, labeled I3lI Rose Bengal with the use of cybernetic techniques was recently elaborated. This has made it possible to reduce the multiparameter diagnostic space, which formerly rendered difficult an unambiguous interpretation of results, to a unidimensional space (numerical classification or a physiologic scale of functions) (Sivoshinsky and Narkevich, 1975; Sivoshinsky et a / . , 1977; Perevodchikova et u / . , 1977). Many hundreds of patients with malignant neoplasms of nonhepatic localization exhibited, before treatment, an increasing depression of the functional state of the liver in the I11 and IV stages of the tumor development. Surgical removal of the tumor or clinically effective antitumor treatment clearly improved the function of the liver. Thus evaluation of the extent of the tumor’s systemic effect on the liver makes it possible to judge the results of the use of therapeutic factors and to test quantitatively the toxic side effects of the drugs. Of course, the above approach, which makes it possible to study only one of the many functions performed by the liver, cannot give an exhaustive appraisal; but as a test method, it has demonstrated its value in practice.
V. Conclusion Here we should like to express some general considerations and assumptions which follow from the materials set forth in this article. It is reasonable to assume that interception of glucose by the malignant tumor from host tissues which creates “hypoglycemic pressure” on the host is a phenomenon triggering a cascade of various metabolic and endocrine disorders in the host. Among them are: (a) hyperfunction of adrenals followed by the stimulation of catabolism of proteins as well as gluconeogenesis from amino acids, the latter becoming less available for tissue protein synthesis: (b) functional impairnient of the insulin produc-
140
V . S. SHAPOT
ing apparatus: and (c) mobilization of lipids with possible blood hypercoagulability as one of the consequences (see Section 11,C). It is most likely that the main form of systemic action of a malignant tumor on the host is its ability to compete successfully with the host tissues for the vitally important metabolites and other factors. Such an advantage may be partly conditioned by the characteristics of the surface membranes of tumor cells which favor unusually rapid transport of these metabolites from the capillary network to the intracellular space. The text contains information to the effect that the rate of growth of transplanted malignant neoplasms directly correlated with their ability to intercept from the host tissues, including proliferating tissues (thymus, spleen), the precursors of nucleic acids-thymidine and uridine. As a result, the level of the latter in these tissues during the very first minutes after injection of the label proves to be particularly low, much lower than normal, and their incorporation of the corresponding nucleic acids is sharply retarded. It may be assumed that similar phenomena also occur with the exogenous amino acids supplied with the food, as well as arising in the process of breakdown of tissue proteins. The assumed characteristics of the surface membranes of tumor cells may be regarded as a most important condition that determines the very possibility of the parasitic growth of the neoplasm in the host, as well as exerting systemic action on it. One more factor may favor the uptake of vital metabolites by the tumor. With regard to glucose, we may take for granted the role of the enormous concentration gradient established between arterial blood and the tissue of the tumor. It is quite possible that a similar process takes place with amino acids and the precursors of nucleic acids, since in the tumor, unlike in normal cells, the catabolism of proteins and nucleic acids is almost completely suppressed. It follows that recyclization of amino acids and nucleotides does not take place in cancer cells (Jewel1 and Hunter, 1971), and this must reduce the pool of these compounds. Moreover, it is not unlikely that the tumor utilizes the host endocrine system to accelerate the transport of glucose, amino acids, and nucleosides from the blood. Receptors for insulin were found in the plasma membranes of Zajdela ascites hepatoma (Shelepov, et a l . , 1978). We consider successful competition of the tumor with the host tissues the main form of its systemic action because it is precisely this form that is, in the final analysis, able to cause not only essential changes in the carbohydrate, nitrogen, and lipid metabolism in the host and transform the protein-synthesizing apparatus of the protein-secreting organs, but also the tension of the compensatory mechanisms, including the functions
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
14 I
of the endocrine system, which often ends in its exhaustion and even the dystrophy of various endocrine glands. Both these systemic effects of the tumor on the host (see Sections I1 and 111) seem to be interwoven. Hypoglycemia developed in those tumor hosts that were not able to sufficiently stimulate gluconeogenesis because of the elevated threshold of the liver to controlling hormones. As we also have seen, it is possible to draw a connection from t h e competitive abilities of the tumor and suppression by it of adenosine deaminase activity in the lymphoid organs to the phenomenon of immunodepression in the host. The changes in the activity of some of the enzymes whose synthesis is regulated by hormones may be the result of changes in the hormonal balance; for example, an increased production of glucocorticoids which induce the synthesis of thymidine kinase in the liver and inhibit it in the thymus and spleen. We should like to place special emphasis on the enormous prospects of using in oncological treatment adequate parenteral nutrition (Wretlind, 1972; Dudrick et al., 1977; Sudjyan, 1973a,b, Sudjyan et al., 1975, 1977) as a potent means of counteracting the impoverishment of the host tissues in vital trophic factors and the disturbance in its homeostasis, including immunodepression. It is now already clear that parenteral alimentation can serve as an essential adjuvant of surgery and chemotherapy, overcoming the toxic side effects of drugs. Regrettably, these possibilities are sometimes underestimated by oncologists. A good deal in the systemic action of tumors is not clear as yet: for example, the elevation of the sensitivity threshold of the target tissues to corresponding hormones and other factors that control their functions. This latter phenomenon involves a weakening or complete disruption of the negative feedback between the central and peripheral endocrine glands which normally safeguard their functional potential. Many so-called complications of the main disease in cancer patients require special analysis. Here pathophysiologists, immunologists, biochemists, and clinicians still have a vast and as yet unexplored field of endeavor. We shall venture to dwell briefly on two of these complications. It is possible that the frequent paradoxical combination of hemorrhages and thromboses in cancer patients is a result of deposition of fibrin in the vascular bed of such tissues as the liver, spleen, and heart valves with subsequent fibrinopenia (McKay el al., 1953; Khato et al., 1977; Goodnight, 1974). Tendency to bleeding may be associated with an increased consumption of platelets and a reduced half-life of fibrinogen (Rasche and Dietrich,
142
V . S. SHAPOT
1977), their turnover rate being four times as high as normal (Schlichter and Harker, 1974). However, at the same time the hypercoagulable state of the blood is a phenomenon most concomitant to cancer. Most patients examined with squamous cell and poorly differentiated lung cancer (Sopotzinskaya, 1978) exhibited elevated levels of blood fibrinogen. The above author found a direct correlation between the elevated concentrations of blood fibrinogen and free cortisol. In addition, excessive amounts of thromboxane following inhibition of the formation of prostacycline as a result of hyperlipidemia may be causatively involved in the phenomenon of hypercoagulability of the blood of cancer patients (see Section 11,C). As for dermatomyositis and polymyositis, there are reasons to suspect their autoimmune nature (Curth, 1974). Cytotoxic action of the lymphocytes of cancer patients with polymyositis with respect to muscle explants in culture, as well as fluorescent antibodies in the bulk of muscle fibers of these patients, has been found (Frion, 1974). Such phenomena might be rendered comprehensible by some experimental observations. Organspecific antigens of renal and muscular tissues have been found in the extracts of several rat hepatomas, and antigens of muscular tissue have been discovered in extracts of renal adenocarcinoma (Fel and Schwemberger, 1968). Johnson et al. (1965) found striation, and the presence of myosin, revealed immunologically, in the Wilms tumor-a neoplasm of renal origin. It is conceivable that malignant human neoplasms of various histogeneses are able to synthesize and eject ectopic muscle-type proteins through the plasma membrane into the blood, thereby inducing formation of corresponding antibodies, which damage the patient's muscles, the foregoing being conditioned by an anomalously functioning immune system. It follows that the emergence of an autoimmune disease in cancer patients is one more indication of impairment of the regulatory mechanisms of their immunologic supervision. An analysis of the multiformity of manifestations of the tumor's systemic effects makes it possible to contemplate the solution of the old controversy of whether a tumor should be considered a local or general disease of the organism. A benign tumor has no systemic action. If it does not, by its bulk and localization, hamper the functions of vitally important organs, as is the case, for example, with benign brain tumors, the life span of the host does not diminish. There is therefore every reason to regard benign neoplasms as a purely local disease. On the other hand, the development of even a nonmetastasizing malignant tumor impairs the activity of all physiologic systems of the host and disturbs its homeostasis. Hence it is obvious that here we have to deal
TUMOR A N D HOST MULTIFORM RELATIONSHIPS
143
with a general disease of the organism, although it is induced and maintained by local neoplastic lesions. Usually the degree of the tumor's malignancy is evaluated, especially by experimenters, by the rate of its growth. It seems to us that the following formulation is more exact: the degree of malignancy is higher the sooner the tumor causes the death of the host, i.e., the shorter its life span. Here we imply both the dissemination and systemic effects of the tumor. We assume that the systemic effects of a tumor is one of the obligatory properties of malignant neoplasms, exactly as are invasiveness and destructive growth. Unlike this, cataplasia and metastases, although characteristic of malignant neoplasms, do not manifest themselves in all cases. It is time we corrected the concept of the tendency toward resemblance of the biochemical characteristics of distant tissues of the host to those of the tumor itself. This rule is far from being always observed. However, if the directions of the changes in the biochemical characteristics of the tissues and tumor really coincide, this coincidence warrants our assumption that the above deviations are not essential to the tumor transformation in the sense that distant tissues do not in these cases acquire the basic property of a neoplasm, i.e., the capacity for uncontrolled growth. Last, there is one more conclusion pertaining primarily to experimenters. Since practically no single organ of the tumor-carrying organism functions normally, it is necessary, in studying any characteristics of a malignant neoplasm, to use, as control, tissues only of a healthy animal, and in no case those of the host. Incidentally, clinicians would also do well to take this factor into consideration. Let us recall here the aforesaid changes in the spectrum of lactic dehydrogenase isozymes in the tissues of oncologic patients not directly affected by the tumor. In conclusion, we should note the amazing feature of the systemic action of tumors, namely, its nonstereotypic character. It cannot be expected that the impairment of function of any one physiologic systems will be similar, even with the same type of tumor. The disorders may differ in the extreme and may not always be predictable.
REFERENCES Abbott, W. E . , Levey, S. V . , and Krieger, H. (1959). Mefaholistn 8, 847-861. Afrikyan, M. N. (1977). Sov. Med. 6, 48-51. Alekhina, R. P., Lichtenstein, V. A., and Shapot, V. S. (1976). Bull. E x p . B i d . Med., 11, 1354- 1357 (in Russian). Allison, A . C., Hovi, T., Watts, R. W. E., and Webster, A . D. B . (1977). It1 "Purine and
144
V . S . SHAPOT
Pyrimidine Metabolism." Ciba Found. Symp. 48, pp. 207-224, Amsterdam, Oxford, New York. Anisimov, V. N., and Ermoshenko, V. S. (1975). Vopr. Oiikol. 21, 3, 56-59. A/intr. friterri. M e ( / . (1973). 79, 597-599. Ardenne, v. M., and Rieger, F. (1972). Arch. Ge.sc~hn~ul.srfor.sch. 40, 51-79. Baker, N . , Hill, V.. and Ookhtens, M. (1978). C a n c e r Res. 38, 2372-2377. Barriere H. (1975). Aiiri. fmerti. M d . 126, 177- 181. Begg, R. W. (1958). Ad,,. Crrnc.c.r R a s . 5, 1-54. Beisel, W. R., and Rapoport, M. J. (1969). N e w Engl. J . M e t / . , 280, 54-546. Benjamin, F. (1974). Ohsret. Gy/ircol. 43, 257-261. Berken, A., and Benacerra, B. (1968). Proc. Soc. E x p . B i d . Met/. 128, 793-795. Blackburn, G. L., Maini, B. S., Bistrian, B. R., and McDermott, W. V., Jr. (1977). Ctr/icer Res. 37, No. 7, part 2, 2348-2353. Blinov, V. A. (1969). Vopr.. O/rkol. 15, 81-85 (in Russian). Blinov, V. A. (1974). "Interrelationships of Carbohydrate and Nitrogen Metabolism in the Tumor's Host," Dr. Sci. Thesis, Kiev (in Russian). Blinov, V. A., and Shapot, V. S. (1974). Bull. Exp. Riol. M e t / . 7, 52-55 (in Russian). Blinov, V. A., and Shapot, V. S. (1974a). Vopr. Onkol. 20, No. 42, 60-65 (in Russian). Blinov, V. A., and Shapot, V. S. (1974b). P(i/hol. Ph.vsiol. Exp. Ther. 4, 49-52 (in Russian). Blinov, V. A., Ananich, N. A., and Shapot, V. S. (1972). Prithophysiol. E S P . Ther. 6, 5154 (in Russian). Blinov, V. A., Rasulov, A. S., and Shapot, V. S. (1973). Bull. E x p . Biol. M c d . 6, 49-51. Blinov, V. A., Blinova, N . V., Lyubimova, N. V., and Shapot, V. S. (1975). P a t h o / . Phy.sio/. Exp. Ther. 5, 59-64 (in Russian). Borzenko, B. G., Vornovitskaya, G. I., Belousov, I. M., Gets, G., Drel, K. A., and Shapot, V. S. (1977). Biokhimici 42, 1266- 1270 (in Russian). Cahill, G. F., Jr. (1970). NCW EngI. J. Met/. 262, 668-675. Carey, R. W., Pretlow. T. C., Ezdinie, E. A., and Holland, J. F. (1966). A m . J. M e d . 40, 458-469. Carson, D. A., Kaye, J., and Seegmiller, J. E. (1977). Proc. Nritl. Accid. Sci. U . S . A . 74, 5677-5681. Chan, S . H . P., and Higgins, E., Jr. (1978). Con. J . Biocheni. 56, 111-116. Chowdburry, F., and Bleicher, S. J. (1973). Metubolism 22, 663-674. Clark, C. M., and Goodlad, C. A. J . (1975). Biochim. Biophys. Actrr, 378, 230-240. r 37, No. 7, part 2, Copeland, E . M., Daly, J. M., and Dudrick, S. J. (1977). C ~ n c e Res., 245 1-2456. Costa, G. (1973). I n "Cancer Medicine" (J. F. Holland and E. Frei, eds.), pp. 1035-1044. Lea and Febiger, Philadelphia. Costa, G. (1977). Cuticer Res. 37, 2327-2335. Curth, H . 0. (1974). Ann. N. J . A Acrid. Sci. 230, 435-442. Davidova, S. Ya., Shapot, V. S., and Drozdova, G. A. (1970). Biochim. Biophys. Acru 220, 206-212. Dedkova, E. M., and Raben, A. S. (1977). "Paraneoplastic Diseases." Moscow (in Russian). Dembinska, A., Gryglewska, T., Zmuda, A., and Gryglewski, R. J. (1977). Prostglnndins 14, 1035- 1042. Dessaint, J. P., Lefebre, J., Adenis, L., Vemeau, J. L., and Linquette, M. (1978). Anti. Endocrinol. 39, 73-74. Dilman, V. M. (1974). f t i "Mammary Cancer and Neuroendocrine Therapy" (B. A. Stoll, ed.), 197-228. Butterworth, London.
TUMOR AND HOST MULTIFORM RELATIONSHIPS
145
Dinescu-Romalo, G., Michel, C., and Vlad, L. (1977). Rev. Roum. Biochim. 14, No. 3, 161-165. Dmitrieva, L. V., Volchek, A. G., Rosen, V. B., Adler, V. V., and Shapot, V. S. (1976). Biokhirnirr 41, 1850- 1858 (in Russian). Dole, V. P. (1956). J . Clin. Invest. 35, 150-154. Dudrick, S. J., MacFadyek, B. V., Jr., Souchon, E. A., Englert, D.-A. M., and Copeland, E. M. (1977). Cancer Res. 37, No. 7, part 2, 2440-2445. Egorova, L. I. (1965). “Therapy with Glucocorticoids and ACTH.” Moscow (in Russian). Edmondson, J . H. (1966). Cancer 19, 277-280. Eisenstein, A. (1973). A m . J . Clin. Nutr. 26, 113- 120. Ekren, T., and Vartin, M. B. (1972). Biochim. Biophys. Acta 281, 263-269. Ellem, K. A. O., and Sheridan, S. W. (1964). Biochem. Biophys. Res. Commun., 16, 505511.
Ertle, N. (1973). Oncology 27, 415-429. Exton, J. H. (1972). Metabolism 21, 945-995. Farron, F. (1972). Enzyme 13, 233-237. Fel, V. Ja., and Schwemberger, I. N. (1968). “Morphological and Immunological Studies of Cytodifferentiation of Experimental Tumors.” Leningrad (in Russian). Felig, P. (1975). Annu. Rev. Biochem. 44,933-955. Feninger, L. D., Waterhouse, C., and Kentman, E. H. (1953). Cuncer 6, 930-941. Ferdinandis, J. A., Morris, H. P., and Weber, G. (1971). Cancer Res. 31, 550-556. Friedman, B., Goodman, E. H., and Weinhouse, S. (1967). J . B i d . Chem. 242, 3620-3627. Frion, G. J. (1974). Ann. N.J. Acad. Sci. 230, 23-55. Gerstein, E. S., Vornovitskaya, G. I., and Shapot, V. S. (1978). Biokhirniu 43, 1303-1310 (in Russian). Giblett, E. R., Scott, C. R., and Chen, S.-H. (1975). In “Combined Immunodeficiency Disease and Adenosine Deaminase Deficiency. A Molecular Defect” (M. J. Menwissen, R. J. Pickering, B. Pollara and I. H. Porter, eds.), pp. 103-110. New York. Giblett, E. R., Anderson, J. E., Cohen, F., Pollara, B., and Menwissen, H . J . (1975). Lancet ii, 1010- 1013. Girman, G., Pees, H., Schwarze, G., and Schenerlen, P. G. (1976). Nature (London) 259, 399-40 1. Gjesdal, K. (1976). Scand. J. Haemotol. 17, 205. Gordon, R. S., Cherkes, A,, and Gates, H.(1957). J. Clin. Invest. 34, 810-824. Gorozhanskaya, E. G., and Shapot, V. S. (1964). Doklady USSR Akad. Nauk 155,947-950 (in Russian). Gorozhanskaya, E. G., and Shapot, V. S. (1971). Vestnik USSR Akad. Nauk, 3, 28-32 (in Russian). Goodnight, S. H., Jr. (1974). Ann. N.J. Acad. Sci. 230, 271-288. Granzov, C., and Beheim, P. (1972). Eur. J . Cancer 8 , 225-230. Greengard, O., and Herzfeld, A. (1977). Cancer Res. 37, 884-891. Greengard, O., and Machovich, R. (1972). Biochim. Biophys. Acta 286, 382-388. Greenstein, J. P. (1947). “Biochemistry of Cancer.” Academic Press, New York. Gudewicz, P. W., and Saba, T. M. (1977). Br. J . Cancer 36, 670-677. Gullino, P. M., Clark, S. M., and Grantham, F. J. (1964). Cancer Res. 24, 780-798. Gullino, P. M., Grantham, F. H., and Courtney, A. (1967). Cancer Res. 27, I03 I- 1040. Harris, J., and Copeland, D. (1974). Ann. N.Y.Acad. Sci. 230, 56-85. Haven, F. L., and Bloor, W. R. (1956). A d v . Cancer Res. 4, 237-314.
146
V . S. SHAPOT
Hawrylewiez, E. D., Blair, W. H., and Giltner, L. W. (1974). Ahstr. I n / . Cancer Coriar. I / (Florence) 4, 624. Heitzman, R. J., Hibbit, E . G., and Mather, F. (1971). Eur. J. B i o c h e ~ n 21, . 411-415. Herrmann, J., and Staib, W. (1969). E w . J . Biochein. 7, 427-433. Herzfeld, A., and Greengard, 0. (1972). Criticer Res. 32, 826-832. Herzfeld, A., and Greengard, 0. (1977). Cancer Res. 37, 231-238. Holden, W. D., Krieger, H., Levey, S. and Abbott, W. E. (1957). Ann. Surg. 146, 563577. Holroyde, C. P., Gabuzds, T . G., and Putnam, R. C. (1975). Cai7cer Res. 35, 3710-3714. Holroyde, C. P., Myers, R. N., Smink, R. D.. Putnarn, R. C., Paul, P., and Reichard, G . A. (1977). Cnncrr Res. 37, 3109-3114. Horvath, I., Aranyi, P., Naray, A., Fuldes, I., and Gyuris, A. (1975). I n / . J. Ctrticer 16, 897-904. Ioannesjanz, I . A., Adler, V. V., Elkina, 2. I., Artamonova, S. I., Kadagidze, 2. G., and Shapot, V. S. (1977). Bull. Exp. B i d . M e t / . 83, 449-452 (in Russian). loannesyants, I. A., Adler, V. V., Narimov, M., Kadagidze, L. G., and Shapot, V. S. (1978). J. Steroid Biochem. 9, 649-656. Jacob, A., Labhart, A., and Troesch, E. A. (1969). I n “Diabetes: Proc. 6th Congress Int. Diabetes Fed.” (J. Ostram, ed.), pp. 898-902. Stockholm. Jaroszewicz, L., Wincewicz, A., and Rzeczycki, W. ( 1976) Nroplr/.s/nrr 23, 259-276. Jewell, W. R., and Hunter, L. (1971). Cancer Res. 31, 257-259. Johnson, W., Jurand, I., and Hiramoto, R. (1965). A m . J. Pntliol. 47, 1139- 1145. Kavetzky, R. E . (1977). “Interaction of Tumor and the Host.” Kiev (in Russian). Kelly, W. M. (1959). Dirrhetes 8, 22-28. Kendysh, 1. N. (1972). U s p . Sovrem. Biol. 74, 368-384. (in Russian). Khato, J., Sato, T., Sato, H., Abe, K., Endo, E., and Ohta, E. (1977). Gunn 68, 797-804. Khuzhamberdyev, M., Romanova, L . A., Neyfakh, E. A., and Gorkin, V. 2. (1973). Vopr. Med. Khim. 19, 415-422 (in Russian). Kigoshi, S., and Ito, R. (1973). Experienticr 29, 1408-1410. Kirikuta, Y . , Simu, H., and Toma, V. (1970). Arch. Geschiviilsrf~rsch.35, 61-112. Klastersky, J . , Danean, D., and Verhest, A. (1972). Eur. J. Crrnc.er 8, 149- 154. Knox, E. W., and Sharrna, C. (1968). E i i i y m . Biol. Clin. 9, 21-30. Kovalevic, Z., and Morris, H. P. (1972). Cmicer Res. 32, 326-333. Kralovic, B . C., Zepp, E. A., and Cenedella, R. J . (1977). Eur. J. Cancer 13, 1071-1079. Krechetova, G. D., Chudinova, I. A., and Shapot, V. S. (1972). Biochim. Bioplrys. Act(/ 227, 161-178. Kushiwagi, K., Tobe, T., Higeshi, T., and Warabioka, K. (1972). G m n 63, 57-62. Lankin, V. 2 . (1971). Binkkiniiyn, 36, 1234- 1238 (in Russian). Lankin, V. 2. (1973a). fib’. Akcrd. Nriitk SSSR 1, 47-51. (in Russian). Lankin, V. 2. (1973b). Thesis. Moscow (in Russian). Lankin, V. Z., and Neyfakh, E. A. (1973). A k t u d n . Vopr. Soivwn. 0 1 7 k d . MGU, Moscow, V.3, pp. 112- 120 (in Russian). Lankin, V. Z., Polyakov, V. M., and Gurevich, S. M. (1977). Lipirly Nlrukci Moscow, 93103. Lankin, V. Z., Tikhase, A. K., and Kotlovzeva, N. V. (1976). Krirdiologia 2, 23 (in Russian). Law, D. K., Dudrick, S. J . , and Abdou, N. I. (1973). Ann. Interii. Med. 79, 545-550. Lawson, D., Paik, W. K., Morris, H . P., and Weinhouse, S. (1977). Cancor R e s . 37, 850856.
TUMOR AND HOST MULTIFORM RELATIONSHIPS
147
Laszlo, J., Harlan, W. R., Bogdanoff, M. D., and Estes, E. H., Jr. (1960). Cliu. Res. 8, 59 (abstr.). LePage, G. A., v. Potter, R., Bush, H., Heidelberger, C. and Hurlberg, R. B. (1952). Cancer Res. 12, 153-157. Levy, M. H., and Wheelock, F. (1974). Adv. Cnncer Rev. 20, 131-163. Lichtenstein, A. V., and Shapot, V. S. (1971). Biochent. J. 125, 225-234. Lichtenstein, A. V., Alekhina, R. P., and Shapot, V. S. (1975). Bioche/n. J . 147, 447-456. Lichtenstein, A. V., Alekhina, R. P., and Shapot, V. S. (1978). Eur. J . C mcer 14, 939947. Lundholm, K. ( 1975). "Skeletal Muscle Metabolism in Cancer Disease." Elanders Boktryckeri Aktiebolag Kungsbacka, Goteborg, Sweden. Lundholm, K., Bylund, A X . , Holm, J., and Scheraten, T. ( 1976). Ew. J. Cancer 12,465473. Madajewicz, S., Haruppe, J., and Kaminsa, J. (1977). Eur. J . Cancer 13, 801-804. Mallick, L., Banergee, S. K., and Shrivastava, G. C. (1968). Br. J . Concer 22, 110-115. MacKay, W. D., Edwards, M. H., Bulbrook, R. D., and Wang, D. Y. (1971). Ltr/icef 2, 1001-1006.
Marks, L. J., Steinke, S., and Podolsky, S. (1974). Ann. N . Y . Actrd. Sci. 230, 147-160. May, N. E., and McCay, P. B. (1968). J. Biol. Chern. 243, 2288-2296. Mays, E. T. (1971a). J . Surg. Oncol. 3, 487-493. McKay, D. G., Mansell, H.,and Hertig, A. T. (1953). Concer, 6, 862-869. Megyesi, K., Kahn, C. R., Roth, J., Neville, D. M., Nissley, S. P., Hurnbell, R. E., and Froesch, E.R. (1975). J. Biol. Chem. 250, 8990-8996. Merskey, C. (1974). Ann. N . Y. Acad. Sci. 230, 289-293. Mertvetzov, N. P. (1969). Biokhiiniu 34, 381-384 (in Russian). Mider, G. B. (1951). Cancer Res. 11, 821-829. Mider, G. B. (1953). Annu. Rev. Med. 4, 187-198. Miller, F., Brooks, R., and White, J . (1969). J. Nut/. Cuncer I n s t . 42, 51-58. Mischenko, I. P. (1940). "On the Processes of the Synthesis and Analysis in Cancer's Host a s Revealed by the Data on Nitrogen Metabolism." Kharkov (in Russian). Mishineva, V. S., Goryukhina, T. A., Burova, T. M., and Seitz, I. F. (1973). Vopr. Otikol. 19, No. 9, 86-91 (in Russian). Monakhov, N. K., Neistadt, E . L., Shavlovskii, M. M., Shvartsman, A. L., and Neifakh, S. A. (1978). J . Ntrtl. Cancer Inst. 61, 27-33. Monastyrskaya, B. I. (1963). Proc. VIII Int. Crinrer Congr. 8, Moscow, Vol. 3, 452-454 (in Russian). Monastyrskaya, B. I. (1963). Proc. V I l l I n t . Concer Cortgr. 8, Moscow, Vol. 3, 452-454 (in Russian). Moncada, S . , Gryglewski, R. J., Bunting, S . , and Vane, J. R. (1976). ProstrrglNndiris 12, 7 15-735. Morgan, W. W., and Cameron, T. L. (1973). Cancer Res. 33, 441-449. Morita, Y., and Munck, A. (1964). Biochirn. Biophys. A r f u 93, 150-157. Morozkina, T. S . , and Shapot, V. S . (1976). Bull. E x p . Biol. Med. 6, 727-729 (in Russian). Morozkina, T. S., Vysotskaya, G. V., and Vysotski, A. M. (1975). Aktunl. Vopr. Onkol. Med. Radio/. Vol. 5 , pp. 351-355 (in Russian). Murgita, R. A., and Tomasi, T. B. (1975a). J. E.rp. Men. 141, 269-286. Murgita, R. A., and Tomasi, T. B. (1975b). J . Exp. Med. 141, 440-452. Mustea, I . (1971). Ortcol. Radio/. 10, 51-56. Nakamura, W., and Hosoda, S. (1968). Biochim. Biophys. Acta 158, 212-218.
148
V. S . SHAPOT
Naray, A., Aranyl, P., Folder, I., and Horvath, I. (1977). J. Nrrtl. Crincer Itist. 59, 12371241. Niebauer, G. (1974). Wein. M e d . Wschr. 124, 683-688. Nissau, S ., Bar-Moor, A., and Shafrir, E. (1968). N e w Engl. J . M e d . 278, 177- 183. Nisselbaum, J. S . (1972). Cuncer Res. 32, 2167-2171. Ogawa, K., Tominaga, K., Taoka, S., Yata, K., and Tsibura, E. (1978). Cann 69,471-475. Papaioannou, A. N. (1966). Srcrg. Gynecol. Ohster. 123, 1093-1109. "Paraneoplastic Syndromes." (1974). Ann. N . J . Acnd. Sci. 230. Pattillo, R. A., Husse, R. O., and Garantis, J . C. (1971). In Vitro 7, 59-67. Paul, D. W., Jacobs, L. S . , Danofrio, R., Burday, S . Z., and Schalch, D. S . (1974). J . Entlocrinol. Mettrhol. 38, 71-82. Pearson, C. M. (1966). Annu. R e v . Merl. 17, 63-82. Perevodchikova, N. I., Spirina, S. K., Smirnova, T. I., and Sivoshinsky, D. S. (1977). In "Experimental and Clinical Pharmacotherapy," Vol. 7, pp. 170- 191. Zinatne, Riga (in Russian). Petrov, R. V., and Khaitov, R. M. (1977). Vestnik Akatl. Med. Nmuk. SSSR, 20, 64-69 (in Russian). Pineo, G. F., Brain, M. C., Gallas, A. S . , Hirsh, J., Hatton, M. W., and Regeczi, E. (1974). A t i t i . N . Y. Act/(/. S C ~230, . 262-270. Polyakov, V. M.. Ldnkin, V. Z . , Arkhangelskaya, A. V . , and Blagorodov, S. G. (1977). Biokhiniicr 42, 499-504. Prost,qlnndins (1977). 14, 201-209. Pushkina, I. P., Krechetova, G. D., and Shapot, V . S. (1976). Biokhiniio 41, 1941-1944 (in Russian). Quan, P. C., and Burtin, P. (1978). Cancer R e s . 38, 286-296. Rasche, H., and Dietrich, M. (1977). Eur. J . Cancer 13, 1053- 1046. Reichard, P. (1968). Errr. J . Biochem. 3, 259-266. Roberts, J . , Holcenberg, J . E., and Dolowy, W. C. (1971). Lifc Sci. 10, part 2, 251-255. Rosso, R., McDonelli, M., and Franchi, G. (1971). Eur. J . Crrncer 7 , 563-577. Saez, S. (1971). Eur. J . Concer 7, 381-387. Saez, S . (1974). In "Mammary Cancer and Neuroendocrine Therapy" (B. A. Stoll, ed.), pp. 101- 122. Butterworth, London. Samundjan, E. M. (1973). "Corticosteroids and Cancer." Kiev (in Russian). Sato, K., and Tsuiki, S. (1972). Cnncer R e s . 32, 1451-1454. Savchenko, N . E., Aleksandrov, M. N., Fradkin, S. Z., Zhavrid, E. A., Mashevsky, A. A., Toropova, T. V . , Istomin, Yu.P., Rubanova, V. Z., Bezruchko, V. I., and Filatovich, L. N. (1977). A k t i i d n . V o p r . Onkol.. M e d . R o d i d . Minsk, Vol. 1, pp. 69-75 (in Russian). I ~ 239-246. P~ Schecter, B., Segal, S . , and Feldman, M. (1977). Int. J . C ~ I J 20, Schersten, T., Wahiquist, L., and Julderas, B. (1971). Canccr 27, 278-284. Schlichter, S. J.. and Harker, L.-A. (1974). A n n . N . J . Acrid. Sci. 230, 252-261. Schreck, B., Holcenberg, J . S., and Batra, K. V . (1973). Proc. Am. As.soc. Ccrncer R e s . 14, 26. Schumm, D. E., and Webb, T. E. (1975). Nirtiire (London) 256, 508-509. Schumm, D. E., Billmire, D. F., and Morris, H . P. (1976). Eur. J . Cancer 12, 689-694. Scott, H., and Goodnight, R., Jr. (1974). Ann N . J . A c d . Sci. 230, 277-288. Seegmiller, J . G., Watanabe. T., and Schreier, M. H. (1977). In: "Purine and Pyrimidine Metabolism." Ciba Found. Symp. 48, pp. 249-276. Amsterdam, Oxford, New York. Shabad, L. M. (1936). .%I,. Vrcichc4m. Zh. 15, 15-31 (in Russian). Shamberger, R. J . , Hozumi, M., and Morris, H. P. (1971). Concer R e s . 31, 1632-1639.
TUMOR AND HOST MULTIFORM RELATIONSHIPS
149
Shapot, V. S. (1968). Vest. Akad. Med. Nauk. SSSR 3, 11-22 (in Russian). Shapot, V. S. (1970). Aktuuln. Vopr. Sovrem. Onkol. MGU, Moscow 2, pp. 111-125 (in Russian). Shapot, V. S. (1972). A d v . Cuncer Res. 15, 253-286. Shapot, V. S. (1975). “Biochemical Aspects of Tumor Growth.” Medicina, Moscow. Shapot, V. S. (1975a). Ad)’.Enzyme Regul. 13, 67-75. Shapot, V . S. ( 1976). In “Oxygen Transport to Tissues-11’’ (J. Grote, D. Reneau and G. Thews, eds.), pp. 581-586. Plenum Press, New York, London. Shapot, V. S. (1978). In ”Problems of Oncology,” pp. 49-65. Moscow. (in Russian). Shapot, V. S., and Berdinskich, N. K. (1975). Vopr. Onkol. 21, 57-62 (in Russian). Shapot, V. S., and Blinov, V. A. (1974). Cuncer Res. 34, 1827- 1832. Shapot, V. S., and Blinov, V. A. (1975). Itogi Nuuki Onko VINITI 8, 150-207 (in Russian). Shapot, V. S., and Lichtenstein, A. V. (1973). Neoplmma 20, 555-557. Shapot, V. S . , Davidova, S.Y., and Drozdova, G. A. (1960). Vopr. Mrd. Khirn. 9, 102104. Shapot, V. S., Davidova, S. Ya., and Drozdova, G. A. (1963). Vopr. M r d . Khirn. 9, 102105 (in Russian). Shapot, V. S., Gorozhanskaya, E. G., and Lubimova, N. V. (1976). Biokhirniu 41, 17661772. Shapot, V. S., Morozkina, T . S., and Chumakov, V. N. (1976). Vopr. Onkol. 22, 7, 43-51 (in Russian). Shapot, V. S., Alekhina, R. P., Zabojkin, M. M., and Lichtenstein, A. V. (1977). Vest. Akad. Meti. Nauk. SSSR, 3, 64-69. Sharoukhova, K. S. (1974). “Use of Characteristics of the Hormonal Status on Therapy of Patients with Hyperplastic Diseases of Mammary Gland and Uterus.” Dr.Sci. Thesis, Moscow (in Russian). Shelepov, V. P., Davydova, S.Ya., and Shapot, V. S. (1978). Biokhirniu 43, 539-544. Sickless, E. A., Young, V. M., Greene, W. H., and Wiernik, P. H. (1973). Ann. Int. M e t / . 79, 528-531. Sivoshinsky, D. S., and Narkevich, B.Ya. (1975). Med. Rucliol. 11, 71-76 (in Russian). Sivoshinsky, D. S., Zakirkhodzha’ev, U. D., and Vittenberg, E. I. (1977). All-Union Congr. Rentgenol. Rudiol. 578-579. Smolyanskaya, A. Z., and Grinenko, G. I. (1976). Vest. Akad. Med. Nauk. SSSR 2, 78-84 (in Russian). Sopotzinskaya, E. B. (1978). In “Mechanisms of the Antitumor Resistance” (K.P. Balitzky, ed.), pp. 58- 112. Kiev (in Russian). Souchon, E. A,, Copeland, E. M., Watson, P., and Dudrock, S. J . (1975). J . Surg. Res. 18, 451-454. Spector, A. A. (1967). Cancer Res. 27, 1580- 1586. Storck, H. (1976). Mrd. Klin. 71, 365-372. Strautmans, I. K., and Schmidt, A. A. (1966). Izv. L c m . SSR Akrctl. N . 10, 97-102 (in Russian). Sudjian, A. V. (1973a). Abstr. S.vn?p. Free Co/nni. Il7t. Congr. Nutrition 10, 243. Kioto, Japan. Sudjan, A. V. (l973b). “Parenteral Nutrition in Onkokhirurgi.” Medicina, Moscow (in Russian). Sudjian, A. V., Lipatov, A. M.,and Kulaevskaya, N. Z. (1975). Vopr. Onkol. 21, 9, 46-52 (in Russian) Sudjian, A. V., Biletov, B. V.. and Buzokina, L. P. (1977). Klin. Mecl. 55, 2, 70-85 (in Russian).
I50
V. S. SHAPOT
Sukhova, T. I., Krechetova, G. I., and Shapot, V. S . (1978). Biokhimicr, 43, 1838-1844. Tannock, I. (1976). 111 "Oxygen Transport to Tissue-11'' (J. Grote, D. Reneau, G. Thews, eds.), pp. 597-604. Plenum Press, New York, London. Teras, L. E., and Isok, M. E. (1974). Vopr. Med. Khirn. 30, I (in Russian). Teras, L. E. (1979). Bull. E r p . Biol. Med. (in press). Toma, V., and Simu, G. (1973). Oncologv 27, 289-293. Turnell, R. W., Clarks, L. H., and Burton, A. A. (1973). Cancer Rev. 33, 203-212. Tyler, H . R. (1974). Ann. N . Y . Acad. Sci. 230, 348-357. Uberti, J . , Johnson, R. M., Talley, R., and Lightbody, J. J. (1976). Clrncer Res. 36, 20462047. Vaupel, P. (1974). Atemgaswechsel und Glucose-Stoffwechsel von Implantationstumoren in vivo." Mainz. Vereschagind, G. V., Zaborski, G. A., Goloskov, N. P., and Sityagin, Y. V. (1977). Sov. Med. 2, 78-82 ( in Russian). Vornovitskaya, G. I., Shapot, V. S., and Nikolskaya, T. I. (1968). Biochim. Biophys. Acttr 166, 596-599. Vornovitskaya, G. I., Alshanetskaya, A. P., Novikova, M. A., Karelina, Morozava, N. S., and Shapot, V. S . (1972). Biokhirnici 37, 940-947. Vornovitskaya, G. I., Dubinina, I. G., Gerstein, E. S. , Grekhova, N. V., and Shapot, V. S. (1979). Bull Exp. Biol. Med. 3, 267-270 (in Russian). Waldman, T. A., Broder, S., and Strober, W. (1974). Ann. N . Y . A c d . Sci. 230, 306-317. Wang, B. S . , Badger, A. M., Nimberg, R. R., Cooperband, S. R., Schmid, K., and Mannick, J . M. (1977). Cancer Res. 37, 3022-3025. Waterhouse, C., Baker, N., and Rostamitt, E. (1969). J . Lipid Res. 10, 487-494. Weber, G . (1968). Riol. Basis Med. 2, 263-307. Weber, G., Convery, H. J., Lea, M. A., and Stamrn, N. B. (1966). Science 154, 13571370. Wretlind, A. (1972). Nutr. Metrrhol. 14, Suppl., 1-57. Wu, C., and Morris, H. P. (1970). Cancer Res. 30, 2675-2684. Wu, 2. C., Roberts, E., and Bauer, J . M. (1965). Cancer Res. 25, 677-684. Wysocki, R. (1971). Schw. Med. Wuchr. 101, 475-478. Yamazaki, M., Shinoda, H., Hattori, R., and Mizuno, D. (1977). Garin 68, 513-516. Yap, S. H . , Strair, R. K., and Shafritz, D. A. (1977). Proc. Nrrtl. Acad. Sci. 74, 5398-5401. Zborovskaya, I . B . , Alekhina, R. P., Lichtenstein, A. V., and Shapot, V. S . (1978). Biohhitnicr 43, 1516-1524. (in Russian). Zmuda, A., Dernbinska-Kiec, A., Chytowski, A., and Gryglewski, R. J. (1977). Prostaglitndins 14, 1025- 1034. "
ADVANCES I N CANCER RESEARCH, VOL. 30
ROLE OF HYDRAZINE IN CARCINOGENESIS Joseph Ba16 Department of Pathological Anatomy, Semmelweis Medical University, Budapest. Hungary
I. Introduction
..........................................................
............ 11. Toxic Effect of Hydrazine ............................. ........................ 111. Hydrazine-Induced Alteration in Rat Liver
........... IV. Occurrence of Hydrazines in the Environm V. The Oncogenicity of Isonicotinylhydrazide ............................... VI. Experiments of H. Druckrey in the Production of Tumors with Hydrazine ............................... Compounds VII. Production o VIII. Hydrazine-Caused Cancer .............................................. IX. Does IN H Produce Tumors in Humans? ................................. X. Methylhydrazine Derivatives, a New Class of Cytotoxic Agents . . . . . . . . . . . . XI. Hydrazine Therapy in Hodgkin's Disease ................................ XII. Summary ............................................................. References ...........................................................
1.51
1.52 152
1.53 1.53 15.5 158 1.58 1.59 160 161 161
I62
I. Introduction
Hydrazine NH,-NH,, as a diamide, is produced when sodium hypochlorite acts in the presence of excess ammonia:
+ NaOCl = NH,CI + NaOH NH&I + NH, = NH,-NH,*HCl
NH,
The hydrochloride of hydrazine, NH,-NH,.HCl is a colorless, oily liquid, which fumes in air, is slightly more viscous than water, and has a penetrating ammoniacal odor. It boils at 113.5"C, freezes at 2"C, and can be stored for years if sealed in glass and kept in a cool, dark place. It is miscible with water, methanol, and ethanol, but is insoluble in ether, chloroform, and benzene. It is caustic to skin and mucous membranes, and is carcinogenic. It produces hydrazine hydrate with water and hydrazine salts with acids. Hydrazine reacts with aldehydes and ketones, forming hydrazones. Metal hydrazides are combinations of metals with hydrazine. Its alkylated derivative, dimethylhydrazine, together with hydrogen peroxide, nitrogen oxide, and saltpeter is used as rocket fuel. Prior to 1875 only hydrazo compounds, derivatives of hydrazine, were known. In that same year Emil Fischer suggested the name hydrazine. 1.51 Copyright @ 1979 by Academic Preps. Inc. All rights of reproduction in any form reserved ISBN n-12-m63oo
I52
JOSEPH B A L 6
Isonicotinylhydrazide or isoniazid (INH) is used in the treatment of tuberculosis. Maleic hydrazide is employed to control suckering of tobacco. Hydrazine has military application in rocket propellants, hydrazine insecticides are common environmental pollutants, and pharmacological effects of hydrazine compounds are being exploited in a group of antidepressive agents which have monoamine oxidase-inhibiting effects (Toth, 1976).
II. Toxic Effect of Hydrazine
Borissow (18941, Poduschka (l900), and Pohl (1902) established that hydrazine exerts toxic effects in experimental animals. In 1908 Underhill and Kleiner reported that hydrazine produced alterations in the liver of bitches. Following injection of doses of 0.1 gm, restlessness was observed, with augmentation of the heart beat, followed by respiratory difficulty and general paresis. Coincident with these symptoms, variable quantities of protein, bile pigment, and allantoin crystals were observed in the urine. Wells (1908) found that hydrazine seems to be almost specific to the cytoplasm of the liver parenchyma. However, surprisingly, hydrazine was not at all toxic for the myocardium, kidney, and red blood cells. After hydrazine intoxication, recovery is ,very rapid causing no anatomical and histological changes, unlike that found with other hepatic toxins. In experimental intoxication there is accumulation of fat in the periportal and midzonal regions, sparing the pericentral region of the liver cells, 24 hours after injection, with loss of glycogen. In fasting animals neither lipid nor glycogen can be found. Parenteral feeding showed that the liver can store glycogen, although its quantity after 24 hours was only 40% of the controls. Gideon Wells, professor of Pathology at the University of Chicago in 1908, published a paper on the pathohistology of hydrazine in the J . Exp. M c ~ d .10, 457. He also poisoned and then examined the tissues of a few other dogs, as well as of cats and guinea pigs. Ill. Hydrazine-Induced Alteration in Rat Liver
Much of the current knowledge of hydrazine toxicity is based on studies of the rat. Amenta and Johnston (1962)found that animals became relatively quiet within 2 hours after injection. Four hours after adminis-
ROLE OF HYDRAZINE IN CARCINOGENESIS
153
tration of hydrazine the liver was pale, and upon microscopic examination droplets of fat could be found in the periportal and midzonal regions. The lipid concentration reached maximal value at 24 hours and then rapidly decreased. Lewis and Izume (1926) showed that parenteral glucose could not be converted into liver glycogen in hydrazine-treated rabbits. Amenta and Johnston (1962) emphasized that hydrazine sulfate intoxication does not produce either liver necrosis or inflammation. IV. Occurrence of Hydrazines in the Environment
In nature hydrazine occurs in tobacco and mushrooms. In tobacco, hydrazine is present in the leaves and in the smoke. Liu and Hoffmann (1973) and Liu et al. (1974) found that tobacco smoke from a single cigarette contained 30 ng of hydrazine. Burning of amino acids or protein also liberates hydrazine. Levenberg (1964) found that the common edible mushroom, Agaricrrs bisporirs found in the United States contains a hydrazine named agaritin. The y-glutamyltransferase dissociates agaritin into L-glutamate and 4hydroxymethylphenylhydrazine. In Europe an indigenous mushroom, Agaricirs liortensis, produces a substance similar to agaritin. A wild mushroom, Gyrornitra escirlenta, contains methylhydrazine. If this mushroom is cooked in an open pot, methylhydrazine evaporates. However, if the lid is placed on the pot the methylhydrazine cannot vaporize, with the consequent effect that the mushroom becomes poisonous (List and Luft, 1968, 1969). Gowing and Leeper (1955) observed that pineapple seedlings sprinkled with ethylhydrazine flowered earlier than usual. Hydrazines find application not only in industry and horticulture but also in medical therapy. INH, though routinely used in the therapy of tuberculosis, is not used for prevention because of its tumorigenicity. Among the diseases of blood-forming organs, in the therapy of polycytemia Vera the use of phenylhydrazine has been introduced. Math6 e t al. (1963) applied methylhydrazine in the therapy of Hodgkin's disease. V. The Oncogenicity of lsonicotinylhydrazide
Domagk ef a!. (1952) in 1951 synthesized the drug Neoteben [isonicotinylhydrazide or isoniazid (INH)], a most effective remedy for tuberculosis. Balo (1959, 1965) and Balo et al. (1961) noted that following the intraperitoneal administration of Neoteben, tumors of the lung arise in
I54
JOSEPH B A L 6
mice. The authors employed INH produced by a pharmaceutical firm in Budapest. In a footnote from the paper by Juhhsz el al. (1957), Hackmann and Hecht, representing the Bayer Company, stated that Neoteben was nontumorigenic. Following a request from the author, however, Domagk provided us with Neoteben from the Bayer Company; it was proved that Bayer INH was equally tumorigenic. The establishment of the carcinogenicity of INH caused surprise in the scientific community, but was confirmed in several laboratories. Mori and Yasuno (1959) produced lung tumors in dd strain mice by oral administration of INH. Mori el al. (1960) found that in five groups of 15-20 mice each, 50%- 100% of mice treated with INH developed lung tumors, whereas in the two control groups, lung tumors were found in only 13%. The authors held that the carbamyl group was responsible for tumorigenicity. Matsumoto et al. (1960) also observed tumors after the oral administration of INH. Wolfart (1960) at the Robert Koch Clinic in Freiburg concluded on the basis of the data at that time that the tumorigenicity of INH could not be denied. On the other hand, Viallier and Casanova (1960) were not in agreement. Schwan (1961, 1962), on the contrary, supported the fact that INH produced tumors. Wagner and Moritz (1962) found that in animal experiments or in tissue culture application of small doses of INH promoted tumor growth, whereas large doses had inhibitory effects. In 1962 Weinstein and Kinoshita using BALB and C5, black mice obtained mainly lung tumors. Biancifiori et al. (1964) experimented with the CBA/Cb/Se strain of mice with the same result. Beer and Schaffner (1959) treated a 65-year-old woman for hypertension. In her last six weeks she was given 600 mg of P-phenylhydrazine and died with symptoms of hepatitis. Engbaek et al. (1965) compared respective dosages and concluded that the tumorigenic doses used in animal experiments did not surpass those commonly used in the treatment of human beings. According to Pompe (1956) lupus vulgaris seldom leads to carcinoma (only 0.5%) but in lupus treated with INH, carcinoma occurred in 4.6%. Biancifiori and Ribacchi (1962) and Biancifiori et al. (1964) endeavored to ascertain which part of the INH compound was carcinogenic. Using mice of the BALB/c strain, they administered sodium isonicotinate to one group, to another, isonicotinic acid hydrazine, and to a third group, hydrazine sulfate. In animals given isonicotinic acid hydrazide or hydrazine sulfate the development of lung tumors occurred. In animals that received sodium isonicotinate few tumors developed. On the basis of these findings it is concluded that the hydrazine moiety of isonicotinic
ROLE OF HYDRAZINE IN CARCINOGENESIS
155
acid hydrazine is primarily responsible for the turnongenic action. According to JuhAsz et al. (1967) hydrazine hydrate dissolved in physiological saline and all compounds from which hydrazine is liberated were carcinogenic. Seven and Biancifiori (1968) also emphasized the carcinogenic effect of hydrazine. Presently there are no data to confirm or deny that the administration of INH is a human cancer hazard. In Haddow’s department, Roe et al. (1965) already stressed that it is necessary to make long-term, repeated INH trials to learn whether INH therapy predisposes to lung cancer or to cancer of other organs. VI. Experiments of H. Druckrey in the Production of Tumors with Hydrazine Compounds
Nitrosamines and their derivatives are generally carcinogenic, with a wide spectrum. The first observations in this field were made by Barnes and Magee (1954) who showed that N-dimethylnitrosamines were potent poison to liver in both man and experimental animals. Moreover, oral administration to rats has resulted in liver adenomas or carcinomas (Magee and Barnes, 1956). Since nitrosamines are very simple aliphatic compounds, with many chemical variations, Druckrey et al. (1961) have used them as models for studies of the relationship between chemical structure and carcinogenicity. They scrutinized systematically the homologous series of dialkylnitrosamines and acyl alkylnitrosamides (Druckrey et af. 1964, 1966, 1967a), as well as many derivatives of N-nitroso compounds, among them the dimethyl- and diethylhydrazine. Among the more than 65 compounds examined, they found that nearly all symmetrically substituted dialkylnitrosamines have produced carcinomas of the liver (Schmahl et a l . 1960). Nonsymmetrical dialkylnitrosamines selectively induced tumors of the esophagus (Druckrey et al. 1964). After subcutaneous injection of cyclic nitrosamines, tumors of the nasal cavity (ethmoturbinalia) have been encountered. Methylnitrosourea was effective in selectively inducing brain tumors in rats (Druckrey and Preussmann, 1964), and treatment with N-methyl, N-nitrosourea has resulted in malignant growth of the brain and the spinal cord (Druckrey and Preussmann, 1965; Thomas, 1965; Thomas and Kersting, 1964). The selective carcinogenic effects of different nitroso compounds provoking tumors in specific organs provided Druckrey et al. (1967a) with the impetus to construct the so-called “diazoalkane theory.” According to this theory, the different alkyl compounds of nitrosamines were transformed to the respective azoalkanes. This proceeded in different organs
156
JOSEPH B A L 6
by the action of organ-specific hydroxylases forming the respective azoxyalkanes. This important step would actually determine organotropy. The proposed biochemical mechanism was supported by the observed carcinogenic effects of azoethane and azoxyethane (Druckrey et a l . , 1965). Assuming that all aliphatic azoderivatives would be dealkylated in higher organisms through oxidation, producing diazoalkanes, the carcinogenic effect of dialkylhydrazine seemed to be evident. The following mechanism was proposed in the degradation of alkylhydrazines: H n
Qli
R-CH,-N
H-N H
4
symmetrical dialkylhydrazine
-R'
I\
lH
specific hydroxylase
1/ d H
a a
R-CH,-NH-NH--C
-R'
I
h ydroxydialkylhydrazine
\ oxidation
-R'
R--CH,-NAN-C
I
OH -H,O
diazoalkane
R-CH,-N=N-CH,-R'
J
oxidation
reduction
= N-CH,-R'
R-CH,-N
\
N-N+-CH,--RR'
hydroxydiazoalkane
\
diazoxyalkane
/
i
heterolysis
+ C+H,-RR'
+ N,
end products
From the completed first step onward all following reactions through the alkylating diazoalkane occur spontaneously. The initial a-hydroxylation is considered, thus, the clue for the observed remarkable organotropic effects. In view of the use of 1,2-dialkylhydrazines as rocket fuels, Druckrey et a l . (1965) started their comprehensive experimental work on hydrazines with 1,2-diethyIhydrazine and 1,2-dimethyIhydrazine. Druckrey et
ROLE OF HYDRAZINE IN CARCINOGENESIS
157
al. (1966) performed the following experiment. Forty-five rats of the BD strain, in 3 groups, obtained weekly 1,2-diethylhydrazine, given subcutaneously in dosages of 100, 50, and 25 mg/kg body weight, respectively, in neutralized aqueous solution. After 126, 191, and 296 days, 43 of the total of 45, animals developed tumors. There were 8 reticulosarcomas, 5 stem cell and blast cell leukemias, 15 esthesioneuromas, ethmoturbinalias, 6 gliomas or gliomas with mesodermal mixed tumors of the brain, 2 adventitial sarcomas of the brain, 1 malignant neuronoma of the cauda equina, 1 subcutaneous sarcoma, and 15 adenocarcinomas of the breast out of 20 female rats. The latter corresponded closely to human breast cancer and could be transplanted as reticulosis and leukemia to animals of the same strain. These results showed remarkable similarities with the carcinogenic effects of azoethane. All the above tumors, except for one did not occur at the subcutaneous site of injection but in remote organs. From this fact it was clear that the carcinogenic effect of diethylhydrazine is exerted not directly but is due rather to a transformed product which arises during its metabolic activation. Druckrey et af. (1967) carried out further experiments with 1,2-dirnethylhydrazine using rats of BD strain, injecting 21 mg/kg, in water (pH 6.5) subcutaneously. Following a mean induction period of 184 ? 30 days, all animals died with multiple tumors of the intestine, as follows: duodenum, 6; small intestine, 5; colon, 10; rectum, 8. The histological diagnosis was adenocarcinoma, which infiltrated the whole thickness of the aforementioned organs. In these experiments, Druckrey et af. (1967) selectively induced intestinal cancer in rats using dimethylhydrazine. It was concluded that neither 1,2-diethylhydrazine nor 1,2-dimethylhydrazine have direct carcinogenic effects, but require metabolic activation through organ-specific hydroxylases as do the nitroso compounds. Finally, a conversion by oxidative dealkylation to diazoalkanes occurs. The organotropy of the hydrazines was further confirmed by circumstantial evidence presented by Druckrey et af. (1967a). They discovered that the symmetrical 1 ,2-dimethylhydrazine (CH,-NH-NH-CH, 1 was one hundred times more carcinogenic, producing tumors selectively in the colon, than the nonsymmetrical analog, 1 , I-N-dimethylhydrazine
158
JOSEPH B A L 6
VII. Production of Polyps and Tumors in the Intestinal Canal
Introduction of dimethylbenzanthracene or benzpyrene into the intestine rectally does not cause polyps or tumors in the bowel. Many years ago pathologists investigated the genesis of papillomas and adenomata of the bowel. Dukes (1947) studied the origin of rectal adenomata and papillomas. Cole and McKalen (1963) attempted to produce polyps of the large intestine experimentally. Bertalanffy ( 1960) examined the occurrence of mitoses in the epithelium of the alimentary canal of rats, and Bertalanffy and Nagy (1958) observed mitoses in human duodenum. In rats, Druckrey et al. (1966) brought about the development of intestinal cancer with 1,2-dimethylhydrazine. In similar investigations by Wiebecke et a l . (1969) with mice and rats subcutaneous injections of dimethylhydrazine produced adenomatous polyps and adenocarcinoma in the intestinal canal. With weekly subcutaneous hydrazine administration, Thurnherr et u1. (1973) produced hyperplasia, polyps, and adenoma of the epithelium of mucous membrane. CFI strain mice received weekly 20 mg/kg of 1,2dimethylhydrazine. In the early stage, focal hyperplasia was observed in the crypts of glands. Troncale et ul. (1971) and Lipkin et a / . (1963) observed the regeneration of rectal epithelial cells. Recently, Tan et al. (1976) reviewed methods of selective in t i f r o cultivation of adenocarcinomas. Thus, subcutaneous administration of hydrazine derivatives produces polyps and tumors of the intestinal canal.
VIII. Hydrazine-Caused Cancer
Cancer caused by hydrazine differs considerably from other cancers in that it does not develop at the site of administration but is absorbed from the circulation, and its carcinogenic effect is exerted in the organs where its metabolism occurs. Targets frequently are the cells of intestinal mucosa, lung, or parenchymal organs. On occasion, tumors arise simultaneously with or following another tumor either in the same or other organs. Schmahl et al. (1976) administered 30 mg/kg I ,2-dimethylhydrazine (DMH) to newborn Sprague-Dawley rats monthly beginning from the second day after birth through 10 months. After 330 days, 72% of the animals developed adenocarcinomas of the large intestine, 17% developed squamous cell carcinomas of the external auditory canal, 13% had adenocarcinomas of the kidney, and I I % had liver carcinomas. Simultaneous administration of immune-depressant or enzyme-activating
ROLE OF HYDRAZINE IN CARCINOGENESIS
159
agents did not influence the development of these different types of cancer. However, on a vegetarian diet the occurrence of liver and kidney tumors diminished significantly and cancers of the intestine and ear duct increased. Druckrey et al. (1966, 1967) stated that DMH selectively produces intestinal cancer. Schmahl et al. (1976) did not agree, however. Their difference of opinion may be explained by the fact that the experiments of Druckrey et al. (1966, 1967) were carried out on 3-month-old animals, whereas Schmahl et al. (1976) used newborn rats. The application of different hydrazine compounds, the age of animals, as well as their state of nutrition, and probably other factors influence differences in the target organ. Druckrey (1966, 1967) and his co-workers found in a comparison of I ,2-diethylhydrazine and I ,2-dimethylhydrazine that this small difference in the structure of the hydrazines brings about considerable disparity in the sites and structures of the tumors. Toth and Wilson (1971) succeeded in producing tumors of the blood vessels in Swiss mice. If they applied symmetrical dimethylhydrazine within 7 days after birth, blood vessel tumors occurred in 95% of the females and 92% of the males, whereas occurrence in control females was only 3% and in males 1%. The blood vessel tumors were mostly angiosarcomas localized in the muscles, perirenal fat, liver, and pararenal, paradidymal tissue. On the other hand, nonsymmetrical dimethylhydrazines yielded mainly lung tumors. The investigation of Reddy et al. (1974) on germ-free rodents indicated that the intestinal microflora plays a modifying role in colon carcinogenesis not only by liberating an active metabolite but also by supplying promoters or accelerators to act on colon mucosa.
IX. Does INH Produce Tumors in Humans? At present it is not yet decided whether the administration of INH to patients with pulmonary tuberculosis increases their risk of developing cancer. Every tuberculous patient who dies is by no means examined at necropsy and no systemic comparison has been made of the incidence of cancer in tuberculous patients treated with INH. Peacock and Peacock (1966) concluded that INH is an acceptable risk when used as a curative drug for tuberculosis, but should not be used prophylactically in healthy infants. It should be regarded as a potential carcinogen, and it would be prudent to assume there is some degree of risk involved in its use. I t
160
JOSEPH B A L 6
should be used only for therapy and, in any case, should be limited, as far as possible, in the duration of its application. X. Methylhydrazine Derivatives, a New Class of Cytotoxic Agents
Zeller rt ul. (1963) when testing a series of hydrazines for another purpose found that I-methyl-2-benzylhydrazine had a pronounced tumorinhibiting effect. Screening of several hundred compounds revealed some forty to be efficient tumor inhibitors, among which I-methyl-2-p-(isopropylcarbamoy1)benzylhydrazine hydrochloride and I-methyl-2-allophanoylhydrobromide were chosen for extended biological and clinical trials. The tumor-inhibiting effects of this new class of cytostatic methyl derivatives are as follows. The growth of the Ehrlich carcinoma in solid ascitic form, the Crocker-sarcoma 256, and the epithelioma of the uterus T 8 are definitely inhibited. I-Methyl-2-p-(isopropylcarbamoyl)benzylhydrazine hydrochloride (I) and 1-methyl-2-p-allophanoylbenzylhydrazine hydrobromide (11) distinguish themselves especially through their strong cytostatic activity. CH, ONH-CH-NCI-HCI CHa
CHa-N H-N H--CH, I
The cytostatically effective methylhydrazine derivative Natulan is described in a series of chemical, experimental, and clinical papers. The metabolism of this tumor-inhibiting methylhydrazine derivative proceeds according to a similar pattern in man, dog, and rat. Initially, a very rapid oxidation of the hydrazine group occurs with the formation of an azo compound, then cleavage and further oxidative degradation follows. The major portion of the drug is excreted in the urine as N-isopropylterephthalamic acid. Rutishauser and Bollag (1967) in their earlier experiments found that methylbenzylhydrazine (MBH) hinders mitosis: later they proposed that this procarbazine impedes the synthesis of DNA.
ROLE OF HYDRAZINE IN CARCINOGENESIS
161
XI. Hydrazine Therapy in Hodgkin’s Disease
According to Math6 et al. (1963) l-methyl-2-p-(isopropylcarbamoyl) benzylhydrazine chlorhydrate, a radiomimetic agent, has proved of considerable value in the treatment of Hodgkin’s disease. This drug was administered to 22 patients and resulted in 7 apparently complete remissions and 8 incomplete remissions, partial failure in 5 patients, and complete failure of disease which had been treated before and was resistant to radiotherapy, alkylating agents, or vinblastine and also in patients in early stages of the disease which have not been treated previously. Some incomplete remissions were obtained in cases of reticulosarcoma and lymphosarcoma.
XII. Summary
The author and his pupils in 1957, and the author later, in 1962, in Perugia (Ba16, 1965)established that isoniazid (INH), a compound used in the treatment of tuberculosis, stimulates the growth of tumors. Juhasz et al. (1966)confirmed this in 1967 and subsequent studies found that INH produces tumors in many species of animals. Roe ef al. (1965) stressed the necessity of ascertaining whether INH therapy may also induce tumors in human beings. On the other hand, many researchers are currently working to find, among the great variety of hydrazine compounds, some that possess therapeutic effects against cancer, including a study on ethylhydrazine being carried out by the Hoffmann La Roche Co. Cancer caused by hydrazine differs from other cancers, such as those caused by tar, which develop at the site of application. Cancer due to hydrazine does not develop at the site of application but rather in the organs in which metabolism occurs. Frequent targets are intestinal mucosa, lungs, parenchymal organs, or even the nervous system. Following development of cancer in one organ, tumors may occur in different organs Weitzel ef al. (1963).Oswald and Kriiger (1969)point out that cancer induction by hydrazines injected into the subcutaneous tissue is influenced by its absorption and modification by local tissues, by age, and by the nutritional state. Toth and Wilson (1971) observed that both symmetry and nonsymmetry of hydrazines influence the development of hydrazine tumors. The hydrazine tumor cannot be attributed to the effect of hydrazine alone or its derivatives, but it is the result of the mutual effects of hydrazine and its products of metabolism. From a variety of studies the conclusion is inescapable that all hydrazines are carcinogenic with the
162
JOSEPH BALO
possible exception of methylhydrazine (Natulan) (Roe r t ul., 1967; Toth, 1975; d’Allesandri et al., 1963). However the studies of Gold (1966, 1973) who observed that hydrazine sulfate and various hydrazides inhibit growth of Walker 256 intravascular carcinoma, B- 16 melanoma, Murphy-Sturm lymphosarcoma, and L- I2 10 solid leukemia, suggest that hydrazines may find a place in tumor therapy (Issekutz, 1968).
REFERENCES Amenta, J. S., and Johnston, E. H. (1962). Lab. Inivst. 11, 956-962. Balo, J . (1959). “Berliner Symposium uber Fragen der Carcinogenese.” Akad. Verlag, Berlin. Ba16, J. (1962). Miinch. M i d . Wschr. 104, 1424-1428. Balo, J. (1965). Qrrrrdr. Cot$ Cancer 3, 623-636. Univ. Perugia, Italy. Balo, J., Juhasz, J., and Kendrey, G. (1961). Z . Kri~bsfbrsch.59, 561-567. Barnes, J . M . , and Magee, P. N. (1954). Br. J . fndrrstr. Med. 11, 167. Beer, D. T.. and Schaffner, F. (1959). J . A m . Med. Assoc. 171, 887-889. Bertalanffy, F. D. (1960). Acta Anat. 40, 130- 148. Bertalanffy, F. D., and Nagy, K. (1958). Anat. Rec. 130, 271-272. Biancifiori, C., and Ribacchi, R. (1962). Narwe (London) 194, 488-489. Biancifiori, C.. Bucciarelli, E., Clayson, D. B., and Santilli, F. E. (1964). Br. J . Cancer 18, 543-550. Borissow, P. (1894). Z. Pliysiol. Chrm. 19, 499-510. Cole, J . W., and McKalen, A. (1963). Cancer 16, 998-1002. d’Alessandri, A., Keel, H. J., Bollag, W., and Martz. G. (1963). Sc/iw. Mrd. Wschr. 93, 1018-1024. Domagk. G.. Offe, R. A., and Siefken. W. (1952). Dtscli. Med. Wschr. 77, 573-578. Druckrey, H., and Preussmann, R . ( I 964). Nntrrrioissi.nsc/iqfifn 51, 144. Druckrey, H . , and Preussmann, R. (1965). Z. KrebsJi,rsch. 66, 389. Druckrey. H., Ivankovic, S . , Mennel, H. D., and Preussmann. R. (1964). Z. Krrbsforscli. 66, 138- 150. Druckrey, H., Preussmann, R.,Schmahl, D., and Muller, N . (I961 ). Naturii,issensc/iurffPn 48, 134-135. Druckrey, H., Preussmann, R., Matzkies, F., and Ivankovic, S. (1966). NatrrrM~issi.tisr./iafL ten 53, 557-558. Druckrey, H., Preussmann, R., Matzkies, F.. and Ivankovic, S. ( I 967). Natririi,issc,nsL./ia/~ I P N 54, 285-286. Druckrey. H . . Preussmann, R.. Ivankovic, S., and Schmahl, D. (l967a). Z. Kreh.~fhrsc/i. 69, 103-201. Druckrey, H., Preussmann, R., Ivankovic, S., Schmidr, C. H., So, B. T., and Thomas, C. (1965). Z . Krebsforsch. 67, 31-45. Dukes, C. E. (1947). Proc. R . Soc. Med. 40, 829-830. Engbaeck, H. C., Bentzon, M. W.,Heegaard. H., and Christensen, 0. (1965). Acta Path. Microbiol. S c a d . 65, 69-83. Gold, J. (1966). Cancer Res. 26, 695-705. Gold, J . (1973). Oncologv 27, 69-80.
ROLE O F HYDRAZINE IN CARCINOGENESIS
163
Gowing, D. P., and Leeper, R. W. (1955). Science 122, 1267. Issekutr, B., Sr. (1968). Onco/ogy 22, 173-184. Juhasz, J., Ba16, J., and Kendrey, G. (1957). Z. Krehsforsch. 62, 188-197. JuhBsz, J., Balo, J., and Szende, B. (1966). Nciture (Loridon) 210, 1377. Juhasz, J., Balo, J., and Szende, B. (1967). Z. Krehsforsch. 70, 150-156. Levenberg, B. (1964). J. B i d . Chein. 239, 2267-2273. Lewis, H. B., and Izume, S. (1926). J . B i d . Chern. 71, 33. Lipkin, M., Bell, B., and Sherlock, P. (1963). J . Clin. Invest. 42, 767-776. List, P. H., and Luft, P. (1968). Arch. Phurm. 301, 294-308. List, P. H., and Luft, P. (1969). Arch. Pharin. 302, 143-146. Liu, Y.Y., and Hoffmann, D. (1973). A n d . Chem. 45, 2270. Liu, Y. Y., Schmeltz, I., and Hoffmann, D. (1974). Anal. Chein. 46, 885-889. Magee, P. N., and Barnes, J. M. (1956). Br. J . Cancer 10, 114. Matht, D., Schweisguth, O., Schneider, M., Amiel, J . L., Berumen, Z., Brule, G., Cattan, A., and Schwarzenberg, L. (1963). Lnncet 2, 1077-1080. Matsumoto, K., Mori, K., and Jasuno, A. (1960). Gann 51, 91. Mori, K., and Jasuno, A. (1959). Gann 50, 107- 110. Mori, K., Jasuno, A., and Matsumoto, K. (1960). Gan/i 51, 83-89. Oswald, H., and Kriiger, F. W. (1969). Arzneimittel Forsch. 19, 1891-1892. Peacock, A., and Peacock, P. R. (1966). Br. J . Cancer 20, 307-325. Poduschka, R. (1900). Arch. Exp. Puthol. Phunn. 44, 59-67. Pohl, J. (1902). Arch. Exp. Pathol. Pharm. 48, 367-375. Pompe, K. (1956). Derin. Wschr. 133, 105-108. Reddy, B. S . , Weisberger, J . N., Narisawa, T., and Wynder, E. E. (1974). Cancer R r s . 34, 2368-2372. Roe, F. T . C., Boyland, E., and Haddow, A. (1965). Br. Med. J . 1, 1550. Roe, F. T. C., Grant, G. A., and Milican, D. M. (1967). Ncrfure (Lo,idon) 216, 375-376. Rutishauser, A., and Bollag, W. (1967). Experientiu 23, 222-223. Schmahl, D., Preussmann, R., and Hamperl, H. (1960). Naturwis.senschrrfte/i 47, 89- 196. Schmahl, D., Danisman, A., Habs, M., and Diehl, B. (1976). Z. Krehsforsch. Klin. Onkol. 86, 89-94. Schwan, S. (1961). Prrthol. Pol. 12, 53. Schwan, S. (1962). Pfithol. Pol. 13, 185. Severi, L., and Biancifiori, C. (1968). J . N u t / . Caricer Inst. 41, 331-340. Tan, M. H., Holyoke, E. D., and Goldrosen, M. H. (1976). J. Nor/. Cancer Inst. 56, 871873. Thomas, C. (1965). Z. Krehsforsch. 67, 1-30. Thomas, C., and Kersting, G. (1964). Nuturwissenschajien 51, 144- 145. Thurnherr, N., Deschner, E. E., Stonehill, E. H., and Lipkin, M. (1973). Cuncer Res. 33, 940- 945. Toth, B. (1975). Cancer Res. 35, 3693-3697. Toth, B. (1976). Cancer R e s . 36, 917-921. Toth, B., and Wilson, R. B. (1971). A m . J. Puth. 64, 585-600. Troncale, F., Hertz, R., and Lipkin, M. (1971). Cancer Res. 31, 463-467. Underhill, F., and Kleiner, I. S. (1908). J . B i d . Chern. 4, 165. Viallier, J., and Casanova, F. (1960). C . R . S . B i d . (Paris) 154, 985-987. Wagner, H., and Moritz, R. (1962). Arch. Geschwulstforsch. 19, 123- 129. Weinstein, H. J., and Kinosita, R. (1962). J. Lab. Clirz. Med. 60, 1025. Weitzel, G., Schneider, F., Fretzdorf, A. M., Seynsche, K., and Finger, H. (1963). Z. Physiol. Chern. 334, 1-25.
164
JOSEPH BAL6
Wells, H. G . (1908). J . B i d . Chem. 4, 165. Wells, H . G . (1908). J . E.rp. Med. 10, 457-464. and , Eder, M. (1969). Z. Ges. Exp. Meci. 149, 277Wiebecke, B., Lohrs, V., Girnmy, .I. 278. Wolfart, W. (1960). Dtsch. Mecl. Wschr. 85, 1655- 1657. Zeller, P., Gutmann. H . , Hegedus, B., Kaiser, A . , Langemann, A., and Muller, M. (1963). Experientiri 19, 129.
ADVANCES IN CANCER RESEARCH. VOL. 3 0
EXPERIMENTAL INTESTINAL CANCER RESEARCH WITH SPECIAL REFERENCE TO HUMAN PATHOLOGY
Kazymir M. Pozharisski, Alexei J . Likhachev, Valeri F. Klimashevski, and J a c o b D. Shaposhnikov Laboratory of Experimental Tumors, N. N. Petrov Research Institute of Oncology, USSR Ministry of Public Health, Leningrad, USSR
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Experimental Models of Intestinal Tumors
B. Transplantable Tumors
....................
..
B. Electron Microscopic Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
........... IV. Factors Modifying Intestinal Carcinogenesis . A . Endogenous Factors .......................................
..............
166 i66 166 169 169 169
174 177 179 184 184 190
s and during Carcin-
V. The Kinetics of Intestinal E ogenesis. . . . . . . . . . . . . . . . . A. Disturbances in Prolifer genesis . . . . . . . . . . . . . . . .
..................
196
ells during Carcino-
...................
C. Stem Epithelial Cells of the Intestines and Their Role in Carcinogenesis . . VI. Biochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Nucleic Acids, Nucleotides, and and Metabolism . . . . . . . . . . . . . . .......... B. Nuclear Proteins . . . . . . . . . . . . . . . . C. Enzymes . . . . . . . . . . . . . . . . . . ................................ ......... VII. Immunology . . . . . . . . . . . . . . . . . . . . . . . . . A. Antigens in Chemical-Induced T B. Antigens in Tumors of the Small Intestine C. Antigens in Nonspecific Lesions of the Col D. Antigens in Intestinal Carcinoma and Estroge E. Cell-Mediated Immunity . . . . . . . . . . . . . . . . . . VIII. Metabolism of 1,2-Dimethylhydrazine and Relate A. Experimental Evidence on DMH Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . B. The Effect of Certain Modifying Factors on DMH Metabolism . IX. Interaction of 1,2-Dimethylhydrazine and Related Compounds with Cell Com....................... ponents . . . . . . . . . . . .. . . . . . . X . Conclusion . . . . . . . . ........................... References . . . . . . . . . . . .
196 20 1 204 206 206 207 208 21 1 21 I 212 213 213 214 216 216 22 I 222 226 227
I65 Copyrighf @ I979 by Academic Press. Inc. All rights of reproduction in any form reserved ISBN o- I?-~XMW-O
I66
KAZYMIR M. POZHARISSKI ET AL.
I. Introduction
The morbidity of malignant tumors of the colon and rectum leads the cancer patient registries in many countries. Colonic and rectal tumors rank second, after lung cancer, in the death toll of malignant disease in these same countries (37,71,I12, If 9 ). Moreover, there is a tendency toward an increased frequency of colonic cancer incidence (155). At the same time, the results of therapy of patients with cancer of this site do not show any noticeable improvement (37 ), for which insufficient knowledge of the pathogenesis of colonic cancer is partly responsible. This situation calls for carrying out experimental studies of the pathogenesis of tumor growth. The solution to this problem would facilitate the elaboration of measures for prevention and early diagnosis of intestinal tumors. Adequate experimental models of intestinal tumors were developed in the 1960s (78,150,287),which laid the methodological foundation to an extensive study of intestinal cancer. Numerous reports on different aspects of this problem have been published since that time. The aim of t k present review is to systematize the data available in the literature on experimental intestinal neoplasms and to compare them with the findings on colonic and rectal tumors in man. Survey of the literature relevant to this review ended in May 1978. The following abbreviations have been used: AM, azomethane; AOM, DMH, 172-dimeazoxymethane; DMBA, 3,2-dimethyl-4-aminobiphenyl; thylhydrazine; MAM, methylazoxymethanol: MNNG, N-methy1-N'nitro-N-nitrosoguanidine;MC, methylcholanthrene; MNU, methyl-Nmethyl-N-nitrosourea, DMNA, N-dimethyl-N-nitrosoamine.
II. Experimental Models of Intestinal Tumors
A. METHODSOF TUMOR INDUCTION Numerous attempts to induce tumors of the small and, particularly, large intestine of experimental animals met with failure for a long time (226,228). Little success, if any, was achieved in the induction of intestinal neoplasms by exposure to such carcinogenic substances as polycyclic hydrocarbons' (I28, 194 1, 2-acetylaminofluorene and its derivatives (19,51,195,196,348),and benzidine and some of its derivatives (12,283). Better results were obtained with 4-aminobiphenyl and its derivatives
'
The only exception seems to have been some strains of hamsters, in which oral administration of methylcholanthrene induced tumors of the small and large intestine at a high incidence (127).
EXPERIMENTAL INTESTINAL CANCER RESEARCH
167
(3, 2’-dimethyl-4-aminobiphenyl:3, 3’-dimethyl-4-aminobiphenyl:3methyl-Caminobiphenyl). Subcutaneous injection of these substances invariably induced multiple adenomatous polyps and invasive carcinomas of rat colon (318,319). 3,2’-Dirnethyl-4-aminobiphenylhas found particularly wide application in the induction of colon tumors and in studies of different aspects of colonic carcinogenesis (45,46,141,205,284-287). Nearly all the animals fed bracken fern, Pteridium uqiiilinum, developed multiple adenomatous polyps and adenocarcinomas of the small intestine (85,122,123,219). It is suggested that bracken fern may also promote the development of gastrointestinal tumors in man (86 ). Some nitrosamines can also induce intestinal tumors. For example, a single intravenous, intraperitoneal, or oral administration of ethyl- or methylnitrosourea and -nitrosourethane is sometimes followed by the formation of intestinal tumors in rats (79,154,274 ). The intrarectal instillations with methylnitrosourea induced colon tumors in mice, rats, and guinea pigs (202-204,250 ). Narisawa et ul. (201 ) induced tumors in the distal part of the colon in rats by rectal instillations with nitrosamide-Nmethyl-N I-nitro-N-nitrosoguanidine (MNNG). Since that time this tumor model has found wide application in different laboratories (138,139, 170,199,200,218,250,256,281,282,334). When MNNG and N-ethyl-N nitro-N-nitrosoguanidinewere administered with drinking water to mice, rats, and dogs, tumors of the duodenum and carcinomas of the glandular stomach often developed (173,174,177,178 ). Among other nitroso compounds capable of causing intestinal carcinogenesis, mention should be made of nitrosobutylurea. A single intraperitoneal administration of this agent, under optimal conditions (dosage and the age of animals), was shown to induce predominantly tumors of the small intestine in mice (324). Finally, the selective induction of intestinal neoplasms in rats by means of a single intraperitoneal injection of methyl (acetoxymethyl) nitrosamine was reported quite recently. An exposure to this substance was followed by the predominant formation of adenocarcinoma of the jejunum and ileum and, much less frequently, tumors of the colon (137). It is suggested that this carcinogenic substance exerts a direct action (322 ). Although the susceptibility of humans to the carcinogenic effect of nitroso compounds has not been proved, such possibility should not be rejected (317), and intestinal cancer does not seem to be an exception in this respect, Etiologically, it is suspected that the so-called “spontaneous” tumors in sheep are caused by nitrosamines contained in fish feed (106). Varghese et al. (315) reported that they isolated from human feces an Nnitroso compound, which is a potential colon carcinogen. Laqueur (I50) demonstrated that cycasin, found in Cycadeceae plants, I-
168
KAZYMIR M . POZHARISSKI ET AL.
and some products of its metabolism (methylazoxymethanol) may sometimes induce adenomatous polyps and colonic carcinoma and tumors at other sites. The analysis of the chemical structure and possible metabolism of cycasin led Druckrey to a suggestion that 1,2-dimethylhydrazine (DMH) can also induce intestinal tumors. Indeed, the experiments conducted by Druckrey ct ( I / . (78) demonstrated that parenteral (subcutaneous) and oral administration of DMH induces tumors in different segments of rat intestines at a high rate of incidence. Subsequently, it was shown that DMH is also capable of inducing intestinal neoplasms in mice,2 Syrian and European hamsters (217,260, 312,33Y,344). At present, the intestinal tumors induced by DMH and homologous compounds (azoxymethane and methylazoxymethanol) are the most popular experimental model used in studies of different aspects of oncology (2,9,17,22,6I,98,152,l76,226,228,289 ). All these authors stress that most of the properties of these tumors are very close to those of human neoplasms of the colon and rectum. Among other hydrazo- and azoxycompounds, capable of inducing intestinal tumors in experimental animals (rats, mice, and hamsters), mention should be made of 1-methyl-2-butylhydrazineand methylazoxybutane (75 ) as well as I , I-dirnethylhydrazine, methylhydrazine, and trimethylhydrazine (305 ). Since 1,l-dimethylhydrazine occurs in tobacco and methylhydrazine in the edible mushroom, Gyrontirrn escrrlmtcr, hydrazines may be suspected to be an etiological factor of colon carcinogenesis in humans (30.5 ). Newberne and Rogers (208 ) reported the induction of colon carcinoma in rats which were kept on a vitamin A-deficient diet and fed aflatoxin B,. In another study, the colon tumor incidence in aflatoxin B,-fed rats ranged from 9% to 4095, with the mean incidence being 20% (323). Considering the above data on the development of colon tumors in persons occupationally exposed to pure aflatoxin, it seems pertinent to inquire if this substance was responsible for these cases of malignant disease (57). Some investigators succeeded in inducing predominantly tumors of the small intestine in different animal species by an exposure to whole-body and local ionizing irradiation as well as oral administration of various isotopes (162,166 1. Recent years have witnessed a fresh interest in this experimental model, which is being employed in the immunological and biochemical studies of tumors of the small intestine (2Y6-298 ). Hirose et Apart from adenomatous polyps and adenocarcinoma of different intestinal segments, DMH can induce squamous cell carcinoma of the anus in mice (6Y,111.303,312).
EXPERIMENTAL INTESTINAL CANCER RESEARCH
169
a / . ( I 24 ) induced rectal tumors with a high incidence in mice by exposing their pelvis to X rays. Similarly, there is evidence on the development of colonic and rectal tumors in humans, following an exposure of the pelvis to radiation (54,248). Rogalski et a / . (262) injected newborn Wistar rats with cell-free filtrates of human desmoid fibromas which arose in the postoperative scar two years after total colectomy performed because of familial diffuse polyposis of the colon. Adenocarcinomas of the small intestine and cecum developed in 4 out of 18 rats. In the authors' opinion, these tumors were caused by a virus responsible for the human neoplasms. Of certain interest, in this connection, are the data reported by Spiegelman's team (53). These researchers discovered 70 S RNA and reverse transcnptase in the particles with a density of 1.16-1.17 gm/ml in 70% of colonic tumors and all rectal carcinomas simultaneously. These parameters correspond to some of the diagnostic characteristics of RNA tumor viruses. These features found in the neoplasms have not been observed in normal mucosa. B. TRANSPLANTABLE TUMORS Progress made in experimental tumor induction led to the development of transplantable strains of intestinal tumors of mice and rats. Primary tumors were induced by exposure to DMH (84,173,180,326 ), MNNG (173 ), methylnitrosourethane and methylnitrosourea (50 ). Transplantable tumors are used currently in studies of the ultrastructure of tumor cells (224 ), in biochemical and immunological investigations (172,246 ), as well as for the development of the techniques of chemo- and immunotherapy of tumors (5O,72,84,I75,292). Ill. Morphology and Morphogenesis of Experimental Intestinal Tumors
A. GROSSAPPEARANCE A N D MICROSCOPIC STRUCTURE Most experimental intestinal neoplasms are generally multiple. They arise in different segments of the intestines; nearly half of them develop in the duodenum and proximal part of the small intestine. Tumors form as frequently in the ascending and transverse colon. The descending colon appears to be the most susceptible and it is affected practically in all animals in some experiments. Less frequent (less than 25% of cases)
170
KAZYMIR M . POZHARlSSKl ET AL.
are the tumors of the cecum, while in the terminal part of the ileum, tumors hardly arise at all. Such a varying susceptibility of the different segments of the intestinal tract, particularly that of the descending colon and ileum, makes this experimental model especially valuable for various comparative studies (biochemical, autoradiographic, etc.) because specific neoplastic changes may be distinguished from nonspecific ones, e.g., toxic lesions. Experimental intestinal tumors have diverse forms of macroscopic appearance. The following anatomical varieties of tumors may be distinguished on the basis of appearance and type of growth (228,230). Macroscopic classification of epithelial tumors is as follows:
I. Exophytic tumors 1. Polypoid tumors (a) pedunculated polyps (b) sessile polyps 2. Cauliflowerlike tumors 3. Fungiform carcinomas 11. Exophytic-endophytic tumors 1. Plaque-shaped carcinomas 2. Saucer-shaped carcinomas 111. Endophytic tumors 1 . Ulcerative-infiltrative carcinomas 2. Circular carcinomas 3. Carcinomas of linitis plastica type 4. Carcinomas of adenomatous diverticulum type As the above types of tumors are described in detail elsewhere (226, 228), only a short commentary on the rare forms of cancer follows below. Carcinoma of the linitis plastica type makes the intestinal wall much thicker and firmer and causes the mucosal folds to disappear completely. This very rare type of carcinoma affects a considerable length of the intestinal tract. Carcinomas of linitis plastica type are very rare in humans also. Only 100 such cases have been reported; as few as 34 of those tumors were primary, while the rest were metastatic lesions (249 ). Another peculiar variety of experimental tumor is the so-called “adenomatous diverticulum” type of carcinoma. At first, they were referred to as tumors of the glandular stomach of the rat (28,299). Unlike these authors, we consider the adenomatous diverticulum to be a malignant tumor. These neoplasms do not seem to occur in humans. Each macroscopic type of tumor tends to prevail in a particular part of the intestines (228,230). For instance, endophytic-exophytic tumors and carcinoma of the adenomatous diverticulum type arise most frequently in the duodenum and jejunum. All types occur in the descending colon, but exo-
EXPERIMENTAL INTESTINAL CANCER RESEARCH
171
phytic ones appear to predominate. The cecum is frequently affected by ulcerative-infiltrative tumors. Carcinoma of the linitis plastica type occurs exclusively in the cecum, and circular carcinoma in the initial portion of the ascending colon. The dynamic observations of the process of tumor induction show that the macroscopic types are rather stable and that endophytic and exophytic-endophytic tumors and their variants do not result from necrosis and ulceration of polypoid tumors. Many authors (19,205,286,287,338), who studied the morphology of experimental intestinal tumors, describe only adenomatous polyps and well-differentiated adenocarcinomas, which do not metastasize. However, King and Varasdi (141), Laqueur (150), Schauer et al. (272), and Wittig et al. (345)observed mucinous and signet-ring cell carcinomas, in addition to adenocarcinomas. The following histological classification of neoplasms was suggested on the basis of the study of a great number of DMH-induced tumors in rats (228): I. Tumors of cylindrical epithelium with a marked tendency to form glandular structures 1. Polyp (a) hyperplastic (b) adenomatous (adenoma) 2. Carcinoma in sitir 3. Superficial cancer 4. Adenocarcinoma (a) tubular (b) papillary (c) mucus-secreting (d) scirrhous (e) villous tumor (f) glandular-villous tumor 11. Tumors incapable of forming glandular structures 1. Mucinous carcinoma 2. Signet-ring cell carcinoma 3. Scirrhous carcinoma 4. Solid carcinoma According to Lingeman and Garner (158), the adenocarcinomas of the first group may be designated as either well or moderately differentiated ones (i.e., as corresponding to grade I or grade I1 in their histological classification), while the tumors of the second group may be referred to as “poorly differentiated” (i.e., as corresponding to grade I11 or grade IV of adenocarcinomas in the classification). However, it should be
172
KAZYMIR M . POZHARISSKI ET AL.
pointed out that the above grades of adenocarcinoma differentiation describe the level of histotypical differentiation only and do not correspond to their cytotypical differentiation (see 244 ). Neoplasms, which resemble the hyperplastic (metaplastic) polyps of human colon, occur very rarely in experimental material. It is widely believed that hyperplastic polyps are not actually tumor lesions (I 97 ). Both their nature and the mechanism of development under experimental conditions and in human pathology remain to be established. Many investigators describe experimental adenomatous polyps (see above). Our results, however, show that adenomatous polyps arise very rarely-on exposure to DMH and also at the later stages of intestinal tumor induction. They are usually observed in combination with multiple carcinoma (230 ). Maskens (I 76 ) claims that all the neoplasms induced by DMH treatment are malignant. Considering the late onset of adenomatous polyp formation and their invariable occurrence in combination with multiple cancer, it may be supposed that these neoplasms result from the reactive alterations in the mucosa in response to the presence of multiple carcinomas. This possibility for humans was discussed by German authors many years ago (see 131 ). The distinguishing feature of carcinoma in sirir consists in that it is confined to the intestinal crypt and is, therefore, literally a “carcinoma in place,” whereas superficial cancer shows invasion, which is, however, limited to the lamina propria of the mucosa. As far as villous tumors are concerned, it should be pointed out that, in spite of their morphologically “innocent” appearance, even very small neoplastic lesions are capable of pronounced invasion. Signet-ring cell and mucinous carcinomas usually exhibit peculiar atypical and pleomorphic features. Mucinous carcinoma often reveals such secondary alterations as calcification and ossification. Among human tumors of the gastrointestinal tract, calcification also generally occurs in mucinous carcinoma. It is suggested that calcium deposition is promoted by glycoproteins which act as an ion-exchange medium (277). Tumors of certain histological types tend to arise in definite segments of the intestines (228,230). For example, nearly 90% of all tumors affecting the duodenum, transverse, descending colon, and rectum are different varieties of adenocarcinoma (tubular, papillary, mucus-secreting, and scirrhous). However, mucinous and signet-ring cell carcinomas arise rather frequently (about 29%) in the jejunum and ascending colon. The prevalent localization of experimental mucinous carcinoma in the ascending colon was reported by Ward (321 ). The same predominance of mucinous carcinoma was also observed in human ascending colon (346 ). Particularly frequent are signet-ring cell and mucinous carcinomas
EXPERIMENTAL INTESTINAL CANCER RESEARCH
173
in the cecum, which acccount for over 52% of tumors. The above data illustrating the selective localization of certain types of carcinoma in certain parts of the intestine provide a basis for a purposeful study of the morphology and morphogenesis of carcinomas of different histological types. It is generally possible to trace a correlation between the macroscopic type of tumor and its microscopic structure (228,230). For example, exophytic neoplasms usually have the structure of a well-differentiated tubular or papillary adenocarcinoma. Cauliflowerlike structures are histologically villous o r glandular-villous tumors. Signet-ring cell, mucinous, and scirrhous carcinomas, capable of pronounced invasive growth, most frequently assume the form of endophytic tumors. Exophytic-endophytic neoplasms develop as a scirrhous o r signet-ring cell carcinoma. Adenomatous diverticulum has the structure of a moderately or well-differentiated tubular adenocarcinoma, often including papillary structures. Many cases of adenomatous diverticulum reveal mucus-secreting adenocarcinomas, which transform to mucinous and signet-ring cell carcinomas in some parts of tumor. Schauer et al. (272) reported one case of a “typical adenocancroid” in rat colon but did not provide any description or illustrations. Squamous cell metaplasia in an experimental carcinoma of the rectum is mentioned in another paper (141). However, in the course of the examination of thousands of experimental intestinal tumors, we have never encountered structures comparable with squamous cell epithelium. These facts seem to suggest that tissue specificity is preserved even during tumor growth and to support the theory of histological determination (42 ). Experimental intestinal tumors often involve metastasis formation. When neoplasms are induced by DMH treatment, metastases occur in more than 50% of cases (226,228). Initially, tumor cells penetrate in lymphatic vessels and perineural spaces. Metastases first appear in the lymph nodes of the ileocecal angle, base of mesentery, greater omentum, and parietal peritoneum. When metastases develop in the diaphragm, they often extend through the thoracic cavity and infiltrate the lymph nodes of the anterior mediastinum (area thymica), visceral, and parietal pleura. In the viscera, only the lungs, liver, adrenals, kidneys, and ovaries are affected by metastases. It is chiefly mucinous and signet-ring cell carcinomas that tend to metastasize (141,226,272,322). The above data testify that the morphology and clinical behavior of intestinal tumors induced in experimental animals resemble their human counterparts very much. Nearly all morphological varieties of human epithelial tumors may be reproduced experimentally. The spread of intestinal tumors in the rat (metastases in regional lymph nodes, implan-
174
KAZYMIR M. POZHARISSKI ET AL.
tation metastases in peritoneal cavity) also bears much resemblance to that in man, although hematogenic metastases are rare in experimental animals. Unlike man, experimental rodents show a very high frequency of tumors of the duodenum and jejunum. However, the localization of neoplasms of the small intestine (predominantly its proximal parts) is similar in rat and man (164). Experimental animals usually develop well- or moderately differentiated adenocarcinomas, while hardly any nondifferentiated anaplastic carcinomas occur generally. Finally, experimental exposures result, as a rule, in the formation of multiple lesions in different intestinal segments, which is rather infrequent in human pathology. On the other hand, intestinal carcinoids, which are rather common in man, have not yet been described in experimental animals. However, the specific features of intestinal tumors and some differences in etiologic factors do not detract from the importance of experimental models and, on the contrary, even lend to their value for studies of the pathogenesis and morphogenesis of tumors of this localization, thanks to such characteristics as the discriminate susceptibility of certain intestinal segments to multiple neoplasms, induction of tumors of definite type, and so on. B. ELECTRON MICROSCOPIC STUDIES There is very scarce evidence in electron microscopic studies of experimental tumors of the intestines. Spjut and Smith (286) carried out a special comparative electron microscopic study of intestinal tumors in man and similar neoplasms induced in rats and came to the conclusion that the cells of both subjects have very much in common, in their ultrastructure. These authors did not find any ultramicroscopic signs of transformation of adenomatous polyps to adenocarcinoma. Haase rr al. (11 1 ) briefly described the ultramicroscopic structure of adenocarcinoma cells in mice and also concluded that it is similar to that described for human neoplasms of the colon and rectum. These investigations were concerned with adenomatous polyps and well-differentiated adenocarcinomas only. The results of the electron microscopic studies reveal that the ultrastructure of cells of experimental intestinal adenocarcinoma undergoes diverse changes (1 11,244,286,306,344). These changes affected both the nuclei of tumor cells and cytoplasm organoids (Fig. I ) . The nuclei of
(a) (b) (C) FIG. 1 . (a) Scheme of cellular ultrastructure in normal mucosa. (b) adenocarcinoma and ( c ) signet-ring cell carcinoma. From K. M. Pozharisski and G. A. Sovost'yanov, (244 ). Used with permission of "Medizina" publisher.
176
KAZYMIR M . POZHARISSKI ET AL.
adenocarcinoma cells are usually relatively large, take up much of the tumor cell space, and are somewhat displaced toward the apex. The outlines of tumor cell nuclei are irregular and have numerous invaginations of cytoplasm. These cells often contain augmented nucleoli and microdispersed, uniformly distributed chromatin. There are more free ribosomes and polysomes and fewer vacuoles in the cells of such tumors. Mitochondria display a considerable variety of shape and size, extent of crista development, and electron density of the matrix. But the most pronounced changes occur in the cell surface (see Fig. I). Rare, often-deformed microvilli may be seen on the apical parts of adenocarcinoma cells. Microvilli and glycocalyx are nearly completely reduced in many cells of such tumors. Wide spaces, filled with an electron-transparent material and interwoven cytoplasmic spikes of tumor cells, are formed between the adjacent sides of cells. These changes represent a morphological manifestation of the disorders in the intercellular contacts of tumor cells. The outgrowths of cytoplasm extending from the basal surface of cells through ruptures in the basal membrane into the submucosa, indicate the invasiveness of adenocarcinoma cells; the basal membrane remains intact in other areas (244 ). Using a scanning electron microscope for examination of experimental adenocarcinomas, Barkla and Tutton (17) observed a varying thickness and uneven surface of the basal membrane but did not see any apparent ruptures. Electron microscopic examination revealed that all organelles of adenocarcinoma cells are evenly pronounced. These data show that adenocarcinoma cells differ considerably from the differentiated columnar cells of the mucosa, but still they have much in common with the proliferating epithelial cells of the deep layers of the crypts (178,224). These observations may be interpreted as the lack of specific functions in the cells of these tumors, which points to a low level of their cytotypical differentiation, though they retain a histotypical differentiation, being the structural elements of organoid formations. In contrast to this, the ultrastructural studies of signet-ring cell carcinoma show that its cells exhibit a very high level of differentiation (244). It is manifested by their specific function of intensified output of glycosaminoglycans (Fig. 1). Such signet-ring cells, which have completed their cycle of differentiation, are round and most of their cytoplasm is filled with secretion. The nucleus and organoids, which have survived in the form of the vacuolar system and few small swollen mitochondria, are displaced toward the cell periphery. The nucleus appears to be small, shriveled, and hyperchromatic. Although signet-ring cell carcinoma reveals a highly pronounced cytotypical differentiation, a histotypical one disappears completely.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
177
The above comparison of the ultrastructure of the cells of different histologic forms in experimental malignant tumors of the intestine is of interest from two points of view: First, it indicates that the degrees of histo- and cytotypical differentiation of cells do not necessarily coincide. Second, it points to the lack of a direct correlation between the disturbances in tumor cell differentiation and malignant growth. This is particularly obvious in the case of signet-ring cell carcinoma, which displays a high rate of growth, high invasiveness, and an ability to produce extensive metastases, on the one hand, and the cytological manifestations of high differentiation, on the other. The regular examinations of the mucosa of rat colon at different terms (1-3 months), after exposure to DMH, failed to detect any ultrastructural changes of a precancerous nature in cells (224). The electron microscopic changes in cells could not be detected until the lesions of the carcinoma in situ and superficial cancer type had developed. These data show that cell ultrastructure undergoes changes in malignant tumors only (244,286 ). This may be accounted for by the fact that a tumor cell does not possess any pathognomonic ultrastructural features, and it is especially true with cells that do not exhibit any signs of malignancy. On the other hand, the morphogenetic studies conducted on the light level show that experimental intestinal tumors may develop de n o w (see Section 111,D).The scanning electron microscopic study conducted recently by Barkla and Tutton ( 1 7 ) revealed such changes, as bulging crypts, in the gut relief in DMHtreated rats before the appearance of developed tumors. On the basis of these findings, the above authors rule out the possibility of cancer development de novo. But they did not endeavor to establish the nature of the “pretumor” changes that they observed. It may be supposed that those changes were actually in siru carcinoma or superficial cancer. It should be mentioned in conclusion that the changes in the ultrastructure of the cells of experimental intestinal adenocarcinoma are identical to those in human malignant tumors of the same site (20,100,125,133, 134,136 ) .
C. HISTOCHEMICAL STUDIES At present, there is an abundant literature on the enzymohistochemical description of colonic and rectal neoplasms in man (55,145,146,167, 182,198,329). However, only few papers deal with histochemical investigations in the enzymes of experimental tumors of the intestines. Vollnagel et al. (316 ) studied succinate dehydrogenase, alkaline phosphatase, nonspecific esterase, and adenosine triphosphatase in the DMH-
178
KAZYMIR M. POZHARISSKI ET AL.
induced carcinomas of rat small intestine and reported that their levels in tumors are reduced, as compared with normal mucosa. Matsuyama et a / . (f 77 ) reported lower levels of lactate dehydrogenase, glucose-6-phosphate dehydrogenase, acid phosphatase, and leucine aminopeptidase in the same tumors. Pozharisski and Kolodin (242 ) studied the activity and localization of 19 enzymes in the DMH-induced tumors of the colon and small intestine of the rat. These enzymes are characteristic of the main metabolic processes, such as glycolysis (hexokinase, glucose-6-phosphate isomerase, a-glycerophosphate- and lactate dehydrogenase), tricarboxylic acid cycle (isocitrate-, succinate-, and malate dehydrogenases), terminal oxidation (cytochromeoxidase, NADH2- and NADPH,-diaphorases), pentose cycle glucoso-Qphosphate- and 6-phosphogluconate dehydrogenases), RNAprecursor synthesis (inosine 5'-phosphate- and guanosine S-monophosphate dehydrogenases), hydrolysis of monoesters of orthophosphoric acid (acid and alkaline phosphatases), amino acid synthesis (glutamate dehydrogenase), oxidative deamination (monoaminoxidase) and alcoholic fermentation (alcohol dehydrogenase). The results of these enzymohistochemical investigations showed that the enzymic activity involved in tissue respiration and Krebs cycle is lowered in experimental adenocarcinomas, while the level of glycolytic enzymes is increased, as compared with normal intestinal epithelium. Moreover, these neoplasms reveal an enhanced activity of the enzymes of the pentose-monophosphate shunt. These changes reflect the metabolic transformations that take place in tumor tissue. In essence, in tumors, energy is generated chiefly as a result of an intensified aerobic glycolysis, while in normal tissues, it is mostly supplied by oxidative phosphorylation. The elevated level of the enzymes of the pentose-monophosphate shunt may be explained by an intensification of the synthesis of nucleic acid precursors, due to the presence of a greater amount of proliferating cells in intestinal neoplasms, as compared with normal mucosa (239 ). A comparison of the experimental findings with clinical material shows that essentially identical metabolic disturbances occur in human adenocarcinomas of the colon and experimental tumors at the same site. It should be stressed that the microscopically unaltered mucosa, both at the early stage of tumor induction and along the periphery of developed neoplasms, does not show any marked differences in histochemical properties from that in nonelicited control animals. The first histochemical disorders are distinguishable only in in situ carcinoma, which is actually identical to invasive adenocarcinoma, as far as these characteristics are concerned (242 ). Vollnagel et a / . (316) and Schauer and Kunze (271 ) could not detect any preneoplastic or precancerous changes in experi-
EXPERIMENTAL INTESTINAL CANCER RESEARCH
179
mental intestinal tumors on the basis of histochemical studies either. It should be pointed out that the comparative histochemical studies of normal mucosa, adenomatous polyps, and colorectal adenocarcinomas in man failed to confirm the current view on the precancerous significance of polyps in carcinogenesis (55,f98,329). Filipe (98 ) undertook a detailed histochemical investigation of mucins in rats exposed to DMH and found that the secretion of sulfomucins is prevalent in the distal part of normal colon, while goblet cells start predominantly to produce sialomucins after DMH injection, but before any tumors develop; there was little or no secretion of mucins in tumor lesions. On the basis of these data, this author concludes that carcinogenesis passes through many stages and claims that disturbances in mucin production may be regarded as an indication of precancerous alterations. These observations are consistent with the earlier findings on the qualitative changes in mucin secretion, established in the vicinity of carcinoma and at some distance from it in human colectomy preparations ( 9 9 ) . However, the conclusions made by Filipe (98 ) seem to be questionable. First, the data presented by this author point to a considerable variation in sulfomucin and sialomucin distribution in different segments of normal intestines. For instance, sialomucins predominated in the proximal segments of rat colon, which, naturally, cannot be interpreted as a precancerous alteration, all the more so because tumor incidence at this site is generally much lower than in the distal part of the colon. Second, histochemical changes-from incipient morphological lesions (dysplasia) to overt malignant tumor-did not show a distinct tendency. Finally, the illustrations provided by Filipe (98 ) clearly demonstrate that in DMHinduced tumors, sialomucins were predominantly observed on tangential sections, which mostly show the superficial layers of the mucosa, while it is precisely in these layers that m u c k are secreted at a high level in normal tissue.
D. MORPHOGENESIS OF EXPERIMENTAL INTESTINAL TUMORS Morphogenesis of human and experimental tumors has not received due attention to date, probably, because of the underestimation of the role of this problem in the overall knowledge about tumors. The few papers dealing with the mechanism of experimental intestinal tumor development may be divided into two groups. Some authors hold traditionally that carcinogenesis is preceded by such morphological changes as hyperplasia of the mucosa and adenomatous polyp development (17, 98,ff2,218,338,345).The others maintain that carcinoma develops with-
180
KAZYMIR M . POZHARISSKI ET AL.
out preliminary changes in the mucosa (139,176,178,271,285,287, 316,321 ). However, all these conclusions are not based on the results of strictly morphogenetic studies. The main principle of morphogenetic studies-a stage-by-stage examination of the material-was not observed in the above investigations. As a rule, such studies are concerned with developed tumors, which are usually classified according to some conjectured stages of carcinogenesis in a rather arbitrary manner. No attempt is generally made to establish the nature of preinvasive changes, on the assumption that all these changes are always benign. The authors, who reject the stage-by-stage development of cancer, make such conclusions because they often encounter endophytic tumors, the genesis of which is hardly suggestive of a preceding stage of adenomatous polyp. However, the current literature contains all the necessary data that may be used as a methodological basis for thorough stage-by-stage studies of the morphogenesis of intestinal tumors. The model of DMH-induced tumors is particularly valuable in this respect. The point is that the treatment of rats with DMH results in the formation of multiple carcinomas of the descending colon (which is only 5-6 cm long) in nearly all animals simultaneously (in about 5-6 months). Numerous specimens, taken from such a short length of the gut for histological examination, provide all the necessary information on the changes that have taken place in the intestinal tissues and even those which preceded the development of invasive carcinoma. Moreover, since the intestine has thin walls, it is very easy to detect the slightest alterations with the naked eye or in a binocular lens, which facilitates a purposeful screening of the material for further microscopic studies (230 ). Pozharisski (230 ) undertook a systematic study of the morphogenesis of intestinal tumors induced in rats by DMH treatment. This investigation involved the examination of the material on a strictly chronological basis. It should be emphasized once again that it is only such a stage-by-stage approach that can ensure an objective interpretation of the sequence of morphologic changes. According to this method, intestinal tissue specimens, taken from 10-70 rats, were examined every successive week of the treatment with the carcinogenic suhtance; a total of 2000 animals were used in the study. The results showed that 1 month after the beginning of the experiment, the proliferative compartment in intestinal crypts becomes wider, as indicated by the arrangement of mitoses and 3H-thymidine-labeled cells, within 1 hour after labeling. For example, while the labeled epithelial cells of normal colon in the rat do not spread beyond two-thirds of the crypt’s length (as seen from the crypt base), some “H-TdR-labeled cells may be found even in the crypt’s mouth after four injections of DMH.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
181
Pozharisski (230) believes this to be a manifestation of the disturbed proliferation and differentiation of epithelial cells; possibly, due to the absence of the “ban” on DNA synthesis in transformed cells, which comes into force as they move along the crypt’s length (160). It is important to point out that the superficial layers of crypts, in which migrating transformed cells continue to proliferate, are precisely those sites where the first morphologically detectable tumor lesions, i.e., in situ carcinoma, develop. In situ carcinomas generally arise by the end of the second month. The rather long interval between the detection of transformed cells outside the proliferative zone and formation of the first foci of in situ carcinoma may probably be accounted for by the elimination of a fraction of transformed cells from the epithelial lining by some defense (most likely, immunological) mechanisms. At subsequent stages, tumor foci extend through the crypts, giving rise to invasion, which, however, is still confined to the lamina propria of the mucosa. Invasion manifests itself in changes in the crypt configuration, fission of tumor glandules, surrounded by developing desmoplastic reaction, and accumulation of wandering cells; changes that are considered to be superficial carcinoma. Subsequently, the remaining mucosa is gradually substituted by tumor. When tumor extends into the submucosa of intestinal wall, invasive adenocarcinoma is formed (Fig. 2). Morphogenetic studies also revealed that adenocarcinoma may stay at the preinvasive stage for a long time, without showing a pronounced atypia and polymorphism. Therefore, it is easy to explain why such lesions are regarded as adenomatous polyps: such conclusions are not based on the precise knowledge of tumor morphogenesis (152,218,321). The development of signet-ring cell carcinoma is signalled by the accumulation of goblet cells filled with mucus in the bottom layers of crypts, in which they assume a signet-ring appearance. This stage of the development of signet-ring cell carcinoma may be considered to be carcinoma in situ. Subsequently, signet-ring cells extend through basal membrane and invade the underlying layers of intestinal wall as far as the serosa. Hence, experimental intestinal carcinoma develops de n o w , not preceded by any hyperplastic alterations in the mucosa or adenomatous polyps. Such a morphogenesis of intestinal carcinoma is also supported by the results of electron microscopic and histochemical investigations (see Sections III,B and C). Intestinal tumors arise invariably after a single administration of carcinogen (230); thus ruling out the existence of factors that may cause multiple changes in the phenotype of cells.
FIG.2. Scheme of morphogenesis of experimental intestinal tumors: ( I ) normal mucosa of colon: (la) widening of proliferative zone within crypts on tip of mucosal fold: ( I b ) in sitir carcinoma on tip of fold: (Ic) superficial cancer on tip of fold: (Id) exophytic adenocarcinoma without invasion of pedicle: ( le) invasive exophytic adenocarcinoma: (2a) widening of proliferative zone in crypts of mucosa between folds: (2b) in sifu carcinoma in mucosa between folds: (2c) superficial cancer in mucosa between folds: (2d) adenocarcinoma extending across entire thickness of mucosa, with submucosal layer still unaffected: (2e) endophytic invasive adenocarcinoma: (3a) accumulation of large quantities of goblet cells in bottoms of crypts: (3b) goblet cells converting to signet-ring cells and penetration of some of them through basal membrane: (3c) development of signet-ring cell carcinoma. From K. M. Pozharisski (230).
EXPERIMENTAL INTESTINAL CANCER RESEARCH
183
On the basis of the results of experiments, which show that intestinal adenocarcinoma develops de n o w (1 76,271,287,321,344), it is possible that adenomatous polyp formation does not play a vital role in the genesis of colorectal tumors and is merely a concomitant pathology (231 ). This suggestion seems to be justified since the possibility of the malignant transformation of these benign tumors is held in doubt by many researchers (4,39,40,129,225,231,288,309,349). Comparative pathology provides additional data which cast doubt upon the role of adenomatous polyp in carcinogenesis. For example, adenomatous polyp occurs frequently in the pig: 13.8% of these animals (or even up to 40% in some breeds) are affected by this lesion (144). Yet, malignant transformation of adenomatous polyps in pigs has never been reported. It should be mentioned that although colon polyps were observed in 10% of C57Bl mice, followed up during their whole life, malignant transformation of these tumors was never reported ( 8 0 ) .On the other hand, there is no connection between cancer, a frequent disease in sheep, and polyps, which do not occur in this species (106,278). Taking into account the data available on the morphogenesis of experimental intestinal tumors, it seemed desirable to conduct a comparative study of human polypoid tumors of the colon and rectum (231 ). When the malignancy of these neoplasms was established, the following morphological signs of atypia received most attention: a sharp decrease or complete absence of goblet cells, pseudostratification or even a multilayer structure of the epithelium, location of mitoses in the superficial layers of tumor and increased diversity of pathological forms of karyokinesis, development of papillary and villous structures and changes in their configuration, bizarre shapes of tumor glandules and their fission, cellular and nuclear polymorphism, appearance of cribriform structures, a great number of plasma cells and karyorrhexis-affected lymphocytes, etc. When these changes were detected in biopsy material, invasion in the submucosal and deeper layers of the removed segment of colon was found in nearly 70% of those patients who were operated on later. This points to the validity of the structural and cytological criteria of malignancy used in the study. This investigation often revealed such early stages of adenocarcinoma development as in sit14 carcinoma and superficial cancer outside polyps, on the one hand, and tiny polypoid lesions (up to 2-3 mm in diameter), which displayed all the signs of malignancy, invasion included, on the other. However, not a single case of malignant transformation of adenomatous polyp was recorded. These findings suggest that intestinal tumors also develop in humans de n o w . Comparative morphologic data on experimental and human intestinal
I84
KAZYMIR M . POZHARISSKI ET AL.
tumors showed that villous and glandular villous (with marked villous component) tumors are malignant ah initio (231 ). Some data in the literature point to essential differences between villous tumors and adenomatous polyps, in pathohistological characteristics and clinical course (4,6,110). The ultra-structure of villous neoplasms is very close to that of well-differentiated adenocarcinoma, whereas adenomatous polyp does not differ from normal mucosa in this respect (100,132,134). Villous tumors have more enzymic and immunologic features in common with adenocarcinoma than with adenomatous polyp (55,SS ). The kinetics of growth of exophytic tumors and villous neoplasms of the colon is the same and seldom differs from that of adenomatous polyp (335). The literature data in their totality support the opinion that the nature of intestinal tumor (benign or malignant) is predetermined at the time of its appearance (49,225). Such an approach transfers the emphasis from undertaking attempts at detection of the so-called "preneoplastic" lesions to the early diagnosis of malignancy proper at the preinvasive stage.
IV. Factors Modifying Intestinal Carcinogenesis
A. ENDOGENOUS FACTORS 1. Genetic Factors
The current literature has abundant data on the differences in susceptibility to intestinal carcinogens in different strains of experimental animals. The investigation of the carcinogenicity of methylcholanthrene (MC) fed to Syrian hamsters of seven inbred strains showed that intestinal tumor yield in some strains was rather high (up to 5% in the small intestine and up to 65% on the large bowel); the susceptibility to MC appeared to be much lower in other strains, while some strains were completely resistant (127 1. Reuber and Thomas (259) found that among different inbred strains of rats (Buffalo, Marshall, ACI, Fischer, and Osborne-Mendel), Buffalo rats were the most susceptible to the carcinogenic action of methylazoxymethanol acetate on the intestines. A varying incidence of duodenal carcinoma in BD-IX, BN, and Lewis rats, treated with MNNG in drinking water, was described by M. S. Martin et al. (174). According to Moon and Fricks (191), BD-IX rats are definitely more susceptible to the carcinogenic effect of DMH than those of BD-I1 strain. Mouse strains also exhibit distinct differences in susceptibility to carcinogens. For example, DMH induces colon tumors in two-thirds of Ha/
EXPERIMENTAL INTESTINAL CANCER RESEARCH
I85
ICR mice but not in C57/B mice (88). Subsequently, it was shown that DMH treatment fails to induce intestinal tumors in DBA/2 mice, while tumor incidence in ICR/Ha mice derived from a random cross is lower than in the inbred strain (87). Diwan et al. (69 ) report the following colon tumor yields for different strains of mice exposed to DMH: SWRL, 83%: P/y, 80%: C57BL/6j, 48%; and GR, 14.2%. No tumors were induced by this treatment in AKR/y mice. However, squamous cell carcinoma of the anal region occurred in nearly all GR mice; only single such tumors arose in the other strains or did not occur at all (C57BL/6J). These data point to differences in the genetically determined susceptibility of tissues to the action of the same carcinogenic substance. Different effects of castration and administration of dihydrotestosterone on intestinal tumor incidence were observed in the rats of BDII and BDIX strains (191 ). Evans et al. (89,90) undertook a genetic analysis of colon tumor induction by DMH on F,, F2, and reciprocal back-cross hybrids derived from a cross between the 100% susceptible ICR/Ha and completely resistant C57BL/Ha mice. This analysis revealed that the single gene inherited by ICR/Ha mice is responsible for their susceptibility to the tumor-inducing effect of DMH. There was no evidence for sex linkage or Y-group linkage. The susceptibility to this carcinogen proved to be dominant, because with the mating scheme used, tumor yield was in line with Mendelian segregation. The data on intestinal tumor induction in C57BL mice are controversial (50,69,87,89,90).It is supposed to be due to the genetic differences between sublines (69 ). Inheritance seems to be an important factor in the carcinogenesis of human colon, particularly in the cases of familial polyposis, which is characterized by a dominant autosomal mode of inheritance (113,143). 2. Sex Hormones The intestinal tumor incidence in Sprague-Dawley female rats, treated with cycasin and its aglycone, is much lower than in the males of the same strain (150). The later formation of intestinal neoplasms in female rats than in males was recorded in the experiments with 2-acetylaminofluorene (19 ) as well as a less frequent development of multiple intestinal tumors in female Buffalo rats than in males fed N , Nr-2,7-fluorenylenbisacetamide (348 ). Duodenal carcinomas were more frequent in male Lewis rats than in females (174). When Sprague-Dawley rats were treated with methyl(acetoxymethy1)-nitrosamine, intestinal tumors were somewhat fewer and less frequent in females than in males: a particularly striking feature of these experiments was a longer life span of the female
186
KAZYMIR M . POZHARISSKI ET AL.
rats (137). Similar data on the sex-associated distinctions in intestinal tumor induction in rats and mice are reported by other authors (13,26, 191,259 1.
Pozharisski and Anisimov (232 ) discovered sex distinctions in intestinal carcinogenesis in noninbred rats and BDIX rats treated with DMH. Female rats had a rectal tumor incidence 2.8-3.1 times that in male rats and a longer life span. Female rats survived for a longer time because tumors arose later, showed less involvement, and metastases were less frequent. The latter phenomenon is probably due to the fact that female rats generally develop fewer tumors of the small intestine and cecum, i.e., at the sites at which extensively metastasizing signet-ring cell carcinomas occur most frequently. However, no sex distinctions were found in the incidence of tumors of Zymbal’s gland, which are often induced by exposure to DMH. Sex distinctions disappeared after castration of both female and male animals in the experiments of Pozharisski and Anisimov. Meanwhile, the castrated animals revealed distinctions of a different nature: tumors of the small intestine, cecum, and Zymbal’s gland arose less frequently in males than in females. Castration of female Wistar rats resulted in a lower frequency of colon tumors induced by 3,2’-dimethyl-4-aminobiphenyl: moreover, tumors arose at other sites (285). On the other hand, the incidence of spontaneous adenomatous polyps of the colon showed a sharp rise in pigs following castration (144 ). According to Moon and Fricks (191 ), castration reduced intestinal tumor incidence in male BDIX rats, which were treated with DMH, beginning from the age of more than 120 days. The administration of an androgen (dihydrotestosterone) to such animals was followed by an increase in colon tumor incidence, which was close to that observed in intact animals. A study on the influence of DMH treatment on the rat reproductive system showed that a single administration of this carcinogen led to a decrease in the weight of the testes and follicle-stimulating activity of the pituitary of male rats. In females, DMH did not affect the weight of the ovaries, uterus, or follicle-stimulating activity of the pituitary. In hemicastrated animals, DMH treatment did not inhibit compensatory ovarian hypertrophy but interfered with the suppression of compensatory ovarian hypertrophy induced by diethylstilbestrolpropionate (232 ). This indicates that DMH treatment raises the threshold of sensitivity of the hypothalamopituitary system to the homeostatic inhibition by estrogens. According to Dilman (67 ), the development of disturbances in the reproductive and energy homeostats is sufficient to trigger a string of diseases, designated as the diseases of compensation, all these factors paving the way
EXPERIMENTAL INTESTINAL CANCER RESEARCH
I87
to carcinogenesis. On the basis of the above data, Pozharisski and Anisimov (232) put forward a suggestion that, apart from the direct action of DMH on the cells of intestinal mucosa (alkylation of their macromolecules), this agent probably exerts a nonspecific effect on the hypothalamopituitary system, thus promoting the proliferation of transformed cells, The literature data show that endocrine disturbances were recorded both in the case histories of most patients with colon tumors and in clinical examination: these disorders seem to have played a certain role in the pathogenesis of malignant disease (310 ). 3. A g e Druckrey and Lange (76) did not observe any correlation between intestinal tumor incidence and age (within 1-60 days) in rats given a single dose of azoxymethane. Moon and Fricks (191 ) believe age to be an important factor in the induction of intestinal tumors by DMH in BDIX female rats older than 210 days at the beginning of the experiment. However, according to Balish ef ul. (13), the induction of intestinal tumors by MNNG administration per rectum was more effective in younger (30-day-old) than in older (60-day-old) Sprague-Dawley rats.
4. Microbiologicul Studies Intestinal flora constitute the most integral ecological factor in multicellular organisms. It is natural, therefore, that the role of bacterial flora in intestinal carcinogenesis should receive much attention. As mentioned above, cycasin is toxic and induces tumors at different sites, the intestines included. However, its administration to germ-free rats was not followed by a toxic or carcinogenic effect. Since cycasin is a glycoside it will be transformed to a carcinogenic metabolite-methylazoxymethanol (cycasin aglyconeronly by microbial P-glycosidase (150 ). When cycasin was administered to germ-free rats, it reached urine and feces unchanged, whereas most was metabolized in conventional animals. Methylazoxymethanol induces tumors both in conventional and germ-free animals, being unaided by the enzymic activity of microbial flora (see 151 ). The results of experiments involving tumor induction by DMH in isolated segments of the intestine suggested a role of microbial flora in the carcinogenic effect of this substance (227). It was demonstrated simultaneously by two teams of researchers (233,257) that DMH-induced intestinal tumors were much less frequent in germ-free rats than in conventional animals and the tumor process was less pronounced. After DMH metabolites are conjugated to glucuronic acid, they are delivered
188
KAZYMIR M . POZHARISSKI ET AL.
from the liver to the intestines via blood circulation (234,245). The pglucuronidase activity of intestinal flora results in the release of active carcinogenic metabolites, which are responsible for carcinogenesis. When mechanisms preventing intestinal carcinogenesis in germ-free animals are considered, it should be borne in mind that the renewal of intestinal epithelium proceeds at a much slower rate in such animals (3 ), while an enhanced proliferation of epithelial cells was shown to promote carcinogenesis (22Y 1. As a result of oral administration of Luctobacillrrs ucidoplzilrrs to DMHtreated rats, tumor yield in the proximal part of the small intestine was higher than in control rats. Presumably, Lactohacillrrs acidophilirs produces p-glucuronidase, thus promoting the release of an active carcinogenic metabolite in the intestinal segment, in which microbial flora is normally poor (223 ). In contrast to the above data, Asano c>t 01. (9)induced multiple intestinal tumors in germ-free rats treated with DMHper os. But these authors did not compare their results with those obtained in conventional animals and, therefore, their conclusion that the role of microorganisms in intestinal carcinogenesis is insignificant is not justified. It was confirmed later that the absence of microbial flora in the intestine reduces the carcinogenic effect of DMH considerably when administered subcutaneously (255 ) or intrarectally (252 ). It was discovered that the rectal administration of azoxymethane, a suspected metabolite of DMH, induces more intestinal tumors and at a higher incidence in germ-free and monocontaminated (Clostridia pevfiingpns) rats than in normal animals (255). The latter effect is attributed by the authors to an increased absorbability of the intestinal tract in germ-free animals. However, it should be pointed out that subcutaneous administration of azoxymethane proved to be more effective in inducing intestinal tumors in germ-free animals than in conventional rats (252 ). Therefore, nothing remains but to assume there are certain differences in the metabolic activation of DMH and azoxymethane. There is evidence that another carcinogenic agent, 2,3’-dimethyl-4aminobiphenyl, does not induce tumors in germ-free rats (47). On the other hand, intrarectal instillations of MNNG, a carcinogen of direct action, induced twice as many colon adenomas in germ-free rats as in conventional animals, whereas, the microbiological status of the animals used in these experiments did not affect the number of adenocarcinomas considerably (257,334 ). However, these authors admit that if the experiments had been continued for a longer time, the number of adenocarcinomas in germ-free rats would probably have increased. Such peculiarities of action of this carcinogenic substance should probably be accounted
EXPERIMENTAL INTESTINAL CANCER RESEARCH
189
for by its absorption or chemical degradation by bacterial flora. Balish ef al. (13 ) also demonstrated a faster development of intestinal tumors in germ-free rats, treated with MNNG and N-methyl-N-nitrosourea per rectum. Studies on the role of microorganisms in the pathogenesis of tumor growth are not limited to experiments on animals, but are also being extended to the elucidation of carcinogenesis mechanisms in humans. It was suggested, for example, that the low incidence of small intestine carcinoma is due to relatively poor bacterial flora in this intestinal segment (34 ). Geographic differences in the frequency of rectum tumors are also well known. These tumors are more frequent in Western Europe and North America and far less frequent in Africa, Asia, and South America. Comparative studies of the fecal flora in the population of these areas revealed that anaerobes which do not form spores prevail in both groups. However, bactericides, bifidobacteria, Cl. parupirtr$?curn and other Closrridia occur much more frequently in the populatim at high risk of rectal cancer, while eubacteria and Enferococcrrs occur less frequently, as compared with those areas where this type of cancer is infrequent. The anaerobe/aerobe ratio increases in proportion to rectum cancer incidence (7,74,/20). The pathogenetic significance of bacteria in human carcinogenesis lies in their ability to synthesize carcinogens and/or cocarcinogens from food components or intestinal secretions (120,121 ). Much importance, in this connection, is attributed to carcinogenic steroids, which are synthesized from bile components under the influence of microbial flora.
5 . Immunologic Sirrveillance BCG injections in rat intestinal wall does not affect the incidence or induction of intestinal tumors following short-term oral administration of DMH. This immunologic stimulation promoted a higher incidence of mucinous carcinoma, which often spread to the lymph nodes of the abdominal cavity. The BCG injections administered directly in large intestinal tumors were not followed by any distinct changes in them (263 ). On the other hand, the immunodepressive effect of the injection of a purified fraction of antilymphocytic serum failed to affect intestinal tumor induction in rats treated subcutaneously with azoxymethane (147 ). According to Schmahl et al. (273 ), additional administration of immunodepressants (cyclophosphamide, hydrocortisone, and methotrexate) and immunostimulators (BCG, albumin, and vitamin A) had no effect on DMH induction of colon carcinomas in rats.
190
KAZYMIR M . POZHARISSKl ET AL.
Hence, it appears that neither immunodepression nor immunostimulation have a distinct effect on intestinal carcinogenesis. However, recent findings cast doubt upon such a conclusion. Bansal et (11. ( 1 5 ) showed that administration of, antithymocytic globulin was followed by an enhanced DMH-induced carcinogenesis. This effect manifested itself in a shorter latent period, higher frequency, and multiplicity of tumors. These authors believe that the normal functioning of the immunological (thyma1 cellular) system can effectively suppress the development of some nascent tumors. When, however, this system is inhibited, tumor microlesions develop into gross neoplasms. Bansal et al. (15 ) maintain that DMH does not affect the immunological system. They drew this conclusion from the data showing that this agent does not affect lymphocyte count in peripheral blood and serum level, nor does it increase the frequency of infection of surgical wounds. But Dilman et ul. ( 6 8 ) reported that administration of four weekly doses (21 mg/kg) of DMH resulted in the formation of distinct immunodepression, as indicated by ( 1 ) suppression of reaction of blast cell transformation to stimulation by phytohemagglutinin or liposaccharides, (2) decrease in the titer of antibodies to sheep blood red cells, and (3) decline in the phagocytic activity of macrophages. These authors also revealed that DMH reduces the catecholamine level in the hypothalamus and carbohydrate tolerance, raises insulin and triglyceride levels in peripheral blood, and causes other disturbances (see Section IV,A,2). It should be mentioned that an antidiabetic drug, phenformin, was found to remove the immunological and metabolic disturbances brought about by DMH treatment. On the basis of these data, Dilman et a / . (68 ) concluded that DMH causes the so-called “metabolic immunodepression. ”
B. EXOGENOUS FACTORS 1. Diet
Epidemiologic studies suggest that colon carcinogenesis in humans is also subject to the influence of exogenous factors. Most importance in this respect is placed on diet, particularly, excessive intake of animal fats and proteins (8,347). Burkitt (31) sees a connection between the high incidence of colon tumors in the developed countries and a high-calorie cellulose-deficient diet. These hypotheses were tested experimentally and it was found that a fat-rich diet contributes to a higher incidence and multiplicity of colon tumors, induced by DMH and azoxymethane (30,213,258,265). Some
EXPERIMENTAL INTESTINAL CANCER RESEARCH
191
researchers observed a higher malignancy of tumors and more frequent development of metastases in animals fed a fat-rich diet (213,342 ). Notably, the type of fat (animal or plant) is irrelevant for the promotion of carcinogenesis (251 ). Intestinal adenocarcinomas occurred more frequently in the DMHtreated rats fed a protein-rich diet; the tumors were relatively larger and much more multiple (253,304). The nature of the protein had no influence on the rate of tumor incidence and growth (253). The results of Schmahl et a / . (273) are at variance with the almost generally accepted opinion that a fat- and protein-rich diet contributes to intestinal carcinogenesis. These authors reported that DMH induction of intestinal tumors in rats was not affected by such factors as a fat- and cholesterol-rich or a vegetarian diet. When the mechanism by which fats and proteins participate in intestinal carcinogenesis is considered, it is generally assumed that it involves (1) an enhanced output of bile acids and cholesterol, (2) changes in the composition of intestinal microflora and its enzymic properties, (3) changes in the enzymic systems of the body, which take part in carcinogen metabolism, and (4)disturbances in the immunological status. Recent experiments on animals showed that an excessive intake of fat is followed by a rise in the levels of sterols and total fecal bile acids. Simultaneously, it is accompanied by a rapid degradation of primary bile acids, resulting in higher concentrations of deoxycholic, mitocholic, and other bile acids in feces (213,256,258). At the same time, it is suggested that bile acids promote intestinal carcinogenesis. Indeed, repeated administrations of chenodeoxycholic, mitocholic, deoxycholic, and taurodeoxycholic acids, following single or several intrarectal instillations with MNNG, led to a sharp increase in the frequency of colon tumors and their number (199,254,256 ). When rats were fed cholestyramine, an elevated level of total bile acids and a higher percentage of secondary bile acids were found in feces (212 1. A combined oral administration of cholestyramine and DMH, azoxymethane, or methylazoxymethanol resulted in a high yield of tumors, particularly, in the large bowel, and accelerated their malignant transformation (9,35,2f 1,212 ). When candicine is administered, the fecal level of cholesterol and its highly degraded metabolites is increased, leading to the development of numerous tumors in the distal part of the small intestine. It is concluded on the basis of these data that bile acids promote carcinogenesis in the colon, while cholesterol and/or the products of its degradation contribute to tumorigenesis in the distal part of the small intestine (212). The promoting effect of bile acids on intestinal carcinogenesis is con-
192
KAZYMIR M . POZHARISSKI ET AL.
firmed by the results of the implantation of the common bile duct in the middle of the small intestine in rats. Under such conditions, azoxymethane treatment induced 2.5 times as many intestinal tumors than in control rats. The fecal level of bile acids in the operated rats was twice that in normal animals (43). DMH treatment after the removal of the gallbladder resulted in a higher frequency of colon tumors in rats. The effect of cholecystectomy in these experiments is associated with the synthesis of secondary bile acids by intestinal bacteria and with the disturbed passage of DMH through the liver, since it is not absorbed by the gallbladder (336 ). Singh ef d.(280) reported that administration of 6-N-propyl-2-thiouracil reduced the yield of intestinal adenocarcinomas induced by azoxymethane. They also observed that this thyrostatic agent lowered the levels of bile acids, neutral steroids, cholesterol, and coprostanol in feces. Therefore, these authors claim that hypothyroidism reduces the carcinogenicity of azoxymethane by decreasing the fecal concentrations of these substances. Reddy et a / . (256 ) presume that primary and secondary bile acids affect the proliferation of intestinal epithelial cells by altering their microenvironment. This effect is likely to account for the mechanism of promotion of carcinogenesis. Changes in the diet appear to modify the composition of microbial flora of the intestines. A meat diet led to an elevation in the activity of pglucuronidase and other enzymes in rats. These changes do not seem to be due to a mere induction of enzyme synthesis: it is more likely that they are brought about by changes in the composition of intestinal microbial flora. An intensified activity of p-glucuronidase promotes the conversion of glucuronides to carcinogenic aglycones (108 ). An elevated level of polyunsaturated fats in the diet was found to change microbial flora composition by increasing p-glucuronidase activity (30 ). Wattenberg (330) showed that the addition of certain chemical substances may affect the course of carcinogenesis by changing the metabolism of the carcinogens involved. This effect was particularly marked in the inhibition of intestinal carcinogenesis by disulfiram (331 ). This compound completely prevented intestinal tumor development in mice exposed to DMH. Our results confirmed these findings, but it appeared that disulfiram prevents the initiation of carcinogenesis in rats but does not affect its promotion (Pozharisski ef a / . , unpublished data). Subsequently, it was shown that DMH-induced carcinogenesis may be inhibited by sodium diethyldithiocarbamate and such pesticides as ethylene bis(dithi0carbamate) manganese and bisethylxanthogen (332 ). Pamukcu et al. (220) reported that disulfiram, butylated hydroxyanisole, and calcium chloride reduced intestinal tumor incidence by 25%-30% in the
EXPERIMENTAL INTESTINAL CANCER RESEARCH
193
experiments involving the use of bracken fern, Pteridium uqirilinum, as a carcinogenic agent. Selenium administered in drinking water cut down the frequency and number of colon tumors induced by DMH. Methylazoxymethanol-induced tumor frequency was not affected by the administration of selenium, but their number decreased (135 ). Presumably, a low level of lipotropic substances (methionine and choline) alters the metabolism of intestinal carcinogens in the liver (265). There are some data showing that the level of vitamin A plays a certain role in intestinal carcinogenesis. For instance, the administration of this vitamin inhibits the development of colon tumors induced by exposure to DMH and aflatoxin B, (208,209,264).However, Narisawa et ul. (200) obtained results which show quite the opposite: an insufficient level of vitamin A was responsible for a considerable inhibition of colon carcinogenesis and a lower incidence of tumors in rats in which MNNG was instilled per rectum. Broitman el ul. ( 3 0 ) reported that transformation of the lymphocytes of the thymus under the influence of mitogens diminished when more polyunsaturated fats were added to the diet. Hence, it is believed that an impaired immunocompetence in conjunction with some other changes due to a modified diet may be responsible for the promotion of intestinal carcinogenesis. The carcinogenic effect of DMH and azoxymethane in rats was decreased when the bulk of feed was increased by addition of cellulose or wheat bran (16,325,342). The effect of these bulk-filling feeds may be accounted for by their high capacity for bile adsorption. A liquid synthetic feed diet, containing a mixture of synthetic amino acids, safflower oil, and 85% of carbohydrates with vitamin and mineral salt ingredients, cut down DMH toxicity and frequency of tumors of the large and small intestine drastically (38). Mori and Hirono ( 1 93 ) reported that administration of coffee in drinking water increased the incidence of tumors, particularly, colorectal adenocarcinomas induced by a single oral dose of cycasin. The above data show that diet is an important factor of modification of intestinal carcinogenesis. The exploration of this problem is just beginning. 2 . Functional State of the Intestines There are separate clinical observations showing that the cutting-off of an intestinal segment affected by carcinoma (by means of colostomy) may lead to regression and resolution of tumor (56,81). More than 20 cases of resolution of rectal polyps in familial polyposis following colec-
194
KAZYMIR M . POZHARISSKI ET AL.
tomy and application of ileorectal anastomosis have been described (340 ). These observations indicate that the functional condition of the intestine and changes in the metabolism of its contents may influence tumor growth at this site. The results of the investigations based on the combined techniques of experimental oncology and surgery yielded some new data. For instance, it became possible to establish the metabolic pathways of intestinal carcinogens as well as the influence of intestinal contents on tumorigenesis of this localization. For example, exposure to 3,2’-dimethyl-4-aminobiphenyl fails to induce intestinal tumors distal to the colostomy in rats. On the basis of these data, it was suggested that the end carcinogenic metabolite is synthesized in the liver and goes to bile (46,205,285).Some authors consider intestinal contents to be cocarcinogenic (345 ). They came to this conclusion on the basis of the results of DMH induction of tumors in rats with “double-trunk” preternatural anus. Intestinal tumors arose distal to the colostomy much less frequently than in the unoperated animals. Pozharisski (227) performed two types of operation in rats: (1) application of “single-trunk” preternatural anus at the level of the splenic angle of the colon, and ( 2 ) isolation of a segment of the descending colon: the proximal end of which was closed by suture, while the distal one in the form of a colostomy was led out to the skin. The permeability of the gut was not obstructed due to the application of an end-to-end anastomosis. The rats of both groups were given subcutaneous injections of DMH. In the rats of the first group, the development of tumors distal to the colostomy was very rare, while in the second group, the isolated segment of the colon appeared to be affected by tumor lesions in all animals. Thus, the application of essentially identical surgical techniques produced opposite results. This means that the contents of the intestines does not by itself play a particularly important role in the induction of intestinal tumors, since colonic segments were isolated from intestinal contents in both experimental groups, but the carcinogenic effect proved to be different. Moreover, these experimental results show that DMH metabolites are delivered to intestinal wall via blood circulation but not by bile, as was suggested earlier (333 ). Campbell et al. (35) observed rare tumors distal to the colostomy in rats treated with azoxymethane and suggested that the following factors are involved: (1) some of the carcinogenic substance is delivered with bile; ( 2 ) bile promotes carcinogenesis; and (3) microbial flora is altered and epithelial cell proliferation is disturbed as a result of t h e absence of intestinal contents. The latter explanation seems to be the most plausible, because total
EXPERIMENTAL INTESTINAL CANCER RESEARCH
195
anaerobic microflora in the isolated segment of the colon is reduced (138), and it reveals hyperplasia with a sharply decreased fraction of proliferating epithelial cells (227 ). When intrarectal instillation of cycasin was performed, there were much fewer tumors distal to the colostomy than in the other segments of rat intestines (327). Notably, tumors were fewer in the isolated segment of the intestine, to which a directly acting carcinogen-MNNG-was administered (138,218,343).This effect seems to be due to a diminished proliferative activity of the epithelium in the nonfunctioning part of the gut. The resection of the proximal part of the small intestine, performed after a course of azoxymethane, resulted in a higher yield of colon tumors but was not followed by development of tumors in the ileum, although its mucosa showed a pronounced hyperplasia (341 1. Opposite results were reported by Werner ef al. (337) who removed the terminal part of rat ileum, the cecum, and ascending colon, and applied an ileotransverse anastomosis. The operated animals were treated with DMH, but the remaining intestines proved to be resistant to this carcinogenic agent. However, when DMH was administered after application of ileotransverse anastomosis (Pozharisski, unpublished data), neoplastic lesions developed in the ileum, in which under ordinary experimental conditions, exposure to this carcinogenic substance usually fails to induce tumors (228,230) . The above data in their totality demonstrate that the functional state of the intestines and changes in their contents may influence the realization of the carcinogenic effect of different agents which induce tumors in this organ. An opposite view is held by Gennaro et al. (105), who transposed segments of the small bowel into the large one, and vice versa. When azoxymethane was administered, tumors arose in the large bowel and its transposed segment, but did not affect the transposed part of the small bowel. These exquisitely designed and finely performed experiments demonstrated that some endogenous factors, which can promote or inhibit tumorigenesis, are at work in different intestinal segments. Yet, it does not detract from the importance of the exogenous factors modifying carcinogenesis. It was suggested that constipation is one of the etiological factors of colon and rectum carcinogenesis in humans (31 ). However, administration of purgative drugs did not affect the yield of tumors induced by 3,2'dimethyl-4-aminobiphenyl in rats appreciably. Since, as was mentioned above, the isolation of an intestinal segment inhibits the development of tumors therein, we decided to find out how such an operation may influence existing tumors. For this purpose, pre-
196
KAZYMIR M . POZHARISSKI ET AL.
ternatural anus was applied in the rats and treated with DMH for 4 months; and tumors of the descending colon were detected by laparotomy. After operation, the treatment with the carcinogen was terminated, but tumor growth in the isolated descending colon did not decrease: on the contrary, it became more pronounced. It is likely that the isolation of an intestinal segment, which brings about a reduction in functional and proliferative activity, prevents tumor initiation but does not affect its promotion. Moreover, under such conditions, cell loss in tumor is reduced considerably and, therefore, tumor mass increases rapidly.
3 . Injiirv qf the Mucosa Chronic injury of intestinal mucosa, which involves a long-term reparative regeneration and intensification of proliferation of epithelial cells, results in a sharp rise in the frequency of tumors at this site in rats exposed to DMH (229 ). Experimentally, the first chemical-induced tumors arise in the region of colon anastomosis and, subsequently, they appear to be most pronounced in this area (227). Oral administration of Citrohacfrr j k u n d i i induces a transient pronounced hyperplasia of colonic mucosa in mice. If coupled with DMH injections, tumors develop at a much faster rate; as early as 1 month after the beginning of the experiment (18 ). It was also discovered that iota-carrageenan (preparation C 16), isolated from Eiicherizu spinosrim, may influence intestinal carcinogenesis. When administered in feed and drinking water in combination with DMH injections, malignant lesions developed from the deeper layers of the crypts and were, as a rule, poorly differentiated mucus-producing carcinomas, tending to extend to the proximal part of the large bowel (132 1. These experimental results are in line with the well-known clinical data on the high incidence of intestinal tumors in patients with chronic ulcerative colitis (148,328,346 ), Crohn’s disease (91,101 ), and some other long-term injuries of intestinal mucosa (foreign bodies, diverticula, fistulae, parasitic lesions, ureterosigmoidostomy, etc.). All these lesions contribute to carcinogenesis but are not associated with it genetically. Hence, they can hardly be considered as precancer (229,231 ). V. The. Kinetics of Intestinal Epithelial Populations in Tumors and during Carcinogenesis
A. DISTURBANCES IN PROLIFERATION A N D DIFFERENTIATION OF CELLSDURING CARCINOGENESIS
Numerous investigators have shown that cells of normal intestinal epithelium both in experimental animals and humans are characterized
EXPERIMENTAL INTESTINAL CANCER RESEARCH
197
by a high rate of proliferation as revealed by a relatively short mitotic cycle. Epithelial cells proliferate within the crypts. As epithelial cells move toward the gut lumen, they acquire the properties of mature differentiated cells but cease to proliferate. The studies of distribution of labeled cells in the microscopically unaltered mucosa of the colon of experimental animals during tumor induction revealed an interesting phenomenon-the widening of the proliferative compartment (~60,240,289,303,3/3,338). This phenomenon was established by detecting labeled epithelial cells 1 hour after labeling with 3H-TdR and mitoses in the upper zones of the crypts and in superficial epithelium. An analysis of curves of labeled cell distribution within the crypt of the descending colon during carcinogenesis revealed the widening of the proliferative compartment in the microscopically unaltered mucosa as early as 1 month after the beginning of DMH treatment (240). These findings point to the development of disturbances in epithelial cell differentiation resulting in the appearance of proliferating cells in those layers of the morphologically unaltered mucosa in which cells do not divide normally and only differentiated cells are to be found. This may result from disturbances in the control of proliferation regulation, followed by the cessation of DNA synthesis in normal tissues, as epithelial cells move toward the surface of intestinal mucosa. It was precisely in those layers of mucosal surface where epithelial cells with disturbed differentiation were detected that primary tumorsin siru carcinoma and superficial cancer-formed at a later stage (beginning from the second month of DMH treatment). These focal lesions were found to consist of cells characterized by an enhanced proliferative activity and disturbed differentiation of cells, as indicated by the localization of DNA-synthesizing cells near the surface of the mucosa, i.e., in the compartment, in which differentiated cells are located in normal tissue (240 ). The kinetics of epithelial cell populations of normal human colon mucosa revealed that the proliferating cells of this segment of the intestinal tract occur in the lower two-thirds of the crypt. As differentiated cells move toward the gut lumen, they cease to proliferate and, finally, are shed. However, it was shown by autoradiographic examinations of operation and biopsy material obtained from patients who suffered from multiple familial polyposis and what was believed to be adenomatous and villous polyps of the colon and rectum that the arrangement of proliferating cells in such tissues is abnormal (48,62-64,66). The cells of the superficial areas of adenomatous and villous polyps were found to incorporate, apart from 3H-leucine, whereas it is normally the bottom twothirds of the crypt of the mucosa that incorporates most of labeled precursors of RNA and protein (62).
198
KAZYMIR M . POZHARISSKI ET AL.
Particular attention has been given to studies of colonic and rectal mucosa adjacent to malignant tissues and yet grossly and microscopically unaltered. It was found that DNA-synthesizing cells are sometimes located in the upper one-third of the crypt and in superficial epithelium of such unaltered tissues ( 4 8 , 6 2 4 4 , 1 5 6 ) .These authors think that an enhanced proliferation of cells in the superficial layers of the mucosa is one of the earliest changes that occur in the mucosa of human colon when polyps are formed. They also believe that the displacement of the proliferative compartment of epithelial cells in intestinal crypts is an important factor of prognosis, indicating the possibility of further formation of polyps. Similar disturbances in the topography of proliferating epithelial cells were detected in the histologically normal colonic mucosa in the relatives of patients suffering from familial intestinal polyposis (65 ). In the seemingly “normal” colonic mucosa of cancer patients and relatives of those suffering from familial polyposis, DNA-synthesizing and proliferating cells are located in the zone in which differentiated cells are normally observed. As these cells move toward the crypt surface, they continue to synthesize DNA and incorporate more 3H-uridine in RNA and 3H-leucine in protein than cells at the corresponding level of the crypt, under normal conditions. Similar biochemical properties are exhibited by the epithelial cells of the superficial layers of colonic neoplasms. The continuation of DNA synthesis in those cells in which it normally ceases is the earliest and most reliable pathological sign of the commencement of neoplastic alterations. Special mention should be made of such enzymes participating in the metabolism of nucleic acids as thymidine kinase, thymidine phosphorylase, and adenine and hypoxanthine phosphoribosyltransferase, which are of interest in connection with the development of lesions in human colon. As the epithelial cells of the intestines in humans and experimental animals move to the upper part of the crypt during normal differentiation of cells, the activity of thymidine kinase is decreased considerably, that of thymidine phosphorylase is reduced to a lesser extent, while the activity of adenine and hypoxanthinephosphoribosyltransferase is increased (159,161,303,307). The development of tumors in human intestines is also accompanied by changes in the activity of the above enzymes. For example, the superficial layer cells of adenomatous polyps, villous papillomas, and carcinomas are characterized by a high content of thymidine kinase, which usually occurs in proliferating cells. Meanwhile, the activity of adenine and hypoxanthinephosphoribosyltransferase, characteristic of cell differentiation processes, progressively decreases in these neoplasms (159,307).
199
EXPERIMENTAL INTESTINAL CANCER RESEARCH
In some cases, a decreased level of enzymatic activity, characteristic of mature cells (159), and a relatively high level of enzymes participating in cell proliferation (268) were observed in the superficial layers of the histologically normal mucosa of human colon, adjacent to tumor tissues. Similar changes in the activity of enzymes, taking part in cell maturation, were found in the morphologically unaltered mucosa of the jejunum of mice treated with DMH (303 >. Comparative studies of the properties of thymidine kinase in the epithelial cells of the intestines of rat embryos, rats of varying ages, and those of DMH-induced tumors suggested that the embryolike properties of tumor enzymes develop before the beginning of tumor formation (270 1. These authors also demonstrated that the activity of ornithine decarboxylase, another enzyme participating in cell proliferation, in the epithelial cells of rat colon increases 50-fold following DMH treatment, as compared with normal tissues. On the basis of a comparative study of the topography of proliferating epithelial cells in patients with solitary polyp, familial polyposis, and colonic carcinoma, and in experimental intestinal tumors induced in animals, Lipkin (160) could distinguish two stages of disturbances in intestinal epithelial proliferation during carcinogenesis. While in phase I , the cells of the grossly and microscopically unaltered mucosa continue to incorporate 3H-TdR as they move toward the gut lumen. The continued proliferation of cells even in superficial epithelium is due to the inadequate suppression of DNA synthesis. However, since the production and loss of epithelial cells are in equilibrium, the overall cell mass of the mucosa is not increased and, therefore, the appearance of the latter is not changed. At phase 2, highly proliferating cells are not shed into the gut lumen: they accumulate in the rnucosa, thus leading to the formation of intestinal tumors. However, the mechanisms responsible for the development of disturbances in cell reproduction are not clear yet. The dynamics of changes in the kinetics of populations of intestinal epithelial cells at the incipient stages of tumorigenesis has been studied in great detail by Pozharisski et ul. (240). The investigation was concerned with the microscopically unaltered mucosa of rat descending colon, examined 1, 2, 3, and 4 months after DMH treatment as well as such early malignant neoplasms arising in this segment of the intestines as in sitri carcinoma, superficial cancer, and slightly invasive, small adenocarcinoma. The evaluation of percentage labeled mitosis (PLM) curves undertaken in this study showed an increase in the duration of the short mitotic cycle from 11 hours under normal conditions to 15-16 hours in the cells of microscopically unaltered mucosa, in sitrf carcinoma, superficial cancers, and small adenocarcinomas. While f, and tG, 4 I,,,
+
200
KAZYMIR M . POZHARISSKI ET AL.
remain practically unchanged, the increase in mitotic cycle duration is due to the increase in GI-period to 3-6 hours, as compared with 2 hours in control. A comparison of PLM curves for microscopically unaltered mucosa and early malignant lesions revealed some changes-a shoulder-in the descending part of the first wave of PLM curve (240). This shoulder becomes increasingly pronounced by month 4 and is particularly well defined on PLM curves for in sit14 carcinoma, superficial cancer, and small adenocarcinoma with incipient invasion. Possible explanations of this shoulder on the descending part of PLM curves are discussed in the report of Pozharisski et a / . , (237,238). The most plausible seems to be the variations in cell population versus mean duration of S-period. The studies of changes in the fraction of labeled epithelial cells in different compartments of the crypt, following multiple injections of 3HTdR, and an evaluation of label cumulation (LC) curves showed that the compartment of maximum cell proliferation in the crypts of normal mucosa of rat descending colon is characterized by a varying mean length of the mitotic cycle of cell population. The latter, in turn, consists of three subpopulations of cells with different mean lengths of mitotic cycle. The bottom of the crypt was shown to comprise a four-component population of cells also made up of subpopulations with different mean lengths of mitotic cycle (237,238). The analysis of LC curves for the microscopically unaltered mucosa of the descending colon of rats, for different terms of DMH treatment, showed that the structure of the proliferating cell population both in the compartment of maximum proliferation and the bottom of the crypt becomes less complicated (240). After 4 months of DMH treatment, 95% of the cells in the cells in the compartment of maximum proliferation of the crypts of the microscopically normal mucosa of the descending colon proliferated through a short (15-16 hour) cycle, while about 5% of the remaining cells either had a long mitotic cycle or were able to enter the resting phase R,. These subpopulations occur in nearly the same ratio in in situ carcinoma, superficial cancer, and small adenocarcinomas. Another indication of neoplastic changes in the intestines is a shift in the ratio of mitotic phases, with the fraction of pathological forms of mitosis increasing (240). For example, prophases constituted 26% in the epithelium of the descending colon in control rats, and metaphases 57%; whereas after 4 weeks of DMH treatment, the fraction of prophases dropped to 6.396, while that of metaphases rose to 74%. There were no significant changes in this ratio at the later stages of the experiment. The epithelium of the descending colon of control rats was found to contain 4% of pathologic mitoses. But their fraction increased up to 51%
EXPERIMENTAL INTESTINAL CANCER RESEARCH
20 1
as early as 1 month after the beginning of the experiment. At later stages, it remained more or less stable within 57% to 60%. Cessation of DMH treatment was not followed by a decline in the number of abnormal mitoses in intestinal adenocarcinomas within the next 2 or 3 months. This observation suggests that the appearance of pathological mitoses was not due to the toxic effect of the carcinogenic substance. Diverse forms of pathologic mitoses (bridges, the lag of chromosomes and their fragments during divergence, chromosome scatter, hollow or three-group metaphases, monocentric and multipolar mitoses, etc.) with an apparent predomination of the lag of chromosomes and their fragments in metakinesis (12%-17%) and K-mitoses were observed for different terms of DMH treatment (in microscopically unaltered mucosa) and in developed adenocarcinomas, too. The iliac epithelium did not show any changes in the ratio of mitotic phases throughout the experiment, nor was the fraction of their pathological forms in excess of 3.5%. These changes in the kinetics of populations of intestinal epithelial cells seem to be manifestations of the malignant transformation of these cells (240 ) .
B. THEPECULIARITIES OF THE KINETICS OF TUMOR CELLS The analyses of tagged cell distribution after pulsed labeling with 3Hthymidine revealed that there is no clear distinction between the compartments of proliferating and nonproliferating cells in colonic neoplasm tissues and, therefore, DNA-synthesizing cells are distributed all over the tumor (139,239,338,348). Thus, the entire tumor is the proliferative zone. The analysis of percentage labeled mitosis curves showed that the mean length of cell cycle of the fastest proliferating subpopulation of cells of invasive tubular adenocarcinoma is 16 hours, as compared with 11 hours for the normal mucosa of the descending colon (239). This increase is due to the prolongation of GI-period, with the values oft, and tG, + f t , being practically the same. The slower rate of proliferation of adenocarcinoma cells, as compared with those in the proliferative compartment of the crypts of normal mucosa, was also reported by other authors (313,348). The evaluation of percentage labeled mitosis curves revealed a heterogeneity of the proliferating cells of adenocarcinoma with respect to the mean duration of S-period (239). Moreover, on the basis of the data showing that the maximum of the first wave of PLM curve remains short of loo%, the authors suggest this also to be an indication of the hetero-
202
KAZYMIR M . POZHARISSKI ET AL.
geneity of tumor cell population with respect to the mean duration of G,period. It may as well indicate even the existence of an R,-subpopulation. The label cumulation curve obtained in this study also points to the increasing heterogeneity of adenocarcinoma cell population with respect to the mean length of mitotic cycle. This curve is identical to that obtained for the crypt bottom of normal descending colon, in which a complicated four-component structure of cell population was established and epithelial stem cells occur (237,238). These findings led these authors to suggest that the properties of tumor cells resemble those of the epithelial stem cells of the descending colon. The similar rates of cell proliferation in adenocarcinoma and the crypt bottoms in the normal mucosa of rat colon were also reported by other authors (313 1. While the determinations of cell cycle parameters from percentage labeled mitosis curves produce leveled indices for separate tumors, the application of the thymidine-colchicine method makes it possible to determine separate parameters of cell cycle for each tumor, both of the same and different histological types (235). Using the latter method, these authors demonstrated that, among such experimental tumors as adenocarcinoma, signet-ring cell, and mucinous cancer, signet-ring cell cancer and adenocarcinoma exhibit the highest proliferative activity (mitotic cycle length, T,, is 47 and 54 hours, respectively), while mucinous carcinoma exhibits the lowest one (T, = 124 hours). These authors give the following explanation for these differences: The cells of signet-ring cancer and glandular structures of adenocarcinoma are located in a highly vascularized stroma so that their requirements in nutrition and oxygen supply are fully satisfied. Mucinous cancer, on the other hand, consists of large lacunas of mucus, in which separate tumor cells or their aggregates are located. It is easy to understand that under such conditions, metabolic processes proceed on a reduced scale in tumor cells. This manifests itself in a varying rate of cell proliferation. The parameters of generation cycle in the tumors of the same histological structure, which formed in one and the same animal, were also found to vary. Marked variations in the magnitude of generation cycle parameters of the same histological type are believed to characterize the rate of cell proliferation, both in the entire tumor and its separate compartments. Variations in the proliferative rate within one and the same tumor were confirmed by the results of the subsequent studies of these authors (Pozharisski and Klimashevski, unpublished data). By means of the thymidine-colchicine method of determination of generation cycle parameters and analysis of the curves of labeled mitoses and those for changes in the fractions of labeled cells, following multiple injections of 3H-TdR into three regions (periphery, center, and zone of invasion) of
EXPERIMENTAL INTESTINAL CANCER RESEARCH
203
tubular adenocarcinoma, it was found that the cells of tumor periphery are characterized by the highest rate of proliferation. A lower rate of cell proliferation was observed in the central part of tumor and in its zone of invasion. The latter finding is rather strange. We should expect that the invasive zone would comprise cells capable of increased proliferation, indispensable for the infiltrative growth of tumor. Two maximum values of label index-near the surface and in the base of the mucosa, at an area of invasive growth into the submucosa-were established in the neoplasms considered to be adenomatous polyps with malignant transformation and incipient infiltrative growth (338). There are very few data available on studies on mitotic cycles in colonic and rectal tumors in humans. This seems to be due to the difficulties involved in taking multiple specimens for further autoradiographic investigations as well as moral considerations. Nevertheless, a few papers contain data on the quantitative parameters of the kinetics of cell populations in human intestinal tumors (2/,25,36,/59,302).Apart from autoradiography, the stathmokinetic method based on the use of the ability of colchicinelike preparations to block cell division at metaphase has been employed for the determination of cell proliferation rates (25.36 ). It is possible to infer from the results of these investigations that the rate of proliferation of tumor epithelial cells is generally lower than that of normal ones. The autoradiographic studies of the kinetic parameters of adenocarcinoma of human colon and rectum, involving the use of double label (3HTdr and 14C-Tdr),showed that DNA synthesis duration in tumor is about twice that of normal mucosa for both sites ( 2 1 ) . These authors attach much importance to this fact and suggest that any increase in the value oft, may be considered sufficient evidence to suspect carcinogenesis. Of similar importance is another finding that a rise in t , magnitude was observed in morphologically normal mucosa with an atypical arrangement of proliferating cells. At the same time, however, the authors are right in saying that more evidence is required to accept the rise in t , as a signal of malignant transformation or precancerous state development. Since the data are scarce, it is impossible to advance any theory on the peculiarities of cell population kinetics in human intestinal neoplasms. Such important factors as the proliferative pool and degree of cell loss were studied in few investigations only and the data obtained vary within a considerable range (36,302). In order to obviate these difficulties in studies of human tumors, the parameters of the mitotic cycle in human colonic and rectal adenocarcinoma were measured, following the transplantation of these tumors to mice, in which immunological reactivity had been suppressed (223). It was established that the lengths of mitotic
204
KAZYMIR M . POZHARISSKI ET A L .
cycle and its separate phases do not differ materially from those determined in relevant tumors of man. However, these authors withhold their judgment on the identity of the kinetic parameters in the cells of human colonic adenocarcinoma and its heterotransplants, owing to the scarcity of literature data on the peculiarities of proliferation in the primary tumors.
C.
STEMEPITHELIAL CELLSOF CARCINOGENESIS
THE
INTESTINES AND THEIRROLEI N
The role of the proliferating cells of organs and tissues in carcinogenesis may be considered well established at the present time (see 214). However, the proliferating cells of tissues may be part of different cell populations undergoing different stages of cell differentiation and, therefore, their contribution to carcinogenesis may be different. Meanwhile, the detection of a factor, responsible for tumorigenesis, on a cell level is vitally important for the elucidation of the nature of tumor growth. The following facts point to the leading role of enterocytes in carcinogenesis (241 ): 1. Intestinal epithelium is a quickly renewing tissue, and the first tumor lesions develop within 2 to 3 months after repeated injections of DMH (230). Since epithelial cells renew several times during this period, a source of carcinogenesis should be sought among more stable elements which can store carcinogenic effects. Stem epithelial cells are probably such elements, because they are located in the crypt bottom (41,238) and, therefore, do not enter the cell flow toward the superficial layers of the mucosa, where they are shed into the gut lumen. DMH treatment affects the hereditary mechanism of stem cells and they produce cells with an impaired heredity. It should be also mentioned that a single administration of DMH is sufficient to induce intestinal tumors. 2. The malignant transformation of enterocytes, as revealed by continued DNA synthesis in cells, which go on dividing outside the proliferative compartment in intestinal crypts, is a rather rare phenomenon. This means that it is actually brought about by a relatively small population of cells, i.e., stem enterocytes, rather than by the bulk of proliferating nonstem cells. 3. In the kinetic parameters of cell populations, the cells of large invasive adenocarcinomas resemble those in the bottom of colonic crypts, where stem cells are located (238). 4. A local stimulation of intestinal epithelium proliferation due to a chronic nonspecific injury of the mucosa results in a sharp rise in exper-
EXPERIMENTAL INTESTINAL CANCER RESEARCH
205
imental tumor incidence at the injured site. An enhanced proliferation of epithelial cells is not likely to be due to a shorter length of mitotic cycle of nonstem cells, because their rate of replication is very high under normal conditions. It is probably due to the entry of a greater portion of the stem cells of the intestinal epithelium into the mitotic cycle. Therefore, it may be inferred that stem enterocytes act as acceptors of carcinogenic influences (236). 5 . This conclusion is further confirmed by the correlation between the rate of tumor incidence in certain segments of the intestines, on the one hand, and the amount of stem cells, their proliferative pool, and the length of their life cycle in these segments, on the other. To illustrate, the descending colon, in which DMH treatment induces tumors in 100% of rats, reveals a greater fraction of stem cells, a higher proliferative pool, and shorter life cycle as compared with the ileum, which is practically insusceptible to tumors (236). 6. The conversion of adenocarcinoma cells into those of signet-ring cell cancer, and retransformation of the latter into the former, support the unitary theory of the origin of all types of epithelial cells (41) and demonstrate that, despite the cytological characteristics of cell differentiation, tumor cells retain stem-like properties. 7. A specific glycoprotein, similar to cancer embryonal antigen (CEA) observed in patients with carcinoma of the colon, was detected in experimental intestinal tumors and blood serum of tumor-bearing animals (216 ). However, this antigen is also found in the blood serum of patients with chronic nonspecific lesions of intestinal mucosa characterized by an intensified proliferation of the epithelium (243 ). As mentioned above, the increase in the fraction of proliferating epithelial cells is caused by the entry of a greater number of stem cells into mitotic cycle. Hence, it may be suggested that CEA is produced by stem cells passing through the mitotic cycle and that the stem cells of intestinal epithelium and the cells of an intestinal tumor possess common antigenic properties. It was suggested recently by Pierce et al. (224), who found the same ultrastructure in least-differentiated cells of normal colon of rats and intestinal tumor transplants in these animals, that stem cells serve as targets in carcinogenesis. On this basis, the mechanism of development of experimental tumors of the colon may be visualized as follows (240). As a result of DMH treatment, stem cells are transformed while in mitotic cycle and the descendants of the altered cells become capable of proliferating in those compartments of crypts in which differentiated epithelial cells are normally located. This means that transformed cells cease to differentiate. They are distributed throughout the entire length of the crypt and even
206
KAZYMIR M . POZHARISSKI ET AL.
reach its mouth. Since stem epithelial cells are very few and their fraction passing through mitosis (i.e., when cells are susceptible to the transforming effect of the carcinogen) is exceedingly small, morphological means fail to detect any malignant alterations at initial stages. It is very likely that some altered epithelial cells are eliminated by immunological mechanisms. Repeated administrations of the carcinogenic substance result in the transformation of a greater amount of stem cells. Subsequently, local disturbances in the steady state of intestinal epithelium occur, which are manifested by the accumulation of atypical cells with altered proliferative properties and formation of in situ carcinoma in the superficial layers of the crypts. The formation of tumor lesions follows as a result of an upset equilibrium between cellular output and loss, because the latter diminishes. As these lesions continue to develop, the lamina propria of the mucosa becomes invaded, and superficial cancer is formed.
VI. Biochemistry
A. NUCLEIC ACIDS, NUCLEOTIDES, A N D ENZYMES PARTICIPATING IN THEIRSYNTHESIS A N D METABOLISM Although there are abundant data on the structure and function of nucleic acids in different experimental tumors, communications on studies of DNA in DMH-induced tumors of the intestines are very scarce (115). Equilibrium centrifugation of the DNA from the cells of mouse colonic tumor revealed only one peak of DNA, which points to the homogeneity of isolated DNA. Its buoyant density corresponded to that of liver DNA (1.700 gm/ml). Much attention is being given to studies of the so-called “second messengers” (cyclic nucleotides), which play an important role in the regulation of mitotic activity, alteration of membrane properties, control of the rates of metabolic reactions, etc. (52). For example, guanylate cyclase-guanosine 3’,5’-monophosphate in the mucosa of rat colon may be activated under the influence of N-methyl-N’-nitro-N-nitrosoguanidine ( 6 0 ) . I n contrast, the level of cyclic AMP in X-ray-induced adenocarcinoma of the jejunum is reduced (298). Similar results are reported for tumors of human colon (59). Adenylate cyclase activity in these tumors was much lower than in normal mucosa. These authors conclude that the low level of cyclic AMP in tumor should be accounted for by a decrease in its synthesis rather than an increase in its degradation. However, the mean level of cyclic AMP in human adenocarcinoma appeared to be lower than in normal tissues, which is consistent with the above experimental findings. In contrast with the data of DeRubertis et al. (5Y), the
EXPERIMENTAL INTESTINAL CANCER RESEARCH
207
results of Minton er al. (188) indicate that cyclic AMP level in colonic tumors does not differ from that of normal mucosa in man. Much more data are available on the role of enzymes participating in nucleic acid synthesis in normal and malignant intestinal tissues. Ball rt al. (14) conducted a study on precancerous lesions, an informative and the only investigation on the subject, in which two enzymes associated with cell division were examined. Ornithine decarboxylase activity is low in the rapidly dividing colonic cells, which are the targets for DMH carcinogenesis, whereas it is rather high in the nondividing cells of the small intestine. DMH administration resulted in an enhanced activity of this enzyme in colonic epithelium but did not affect it in the liver: this points to the specificity of this effect. The study of the other enzymethymidine kinase-revealed that the fetal and tumor enzymes are alike in possessing a high level of activity and an ability to be stimulated by phospholipase C. The properties of the colonic tissues of the rats exposed to DMH range between those of normal and tumor tissues. The activity and regulation of thymidine kinase in normal and malignant intestinal epithelium in humans were studied Pt the same laboratory (269). Unlike in rats, the difference in these characteristics between tumor and fetal intestinal tissues proved to be considerable: in carcinoma, the enzymic activity was higher and was not stimulated by phospholipase C, while in the fetus, it was 7-10 times as high, and in the normal mucosa of adult humans 2.5 times as high. It was inferred from these results that cell division regulation in gastrointestinal tumors is not identical to that of the rapidly dividing cells of the crypts and fetal intestine. The activity of thymidine kinase and thymidine phosphorylase appeared to be maximal in tumors and young proliferating cells, declining as the degree of cell differentiation increased and as cells migrated toward the tissue surface. The activities of adenine- and hypoxanthine phosphoribosyltransferases in tumor and young proliferating cells were minimal, and increased as cells became more differentiated and migrated toward the surface (307). The DMH-induced tumors of mouse colon show an enhanced activity of some enzymes, which play a key role in RNA synthesis: nuclear RNA polymerase, nucleolar RNA polymerase, ATP-tRNA-adenylyltransferase, L-methionine-tRNA-ligase, L-leucine-tRNA-ligase, and L-lysinetRNA-ligase (222 ).
B. NUCLEAR PROTEINS Some noteworthy results have been reported in papers dealing with the studies of nonhistone protein behavior during carcinogenesis and in
208
KAZYMIR M . POZHARISSKI ET AL.
DMH-induced tumors of rat colon. The basic difference was found to consist in a sharp rise in the levels of the proteins with molecular weights of about 44,000 and 62,000 daltons in the course of carcinogenesis (24 ). Selective synthesis of proteins sets in long before the development of any morphological signs of neoplasms. It is likely that the accumulation of these acidic proteins is not simply due to the increase in the fraction of epithelial cells passing through the proliferative phase (22 1. Since the program of cell growth and differentiation under normal conditions and, therefore, that of cell dedifferentiation in tumor tissue are determined to a considerable degree by the interaction of DNA and firmly bound chromosomal proteins, which regulate gene activity, it is easy to appreciate the significance of the observation of a considerable selective accumulation of the same proteins in HT-29 cells, which originate from the adenocarcinoma cells of human colon ( 2 2 ) . Their level in human tumors of the colon is also raised ( 2 2 ) . These proteins appear in the adenocarcinomas which arise in patients suffering from familial polyposis long before any signs of malignant transformation develop (23). This is correlated with similar results in experimental tumors ( 2 4 ) and may be of diagnostic value. The protein with a molecular weight of 62,000 daltons, isolated from rat tumors, does not bind to DNA in vitro, while the other one (44,000 daltons) shows a great affinity for DNA, which points to its association with the active part of the genome (23 ). A two-dimensional electrophoresis showed a considerable difference in the isoelectric points and molecular weights of these proteins (23 1. It was also shown in this laboratory that sodium cyanate selectively inhibits 3H-labeled amino acid incorporation into the cytoplasmic and nuclear proteins of DMH-induced tumors. This effect was not observed in the HT-29 cell culture (5 ).
C. ENZYMES The cell membranes of mammals contain glycoproteins, glycolipids, and, sometimes, glycosoaminoglycans, and fucolipids in the microvilli of intestinal cells. All these components may be responsible for such functions of the cell as adhesion, recognition, “contact inhibition,” agglutination, growth regulation, etc. Hence, cell membranes and their constituents have aroused considerable interest. For instance, La Mont et (11. (149) demonstrated that the malignant transformation of rat intestinal cells is accompanied by changes both in their agglutination by lectins (cells of the tumors of small intestine are agglutinated by concanavalin
EXPERIMENTAL INTESTINAL CANCER RESEARCH
209
A and those of colonic tumors by wheat germ agglutinin) and the activity of glycosyltransferase of cell surface (activity in tumors of small intestine was higher and in colonic tumors lower than in normal homologous organs). Similar results for a number of ,glvcosyltransferases (sialyl-, galactosyl-, fucosyl- and polypeptidyl-N-acetyl galactosaminyltransferases) were obtained by Kim and Isaaks (140) who studied colonic tumors in humans. The same authors reported the absence of any changes in glycosidase activity and reduced levels of fucose, N-acetylglucosamine, N-acetylgalactosamine, and N-acetylneuraminic acid in the membranes of tumor cells. These findings are at variance with the data of Bossmann and Hall (27) who discovered elevated levels of activity of glycosidases, some pro tease^,^ and sialyltransferase. Taking into consideration the use of the same techniques for enzymic activity determinations by the authors of the two latter publications, it is hard to find an explanation for this discrepancy. Unlike Kim and Isaaks (140), Mian and Cowen (184) found that the activity of glycosidases (N-acetyl-8-D-glucosaminidase, N-acetyl-p-Dgalactosaminidase, p-D-galactosidase, and a-L-fucosidase) in the DMHinduced colonic tumors of mice and rats is 2 to 4 times that of normal epithelium. Subsequently, Mian et al. (185) reported that K,'s for these enzymes in tumors are similar to those in embryonic tissue and differ from those in normal intestinal tissue. Moreover, Mian and Nutman (186) showed that the activity of phexosaminidase A in tumors induced by DMH in rats is lower than in normal intestinal epithelium. This observation is supported by the data of Brattain et ol. (29) on a relatively higher level of P-hexosaminidase B and a lowered level of isozyme A in tumors of human colon, while this ratio is reversed in normal tissue. Both isozymes have identical K,,, and V,,,,, although thermal stability and optimal pH are different. Schwartz (275) reported that the activity of enzymes which catalyze the incorporation of N-acetylneuraminic acid, galactose, fucose, and N acetylamine in the glycoproteins of colonic carcinoma in humans is relatively high: and that glycosidase activity in DMH-induced tumors of rats is elevated. These data point to changes occurring in the structure and function of cell membranes. They are further corroborated by the evidence on the altered permeability of the cell membranes of the DMH-induced tumors It should b e mentioned that in spite of this, the levels of inhibitors of proteases in human tumors are either identical to those in normal tissues (acid a,-glycoprotein, aPmacroglobulin, and antithrombin 111) (314 ) or even elevated (a,-antitrypsin, acid a,-glycoprotein, haptoglobin, and prealbumin) (320 ).
210
KAZYMIR M . POZHARISSKI ET AL.
of rat colon (189). The permeability of cell membranes was assessed by the appearance of hexokinase-a key enzyme in glycolysis-in serum. This enzyme is not detected in the serum of intact animals, while, in DMH-treated rats, it was found 1 month after the beginning of carcinogen treatment and reached a maximum at month 5 . Simultaneously with the intensification of enzyme synthesis by tumor cells and its release into the blood, the activity of the isozyme to migrate fastest to the anode disappeared or pronouncedly decreased in experimental tumor. Only one of three isozymes, characteristic of normal intestinal tissue, remained in the metastases of signet-ring cell carcinoma. The K,, in tumors was 2-3 times less than in normal tissues. These results are consistent with the previous data on colonic tumors in humans (206). Changes in the isozyme composition of adenosine deaminase were reported recently by Trotta and Balk (308). The colonic tumors induced in rats by exposure to methylazoxymethanol acetate showed a rise rather than a decrease in the number of isozymes. Moreover, normal and tumor variants of enzymes differed in substrate specificity and responsiveness to the inhibition by 9-erythro(2-hydroxy-3-nonyl)adenine.In the opinion of these authors, the latter fact may suggest application of this preparation for chemotherapy. It is to be noted that, in normal mucosa of the jejunum, there were two enzymes, as in tumor of the colon with molecular weights of 37,000 and 33,500daltons, and PI 4.80 and 4.85, respectively. Pegg (221 ) and Pegg and Hawks (222) revealed an enhanced activity of tRNA-methylases in the DMH-induced tumors of rat colon. Of considerable interest are the results of Fiala et al. ( 9 7 ) who found an increased activity of y-glutamyltranspeptidase in the DMH-induced carcinoma of the colon and the intestines of newly born animals. Due to the simultaneous detection of CEA and an enhanced activity of this enzyme, it became possible to distinguish the metastases of colonic tumor in the liver from local recurrences and other metastases in human patients (2Y5), and the metastases of colonic carcinoma in the uveal tract from a genuine uveal melanoma (187). The mucosa of adenocarcinoma of the jejunum of rats treated with DMH shows a reduced concentration of retinyl ethers and contains no free retinal at all (300), unlike normal tissue of the small intestine. In summary, it may be concluded that the available results of the biochemical studies of experimental intestinal tumors still leave many gaps in the present-day knowledge of the subject. The situation is somewhat better in the field of biochemical peculiarities of human intestinal neoplasms. Experimental investigations in those biochemical parameters, which have been studied in man, seem to have considerable promise. They may, for instance, include the activity and isozymic spectrum of
EXPERIMENTAL INTESTINAL CANCER RESEARCH
211
lactate hydrogenase (301 ), which are generally different in the tumors of other sites, as well as p-glucuronidase and other glycosidases (73 ), important for the etiology of colonic tumors (see Sections IV,A,4 and VII1,A) and polyamines (putrescine, spermidine, spermine) (see 267 ), which play a role in nucleic acid and protein synthesis.
VII. immunology
In 1965, Gold and Freedman (107) found that human tumors of the colon produce an embryo-specific glycoprotein, which was designated as a cancer-embryonal antigen (CEA). Later on, the presence of CEA in patients with tumors of the gastrointestinal tract was confirmed by numerous reports (for references see the reviews 33,102,126,142). Today, CEA is the only biochemical marker for which the U.S. Food and Drug Administration granted a license to Hoffman-La Roche, Inc. for production and sale of kits for radioimmunoassay (179). However, no analog of CEA in experimental animals with intestinal tumors has been found so far; and this fact presents certain difficulties in conducting a detailed study of the pathogenic nature of this phenomenon.
A. ANTIGENS IN CHEMICAL-INDUCED TUMORS OF RAT COLON The earliest investigations involving the use of precipitation techniques detected an antigen in DMH-induced tumors, which was also present in embryonal tissues but was not detected in the intestinal tissues of adult rats. This antigen seems to circulate in tumor-bearing animals (103,104). Subsequently, the antigen was found in spleen extracts and, by immunohistochemical means, in marrow, macrophages, and polymorphonuclear leukocytes, but not in tumor cells; its presence was supposed to be due to tumor infiltration by leukocytes (168,f7 0 ) . The fact that Martin et nl. (170) failed to discover the antigen in perchloric acid extracts indicates that it differs from the one found by Abeyounis and Milgrom ( 2 ) and Okulov and Pozharisski (216) (see below). Moreover, experimental intestinal tumors generally do not reveal any pronounced infiltration by leukocytes, except necrotic foci (228,230). However, such neoplasms often develop in the regions of Peyer’s patches and lymphoid follicles (230,264),and it may be supposed that the antigen, which is not directly associated with tumor elements, was found only because specimens were taken from these areas.
2 12
KAZYMIR M . POZHARISSKI ET AL.
The above investigation of Martin rt al. (170) revealed two antigens in tumor cells and, at a low concentration, in normal intestinal tissues. Finally, when rabbits were immunized by cultured rat adenocarcinoma cells and the antiserum was absorbed in vivo in syngeneic rats, there was revealed a membrane-associated antigen which was independent of the strain of rats, tumor site in the intestine, and carcinogens used for the induction of these tumors (169). These authors suspect the antigen to be tumor-specific and consider it to be a homolog of CEA in patients with colonic tumors. The sera of rats with transplanted intestinal tumors were shown to contain antibodies to this membrane-associated antigen (172). Okulov and Pozharisski (216) and Abeyounis and Milgrom (2) detected an antigen in the perchloric acid extracts of DMH-induced tumors of rats. It appeared to resemble human CEA in some physicochemical and biological properties (thermal stability, glycoprotein composition, immunoelectrophoretic migration, similar to p-globulins, and high level in embryonal and tumor intestinal tissues). The general consensus is that the antigens from experimental and human intestinal neoplasms do not cross-react (2,103,104,170).
B. ANTIGENS I N TUMORS OF
THE
SMALL INTESTINE
A recent investigation of X-ray-induced tumors of the ileum and jejunum revealed an antigen not detected in the tissue homogenates of the liver, kidney, colon, spleen, lungs, urine, and feces of tumor-bearing animals (297). At the same time, its immunological properties were identical to those of protein found in the 17- to 19-day-old rat embryos. This antigen, however, was not detected in irradiated rats, in which tumors failed to be induced (296). In such characteristics as thermal stability and mobility in immunoelectrophoresis similar to p-globulins, it resembles CEA. However, when extracted with perchloric acid, some of its antigenic determinants were hydrolized. They were completely hydrolized when extracted in 0.1 N HCI, but, 0.1 N NaOH did not affect the immunological activity of the antigen (296). The antigen was also observed in blood (83). A superficial tumor-associated antigen was detected in the cells of carcinoma of human small intestine by immunohistochemical means (210). For this purpose, an antiserum obtained by immunization of rabbits with the tissue of well-differentiated adenocarcinoma of human colon was used. This seems to point to the common nature of the antigenic composition of tumors of the large and small intestine.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
213
C. ANTIGENS I N NONSPECIFIC LESIONSOF THE COLON An antigen, similar to the one described above in connection with DMH-induced tumors of the colon (216), was discovered in rat serum immunoprecipitation in agar, with nonspecific injury of the cecum (243 ). This means that the production of a CEA-like antigen in adult organism is not limited to neoplastic tissues only, and it is probably also synthesized in normal intestinal epithelium. This supposition is supported by the literature data showing that, although in small concentration (not more than 2.5 ng/ml), CEA was also found in the serum of healthy subjects (44,126,142). The appearance of the antigen in serum corresponds under experimental conditions to the period of the most enhanced reparative proliferation of intestinal epithelial cells. Since the increase in the proliferative activity of the epithelium is probably due, chiefly, to the entry of stem enterocytes into mitotic cycle (22Y), it may be suggested that CEA is produced precisely by these cells. On the other hand, immunohistochemical data show that CEA is located mainly in the glycocalyx of the most-differentiated cells of adenocarcinoma and normal intestinal mucosa in humans (130). But it seems that CEA is only absorbed on the surface of enterocytes (261), while it is actually synthesized in the cytoplasm of nondifferentiated cells (130). In general, the manner in which CEA is produced and released has very much in common with that of the mucopolysaccharides of intestinal goblet cells (207). The detection of a CEA-like substance in the experimental animals with nonspecific lesions of the intestine may account for the false positive results of the diagnosis of colonic carcinoma in man when CEA is used (44), as well as its high concentration in the blood serum of patients suffering from chronic ulcerative colitis of different etiologies (192,303). D. ANTIGENS I N INTESTINALCARCINOMA A N D ESTROGENS It has been shown recently that the administration of estrogens-propionate diethylstilbestrol-may cause an increased release of a CEA-like substance into blood serum in rats bearing DMH-induced tumors and in animals with chronic nonspecific injury of intestinal mucosa (215). This seems to be due to the presence of estrogen receptors in colonic neoplasms similar to those found in cancer patients (181 ). These data suggest the possibility of administration of estrogens to stimulate an increased release of CEA into blood serum, when it is not detectable in patients with intestinal tumors, for diagnostic purposes (215 ).
214
KAZYMIR M. POZHARISSKI ET AL.
E. CELL-MEDIATED IMMUNITY The cytotoxic action of immune lymphocytes on the cell cultures of colonic tumors induced in rats by N-methyl-N’-nitro-N-nitrosoguanidine, 3,2’-dimethyl-4-aminobiphenyl, and DMH was described by Steele and Sjogren (290). These authors claimed that chemical-induced tumors of rat colon have a common tumor-specific superficial antigen(s). At the same time Steele and Sjogren (291 ) revealed an antigenic cross-reaction between colonic tumors and fetal intestine, which points to the presence of a fetal organospecific antigen (or antigens) in carcinoma. A suggestion was made that tumor cells produced neoantigens which were the products of reexpressed fetal genes (291). Moreover, it was shown that lymphocytes in rats with X-ray-induced tumors of the small intestine exerted a cytotoxic effect on the cells of small bowel adenocarcinoma ( 8 2 ) . The cytotoxicity of the cells of lymph nodes obtained from rats with colonic tumors regarding target cells showed a sharp decrease on addition of a soluble fraction of cytoplasmic proteins and solubilized tumor membranes prepared by treatment of the cells with 3 M KCl or papain (293),as well as the serum of tumor-bearing animals (290,291,293,294). It is to be noted that the cytotoxic effect of lymphocytes and the blocking activity of the serum of rats with intestinal tumors become apparent before or at the earliest stage of tumor detection by doublecontrast X-ray examination (294). The blocking effect of serum was not observed immediately after the removal of tumors, unless any recurrences developed (294 ). Sjogren and Steele (281) give the following description of the tissue and embryonal specificities of antigens in the chemical-induced tumors of the colon: The former include the cytotoxic action of the cells of lymph nodes and blood lymphocytes in rats with intestinal tumors on the cells of intestinal carcinoma. They do not, however, exert this effect on the cells of other tumors or adult animals. The cytotoxic effect of lymph node cells in tumor-bearing animals on carcinoma cells was specifically inhibited by the serum of tumor-bearing animals. The cytotoxic effect of lymphocytes and the blocking activity of serum become manifest before or at the earliest stage of colorectal cancer detection. The blocking activity of serum ceases immediately on removal of primary tumors. The embryonal specific characteristics include the cytotoxic effect of the cells of lymph nodes of multiparous (more than three litters) pregnant rats on those of colonic carcinoma but not on the cells of adult animals. The lymphocytes of tumor-bearing rats are cytotoxic on the cells of fetal gut but not on those of fetal kidney, liver, lungs, or adult rats, intestinal mucosa included. The serum of pregnant rats can block the cytotoxic
EXPERIMENTAL INTESTINAL CANCER RESEARCH
215
action of lymphocytes on fetal intestine and intestinal carcinoma. Apart from the intestine-specific embryonal antigen, there are the so-called widely distributed embryonal antigens. Antigens expressed on the cells of fetal intestine are observed in a solubilized tumor material. In 1972, Hellstrom and Hellstrom (118) demonstrated that the lymphocytes of patients with colonic tumors are cytotoxic with respect to tumor cells in vitru but do not damage the normal epithelial cells of the intestines of the same patient. The lymphocytes from patients with other neoplasms as well as from healthy donors do not damage colonic tumor cells in vitro. However, both autochthonous and allogeneic lymphocytes of patients with colonic tumor can destroy tumor cells. On the basis of these data, it was concluded that tumors of human intestines have a common antigen. It was also shown that the lymphocytes of patients suffering from tumors of the colon can destroy the epithelial cells of fetal gut but not those of fetal kidneys or normal intestinal tissue of adult humans. The response of peripheral blood lymphocytes of patients with colonic carcinoma to phytohemagglutinin in vitro is markedly decreased (183 ). On removal of tumor the level of growth stimulation is restored. These authors suggest the following explanations: (1) the decline in the ratio of circulating T and B lymphocytes, which reflects the quantitative rather than qualitative anomaly of lymphocytes: (2) the production in patients of a factor, capable of interacting with “receptor sites for phytohemagglutinin” on lymphocyte surface: and (3) the rate and duration of the action of tumor-specific antigens on lymphocytes may change the ability of these cells to respond to other exogenous factors. However, each of these suppositions should be experimentally tested. In conclusion, owing to absolutely different approaches available at present, it is difficult to compare antigens which contribute to humoral and cell-mediated immunity developing in cases of colonic carcinoma, although, according to Martin et a / . (171), they may be identical. On the other hand, Hellstrom and Hellstrom (118) showed that lymphocytes obtained from patients with colonic tumors inhibit the growth of the cells of intestinal tumors but not those of the normal tissues of other organs or tumors of other sites. Therefore, the antigen (or antigens) which contributes to antitumor immunity seems to be strictly organspecific. However, CEA, nonspecific cross-reacting antigens and membrane-associated tissue antigens do not possess such a specificity. All these facts pose problems, the solution of which is of vital importance for tumor therapy: Why does colonic carcinoma progress quickly in v i w , though its cells may be killed by the lymphocytes of cancer patients in vitro? This may be due to the action of the above blocking elements of serum, the nature of which has not yet been identified.
216
KAZYMlR M. POZHARISSKI ET AL.
Therefore, a better immunotherapeutic effect in the management of tumors should be, probably, achieved by reducing the blocking activity of serum rather than by bolstering antitumor cell-mediated immunity (118).
VIII. Metabolism of 1,2-Dirnethylhydrazine and Related Substances
The fact that DMH and homologous substances induce tumors in the intestine only but never at the sites of administration suggests that these agents undergo certain metabolic changes before they exert their carcinogenic effect. On the basis of analogy with the metabolism of nitrosocompounds and some speculation, Druckrey and co-workers (75,247) suggested the following mechanism of DMH metabolism in a living organism: At first, it is oxidized, probably, without participation of enzymes, to azomethane (AM). This reaction is catalyzed by traces of heavy metals (247). Azomethane is further oxidized to azoxymethane (AOM). Then azoxymethane is subjected to enzymatic a-hydroxylation to form methylazoxymethanol (MAM). MAM may, in turn, be dealkylated by means of concerted reactions, forming an alkylating metabolite-methyldiazonium. Since the metabolic transformations of AOM and dimethylnitrosamine are similar, according to the above authors, it might be expected that they should produce similar carcinogenic effects. In actuality, they are characterized by a totally different organotropism, and the above mechanism does not account for these differences. Preussmann rr ul. (247) and Druckrey (75) believe that such differences in organ specificity are due to the presence of specific hydrolases in a target organ. In an attempt to explain the organ-specific carcinog :nic effect of DMH treatment, Weisburger (333) suggested that MAM is conjugated with glucuronic acid in the liver and then is transported in this form, together with bile, to the intestinal tract, where it is freed again by microbial enzymes and forms the final alkylating metabolite, which interacts with the macromolecules of enterocytes. The data available on the carcinogenicity of cycasin, a glycoside of MAM, which exhibits a similar organotropism, point to the likelihood of such a mechanism responsible for the organotropism of DMH. A. EXPERIMENTAL EVIDENCE ON DMH METABOLISM The experimental data supporting the above hypothesis on the metabolism of DMH and homologous compounds, capable of the induction of
EXPERIMENTAL INTESTINAL CANCER RESEARCH
217
intestinal tumors, may be divided into two groups. One of them includes the evidence on the detection of the postulated metabolites of DMH in experimental animals. Numerous studies of DMH metabolites involving the use of labeled carcinogenic agents have been carried out recently. For example, the air exhaled by l4C-DMH-tieated rats was shown to contain some AM, in addition to CO, (92,94,95):its level varied, depending on the dose. When 21 mg/kg of DMH were administered, exhaled air contained 14% of injected radioactivity, as AM (94,95).As DMH dosage was increased to 200 mg/kg, the level of AM in exhaled air rose considerably (up to 23% of injected radioactivity) (95 ) , Simultaneously, the fraction of exhaled CO, was found to decrease in such animals considerably-from 11%16% to 4% (95,116).This seems to show incomplete metabolism of DMH when it is administered in high doses. While AM is detected in exhaled air immediately after DMH administration and most of it is excreted within 4 to 5 hours, COPbegins to be exhaled at a later stage ( 9 5 ) . Such a delay in CO, formation is probably due to a complete oxidation of DMH metabolites or the induction of relevant enzymes ( 9 5 ) .The rate of DMH metabolism in mice seems to be higher, because the treatment with the same dosage of this substance results in CO, formation in an amount twice that in rats (116). Apart from AM and C 0 2 , the air exhaled by DMH-treated rats was found to contain monometh ylhydrazine and methylamine. A total of 45%-60% of radioactivity was shown to be excreted with air within 24 hours ( 9 6 ) . When 21 mg/kg of 3H-DMH was injected into rats, 3.7% of radioactivity was excreted in bile during 24 hours (245). However, when 14C-DMH was administered to rats, only 0.4%-0.9% of radioactivity went to bile, irrespective of carcinogen dosage (96,f16 ). All radioactivity was found in unmetabolized DMH when thin-layer chromatography was used (116 ). However, other investigators revealed different metabolites of DMH in rat bile. For example, chromatographic determinations on Sephadex LH-20 detected six different metabolites of DMH, including AOM, in bile ( 9 6 ) . Since AOM is eluted together with MAM in this chromatographic procedure, high-pressure liquid chromatography was carried out, which showed bile to contain AOM also. Unconjugated MAM was not detected. Apart from AOM this chromatographic assay showed some other peaks which probably represented MAM glucuronides and sulfates ( 9 2 ) . Another study was concerned with determinations of AOM, DMH, and AM in bile ( 9 4 ) . As much as 66% of radioactivity went to urine in the rats treated with 3H-DMH (245). Sephadex LH-20 chromatography detected DMH, MMH, and AOM in urine: 12% to 15% of radioactivity (14C-DMH)was excreted with these compounds (96 ). High-pressure liq-
2 18
KAZYMIR M. POZHARISSKI ET AL.
uid chromatography revealed the presence of DMH, AM, AOM, and MAM in urine ( 9 4 ) . MAM was also found in the urine of the AOMtreated rats ( 9 3 ) . Hence, the results of these studies confirm the hypothesis of Druckrey, Preussmann, and Weisburger on the metabolic pathways of DMH, because all the substances supposed to be DMH metabolites were detected in the body after administration of this carcinogenic agent. The ability of the substances supposed to be DMH metabolites to induce intestinal tumors selectively is another point in support of the above hypotheses on metabolic pathways. For instance, tumors of this localization are invariably induced by AOM, MAM, and a related carcinogenic substance-cycasin (75,247). On the other hand, carcinogenic agents, which belong to hydrazo-, azo-, and azoxycompounds but are not metabolized to intermediary products identical to those of DMH, do not induce intestinal tumors. For example, 1,2-diethylhydrazine and its derivatives (azo- and azoxyethane) exhibit a specificity for other organs, inducing tumors of the nervous system, mammary gland, liver, and leukemia (75). Preussmann E t al. (247) claim that intestinal carcinogenesis is caused by those hydrazo compounds, which metabolize, forming a methylcarbonium ion. It is natural, therefore, that such substances as I-methyl-2ti-butylhydrazine and 1-methyl-2-benzylhydrazine should induce both neurinomas and intestinal neoplasms (247). On the basis of the above data on DMH metabolism and its reactivity, Pozharisski et al. (245) suggested a scheme of metabolism of this agent, which is intended to complement and detail the available concepts of Preussmann et al. (247), Weisburger (333), and Druckrey (75). According to this scheme, after DMH has found its way into the circulation, it is dehydrogenated to form AM. Subsequently, this product may undergo the three following changes: (I) homolytic degradation to nitrogen and methyl radical: (11) N-oxidation, forming N-oxyazomethane (AOM): or (111) a-C-hydroxylation (Fig. 3). There is some evidence which precludes the first possibility. For instance, the macromolecules of enterocytes fail to be labeled following total hepatectomy (234). It is significant that tumors do not arise at the sites of DMH administration and in the lungs, i.e., at the site where most of AM is excreted. Moreover, there are data demonstrating that the enzymic systems of the liver participate in the metabolism of DMH and its derivatives (93,247). The other two pathways of azomethane oxidation in biological systems are equally probable: Two maxima of label incorporation into the DNA and protein of different organs recorded at different times after injection
219
EXPERIMENTAL INTESTINAL CANCER RESEARCH
homolyaie
1,E-Dimethylhydrazine
Yethylazomethanol
II
.
N-oxidation1 in liver CH N=NCHJ trena- and cis-Nethylazoxymethane
3'
CH,,~NCH~OH trenm- end cia-NethylO"N-esoxymeth.no1
0
r(
C-hydroXY1etion W
h*temlyeis
-Hcwo
rCHjn.nOn
-pmo2
Yethyldiuo, hydroxide
uthylnitruine
Alk-1-CH2COOH I
ma3
1Jsthyl-3-lkyl-3cuboxymthyl-trlezene
FIG.3. Possible metabolic pathways of DMH. *Able to form glucuronides. From K . M. Pozharisski et ul. (245 ).
of 3H-DMH (234) suggest that DMH metabolism may occur by these two pathways simultaneously, though, probably, at different rates. a-C-hydroxylation of azomethane (Pathway 111) may result in the formation of methylazomethanol which may degrade to formaldehyde and methyldiimine, the latter, probably, forming a methyl radical, as a result of homolysis. When DMH is metabolized, forming methylazoxymethane with its subsequent hydroxylation (Pathway 11), unstable products-derivatives of MAM-are formed. The possibility of identification of the site at which this reaction takes place is confirmed by the results of in virro experiments which demonstrated hydroxylation of AOM, forming MAM, to occur due to hepatic homogenates and microsomes in the presence of NADPH. At the same time, this reaction failed to take place on addition of homogenates or microsomes of colonic mucosa (93,94). MAM derivatives are subsequently converted to methyldiazohydrate. Since these products form, as a result of DMH metabolism in the liver, while tumors arise mostly in the intestine, it was assumed that the liver is the site of the formation of DMH metabolites, which are released in target organ, as a result of the action of its specific factors. The most likely in this connection seems to be the formation of glucuronides, which synthesize a carcinogenic metabolite, probably, methylazoxymethanol, in the presence of bacterial P-glucuronidase. Recent investigations have revealed that tissue enzymes participate in MAM metabolism, because
220
KAZYMIR M . POZHARISSKI ET AL.
the cytosols of the colon and cecum and, particularly, that of the liver reduce NAD to NADH in the presence of MAM acetate as a substrate. Meanwhile the cytosol of the jejunum and ileum, which are resistant to MAM, exhibits a low NAD-dependent alcohol dehydrogenase activity in the presence of MAM (109). These findings point to MAM being metabolized by alcohol dehydrogenase-like enzymes and suggest that NADdependent enzymic reactions, occurring in tissue cytolsol, may determine the organotropism of the carcinogenicity of MAM in rats also. These results further support the concepts of Preussmann et al. (247) and Druckrey (75) who think that MAM can be dealkylated spontaneously. Most of the labeled metabolites of DMH (96%)enter the intestine with bile (245),while as little as 4% of all radioactivity detected in the intestine is delivered there by blood flow. Nevertheless, tumors were also induced in isolated fragments of the intestine (227). Therefore, it is from the circulation that carcinogenic metabolites penetrate into the intestinal wall. As mentioned above (Y2), AM and AOM were found in bile, in addition to DMH, while other metabolites of this agent have not yet been identified. Pozharisski et al. (245) suggested that, apart from glucuronides, other synthesized in the methyl group-carrying substances-triazenes-are liver. These substances may be the products of a reaction of methyldiazonium and protein amino acids. These authors believe that triazenes are likely, by alkylation, to be responsible for tumor formation in germfree rats (233,257) as well as in the duodenum, in which bacterial flora are scare. It may as well be attributed to the effect of P-glucuronidase, which is contained in intestinal epithelium at high concentration. The metabolism of DMH, forming AOM (Pathway II), and a-C-hydroxylation of AM (Pathway 111) would result in different end productsa carbonium ion (CH,+) or a methyl radical (CH3.). They are different in chemical nature and reactivity. Since Pathway I11 of DMH metabolism produces a derivative of monomethyl hydrazine, which is a poor alkylator and does not induce intestinal tumors (116), it is most likely that the carcinogenic effect of DMH is due to its metabolism by Pathway 11, forming a carbonium ion. It was shown quite recently that, following a pretreatment with disulfiram which blocks N-oxidation of AM (94), the bases of DNA were not alkylated, while the label from 3H-DMH was incorporated into unmethylated purines ( 1 5 7 ~ )These . findings prove that DNA is methylated by the carbonium ion alone, rather than by the methyl radical. Thus, on the basis of the available data, Pozharisski et ul. (245) distinguish the following stages of DMH metabolism, leading to the initiation
EXPERIMENTAL INTESTINAL CANCER RESEARCH
22 1
of carci nogenesis : (a) formation of “active” metabolites (methylazoxymethanol or methyldiazohydrate) in the liver; (b) binding of the metabolites to glucuronic acid: (c) delivery of conjugates to the intestine via blood flow: (d) liberation of “active” metabolites due to enzymic activity of intestinal flora (P-glucuronidase): (e) formation of carbonium ion; and (f) specific methylation of enterocyte macromolecules.
B . THEEFFECTOF CERTAIN MODIFYING FACTORS ON DMH METABOLISM The suppression of the carcinogenic effect of DMH treatment in mice by administration of disulfiram (331) has stimulated studies of the mechanism of this phenomenon. A pretreatment with disulfiram (94) or some of its metabolites-diethyldithiocarbamate, bis(ethylxanthogen), and carbon disulfide-to rats before injection of 14C-DMH(94 ) appeared to result in a trebled level of azomethane in exhaled air, while carbon dioxide concentration fell by 65%-80%, as compared with the animals treated with DMH alone. AOM and MAM concentrations in the urine of these rats either dropped considerably or were not detected at all. The levels of 14C incorporation into the liver, kidneys, lungs, spleen, and colonic mucosa decreased by 43%-70%. As a result, Fiala rt a / . (94 ) concluded that these substances block the N-oxidation of AM and, therefore, prevent its conversion to AOM. However, the pretreatment with disulfiram or carbon disulfide also blocked the oxidation of the AOM administered to rats, which manifested itself in a complete suppression of carbon dioxide formation within 6-8 hours. The concentration of AOM in exhaled air, however, was rather high. Moreover, disulfiram and carbon disulfide reduced the level of MAM and raised that of AOM in the urine of these rats (93). The decrease in the carbon dioxide level following the administration of disulfiram and its metabolites was probably due to the diminished formation of formaldehyde, a decline in its oxidation, or a combination of both these processes ( 9 4 ) . Therefore, disulfiram and its derivatives not only block the N-oxidation of AM but inhibit further metabolism of AOM. It is most likely that DMH metabolism becomes blocked by disulfiram and homologous compounds as a result of the inhibition of the activity of the microsomal cytochrome P-450-dependent oxygenases.
222
KAZYMlR M . POZHARlSSKl ET AL.
Unlike disulfiram, a preliminary administration of phenobarbital to rats, followed by subcutaneous injection of AOM, resulted in a 1.5-fold increase in the level of I4CO2exhaled, while the level of MAM excreted with urine increased tens of times. At the same time, the amount of unmetabolized AOM in urine decreased ( 9 3 ) . All these data point to an intensified metabolism of the carcinogenic substance, following the administration of phenobarbital. Chrysene exerted a still more pronounced effect (93 1. IX. Interaction of 1,2-DimethyIhydrazine and Related Compounds with Cell Components
On the basis of the above data on the pathways of DMH metabolism, leading to the formation of an alkylating metabolite, one might expect a priori that DMH treatment of animals would result in the alkylation of different macromolecules. The alkylating potential of DMH was confirmed by Hawks et al. (f 1 7 ) , who by Dowex-50 chromatography of hydrolyzates of liver and colon DNA and RNA of mice treated with DMH 6 or 24 hours before, detected 7-methylguanine in all samples studied. Hawks and Magee (116) found that within 6 hours after injecting 15 mg/kg of 14C-DMH into mice, 7methylguanine was present in the nucleic acids of all organs studied. Its level ranged from 1.5% (small intestine and lungs) to 4%-5% (large intestine, kidneys, and spleen) of that in liver DNA. In the RNA from the same tissues, the levels were similar. The distribution pattern in the rats sacrificed 6 hours after an injection of 200 mg/kg of the carcinogenic substance was somewhat different: 7-methylguanine in intestinal DNA and RNA was more than 50% and, in the nucleic acids of kidney tissue, about 12% of that of liver. The rate of alkylation of nucleic acids in mice appeared to be higher than in rats, and like nitrosamines, DMH methylated more RNA than DNA. No alkylated products other than 7-methylguanine were found in this study. Although these investigations confirm the alkylating potential of DMH, they do not account for the organotropism of DMH. Both in rats and mice, maximal alkylation was observed in liver nucleic acids, while tumors are induced by this agent mostly in the colon. To establish possible correlations between the organotropism of DMH and its interactions with DNA, RNA, and proteins, a time-course study was conducted (234).In all tissues under study, the specific radioactivity of DNA appeared to have two maxima: the first peak of different segments of the intestines (duodenum, small intestine, ascending and de-
EXPERIMENTAL INTESTINAL CANCER RESEARCH
223
scending colon) was observed within 1 hour and that of liver and kidney DNA 3 hours after carcinogen treatment. Subsequently, the radioactivity level decreased followed by a rise by hour 9. By hour 12, there was a sharp fall in specific radioactivity level in the DNA of all tissues, followed by a gradual decline until day 3 after exposure. By day 7, the radioactivity rose again, but by four weeks after 3H-DMH treatment, no tissues contained any radioactivity. It was suggested that the rise in radioactivity of DNA between days 3 and 7 was due to the incorporation of methylated precursors from the cell pool into the newly synthesized DNA, which is consistent with other observations. In all RNA samples, radioactivity increased within the first 6-9 hours. By hour 12, there was a sharp fall, followed by a gradual decrease. Between days 3 and 7, radioactivity, except for the liver and small intestine, remained practically unchanged. Like DNA, this may be due, probably, to incorporation of labeled precursors. Radioactivity was negligible by week 4. Throughout the entire experiment, radioactivity levels in the DNA and RNA from the liver and kidneys were higher than those of different intestinal segments. The radioactivity in total protein was maximal within 1 hour after treatment with the carcinogen and its levels in all segments of the intestine were much higher than in parenchymatous organs. By hour 3, the protein radioactivity fell in all organs, but at hour 9, it increased in protein of the intestines. Within the period from hour 12 to day 7 , protein radioactivity remained practically unchanged and it could not be detected 4 weeks after treatment. In control rats injected with a substance homologous to DMH-3H-1,2diethylhydrazine (3H-DEH), no label was detected in either DNA, RNA, and protein of the intestine, liver, or kidneys. However, radioactivity levels in all fractions of tissues of DEH-target organs (thymus, spleen, and brain) were high. Thus, the experiments of Pozharisski et al. (234) did not reveal any correlation between the radioactivity level of the macromolecules of different organs of rats and their susceptibility to the carcinogenic effect of DMH. These results as well as those of Hawks et al. (117), Hawks and Magee (116), and Shimitzu and Toth (276), who autoradiographically revealed a maximum incorporation of label from 14C-DMH by mouse hepatocytes rather than enterocytes, fail to explain why DMH-induced tumors arise predominantly in the colon. Pozharisski et al. (234) suggested that in those organs in which DMH treatment fails to induce tumors a “nonspecific” methylation or participation of the methyl groups of 3H-DMH in the normal metabolism of nucleic acids occurs. “Specific”
224
KAZYMIR M. POZHARISSKI ET AL.
methylaction probably occurs in the tissues in which DMH induces tumors predominantly (colon). It has recently become obvious that the level of alkylation of DNA guanine in the 7 position is not correlated with the organotropism of the carcinogenicity of many alkylating agents. On the other hand, there is much evidence that the carcinogenicity of alkylating agents may be caused by 0 6-methylguanine formation in DNA. It may be the persistence of 06-methylguanine in DNA rather than the initial level that is actually responsible for its carcinogenic effect (153,165 ). Loveless (163 ) suggested that, being a potential mutagenic base, 0 6-methylguaninecan block base pairing with cytidine and cause the misincorporation of thymine during DNA replication (see 153). This effect was confirmed in in vifro experiments, using the RNA-dependent RNA polymerase and an 0 %nethylguanine-containing template. The product formed in this system incorporated uracil, an analog of thymine but not cytosine, whereas 7methylguanine formed a normal pair with cytosine. As a result, 0 6 methylguanine may cause a mutation (see 153). Likhachev et al. (157) carried out a detailed study of the alkylated products of the DNA of different tissues of rats killed 3 hours after 3HDMH injection. Similarly to the results of Hawks and Magee ( 1 1 6 ) , the highest level of 7-methylguanine was found in liver DNA, a lower level in colon DNA (14% of that in the liver), and levels in the DNA of kidneys, testes, lungs, and small intestine of 1.7%-3.2% of that in liver. At the same time, the DNA of all organs incorporated radioactive products of DMH breakdown, as a result of de nova synthesis of its precursors. The uptake of radioactivity in adenine varied greatly and the highest level was detected in the adenine of the DNA of the duodenum, in which alkylated purines were not found at all. Therefore, the specific radioactivity of total DNA cannot be used as an indication of the degree of alkylation. Besides 7-methylguanine, the DNAs of the liver and colon had 0 6 methylguanine, which points to a high potency of DMH as an alkylator, like MNU and DMNA (153). Likhachev et ul. (157) showed that the 0 6 methylguanine to 7-methylguanine ratio in the DNA of the target organthe colon-to be more than four times that in the DNA of the liver, in which tumors do not arise after a single exposure to DMH. Unlike in other organs, the level of 06-methylguanine of the colon DNA did not vary within 3 days (190), and O6-rnethylguanine was not detected at all in the mucosa of the ileum, where practically no tumors arise. Since the initial 0 6-methylguanine to 7-methylguanine ratio of DNA is determined by the type of the carcinogen used but not by the target organ (153), this
EXPERIMENTAL INTESTINAL CANCER RESEARCH
225
evidence demonstrates that the colon’s ability to eliminate 0 fi-methylguanine from DNA is much lower than that of the liver. Hence, the results support the hypothesis of Pozharisski et ul. (245) that although the levels of total radioactivity of DNA in different segments of the intestine are similar, the “specific” alkylation of DNA occurs in the colon only. The findings of Likhachev et al. (157) were confirmed by experiments of Rogers and Pegg (266 ), who studied the alkylation of rat liver, kidney, and colon DNA at intervals after intraperitoneal or subcutaneous administration of 14C-DMH.After intraperitoneal administration of the carcinogenic agent, the formation of 7-methylguanine and 0 6-methylguanine was observed in the DNA of all organs under study. Moreover, liver DNA contained small amounts of 1-methyladenine, 6-methyladenine, 3methyladenine, and 3-methylguanine. The highest levels of alkylated bases were detected in all three organs 6 hours after DMH treatment. At hour 48, 0 fi-methylguaninelevel in liver DNA dropped by nearly twothirds, while that of 7-methylguanine by half. 7-Methylguanine was eliminated from kidney DNA at the same rate as from liver DNA, while Ofimethylguanine level dropped by only 40% within the same period. The level of 7-methylguanine in colon DNA decreased fourfold, while that of 0 s-methylguanine threefold. Thus, within 48 hours after intraperitoneal injection of DMH, the 0 s-methylguanine to 7-methylguanine ratio in liver DNA decreased, whereas it grew in the DNA of the kidneys and colon. A somewhat different situation developed after subcutaneous injection of I4C-DMH. While the 7-methylguanine level in rat liver DNA was identical to that after intraperitoneal administration, it was twice as high in kidney DNA, and somewhat lower in colon DNA. During a 48-hour interval after subcutaneous injection of DMH, there was a rise in Ofimethylguanine level in kidney and colon DNA, indicating that DNA methylation proceeded throughout the experiment, although exhalation of radioactivity and, therefore, DMH metabolism, actually ceased in these rats by hour 24. The 0 fi-methylguanineto 7-methylguanine ratio increased in the DNA of all three organs within 48 hours. However, it rose only 1.3-fold in liver DNA, but 3.4-fold in kidney DNA and at least 9-fold in colon DNA. After DMH administration by both routes, 7-methylguanine and 0 fimethylguanine levels in liver DNA were about 10-fold higher than in the DNA of the kidneys and colon. Therefore, the 0 fi-methylguanine to 7-methylguanine ratio in the DNA of the target organ (colon) was higher than those of other organs. The fact that the 0 6-methylguanine level in kidney and colonDNA was
226
KAZYMIR M . POZHARISSKI ET A L .
increasing during the whole 48-hour period after exposure to DMH, when the latter had been excreted, is interpreted by us as a manifestation of either a continued liberation of an alkylating metabolite from the corresponding glucuronide or an incorporation of alkylated precursors into the newly synthesized DNA. Unfortunately, Rogers and Pegg (266 ) were not concerned with a comparative study of DNA methylation in the small intestine, in which tumors fail to be induced by DMH. As a consequence, it is impossible to adequately appraise the role of DNA alkylation in the tropism of carcinogenesis of DMH in different parts of the intestine. Unlike the data reported by Likhachev et al. (157), Rogers and Pegg (266) found that the level of alkylation of colon DNA and, particularly, that of 06-methylguanine was one order higher than methylpurine level in liver DNA. This discrepancy may be due to the fact that Rogers and Pegg (266 ) determined the alkylation of purines of the DNA isolated from the entire mass of the colon, but not from its mucosa, which is the target tissue for DMH and constitutes only a small portion of the whole mass of the intestinal wall. Some recent reports demonstrate that the alkylation by DMH may be studied in cultured tissues of humans and animals. For instance, after incubation of cultured human colon with 14C-DMH,the labeling of DNA and protein was much higher than with 14C-dimethylnitrosamine and, particularly, 14C-benzo[a]pyrene. After a 24-hour incubation with DMH, 3-methyladenine, 7-methylguanine, and 0 6-methylguanine were found in colon DNA (11 ) and also in DNA after a 24-hour incubation of slices of human bronchi in the presence of DMH. 0 s-Methylguanine level was the highest-three times that of 7-methylguanine (114). DMH metabolites also bind to macromolecules of the epithelium of rat intestine cultured i n vitro (10). Thus these tissues of human and animal intestines and bronchi contain enzymes indispensable for the synthesis of an alkylating metabolite of DMH, which methylates the DNA of these organs. In conclusion, it should be mentioned that apart from 0 6-alkylguanine formation, alkylation of other positions in the purine and pyrimidine bases of DNA may also play a role in carcinogenesis (153,279). For instance, 0 4-methylthyminewas also shown to form a mispair with guanine, instead of adenine, during DNA synthesis in vitro (I ). X. Conclusion
The study of various aspects of intestinal carcinogenesis has become possible owing to the development of relevant experimental models. Considerable advances have occurred in the study of morphology and
EXPERIMENTAL INTESTINAL CANCER RESEARCH
227
morphogenesis of intestinal cancer, which contribute to the elucidation of the mechanisms of human colon carcinogenesis. It has been shown that the malignant potency of adenomatous polyps was exaggerated; however, local lesions of intestinal mucosa, which promote carcinogenesis, though not determined genetically, have been detected. Experimental intestinal tumors are finding a wide application in the studies of the immunology and kinetics of tumor cell populations, opening up new vistas in the explorations of the nature of tumor growth. Such animal models are being used for the improvement of X-ray (32,260) and clinical (70) diagnosis of intestinal tumors. The available data show that the pathogenesis of experimental intestinal tumors is a complex of processes, taking place at all levels-from the molecule to the body, including multistage metabolism of the carcinogen involved. This metabolism is determined, to a considerable degree, by the enzymic function of the liver, specific transport of carcinogenic metabolites to the target organ, and enzymic activity of intestinal bacterial flora, terminating in the alkylation of DNA bases, which persist for a long time. Of vital importance are the cell types, which interact with the end metabolite, and the phase of their life cycle. Moreover, such factors as genetic characteristics, age and hormonal status of the organism, and diet also contribute.
ACKNOWLEDGMENT The authors consider it their pleasant duty to thank Prof. N . P. Napalkov for his unflagging interest to this study and valuable advice, Prof. H. Druckrey for discussions on certain aspects of the problem, and Mr. I. K. Arkhipov for assistance in the preparation of the manuscript for publication.
REFERENCES 1. Abbott, P. J., and Saffhill, R. (1977). N!4ct. Acids Res. 4, 761-769. 2. Abeyounis, C. J., and Milgrom, F. (1976). J. Irnrnunol. 116, 30-34. 3 . Abrams, G. D., Bauer, H., and Sprinz, H. (1963). Lab. Invest. 12, 355-364. 4 . Ackerman, L. V., and Spratt, J. S., Jr. (1963). Gastroenterology 44, 905-909. 5 . Allfrey, V. G., Boffa, L. C., and Vidali, G. (1977). Cancer 40, 2692-2698. 6 . Andreassen, J. C. (1%9). Actu Chir. Scand. Suppt. 396, 123-169. 7. Aries, V . , Crowther, J. S.,and Drasar, B. S. (1969). Gut 10, 334-335. 8 . Armstrong, B., and Doll, R. (1975). In?. J . Cancer 15, 617-63 I . 9. Asano, T., Pollard, M., and Madsen, D. C. (1975). Proc. SOC. Exp. B i d . Med. 150, 780-785. 10. Autrup, H., Hanis, C. C., Stoner, G. D., and Fugaro, S. (1977). In vitro 13, 192. 11. Autrup, H . , Hams, C. C., Stoner, G . D., Jesudason, M. L., and Trump, B. F. (1977). J . Nail. Cancer Inst. 59, 351-354.
228
KAZYMIR M. POZHARISSKI ET AL.
l 2 . Baker, R. K. (1953). Cancer Res. 13, 137-140. 13. Balish, E., Shih, C. N., Croft, W. A., Pamukcu, A. M., Lower, G., Bryan, G . T., and Yale, C. E. (1977). J . Natl. Crrncer fnst. 58, 1103-1 106. 14. Ball, W. J., Jr., Salser, J. S . , and Balis, M. E. (1976). Cancer Res. 36, 2686-2689. I S . Bansal, S. C., Mark, R., Bansal, B., and Rhoads, J. E. (1978). J . Natl. Crrncer I n ~ t . 60, 667-675. 16. Barbolt, T., Ringwood, N., and Abraham, R. (1977). Tuxirol. Appl. Phannrrcol. 41, 157. 17. Barkla, D. H., and Tutton, P. J. M. (1977). C m c e r Res. 37, 262-271. 18. Barthold, S. W., and Jonas, A. M. (1977). Cancer Res. 37, 4352-4360. 19. Bielschowsky, F. (1946). Br. J . E x p . Path. 27, 135-139. 20. Birbeck, M. S . C., and Dukes, C. E. (1963). Proc. R. Soc. Med. 56, 793-797. 21. Bleiberg, H., and Galand, P. (1976). Ccrncer Res. 36, 325-328. 22. Boffa, L. C., and Allfrey, V. G. (1976). Cancer Re.?. 36, 2678-2685. 23. Boffa, L. C., and Allfrey, V. G. (1977). Cuncer 40, 2584-2591. 24. Boffa, L. C., Vidali, G., and Allfrey, V. G. (1976). Exp. Cell Res. 98, 396-410. 25. Bone, G., Camplejohn, R. S. (1975). Am. J . Procrol. 26, 65-76. 26. Bonser, G. M., Clayson, D. B., and Jull, J. W. (1956). Br. J . Cajicer 10, 653-667. 27. Bosmann, H. B., and Hall, T. C. (1974). Proc. Natl. Acad. Sci. U . S . A . 71, 18331837. 28. Bralow, S. P., Gruenstein, M., Meranze, D. R., Bonakdarpour, A., and Shimkin, M. B. (1970). Cancer Re.\. 30, 1215-1222. 29. Brattain, M. G., Kimball, P. M., and Pretlow, T. G., 11. (1977). Crrncer Res. 37, 731735. 30. Broitman, S. A., Vitale, J. J., Vavrousek-Jakuba, E., and Gottlieb, L. S. (1977). Clrncer 40, 2455-2463. 31. Burkitt, D. P., (1974). Dis. Colon Rectrun 17, 456-461. 32. Burkle, G . , von. (1975). Fortschr. Geh. Riientgemtr. Nrikleurmrd. 122, 352-364. 33. Burtin, P. (1974). Cancer 34, 829-834. 34. Calman, K. C. (1974). Gut 15, 552-554. 35. Campbell, R. L., Singh, D. V., and Nigro, N. D. (1975). Cancer Res. 35, 1369-1371. 36. Camplejohn, R. S . , Bone, G., and Aherne, W. (1973). B r r . J . Cuncer 9, 577-582. 37. Carter, S. (1976). Semin. Oncol. 3, 433-443. 38. Castleden, W. M. (1977). Br. J . Cuncer 35, 491-495. 39. Castleman, B., and Krickstein, H. I. (1962). N. Engl. J . Med. 267, 469-475. 40. Castleman, B., and Krickstein, H. I. (1966). Ga.strcietitero/og~51, 108-1 12. 41. Cheng, H., and Leblond, C. P. (1974). Am. J. Ancrf. 141, 537-562. 42. Chlopin, N. G. (1946). “Biological and Experimental Fundamentals of Histology.” Publ. House of the U.S.S.R. Academy of Sciences. 43. Chomchai, C., Bhadrachari, N., and Nigro, N. D. (1974). Dis. Colon Rectum 17,3103 12. 44. Chu, T. M., Reynoso, G., and Hansen, H. J. (1972). Ncrtrire (London) 238, 152-153. 45. Cleveland, J. C., and Cole, J. W. (1969). Cancer 23, 1200-1203. 46. Cleveland, J. C., Litvak, S. F., and Cole, J. W. (1967). Cancer Res. 27, 708-714. 47. Cole, J. Quoted by Wynder, E. L., Kajitani, T., Ishikawa, S . , and Takano, A. (1969). Cancer 23, 1210-1220. 48. Cole, J . W., and McKalen, A. (1963). Crrncer 16, 998-1002. 49. Colvert, J. R.. and Brown, C. H. (1948). Am. J . Med. Sci. 215, 24-32. 50. Corbett, T. H., Griswold, D. P., Jr., Roberts, B. J., Peckham, J. C., and Schabel, F. M., Jr. (1975). Cancer Res. 35, 2434-2439.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
229
51. Cox, A. J., Wilson, R. H., and DeEds, F. (1947). CuIrcer Res. 7, 647-657. 52. Criss, W. (1974). O / i c d o g y 30, 43-80. 53. Cuatico, W., Cho, J.-R., and Spiegelman, S. (1974). Proc. N o t / . A c d . Sci. U . S . A . 71, 3304-3308. 54. Cunningham, M. P., and Wilhoite, R. (1973). Dis. Co/o,7 Recrrrni 16, 145-148. 55. Czernobilsky, B., and Tsou, K-C. (1968). C o w e r 21, 165-177. 56. Daland, E. (1954). N . &a/. J. Med. 251, 664-669. 57. Deger, G. E. (1976). Anu. Intc~rn.Mid. 85, 204. 58. Denk, H., Holzner, J . H., and Obiditsch-Mayer. I. (1975). Verli. Dtsch. Ccs. Ptrthol. 59. Tug. Stuftgurt 490. 59. DeRubertis, F. R., Chayoth, R., and Field, J . B. (1976). J. C/i/r./ / w e s t . 57, 641-649. 60. DeRubertis, F. R., and Craven, P. A. (1977). Cancer 40, 2600-2608. 61. Deschner, E. E. (1974). Cuncer 34, 824-828. 62. Deschner, E. E., and Lipkin, M. (1970). J . N u t / . Cancer I n s t . 44, 175-187. 63. Deschner, E. E., and Lipkin, M. (1976). C h . Gus/roe/r/ero/.5, 543-561. 64. Deschner, E. E., Lipkin, M., and Solomon, C. (1966). J. N u t / . Cu~icc~v I r r s t . 36, 849858. 65. Deschner, E. E., Long, F. C., and Katz, S. (1975). Am. J. D i g . Dis. 20, 418-424. 66. Deschner, E. E., Winawer, S. J., Long, F. C., and Boyle, C. C. (1977). Crrrrwr 40, 2625-263 1. 67. Dilman, V. M. (1971). Luncer 1, 1211-1219. 68. Dilman, V. M.. Sofronov, B. N., Anisimov, V. N., Nasarov, P. G., Lvovich, E. G . , and Polushina, L. I. (1977). Vopr. Oticol. 23, 50-54. 69. Diwan, B. A., Meier, H., and Blackman, K. E. (1977).J . N u t / . CUMPT/ m i . 59, 455458. 70. Dixon, M. F., Cowen, D. M., and Cooper, E. H. (1975). Bionirdicine 23, 247-252. 71. Doll, R., Muir, C., and Waterhouse, J . (eds.) (1970). "Cancer Incidence in Five Continents." UICC, Geneva. 72. Double, J. A., and Ball, C. R. (1975). Cuncczr C h ~ m o f h ~Rrc.p . 59, 1083-1089. 73. Double, J. A., Cooper, E. H., and Goligher, J. C. (1977). Biomcdici/rc. 27, 11-13, 74. Drasar, B. S . , and Jenkins, D. J. A. (1976). Am. J. C/;/t.Nritr. 29, 1410-1416. 75. Druckrey, H. (1972). 111 "Topics in Chemical Carcinogenesis. 2nd International Symposium of the Princess Takamatsu." Univ. of Tokyo Press, Tokyo. 76. Druckrey, H., and Lange, A. (1972). Fed. Proc. 31, 1482-1484. 77. Druckrey, H., Preussmann, R., Ivankovic, S., and Schmahl, D. (1967). Z. Krebs,forsch. 69, 103-201. 78. Druckrey, H., Preussmann, R., Matzkies, F., and Ivankovic, S. (1967). Nrrtrir~~,is.senschuften 54, 285-286. 79. Druckrey, H . , Steinhoff, D., Preussmann, R., and Ivankovic, S . (1964). Z. Krrbsjbrsch. 66, 1-10. 80. Dunn, T. B. (1965). I n "Carcinoma of the Alimentary Tract" (W. J . Burdette, ed.), pp. 45-54. Univ. of Utah Press, Salt Lake City, Utah. 81. Dunphy, J. E., Patterson, W. B., and Legg, M. A. (1959). Ann. Srirg. 150, 488-498. 82. Englund, C. W., Stevens, R. H., Chan, S. W., and Osborne, J. W. (1976) Radicrt. Res. 67, 568. 83. Englund, C . W . , Stevens, R. H., Cheng, H. F., and Osborne, J. W. (1975). Rudiut. Res. 62, 603. 84. Enker, W. E., and Jacobitz, J. L. (1976). Surgery 80, 164-170. 85. Evans, I. A., and Mason, J. (1965). Nature (London) 208, 913-914.
230
KAZYMIR M . POZHARISSKI ET A L .
86. Evans, I. A,, Widdop, B., Jones, R. S., Barber, G. D., Leach, H., Jones, D. L., and Mainwaring-Burton, R. (1971). Biochem. J. 124, 28P. 87. Evans, J. T., Hauschka, T. S., and Mittelman, A. (1974). J . Nail. Cancer Inst. 52, 999- 1000. 88. Evans, J. T., Lutman, G., and Mittelman, A. (1972). J. M e d . 3, 212-215. 89. Evans, J. T., Mittelman, A., and Hauschka, T. S. (1975). Proc. Am. Assoc. C u ~ c e r Res. A m . Soc. Cliii. Oncol. 16, 160. YO. Evans, J . T., Shows, T. B., Sproul, E. E., Paolini, N. S., Mittelman, A., and Hauschka, T. S. (1977). Cancer Res. 37, 134-136. 91. Farmer, R. G., Hawk, W. A., and Turnbull, R. B., Jr. (1971). Cancer 28, 289-292. 92. Fiala, E. S. (1975). Cancer 36, 2407-2412. 93. Fiala, E. S. (1977). Cuncer 40, 2436-2445. 94. Fiala, E. S . , Bobotas, G., Kulakis, C., Wattenberg, L. W., and Weisburger, J. H. (1977). Biochem. Phurmtrcol. 26, 1763-1768. 95. Fiala, E. S., Kulakis, C., Bobotas, G., and Weisburger, J. H. (1976). J . Natl. Cuncer Inst. 56, 1271-1273. 96. Fiala, E. S., and Weisburger, J. H . (1975). Toxicol. Appl. Pharnracol. 33, 178. 97. Fiala, S., Fiala, A . E . , Keller, R. W., and Fiala, E. S. (1977).Arch. Grsc1iwrtlsrfor.sch. 47, 117-122. 98. Filipe, M. I. (1975).Br. J . Cuncer 32, 60-77. 99. Filipe, M. I., and Branfoot, A. C. (1974). Cancer 34,282-290. 100. Fischer, E. R., and Sharkey, D. A. (1962). Cancer 15, 160-170. 101. Frank, J. D., and Shorey, B. A. (1973).Gut 14, 120-124. 102. Freedman, S. 0. (1972).J . Allergy C1iIi. Immrrnol. SO, 348-357. 103. Garmaise, A.B-K., Rogers, A., Newberne, P. M., Saravis, C. A., Kupchik, H . 2.. and Zamcheck, N. (1974). Nurure (London) 248, 706-707. 104. Garmaise, A.B-K., Rogers, A. E., Saravis, C. A., Zamcheck, N., and Newberne, P. M. (1975).J. Natl. Cancer I / i s t . 54, 1231-1235. 105. Gennaro, A. R., Villanueva, R., Sukonthaman, Y., Vathanophas, V., and Rosernond, G. P. (1973). Crrncer Ros. 33, 536-541. 106. Georgsson, G., and Vigfusson, H. (1973).Actu Vet. Scaird. 14, 392-409. 107. Gold, P., and Freedman, S. 0. (1965).J. Exp. Med. 122, 467-481. 108. Goldin, B. R., and Gorbach, S. L. (1976).J. Nut/. Cuncer Inst. 57, 371-375. 109. Grab, D. J., and Zedeck, M. S. (1977). Fed. Proc. Fed. Am. Soc. Exp. B i d . 36, 348. 110. Grinnel, R. S . , and Lane, N. (1958). Irrt. Ahstr. Srirg. 106, 519-538. 1 1 1 . Haase, P., Cowen, D. M., Knowles, J. C., and Cooper, E . H. (1973). Br. J . Cunccv 28, 530-543. 112. Haenszel, W., and Correa, P. (1973). [ A s . Colon Rectum 16, 371-377. 113. Harber, P. S. (1975). Schweb. Mcd. Wochenschr. 105, 564-569. 114. Harris, C. C., Autrup, H . , Stoner, G. D., McDowell, E. M., and Trump, B. F. (1977). Cancer Res. 37. 2309-23 11. 115. Hawks, A., Farber, E., and Magee, P. N . (197U1972). Chem.-Bio/. Ititeruct. 4, 144148. 116. Hawks, A., and Magee, P. N. (1974). Br. J. Cuncw 30,440-447. 117. Hawks, A., Swann, P. F., and Magee, P. N . (1972). Biocheni. Pliurmucol. 21, 432433. 118. Hellstrom, I., and Hellstrom, K. E. (1972). Dis. Colotr Rectum 15, 100-105. 119. Higginson, J . (1967). 111 "Natl. Cancer Inst. Monograph," Vol. 25, pp. 191-198. Washington, D.C. 120. Hill, M. J. (1975). Crrnwr Ros. 35, 3398-3402.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
23 1
121. Hill, M. J., and Aries, V. C. (1971). J . Pathol. 104, 129-139. 122. Hirono, I., Fushimi, K., Mori, H., Miwa, T., and Haga, M. (1973). J. Nail. C ~ n c e r Inst. 50, 1367-1371. 123. Hirono, I., Sasaoka, I., Shibuya, C., Shimizu, M., Fushimi, K., Mori, H., Kato, K.. and Haga, M. (1975). Gann Monogr. Cancer Res. 17, 205-217. 124. Hirose, F., Fukazawa, K., Watanabe, H., Terada, Y., Fujii, I., and Otsuka, S. (1977). Gunn 68, 669-680. 125. Hollmann, K. H. (1964). Arch. Ma/. Appiir. Dig. 53, 975-994. 126. Holyoke, D., Reynoso, G., and Chu, T. M. (1972). Ann. Siirg. 176, 559-563. 127. Hornburger, F., Kerr, C. S., and Hsueh, S. S. (1971). Narirre (London) New Biol. 234, 28-29. 128. Horava, A., and Haam, E., von. (1958). Cancer Res. 18, 764-767. 129. Horn, R. C. (1971). Cuncer 28, 146-152. 130. Huitric, E., Laumonier, R., Burtin, P., Kleist, S., von, and Chanavel, G., (1976). Lab. Invest. 34, 97-107. 131. Hultborn, K. A. (1954). Acia Radio/. (Stockholm) Siippl. 113, 1-71. 132. Iatropoulos, M. J., Golberg, L., and Coulston, F. (1975). Exp. M o l . Pathol. 23, 386401. 133. Imai, H., Saito, S., and Stein, A. A. (1965). Gustroenterology 48, 188-197. 134. Ioachim, N.-J., Delaney, W. E., and Madrazo, A. (1974). Cancer 34, 586-596. 135. Jacobs, M. M., Jansson, B., and Griffin, A. C. (1977). Cuncer Lerr. 2, 133-138. 136. Jimenez, C. J. M., Diener, K. A . , Alvarez, E. E . , Elizabeth, Q., and Maldonado, M. (1971). Ani. J . Proctol. 22, 301-307. 137. Joshi, S . R., Rice, J . M., Wenk, M. L., Roller, P. P., and Keefer, L. K. (1977). J . Nail. Cancer Insf. 58, 1531-1535. 138. Kanazawa, K., Yarnamoto, T., and Sato, S. (1975).Jupun. J . Exp. Med. 45,439-456. 139. Kikkawa, N . (1974). Med. J . Osaha Univ. 24, 293-314. 140. Kim. Y. S . , and Isaaks, R. (1975). Cancer Res. 35, 2092-2097. 141. King, E. S . J . , and Varasdi, G. (1959). Aiist. N . 2. J . Surg. 29, 38-53. 142. Kleist, S., von. (1977). Z. Krehsjbrsch. 90, 1-11. 143. Knudson, A. G . , Jr. (1973). A d v . Cancer Res. 17, 317-352. 144. KorpBssy, B., and Tiboldi, T. (1957). Oncologici 10, 330-335. 145. Koudstaal, J., Makkink, B., and Overdiep, S. H. (1975). Eirr. J. Cancer 11, 105-109. 146. Koudstaal, J., Makkink, B., and Overdiep, S. H. (1975). Eirr. J . Cunccv 11, 111-115. 147. Kroes, R., Berkvens, J. M., and Weisburger, J. H. (1975). Cancer Res. 35, 26512656. 148. Kronman, B. (1971). Cancer 28, 82-88. 149. L a Mont, J . T., Weiser, M. M., and Isselbacher, K. J. (1974). Cuncer Res. 34, 32253228. 150. Laqueur, G. L. (1965). Virchows Arch. 340, 151-163. 151. Laqueur, G. L., and Spatz, M. (1975). Cann Monogr. 17, 189-204. 152. Laumonier, R. (1975). Biomedicine 22, 358-367. 153. Lawley, P. D. (1974). Miifui. Re.\. 23, 283-295. 154. Leaver, D. D., Swann, P. F., and Magee, P. N. (1969). Br. J . Ciinci,r 23, 177-187. 155. Lee, J. A. H . (1976). I n f . J. Epidemol. 5, 187-194. 156. Lieb, L. M., and Lisco, H. (1966). Cancer Res. 26, 733-740. 157. Likhachev, A. J . , Margison, G. P., and Montesano, R. (1977). Chem. Biol. Interoct. 18, 235-240. 1 5 7 ~ .Likhachev, A. J., Petrov, A. S., Prvanova, L. G., and Pozharisski, K. M. (1978). But/. Exp. B i d . Mrif. 86 (121, 679-681.
232 158. 15Y. 160. 161.
162. 163. 164. 165. 166. 167. 168. 16Y. 170.
KAZYMIR M . POZHARISSKI ET AL.
Lingeman, C . H., and Garner, F. M. (1972). J. Noil. C o w e r I n s / . 48, 325-346. Lipkin, M. (1971). Cancer 28, 38-40. Lipkin, M. (1974). Cancer Srtppl. 34, 878-888. Lipkin, M., Bell, B., Stalder, G., and Troncale, F. (1971). In "Carcinoma ofthe Colon and Antecedent Epithelium" (W. J. Burdette, ed.), pp. 213-221. Thomas, Springfield, Massachusetts. Lisco, H., Brues, A. M., Finkel, M. P., and Grundhauser, W. (1947). Ctiticer R e s . 7, 721. Loveless, A. (1969). Ntrrrrrc. (London) 223, 206-207. Lowenfels, A. B., and Sonni, A. (1977). C a m w L e t / . 3, 83-86. Magee, P. N . , Montesano, R., and Preussmann, R. ( 1976).A m . Coric.er Soc. Moriogr.. S e r . 173, 491-625. Marks, S., and Sullivan, M. F. (1960). Noirrrc (Lomfon)188, 953. Marsden, J. R., and Dawson, J. M. P. (1974). Giii 15, 783-787. Martin, F. (1976). J. Nrril. Cuwcer Insi. 56, 212. Martin, F., Knobel, S . , Martin, M., and Bordes, M. (1975). Ctrticcv R e s . 35, 333-336. Martin, F., Martin, M. S., Bordes, M., and Knobel, S . (1975). / t i t . J . Crrncer 15, 144-
151. 171. Martin, F., Martin, M. S. , and Knobel, S . (1975). Proc. I n / . Ctrnccr C o ~ t g r ./ / / / I , F lo r i~n c1~~ Y74 I , 300-303. 172. Martin, F., Martin, M., Lagneau, A., Bordes, M., and Knobel, S. (1976). Crtncc,r R P ~ .
36, 3039-3042. 173. Martin, M. S., Bastien, H., Martin, F., Michiels, R., Martin, M. R., and Justrabo, E. (1973). BionitdiciniJ 19, 555-558. 174. Martin, M. S., Martin, F., Justrabo, E., Michiels, R., Bastein, H., and Knobel, S .
(1974).J . N a i l . Ctrncer I n s t . 53, 837-840. 175. Martin, M. S., Martin, F., Justrabo, E., Michel. M. F., and Lagneau, A. (1977). Git/
18, 232-235. 176. Maskens, A. P. (1976). C a ~ t c e rR e s . 36, 1585-1592. 177. Matsuyama, M., [to, M., Yamada, S., Nakamura, T., and Nagayo, T. (1976).J. Ntrrl. Ct//iccJrI / i s t . 56, 791-795. 178. Matsuyama. M., Nakamura, T., Suzuki, H., and Nagayo, T. (1975). Ctrriii Moriogr.
17, 269-281. 179. Maugh, T. H. (1977). StYcvice 197, 544. 180. McCall, D. C . , and Cole, J . W. (1974). Cancer, 33, 1021-1026. 181. McClendon. J. E., Applely, D., Clandon, D. B., Donegan, W. L., and DeCosse, J . J. (1977). Arch. Srrrg. 112, 240-241. 182. McGinty, F., Deliges, G., and Harrison, D. (1973). G I / / 14, 502-505. 183. Mclllmurrey, M. B., Gray, M., and Langman, M. J . S. (1973). Girt 14, 541-544. 184. Mian, N., and Cowen, D. M. (1974). Br. J . Cancer 29, 438-446. 185. Mian, N., Herries, D. G., Cowen, D. M., and Batte, E. A. (1975). Biochim. Biop1t.v.s. A c / a 391, 179-188. 186. Mian, N., and Nutman, C. A. (1975). Br. J. Ctrncer 32, 252-253. 187. Michelson, J . B., Felberg, N. T., and Shields, J . A. (1977). Arch. Oph//ictlnrol. 95,
692-694.
P.,Matthews, R. H.. and Wisenbaugh, T . W. (1976). J. N u t / . C t r m w / U S / . 57, 39-4 I . 189. Monakhov, N . K., Pozharisski, K. M., and Schwartzman, A. L. (1975). Bull. E,vp. B i d . M e t l . 79, 53-56. /YO. Montesano, R., Margison, G. P., and Likhachev, A. J. (1977). //7 "Pathophysiology 188. Minton, J.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
233
of Carcinogenesis in Digestive Organs” (E. Farber et u I . , eds.), pp. 221-231. Univ. of Tokyo Press and Univ. Park Press, Tokyo and Baltimore. 1Y1. Moon, R. C., and Fricks, C. M. (1977). Cancer 40, 2502-2508. 192. Moore, T., Kantrowitz, P. A . , and Zamcheck, N. (1972). J . AMA 222, 944-947. 193. Mori, H., and Hirono, I. (1977). Br. J. Cuncer 35, 369-371. 194. Mori, K., Ichii, S., and Shigeta, Y. (1955). Cunn 46, 631-634. 195. Moms, H. P., Wagner, B. P., Ray, F. E., Snell, K. C., and Stewart, H. L. (1961). N u t / . Cancer Inst. Morrogr. 5, 1-53. 196. Morris, H. P., Wagner, B. P., Ray, F. E., Stewart, H. L., and Snell, K. C. (1963). J . N u t / . Cancer Inst. 30, 143-161. 1’97. Morson, B. C. (1976). ”Histological Typing of Intestinal Tumours.” International Histological Classification of Tumours. No. 15, World Health Organization, Geneva. 198. Nachlas, M. M., and Hannibal, M. J. (1961). Siwg. Gynec. Obstet. 112, 534-542. I Y Y . Narisawa, T., Magadia, N. E., Weisburger, J. H., and Wynder, E. L . (1974). J . N u t / . Cuncer I n s f . 53, 1093-1097. 200. Narisawa, T., Reddy, B. S . , Wong C.-Q., and Weisburger, J. H . (1976). Cancer Res. 36, 1379-1383. 201. Narisawa, T., Sato, T., Haykawa, M., Sakuma, A., and Nakano, H. (1971). Gunn 62, 231-234. 202. Narisawa, T . , and Weisburger, J. H. (1975). Proc. Soc. Exp. Biol. Mrd. 148, 166-169. 2 W . Narisawa, T., Wong, C.-Q., Maronpot, R. R., and Weisburger, J. H. (1976). Cancer Res. 36, 505-510. 204. Narisawa, T . , Wong, C.-Q., and Weisburger, J. H . (1976). Gunn 67, 41-46. 205. Navarrete, A., and Spjut, H. J. (1967). Cnncer 20, 1466-1472. 206. Neifakh, S. A., Monakhov, N. K., and Schwartzman, A. L. (1972). Dokl. AN U.S.S.R. 206, 743-746. 207. Neutra, M., and Leblond, C. P. (1966). J . C d Biol. 30, 119-136. 208. Newberne, P. M., and Rogers, A. E. (1973). J. N u t / . Cancer I n s t . SO, 439-448. 20Y. Newberne, P. M., and Suphakarn, V. (1977). Cuncer 40, 2553-2556. 210. Newman, R., Bennett, S. J., and Pascal, R. R. (1975). C a n c i v 36, 2016-2019. 211. Nigro, N. D., Bhadrachari, N., and Chomchai, C. (1973). Dis. Colon Rectum 16,438443. 212. Nigro, N. D., Campbell, R. L., Gantt, J . S., Lin, Y. N., and Singh, D. V . (1977). Cuncer Res. 37, 3198-3203. 213. Nigro, N. D., Singh, D. V., Campbell, R. L., and Pak, M. S. (1975). J. N u t / . Cuncer Inst. 54, 439-442. 214. Oehlert, W. (1973). Cell Tissue Kinet. 6, 325-335. 215. Okulov, V. B., Anisimov, V. N., and Pozharisski, K . M. (1977). Bull. E.rp. B i d . M d . 83 (5), 582-583. 216. Okulov, V. B., and Pozharisski, K. M. (1975). B i d / . Exp. B i d . Mid. 80 (11). 78-81. 217. Osswald, H., und Kriiger, F. W. (1969). Arineiwr. Forsc/i. 19, 1891-1892. 218. Otto, H. F., Winkler, R., and Heitmann, C. (1976). Z . K r c h f o r s c l r . 87, 333-342. 21Y. Pamukcu, A. M., and Price, J . M. (1969). J . Nut/ Cunccr Inst. 43, 275-281. 220. Pamukcu, A. M., Yalqiner, S., and Bryan, G. T. (1977). Crincer 40, 2450-2454. 221. Pegg, A. E. (1974). Biockcm. J. 137, 239-248. 222. Pegg, A. E., and Hawks, A. M. (1974). Bioc/rc,m. J. 137, 229-238. 223. Pickard, R. G., Cobb, L. M., and Steel, G. G. (1975). Br. J. C n n c ~ r31, 36-45. 224. Pierce, G. B., Nakane, P. K., Martinez-Hernandez, A., and Ward, J . M. (1977). J. Nntl. C i i m ~ i I~nrs t . 58, 1329- 1345. 225. Potet, F., and Soullard, J . (1971). Girt 12, 468-482.
234
KAZYMIR M . POZHARISSKI ET AL.
226. Pozharisski, K. M. (1972). V o p r . Oticol. I8 ( I ) , 64-71. 227. Pozharisski, K. M. (1973). V r s t n . Chit-. 111 (8), 52-57. 228. Pozharisski, K. M. (1973).In ”Pathology of Tumours in Laboratory Animals” (V. S. 22Y. 230. 231. 232. 233. 234. 235. 236. 237. 238. 23Y. 240. 241.
242. 243. 244. 245. 246. 247. 248. 249. 250. 25l. 252. 253. 254. 255.
Turusov, ed.), Vol. I, pt. 1 , pp. 119-140. IARC Scientific Publication N o 5, Lyon. Pozharisski, K. M. (1975). Cnt~crrRrs. 35, 3824-3830. Pozharisski, K. M. (1975). J. Natl. Cuncrr Inst. 54, 1115-1 135. Pozharisski, K. M. (1978).Arch. Pathol. 40 ( 5 ) . 76-84. Pozharisski, K. M . , and Anisirnov, V. N. (1975). Pathol. Physiol. I , 47-50. Pozharisski, K. M., Dushkin, V. A., and Podoprigora, G. I. (1974). Bull. E x p . Biol. Med. 78 (lo), 81-84. Pozharisski, K. M., Kapustin, Yu. M., Likhachev, A. J., and Shaposhnikov, J. D. (1975). Int. J . Cuticer IS, 673-683. Pozharisski, K . M., and Klimashevski, V. F. (1974). Exp. Puthol. 9, 88-98. Pozharisski, K. M., Klimashevski, V. F., and Gushchin, V. A. (1975). Do/,/. AN U . S . S . R . 220 ( l ) , 216-219. Pozharisski. K. M., Klimashevski, V. F., and Gushchin, V. A. (1977). Tsitologiiu 19, 303-317. Pozharisski, K. M., Klimashevski, V. F., and Gushchin, V. A. (1977). Tsitologiirr 19, 318-328. Pozharisski, K. M., Klimashevski, V. F., and Gushchin, V. A. (1977). Tsitologiiu 19, 537-544. Pozharisski, K. M., Klimashevski, V. F., and Gushchin, V. A. (1977). Tsitologiiu 19, 768-780. Pozharisski. K. M., Klimashevski, V. F., Gushchin, V. A., and Okulov, V. B. (1977). I n “The Role of Stem Cells in Leucoso- and Carcinogenesis” (P. E. Kavetsky, ed.), pp. 21-24. Naukova Dumka, Kiev. Pozharisski, K. M., and Kolodin, V . I. (1975). Vopr. Oncol. 21 ( 5 ) . 80-86. Pozharisski, K. M., and Okulov, V. B. (1976). B d l . Exp. B i d . Med. 81, 467-470. Pozharisski, K. M., and Savostjanov, G. A. (1976). Vopr. Oncol. 22 ( I ) , 59-68. Pozharisski, K. M., Shaposhnikov, J. D., Petrov, A. S., and Likhachev, A. J. (1976). 2. Krehs:forsch. 87, 67-80. Pretlow, T. P., Glover, G. L., and Pretlow, T. G., 11. (1977). J . N u t / . Cuncer I n s t . 59, 98 1-987. Preussmann, R., Druckrey, H., Ivankovic, S., and Hodenberg, A., von. (1969). Ann. N . Y . A C N ~Sci. . 163, 697-716. Qizilbash, A. H. (1974). Arch. Puthol. 98, 118-121. Qizilbash, A. H. (1974). Uis. Colon Rectum 17, 540-549. Reddy, B. S., Narisawa, T., Maronpot, R., Weisburger, J. H., and Wynder, E. L. (1975). Cancer Res. 35, 3421-3426. Reddy, B. S., Narisawa, T., Vukusich, D., Weisburger, J . H., and Wynder, E. L. (1976). Proc. Soc. Exp. Med. 151,237-239. Reddy, B. S., Narisawa, T., and Weisburger, J. H. (1976). Cancer Res. 36, 28742876. Reddy, B. S.. Narisawa, T., and Weisburger, J. H. (1976). J . N u t / . Cuncer Inst. 57, 567-569. Reddy, B. S., Narisawa, T., Weisburger, J. H., and Wynder, E. L. (1976). J . Natl. Cancer Inst. 56, 441-442. Reddy, B. S., Narisawa, T., Wright, P., Vukusich, D., Weisburger, J. H., and Wynder, E. L. (1975). Cancer Res. 35, 287-290.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
235
256. Reddy, B. S . , Watanabe, K., Weisburger, J. H., and Wynder, E . L. (1977). Cancer Res. 37, 3238-3242. 257. Reddy, B. S., Weisburger, J . H., Narisawa, T., and Wynder, E. L . (1974). Cancer Rrs. 34, 2368-2372. 258. Reddy, B. S . , Weisburger, J . H., and Wynder, E . L. (1974). J. Narl. Cancerlnst. 52, 507-51 1 . 259. Reuber, M. D., and Thomas, A. (1973). Proc. A m . Assoc. Cancer Res. 14, 5 5 . 260. Reznik, G., Eckel, H., Green, U . , Rippel, W., and Mohr, U. (1976). 2. Krebsforsch. 87, 321-326. 261. Rogalsky, V. J . (1975). J. Nutl. Cancer Inst. 54, 1061-1071. 262. Rogalsky, V. J., Konovich, E. A., Rivkin, V. L., and Kapuller, L. L. (1972). Vopr. Oncol. 18 (4) 58-63. 263. Rogers, A. E., and Gildin, J . (1975). J. N d . Cancer Inst. 55, 385-391. 264. Rogers, A. E., Herndon, B. J., and Newberne, P. M. (1973). Cancer Res. 33, 10031009. 265. Rogers, A. E., and Newberne, P. M. (1973). Nutiirt (London) 246, 491-492. 266. Rogers, K . J., and Pegg, A. E. (1977). Cancer Res. 37, 4082-4087. 267. Russel, D. H. (1977). Clin. Chem. 23, 22-27. 268. Salser, J. S., and Balis, M. E. (1973). Cancer Res. 33, 1889-1897. 269. Salser, J. S . , and Balis, M. E. (1974). Cancer 34, 889-895. 270. Salser, J. S . , Ball, W. J., and Balis, M. E. (1976). Cancer Res. 36, 3495-3498. 271. Schauer, A., and Kunze, E. (1971). Sci. Tech. Inf. 11, 45-49. 272. Schauer, A., Vollnagel, Th., and Wildanger, F. (1969). Z. Gesamfe Exp. M t d . 150, 87-93. 273. Schmahl, D., Danisman, A., Habs, M., and Diehl, B. (1976). Z. Krehsforsch. 86, 8994. 274. Schoental, R., and Bensted, J . P. M. (1968). Br. 1. Cancer 22, 316-323. 275. Schwartz, M. K. (1977). Cancer 40, 2620-2624. 276. Shimizu, H . , and Toth, B. (1975). Gann 66, 589-601. 277. Shockman, A. T. (1973). Am. J . Gastroentrrol. 60, 178-184. 278. Simpson, B. H., and Jolly, R. D. (1974). J . Pathol. 112, 83-92. 279. Singer, B. (1976). Nature (London) 264, 333-339. 280. Singh, D. V., Campbell, R. L., Lin, Y. N., and Nigro, N. D. (1977). Experientia 33, 1516-1518. 281. Sjogren, H.O., and Steele, G. (1975). Cancer 36, 2469-2471. 282. So, B. T., Magadia, N. E., and Wynder, E. L. (1973). J. Narl. Cancer Inst. 50, 927932. 283. Spitz, S., Maguigan, W. H., and Dobriner, I(.(1950). Cancer 3,789-804. 284. Spjut, H. J. (1972). Dis. Colon Recium 15, 94-99. 285. Spjut, H. J., and Noall, M. W. (1971). Cancer 28, 29-37. 286. Spjut, H. J., and Smith, M. N. (1967). Exp. Mol. Pathol. 6, 11-24. 287. Spjut, H . J., and Spratt, J . S. (1965). Ann. Surg. 161, 309-324. 288. Spratt, J. S . , Ackerman, L. V., and Moyer, C. A. (1958). Ann. Surg. 148, 682-698. 289. Springer, P . , Springer, J., and Oehlert, W. (1970). Z. Krebsforsch. 74, 236-240. 290. Steele, G., Jr., and Sjogren, H. 0. (1974). Cancer Res. 34, 1801-1807. 291. Steele, G . , Jr., and Sjogren, H. 0 . (1974). In?. J . Cancer 14, 435-444. 292. Steele, G., Jr., and Sjogren, H. 0. (1977). Surgery 82, 164-169. 293. Steele, G . , Jr., Sjogren, H. O., and Price, M. P. (1975). Inr. J. Cancer 16, 33-51. 294. Steele, G . , Jr., Sjogren, H. O., Rosengren, J. E., Lindstrom, C., Larsson, A., and Leandoer, L. (1975). J. Narl. Cancer Insr. 54, 959-967.
236
KAZYMIR M. POZHARISSKI ET AL.
2Y5. Steele, L., Cooper, E. H., MacKay, A. M., Losowsky, M. S., and Goligher, J . C. (1974). Br. J. Ccr/iccr 30, 319-324. 2Y6. Stevens, R. H . , Englund, C. W., Osborne, J. W., Cheng, H. F., and Hoffman, K . L. (1976). Cuticrr Res. 36, 3260-3264. 2Y7. Stevens, R. H., Englund, C. W., Osborne, J. W., Cheng, H. F . , and Richerson, H. B. (1975). J. N u / / . Ccrncrr Irisf. 55, 1011-1013. 2Y8. Stevens, R. H . , Smith, D. D., Osborne, J. W., and Oberley, L. W. (1976). J . Natl. Cuncrr I r i s / . 57, 43-45. 299. Stewart, H. L. (1953). I n “The Physiopathology of Cancer” (F. Homburger and W. H . Fishman, eds.), pp. 3-45. Cassell, London. 300. Sundaresan, P. R., and De Luca, L. M. (1977). J. Natl. C ~ n c e Ir/ i s t . 58, 1643-1645. 301. Tayler, R., Cumberland, V. H., and Piper, D. W. (1977). Grit 18, 45-47. 302. Terz, J. J., Curutchet, H. P., and Lawrence, W., Jr. (1971). Crrncrr 28, 1100-1 110. 303. Thurnher, N., Deschner, E. E., Stonehill, E. H., and Lipkin, M. (1973). Cancw R i v . 33, 940-945. 304. Topping, D. C., and Visek, W. J. (1976). J. Nirtr. 106, 1583-1590. 305. Toth, B. (1977). Cancer 40, 2427-2431. 306. Toth, B., Malick, L., and Shimizu, H. (1976). Am. J . Pufli01. 84, 69-86. 307. Troncale. F., Hertz, R., and Lipkin, M. (1971). Ccrticer R r s . 31, 463-467. 308. Trotta, P. P.. and Balis, M. E. (1977). Crrncer 40, 2592-2599. 30Y. Turell, R. (1973). Cu 23, 91-93. 310. Turkevitch, N. M., and Samundjan, E. M. (1975). “Hormones and Carcinogenesis.” Naukova Dumka, Kiev. 311. Turner, M. D., Kleinman, M. S., and Thayer, W. (1973). Digestiori 9. 116-123. 312. Turusov, V . S . , Lanko, N . C., and Bazlova, L. S. (1977). V o p r . Onkol. 23 (7). 39-42. 313. Tutton, P. J. M., and Barkla, D. H. (1976). VirchoMBs Arch. B: Cell Pntliol. 21, 147160. 314. Twining, S. S . , and Brecher, A. S. (1977). Clin. Chim. Acfri 75, 143-148. 315. Varghese, A. J., Land, P., Furrer, R., and Bruce, W. R. (1977). Proc. A m . Assoc. Criiicrr Res. IS, 80. 316. Vollnagel, T., Wildanger, F., and Schauer, A. (1969). Z. Giwimtc, Exp. Met!. 151, 208-2 13. 317. Walker, E. A., Bogovski, P., and Griciute, L. (eds.) (1976). ”Environmental N-nitroso Compounds. Analysis and Formation.” IARC Scientific Publications No. 14, Lyon. 318. Walpole, A. L., Williams, M. H., and Roberts, D. C. (1952). Br. J. Ind. Med. 9, 255263. 319. Walpole, A. L., Williams, M. H., and Roberts, D. C. (1955). Br. J . Cancer 9, 170176. 320. Ward, A. M., Cooper, E. H., Turner, R., Anderson, J . A., and Neville, A. M. (1977). Br. J. Concer 35, 170-1723, 321. Ward, J. M. (1974). Lab. Invest. 30, 505-513. 322. Ward, J. M., Rice, J. M., Roller, P. P., and Wenk, M. L. (1977). Cancer Res. 37, 3046-3052. 323. Ward, J. M., Sontag, J. M.,Weisburger, E. K., and Brown, C. (1975).J. N u / / .Cancer I m t . 55, 107-113. 324. Ward, J. M., and Weisburger, E. K. (1975). Cancer Res. 35, 1938-1943. 325. Ward, J. M., Yamamoto, R. S., and Weisburger, J. H. (1973). J. Nritl. Crrncrr I/i.st. 51, 713-715. 326. Ward, J. M.,Yamamoto, R. S., Weisburger, J. H., and Benjamin, T . (1973). J. Nut/. Cancer I n s t . 51, 1997-1999.
EXPERIMENTAL INTESTINAL CANCER RESEARCH
237
327. Watanabe, K., Yoshii, H., Iwashita, H., Muta, K., Hamada, K., Isaka, H., and Nishi. M. (1975). Gutin 66, 449-453. 328. Watson, D. W. (1974). C(117cer34, 867-871. 329. Wattenberg, L. W. (1959). Corrcer Res. 19, 1118-1 123. 330. Wattenberg, L. W. (1974). Am. J . Dig. Die. 19, 947-953. 331. Wattenberg, L. W. (1975). J. NNil. Ctrticcv Inst. 54, 1005-1006. 332. Wattenberg, L. W., Lam, L. K. T., Fladmoe, A. V., and Borchert, P. (1977). C'mcc,r40, 2432-2435. 333. Weisburger, J . H. (1971). Ctr/icer 28, 60-70. 334. Weisburger, J . H., Reddy, B. S., Narisawa, T., and Wynder, E. L. (1975). Proc. SOC. Exp. B i d . Mad. 148, 1119-1121. 335. Wellin, S., Yonker, J., and Spratt, J. S. , Jr. (1963). Am. J . R O O / l f g ~ ~ 090, / . 673-687. 336. Werner, B . , de Heer, K., and Mitschke, H. (1977). Z . Kwbsfbrsc/i. 88, 223-230. 337. Werner, B., de Heer, K., Mitschke, H., Knipper, A,, Thoran, P.. and Dahm, K. (1977). Z. Krehsforsch. 89, 53-60. 338. Wiebecke, B., Krey, U., Lohrs, U., and Eder, M. (1973). V i r c ~ l i o ~Arch. i ~ s A: PNIIIOI. Anni. 360, 179-193. 339. Wiebecke, B., Lohrs, U., Gimmy, J., and Eder, M. (1969). Z. Grscwiic, Exp. M r d . 149, 277-278. 340. Williams, R. D., and Fish, J. C. (1966). Am. J . Surg. 112, 846-849. 341. Williamson, R. C. N., Bauer, F. L. R., and Malt, R. A. (1977). Br. J. Cruicer- 36,401. 342. Wilson, R. B., Hutcheson, D. P., and Wideman, L. (1977). Ani. J . Clin. Nictr-. 30, 176- 18 1. 343. Winkler, R., Heitmann, C., and Otto, H. F. (1977). Chir. Gn,stroe/iterol. 11, 105-108. 344. Winneker, R. C., Tompkins, M., Westenberger, P., and Harris, J . (1977). Exp. Mol. Pothol. 27, 19-34. 345. Wittig, G., von, Wildner, G. P., and Zielbarth, D. (1971). Arch. G~~.sc/rii~/il.sffbr.sc~h. 37, 105-115.
346. Wood, D. A. (1967). "Tumors of the Intestines. Atlas of Tumor Pathology," Sect. 6, Fasc. 22. U.S. Armed Forces Institute of Pathology, Washington, D. C. 347. Wynder, E. L., Reddy, B. S., McCoy, G . D., Weisburger, J. H., and Williams, G. M . (1976). Clin. Gustrornterol. 5, 463-482. 348. Yamada, S., Ito, M., and Nagayo, T. (1971). Gaiiri 62, 471-478. 34Y. Yamamoto, T., Kato, H., and Smith, G . S. (1975). Gomi 66, 623-630.
This Page Intentionally Left Blank
ADVANCES IN CANCER RESEARCH, VOL. 30
THE MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES*
Bill Sugden, Christopher R. Kintner, and Willie Mark The McArdle Laboratory tor Cancer Research, University of Wisconsin, Madison, Wisconsin
I. Introduction ........................................................... 11. A Brief Survey of Lymphotropic Herpesviruses . B. Persistent and Latent Infections ......................................
EBV ............................................. A . Proteins in the Virion ...............................................
References
............................................................
239 240 244 244 246 248 249 249 25 I 25 1 258 259 259 262 265 266 268 268
I. Introduction
Lymphotropic herpesviruses are those members of the Herpesvirus family that infect lymphoid cells of their natural hosts. These herpesviruses frequently cause lytic infections as well as an infection that leads to a self-limiting lymphoproliferative disease or a malignant lymphoma. The ability of these viruses to cause these diseases is probably derived, at least in part, from their capacity to induce and maintain cell division in the infected cell. A major goal of research into the molecular biology of lymphotropic herpesviruses is to learn how they induce these growth changes in their target cells. If we can achieve this goal, we may gain insight into mechanisms of both oncogenesis and the control of normal development.
* We dedicate this paper to Dr. Harold Rusch, under whose leadership McArdle Laboratory has become a communal research laboratory where our science profits immensely from our environment. 239 Copyright @ 1979 by Academic Press. Inc. All rights of reproduction in any form reserved ISBN 0-I?-006630-0
240
BILL SUGDEN ET AL.
The bulk of studies on the lymphotropic herpesviruses to date has focused on three viruses: Marek’s disease virus (MDV), found in domestic chickens; H e r p e s v i r ~ ssaimiri (HVS), found in squirrel monkeys; and Epstein-Barr virus (EBV), found in human beings. MDV causes a lymphoma in chickens which progresses rapidly to kill the animals (Witter, 1971; Nazerian, 1973). MDV is transmitted horizontally, and its DNA contains no detectable nucleic acid homology with chicken cell DNA (Lee et al., 1971; Tanaka et al., 1978). These findings distinguish MDV mechanistically from the transformation-defective Rous-associated viruses, which probably induce lymphoid leukosis by a process involving exchange of information with cell DNA with which they share sequence homology (Temin, 1971). We conclude that MDV and, perhaps, other lymphotropic herpesviruses induce tumors by themselves introducing the genetic information needed to render an infected cell malignant. EBV is necessary and sufficient to transform fetal and adult human lymphocytes into permanently dividing cultures of lymphoblasts (Pope et al., 1968). If we can understand how EBV changes an apparently differentiated end cell into a permanent blast cell, we may also learn how some cellular decisions are made in normal development. In this review we shall first present a general survey of the lymphotropic herpesviruses. We shall then describe our current knowledge of the molecular biology of the three prototypic viruses: MDV, HVS, and EBV. We shall also present an overview of the relationship of these viruses to their hosts it7 virw. It is these relationships that we hope eventually to understand at the molecular level. II. A Brief Survey of Lymphotropic Herpesviruses
A survey of the salient (for the purpose of this review) features of the identified lymphotropic herpesviruses is presented in Table I . MDV, HVS, and EBV will be discussed in detail later. We will now briefly describe the other lymphotropic herpesviruses. Herpesvirus s y l v i l a g ~ swas isolated from naturally infected, wild cottontail rabbits (Hinze, 1971a). This virus can be grown in tissue culture, but it has not been well characterized. When rabbits are infected experimentally with this virus, they may develop a lymphoproliferative disease (Hinze, 1971b). Herpesvirus of turkeys (HVT) is antigenically related to MDV (Witter et al., 1970). It is avirulent in chickens and has been used for vaccination of chickens against Marek’s disease (Okazaki et al., 1970; Eidson and Anderson, 1971). The virus is less frequently bound to cell debris than MDV when grown in tissue culture. This property of HVT permits the
MOLECULAR BIOLOGY OF LY MPHOTROPIC HERPESVIRUSES
24 1
relative ease of purifying cell-free virus needed for vaccine preparation. Little is known about the molecular biology of this virus. Most of the lymphotropic herpesviruses so far identified have been isolated from nonhuman primates, and more are likely to be found among them. Four species of New World monkeys have been found to have antibodies to HVS-associated antigens (Deinhardt et al., 1974). Eight species of Old World monkeys and four species of the great apes can carry antibodies that cross-react with EBV-induced antigens (Frank et al., 1976). These findings indicate that monkeys in the wild are exposed to herpesviruses that are related to EBV and HVS. It seems likely that these species of primates harbor the viral agent(s) and act as reservoirs for their horizontal transmission. The age-dependent acquisition of antibodies by these animals seems to support this conclusion (Landon and Malan, 1971; Deinhardt et al., 1974). In addition, when monkeys are housed together with seropositive animals, they develop antibodies that cross-react with EBV- or HVS-induced antigens (Deinhardt et al., 1974). Two of the most studied simian lymphotropic herpesviruses are HVS and Herpesvirus ateles (HVA). Melendez and colleagues first isolated HVS and HVA from fibroblast cultures derived from clinically healthy host animals (Melendez et al., 1968: Melendez et al., 1972). Subsequently, HVS and HVA have been isolated from peripheral blood leukocytes of their natural hosts by cocultivation with permissive cells (Falk et a/., 1973; Falk et al., 1974b). Both viruses grow well in tissue culture and can be propagated in several species of primate fibroblast cell lines. The occurrence and isolation of these two herpesviruses have been reviewed by Deinhardt et al. (1974). In this review, HVS will be used as a prototype virus for discussion. There is no proof yet that the more recent isolates of herpesviruses from other primates are distinct viruses. Herpesviruses have been identified in spontaneously transformed lymphoblastoid cell lines of chimpanzees (Landon et al., 1968) and baboons (Gerber e[ al., 1977). Antigenically, these two herpesviruses are related to EBV, since cell lines harboring the virus are stained by antisera that are specific for EBVinduced antigens (Gerber et al., 1976: Goldman et al., 1968). The viral DNAs present in these cell lines share sequence homology with EBV DNA (Gerber et al., 1976: Falk et al., 1977). Neither virus has been well characterized as a result of the difficulty in obtaining cell-free virus from the spontaneously transformed cell lines. Several attempts have been made to rescue the virus by transforming one species’ leukocytes with cell-free virus from the other (Gerber ef al., 1977, Falk et al., 1976). Transformed cell lines were established in these experiments. Under these circumstances, it is essential to demonstrate which virus is harbored in the established cell line. If the animal that donated the target cells for
242
BILL SUGDEN ET AL.
TABLE I SURVEY OF LYMPHOTROPIC
Virus Herpesvirus sylvilagus
Natural host (reference)
Related herpesviruses" (reference)
Cottontail rabbit (Hinze. 1971a)
Marek's disease virus (MDV)
Chicken (Churchill and Biggs, 1967: Biggs ei a l . , 1968)
Herpesvirus of turkeys (HVT
Turkey (Kawamura ei a / . , 1969)
Herpesvirus saimiri (HVS)
Squirrel monkey (Melendez ei a / . . 1968)
Herpesvirus ateles (HVA)
Spider monkey (Melendez ef a/.,1972)
Epstein-Barr virus (EBV)
Human (Epstein and Barr,
MDV (Kamamura ef a / . , 1969)
HVS (25%) (Fleckenstein r i a/., 1978)
1964)
Herpesvirus papio
Baboon (Falk ei a/. , 1976)
EBV (Falk et a l . , 1976)
Herpesvirus of chimpanzee
Chimpanzee (Landon ei a / . , 1968)
EBV (35-45%) (Gerber ei a/., 1976)
Herpesvirus pongo
Orangutan (Rasheed e / al., 1977)
EBV (30-40%) (Rabin ei al., 1978)
a As measured antigenically or by DNA sequence homology. Only one of each homolpairs.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
243
HERPESVIRUSES Host of leukocytes for transformation in vifra
(reference)
Host for induction of lymphoid tumors (reference) Cottontail rabbit (Hinze, 1971b) Chicken (Biggs et a / ., 1968: Calneck et a / . , 1970)
Cotton-topped marmoset (Hunt et a / . , 1970) Owl monkey (Deinhardt ef a / . , 1974) New Zealand rabbits (Daniel et al., 1975) Cotton-topped marmoset (Falk et a l . , 1974a, 1978)
Cotton-topped marmoset (Melendez et al., 1972)
Human (Pope ef al., 1968) Cotton-topped marmoset (Miller and Lipman, 1973) Other primate monkey (Frank et a / ., I976
Cotton-topped marmoset (Frank et a / . , 1976: Miller et a / ., 1977a)
Human (Gerber et a / ., 1977) Cotton topped marmoset (Rabin et al., 1977) Some other primates (Falk et a / ., 1977) Human, baboon, and some other primates (Gerber et a/. , 1977) Gibbon (Rabin et a / . , 1978) ogous set is noted: parentheses indicate percent of DNA sequence homology between
244
BILL SUGDEN ET
AL.
the transformation experiments harbored cells already transformed in vivo by its herpesvirus, then those cells could also grow out as lines. Recently, a lymphoblastoid cell line derived from a leukemic orangutan was shown to contain a herpesvirus (Rasheed et al., 1977). This virus is related serologically and by DNA sequence homology to EBV (Rabin et al., 1978). However, the orangutan had been in captivity for ten years in a city zoo. It is possible that it had been in contact with human EBV carriers or other naturally infected primates during its captive life. Again it is important to determine the distinguishing properties of new isolates of viruses before we classify them as distinct lymphotropic herpesviruses.
Ill. MDV, HVS, AND EBV in Their Natural Hosts
It is apparent that infection by a lymphotropic herpesvirus can affect the host variously. We know that some strains of MDV are able to cause Marek's disease (MD), a malignant lymphoma, at a high efficiency in genetically susceptible chicken. In other conditions MDV causes a persistent infection. We know EBV in some cases causes one form of infectious mononucleosis in people and that it is possible but not proven that EBV is necessary to cause most cases of Burkitt's lymphoma and undifferentiated nasopharyngeal carcinoma. Finally, EBV infects so many people without causing overt disease that most human beings harbor it in a latent state. In order to identify the possible role that these viruses can play in causing disease, we must ascertain both if the virus is required causally and what modifying factors (for example, virus strains, genetic make-up, and immunological status of host) participate in determining the outcome of infection. A. VIRALDISEASES Marek's disease, in its acute form, is a lymphoproliferative disease that results in the development of lymphoid tumors in chickens. Animals are usually exposed to the virus early in life and become infected by 46 weeks ofage (Biggset al., 1968; Witter, 1971, 1972). The virus replicates efficiently in the feather follicle epithelium of the bird (Calnek and Hitchner, 1969: Nazerian and Witter, 1970): this probably accounts for the infectiousness of the disease. In some fashion lymphoid cells are infected by the virus, proliferate, and infiltrate many organs of the animal to give rise to the neoplastic disease. Several lymphoblastoid cell lines have been established from Marek's disease tumor tissue, all carry T-cell markers (Payne et a / . , 1974; Nazerian and Sharma, 1975: Matsuda et al., 1976),
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
245
and two have been shown to contain multiple copies of MDV DNA (Nazerian and Lee, 1974; Tanaka et al., 1978). These findings are consistent with the notion that the tumor cell in Marek's disease is t h e MDVharboring T lymphoblast. Not all chickens are susceptible to MD when exposed to the virus. There exist strains of inbred chickens that are genetically resistant to MDV-induced lymphoma (Cole, 1972). Longenecker and co-workers have found that resistance is associated with the B locus of the major histocompatibility complex (MHC) of the chicken (Longenecker et al., 1976). Briles et aI. (1977) have since confirmed this finding. Recently, Longenecker and his colleagues have discovered a second locus in the MHC that may also help to confer resistance to MD (Fredericksen et a l . , 1977). The mechanism of resistance is not known. Chickens that are not susceptible to neoplasia can harbor the virus in a latent form. In addition to host-related resistance to MD, there are strains of MDV that are not pathogenic in susceptible chickens. Some of these strains have been used to vaccinate birds against subsequent challenge with virulent strains of MDV (Churchill et al., 1969a). The mechanism of protection against disease by vaccination with avirulent MDV is not yet understood. However, this acquired immunity does not protect the host from persistent infections by MDV; that is, MDV can be isolated from these birds throughout life. Whereas HVS is not known to cause overt disease in its natural host, EBV does. The relationship between EBV and human disease cannot be as directly studied as can that of MDV in chickens. However, we accept the notion that EBV causes the heterophile antibody-positive form of infectious mononucleosis. The evidence for this etiologic role of EBV is necessarily indirect but it is still compelling. The varied evidence runs the gamut from an early demonstration of transmission of infectious mononucleosis by transfusion (Wising, 1942) to careful seroepidemiologic studies (Neiderman et al., 1968), which found a primary induction of antibodies to EBV-associated antigens soon after clinical recognition of the disease. Additional evidence includes the demonstration that a population of killer cells specific for cells harboring EBV information is induced i n vivo and remains detectable only during the acute stage of infectious mononucleosis (Svedmyr and Jondal, 1975)'; that about 1 per 1000 of the peripheral leukocytes of patients in this acute stage is infected This study used EBV-harboring lymphoblastoid cell lines as targets for the killer cells isolated from patients' blood. In most if not all experiments, therefore, the target and killer cells did not share histocompatibility antigens. The observed killing apparently does not require identity of some histocompatibility antigens, although such a requirement has been found for specific killing of murine cells lytically infected with several different viruses (Doherty et ol.. 1976).
246
BILL SUGDEN ET
AL.
by EBV (Klein et al., 1976); and that this number drops dramatically in the convalescent patient (Hinuma and Katsuki, 1978). Most patients early in the course of infectious mononucleosis secrete EBV orally and can continue to do so for many months thereafter (Chang and Golden, 1971; Niederman et al., 1976). These findings are consistent with the idea that the heterophile antibody-positive form of infectious mononucleosis results from the oral exposure of a susceptible person (usually a EBV seronegative adolescent or young adult) to EBV. The ensuing infection and proliferation of EBV-carrying B lymphoblasts lead to the stimulation of killer cells that eventually limit the B-lymphoblast proliferation and then decline in number as a result of the lack of further stimulation. This satisfying (and, inevitably, oversimplified) description of the role played by EBV in causing one form of infectious mononucleosis does not yet have its equally satisfying counterpart in relating EBV to Burkitt’s lymphoma and nasopharyngeal carcinoma. Several reviews describing the evidence that implicates EBV as being necessary to cause these two human tumors have been published (Klein, 1973; Miller, 1974; zur Hausen, 1975). We have presented major pieces of the evidence in Table 11. In general, if a tumor is monoclonal and the tumor cells all contain viral information, then that virus had to be present at the onset of tumor formation and might have played an active role in that formation event. There are sufficient uncertainties in the data for EBV to force us to regard this logical argument as speculative and in need of further evidence for the case of EBV and Burkitt’s lymphoma. Other kinds of evidence for the etiologic role of EBV are less direct than that presented in Table 11. EBV can induce malignant lymphomas in some New World monkeys (Shope et al., 1973) and transforms human B lymphocytes into “immortalized” blast-cell lines (Pope et al., 1968). These facts must be reconciled with the observations that both Burkitt’s lymphoma and nasopharyngeal carcinoma are rare diseases on a world-wide basis but occur frequently in restricted regions of the world, whereas infection with EBV occurs in 80% or more of people throughout the world. If EBV is necessary for these two human neoplasms, it is not alone sufficient to cause them.
A N D LATENTVIRALINFECTIONS B. PERSISTENT
MDV, HVS, and EBV can give rise to persistent or latent infections in their respective hosts. Vaccine strains of avirulent MDV yield life-long persistent infections (Jackson et al., 1974). More than 8% of clinically healthy squirrel monkeys exhibit antibodies to HVS antigens, and the virus can usually be isolated from these animals by cocultivating their
TABLE I1 EVIDENCE RELATINGEBV TO Two HUMANTUMORS ~~
Evidence
Burkitt's lymphoma
References
Nasopharyngeal carcinoma
~
References
I . Seroepidemiologic studies find higher titers of antibodies against virus-associated antigens in patients as compared with controls 2. Multiple copies of viral DNA are found in the tumor cell
+
Henle et a / . (1969)
+
Henle er a/. (1970a)
+
+
zur Hausen et a/. (1970):
3. The tumor is monoclonal in origin" 4. In some genetically immunodeficient people the viral infection appears to lead to tumor formation in several weeksb 5 . A prospective seroepidemiologic survey finds that children in Uganda with abnormally high titers of antibodies to some EBV-associated antigens have a 30-fold increased risk of contracting the neoplasm
+ +
zur Hausen et al. (1970); Nonoyama and Pagano (1973) Fialkow et a/. (1970) Bar ef a/. (1974): Purtilo et a / . (1978)
+
Anderson-Anvret et a / . (1977)
de-The et a/. (1978)
It has t o be noted that rapid growth of a single clone of a multiclonal tumor early in the history of the tumor could yield a tumor that by isozyme markers appears to be monoclonal in origin. It is not proven in the few examples of these patients if EBV is necessary for tumor formation. (1
248
BILL SUGDEN ET AL.
peripheral leukocytes with susceptible fibroblasts (Deinhardt et al., 1974). Infection by EBV appears to persist after infectious mononucleosis, because virus can frequently be isolated from throat washings 6 to 12 months after the clinical symptoms have passed (Niederman et al., 1976). After that time the virus remains in a latent state in that a small number of EBV-harboring B lymphocytes can be detected by growing them in tissue culture, whereas the large mass of lymphoid cells not carrying EBV fail to grow (Moore et al., 1967; Gerber and Monroe, 1968). In the majority of people infected by EBV the entire infection remains asymptomatic and the virus is generally detected only by the specific antibodies it stimulates or by cultivating virus-harboring cells in tissue culture. Additional evidence for the latent infection of seropositive people by EBV is found in studies on immunosuppressed adults, who usually begin to secrete detectable quantities of EBV in their saliva within several weeks of administration of immunosuppressive drugs (Strauch et d., 1974).
IV. Experimental Tumor Studies
MDV, HVS, and EBV can each induce lymphomas in experimental animals. Some of the accumulated data for tumor induction by these viruses are presented in Table I . These data indicate that MDV induces tumors in the chicken, its natural host, and that HVS can induce tumors in two kinds of marmosets as well as in rabbits but not in its natural host, squirrel monkeys. It should be noted, however, that the number of squirrel monkeys studied in captivity is tiny compared with the number of people observed medically. If tumors induced by HVS occurred as rarely in its natural host as Burkitt’s lymphoma and nasopharyngeal carcinoma occur in people, then it is statistically likely that these rare tumors in squirrel monkeys probably would not have been identified. The ability of EBV to induce tumors in cotton-topped marmosets is important but may not be relevant to its potential etiologic role in human tumors for two reasons. First, as expected, there are significant differences in the virusassociated tumors in the different hosts. For example, while Burkitt’s lymphoma appears to be monoclonal in origin (Fialkow et al., 1970). the EBV-induced lymphoma in marmosets appears to be polyclonal in origin (Deinhardt et al., 1975). Second, there are several precedents of viruses that apparently do not induce tumors in their natural hosts (human adenoviruses and simian virus 40, for example), but can induce tumors in experimental animals. A second mode of tumor formation has been used in the study of EBV-
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
249
related tumors. Biopsies or cell lines of Burkitt’s lymphoma and nasopharyngeal carcinoma biopsies grow as tumors following xenotransplantation into homozygous nu/nu mice (Nilsson et al., 1977; Klein et al., 1974). This foreign graft experiment provides a way of measuring tumorigenicity. It permits distinction among four kinds of cells harboring EBV: ( I ) those recently transformed in vitro, (2) those recently established from normal donors, (3) those passaged in culture for long times, and (4) those freshly established from tumors. The latter two yield tumors upon subcutaneous transplantation; the former two do not (Nilsson et al., 1977). V. Studies of MDV, HVS, and EBV in Tissue Culture
A. TYPESOF CELLSSUSCEPTIBLE TO INFECTION Types of cells that are susceptible to infection by MDV, HVS, and EBV in vivo and in vitvo are presented in Table 111. Although we know that M D V replicates efficiently in the feather follicle epithelium of the chicken, we do not know the sites for replication of HVS and EBV in their natural hosts. It is obviously important to identify sites of replication of EBV in vivo because in so doing we may be able to establish in culture a productive host cell for that virus. Detection of EBV in throat washings of patients with infectious mononucleosis (Chang and Golden, 197 I ; Niederman et al., 1976) and of EBV-associated antigens in a high percentage of B lymphocytes carrying complement receptors in tonsils (Veltri et al., 1977) indicates that the throat probably harbors a site for viral replication, but it has not yet been identified. As determined by xenotransplantation of nasopharyngeal carcinoma tissue into nude mice, epithelial tumor cells harbor viral DNA (Klein et al., 1974). Cells with epithelial morphology isolated from throat scrapings of patients with infectious mononucleosis harbor EBV DNA as detected by in sitrr hybridization (Lemon et al., 1977). However, it is not known if the virus can mature in and be released from these cells. MDV and HVS lytically infect the appropriate fibroblast cultures in vitro, but no such productive host for EBV has been identified. The lack of a productive host cell has severely restricted research on EBV in two ways. First, MDV and HVS can be genetically purified by repeated plaque formation on susceptible target cells, whereas EBV cannot. The potential heterogeneity of all stocks of EBV precludes some experiments and renders others difficult to interpret. Second, large-scale production and isotopic labeling of the virions in tissue culture are practicable for HVS but probably not for MDV and certainly not for EBV. The release
250
BILL SUGDEN ET AL.
TABLE 111 TYPESOF CELLS OBSERVED TO
INFECTION BY LYMPHOTROPIC HERPESVIRUSES
BE SUSCEPTIBLE TO
I n Viiw
Virus MDV
HVS
EBV
Lytic Infection (references)
Transformation (references)
Chicken feather Chicken T follicle lymphocytes (Calneck and (Payne el a / . , Hitchner, 1969: 1974: Nazerian Nazerian and and Sharrna, Witter, 1970) 1975: Matsuda P / ul. , 1976) Owl monkey T None lymphocytes (Wallen PI ul. , 1973) Marmoset and owl monkey lymphocytes (Deinhardt rt d.,1974) Human B None lymphocytes (Jondal and Klein, 1973: Huber et ul., 1976) Human epithelial cells of the nasopharynx and throat (Klein C I a l . , 1974: Lemon rt ul., 1977) Cot ton-topped marmoset lymphocytes (Frank ~t a / . , 1976)
Ill
Lytic infection (references)
Vi1r.o
Transformation (references)
Chick and duck embryo fibroblasts (Witter P I ul., 1969)
None
Green monkey kidney and owl monkey kidney cell lines (Deinhardt C I 01.. 1974)
None
None
Human B lymphocytes (Pattengale C I ul.? 1973: Mizuno ('1 n l . , 1974) Lymphocytes of 7 nonhuman primates (Frank PI ul., 1976)
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
25 1
of MDV into the extracellular fluid is too inefficient for large-scale virus production, presumably because the virus remains cell-associated. Therefore, many experiments on the molecular biology of MDV and EBV are not feasible. Neither MDV nor HVS has been shown to transform lymphocytes in culture. This lack may result from our inexperience in working with uninfected avian and primate lymphocytes in culture, or it may result from the fact that those cells that are susceptible to transformation in vivo may not be present in adequate concentrations in peripheral blood to be transformed in vitro. EBV does transform human B lymphocytes in culture, and clonal transformation assays for EBV have been developed with fetal (Yamamoto and Hinuma, 1976) and adult (Sugden and Mark, 1977) lymphocytes. These assays should permit quantitative studies of the transformation of target cells by EBV.
B. LYTICINFECTIONS A s shown in Table 111, MDV and HVS infect some fibroblastic cultures in vitro, and HVS grows productively in them. N o lytic host has as yet been identified for EBV. The molecular events that occur in the lytically infected cell, including uncoating, transcription and translation of viral products, viral DNA synthesis, and virion maturation, have not been elucidated for MDV or HVS. We do know that the DNA of HVS is infectious in tissue culture (Fleckenstein er al., 1975) as is herpes simplex DNA (Graham et a/., 1973; Sheldrick et al., 1973). This finding indicates that HVS requires no virion proteins for complete infection of cells and, by analogy only, that MDV and EBV probably do not require virion proteins for their successful infection of their target cells. Although a productive host cell in tissue culture has not been found for EBV, a number of cell strains and lines have been tested to determine whether they could serve in that capacity. A list of these negative results is given in Table IV, not to discourage but to guide investigators in testing other cells in culture.
C. TRANSFORMATION MDV and HVS have not yet been shown to transform cells in tissue culture. EBV, however, does transform both human and nonhuman lymphocytes in culture. This transformation of lymphocytes by EBV is distinct from transformation of fibroblasts by other animal DNA viruses (see Tooze, 1973, for a review). Transformation by EBV does not result in altered growth morphology, reduced nutrient requirements, or loss of
252
BILL SUGDEN ET AL.
TABLE IV CELLSTESTEDA N D FOUND INCAPABLE OF SERVING AS LYTICHOSTSFOR EBV Primary cells
Human embryonic tissues" Human placenta" Human epithelial cells from adenoids and tonsilsu Human macrophages' Human lymphocytes from patients with chronic and acute lymphocytic leukemia" Marmoset kidney, skin, and muscle" Squirrel monkey kidney, lung, and heart" Rhesus monkey kidney" Callothrix skin and muscle" Hamster embryo fibroblasts" Chicken embryo fibroblasts"*b Turkey embryo fibroblastsb Pheasant embryo fibroblastsb Duck embryo fibroblastsb Lymphoid cell cultures" of cotton-topped and white-lipped marmosets, squirrel monkey, woolly monkey, callothrix, slow loris, hamster, and guinea pig. Cell 1inc.s or cell strains
Human fibroblast and epithelial:
W13XR,HeLa", Hep2"ab, KBb, RPMI-26SOb~', A-2S3b*P, A-38Sb.' Human lymphoid: NC37". F26Sa, WILZ", DEMO, MOLT4cr,BJAB" Green monkey kidney: VERO"sb, BSC-I"*b.JR", CV-Ib Squirrel monkey lung" Owl monkey lungU Woolly monkey lung" Hamster: Iotau, BHK-21b Mouse: Lb, Friend virus erythroleukemias, GM-X6b and T3-CI Mink: normal fibroblastsb Cells tested by Dr. George Miller and his colleagues of Yale University who kindly provided us with their unpublished results. They screened EBV-exposed cultures for cytopathic effects, transformation, and expression of viral capsid antigens. Cells we tested. We screened EBV-exposed cultures for cytopathic effects, for expression of viral capsid antigens, and (after several passages) for the presence of EBV DNA by renaturation kinetics. Human epithelial tumor cell lines originating from the nasopharynx kindly provided us by Dr. Jorgen Fogh of Sloan-Kettering. (I
"topoinhibition," because these changes can affect only dividing cells. EBV infects and transforms a nondividing lymphocyte and converts it into a permanently dividing lymphoblast. Several kinds of assays have been established to monitor transformation by EBV: transformation in mass culture (Pope r t al., 1968), by end-
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
253
point dilution (Moss and Pope, 1972; Henderson et al., 1977), and by colony formation in a semisolid medium (Mizuno er al., 1976: Yamamoto and Hinuma, 1976: Sugden and Mark, 1977). Each of these assays has advantages and disadvantages. One disadvantage of transformation of total peripheral leukocytes in mass culture and in end-point dilution experiments is that many cells that are destined not to be transformed are close to and can interact with cells that will be transformed. This problem may be only academic for fetal cells, but is acute when adult leukocytes are used as targets, because adult T cells inhibit transformation of autochthonous B cells (Thorley-Lawson et ul., 1977). The problem can be avoided by first separating lymphocyte populations and using only B cells as targets. Clonal transformation techniques also avoid this problem but require that the transformed cells form colonies in a semisolid medium. For maximum efficiency both end-point dilution assays (Henderson et al., 1977) and clonal transformation assays (Sugden and Mark, 1977) require feeder layers of fibroblasts but are not affected by mercaptoethanol. Several quantitative characteristics for transformation of fetal lymphocytes as measured by end-point dilution (Henderson er al., 1977) and for adult lymphocytes as measured by clonal transformation (Sugden and Mark, 1977) have been determined and agree remarkably well: ( I ) one particle of EBV is sufficient to transform a cell: (2) between 3% and 10% of the total leukocyte population is susceptible to transformation by EBV; (3) the plating or cloning efficiency of the transformed cells ranges between 1% and 1%. We have stated that the human target cell for transformation by EBV is the B lymphocyte (Pattengale et al., 1973; Mizuno et al., 1974: Robinson et al., 1977; Katsuki et ul., 1977; and Thorley-Lawson et al., 1977). Hinuma and his colleagues have tried to determine whether only a fraction of human B lymphocytes are susceptible to transformation by EBV (Katsuki et al., 1977) and have found that the number of cells in a target population that are IgM-positive correlates best with the number of cells that become transformed (Katsuki er af., 1977). They find only '/a to '/z of the percentage of the total lymphocytes that bear IgM found by other investigators (Rowe er al., 1973; Kumagai et al., 19751, but they find that all cell lines established by transformation with EBV in vitro carry IgM (Katsuki and Hinuma, 1975). We find the suggestion exciting that peripheral leukocytes that express IgM on their surface are the necessary targets for EBV transformation. The majority of IgM-positive cells also carry IgD (Rowe et al., 1973), and it has been suggested that surface IgD serves as a trigger that permits
254
BILL SUGDEN ET AL.
maturation of the lymphocyte to a memory or plasma cell after encountering antigen (Vitetta and Uhr, 1975). It would be desirable to know if EBV preferentially transforms IgD-positive cells and if the transformed cell lines continue to express IgD. This information might help us to understand if EBV’s transforming a lymphocyte into a blast cell is a carefully regulated change that in part follows the normal ontogeny of lymphocyte maturation or represents a grosser level of virus-imposed regulation that overrides normal developmental controls. Miller and his colleagues have studied the surface characteristics of cell lines established by transforming nonhuman primate leukocytes with EBV (Robinson et ul., 1977; Andiman and Miller, 1978). In addition, they have separated populations of lymphocytes from cotton-topped marmosets and found that, although the population of cells enriched for those bearing complement receptors is most susceptible to transformation, the transformed cell lines that grow out lack complement receptors (Robinson et al., 1977). They favor the hypothesis that transformation of marmoset cells by EBV leads to alteration of specific surface properties of the target cells. An alternative explanation of these findings comes from the work of Thorley-Lawson et al. (1977), who found that adult human T cells inhibit transformation of autochthonous B cells by EBV. I n the experiments of Robinson et al. (1977) in which different fractions of marmoset leukocytes were tested for their susceptibility to being transformed, the T cells would largely copurify with those cells that lacked complement receptors. It is possible that these T cells would prevent transformation of the bulk of complement receptor negative cells with which they copurify . Those cells that lack complement receptors but contaminate the population both enriched with cells bearing that receptor and depleted of T cells could still be the targets for transformation by EBV. The above discussion illustrates a difficulty of transformation studies with EBV. Not only are all virus stocks potentially heterogeneous, but also all target cell populations are heterogeneous and the multiplicity of potential interactions that may take place makes it difficult to interpret results. Epstein and his colleagues use heterogeneous cell populations while studying whether cells harboring EBV in vivo are capable of growing in vitro and, if not, whether they release EBV, which then infects other cells that do grow in vitro (Rickinson et al., 1975; Rickinson et d., 1977). The authors favor the second explanation, which would require two steps to establish EBV-transformed cell lines from the peripheral blood of patients with infectious mononucleosis. The experiments involve cocultivating fetal cells with peripheral leukocytes of patients with infectious mononucleosis, and these investigators find that the transformed
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
255
cells that grow out are usually of fetal origin (Rickinson et al., 1975, 1977). Again the findings of Thorley-Lawson et al. (1977) provide a different possible explanation for the observed prevalence of this twostep transformation. Adult T cells from the infectious mononucleosis patient will inhibit the outgrowth of autochthonous EBV-transformed cells but do not influence the transformation of those fetal leukocytes that have been infected with virus released from the adult transformed cells (in general, the fetal and adult cells will not share histocompatibility antigens). In all events, although the work of Epstein and his colleagues shows that transformation can take place in two steps (that is, that some peripheral leukocytes of infectious mononucleosis patients do release EBV), the work of Hinuma and Katsuki (1978) demonstrates that two steps are not required. They found that at least 1 in lo5 peripheral leukocytes from patients with infectious mononucleosis can form colonies when plated directly into agarose (Hinuma and Katsuki, 1978). The early events that take place during transformation of human lymphocytes by EBV are difficult to study and to interpret not only because of heterogeneities of virus and target cells, but also because of the potential temporal asynchrony of those events. Nevertheless, several investigators are trying to define those early events. Using fetal cells infected with EBV, labeling them with tritiated thymidine, and determining radioactive uptake by autoradiography , two groups have found that one EBV-associated nuclear antigen, EBNA, is expressed about 12 to 16 hours after infection, although cell DNA synthesis is not detected in those cells within 30 to 36 hours after infection (Einhorn and Ernberg, 1978; Takada and Osato, in press). On the assumption (as yet unsupported) that EBNA is coded at least in part by EBV, these findings indicate that expression of some viral genes precedes both the stimulation of cell DNA synthesis and blast formation (Einhorn and Ernberg, 1978; Takada and Osato, in press). Infection with EBV stimulates cell DNA synthesis in a biphasic fashion with time, and UV-irradiation of the virus abolishes this stimulation (Thorley-Lawson and Strominger, 1978). The first stimulatory phase is not inhibited by the addition of phosphonoacetic acid (PAA), but the second is (Thorley-Lawson and Strominger, 1978). Growth of potentially transformed cells is inhibited by PAA (Thorley-Lawson and Strominger, 1976). PAA inhibits some DNA polymerases by binding to the site where pyrophosphate is released (Leinbach et al., 1976). In particular, PAA inhibits DNA polymerases specifically induced by herpes simplex virus (Mao et al., 1975) and MDV (Lee et al., 1976). Investigators who have used PAA to study transformation by EBV have tended to interpret observed inhibition by PAA to be the result of the drug’s inhibiting an
256
BILL SUGDEN ET AL.
EBV-induced DNA polymerase that performs a necessary role during transformation. This interpretation may or may not be correct because PAA also inhibits one DNA polymerase of uninfected cells (Lee er al., 1976), and that cell enzyme is necessary to synthesize the DNA molecules of other animal viruses (Edenberg et a l . , 1978; Otto and Fanning, 1978). Therefore, PAA may serve as a rather blunt scalpel to dissect early events during transformation by EBV. However, these experiments (Einhorn and Ernberg, 1978; Takada and Osato, in press; Thorley-Lawson and Strominger, 1978) permit us to reconstruct a likely sequence of events leading to the first stirnulatory phase early in the transformation of B lymphocytes by EBV: first, penetration and uncoating of the virus particle takes place; then transcription by host enzymes of some viral messages occurs; finally those messages are translated into viral proteins that act directly or that induce cell proteins to stimulate the initiation of cell DNA synthesis by a PAA-insensitive host DNA polymerase ( p DNA polymerase?). Further experiments are required to support or to refute this putative sequence of early events during transformation of cells by EBV. The amount of viral information required for the initiation and maintenance of the transformed state has been studied by inactivation experiments (Henderson et al., 1978; W. Mark and B . Sugden, unpublished). The slope of the curve of inactivation of the transforming capacity of EBV with doses of X rays over one decade (Henderson rr al., 1978) or with y rays over two decades (W. Mark and B. Sugden, unpublished) is the same as that for similar inactivations of plaque formation by herpes simplex virus type I (HSV-I). These findings indicate that the target size for transformation by EBV is the same as the target size for plaque formation by HSV-I when measured by inactivation with X or y rays. The findings differ strikingly from those resulting from similar studies performed with other DNA-transforming viruses. By X ray inactivation experiments, Benjamin (1965) found that only one-half of the polyoma genome was required to transform cells, and Graham el ul. (1974), by fractionating sheared viral DNA, found that less than 10% of Adenovirus 2 information was required to transform cells. There are two possible interpretations for these findings from inactivation studies of EBV. One is that all of the viral information must be expressed in order for EBV to transform a cell. The second interpretation results from the recognition that the major unrepaired lesion introduced by X and y rays into DNA is a double-stranded break. Breaking the linear viral DNA molecule into pieces may make it impossible for the virus to maintain its genome in the transformed cell. One may be able to distinguish between these two interpretations by comparing the slope of inactivation of transformation
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
257
by EBV with that for inactivation of plaque formation by HSV-1 when point mutagens are used as the inactivating agents. The concentration of viral DNA and release of mature virions have been studied in transformed cells after the first 20 to 30 cell divisions following exposure of cells to EBV. It appears certain that viral DNA undergoes disproportionate replication relative to cell DNA during this time (Sugden et al., J . Virol., in press). This conclusion was reached by infecting adult leukocytes with 0.01 to 0.1 viral particles per cell and cloning the infected cells in agarose immediately. When the clones had grown enough to have their viral DNA content determined by renaturation kinetics, the average number of copies of viral DNA per cell varied between 10 and 1000 among the different clones (Sugden et al., J . Virol. in press). This finding indicates that early in the life of the EBV-transformed cell, viral DNA synthesis is not tightly coupled with cell DNA synthesis: however, later in the history of the cell some kind of equilibrium is reached, because those clones with an average number of copies of viral DNA per cell between 100 and 1000 stabilize at their respective numbers. Most clones of transformed cells harbor small populations of cells that release infectious virus. This phenomenon is detectable several cell divisions after exposure of the target cells to the virus. I n those clones described above, the number of virus-producing cells in each clone does not correlate with the average number of copies of viral DNA per cell, and for most of the clones about I cell per 2000 to 1 per 10,000 releases virus (Sugden et al., J . Virol., in press). When cells other than adult human leukocytes are transformed by EBV, different percentages of their populations release virus. In particular, cotton-topped marmoset cell lines transformed by EBV in vitro uniformly release more virus than do their human counterparts (Miller and Lipman, 1973). U p to 10% of the marmoset cells may contain viral capsid antigens and, presumably, most of those cells will release virus. This difference between adult human and marmoset transformed cells indicates that the regulation of expression of viral functions in the cell is at least in part determined by that cell. The expression of viral functions in some EBV-transformed cells can be modulated exogenously in three ways: ( I ) infecting the cell with a nontransforming strain of EBV (Henle et ul., 1970b: Traul et al., 1977: Yajima et a / . , 1978): (2) exposing the cells to halogenated pyrimidines (Gerber, 1972: Hampar et a / . , 1972): (3) exposing the cells to the tumorpromoting agent, tetradecanoyl phorbol acetate (TPA) (zur Hausen et al., 1978). All three methods lead to the new expression of EBV-associated an-
258
BILL SUGDEN ET AL.
tigens in some of the affected cells, and all can lead to new or increased release of some infectious virus. The addition of TPA to EBV-transformed cotton-topped marmoset cell lines is particularly useful in stimulating many of the cells to release infectious virus (zur Hausen ef a l . , 1978). This finding permits sufficient EBV to be harvested so that some formerly impractical molecular biological experiments now become feasible. The mechanisms of modulating viral expression in EBV-transformed cells are not now understood.
VI. Identification and Properties of Virus-Related Products for MDV, HVS, and EBV
Many viral-induced antigens in infected cells have been detected. Some of these antigens are likely to be coded for by their respective viruses, but which ones are virus-coded has not been established. Few of the antigens have been purified, so that the antisera used to define most antigens come from infected hosts: for EBV this means that most of the antiserum comes from tumor patients who may be immunosuppressed and will be infected with many viruses. Much of this work has been reviewed recently (Nazerian, 1973; Klein, 1973, zur Hausen, 19751, so that we shall focus on known viral components, on those antigens that our current biases lead us to think are most likely to be coded for by their respective viruses, and on those virus-associated antigens that have recently been identified or purified. Most studies on defined viral products done on HVS and EBV have focused on the structure of the viral DNA that has been extracted from virions or from transformed cells. For HVS and EBV the DNA is the one viral product that can be unambiguously identified and routinely purified. The viral DNA of MDV is difficult to purify because its density does not permit ready separation from host cell DNA. Some work cataloging the number and map positions of viral RNA sequences in EBV-infected cells has been undertaken (Hayward and Kieff, 1976: Orellana and Kieff, 1977; Powell et ul., in press). These workers have found that the complexity of viral RNA expressed in cells that do not release virus ranges between 25% and 50% of that found in the cells from which they harvest virus. This work is obviously difficult both because of the complexity of EBV DNA and because so little of the DNA is available. Our recognition that animal viruses construct their messenger RNA molecules by splicing together sequences transcribed from distant regions of the DNA template adds yet another level of difficulty to the study of viral RNA in cells infected with lymphotropic herpesviruses (Berget et ul., 1977; Chow et al., 1977; Klessig, 1977).
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
259
A. PROTEINS I N THE VIRION Some work following a similar design has been done on the virion proteins of both MDV and EBV. Infected cells have been labeled with radioactive amino acids or sugars, and extracellular virions have been partially purified by various methods of differential centrifugation. The recovered virions were disrupted in mercaptoethanol and sodium dodecylsulfate, and labeled polypeptides were separated by electrophoresis in polyacrylamide gels and detected by cutting and counting the gels or by autoradiography. Chen et al. (1972) found eight polypeptides, two of which were glycosylated, in their preparations of MDV. Kieff and his colleagues (Dolyniuk et al., 1976a, b) found 33 polypeptides, 12 of which were glycosylated, in their preparations of EBV. These experiments serve as the foundation for further work to elucidate which of the identified polypeptides are constituents of the virions. It is likely that there are more polypeptides (particularly in the case of MDV) that are also components of the virions. Because herpesviruses are notoriously difficult to purify, further work must incorporate several independent criteria for identifying any polypeptide as a constituent of either an MDV or EBV viral particle.
B. DNA
I N THE
VIRION
The density of MDV DNA (and that of HVT DNA) is 1.706 gm/cm3 and its molecular weight 1 x lo8. In conjunction with sedimentation studies in alkaline sucrose, these findings indicate that the viral DNA is a double-stranded, linear molecule (Lee et al., 1971; Lee, 1972), containing 46% guanine plus cytosine residues. This value is quite close to the 42% G + C content of chicken and duck DNA. This similarity in density provides a major difficulty in purifying MDV DNA free from contaminating host sequences by equilibrium sedimentation, which is a common method for purifying the viral DNA of different herpesviruses. HVS DNA is the most thoroughly characterized of the three prototypic lymphotropic herpesviruses. Fleckenstein and Wolf (1974) found that the density of HVS DNA was 1.709 gm/cm3 and its molecular weight was 0.91 f 0.05 x lo8 daltons. They also observed that, after shearing, two populations of viral molecules were generated with remarkably different densities: a heavy moiety with a density of 1.729 gm/cm3, and a light one of 1.694 gm/cm3. Later work has shown that only the full-length molecule of intermediate density, containing both covalently attached light and heavy moieties, is infectious in tissue culture (Fleckenstein et al., 1975). In addition, the moiety of higher density was found to contain repetitious DNA (Fleckenstein et al., 1975). A detailed study of the structure of
260
BILL SUGDEN ET A L .
HVS DNA isolated from virions has been carried out with denaturation mapping by electron microscopy and separation of restriction endonuclease digestion products by gel electrophoresis (Bornkamm et al., 1976). The full-length, infectious molecule is a linear, double-stranded DNA of 1.01-1.05 x lo8 daltons. This molecule is composed of a low-density moiety, which has a mass of 0.71 x lo8 daltons and is flanked on both sides by differing lengths of the high-density moiety such that the sum of the masses of both high-density moieties on each molecule varies between 0.3 and 0.35 x lo8 daltons (see Fig. I ) . This means 30% of the mass of HVS DNA is devoted to its terminal repetitions. This more recent study (Bornkamm et al., 1976) has found the repeat length of the high-density moiety DNA to be 0.83 X los daltons. Some preliminary results have shown that treatment of the viral DNA with an exonuclease can lead to its intramolecular circularization (Bornkamm et al., 1976). It is not known if this repeated DNA is transcribed (it could code for up to 3 x lo4 daltons of protein) or whether it performs only a structural role in the life cycle of HVS. Purified populations of HVS DNA contain both infectious and noninfectious molecules. Among the noninfectious molecules there is a population of defective molecules that consists solely of the repetitive se-
HVS
UNIOUE DNA
EBV
TR t
l-RDl-1 .----I-
TR i
FIG. I . These drawings represent the structures of the duplex DNAs isolated from virions of HVS and EBV. Sources of the data used to construct these representations are cited in the text. Both viral DNAs are approximately IOR daltons in mass. HVS: The density of repetitious DNA (RD,,, RDI.) is 1.729 gm/cm3. The density of unique DNA is 1.6V4gm/cmD. RDI and RD,, vary in length. The sum of RD,, and RD,. is constant and represents 3W of the H V S genome. E B V The density of EBV DNA is 1.717 gm/cm3. Repetitious DNA (RDJ represents 21% of the EBV genome. The terminal repeat (TR) is about 200- 1000 base pairs long.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
26 1
quences of high density normally found at the ends of the infectious HVS molecules. The generation of HVS defective molecules differs from that of defective molecules found in other animal viruses, because the HVS defective molecules arise at high frequency and independently of the multiplicity of infection. It has been proposed that some structural feature of the repetitive DNA is responsible for the ease of generation of defective molecules (Fleckenstein et al., 1978). Much work has been done to determine the structure of EBV DNA, but progress has been slow. This slow progress surely results from the small quantities of viral DNA available and from our inability to label it isotopically in vivo to a high specific activity. These problems again are consequences of our lack of a productive host cell for EBV. The virion DNA of two strains of EBV, P3HRl (Hinuma ef al., 1967) and B95-8 (Miller and Lipman, 1973) has been studied. The DNA isolated from a nontransforming strain of EBV, P3HR1, has been analyzed by means of renaturation kinetics, digestion with restriction endonuclease, and denaturation mapping (Pritchett et a / . , 1975: Sugden et al., 1976: Hayward and Kieff, 1977: Rymo and Forsblom, 1978: Delius and Bornkamm, 1978). This work in sum indicates that the DNA of the P3HRI strain is heterogeneous. Molecules with different sequence arrangements are found in each preparation of P3HRI DNA. In addition, biological experiments indicate that pools of this strain of EBV are heterogeneous in that exposure of homogeneous target cells to a highly diluted inoculum leads to expression of different virus-associated antigens in different individual cells (Fresen et al., 1977; zur Hausen and Fresen, 1977). We shall review only the work carried out on the B95-8 strain of EBV because that strain transforms cells, has a high infectivity-to-particle ratio (Henderson et al., 1977: Sugden and Mark, 1977), and its DNA is homogeneous as determined by partial denaturation mapping (Delius and Bornkamm, 1978). EBV DNA extracted from B95-8 virions is a double-stranded, linear molecule with a density of 1.716-1.718 gm/cm3 [as was first determined for the P3HRl strain (Schulte-Holthausen and zur Hausen, 1970)] and has a molecular weight of 1.01-1.05 x lo8 daltons as determined by length measurements in the electron microscope (Pritchett et a / . , 1975). The linear molecule is terminally repetitious. The repetitions have been demonstrated by digesting the molecules with A-exonuclease, renaturing, and observing circular molecules in the electron microscope (Kintner and Sugden, Cell, in press). When an average of 0.75% of each molecule is digested with A-exonuclease, up to 40% of the molecules observed are circles. This observation indicates that the length of the terminal repeat is approximately 1000 base pairs on each end of the molecule. Measurement of the joint region where the reannealed ends of extensively digested
262
BILL SUGDEN ET AL.
molecules overlap indicates that the length of the terminal redundancy is on the order of 200-1000 base pairs. Sugden (1977) cleaved viral DNA with the Bam restriction endonuclease and separated the products electrophoretically in agarose gels. The results indicated that one fragment of 1.8 x lo6 daltons was present in several copies per molecule. Rymo and Forsblom (1978), using a sensitive method of quantification, found I I copies of that fragment per molecule. Therefore, 20% of EBV DNA is repetitious with a complexity of 1.8 x lo6 daltons or less. I n addition, Rymo and Forsblom (1978) found that the repetitious DNA sequences all lay within one DNA fragment generated by digestion with the Eco RI restriction endonuclease, and that fragment has subsequently been mapped to lie between 8% and 32% from one end of the molecule (Given and Kieff, 1978). All of these findings are roughly consistent with a denaturation map of the B95-8 strain of EBV DNA produced by Delius and Bornkamm (1978). In sum, the data indicate that the DNA of the B95-8 strain of EBV is a linear molecule with terminal repetitions of approximately 200-1000 base pairs at each end and with a tract of internal repeated sequences that begins 8% from one end of the molecule and proceeds to about 28% from that end. It is important to recognize that the structure of EBV DNA is similar to that of HVS DNA in that it contains a significant amount of repetitious sequences (20% of EBV versus 30% of HVS), but is strikingly different in that its repetitious DNA lies as a single tract bounded at both ends by nonrepeated sequences. A schematic drawing comparing HVS and EBV DNA is presented in Fig. I . Recent findings indicate that a portion of the repeated DNA sequences in EBV are transcribed (Powell er af., submitted) but, as with HVS, it is not known if they are translated.
C. VIRUS-INDUCED ANTIGENS A N D PROTEINS I N INFECTED CELLS Virus-associated antigens were detected in MDV-infected cells soon after tissue culture methods were developed for this virus (Churchill et af., 1969b). One of these, the A antigen, was present in extracellular fluid and was lost from the infected cultures during prolonged passage in vitro (Churchill er al., 1969b). This passaging in vitro also led to attenuation of the virus. It was later found that some strains of MDV can be pathogenic and not express the A antigen (Purchase el af., 1971; von Biilow, 1971). Therefore, it appears that the A antigen is not required for the pathogenicity or for the infectivity of MDV. This antigen was found to be a glycoprotein (Ross er al., 1973), and subsequent work has found its molecular weight to be about 45,000 (Long er al., 1975a; Long et af.,
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
263
1975b). The B antigen has been partially purified from infected cells and also found to be a glycoprotein with a molecular weight of 55,000 (Velicer et al., 1978). The function of these antigens and their relationship to MDV are not known. A Marek’s disease tumor-associated surface antigen (MATSA) has been identified by antisera raised in rabbits or chickens against established MDV tumor cell lines (Witter et af., 1975; Matsuda et al., 1976). After adsorption these sera stain MDV tumor cells and tumor cell lines in immunofluorescence assays but do not stain normal tissue or fibroblast cultures lytically infected with MDV (Witter et al., 1975; Matsuda ef al., 1976). The antigen, therefore, is not necessarily associated with the presence of MDV but rather with MDV-induced tumor cells. Accordingly, MATSA is not found on other types of lymphoid tumors of the chicken. For example, it is not present on cells of a transplantable lymphoid tumor induced by an RNA leukosis virus (Witter er al., 1975). The variety of MDV tumor cells on which MATSA has been found indicates that it is not likely to be a histocompatibility antigen. The MATSA antigen may be an alloantigen that is specific to the differentiated cells infected by MDV. Tumor virus-associated alloantigens have been defined in the mouse (Stockert et af., 1971). An MDV-induced DNA polymerase activity has been identified and partially purified from infected duck embryo fibroblast cultures (Boezi et al., 1974). The enzyme has an apparent molecular weight of 100,000 and prefers activated DNA as a template. These two characteristics, in addition to the enzyme’s chromatographic behavior on phosphocellulose columns, distinguish it from the DNA polymerase activities present in uninfected cells. All herpesviruses studied have been shown to induce a new DNA polymerase activity in infected cells. These observations have conditioned virologists to view favorably the notion that herpesviruses code for a DNA polymerase. This idea, however, has so far received strong support only from studies of herpes simplex virus that have employed temperature-sensitive mutants (Aron et al., 1975). Of the three prototypes of lymphotropic herpesviruses, it would seem that HVS lends itself best to the study of its associated proteins and antigens. It infects established cell lines efficiently, has a reasonably short growth cycle, and is released into the extracellular fluid. However, of the three, it is the most recent to be studied and, consequently, little is known about its associated proteins. On the other hand, a great many observations have been made in EBVinfected cells on virus-associated antigens that are detected by patients’ antisera. These data have been reviewed by Klein (1975) and zur Hausen (1975). Recently work has progressed on characterizing the EBV nuclear
264
BILL SUGDEN ET AL.
antigen (EBNA) that was first identified by anticomplement immunofluorescence techniques (Reedman and Klein, 1973). This antigen is particularly useful in that it appears to be uniformly associated with all cells infected by EBV (Lindahl et al., 1974). EBNA has been purified to apparent homogeneity (Luka et al., 1978) and has been partially characterized (Lenoir et d . , 1976; Matsuo et d . , 1977; Luka et d . , 1977, 1978; Baron and Strominger, 1978). The molecular weight of the native antigen is about 200,000, whereas in denaturing gels the most purified preparations contain only one polypeptide with a mobility that corresponds to a molecular weight of 48,000 (Luka et al., 1978). The active species, therefore, appears to be a tetramer. It binds to double-stranded DNA cellulose columns in the presence of 0.2 M NaCl but is eluted at 0.4 M NaCl at a pH of 7.4. In two of its characteristics EBNA is reminiscent of the simian virus 40 (SV40) large T antigen. Both antigens are found in the nucleus of all virally infected cells, and both bind to double-stranded DNA. The SV40 large T antigen binds preferentially to a specific sequence of DNA near or at the site for initiation of D N A synthesis on the SV40 DNA molecule (Tijan, 1977). It is not now known if EBNA shows any sequence specificity of binding to double-stranded DNA. An EBV-associated cell surface antigen has been identified with rabbit antisera raised against two Burkitt’s lymphoma cell lines (Sakamoto and Hinuma, 1978). After the sera were adsorbed with normal cells, at least some of the cells of most lines infected by EBV were stained in immunofluorescent assays (Sakamoto and Hinuma, 1978). In this respect the EBV-associated cell surface antigen apparently differs from MATSA, which is found only on tumor cells. However, only EBV-infected lymphoid cells have been studied, so that the virus-associated surface antigen may also be a lymphoid differentiation antigen or, perhaps, a candidate for the target recognized by the immune killer T cells found in patients during the acute stage of infectious mononucleosis (Svedmyr and Jondal, 1975)* Three enzymic activities have been studied in cells that have been induced to express EBV-associated antigens by treating them with halogenated pyrimidines or by superinfecting them with the nontransforming strain of EBV. A ribonucleotide reductase activity resistant to hydroxyurea (Henry et al., in press), a thymidine kinase activity with a distinct electrophoretic mobility (Chen et al., 1978), and a DNA polymerase activity active in 50 mM ammonium sulfate (Miller et a / . , 1977b) have been shown to differ in some way from the corresponding activities in uninduced cells. These enzymic activities, found in the induced cells, have their counterparts in cells infected by other herpesviruses and, therefore, may be host enzymes that are generally induced by different herpesviruses or may be encoded by the viruses themselves.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
D. VIRAL DNA
IN
265
TRANSFORMED CELLS
Viral DNA has been detected in tumor cells, in lines established from tumor biopsies, and in cells transformed in vitro by MDV, HVS, and EBV. Nazerian and Lee (1974) found one MDV tumor cell line to harbor 60-90 genome equivalents of MDV DNA by using the technique of hybridizing complementary viral RNA to cell DNA immobilized on nitrocellulose filters. Nonoyama and his colleagues have studied the structure of this intracellular viral DNA. Using velocity sedimentation analyses in neutral and alkaline glycerol gradients and equilibrium sedimentation in gradients of CsCl plus ethidium bromide, they have shown that at least 80%-90% of the viral DNA is extrachromosomal and behaves as a covalently closed, supercoiled molecule of full length (Tanaka et a/., 1978). It is not yet known if all of the nucleotide sequences of the virion DNA are present in these extrachromosomal molecules. Multiple copies of HVS DNA have been found in tumor biopsy cells and in nonproducer tumor cell lines with the technique of renaturation kinetics (Fleckenstein et a / . , 1976, 1977). The structure of HVS DNA in one nonproducer cell line has been carefully studied by means of denaturation mapping by electron microscopy and analysis of cleavage products generated by digestion with restriction endonucleases (Werner et a / . , 1977, in press). The viral DNA again has been found to be present as an extrachromosomal, covalently closed, supercoiled molecule. However, it has a molecular weight of l .3 x lo8, which is 30% larger than the virion’s molecule, and is composed of two low-density and two highdensity moieties with a portion of each of the low-density moieties being deleted from the intracellular viral DNA (Werner et al., 1977, in press). The study of intracellular viral DNA of lymphotropic herpesviruses was first begun with EBV when zur Hausen and Schulte-Holthausen (1970) detected viral nucleotide sequences in a cell line established from a Burkitt’s lymphoma biopsy. Multiple copies of EBV DNA were detected in both Burkitt’s lymphoma and nasopharyngeal carcinoma biopsies (zur Hausen el a/., 1972: Nonoyama and Pagano, 1973). Nonoyama and Pagano (1973) introduced the technique of isotopically labeling EBV DNA to high specific activity by “nick translation,” using DNA polymerase I from E . coli (Kelly et a / . , 1970). This technique is particularly important for the detection of EBV (and MDV) DNA because it is not practicable to label the viral DNA in vivo, and complementary RNA synthesized in vitro with DNA-dependent RNA polymerase is probably not homologous to all viral DNA sequences and is certainly not a uniform distribution of those sequences (B. Sugden, unpublished results). The location of the bulk of intracellular EBV DNA is extrachromosomal (Nonoyama and Pagano, 1972: Tanaka and Nonoyama, 1974) and consists
266
BILL SUGDEN ET AL.
of covalently closed, supercoiled molecules of about full length (Lindahl 1976). Within the limit of analysis of DNA by electrophoretic separation of cleavage products of restriction endonucleases, intracellular viral DNA in Burkitt’s lymphoma biopsies and in vitro transformed cell clones contains all the nucleotide sequences present in virion DNA (Sugden, 1977). This finding is strengthened by the observation that most clones of cells transformed in vitro by EBV release infectious particles, albeit at a very low level (Sugden et af.,J . Virol., in press), In some cell lines transformed by EBV a small portion of viral DNA is integrated into the cell genome (Lindahl et al., 1976). This conclusion is based on experiments that separate free viral from cell DNA by centrifugation in equilibrium sedimentation gradients. After three consecutive centrifugations about 1% of the viral nucleotide sequences are found in the position of cell DNA (Kaschka-Dierich et al., 1977). Upon shearing and recentrifugation these viral sequences shift their position to the density characteristic of free viral DNA. With similar techniques, tumor biopsy cells, tumor cell lines, and cell lines established from EBV-seropositive donors were found to harbor a small percentage of their viral DNA as integrated sequences (Lindahl et al., 1976; Adams et al., 1977; Kaschka-Dierich et al., 1977). Lindahl and his colleagues have identified one in vitro transformed cell line that lacked detectable integrated viral DNA (Kaschka-Dierich el al., 1977). Because so little DNA is integrated, it is technically difficult to demonstrate directly the presence or absence of integrated viral nucleotide sequences in a given cell line. Without direct experiments to measure the complexity of integrated viral DNA, it remains difficult to evaluate the role this information might play in transformation by EBV.
et al.,
VII. Conclusion
Are there reasons other than convenience to discuss lymphotropic herpesviruses separately from other members of the group? Two reasons are immediately apparent: ( I ) only the lymphotropic herpesviruses infect and transform lymphoid cells in their natural host; (2) only the lymphotropic herpesviruses induce lymphoid tumors in natural or experimental hosts. [With the possible exception of the Luck6 frog virus (Granoff, 19731, the lymphotropic herpesviruses are the only ones of this family known to induce tumors directly in any hosts.] These distinct features may be a function only of the tissue tropism of MDV, HVS, EBV, and their close relatives or may reflect genetic information peculiar to them. MDV, HVS, and EBV permit us to establish lymphoid cell lines of
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
267
chickens, cotton-topped marmosets, and human beings. Most established chicken cells of any kind are MDV-transformed T cells; most cottontopped marmoset lymphoid cell lines are transformed by HVS o r EBV; all B-lymphoid cell lines established from normal human donors and most lymphoid lines established from human tumor patients are transformed by EBV. Transformation by EBV in vitro is functionally equivalent t o the establishment of B-lymphoblast cell lines. It bears repeating that transformation by EBV is qualitatively distinct from transformation by papova-, adeno-, and other herpesviruses. In transformation by EBV, a B lymphocyte is converted into a permanently dividing B lymphoblast. The state of the viral DNA in cells transformed by MDV, HVS, and EBV is strikingly different from that usually found in cells transformed by papova-, adeno-, and other herpesviruses. In the latter cases, one o r a few copies of the viral DNA are usually found to be integrated into the host cell DNA (Sambrook et al., 1968), and only fragments of the viral DNA with as small a complexity as 1-2 x lo6 daltons may be present (Sambrook et al., 1974; Graham et al., 1974; Kraiselbard et al., 1975). Thus only 50% of the papovaviruses, 5%-1% of the adenoviruses, and 1%-2% of herpes simplex virus at most are required to maintain the transformed state. This information is maintained in the cell, presumably by dint of its being covalently linked to cell DNA. With MDV and EBV, copies of all or nearly all of the viral DNA are extrachromosomal; with HVS, in two cases, the majority but not all of the viral DNA is present extrachromosomally in multiple copies. The multiple copies of viral DNA in MDV-, HVS-, and EBV-transformed cells may result from a mechanism evolved to guarantee that both daughter cells would receive at least one copy. Obviously, in view of the limitations of our knowledge, other hypotheses involving gene dosage might also explain the observed features of viral DNA in cells transformed by lymphotropic herpesviruses. The viral DNAs of MDV, HVS, and EBV, within the limits of detection (1%-5% for HVS and EBV, more for MDV), share no sequence homology with their hosts’ DNA. In addition, within our ability to discern it, the virus and viral DNA in EBV tumor cells from human patients are not different from the virus and its DNA in cells cultured in the laboratory. We conclude, therefore, that EBV and, by analogy, MDV carry sufficient information to transform lymphocytes, that is, to infect some precursor cell in vivo and alter it such that it can replicate permanently in vitro. In addition, from among the cells transformed by MDV, neoplastic progeny are generated efficiently in genetically susceptible animals. Prominent goals in working with MDV, HVS, and EBV are to identify those gene products required for the initiation and maintenance of the transformed state, t o elucidate their mechanisms of action, and to deter-
268
BILL S U G D E N E T AL.
mine what role they play in contributing to the neoplastic phenotype. We need to overcome several major problems in order to achieve these goals. We need much information on normal lymphoid cells in vivo and in vitro, including how to identify different classes of leukocytes and their precursors and how to culture their precursors in vitro. We need in vitro transformation assays for MDV and HVS. Finally, we need to grow MDV in large quantities in a cell-free form and to be able to plaquepurify EBV and to propagate it efficiently in a productive host. These needs will be difficult to fulfill, but without them progress toward our goals will be painfully slow and confusingly indirect.
ACKNOWLEDGMENTS We thank our many colleagues who kindly sent us their papers prior to publication so that we might include their findings in this review. We thank Dr. George Miller of Yale University in particular for providing us with his unpublished findings which constitute the bulk of Table IV. Finally, we thank Drs. Ilse Riegel and Jeff Ross of the McArdle Laboratory for critically helping us with this manuscript. We were supported by National Institutes of Health Grants CA 07175, CA 22443, and CA 09135.
ADDENDUM In our discussion of transformation by EBV we referred to the work of Thorley-Lawson et ul. (1977), who found that the presence of adult T-cells inhibited transformation of
peripheral leukocytes by EBV. Those authors used target cells from both EBV seronegative and seropositive donors and did not distinguish between them in their protocols nor in their results. In a recent similar study by Moss et ul. (1978). these workers found that T-cells inhibited transformation of autochthonous cells only from donors who were seropositive for antibodies to EBV. These new findings obviate our alternative interpretation of the work of Robinson et a / . ( 1977) with cells of seronegative marmosets but do not change our reflections on the work of Epstein and his colleagues, who worked with cells of infectious mononucleosis patients [Rickinson rt a / . (1975): Rickinson (11. ( 1977)l.
ADDENDUM REFERENCE Moss, D. .I.Rickinson, A. B., and Pope, J. H (1978), Long-term T-cell mediated immunity to Epstein-Barr virus in man. I. Complete regression of virus-induced transformation in cultures of sero-positive donor leukocytes. I u t . J . Cuncer 22, 662-668
REFERENCES Adams, A., Bjursell, G., Kaschka-Dierich. C., and Lindahl, T. (1977). Circular EpsteinBarr virus genomes of reduced size in a human lymphoid cell line of infectious mononucleosis origin. J . Virol. 22, 373-380.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
269
Andersson-Anvert, M., Forsby, N., Klein, G., and Henle, W. (1977). Relationship between the Epstein-Barr virus and undifferentiated nasopharyngeal carcinoma: Correlated nucleic acid hybridization and histopathological examination. Int. J. Cuncer 20, 486494. Andiman, W. A., and Miller, G. (1978). Properties of Epstein-Barr virus transformed woolly monkey lymphoblastoid cell lines. Proc. Soc. Exp. B i d . Med. 157, 489-493. Aron, G . M., Purifoy, D. J . M., and Schaffer, P. A. (1975). DNA synthesis and DNA polymerase activity of herpes simplex virus type I temperature-sensitive mutants. J . Virol. 16, 498-507. Bar, R. S., DeLor, C. J., Clausen, K. P., Hurtubise, P., Henle, W., and Hewetson, J . F. (1974). Fatal infectious mononucleosis in a family. N . Engl. J. Met/. 290, 363-367. Baron, D., and Strominger, J . L. (1978). Partial purification and properties of the EpsteinBarr virus-associated nuclear antigen. J. B i d . Chem. 253, 2875-288 I . Benjamin, T. L. (1965). Relative target sizes for the inactivation of the transforming and es of polyoma virus. Proc. Nutl. Acrid. Sci. U . S . A . 54, 121-124. Berget, S. M., Moore, C . , and Sharp, P. A. (1977). Spliced segments at the 5 ‘ terminus of adenovirus 2 late mRNA. Proc. Natl. Acud. Sci. U . S . A . 74, 3171-3175. Biggs, P. M., Churchill, A. E., Rootes, D. G., and Chubb, R. C. (1968). The etiology of Marek’s disease-an oncogenic herpes-type virus. In ”Perspectives in Virology” ( M . Pollard, ed.), pp. 21 1-237. Academic Press, New York. Boezi, J . A,, Lee, L. F., Blakesley, R. W., Koenig, M., and Towle, H. C. (1974). Marek’s disease herpesvirus-induced DNA polymerase. J. Virol. 14, 1209- 12 19. Bornkamm, G. W., Delius, H., Fleckenstein, B., Werner, F.-J., and Mulder, C. (1976). Structure of Herpesviriis suimiri genomes: Arrangement of heavy and light sequences in the M genome. J. Virol. 19, 154-161. Briles, W. E., Stone, H . A., and Cole, R. K. (1977). Marek’s disease: Effects of B histocompatibility alloalleles in resistant and susceptible chicken lines. Science 195, 193- 195. Calnek, B. W., and Hitchner, S. B. (1969). Localization of viral antigen in chickens infected with Marek’s disease herpesvirus. J . Natl. Cancer Inst. 43, 935-949. Calnek, B. W., Adldinger, H. C., and Kahn, D. E. (1970). Feather follicle epithelium: A source of enveloped and infectious cell-free herpesviruses from Marek’s disease. AiYrrn Dis. 14, 219-233. Chang, R. S., and Golden, H. D. (1971). Transformation of human leucocytes by throat washing from infectious mononucleosis patients. Nutitre (London)234, 359-360. Chen, J . H., Lee, L. F., Nazerian, K., and Burmester, B. R. (1972). Structural proteins of Marek’s disease virus. Virology 47, 434-443. Chen, S.-T., Estes, J . E., Huang, E.-S., and Pagano, J . S . (1978). Epstein-Barr virusassociated thymidine kinase. J. Virol. 26, 203-208. Chow, L. T., Gelinas, R. E., Broker, T . R., and Roberts, R. J . (1977). A n amazing sequence arrangement at the 5’ ends of adenovirus 2 messenger RNA. Cell 12, 1-8. Churchill, A. E., and Biggs, P. M. (1967). Agent of Marek’s disease in tissue culture. Nutitre (London) 215, 528-530. Churchill, A. E., Payne, L. N., and Chubb, R. C. (1969a). Immunization against Marek’s disease using a live attentuated virus. Natirre (London) 221, 744-747. Churchill, A. E., Chubb, R. C., and Baxendale, W. (1969b). The attentuation, with loss oncogenicity, of the herpes-type virus of Marek’s disease (strain HPRS-16) on passage in cell culture. J. Gen. Virol. 4, 557-564. Cole, R. K. (1972). The genetics of resistance to Marek’s disease. I n “Oncogenesis and Herpesviruses” (P. M. Biggs, G . de-The, and L. N. Payne, eds.), pp. 123-128. International Agency for Research o n Cancer, Lyon.
270
BILL SUGDEN ET AL.
Daniel, M. D., Hunt, R. D.. Dubose, D., Silva, D., and Melendez, L . V. (1975). Induction of Herpesvirus saimiri Lymphoma in New Zealand white rabbits inoculated intravenously. I n "Herpesviruses and Oncogenesis, 11" (G. de-The, M . A. Epstein, and H. zur Hausen. eds.), pp. 205-208. International Agency for Research on Cancer, Lyon. Deinhardt, F. W., Falk, L. A., and Wolfe, L. G . (1974). Simian herpesviruses and neoplasia. Adv. Cciiicer Res. 19, 167-205. Deinhardt, F., Falk. L . , Wolfe, L. G., Paciga, J . , and Johnson, D. (1975). Response of marmosets to experimental infection with Epstein-Barr virus. I n "Oncogenesis and Herpesviruses 11'' (G. de-The. M. A. Epstein, and H. zur Hausen. eds.), pp. 161-168. International Agency for Research on Cancer, Lyon. Delius, H., and Bornkamm, G. W. (1978). Heterogeneity of Epstein-Barr virus. 111. Comparison of a transforming and a nontransforming virus by partial denaturation mapping of their DNAs. J . Virol. 27, 81-89. de-ThC, G., Geser, A., Day, N . E., Tukei, P. M., Williams, E . H., Beri, D. P., Smith, P. G . . Dean. A. G.. Bornkamm, G. W., Feorino, P., and Henle, W. (1978). Epidemiological evidence for causal relationship between Epstein-Barr virus and Burkitt's lymphoma: Results of the Ugandan prospective study. Nutrrrc, (London)274, 756-761, Doherty, P. C., Blanden, R. V., and Zinkernagel, R. M. (1976). Specificity of virus-immune effector T cells for H-2K and H-2D compatible interactions: Implications for H-antigen diversity. Trcoisplan/. Rei,. 29, 89-124. Dolyniuk, M.. Pritchett, R., and Kieff, E. (1976a). Proteins of Epstein-Barr virus I . Analysis of the polypeptides of purified enveloped Epstein-Barr virus. J . Virol. 17, 935-949. Dolyniuk, M., Wolff. E., and Kieff, E. (1976b). Proteins of Epstein-Barr virus 11. Electrophoretic analysis of the polypeptides of the nucleocapsid and the glucosamine- and polysaccharide-containingcomponents of enveloped virus. J . Virol. 18, 289-297. Edenberg. H. J., Anderson, S . , and De Pamphilis, M. L. (1978). Involvement of DNA polymerase (I in simian virus 40 DNA replication. J . Biol. Chern. 253, 3273-3280. Eidson, C. S . , and Anderson, D. P. (1971). Immunization against Marek's disease. Avirrrr Dis. 15, 68-81. Einhorn, L., and Ernberg, I . (1978). Induction of EBNA precedes the first cellular S-phase after EBV-infection of human lymphocytes. I n t . J . C a n w r 21, 157-160. Epstein, M. A., and Barr, Y. M. (1964). Cultivation in iitro of human lymphoblasts from Burkitt's malignant lymphoma. Lancet 1, 252. Falk, L. A., Nigida, S . . Deinhardt, F., Cooper, R. W., and Hernandez-Camacho, J. I . ( 1973). Oral excretion of herpesvirus saimiri in captive squirrel monkeys and incidence of infection in feral squirrel monkeys. J . N u t / . Cancer Inst. 51, 1987-1989. Falk, L.. Wright, J., Wolfe, L., and Deinhardt. F. (1974a). Herpesvirrrs arrlrs: transformation in vitro of marmoset splenic lymphocytes. I n / . J . Cuncer 14, 244-25 I. Falk, L . A., Nigida, S. M., Deinhardt, F., Wolfe. L. G., Cooper, R. W., and HernandezCamacho, J. I . ( l974b). Herpesvirus atrles: Properties of an oncogenic herpesvirus isolated from circulating lymphocytes of spider monkeys (ateles sp.). I n t . J . C'crricer 14, 473-482. Falk. L., Deinhardt, F.. Nonoyama. M., Wolfe, L. G.. Bergholz, C . , Lapin, B., Yakovleva, L., Agrba, V., Henle, G., and Henle, W. (1976). Properties of a baboon lymphotropic herpesvirus related to Epstein-Barr virus. I n / . J . Cancer 18, 798-807. Falk, L. A., Henle. G., Henle, W., Deinhardt, F., and Schudel, A. (1977). Transformation of lymphocytes of Herpesvirris papio. Int. J . Cancer 20, 219-226. Falk, L. A., Johnson, D., and Deinhardt, F. (1978). Transformation of marmoset lymphocytes in iti/ro with Hrrpc,svirns uteles. I n t . J . Cancer 21, 652-657.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
27 1
Fialkow, P. J., Klein, G., Gartler, S. M., and Clifford, P. (1970). Clonal origins of individual Burkitt tumors. Lanccjt 1, 384-386. Fleckenstein, B., and Wolf, H. (1974).Purification and properties of Hivpesiiirus saimiri DNA. Virology 58, 55-64. Fleckenstein, B., Bornkamm, G . W., and Ludwig, H . (1975). Repetitive sequences in complete and defective genomes of Herpesvirus saimiri. J. Virol. 15, 398-406. Fleckenstein, B., Bornkamm, G. W., and Werner, F. J. (1976). The role of Herpesiirus saimiri genomes in oncogenic transformation of primate cells. I n “Comparative Leukemia Research 1975, Bibl. Haemat., No. 43” (J. Clemmesen and D. S. Yohn, eds.), pp. 308-312, Karger, Basel. Fleckenstein, B., Miiller, I., and Werner, J. (1977). The presence of Herpesvirus saimiri genomes in virus-transformed cells. Int. J. Cancer 19, 546-554. Fleckenstein, B., Bornkamm, G. W., Mulder, C., Werner, F.-J., Daniel, M. D., Falk, L. A., and Delius, H. (1978). Herpesvirus ateles DNA and its homology with Herpesvirus saimiri nucleic acid. J . Viorl. 25, 361-373. Frank, A., Andiman, W. A,, and Miller, G. (1976). Epstein-Barr virus and nonhuman primates: Natural and experimental infection. Adv. Cancer Res. 23, 171-201. Frederickson, T. L., Longenecker, B. M., Pazderka, F., Gilmour, D. G., and Ruth, R. F. (1977). A T-cell antigen system of chickens: Ly-4 and Marek’s disease. Inimunogcnetics 5, 535-552. Fresen, K.-0.. Merkt, B., Bornkamm, G.. and zur Hausen, H. (1977). Heterogeneity of Epstein-Barr virus originating from P3HR-I cells. I. Studies on EBNA induction. I n t . J. Cancer 19, 317-323. Gerber, P. ( 1972). Activation of Epstein-Barr virus by 5-bromodeoxyuridine in “virusfree” human cells. Proc. Natl. Acad. Sci. U . S . A . 69, 83-85. Gerber, P.. and Monroe, J. H. (1968). Studies on leukocytes growing in continuous culture derived from normal human donors. J . Natl. Cancer Inst. 40, 855-866. Gerber, P., Pritchett, R. F., and Kieff, E. D. (1976). Antigens and DNA of a chimpanzee agent related to Epstein-Barr virus: J . Virol. 19, 1090-1099. Gerber, P., Kalter, S. S., Schidlovsky, G., Peterson, W. D., Jr., and Daniel, M. D. (1977). Biologic and antigenic characteristics of Epstein-Barr virus-related herpesviruses of chimpanzees and baboons. fnr. J . Cancer 20,448-459. Given, D., and Kieff, E. (1979). DNA of Epstein-Barr virus I V . Linkage map of restriction enzyme fragments of 895-8 and W91 strains of EBV. J. Virol. 28, 524-542. Goldman, M., Landon, J. C., and Reisher, J. I. (1968). Fluorescent antibody and gel diffusion reactions of human and chimpanzee sera with cells cultured from Burkitt tumors and normal chimpanzee blood. Cancer Res. 28, 2489-2495. Graham, F. L., Veldhuisen. G., and Wilkie, N. M. (1973). Infectious herpesvirus DNA. Nature (London) New B i d . 245, 265-266. Graham, F. L., A b r a h a m , P. J., Mulder, C.. Heijneker, H. L., Warnaar, S. O., devries, F. A. J., Fiers. W., and Van der Eb, A. J. (1974). Studies on in vitro transformation by DNA and DNA fragments of human adenoviruses and Simian virus 40. Cold Spring Harbor Svmp. Qctanr. B i d . 39, 637-650. Granoff, A. (1973). The Luck6 renal carcinoma of the frog. In “The Herpesviruses” (A. Kaplan, ed.). pp. 627-640. Academic Press, New York. Hampar, B., Derge, J. G., Martos, L. M., and Walker, J. L. (1972). Synthesis of EpsteinBarr virus after activation of the viral genome in a ”virus-negative” human lymphoblastoid cell (Raji, made resistant to 5-bromodeoxyuridine. Proc. Nafl. Acail. Sci. U . S . A . 69, 78-82. Hayward, S. D., and Kieff, E. (1976). Epstein-Barr virus-specific RNA. I . Analysis of viral
272
BILL SUGDEN ET AL.
RNA in cellular extracts and in the polyribosomal fraction of permissive and nonpermissive lymphoblastoid cell lines. J. V i d . 18, 518-525. Hayward, S. D., and Kieff, E. (1977). The DNA of Epstein-Barr virus. 11. Comparison of the molecular weights of restriction endonuclease fragments of the DNA of strains of EBV and identification of end fragments of the B95-8 strain. J . Virol. 23,421-429. Henderson, E., Miller, G., Robinson, J., and Heston, L. (1977). Efficiency of transformation of lymphocytes by Epstein-Barr virus. Virology 76, 152- 163. Henderson, E., Heston, L., Grogan, E., and Miller, G. (1978). Radiobiological inactivation of Epstein-Barr virus. J . Virol. 25, 51-59. Henle, G.. Henle, W., Clifford, P., Diehl, V., Kafuko, G. W., Kirya, B. G.. Klein, G., Morrow, R. H., Munube, G . M. R., Pike, P., Tukei, P. M., and Ziegler, J. L. (1969). Antibodies to Epstein-Barr virus in Burkitt’s lymphoma and control groups. J . N a i l . Cancer Inst. 43, 1147-1 157. Henle, W.. Henle, G., Ho, H.-C., Burtin, P., Cachin, Y., Clifford, P., de Schryver. A , . de-The, G., Diehl. V., and Klein, G. (1970a). Antibodies to Epstein-Barr virus in nasopharyngeal carcinoma, other head and neck neoplasms, and control groups. J . Niitl. Cancer Inst. 44, 225-23 I . Henle, W., Henle, G . . Zajac, B. A., Pearson, G., Wdubke, R., and Scriba, M. (1970b). Differential reactivity of human serums with early antigens induced by Epstein-Barr virus. Sciencr 169, 188- 190. Henry, B. E., Glaser, R., Heuetson, J . , and O’Callaghan, D. (19781. Expression of altered ribonucleotide reductase activity associated with the replication of the Epstein-Barr virus. I n press. Hinuma, Y.,Konn, M . , Yamaguchi, J., Wudarski, D. J . , Blakeslee, J . R., Jr., and Grace, J. T., Jr. (1967). lmmunofluorescence and Herpes-type virus particles in the P3HR-I Burkitt lymphoma cell line. J. Virol. 1 , 1045-1051. Hinuma, Y., and Katsuki, T. (1978). Colonies of EBNA-positive cells in soft agar from peripheral leukocytes of infectious mononucleosis patients. Int. J . C n n w r 21,426-43 I . Hinze, H. (1971a). New member of the herpesvirus group isolated from wild cottontail rabbits. Infect. Irnmnn. 3, 350-354. Hinze, H. C. (1971b). Induction of lymphoid hyperplasia and lymphoma-like disease in rabbits by Hivp,svirr/.s syhiliigrrs. Int. J . Cnncrr 8, 5 14-522. Huber, C., Sundstriim, C. C.. Nilsson, K., and Wigzell, H. (1976). Surface receptors on human haematopoietic cell lines. Clin. Exp. Immrtnol. 25, 367-378. Hunt, R. D., Melendez, L. V . , King, N. W., Gilmore, C. E., Daniel, M. D., Williamson, M. E., and Jones, T. C. (1970). Morphology of a disease with features of malignant lymphoma in marmosets and owl monkeys inoculated with Hc,rpi,.si’ir//s suiniiri. J . Natl. Cunccv Inst. 44,447-465. Jackson, C. A. W.. Biggs, P. M., Bell, R. A , , Lancaster. F. M., and Milne. B. S. (1974). A study of vaccination against Marek’s disease with an attenuated Marek‘s disease virus. Ai~icrnPutIiuI. 3, 123-144. Jondal, M., and Klein, G. (1973). Surface markers on human B and T lymphocytes. I J . Presence of Epstein-Barr virus receptors on B lymphocytes. J . Exp. Med. 138, 136S1378. Kaschka-Dierich. C . , Falk. L.. Bjursell, G., Adams, A., and Lindahl, T. (1977). Human lymphoblastoid cell lines derived from individuals without lymphoproliferative disease contain the same latent forms of Epstein-Barr virus DNA a s those found in tumor cells. I n / . J . Cirnciv 20, 173-180. Katsuki, T., and Hinuma, Y . (1975). Characteristics of cell lines derived from human leukocytes transformed by different strains of Epstein-Barr virus. 1/11. J . C u n c w 15, 203-210.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
273
Katsuki, T., Hinuma, Y.,Yamamoto, N., Abo, T., and Kumagai, K. (1977). Identification of the target cells in human B lymphocytes for transformation by Epstein-Barr virus. Virology 83, 287-294. Kawamura. H., King, D. J., and Anderson, D. P. (1969). A herpesvirus isolated from kidney cell culture of normal turkeys. Avian Dis. 13, 853-863. Kelly, R. B., Cozzarelli, N. R., Duetscher, M. P., Lehman, 1. R., and Kornberg, A. (1970). Enzymatic synthesis of deoxyribonucleic acid. XXXII. Replciation of duplex deoxyribonucleic acid by polymerase at a single strand break. J. B i d . C h e m . 245, 39-45. Klein, G. (1973). The Epstein-Barr virus. In “The Herpesviruses” (A. S. Kaplan, ed.), pp. 521-555. Academic Press, New York. Klein, G., Giovanella, B. C., Lindahl, T., Fialkow, P. J., Singh, S., and Stehlin, J. S. (1974). Direct evidence for the presence of Epstein-Barr virus DNA and nuclear antigen in malignant epithelial cells from patients with poorly differentiated carcinoma of the nasopharynx. Proc. N a i l . Acad. Sci. U . S . A . 71, 4737-4741. Klein, G. (1975). Virus-induced antigens-a review. In “Oncogenesis and Herpesvirus 11” (G. de-The, M. A. Epstein, and H. zur Hausen, eds.), Pt. I , pp. 293-308. International Agency for Research on Cancer, Lyon. Klein, G., Svedmyr, E., Jondal, M., and Persson, P. 0.(1976). EBV-determined nuclear antigen (EBNA)-positive cells in the peripheral blood of infectious mononucleosis patients. lnt. J. Cancer 17, 21-26. Klessig, D. F. (1977). Two adenovirus mRNAs have common 5 ‘ terminal leader sequence encoded at least 10 Kb upstream from their main coding regions. Cell 12, 9-21. Kraiselbard, E., Gage, L. P.,and Weissbach, A. (1975). Presence of a herpes simplex virus DNA fragment in an L cell clone obtained after infection with irradiated herpes simplex virus 1 . J . Mol. B i d . 91, 533-542. Kumagai, K . Abo, T., Sekizawa, T., and Sasaki, M. (1975). Studies of surface immunoglobulins on human B lymphocytes. I. Dissociation of cell-bound immunoglobulins with acid pH or at 37°C. J . Inirnunol. 115, 982-987. Landon, J . C., and Malan, L. B. (1971). Seroepidemiologic studies of Epstein-Barr virus antibody in monkeys. J . N a t l . Cancer Inst. 46, 881-884. Landon, J. C., Ellis, L. B. Zeve, V. H., and Fabrizio, D. P. A. (1968). Herpes-type virus in cultured leukocytes from chimpanzees. J. Natl. Cancer Inst. 40, 181-192. Lee, L. F. (1972). Induction of deoxyribonucleic acid synthesis and the oncogenicity of Marek’s disease virus. J . Virol. 10, 167-170. Lee, L. F., Kieff. E. D., Bachenheimer, S. L., Roizman, B.. Spear, P. G., Burmester, B. R., and Nazerian, K. (1971). Size and compostion of Marek’s disease virus deoxyribonucleic acid. 1. Virol. 7 , 289-294. Lee, L. F.. Nazerian, K . , Leinbach, S. S . , Reno, J. M., and Boezi, J . A. (1976). Effect of phosphonoacetate on Marek’s disease virus replication. J. Nut/. C a n w r Inst. 56, 823827. Leinbach, S. S . , Reno, J . M., Lee, L. F., Isbell, A. F., and Boezi, J. A . (1976). Mechanism of phosphonoacetate inhibition of herpesvirus-induced DNA polymerase. Biochemistry 15, 426-430. Lemon, S. M., Hutt, L. M . , Shaw, J . E., Li, J.-L. H., and Pagano, J . S. (1977). Replication of EBV in epithelial cells during infectious mononucleosis. Nurure (Londo17)268, 268270. Lenoir. G . , Berthelon, M.-C., Favre. M.-C., and de-The, G. (1976). Characterization of Epstein-Barr virus antigens. I . Biochemical analysis of the complement-fixing soluble antigen and relationship with Epstein-Barr virus-associated nuclear antigen. J . Virol. 17, 672-674. Lindahl, T., Klein. G., Reedman, B. M., Johansson, B., and Singh. S. (1974). Relationship
274
BILL SUGDEN ET AL.
between Epstein-Barr virus (EBV) DNA and the EBV-determined nuclear antigen (EBNA) in Burkitt lymphoma biopsies and other lymphoproliferative malignancies. Int. J. Cancer 13,764-772. Lindahl, T . , Adams, A, Bjursell, G., Bornkamm. G. W., Kaschka-Dierich, C.. and Jehn, U. (1976). Covalently closed circular duplex DNA of Epstein-Barr virus in a human lymphoid cell line. J . Mol. Biol. 102, 51 1-530. Long, P. A., Kaveh-Yamini, P.. and Velicer, L. F. (197Sa). Marek’s disease herpesviruses. I . Production and preliminary characterization of Marek‘s disease herpesviruses A antigen. J. Viro/. 15, 1182-1191. Long, P. A., Clark, J. L., and Velicer, L . F. (1975b). Marek’s disease herpesviruses. 11. Purification and further characterization of Marek’s disease herpesvirus A antigen. J. viroi.15,1 1 9 2 - i m . Longenecker, B. M., Pazderka, F.. Gavora, J . S . , Spencer, J . L., and Ruth, R. F. (1976). Lymphoma induced by herpesvirus: Resistance associated with a major histocompa~ l 401-407. ~ l 7 ~ ~ ~ ~ . ~ tibility gene. / ~ l l ~ l / / l l ~ J 3, Luka. J . , Siegert, W., and Klein, G. (1977). Solubilization of the Epstein-Barr virusdetermined nuclear antigen and its characterization as a DNA-binding protein. J. Virol. 22, 1-8. Luka, J., Lindahl, T., and Klein, G. (1978). Purification of the Epstein-Barr virus-determined nuclear antigen from Epstein-Barr virus-transformed human lymphoid cell lines. J . Virol. 27, 604-6 I I. Mao, J. C.-H., Robishaw, E. E., and Overby, L . R. (1975). Inhibition of DNA polymerase from herpes simplex virus-infected Wi-38 cells by phosphonoacetic acid. J. Virol. 15, I28 1- 1283. Matsuda, H . , Ikuta. K., and Kato, S. (1976). Detection of T-cell surface determinants in three Marek‘s disease lymphoblastoid cell lines. Biken J. 19, 29-32. Matsuda, H.. Ikuta, K., Miyamoto, H.,and Kato, S. (1976). Demonstration of a Marek’s disease tumor-associated surface antigen (MATSA) on six cell lines derived from Marek’s disease lymphomas. Bikrn J. 19, 119-123. Matsuo, T., Nishi, S., Hirai. H.,and Osato, T. (1977). Studies on Epstein-Barr virusrelated antigens. 11. Biochemical properties of soluble antigen in Raji Burkitt lymphoma cells. Int. J. Cancer 19, 364-370. Melendez, L. V.. Daniel, M. D., Hunt, R. D., and Garcia, F. G. (1968). An apparently new herpesvirus from primary kidney cultures of the squirrel monkey. Lab. Aiiiin. Care 18, 374-38 I . Melendez, L. V., Hunt, R. D., King, N. W., Barahona, H.H., Daniel, M. D.. Fraser, C. E. 0..and Garcia. F. G. (1972).Herpesvirm atrles, a new lymphoma virus of monkeys. Nature (LOnd(J11)N f w Biol. 235, 182- 184. Miller, G. (1974). The oncogenicity of Epstein-Barr virus. J . Infkct. Dis. 130, 187-205. Miller, G., and Lipman, M. (1973). Release of infectious Epstein-Barr virus by transformed marmoset leukocytes. Proc. Natl. Acad. Sci. U.S.A. 70, 190-194. Miller, G., Shope. T., Coope, D., Waters, L., Pagano. J . , Bornkamm, G. W., and Henle, W. ( 1977a). Lymphoma in cotton-top marmosets after inoculation with Epstein-Barr virus: Tumor incidence, histologic spectrum antibody responses, demonstration of viral DNA, and characterization of viruses. J . Exp. Med. 145, 948-967. Miller. R. L., Glaser, R., and Rapp, F. (1977b). Studies of an Epstein-Barr virus-induced DNA polymerase. Virology 76, 494-502. Mizuno, F..Aya, T., and Osato, T . (1974). B-Lymphocytes as target cells for E.B. virus transformation. Br. Med. J. 3, 689. Mizuno, F.. Aya, T., and Osato, T . (1976). Growth in semisolid agar medium of human
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
275
cord leukocytes freshly transformed by Epstein-Barr virus. J . Nail. Cancer Inst. 56, I7 1- 173. Moore, G . E., Gerner, R. E., and Franklin, H. A. (1967). Culture of normal human leukocytes. J . Am. Med. Assoc. 199, 519-524. Moss, D. J . , and Pope, J . H. (1972). Assay of the infectivity of Epstein-Barr virus by transformation of human leucocytes in Vi/ro.J . Cen. Virol. 17, 233-236. Nazerian. K. (1973). Marek’s disease: A neoplastic disease of chickens caused by a herpesvirus. A h * . Cancer Res. 17, 279-3 15. Nazerian, K., and Lee, L. F. (1974). Deoxynbonucleic acid of Marek’s disease virus in a lymphoblastoid cell line from Marek‘s disease tumours. J . Cen. Virol. 25, 3 17-32 I . Nazerian, K., and Sharma, J . M. (19751. Detection of T-cell surface antigens in a Marek‘s disease lymphoblastoid cell line, J . Nail. Cancer Inst. 54, 277-279. Nazerian, K., and Witter, R. L. (1970). Cell-free transmission and in Vivo replication of Marek’s disease virus. J . Virol. 5, 388-397. Niederman, J. C., McCollum, R. W., Henle, G., and Henle, W. (1968). Infectious mononucleosis: Clinical manifestations in relation to EB virus antibodies. J . A m . Med. Assoc. 203, 205-209. Niederman, J . C . , Miller, G., Pearson, H. A., Pagano, J . S . , and Dowaliby, J. M. (1976). Infectious mononucleosis: Epstein-Barr- virus shedding in saliva and the oropharynx. N . Enyl. J . Med. 294, 1355-1359. Nilsson, K., Giovanella, B. C., Stehlin, J . S . , and Klein, G . (1977). Tumorigenicity of human hematopoietic cell lines in athymic nude mice. int. J . Cancer 19, 337-344. Nonoyama, M., and Pagano, J . S. (1972). Separation of Epstein-Barr virus DNA from large chromosomal DNA in non-virus-producing cells. Nature (Londun)New Biol. 238, 169- I7 I . Nonoyama, M., and Pagano, J . S. (1973). Homology between Epstein-Barr virus DNA and viral DNA from Burkitt’s lymphoma and nasopharyngeal carcinoma determined by DNA-DNA reassociation kinetics. Natirre (London)242, 44-47. Okazaki, W., Purchase, H. G . , and Burmester, B. R. (1970). Protection against Marek’s disease by vaccination with a herpes virus of turkeys. Avian Dis. 14, 413-429. Orellana, T., and Kieff, E. (1977). Epstein-Barr virus specific RNA. I I . Analysis of polyadenylated viral RNA in restringent, abortive, and productive infection. J . Virol. 22, 321-330. Otto, B.. and Fanning, E. (1978). DNA polymerase a is associated with replicating SV40 nucleoprotein complexes. Nucl. Acids Res. 5 , 1715-1728. Pattengale, P. K . , Smith, R. W., and Gerber, P. (1973). Selective transformation of B lymphocytes by E. B. virus. Lancet 2, 93-94. Payne, L. N., Powell, P. C . . and Rennie, M. (1974). Response of B and T lymphocytes and other blood leukocytes in chickens with Marek’s disease. Cold Spring Harbor Svmp. Qrtuni. Biol. 39, Pt. 2, 817-826. Pope, J . H., Horne, M. K., and Scott, W. (1968). Transformation offoetal human leukocyte in Vitro by filtrates of a human leukaemic cell line containing herpes-like virus. int. J . Cancer 3, 857-866. Powell, A. L. T., King, W., and Kieff, E. (1979). Epstein-Barr virus specific RNA. 111. Mapping of DNA encoding viral RNA in restringent infection. Submitted for publication. Pritchett, R. Hayward, S. D., and Kieff, E. (1975). DNA of Epstein-Barr virus. I . Comparison of DNA of virus purified from HR-I and B95-8 cells. J . Virol. 15, 556-569. Purchase, H. G . , Burmester, B. R . , and Cunningham, C. H. (1971). Pathogenicity and antigenicity of clones from strains of Marek’s disease virus and the herpesvirus of turkeys. infect. Immiin. 3, 295-303.
276
BILL SUGDEN ET AL.
Purtilo, D. T., Hutt, L., Bhawan, J., Yang, J . P. S., Cassel, C., Allegra, S., and Rosen, F. S. (1978). Immunodeficiency to the Epstein-Barr virus in the X-linked recessive lymphoproliferative syndrome. Clin. Immuttol. Jmtnunopathol. 9, 147- 156. Rabin, H., Neubauer. R. H., Hopkins, R. F., Dzhikidze, E. K., Shevtsova, Z. V., and Lapin, B. A. ( 19771. Transforming activity and antigenicity of an Epstein-Barr-like virus from lymphoblastoid cell lines of baboons with lymphoid disease. Intcwirology 8, 240-249. Rabin, H . , Neubauer, R. H., Hopkins, R. F., and Nonoyama, M. (1978). Further characterization of a herpesvirus-positive orang-utan cell line and comparative aspects of in Vitro transformation with lymphotropic old world primate herpesviruses. Int. J. Cunccr 21, 762-767. Rasheed, S., Rongey, R. W., Bruszweski, J., Nelson-Rees, W. A., Rabin, H., Neubauer, R. H., Esra, G., and Gardner, M. B. (1977). Establishment of a cell line with associated Epstein-Barr-like virus from a leukemic orangutan. Scirncr 198, 407-409. Reedman. B. M., and Klein, G. (1973). Cellular localization of an Epstein-Barr virus (EBV)-associated complement-fixing antigen in producer and non-producer lymphoblastoid cell lines. I n t . J . Cancer 11, 499-520. Rickinson, A. B., Jarvis, J . E., Crawford, D. H., and Epstein, M. A. (1975). Observations on the nature of Epstein-Barr virus infection of peripheral lymphoid cells in infectious mononucleosis. I n “Oncogenesis and Herpesviruses I I ” (G. de-ThC, M. A. Epstein, and H. zur Hausen, eds.), Pt. 2, pp. 169-177. International Agency for Research on Cancer, Lyon. Rickinson, A. B., Finerty, S., and Epstein, M. A. (1977). Mechanism of the establishment of Epstein-Barr virus genome-containing lymphoid cell lines from infectious mononucleosis patients: Studies with phosphonoacetate. Int. J. Cancer 20, 861-868. Robinson, J . E., Andiman, W. A., Henderson, E., and Miller, G. (1977). Host-determined differences in expression of surface marker characteristics on human and simian lymphoblastoid cell lines transformed by Epstein-Barr virus. Proc. Natl. Accid. Sci. U . S . A . 74, 749-753. Ross, L. J . N., Biggs, P. M.. and Newton, A. A. (1973). Purification and properties of the ‘A’ antigen associated with Marek’s disease virus infections. J . Gcn. Virol. 18, 291304. Rowe, D. S., Hug, K., Forni, L., and Pernis, B. (1973). Immunoglobulin D a s a lymphocyte receptor. J. Exp. Med. 138, 965-972. Rymo, L., and Forsblom. S. (1978). Cleavage of Epstein-Barr virus DNA by restriction endonucleases EcoRl, Hind111 and B a d . Nucl. Acids Res. 5 , 1387-1402. Sakamoto, K., and Hinuma, Y. (1978). Demonstration of a surface antigen on EpsteinBarr virus genome-carrying lymphoid cells: Distinction from the virus-determined membrane antigen. J . Immunol. 120, 1702-1708. Sambrook, J., Westphal, H., Srinivasin, P. R., and Dulbecco, R. (1968). The integrated state of viral DNA in SV40-transformed cells. Proc. Nut/. Acad. Sci. U.S.A. 60, 12881295. Sambrook, J., Botchan, M., Gallimore, P., Ozanne, B., Pettersson, U . , Williams, J . , and Sharp, P. A. (1974). Viral DNA sequences in cells transformed by simian virus 40, adenovirus type 2 and adenovirus type 5 . Cold Spring Harbor Symp. Quunt. Biol. 39, Pt. I , 615-632. Schulte-Holthausen, H., and zur Hausen, H. (1970). Parital purification of the EpsteinB a n virus and some properties of its DNA. Virology 40, 776779. Sheldrick, P., Laithier, M., Lando, D., and Ryhiner, M. L. (1973). Infectious DNA from herpes simplex virus: Infectivity of double-stranded and single-stranded molecules. Proc. Natl. Acad. Sci. U . S . A . 70, 3621-3625.
MOLECULAR BIOLOGY OF LYMPHOTROPIC HERPESVIRUSES
277
Shope, T., Dechairo, D., and Miller, G. (1973). Malignant lymphoma in cottontop marmosets after inoculation with Epstein-Barr virus. Proc. Natl. Acud. Sci. U . S . A . 70, 2487-249 I . Stockert, E., Old, L. J . , and Boyse, E. A. (1971). The GlXsystem../. Exp. M u d . 133, 13341355. Strauch, B., Siegal, N . , Andrews, L.-L., and Miller, G. (1974). Oropharyngeal excretion of Epstein-Barr virus by renal transplant recipients and other patients treated with immunosuppressive drugs. Luncet 1, 234-237. Sugden, B. (1977). Comparison of Epstein-Barr viral DNAs in Burkitt lymphoma biopsy cells and in cells clonally transformed in Vitro. Proc.. Nut/. Acud. Sci. U . S . A . 74, 465 1-4655. Sugden, B., and Mark, W. (1977). Clonal transformation of adult human leukocytes by Epstein-Barr virus. J. Virol. 23, 503-508. Sugden, B., Summers, W. C., and Klein, G. (1976). Nucleic acid renaturation and restriction endonuclease cleavage analyses show that the DNAs of a transforming and a nontransforming strain of Epstein-Barr virus share approximately 90% of their nucleotide sequence. J . Virol. 18, 765-775. Svedmyr, E., and Jondal, M. (1975). Cytotoxic effector cells specific for B cell lines transformed by Epstein-Barr virus are present in patients with infectious mononucleosis. Proc. Natl. Acud. Sci. U . S . A . 72, 1622-1626. Takada, K., and Osato, T. (1978). Analysis of the transformation of human leukocytes by Epstein-Barr virus: Sequential occurrence from the virus-determined nuclear antigen synthesis to blastogenesis to DNA synthesis. In press. Tanaka, A., and Nonoyama, M . (1974). Latent DNA of Epstein-Barr virus: Separation from high-molecular-weight cell DNA in a neutral glycerol gradient. Proc. Nut/. Acad. Sci. U . S . A . 71, 4658-4661. Tanaka, A., Silver, S., and Nonoyama, M . (1978). Biochemical evidence of the nonintegrated status of Marek’s disease virus DNA in virus-transformed lymphoblastoid cells of chicken. Virology 88, 19-24. Temin, H. (1971). The role of the DNA provirus in carcinogenesis by R N A tumor viruses. I n “The Biology of Oncogenic Viruses” (L. G. Silvestri, ed.), pp. 17h-187. NorthHolland Publ., Amsterdam. Thorley-Lawson, D. A., and Strominger, J . L. (1976). Transformation of human lymphocytes by Epstein-Barr virus is inhibited by phosphonoacetic acid. Nutrire (Londorr) 263, 332-334. Thorley-Lawson, D. A., Chess, L., and Strominger, J. L. (1977). Suppression of in Vitro Epstein-Barr virus infection. A new role for adult human T lymphocytes. J. Exp. Med. 146,495-508. Thorley-Lawson, D. A., and Strominger, J . (1978). Reversible inhibition by phosphonoacetic acid of Human B lymphoctye transformation by Epstein-Barr virus. Virology 86, 423-43 I . Tijan, R. (1977). The binding site on SV40 DNA for a T antigen related protein. Cell 13, 165- 179. Tooze, J. (ed.) (1973) “The Molecular Biology of Turnour Viruses.“ Cold Spring Harbor, New York. Traul, K. A., Stephens, R., Gerber, P., and Peterson, W. D. (1977). Productive EpsteinBarr viral infection of the human lymphoblastoid cell line 6410 with release of early antigen inducing and transforming virus. In/. J . Cancer 20, 247-255. Velicer, L. F., Yager, D. R., and Clark, J . L. (1978). Marek’s disease herpesviruses. 111. Purification and characterization of Marek’s disease herpesvirus B antigen. J. Virol. 27, 205-217.
278
BILL SUGDEN ET AL.
Veltri, R. W., Heyl, L. W., and Sprinkle, P. M. (1977). Epstein-Barr virus genome carrying lymphoctye subpopulations of human palatine tonsils. Proc. Soc. Exp. Biol. Med. 156, 282-286. Vitetta. E. S . , and Uhr. J. W. (1975). Immunoglobulin-receptors revisited: A model for the differentiation of bone marrow-derived lymphocytes is described. Science 189, 964969. von Biilow, V. (1971). Diagnosis and certain biological properties of the virus of Marek's disease. Am. J. Vet. Res. 32, 1275-1288. Wallen, W. C., Neubauer, R. H., Rabin, H., and Cicmanec, J. L. (1973). Nonimmune rosette formation by lymphoma and leukemia cells from herpesvirus saimiri-infected owl monkeys. J. Narl. Cancer I n s t . 51, 967-975. Werner, F.-J., Bornkamm, G. W., and Fleckenstein. B. (1977). Episomal viral DNA in a Irrrpes~~irus sainriri-transformed lymphoid cell line. J. Virol. 22, 794-803. Werner, F. J., Desrosiers, R. C., Mulder, C., Bornkamm, G. W., and Fleckenstein, B. ( 1978). Physical mapping of viral episomes in Herpesvirr4.v saimiri transformed lymphoid cells. UCLA Meeting on Persistemt Viruses, Keystone, Colorado (in press). Wising, P. J. (1942). Successful transmission of infectious mononucleosis to man by transfusion of heparinized blood'! Acta Med. Sccind. 109, 507-513. Witter. R . L. (1971). Marek's disease research-history and perspectives. Porrlt. Sci. 50, 333-342. Witter, R. L. (1972). Epidemiology of Marek's disease-A review. I n "Oncogenesis and Herpesviruses" (P. M . Biggs, G. de-The, and L. N . Payne, eds.), pp. I 11-122. International Agency for Research on Cancer, Lyon. Witter, R. L.. Solomon, J . J., and Burgoyne. G. H. (1969). Cell culture techniques for primary isolation of Marek's disease-associated herpesvirus. Airinn Dis. 13, 101- 118. Witter, R. L., Nazerian. K . , Purchase, H. G., and Burgoyne, G. H. (1970). Isolation from turkeys of a cell-associated herpesvirus antigenically related to Marek's disease virus. Ain. J. Vet. R e s . 31, 525-538. Witter, R . L.. Stephens, E. A., Sharma, J. M., and Nazerian, K. (1975). Demonstration of a tumor-associated surface antigen in Marek's disease. J . Iinmunol. 115, 177-183. Yajima. Y.. Marczynska, B., and Nonoyama, M. (1978). Transforming activity of EpsteinBarr virus obtained by superinfection of Raji cells. Proc. Nut/. Acud. Sci. U.S.A. 75, 2008-20 10. Yamamoto, N., and Hinuma, Y. (1976). Clonal transformation of human leukocytes by Epstein-Barr virus in soft agar. f l i t . J. Cancer 17, 191-196. zur Hausen. H. (1975). Oncogenic herpes viruses. Biochitn. Biophvs. Actu 417, 25-53. zur Hausen, H.. and Fresen, K.-0. (1977). Heterogeneity of Epstein-Barr virus. 11. Induction of early antigens (EA) by complementation. Virology 81, 138-143. zur Hausen, H., and Schulte-Holthausen, H. (1970). Presence of EB virus nucleic acid homology in a "virus-free" line of Burkitt tumour cells. Natrrri, (London) 227, 245248. zur Hausen, H., Schulte-Holthausen, H., Klein, G., Henle, W., Henle, G., Clifford, P., and Santesson, L. (1970). EBV DNA in biopsies of Burkitt tumors and anaplastic carcinomas of the nasopharynx. Nature (London) 228, 1056-1058. zur Hausen, H., Diehl, V., Wolf, H., Schulte-Holthausen, H., and Schneider, U. (1972). Occurrence of Epstein-Barr virus genomes in human lymphoblastoid cell lines. Natrrrc (London) N i w B i d . 237, 189- 190. zur Hausen, H.. O'Neill, F. J., Freese, U . K., and Hecker, E. (1978). Persisting oncogenic herpesvirus induced by the tumor promoter TPA. Nature (London) 272, 373-375.
ADVANCES I N CANCER RESEARCH. VOL. 30
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
Hiroshi Kobayashi Laboratory of Pathology, Cancer Institute. Hokkaido University School of Medicine, Sapporo, Japan
I. Introduction ...........................................................
279
.......... . . . . . . . . . . . . . . . . 280 A. Evidence for Acquisition of a New Antigen ............................ 280
11. Acquisition of a Virus-Specific Antigen
B. Acquisition of a Virus-Specific Antigen Resulting in the Immunological Regression of Tumor Cells ........................................... C. Definition of the Xenogenization of Tumor Cells . . . . .... 111. Increase in the Antigenicity of a Tumor-Specific Antigen . . . . . . . . . . . . . . . . . . . A. Increase in the Irnmunogenicity of Tumor Cells by Infection with a Lytic Virus . . . . . . . . . . .
280 286 287 287 290
C. Increase in th
A. Rejection of Xenogenized Tumor Cells and the Immune Response . . . . . . . B. The Mechanism Responsible for the Increase in the Antigenicity of Tumor V. Summary. . . . . . . . . . . . . . . . References . . . . . . .
................
................
290 29 I 292 292 293 295 297
I. Introduction
One approach to tumor immunotherapy is to modify tumor cells so that they become more foreign to the host. This review will discuss how investigators have attempted to make tumor cells foreign to the host by increasing their antigenicity. The review has been divided into two parts: the first part reviews attempts to induce new foreign proteins on tumor cells, and the second part reviews attempts to increase the antigenicity of existing tumor-specific or tumor-associated antigens (TSA or TAA) on tumor cells. By both of these methods tumor cells may be made more highly immunogenic or more easily immunosensitive to the immune responses of the host. 279 Copyright 0 1979 by Academic Press. Inc.
All rights of reproduction in any form reserved ISBN OI-I?-OOM30-0
280
HIROSH I KOBAYASHl
II. Acquisition of a Virus-Specific Antigen
A. EVIDENCE FOR
ACQUISITION OF A
NEWANTIGEN
Several investigators have described the acquisition of new antigens by tumor cells following infection with viruses. These virus-specific new antigens have been detected in mice and hamsters by use of the complement-dependent cytotoxicity assay (Stuck et af., 1964), transplantation resistance (Hamburg and Svet-Moldavsky, 1964; Sjogren and Hellstrom, 1965; Hosokawa et al., 1967; Eiselein and Biggs, 1970), or the immunofluorescence technique (Pasternak and Pasternak, 1967). Svet-Moldavsky and Hamburg (1964) performed experiments which showed that several virus-infected tumor cell lines were rejected by a host which had been immunized against the virus. Svet-Moldavsky labeled this phenomenon “heterogenization,” while Stuck et al. called it “antigenic conversion” and Sjogren and Hellstrom referred to it as the “acquisition of a virusspecific new antigen.” Many studies have since confirmed these first studies and these are summarized in Table I (A).
B. ACQUISITION OF A VIRUS-SPECIFIC ANTIGENRESULTINGI N THE IMMUNOLOGICALREGRESSION OF TUMOR CELLS[Table I (B)] If tumor cells are infected with a virus, they generally express the viral antigen on their cell surface. If the virus used to infect the tumor cells is foreign to the host, then the virus-infected tumor cells often undergo regression. Kobayashi et al. were successful in inducing the immunological regression of rat tumors which had been artificially infected with murine leukemia viruses, such as the Friend virus or the Gross virus (Kobayashi et al., 1969a,b, 1970; Sendo et al., 1970). Virus-infected transplanted tumors showed no tumor growth when inoculated into either a syngeneic or an autochthonous host. In autochthonous hosts, the tumors were surgically removed, artificially infected with the virus either in vivo or in vitro, and then reinoculated into the animal. The tumors occasionally showed some tumor growth, but eventually regressed [Tables I(B), 11; Figs. 1 , 2). It was observed that rat tumor metastases regressed if the tumor cells had been infected with murine leukemia virus (Kodama et al., 1974). The only exception to this rule of virus-induced regression was a transplantable leukemia and very rapidly growing carcinomas. In the carcinomas, it was felt that tumor proliferation was so rapid that it did not allow time for the effects of the immune response to
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
28 1
,,,-------t ,./'
15
20
25
DLT
30
35
Days after transplantation
FIG. I . Growth curve of Friend virus-infected and noninfected lung tumor (DLTI cells in Donryu rats.
occur, whereas with the leukemia there was initial regression of the injected cells but a subsequent relapse. If the host was first made tolerant through the neonatal injection of Friend or Gross virus, then the virus-infected tumor grew and eventually killed the tolerant host. The virus-infected tumor also grew progressively in a host which had been immunosuppressed by chemicals or irradiation. The explanation for regression of rat tumors infected with murine leukemia virus was that the tumors had acquired a highly antigenic virusspecific antigen (VSA or VAA). Since this antigen was recognized as foreign to the host, a strong immune response occurred against it, resulting in the death of the virus-infected tumor cells. The author called this phenomenon the "xenogenization" of tumor cells (Kobayashi et al., 1969a,b). The observation that tumors regress after they have been artificially infected with a virus has not been limited to the rat [Table I(B)]. Holtermann and Majde (1971) reported that the lethal growth of an adenocarcinoma in SWWJ mice was decreased 50% if the tumor cells were first infected with the LCM virus. Barbieri et al. (1971) reported that the lethal growth of mouse tumors was decreased by 46%-77% when the tumor cells were infected with Rauscher leukemia virus. Greenberger and Aaronson (1973) reported the regression of tumors which had originally been induced by Moloney sarcoma virus (MSV) (nonproducer) and also reported the regression of tumors which had been infected in tissue culture with the C-type helper virus. Similar results have been observed
ACQUISITION OF
A
TABLE I VIRUS-SPECIFIC NEW ANTIGENI N TUMOR CELLS
Animal
N
w
Species
Strain
(A) Inhibition of tumor growth after previous immunization: Mouse Stuck et a / . (1964)
Hamburg and Svet-Moldavsky (1964)
Hamster and mouse
Sjogren and Hellstrom
Mouse
Virus
Tumor
EL4
Rauscher
Herpes, polyoma, SV,, adeno Lymphoma
Polyoma
OMT-6
Friend virus
(1965)
Hosokawa ct al. (1967) Pasternak and Pasternak (1967) Eiselein and Biggs ( 1970)
Mouse Mouse Mouse
ddiOm
Note Antigenic conversion (cytotoxicity test) Heterogenization (transplantation) Acquisition of a virus-specific new antigen Fluorescence
Ehrlich
LCM
Regression of tumors without previous immunization: Kobayashi ef a / . (1969) Rat WKA, Donryu
KMT-68, WST-5 (sarc.), Takeda (sarc.), DLT (sq. carc.), AH-I09 MCA primary tumor Adenocarcinoma (spont.) P4bic (sarc.), TBLC2 ( MCA-sarc.) SSI (sarc.) G M*
Friend
Sendo et a / . (1970) Holtermann and Majde (1971) Barbieri ef af. (1971)
Rat Mouse
WKA SWWJ
Mouse
C57BL
Basombrio (1972) Yamada and Hatano ( 1972) Greenberger and Aaronson (1973) Kuzumaki and Kobayashi ( 1976)
Mouse Hamster
BALBIc
Mouse
NIH, Swiss BALB/c
MSV tumor
Mouse
C3HlHe
Kuzumaki et a / . (1978)
Rat
BDIX
Takeichi et a / . (1978) Takeyama ef al. (1978) Ishimoto (1978) Reed (1978) Kodama (1978)
Mouse Mouse Hamster Mouse Rat
Wi starIFu BALBIc BDF, nude WKA
MH134 (hepatoma), RaLV, MSV, Human Meth A (MCA-sarc.) measles virus, LLFV MBDB (fibrosarc.), 290-T (neurogenic) Endogenous mouse PW 41 (sarc.) E4 (sarc.) MSV L1210 (lymphoma) HVJ-pi THK Friend BHK, HeLa Mumps DEAE-D + Friend KMT-17
Xenogenization (transplantation, cytotoxicity test, etc.)
LCM Rauscher MSV HVJ, rubella RNA C-type
Regression of growing tumor
284
HIROSHI KOBAYASHI
FIG. 2. (a) Friend virus-infected KMT-17 rat tumor tissue 4 days after inoculation showing an ordinary pattern of KMT-17 tumor cells ( ~ 1 5 0 ) . (b) 7 days after inoculation. Note degenerative changes of tumor cells and infiltration of fibroblasts accompanied by lymphoid cells ( X 100).(c) 10 days after inoculation. Tumor cells have completely disappeared ( x 150).
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
285
TABLE I1 LATERAL GROWTHOF VARIOUSRAT TUMORS ARTIFICIALLY INFECTED W I T H FRIEND VIRUS I N SYNGENEIC RATS Tumors infected with virus Spontaneous sarcoma WST-5 Takeda KST- I MCA sarcoma KMT-17 KMT-I9 KMT-50 K MT-68 AMC-60 4NQ0 lung cancer DLT N B U breast cancer KBT- I KBT-2 DAB hepatoma AH I09 KDH-8 NBU leukemia KNL- 1 KNL-2 Total
Lethal growth (96) 0126 6123 0125 01443 0180 0/17
0/24 0110
0128
0112 018
218 0158
(25.0) (0)
0115
014 8/781
Tumors recurred 2-3 months after the regression in all the cases.
by Basombrio (1972), Kuzumaki and Kobayashi (1976), and Takeichi ef al. (1978a) by infecting several mouse tumors with MSV. Ishimoto (personal communication) demonstrated that the hamster tumor, THK, regressed after infection of the tumor with Friend virus. The xenogenization phenomenon, therefore, has been observed not only in rats but also in mice and hamsters. One could generalize and say that the immunological regression of virus-infected tumor cells might be observed in many species. Two conditions must be met if the virus-infected tumor cells are to undergo regression. The first is that the virus must be a nonlytic surfacebudding virus, and the second is that the host must be capable of mounting an immune response against the virus-associated new antigen. It would be desirable to use a nononcogenic virus to produce the immunological regression of tumors. There have been no reports of neoplasms produced by the murine oncornavirus in adult xenogeneic o r allogeneic hosts, but it must still be considered theoretically possible that
286
HI ROSH I KOBAY ASH1
oncornavirus could produce neoplasms in the normal adult host. A few recent reports have described the regression of tumor cells after infection with a nononcogenic virus. Holtermann and Majde (1971) have reported that infection with the LCM virus caused regression of tumors, but that the virus occasionally caused lethal inflammatory changes in the brain of the host. Yamada and Hatano (1972) used nononcogenic viruses in the hamster model system. They showed a decrease in the lethal growth of tumors following infection with the HVJ virus and the regression of tumors if they were infected with the rubella virus. Takeyama et ul. (1978) infected L1210 leukemia cells with HVJ-pi virus and showed that tumors regressed in about one-half of inoculated BDF, mice. Kuzumaki et al. (1978) observed that rat tumor cells regressed when artificially infected with an endogenous mouse virus obtained from nude mice. M. L. Reed (personal communication) infected HeLa cells with the mumps virus and transplanted them into nude mice. The mumps virus-infected viable cells grew only if the nude mice had been X-irradiated. The possible use of xenogenization of human tumor cells has great potential clinical application for cancer treatment. Murine leukemia virus has been shown to be infectious for human neoplastic cells, This was done using the ferritin conjugate technique (Wright and Korol, 1969: Kodama rt ul., 1970). It is interesting to speculate whether it would be possible to xenogenize or virally infect an existing tumor, so that regression would occur without its surgical removal. For this to succeed, the virus would have to infect the growing tumor cells rapidly and efficiently since it is expected that the virus would be quickly eliminated by the host immune response. Kodama ( 1978) has observed that the intralesional inoculation of DEAED before direct infection with a virus produces a higher rate of regression of primary tumors. It is also interesting to speculate whether certain chemicals might in some way modify tumor antigenicity. If this were possible, then such chemicals could extend the concept of xenogenization. N o report has described the regression of inoculated tumors first modified in vitro by exposure to chemicals. Even if a foreign antigen were produced on a tumor cell following exposure to a chemical, the antigen would most likely be eliminated after a number of tumor cell divisions.
c. DEFINITION OF THE XENOGENIZATION OF TUMOR CELLS The word "xenogenization" was used in 1969 to define the immunological regression of tumors after infection of tumor cells with a nonlytic budding virus (Kobayashi et ul., 1969b, 1970, 1977: Sendo et al., 1970).
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
287
Xenogenization more recently has been widely used to designate an acquisition of foreign antigen regardless of the regression of tumors, as well as to designate an increase in the antigenicity of existing tumorspecific antigens as is described in Section 111. The term xenogenization, therefore, has become a general term used to describe any foreignness that tumor cells have acquired due to either a n acquisition of a new foreign antigen or to an increase in the antigenicity of the existing tumorspecific antigen. Xenogenization in its broadest sense may be very similar to SvetMoldavsky’s term heterogenization (Svet-Moldavsky and Hamburg, 1964, 1967). The main aim of the heterogenization experiments was to produce a new antigen, whereas xenogenization has attempted to increase the antigenicity of the already existing tumor-specific antigens in addition to the acquisition of a new antigen. Both processes, either the production of a new antigen or the increase in the antigenicity of weak tumor-specific antigens, should facilitate the recognition of the tumor as foreign to the host. The term heterogenization could be applied to the regression of tumors secondary to inoculation with BCG and other microorganisms. In this case, the growth of tumor cells may be interfered with by a reaction to the microorganisms which occurs in the stroma of the tumor tissues (Svet-Moldavsky, 1978). The definition of xenogenization may then be limited to a phenomenon that occurs on the surface of the tumor cells, and does not include the suppression of tumor growth caused by the inoculation and mixture of microorganisms.
Ill. Increase in the Antigenicity of a Tumor-Specific Antigen
A. INCREASE I N THE IMMUNOGENICITY OF TUMOR CELLSBY INFECTION WITH A LYTICVIRUS[Table III(A)] Lindenmann and Klein (1967), and Lindenmann (1974) showed that the immunogenicity of tumor cells increased after virus infection. They immunized mice with either a lysate obtained from influenza-infected Ehrlich tumor cells (oncolysate) or a lysate from noninfected Ehrlich tumor cells. They observed an increased level of resistance to rechallenge in the animals immunized with the oncolysate. This was confirmed by Eaton et al. (1967, 1973) using Newcastle virus-infected lymphoma cells in mice. Boone et al. (1972a,b, 1974) infected SV40 transformed BALB/3T3 cells (E4 cells) with influenza virus and demonstrated a higher immunogenicity of homogenates from virus-infected cells as compared to homogenates from noninfected cells.
INCREASE IN
THE
TABLE 111 ANTIGENICITY OF TUMOR-SPECIFIC ANTIGENBY INFECTIONWITH
A
VIRUS
Animal Species N W 00
Strain
( A ) Increase in the immunogenicity of tumor cells by infection with a lytic virus Lindenmann and Klein Mouse A2G (1%7)
Tumor
Ehrlich
Eaton et a/. ( 1967)
Mouse
C3H
Lymphoma
Boone and Blackman (1972) Beverley et a / . ( 1973) Takeichi et a / . ( 1978)
Mouse
BALB/c
SV3T3-T4
Mouse Mouse
CBA. BALBIc BALB/c
337 E4
Materials and methods
Note
Influenza A virus (oncoly- Immunization with sate) oncolysate or crude membrane Newcastle disease virus (oncolysate) Influenza A virus (homogenate) Newcastle disease virus MSV, influenza (crude membrane)
(B) Increase in the immunogenicity of tumor cells by infection with a nonlytic budding virus: Kobayashi et a / . (1970) Rat WKA Donryu
WST, KMT DLT
Friend virus
Immunization with viable xenogenized tumor cells
Takeichi rt a / . (1978) Kuzumaki ef al. (1978) Takeyama ef a/. ( 1978) Katoh et a / . ( 1978)
Mouse Rat Mouse Rat
BALBIc BDIX BDF, WKA
E4 MBDB, 290-T L1210 KMT-17
MSV Endogenous mouse virus HVJ-pi Friend virus (inactivated by MMC, glutaraldehyde, and formalin)
Increase in the immunosensitivity of tumor cells by infection with a nonlytic budding virus: Rat Donryu Shirai r f a / . (1971)
DLT
Friend virus
Gotohda ct a/. ( 1978)
Rat
WKA
KMT-17
Moriuchi r f ctl. ( 1978)
Rat
WKA
KMT-I7
Hosokawa ef a / . (1978)
Rat
WKA
KMT-17
Takeichi rt a / . (1978)
Mouse
BALBlc
E4
Friend virus (lateral fluidity) Friend virus (antigenic modulation of TSA) Friend virus (51Cr release assay) MSV
Complement-dependent cytotoxicity test
Cell-mediated immunity
290
HlROSHl KOBAYASHI
The difference in the immunogenicity of oncolysates as compared to tumor lysates has been quite small. Most likely this has been due to the fact that homogenates of tumor cells frequently result in a destruction of the membrane-bound tumor-specific antigen with a major loss in the immunogenicity of such preparations. There have been several reports on the clinical application of oncolysates in combination with chemotherapy (Sauter et ul., 1972: Murray er al., 1977: Sinkovics et al., 1977: Wallack and Steplewski, 1977), but results do not as yet seem satisfactory.
B.
INCREASE I N THE IMMUNOGENICITYOF TUMORCELLSBY INFECTION W I T H A NONLYTIC BUDDING VIRUS[Table III(B)]
Kobayashi et a / . (1969a) have studied the immunogenicity of virusinfected tumor cells. The immunogenicity of virus-infected and nonvirusinfected cells was compared after the tumor cells had been killed by exposure to various chemicals. Katoh rt al. (1978) used mitomycin C , glutaraldehyde, and formalin to treat tumor cells and showed that there was a slightly higher immunogenicity with virus-infected tumor cells. Kuzumaki et a / . (1978) treated virus-infected rat tumor cells with irradiation and showed that these cells were more antigenic than nonvirusinfected tumor cells. A characteristic of xenogenized tumor cells defined by Kobayashi et al. is that viable tumor cells can be used to immunize the host as the tumor cells grow and regress. Following the process of regression of tumor cells, a strong specific immunity is produced. LTD,, when immunized with viable xenogenized tumor cells was approximately 1000 times higher than when not immunized. LTD,, when immunized with viable xenogenized tumor cells is further enhanced from 1000 to 1800 times by the previous administration of cyclophosphamide (Kobayashi et al., unpublished data). Attempts of tumor immunotherapy with viable xenogenized tumor cells are described in Section I I 1 , D .
c . INCREASE IN THE IMMUNOSENSITIVITY OF TUMOR CELLS BY BUDDINGVIRUS[Table III(C)] INFECTION WITH A NONLYTIC
Many studies have been performed to attempt to increase the immunogenicity of tumor cells. There have been a few reports that have described the immunosensitivity of tumor cells infected with a nonlytic virus. Shirai r t a / . (1971) first reported an increase in the immunosensitivity of tumor cells infected with a nonlytic virus. They reported that
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
29 I
nonlytic budding virus-infected tumor cells had a higher sensitivity to antitumor immune cells than noninfected tumor cells using the complement-dependent cytotoxicity assay. Gotohda er al. (1978) and Moriuchi et al. (1978) investigated the lateral mobility of tumor-surface antigens as it is related to the immunosensitivity of tumor cells. Virus-infected tumor cells have been reported to have a higher immunosensitivity to antitumor lymphocytes. Hosokawa er al. ( 1978) demonstrated that Friend virus- or Gross virus-infected rat tumor cells had an increased level of cytotoxicity compared to noninfected tumor cells when incubated with T lymphocytes obtained from tumor-bearing rats. Takeichi er al. (1978b) made similar observations using MSV-infected E4 tumor cells. It can be suggested that virus-infected tumor cells may be a valuable aid in the detection of weaker tumor antigens by using either the cellular or humoral assay system.
D. APPLICABILITY OF VIABLE XENOGENIZED TUMORCELLSI N THE TREATMENT OF CANCER Xenogenization as defined by Kobayashi et al. includes two phenomena; the first is that tumor cells infected with a nonlytic budding virus undergo regression which is immunologically mediated, and the second is that virus-infected cells when compared to noninfected cells are capable of inducing an increased antitumor immune response. Xenogenized tumor cells, because they are good immunogens, may be useful in im-
FIG.3. Ferritin labeling on the membrane surface of a Friend virus-infected KMT-17 cell treated with an antivirus-specific antigen (VSA) or VAA serum absorbed with purified plasma viruses. Mature type-C virus particles are not labeled (x50,OoO).
292
HlROSHl KOBAYASHI
munotherapy. Xenogenized tumor cells have been used in animal immunotherapy experiments (Kaji et a / . , 1969; Kobayashi et al., 1970) but with limited effects. Immunization had to be started in the earliest stage of tumor growth when there was a minimal tumor burden present. Studies in a rat tumor model showed that surgical removal of a tumor followed by immunotherapy with viable xenogenized tumor cells resulted in a higher survival rate when compared to animals having had only surgical or imrnunotherapy treatment (Kobayashi et al., 1975). A combination of chemotherapy and immunotherapy with viable xenogenized tumor cells was effective in causing the tumor growth inhibition of a slow growing tumor (Hosokawa et ul., 1971). When a rapidly growing tumor was studied, most rats died of tumors when immunotherapy or chemotherapy alone was performed, while the combination of these treatments resulted in a definite increase in the antitumor effect. Gotohda et al. (1974) performed studies with a rapidly growing rat tumor, and showed that active immunization with xenogenized tumor cells and the passive transfer of lymphocytes obtained from rats immunized with xenogenized tumor cells used either with or without chemotherapy resulted in an increased survival rate. Xenogenized tumor cells could be used for human immunotherapy experiments. It would be desirable to use viable xenogenized tumor cells as they are a more efffective immunogen than killed xenogenized tumor cells. There is, however, the remote possibility that viable xenogenized tumor cells might grow in immunologically compromised tumor patients, and therefore, much safer irradiated xenogenized tumor cells should first be used to immunize such patients. IV. Immune Responses against Xenogenized Tumor Cells
A. REJECTION OF
XENOGENIZED
TUMOR CELLS
A N D THE IMMUNE
RESPONSE The rejection mechanism responsible for the death of viable xenogenized tumor cells has not been completely investigated. Host reaction to the foreign virus-specific antigen (VSA) present on the surface of the tumor cells may be the most likely explanation as to why xenogenized tumor cells are killed. Although the host may become immunized against the weak tumor-specific or tumor-associated antigens, it is unclear how much this response contributes to the rejection of the tumor. Kobayashi et al. have demonstrated that xenogenized tumor cells can be inhibited by the humoral immune response. Xenogenized tumor cells were placed in a three-layer diffusion chamber and then inserted into the abdominal
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
293
cavity of rats. As the animals became immunized, the number of xenogenized tumor cells in the chamber gradually decreased and eventually all of the tumor cells disappeared. Repeated transfer of immune serum also retarded the growth of xenogenized tumor cells grown in immunologically tolerant rats (Itaya et af.,unpublished data). These experiments indicated that the humoral immune response was responsible for the rejection of xenogenized tumor cells. However, the cellular immune response may also play a very important role in rejection of xenogenized tumor cells. Hosokawa et al. (1978) showed with in vitro studies that xenogenized tumor cells were rejected by lymphocytes from the tumor-bearing host. Experiments by Hosokawa et al. showed that the growth of xenogenized tumor cells was inhibited in irradiated rats which had received an intravenous transfer with normal spleen cells, but not in animals transferred with normal thymus o r bone marrow. However, it has not yet been determined what types o r subsets of immune-competent cells play a major role in inhibiting xenogenized tumor cells. It is of interest to note that repeated immunization with viable xenogenized tumor cells has not produced as strong a level of immunity as has the surgical excision of a tumor mass with subsequent repeated immunization with viable nonxenogenized tumor cells. From this data, one can conclude that there are many unknown factors that can influence the development of tumor immunity following immunization with viable xenogenized tumor cells. Some preliminary studies by Kobayashi et al. have shown that suppressor cells may be produced during the regression of viable xenogenized tumor cells and this may be one explanation for the above phenomenon. Further studies to elucidate what might be the most effective means of immunization with xenogenized tumor cells and what mechanisms are responsible for such immunization are necessary.
B. THEMECHANISMRESPONSIBLE FOR THE INCREASE IN THE ANTIGENICITY OF TUMOR CELLSAFTERINFECTIONWITH A NONLYTIC BUDDING VIRUS The ability to produce immunity against VSA is necessary to produce regression of xenogenized tumor cells and to indirectly develop immunity against the TSA in the normal host. Rats immunized with xenogenized tumor cells develop both TSA and VSA immunity, and indeed the level of tumor-specific immunity has been high in animals immunized with xenogenized tumor cells. The mechanism responsible for the increased tumor-specific immunogenicity of virus-infected tumor cells is not known
294
HlROSHl KOBAYASHI
(Boone et al., 1974, 1978: Lindenmann, 1978; Bromberg et af., 1978: Hamaoka et af., 1978; Lachmann and Sikora, 1978). There are several possible explanations for this phenomenon. The first is that there is a quantitative and qualitative change of the tumor-specific antigen. Absorption experiments performed using tumor-specific antiserum have not been able to detect a quantitative change in the amount of tumor-specific antigen present on xenogenized or nonxenogenized tumor cells. The possibility that there is a qualitative change of the tumor-specific antigen has not been studied. The second possibility is that there is an adjuvant effect of xenogenized tumor cells which results in an increase in the immune response. There are reports that have described the adjuvant activity of viruses, but none of these experiments has shown an increased transplantation resistance following viral xenogenization (Kobayashi ef al., unpublished data). The third possibility is the "hapten carrier effects" of the tumor-specific antigen conjugated with a carrier antigen (Hamaoka et al., 1978; Lachmann and Sikora, 1978). The virus-specific new antigen present on xenogenized tumor cells may stimulate the immune response against the haptenic tumor-specific antigen. There are several reports that have described the hapten carrier effect, but this hypothesis has not always been considered (Beverley ef al., 1973: Evermann and Burnstein, 1973, particularly in the viral xenogenization experiment (Kobayashi et d . , unpublished data). Associative recognition of virus-specific and tumor-specific antigens might be in some way responsible for the stronger immunogenicity produced by xenogenized tumor cells (Boone et d . , 1978: Lindenmann, 1978; Bromberg ef af., 1978). In other words, virus-specific antigen may play a role as a helper determinant. The final possible explanation as to why xenogenized tumor cells are more antigenic, particularly more immunosensitive, may be that changes in the mobility of the tumor cell surface occur in xenogenized tumor cells. Tumor-specific antigens may be able to move more easily and may actually cluster if the tumor cells have first been infected with a virus. That clustering of tumor-specific antigen occurs on xenogenized tumor cells has been shown by the complement-dependent cytotoxicity assay, using monospecific syngeneic antiserum to the tumor-specific antigen (Moriuchi et al., 1978). Immunoelectron microscopic examination of xenogenized tumor cells using ferritin-labeled tumor-specific antibody also has shown a clustering of tumor antigenic determinants on xenogenized tumor cells as compared to nonvirus-infected tumor cells (Figs. 4, 5 ) . It is provable that the clustering of the tumor antigens may be responsible for the increase in the immunosensitivity of tumor-specific antigens on xenogenized tumor cells to antitumor lymphocytes and tumor-specific
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
295
antiserum. Further studies are needed to define more precisely why there is an increased cell-mediated cytotoxicity of virus-infected tumor cells.
V. Summary
Tumor cells have been shown to be of low antigenicity and as such are not readily recognized as foreign by the host. This review has summarized studies that have explored how the antigenicity of tumor cells can be
FIG.4. (a) TSA on the surface of a KMT-17 cell distributed sparsely ( ~ 6 0 , 0 0 0 )(b) . TSA (arrow) on the surface of a Friend virus-infected KMT-17 cell easily clustered ( ~ 6 0 , 0 0 0 ) .
296
HIROSHI KOBAYASHI
tumor cell infected with Friend virus
tumor cell
FIG.5 . Clustering of TSA with antibody in Friend virus-infected KMT-17 tumor cells.
increased by biological modification. By the process of xenogenization of tumor cells infected with a nonlytic budding murine virus (Friend or Gross virus), the author has been able to cause the regression of tumors when inoculated into normal hosts owing to an acquisition of virus-specific antigen that is foreign in the host. Concurrently, the inoculated animals developed a level of tumor-specific immunity that was higher than that observed when immunization was done with X-irradiated nonxenogenized tumor cells. Both in vitro and in vivo studies provided additional evidence that the tumor-specific antigenicity of viral infected tumor cells was increased (Table IV). It is hoped that in the future such modified tumor cells will prove to be more effective in both immunotherapy and in the immunodiagnosis of tumors. The primary focus of this review has been to summarize studies performed on tumor cells that
TABLE IV OF XENOGENIZED TUMOR CELLS ANTIGENICITY Ordinary tumor cells Acquisition of a new antigen (VAN Regression (primary, metastasis) lmmunogenicity (TAA) with viable cells with inactivated cells Immunosensitivity (TAA) against humoral antibody against lymphocytes
Xenogenized tumor cells
None
Yes
None
Yes
Difficult to perform Low
Very strong Slightly increased
Low Low
Increased Increased
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
297
have been infected with a nonlytic budding virus (xenogenized tumor cells). Xenogenization was the general term that was used to describe the process of making a tumor cell antigenically foreign (Greek, x e n m foreign) to the host.
REFERENCES Barbieri, D., Belehradek, .. and Barski, G. (1971). t . J . Cancer 7 , 4-37 Basombrio, M. A. (1972). Proc. Am. Assoc. Cancer Res. 13, 74. Beverley, P. C. L., Lowenthal, R. M., and Tyrrell, D. A. J. (1973). Inr. J . Cancer 11,212223. Boone, C. W.. and Blackman, K. (1972a). Cancer Res. 32, 1018-1022. Boone, C. W., and Blackman, K. (1972b). Nail. Cancer Ins/. Monogr. 35, 301-307. Boone, C. W., Paranjpe, M., Orme, T., and Gillette, R. (1974). In?. J. Cuncer 13,543-551. Boone, C . W., Takeichi, N., Austin, F. C., Gotohda, E., Oikawa, T., and Gillitte, R. ( 1978). Gann Monogr. Cancer Res. 23 (in press). Bromberg, J . , Brenan, M., Clark, E. A., Lake, P., Mitchison, N. A., Nakashima, I., and Sainis. K. B. (1978). Gann Monogr. Cancer Res. 23 (in press). Eaton, M. D., Levinthal, J. D., and Scala, A. R. (1967). J . Natl. Cancer I n s t . 39, 10891097. Eaton, M. D., Heller, J . A., and Scala, A. R. (1973). Cancer Res. 33, 3293-3298. Eiselein, J., and Biggs, M . W. (1970). Cancer Res. 30, 1953-1957. Evermann, J. F., and Burnstein, T. (1975). Int. J. Cancer 16, 861-869. Gotohda, E., Sendo, F., Hosokawa, M., Kodarna, T., and Kobayashi, H. (1974). Cuncer Res. 34, 1947-1951. Gotohda, E., Sendo, F., Nakayarna, M., Hosokawa, M., Kawarnura, T., Kodama, T., and Kobayashi, H. (1975). J. Nut/. Cancer Inst. 55, 1079-1083. Gotohda, E., Moriuchi, T., Kawamura, T., Akiyama, J., Oikawa, T., Sendo, F., Hosokawa, M., Kodama, T., and Kobayashi, H. (1978) (accepted by J. Nut/. Cancer I n s t . ) . Greenberger, J. S., and Aaronson, S. A. (1973). J. N u / / . Cancer Inst. 51, 1935-1938. Harnaoka, T., Fujiwara, H., Tsuchida, T., Kinouchi, T., and Aoki. H . (1978). Gunti Monogr. Cancer Rvs. 23 (in press). Hamburg, V. P.. and Svet-Moldavsky, G. J . (1964). Natirre (London) 203, 772-773. Holterrnann, 0. A., and Majde, J. A. (1971). Transplantation 11, 20-29. Hosokawa, M., Kodama, T., Sendo, F., Takeichi, N., and Kobayashi, H . ( 1967).J. Cancer Immirnopathol. 3, 42-46. Hosokawa, M., Sendo, F., Gotohda, E., and Kobayashi, H. (1971). Cann 62, 57-60. Hosokawa, M., Kasai, M., Yamaguchi, H., and Kobayashi, H. (1978). Gunn Monogr. Cancer Res. 23 (in press). Ishimoto, A. Personal communication. Kaji, H., Sendo, F., Shirdi, T., Saito, H., Kodama, T., and Kobayashi, H. (1969). Mod. Med. 24, 1329-1333. Katoh, H., Ikeda, K., Minarni, A., Hosokawa, M., Kodama, T., and Kobayashi. H. (1978). ( i n preparation). Klein. G., and Klein, E. (1977). Proc. Narl. Acud. Sci. U.S.A. 74, 2121-2125. Kobayashi, H. (1970). In "Immunity and Tolerance in Oncogenesis" (L. Severi. ed.), pp. 637-659. 4th Perugia Quadr. Int. Conf. on Cancer, 1969, Perugia.
298
HIROSHI KOBAYASHI
Kobayashi, H., Sendo, F., Shirai, T..Kaji, H., Kodama, T., and Saito, H. (1969a).J . Nail. Cancer Inst. 42, 413-419. Kobayashi, H . , Kodama. T., Shirai, T., Kaji, H., Hosokawa, M., Sendo, F., Saito, H., and Takeichi, N . (1969b). Hohhaido J . Med. Sci. 44, 133-134. Kobayashi, H., Sendo, F., Kaji, H., Shirai, T., Saito, H., Takeichi, N . , Hosokawa, M., and Kodama, T. (1970). J . Nut/. Cancer 1 m t . 44, 11-19, Kobayashi, H., Gotohda, E.. Hosokawa, M., and Kodama, T . ( 1975). J . Nail. Cuncer I n s t . 54, 997-999.
Kobayashi, H., Kodama, T., and Gotohda, E. (1977). 111 "Xenogenization of Tumor Cells," Vol. 9. pp. 1-124, Hokkaido University Med. Libr. Ser. Hokkaido University School of Medicine, Sappporo. Kodama, T. (1978). Gunn Monogr. Cancer Rc,s. 23 (in press). Kodama, T., Kobayashi, H., Saito, H., Shirai, T., and Matsumiya, H. (1970). Gunn 61, 2 19-22 I . Kodama, T., Gotohda, E., Takeichi. N . , Kuzumaki, N . , and Kobayashi, H. (1974). J. Natl. Cancer I m t . 52, 931-939. Kodama, T., Katoh, H., Gotohda, E., Kobayashi, H., and Sendo, F. (1978). J . Nut/. Cancer I n s t . (brief communication; in press). Kuzumaki, N., and Kobayashi, H. (1976). Transp/antation 22, 545-550. Kuzumaki, N . , Fenyo, E. M., Giovanella. B. C., and Klein, G. (1978). 1nt. J . Cuncc,r 21, 62-66.
Lachmann, P. J., and Sikora, K. (1978). Native (London) 271, 463-464. Lindenmann, J. (1974). Biochim. Biopliys. Acra 335, 49-75. Lindenmann. J. (1978). Gann Monogr. Cancer R e s . 23 (in press). Lindenmann, J., and Klein, P. A. (1967). J. Exp. Med. 126,93-108. Moriuchi, T., Gotohda. E., Hosokawa, M., Kodama, T., and Kobayashi, H. (1978). J . Null. Cancer I n s t . 62, 579-583. Murray, D. R., Cassel, W. A., Torbin, A. H., Olkowski, Z. L., and Morre, M. E. (1977). Cancer 40, 680-686. Pasternak, G., and Pasternak, L. (1967). J. Narl. Cancer Inst. 38, 157-168. Reed, M. L. Personal communication. Sauter, C., Gerber, A., Lindenmann, J., and Martz, G. (1972). Schweiz. Med. Wochensclir. 102, 285-290.
Sendo. F., Kaji, H., Saito, H.,and Kobayashi, H. (1970). Gann 61, 223-226. Shirai, T., Kaji, H., Takeichi, N.. Sendo, F., Saito, H., Hosokawa, M.. and Kobayashi, H. (1971). J . Nut/. Cancer Inst. 46, 449-460. Sinkovics, J. G.. Plager, C.. McMurtrey, M. J., Romero, J. J., and Romsdahl, M. M. (1977). Proc. A m . Assoc. Cancer Rcs. 18, 86. Sjogren, H. O., and Hellstrom, I. (1965). Exp. Cell Res. 40, 208-212. Stuck, B., Old, L. J., and Boyse, E. A. (1964). Nature (London) 202, 1016-1018. Svet-Moldavsky, G. J. (1978). Gann Monogr. Cancer Res. 23 (in press). Svet-Moldavsky, G. J., and Hamburg, V. P. (1964). Nature (London) 202, 303-304. Svet-Moldavsky, G. J., and Hamburg, V. P. (1967). UICC Monogr. 2 (Specific Tumor Antigens), 323-327. Takeichi, N . , Boone, C. W., Holden, H. T., and Herberman, R. B. (1978a). Int. J . Cancer 21, 78-84.
Takeichi, N . , Austin, F. C . , Oikawa, T., and Boone, C. W. (1978b). Cancer Res 38,45804584.
VIRAL XENOGENIZATION OF INTACT TUMOR CELLS
299
Takeyama, H . , Kawashima, K . , Yamada, K . , and Ito, Y . (1978). (submitted for publication). Wallack, M. K . , and Steplewski, Z. (1977). Proc. A m . Assoc. Cancer Rrs. I%, 18. Wright, B. S., and Korol, W . (1969). Cancer Rrs. 29, 18861888. Yamada, T., and Hatano, M . (1972). Gann 63, 647-655.
This Page Intentionally Left Blank
ADVANCES IN CANCER RESEARCH, VOL. 30
VIRUS AUGMENTATION OF THE ANTIGENICITY OF TUMOR CELL EXTRACTS
Faye C. Austin a n d Charles W . Boone Cell Biology Section, Laboratory of Viral Carcinogenesls, National Cancer Institute. National Institutes of Health, Bethesda. Maryland
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Virus Therapy of C a n c e r . . . . . . . . . . . . . A. Virus Therapy in Animals.. .......................................... B. Virus Therapy in Human Cancer Patients. 111. Augmented Immunogenicity of Virus-Infected .......... A. Animal Model Studies ............ B. Immunotherapy of Human Cancer: s ...................... C. Immunodiagnosis of Human Cancer with Virus-Augmented Skin Test Antigens: Clinical Trials ................................................ IV. Mechanisms of Virus Augmentation of TATA Activity . . . . A. Relationship of Viral Antigens to Host Cell Antigens . . B. Augmented Induction of the Primary Antitumor Immune Response C. Virus-Augmented Delayed Hypersensitivity Skin Tests . . . . . . . . . . . . . . . . . . V. Prospects for the Application of Virus-Augmented Tumor Antigens in Immunodiagnosis and Immunotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Immunodiagnosis ...... B. Immunotherapy ................................................. VI. Summary.. . . . . . . . ..................................... References ............................................................
30 I 303 303 305 308 308 320 327 329 329 334 338 338 338 339 339 340
I. Introduction
It has been known for many decades that certain ascites tumors of mice can be made to disappear completely by inoculating them with a cytolytic virus such as influenza virus. More recently, it was noted that mice which survived such “viral oncolysis” therapy were remarkably immune to an additional challenge with viable tumor cells. Given the facts of viral oncolysis and “postoncolytic immunity,” Lindenmann and Klein (1967a) took the next logical step and showed that normal mice could be immunized against tumor graft challenge by inoculating them with influenza virus oncolysate (residual membranous debris remaining in the peritoneal cavity after destruction of ascites tumor by influenza virus) but not by inoculating them with the membranous debris of mechanically disrupted tumor cells. Lindenmann later (1974, 1977) applied 301 Copyright @ 1979 by Academic Press. Inc. All rights of reproduction in any form reserved ISBN 0-12-006630-0
302
FAYE C. AUSTIN A N D CHARLES W . BOONE
the term “viral adjuvanticity” to this effect to emphasize that although immunization is performed by inoculating a combination of “tumor-associated transplantation antigen” (TATA) and virus, the improved immune response of interest is directed against the TATA alone. “Adjuvanticity” carries with it the connotation that the virus augmentation effect might occur by a mechanism akin to the way BCG or Freund’s adjuvant operates, and that a simple mixture of virus with tumor homogenate might be effective. This is not true, however. Only the membranes of tumor cells actively infected with the virus exhibit augmented immunogenicity. We therefore prefer to use the term “virus augmentation of TATA” for the phenomenon described by Lindenmann. To emphasize the fact that the virus antigen on the tumor cell membrane must be in the molecular vicinity of the TATA to enhance the immunogenicity of the latter, the term “vicinal adjuvanticity” is also useful. There are two quite different groups of investigators interested in the use of viruses to augment the immunogenicity of TATA. One group, whose interests form the subject of this review, bases its activity on the initial findings of L.indenmann (described above) and deals exclusively with tumor cell extracts or simple homogenates of virus-infected tumor cells. This group has as one of its major goals the purification of virusaugmented TATAs to high specific activity, thereby increasing their potential usefulness in immunodiagnosis (as skin test antigens) and in immunotherapy. The other group deals exclusively with viable proliferating tumor cells and seeks to improve their immunogenicity for use in immunotherapy by productively infecting them with a surface-budding noncytopathic virus, frequently a type-C retrovirus. The details of this approach are described by H. Kobayashi (1979) in another chapter in this volume. Kobayashi Pf NI. (1969) were apparently the first to demonstrate the virus augmentation of TATA on intact tumor cells. They inoculated rats with syngeneic methylcholanthrene (MCA)-induced tumor cells that had previously been infected with Friend leukemia virus. Whereas uninfected tumor cells grew to lethal size, the Friend virus-infected tumor cells produced tumors which enlarged for about 2 weeks and then regressed completely. The “postregression” rats were strongly immune to challenge with uninfected tumor cells. Here, as in immunization with membrane extracts of influenza virus-infected tumor cells, primary immunization with the combination of TATA and virus produced an augmented immune response to the TATA. Kobayashi introduced the term “artificial xenogenization” to describe this phenomenon. Unfortunately, his term is easily confused with one used earlier by Svet-Moldavsky and Hamburg (1964), ”artificial heterogenization,” which they used to describe the
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
303
different concept of initiating a virus infection in a tumor as it is growing in the host, thereby inducing an immune response to the virus antigen in the tumor cell membrane which brings about the destruction of the tumor cell. Note that in this case the presence of a TATA is not required. A naturally occurring virus infection in a growing tumor, with its attendant alteration of the cell surface by insertion of a virus antigen, Svet-Moldavsky ( 1979) termed “natural heterogenization.” Kobayashi has described still another phenomenon using the term “natural xenogenization” (Kobayashi et al., 1977) which he defines in relation to the regression of murine leukemia virus-induced tumors in rats. This confusion of terms has been further compounded by a tendency of both SvetMoldavsky and Kobayashi, and others [including Boone (1972)] to use the words “heterogenization” and “xenogenization” in their literal sense, i.e., “to make different o r foreign.” Thus, there appear terms such as viral xenogenization (or heterogenization), chemical xenogenization (or heterogenization), xenogenization with enzymes, etc. We feel that eventually the term “xenogenization” in the general sense of “to make foreign” may take hold, but for the present and also to emphasize that this review concerns the enhanced immunogenicity of tumor cell extracts and not whole cells, we will continue to use the term “virus augmentation of TATA.” II. Virus Therapy of Cancer
Awareness of the phenomenon of virus augmentation of TATA immunogenicity evolved directly from attempts to treat tumors with cytopathic viruses. It is therefore germane to describe the development and current status of virus therapy of cancer in animal models and in humans.
A. VIRUSTHERAPYI N ANIMALS The oncolytic properties of several groups of viruses [e.g., arboviruses (togaviruses, rhabdoviruses, Bunyamwera viruses), myxoviruses, pox viruses (vaccinia)] have been demonstrated in various animal systems. Early experiments, which have been reviewed by Moore (1954, 1960), Southam (1960), and Lindenmann and Klein (1967b), demonstrated a lack of consistency in virus-tumor interactions: i.e., no tumors were uniformly susceptible to oncolysis by all viruses, and no viruses were uniformly oncolytic for all tumors. In early studies, Sharpless et al. (1950) noted that certain neurotropic
304
FAYE C . AUSTIN A N D CHARLES W . BOONE
arboviruses had an antagonistic effect on lymphoid tumors in chickens, with resulting tumor immunity. Koprowski et ( i f . (1957) studied the oncolytic effect of several arboviruses on 12 ascites tumors of mice and found that not all types of tumors supported virus growth and only certain types of viruses would thrive on malignant cells. However, neoplastic tissues showed enhanced susceptibility to virus infection. They noted that mice which had survived viral oncolysis were more immune to subsequent tumor graft challenge. Lindenmann (1963) reported that after influenza virus oncolysis of growing Ehrlich ascites (EA) tumors in mice, a strong postoncolytic immunity resulted. Oncolysis of EA tumors by vesicular stomatitis virus was also demonstrated (Lindenmann, 1970). Although the above studies used oncolytic viruses with marked neurotropism, Cassel and Garrett (1965, 1966) demonstrated that a strain of Newcastle disease virus (NDV) with minimal neurotropism could effect a cure of mouse ascites tumors which had reached up to 41% of their total possible development, with resultant tumor immunity. Studies of oncolysis with NDV, vaccinia, and influenza virus led Castle and Garrett ( 1967a,b) to postulate a relationship between viral neurotropism and oncolytic activity. Klein (1967) reported the induction of antibody-mediated immunity to transplantable tumors following oncolysis with reovirus, the mechanism of which is uncertain since reovirus is not a surface-budding virus and is not known to induce cell surface antigens in the infected cell. Interpretation of these experiments is difficult because they have not been repeated in a syngeneic system. Taylor ct uf. (1970, 1971) and Sedmak et ul. (1972) determined that bovine enterovirus caused a rapid regression of nonspecific transplanta. ble tumors in outbred mice. The mechanism of initial tumor cell destruction was due to direct viral cytolysis, while the destruction of any residual tumor was attributed to the development of postoncolytic immunity. The specificity of virus for tumor cells i n vilw was shown to correlate with cytopathic effect (CPE) observed in tumor cells infected in vitvo, which might allow for the in vitvo screening of various viruses to assess their oncolytic potential. In specificity studies, bovine enterovirus was shown to adsorb only to tumor cells of other species and to normal bovine cells. The authors postulated that this specificity of adsorption to the membranes of tumor cells might possibly be due to their increased negative charge. Rukavishnikova and Alekseyeva (1976) studied the use of a vaccine strain of influenza A virus to inhibit the growth of ascites tumor cells in outbred rats and inbred mice. They produced complete regression of 35% of tumors in outbred rats by daily inoculation of virus intraperitoneally
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
305
(i.p.1 and intramuscularly (im.) for 1 week. With a syngeneic mouse ascites lymphoma, however, inoculations of tumor with influenza virus resulted only in prolongation of the survival time of tumor-bearing mice. Recently, Eiselein et al. (1978) reported on the treatment of transplantable mouse mammary tumors, both in solid and ascites form, with an unidentified virus isolated from an ascites form of mammary tumor which had unexpectedly regressed without treatment. This viral agent was adapted to grow in tumor cells in v i m , and has been shown to cause oncolysis of ascites tumors when injected i.p., but not when injected intravenously (i.v.), i m . , or subcutaneously (s.c.). Injection of the virus into solid tumors did not produce the dramatic lysis observed with ascites tumors. However, the addition of cyclophosphamide (CTX) to the treatment schedule significantly improved the cure rate (14% cures with virus alone, 83% cures with virus plus CTX). The virus treatment was effective even with tumors containing several grams of tissue. The authors postulated that in order for this therapy to be effective against tumor metastases, it will be necessary to develop a way to deliver the virus to the metastatic lesions. Work is in progress to identify the virus.
B. VIRUSTHERAPY I N HUMANCANCER PATIENTS The wave of initial interest in treating human cancer with viruses that occurred between 1950 and 1967 has been reviewed by Moore (1954, 1960), Southam (1960), and Lindenmann and KIein (1967b). Many viruses, primarily arboviruses, were used to treat patients with advanced cancers who were not responsive to any established method of treatment. In general, clinical applications of virus therapy as presented in these early reviews did not produce significant improvement in the clinical course of the disease. Smith rt a / . (1956) reported on the use of adenovirus to treat carcinoma of the cervix, which resulted in sloughing of the tumor after viral oncolysis, but with no appreciable modification in the course of the disease. Suskind rt a / . (1957) studied the oncolytic effect of Coxsackie B virus, an enterovirus, on human cancer cells (HeLa and KB) which had been heterotransplanted into irradiated, cortisone-treated rats. Evidence indicated that the virus did not proliferate in tissue other than the tumor in this system and adaptation of the virus to the tumor tissue enhanced the oncolytic effect. According to Southam (1960), Egypt 101 strain of West Nile virus was the only virus among many tested which produced any convincing evidence of oncolytic effect in clinical trials. Cassel and Garrett (1965) attempted virus therapy of a cervical carcinoma with New-
306
FAYE C . AUSTIN A N D CHARLES W . BOONE
castle disease virus, chosen because of its low degree of neurotropism. Inoculation of the virus into a carcinoma of the cervix resulted in extensive sloughing of the tumor and shrinkage of the lymph node metastases, with no evidence of the virus attacking nerve tissue. Webb et al. (1966) reported results of a clinical trial of virus therapy with Langat and Kyasanur Forest disease viruses. The patients in these studies had advanced disease and little effect was noted. Later, Webb and Gordon Smith (1970) discussed the possible beneficial effects of virus therapy in the treatment of human cancer. Examples of tumor-regressive effects of virus infections in cancer patients continue to be reported (Csatary, 1971; Pasquinucci, 1971: BIuming and Ziegler, 1971). The replication in vitro of surface-budding viruses in human cancer cells has been studied (Gerber et d.,1973a,b; Sauter P I a / . , 1973, 1975: Illiger et d., 1975). Sauter et al. (1972) treated a patient with acute myelogenous leukemia (AML) with avian influenza virus adapted to leukemia cells and reported that although the patient had a poor prognosis, a striking hematological improvement was produced. Clinical trials with virus-assisted immunotherapy are continuing. Lindenmann (l979a) and colleagues tested avian influenza virus, adapted in iyifro to the particular cell type involved, for its ability to replicate in i ~ h ~ o when injected into a large tumor mass. In one patient, the virus was inoculated into a carcinomatous pleural effusion, causing regression of the pleural effusion and palpable lymph nodes (Sauter and Lindenmann, 1976). I n another patient (C. Mayer, in preparation, cited in Lindenmann, 1979a) virus therapy appeared to have a positive effect on a pleural exudate and tumorous infiltrates around the neck. Although objective improvement was not long lasting, viral replication was demonstrated in \ t i i v , with effects on the tumor at both local and distant sites. In the U.S.S.R., Muceniece (1972, 1978) studied the suitability of human enteroviruses for viral therapy of tumors and established that about 70% of these viruses possessed oncotropic and oncolytic properties in heterotransplants of human angiosarcoma and lymphoma. Experiments in vitr’o confirmed that various human tumors adsorbed human enteroviruses. The spectrum of viruses adsorbed depended on the individual tumor. The oncotropic and oncolytic activities of human enteroviruses were then assessed in patients with advanced cancer. After i.m. injection, viral antigens and cytopathic changes typical of enteroviruses were detected in about 50% of excised tumor cells. No infectious virus could be isolated. Susceptibility to oncolysis was found to be dependent on tumor type. “Gastrointestinal tract cancer,” some melanomas, and certain sarcomas were found to be sensitive to oncolytic enteroviruses, while breast and lung cancers were insensitive. Viral oncolytic activities ceased one
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
307
week after therapy due to the induction of antiviral immunity. Muceniece (1978) pointed out that virus therapy can be used in conjunction with chemotherapeutic agents if the latter are used in subtherapeutic doses that are not immunosuppressive. In fact, it has been shown in animal studies that virus-infected tumor cells are more susceptible to chemotherapy (Eiselein and Biggs, 1970). Svet-Moldavsky (1974, 1979) reported negative results in attempts to treat patients having widely disseminated cancer with adenovirus therapy. Roenigk et al. (1974) reported positive results of immunotherapy of malignant melanoma with intralesional inoculation of vaccinia virus. Major regression of the tumor was observid in 8 out of 8 patients with stage I1 disease. In studies of patients with stage I11 disease, however, little or no regression occurred. Other studies of vaccinia virus therapy have demonstrated positive results, primarily in patients without extensive tumor involvement (Belisario and Milton, 1961; Milton and Brown, 1966; Hunter-Craig et ul., 1970). In Japan, Asada (1974) described the oncolytic effect of mumps virus on rat tumors. His results appeared to warrant performing mumps virus therapy in humans with advanced cancer. Of 90 patients with terminal cancer of various types, treatment was assessed as very good in 37 (tumor disappeared or decreased in size to less than half of initial size, subjective symptoms improved) and good in 42 (some suppression of tumor growth, subjective symptoms improved). Depending on the patient’s general condition and the type, size, and location of the tumor, mumps virus was applied either externally, locally, orally, rectally, intravenously, or by inhalation. Administration of virus produced few side effects. The initial antineoplastic effect of mumps virus therapy seemed to occur rapidly and strongly in proportion to the growth rate of the tumor, an observation possibly related to increased virus replication in rapidly growing cells (DeMarchi and Kaplan, 1977). The oncolytic effect of the virus came to an end in a short time because of the induction of antiviral immunity. However, there was little additional proliferation of the tumor cells for a long period of time, even if the remaining tumor was left untreated, a phenomenon probably related to postoncolytic immunity. The recent reports described above indicate that the use of surfacebudding viruses in virus therapy may possibly involve the host’s immune system in active immunization, resulting in postoncolytic immunity to residual tumor cells. An enhanced immunological response to the tumor antigen during the course of viral oncolysis would obviously be of great advantage during virus therapy, for the host’s immune system will soon interfere with continued oncolysis by the virus. There are many hazards involved in virus therapy, especially the possibility of viral pathogenicity
308
FAYE C. AUSTIN A N D CHARLES W . BOONE
in an immunosuppressed patient. The selection of a virus is critical, for a successful agent must be oncotropic, oncolytic, and nonpathogenic. Preimmunization of the patient with inactivated virus would diminish any possibility of viremia (Lindenmann, 1963). The success of potential virus therapy will also depend on careful attention to dose, route, and timing of inoculations: the effect of preimmunization with virus: and coordination with other types of therapy. Ill. Augmented lmmunogenicity of Virus-Infected Tumor Cell Extracts
A. ANIMALMODEL STUDIES
I . Criidc Mrmbranr Extracts The first extended experiments to evaluate postoncolytic immunity were performed by Lindenmann and Klein (1967a). When the WSAIAo strain of influenza virus was inoculated i.p. into EA tumors in A2G mice, which are genetically resistant to infection by the virus, the ascites collapsed within 48-72 hours and an agglutinated membranous mass of lysed tumor cells, the oncolysate, was found in the peritoneal cavity. Dilutions of the mechanically dispersed oncolysate up to 1/15 protected normal mice against challenge by lo3 tumor cells, whereas mechanical lysates of the EA tumor gave no protection at all. A number of significant experiments gave insight into the mechanism of the virus augmentation, or viral adjuvanticity , phenomenon..Postoncolytic immunity could be transferred to normal mice by either serum or lymphoid cells (Lindenmann, 1964). The serum was found to contain a high titer of antibodies directed against an alloantigen ( E antigen) present in the tumor cell membrane and in normal tissues of other mouse strains (Klein and Lindenmann, 1965). Later, when other investigators performed similar experiments in syngeneic tumor systems, rather than with the nonspecific allogeneic EA tumor, the immunity produced by virus-augmented TATA was transferrable with lymphoid cells, but never with serum. In spite of the fact that the EA tumor was allogeneic, Lindenmann and Klein (1967a) were able to determine a number of important facts about virus augmentation that were later confirmed in syngeneic systems. Viral oncolysates did not lose their augmented immunogenicity after fixation with 0.08% formaldehyde, which completely inactivated virus infectivity. Simple mixtures of egggrown virus with mechanical lysates of tumor cells were not immunogenic: Actual virus infection of the tumor cells was required before
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
3 09
augmentation of immunogenicity occurred. Two different types of experiments both appeared to show that the immune system was involved in the virus augmentation phenomenon: ( 1 ) Preimmunization with egggrown virus enhanced even more the immune response to immunization with oncolysate. (2) When viral oncolysates were mixed with rabbit antiserum against egg-grown influenza virus, the resultant mixture did not induce either antiviral or antitumor responses. Immunization with egg-grown virus alone produced no tumor protective effect. Thus, the mechanism of virus augmentation was not due to a nonspecific adjuvant effect nor to any changes produced by live virus. Lindenmann (1970) extended his studies of TATA augmentation to include vesicular stomatitis virus (VSV), a rhabdovirus which also matures by budding at the cell surface and which has only one envelope antigen, a hemagglutinin, compared to two, a hemagglutinin and a neuraminidase, possessed by influenza virus. VSV was chosen because it is easily adaptable to many different tissues, in contrast to influenza virus, and is known to contain antigens derived from the host cell. The finding that inbred A2G mice could be immunized against the EA tumor by inoculation with viral oncolysates prepared from VSV-infected EA tumor cells led to the conclusion that VSV acted as an immunological carrier for EA tumor cell antigens which had become incorporated into the coat of the mature virion. In these experiments, it is noteworthy that although the infectivity titer of VSV in viral oncolysates reached a maximum within 24 hours, the immunogenicity of the oncolysates was greatest at 48 hours. It appeared that immunogenic material had accumulated between 24 and 48 hours after infection without concomitant increase in the number of infectious virus particles. Lindenmann (1970, 1973) speculated that a suitable virus could be grown in cancer tissue obtained from a patient and then, after inactivation, be injected into the patient to increase his defenses against the specific antigens of his own tumor. The patient could be protected from any adverse viral infection by a previous vaccination with virus. How much of Lindenmann's speculation has come to pass will be presented in a later section on the application of virus-augmented tumor antigens to human cancer immunotherapy. Since these initial studies, many reports of virus-augmented TATA in animal model systems have appeared (Table I). Hakkinen and Halonen (1971) showed that oncolysates could be prepared from EA tumor cells grown in suspension culture and infected in v i m with either influenza virus or VSV. Unconcentrated cell culture supernatant, purified concentrated virus, and a 10% suspension of virus-infected EA tumor cells all were able to induce antitumor immunity.
S U M M A R Y OF PUBLISHED
Virus
TABLE I DATAON VIRUS-AUGMENTED TUMOR CELL EXTRACTS (ANIMALSTUDIES)
Preparation of infected cells
Tumor
Host
Influenza
In iYi-o
Ehrlich ascites (EA)
A2G mice
Influenza (Hong Kong strain)
117 i.iiw
EA
A2G mice
Influenza
In
EA
BALB/c mice
In ritro
SV4013T3
BALB/c mice
IIT isitro In i'itro
MCA-induced fibrosarcoma MCA-induced fibrosarcoma Lewis lung carcinoma WEHI-I 1 fibrosarcoma
BALB/c mice C57BL/6 mice C57BL/6 mice BALB/c mice
vitro
111 i i t r o Iri
isitru
Reference Lindenmann and Klein (1967a): Lindenmann (1971, 1973, 1974, 1977) Haller and Lindenmann (1975) Hakkinen and Halonen (1971) Boone et 01. (1971, 1973, 1974, 1979a): Boone (1972, 1974); Boone and Blackman (1972); Boone and Gillette (1977): Takeichi et al. 1978 Boone ( 1972) Klein (1974) Griffith e / a / . (1973 Griffith et a / . (1975)
Newcastle disease viius (NDV)
SV40-transformed human cells Sarcoma-37 (S 37) Gross ascites lymphoma
Hamster
Axler and Girardi (1970)
CBA mice C3H/Bi mice
Beverley et ai. (1973) Eaton et a / . (1973); Eaton and Almquist (1975)
EA
A2G mice
Lindenmann (1970)
EA
BALB/c mice
SV4W3T3
BALB/c mice
SV40-transformed TSV-5 CI-2
Inbred Syrian hamsters
L1210 leukemia
C57BL16 mice
Hakkinen and Halonen (1971) Boone et al. (1974): Gillette and Boone ( 1976) Ansel(l974); Ansel er ai. (1977); Huet and Ansel ( 1977) Wise (1977)
Sendai virus
Gross ascites lymphoma
C3H/Bi mice
Eaton et al. (1973)
Semliki Forest virus
WEHI-11 fibrosarcoma
BALB/c mice
Grifith el ai. (1975)
Vaccinia
SV40-transformed peritoneal macrophage tumor
BALB/c mice
Wallack et a/.(1977)
Vesicular stomatitis virus (VSV)
(I
lmmunogen consisted of purified VSV grown in the tumor cells.
3 12
FAYE C. AUSTIN A N D CHARLES W . BOONE
In the first studies of virus augmentation in a syngeneic tumor system, Boone et al. (1971) demonstrated that BALB/c mice could be made immune to tumor challenge with syngeneic SV4O-transformed fibrosarcoma cells (E4) by vaccinating them with homogenates of these cells that had been infected in vitro with influenza virus. No immunity was induced by vaccinating mice with homogenates of either uninfected cells or of influenza virus-infected normal (3T3) cells. The syngeneic cultured tumor cell line (E4, formerly called SV3T3-T4) was obtained as an explant from a solid fibrosarcoma in a BALB/c mouse inoculated with SV40-transformed Balb/3T3 cells. The use of the E4 cell line allowed for a direct comparison with its normal counterpart, Balb/3T3 cells. Boone et al. (1971) used the same WSA strain of influenza A virus that had been used by Lindenmann but with further adaption to growth in E4 cells. The SV40 TATA is a strong antigen, and immunization with X-irradiated E4 cells induced complete immunity to a tumorigenic challenge dose of viable E4 tumor cells. Inoculation of normal 3T3 cells in the same manner did not. The immunogenicity of SV40 TATA is destroyed if the tumor cells are homogenized, but retained in homogenates of tumor cells previously infected with influenza virus. A typical tumor protection experiment is illustrated in Table 11. Tumor cell challenge after immunization with virus-augmented E4 cell extracts resulted in a tumor incidence of 24% as compared to 85% for control mice, and 82% for mice immunized with a homogenate of uninfected E4 cells. The immunogenicity of virusinfected crude membranes (CM) was still less than that of X-irradiated intact tumor cells. The immunizing dose of CM was equivalent to apTABLE I1 AUGMENTED TATA ACTIVITY OF CRUDE MEMBRANES OF E4 TUMOR CELLSINFECTED W I T H INFLUENZAVIRUS Immunogen"
lofi X-irradiated (4000 rad) E4 cells Uninfected E4 CMb, 500 pg FIu-E~CM,b*'500 pg Saline
Tumor incidence after challenge"
'
1/20 (5%) 14/17 (82%) 4/17 (24%) 17/20 (85%)
Immunization (s.c.) on days 0 and 3. Crude membranes prepared by centrifugation (100.000 g , I hour) of a nucleus-free Dounce homogenate. Protein concentration determined by Lowry method. Cells harvested 24 hours after infection with influenza virus, multiplicity of infection: 20. HA titer of infected CM: 4096. Ratio of mice with tumor to mice challenged with lofi E4 cells 14 days after immunization. Tumor incidence recorded 28 days after challenge. a
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
313
proximately lo7 cells. Virus-augmented tumor cell homogenates retained their immunogenicity after inactivation with 3% formalin (Boone and Blackman, 1972). To evaluate the relative immunogenicity of various subcellular fractions, influenza virus-infected E4 tumor cells were homogenized and fractionated by sucrose density gradient centrifugation. Maximum immunogenic activity was found in the 33%-40% (plasma membrane fragments, intact influenza virus) and 40%-45% (plasma membrane fragments, rough endoplasmic reticulum, ribosomes, mitochondria, lysosomes) sucrose fractions. It was postulated that most of the augmented TATA in the 33%-40% fraction may exist as virus antigen-containing plasma membrane microvesicles, rather than as virus with hostcell TATA incorporated into its envelope (Boone, 1972; Boone ef al., 1973). The relative roles of viral antigens incorporated into the tumor cell membrane versus tumor antigens incorporated into the virus envelope in the virus augmentation phenomenon is still unresolved. (See Section 111,A,3 .) In further studies, Boone et al. (1974, 1979a) demonstrated the augmented immunogenicity of homogenates of tumor cells infected with the Hong Kong strain of influenza virus and with VSV. UV inactivation of the virus-infected tumor cell homogenates did not destroy TATA activity. Evidence was presented in support of the “helper antigen” hypothesis: that the viral antigen works through the host’s immune system to induce an augmented response to the TATA of the tumor cell. It was also noted that although homogenates of uninfected tumor cells elicited a delayed hypersensitivity (DH) response when injected into the footpads of tumorimmune mice, homogenates of virus-infected tumor cells produced a more intense footpad DH reaction. This observation was later confirmed (Austin el al., 1977) and formed the basis for the development of virusaugmented human skin test antigens (Boone rf al., 1978, 1979b; see Section 111,C). Influenza virus was also shown to augment the immunogenicity of homogenates of MCA-induced tumor cells, but the effects seen with this chemically induced tumor were not as striking as those seen for the SV40-transformed E4 tumor cell line (Boone, 1972). This may have been because the influenza virus stock had not been fully adapted to grow in this cell line. In another study of chemically induced tumors, Klein (1974) found that the same WSA strain of influenza virus which had been used in earlier studies with EA tumors failed to grow consistently in MCA-induced tumors of C57BL/6 mice. Following a period of adaptation, however, this virus could be grown to high titers in most but not all lines, resulting
3 14
FAYE C . AUSTIN A N D CHARLES W . BOONE
in lysis of the cultured tumor cells. VSV and encephalomyocarditis virus grew in all tumor lines tested without adaptation. Klein (1974) reported the augmented immunogenicity of one line of cultured MCA-induced tumor cells following influenza virus infection. This immunoprophylaxis experiment was performed in virus-immunized mice because earlier observations had indicated that virus-primed mice are more sensitive indicators of virus-augmented tumor cell immunogenicity (Lindenmann and Klein, 1967a). Although Lindenmann had originally stressed the necessity for working with viruses that were fully adapted to grow in tumor cells, Haller and Lindenmann (1975) reported the augmentation of antigens of EA tumor cells produced by infection with the Hong Kong strain of influenza A virus that had not been previously adapted to this tumor. Infection of tumor cells in i i i r v with this strain of influenza virus resulted in oncolysis, and the resultant oncolysates had high levels of hemagglutinin, but low titers of infectious virus, indicating incomplete virus growth. Nevertheless, these oncolysates induced solid antitumor immunity against EA tumors in various mouse strains. This virus strain had also been shown to augment TATA immunogenicity in a syngeneic tumor system, although the yield of infectious virus in this system was not determined (Boone et id., 1974, see above). Of course, augmentation of TATA immunogenicity by an incomplete virus growth cycle would eliminate the complications arising from the presence of large quantities of infectious virus. This approach to the preparation of virus-augmented tumor antigens has obvious potential in human cancer immunotherapy. In other basic studies, Beverley et ul. (1973) immunized mice with NDV-infected tumor cells or cell membrane fractions of a transplantable ascites sarcoma and found that the virus-infected cell membranes were more immunogenic than similar fractions prepared from uninfected cells. The immunity conferred was not very strong. Complete virus replication did not occur in the tumor cells and the authors considered the possibility that the infection by NDV was aborted at a stage prior to the development of a hapten carrier association. In this system preimmunization with virus abrogated the virus-augmented immunogenicity of tumor cell extracts. Eaton et ul. (1973) also studied the effect of NDV, as well as parainfluenza 1 (Sendai strain), to augment the immunogenicity of a Gross virusinduced ascites lymphoma in syngeneic C3H/Bi mice. Multiple injections of NDV-infected cell extracts were required for successful immunization. If the first injection was mixed with complete Freund’s adjuvant (CFA), it was found that fewer injections of antigen were required for immunization. The result that one or two injections of extract with CFA failed to immunize was consistent with the weak immunogenicity of NDV-
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
3 15
augmented extracts reported by Beverley et al. (1973). Experiments with cells treated with nonreplicating virus again confirmed that the growth of virus was essential to obtain virus-augmented TATA immunogenicity. Tests of various cell fractions indicated that the major portion of TATA activity was in the membrane fraction, with none in the cytoplasmic supernatant. Virus and other small particles concentrated from the tissue culture supernatant by ultracentrifugation were less effective than membranes in inducing tumor immunity. Further studies (Eaton and Almquist, 1975) of NDV-infected tumor cells demonstrated that tumor immunity was not produced by the high-speed supernatant of infected cell homogenates, by the virus fraction of tissue culture supernatants, or by detergent extracts of the membrane fractions of infected tumor cells. Griffith et al. (1975) used influenza virus or Semliki Forest virus (SFV) infection to augment the immunogenicity of extracts of cultured WEHI11 cells, a fibrosarcoma of BALB/c mice. In addition, the life span of C57/B1 mice inoculated with Lewis lung carcinoma cells was prolonged if the mice were preimmunized with membranes of the cells which had been infected in vitro with influenza virus. The influenza virus-infected membrane preparations were inactivated by exposure to UV light. Membrane preparations containing infectious SFV were inactivated by heating at 56°C for 1 hour or by treatment with 0.1% sodium deoxycholate. In the U.S.S.R., Rukavishnikova and Alekseyeva (1976) demonstrated that vaccine strains of influenza A virus inhibited the growth of a syngeneic virus-induced ascites tumor in BALB/c mice. Also, immunization with the viral oncolysates protected recipient mice against subsequent tumor challenge. Wise (1977) studied the potentiation of immunogenicity of L1210 mouse leukemia cells after in vitro infection with VSV and also demonstrated the virus-augmented immunogenicity of extracts of VSV-infected tumor cells. All homogenates were UV-inactivated in these experiments. Various aspects of the interaction of the virus with cellular antigens in this study will be discussed in subsequent sections (Sections 111,A,4 and IV,A). Wallack et al. (1977) demonstrated the use of vaccinia virus to augment the immunogenicity of an SV40-transformed mouse peritoneal macrophage tumor. In searching for a model system for the development of immunogenic viral oncolysates with application to the treatment of human cancer, they tested six live virus vaccines (measles, mumps, smallpox, rubella, yellow fever, rabies) against four human tumor cell lines (ovary, lung, melanoma, colon) and noted that only vaccinia (smallpox vaccine) duplicated the oncolytic action of influenza virus on EA tumor cells as previously described by Lindenmann (1963). Vaccinia has
3 I6
FAYE C. AUSTIN A N D CHARLES W . BOONE
also been used in attempts to treat melanoma patients, with some positive results (Burdick, 1960; Burdick and Hawk, 1964; Roenigk et a / . , 1974: see Section 11,B).There was no need to inactivate infectious virus in the vaccinia virus oncolysate since the safety of the vaccine strain had been established. The choice of vaccinia virus to augment tumor cell extract immunogenicity is at first surprising, since vaccinia is not known to be a surface-budding virus. However, vaccinia appears to enter the host cell by fusion of the virus envelope with the plasma cell membrane, followed by distribution of vaccinia antigens in the entire cell surface membrane (Chang and Metz, 1976). In addition, during the course of virus replication, virus-induced cell surface antigens do appear (Burns and Allison, 1977) and can function as helper antigens for the TATA of the tumor cell. Wallack demonstrated that mice which were immunized with the vaccinia oncolysate were rendered solidly immune to challenge with viable tumor cells. Preimmunization with vaccinia virus alone prior to treatment with vaccinia oncolysate did not interfere with the induction of tumor immunity by viral oncolysates. In addition to immunoprophylaxis, Wallack demonstrated therapeutic effects of virus-augmented tumor cell extracts on growing tumors in mice. These successes led him to attempt active specific immunotherapy of human cancer with vaccinia oncolysates. The results of his clinical studies, and other ongoing clinical trials with virusaugmented tumor cell extracts, will be described in Section III,B.
2. Solirble Extructs c,f Virus-Infected Tumor Cells The preparation of soluble TATAs has been the object of a vast amount of research in recent years, impelled by the hope that they might provide standardized reagents for immunotherapy and also facilitate understanding of their biological role in carcinogenesis. Progress has been generally frustrating and only marginally successful because of the heterogeneity and striking lability of most TATAs (for review, see Law and Appella, 1975). Methods that were successful for purification of histocompatibility antigens have also been applied to the isolation of tumor antigens (e.g., 3 M KCl extraction, nonionic detergent solubilization, etc.). In comparing various methods of antigen solubilization, we have found that soluble extracts of influenza virus-infected tumor cells produced by high-speed centrifugation of Dounce homogenates, by sonication of membranes, or by 3 M KCI extraction of cells still exhibit augmented TATA activity (Boone et a / . , 1979a; Austin et ul., in preparation). The augmented immunogenicity of a representative soluble antigen preparation obtained after high-speed centrifugation (lO0,OOO g , I hour) of a nucleus-free Dounce hornogenate is shown in Table 111. The homogenate was sepa-
3 17
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
TABLE 111 VIRUS-AUGMENTED TUMOR ANTIGENS BY HIGH-SPEED ( 100,000n ) OF NUCLEUS-FREE DOUNCE HOMOGENATES
PREPARATION OF SOLUBLE CENTRIFUGATION
Results of tumor challengeb Immunogena
Tumor incidencec
Mean tumor weight(gm) 2 S E
loBX-irradiated E4 cells E4 homogenate Flu-E4 homogenate E4 crude membranes Flu-E4 crude membranes E4 soluble Flu-E4 soluble Saline
0120 (0%) 8/13 (62%) 3/13 (23%) 17/20 (85%) 4120 (20%) 13/20 (65%) 4114 (2%) 17/20 (85%)
0.39 ? 0.18 0.45 2 0 . I I 0.67 0.13 0.09 0.06 0.67 2 0.14 0.36 ? 0.25 1.03 ? 0.13
* *
pd
< .01 < ,005 < .05 < ,0005
< .05 < ,025
Mice were immunized with lo6 X-irradiated (4000 rad) E4 cells or 500 pg of antigen S . C . on days 0 and 3. Mice challenged with 106 viable E4 cells, s.c., on day 14. Ratio of mice with tumors to mice challenged, 28 days after tumor challenge. " Comparison of mean tumor weight (of animals with tumors only) with that of control group (saline) by Student r test.
rated into a cytoplasmic soluble supernatant fraction and a crude membrane pellet fraction, both of which were tested for immunogenicity according to the protocol in Table 111. Soluble extracts from virus-infected tumor cells were more immunogenic than those from uninfected cells. Augmented TATA activity of the soluble supernatant had not been found in early studies (Boone et al., 1973). However, the present studies were performed with improved virus stocks that grew more efficiently in the tumor cells and may have produced a stronger augmenting effect. Our criteria for solubility have been lack of precipitation after centrifugation at 100,000g for 1 hour, or inclusion in the bed volume of 200,000 molecular weight cutoff gel filtration media. We have also solubilized tumor antigens by physical (sonication of cell membranes) and chemical (3 M KCI extraction) methods and found that after further fractionation by gel filtration, some fractions retained virus-augmented TATA immunogenicity (Austin et al., in preparation). Further study of the chemical nature of the virus-augmented soluble antigens should shed additional light on the mechanism of virus augmentation of TATA.
3 . Antigenicity of Tumor-Grown Viruses
To assess the role of host cell TATA incorporated into the envelope of surface-budding viruses, Ansel (1974) studied the immunogenicity of
318
FAYE C. AUSTIN A N D CHARLES W . BOONE
VSV which was grown in SV40-transformed hamster cells. When this virus was highly purified and inactivated, it possessed SV40 TATA activity as demonstrated by its ability to induce specific transplantation immunity to tumor cells which possessed the SV40 TATA. The presence of host cell antigens in this purified virus was also demonstrated by the ability of rabbit antiserum against purified tumor-grown virus to react specifically with SV4O-transformed cells in a complement-dependent 51Cr cytolysis assay. Ansel concluded that the presence of SV40 TATA activity in tumor-grown viruses was due to the incorporation of TATA into the virus envelope. In further studies (Ansel et al., 1977; Huet and Ansel, 1977), treatment of tumor-grown VSV with proteolytic enzymes did not destroy the virus-augmented TATA activity associated with the virus, although the VSV lost its ability to induce virus-neutralizing antibody, thus demonstrating that the augmented TATA activity was not dependent on the presence of the viral glycoprotein. Phospholipase C, however, did suppress the TATA activity of tumor-grown VSV, but not the antigenicity of the glycoprotein of the virus. Since the protein components of virus membranes are thought to be entirely virus specific, while lipids, glycolipids, and the carbohydrate moiety of the glycoprotein originate from the host, these results led to the hypothesis that SV40 TATA activity may be conveyed by the lipid core of the virus envelope. In contrast to Ansel’s experiments, Griffith et a / . (1975) found that partially purified SFV which was grown in WEHI-11 fibrosarcoma cells also protected mice from tumor cell challenge. However, neither highly purified SFV nor the glycoprotein from the envelope of this virus conferred protective immunity to mice. From these results, it was concluded that SFV-augmented immunogenicity was not due to covalent linkage of TATA to the viral envelope protein, but was most probably due to the apposition of viral glycoprotein and cellular TATA in the plasma membrane of the tumor cells. These results, which are in conflict with those of Ansel, may be due to a difference in the molecular structure of the TATA in these different tumors or to a difference in the composition of viral envelopes of the different viruses studied. Other studies have shown partially purified virus to be less effective than crude infected cell homogenates for the induction of antitumor immunity (Hakkinen and Halonen, 1971: Beverley et ul., 1973; Eaton et d..1973). 4. The Ncitirre of Virrrses Thut Augment TATA Activity Table I lists the several viruses that have been shown to augment TATA activity of tumor cell extracts by their growth in the cells prior to their disruption. The viruses listed are known to induce virus-specific
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
319
cell-surface antigens during their replicative cycle. In a study using the SV40-transformed cell line E4 described above, comparison was made of the ability of influenza virus and VSV to augment the immunogenicity of SV40-induced TATA (Gillette and Boone, 1976). Membranes from influenza virus-infected cells were immunogenic at ooth the dose required for membranes from VSV-infected cells. Subcutaneous inoculation was better than the intraperitoneal route of administration. Maximum protection against tumor cell challenge was afforded by two inoculations of VSV-infected cell membranes spaced 3 days apart or by a single inoculation with influenza virus-infected cell membranes. The use of Freund’s adjuvant produced either no effect or a slight enhancement of tumor growth. The viruses listed in Table I differ from those that have been used in studies of viral xenogenization of intact tumor cells (see H. Kobayashi, this volume) in that they were selected for their oncolytic effect on tumor cells, whereas the viruses chosen to augment the TATA activity of intact tumor cells produce a persistent infection with little or no cytopathic effect on the host cell. Augmenting viruses in the latter category include Sendai virus (Eaton et al., 1973), measles virus (Evermann and Burnstein, 1975), lymphocytic choriomeningitis virus (Eiselein and Biggs, 1970), and type-C retroviruses (Kobayashi et a / ., 1969, 1970; Kobayashi, 1970; Sendo et al., 1970; Kuzumaki and Kobayashi, 1976; Al-Ghazzouli et al., 1976; Kuzumaki et al., 1978; Kobayashi and Sendo, 1979). Most of this work is reviewed by Kob’ayashi et al. (1977) and in this volume (H. Kobayashi, 1979). To evaluate the effectiveness of type-C retroviruses in augmenting the immunogenicity of TATA in tumor cell extracts as contrasted to virusaugmented TATA on intact tumor cells, Takeichi et af. (1978) compared the augmented immunogenicity of E4 tumor cell membranes prepared after infection of the cells with influenza virus as opposed to infection with Moloney sarcoma virus (MSV). As expected, crude membrane extracts from influenza virus-infected E4 cells were markedly more immunogenic than extracts from uninfected cells, as measured either by the ability to induce protective immunity against tumor graft challenge or by heightened lymphocyte-mediated cytotoxicity against tumor cells in vitro. Next it was shown that E4 cells productively infected with MSV were rendered so immunogenic that they would grow only in X-irradiated syngeneic mice. In spite of this, the crude membrane extracts from these MSV-infected cells showed no augmented TATA activity such as was seen after infection with influenza virus. Thus, it appears that MSV, and possibly other type-C viruses, may augment the TATA activity of tumor cell extracts in a different way than lytic surface-budding viruses do.
320
FAYE C. AUSTIN A N D CHARLES W. BOONE
Other reports also point to this possibility. Eaton et al. (1967) had shown that syngeneic mice could survive repeated injections of NDVinfected Gross virus-induced lymphoma cells, for they were no longer capable of producing tumors. However, repeated injection of the virusmodified intact tumor cells did not induce immunity against challenge with uninfected Gross virus-induced lymphoma cells. Al-Ghazzouli et al. (1976) reported that the enhanced immunogenicity of type-C virus-infected intact tumor cells was destroyed after X-irradiation, but Kuzumaki et af. (1978) showed that some but not all tumors infected with endogenous type-C mouse virus retained augmented antigenicity after X-irradiation. Of interest is the report by Wise (1977) which demonstrated that although a single injection of X-irradiated L1210 leukemia cells protected mice against tumor cell challenge, the same dose of VSV-infected cells was not immunogenic after X-irradiation. On the other hand, homogenates of VSV-infected L1210 cells were immunogenic, but those of uninfected cells were not.
B. IMMUNOTHERAPY OF HUMAN CANCER: CLINICAL TRIALS The positive results obtained in immunoprophylaxis (tumor protection) experiments with virus-augmented tumor extracts, described in preceding sections, have led several investigators to attempt immunotherapy in human cancer patients with virus-augmented tumor cell extracts (viral oncolysates). A survey of the published reports of clinical trials to date (Table IV) reveals that initial results have not been particularly striking. However, the studies were performed in patients with advanced disseminated disease, where regressive effects on tumor growth are harder to detect. Sinkovics and his colleagues (Sinkovics, 1977; Sinkovics et al., 1974a,b, 1977: McMurtrey ef ul., 1976) implemented an immunotherapy protocol, in conjunction with chemotherapy, in attempts to stimulate the patient’s immune responses to his own tumor. They have established in culture many cell lines from human sarcomas and have prepared viral oncolysates from most of them by infecting them with the PR8 strain of influenza A virus. Initial studies showed an increase in antitumor cell cytotoxic lymphocytes and conversion from blocking to potentiating serum factors in the peripheral blood of patients who were receiving immunotherapy with either X-irradiated allogeneic sarcoma cells or with viral oncolysates of allogeneic sarcoma cells. In further studies, the effect of tumor-specific active immunization with viral oncolysates in combination with chemotherapy and nonspecific immunization with BCG was investigated (Sin-
TABLE IV SUMMARY OF PUBLISHED REPORTSOF IMMUNOTHERAPY WITH VIRUS-AUGMENTED H U M A NTUMOR EXTRACTS ~
Virus
Tumor
Viral oncolysate
Influenza (PR8)
Sarcomas, melanoma
Allogeneic, UVinactivated
Influenza (fowl plague virus) Influenza (A2 or B)
AML"
Newcastle disease virus
Melanoma
Allogeneic, formaldehydeinactivated* Autologous, and allogeneic, formalininactivatedc Allogeneic, noninactivated'
Vaccinia
Carcinomas, melanoma
Osteosarcoma
Acute rnyelogenous leukemia. Given along with chemotherapy. Given along with chemotherapy in some patients.
Autologous, noninactivated
Effect of treatment Increased lymphocytemediated cytotoxicity: slowed progress of disease No apparent effect Increase in cellular and humoral antitumor responses Increased lymphocyte cytotoxicity: decreased size of skin nodules and/or diseases lymph nodes (6 of 13 patients): no effect on visceral disease Evidence of control of tumor growth (9 of 29 patients)
Reference Sinkovics (1977): Sinkovics et al. (1974a,b, 1977): McMurtrey et al. (1976) Sauter et al. (1978); Lindenmann (1979a) Green et al. (1976)
Cassel et al. (1977): Murray et al. (1977)
Wallack (1979a,b): Wallack et al. (1977)
322
FAYE C. AUSTIN A N D CHARLES W . BOONE
kovics, 1977; Sinkovics et ul., 1977). Patients with melanoma or metastatic soft tissue sarcomas received intermittent chemotherapy with or without BCG (6 x IOR viable units scarified twice monthly) or identical chemotherapy, BCG, and viral oncolysate [equivalent of 10' tumor cells injected intracutaneously (i.c.) in the region of BCG scarification twice monthly]. Chemotherapy was given on days 1 through 5; immunotherapy was given on days 17 and 24. Courses were repeated at 28-day intervals. Viral oncolysates were prepared from allogeneic cultured sarcoma cells infected in \iituo with influenza virus (PR8). At the end of the first year of treatment, the results were as follows: of 48 patients with metastatic sarcomas receiving chemotherapy only, 10 had complete remission (21%), and 6 had partial remission or stable disease status (13%). In 32, the disease progressed, often resulting in death (66%). Of 10 patients with metastatic sarcomas receiving chemotherapy and BCG, one had complete remission (lo%), 5 had partial remission or stable disease status (50%), and 4 had disease progression, some resulting in death (4%). Of 14 patients with metastatic sarcomas receiving chemotherapy, BCG, and viral oncolysates, 6 had complete remission (43%), 6 had partial remission or were in stable disease status (43%), and 2 had disease progression (14%). Similar results were reported following treatment of melanoma patients. Of 12 patients receiving chemotherapy and BCG, 6 had disease progression (50%), whereas only 7 out of 27 patients (26%) receiving comparable chemotherapy with BCG and viral oncolysates had disease progression. These results clearly indicated a trend favoring active tumorspecific immunization with influenza virus-augmented allogeneic tumor cell extracts and should be confirmed by observation of these patients for a longer period and by prospective randomized trials. Sauter et ul. (1978) and Lindenmann (1979a) performed a randomized clinical trial to assess the effect of immunization with viral oncolysates in AML patients in remission. They used avian influenza A virus (fowl plague virus) which had been adapted to several types of human tumor cells including human leukemic myeloblasts (Sauter Pt ul., 1973, 1975: Illiger ef u l . , 1975). Patients with AML were chosen for this study because, as soon as they had achieved complete remission induced by chemotherapy, they retained only a minimum residual tumor mass. Also, since the median remission time in this disease is less than 1 year, results of additional therapy would be seen within a short period of time. Patients achieving complete remission were randomized into two groups: one receiving chemotherapy alone, and the other receiving identical chemotherapy plus immunization with viral oncolysates. Chemotherapy was given for a 5-day course every 4 weeks throughout duration of this study. In addition, patients receiving immunotherapy received, on day 15 of
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
323
each 4-week cycle, 1.0 ml of viral oncolysate (0.2 ml i.c. and 0.8 ml s.c.). One milliliter of viral oncolysate was equivalent to log leukemic myeloblasts. The virus was inactivated with formaldehyde (0.2%). No side effects attributable to the injection of viral oncolysate were observed; however, DH skin reactions were observed in most patients at the first injection of viral oncolysate. After an average follow-up period of 13.5 months, there was no significant difference between the two groups with respect to the probability of staying in remission or staying alive. In a critique of this work, Lindenmann (1979a) pointed out that although formalin treatment did not impair the immunogenicity of viral oncolysates in animal model studies, it is not known whether this finding can be extrapolated to human tumor antigens, which may be destroyed by formalin. Inactivation was deemed necessary to avoid the possibility of producing recombinant strains of influenza virus in patients who might accidentally acquire an infection with a common human influenza virus strain. Although it is theoretically possible that such a recombinant could give rise to a new pandemic variant, Lindenmann now feels that this is not a great danger. To investigate the possibility that the allogeneic immunizing material may not have had the same TATA as the patient’s own tumor, a pilot study is planned in which only autologous viral oncolysates will be used. It should be noted that Sinkovics and colleagues (described above) used a more intensive immunotherapy program and a different method of virus inactivation than were used in these studies. Green et af. (1975, 1976) designed a Phase 1 study to determine the toxicity and general effectiveness of a vaccine prepared from influenza virus-infected osteosarcoma tumor cells in osteosarcoma patients. The study included patients with measurable tumors who were resistant to several chemotherapeutic drugs. Twelve patients were chosen: 6 did not receive antitumor agents during or after the administration of the vaccine: the remaining 6 received intravenous methotrexate every 3 weeks while receiving the vaccine. The tumor lines used were established in culture from osteosarcoma tumor tissue taken from three of the patients in the study. The viral oncolysates used were produced with the AYPort Chalmers and B/Michigan strains of human influenza virus. No attempt was made to adapt virus strains to replicate in the tumor cells. The viral oncolysates were inactivated by treatment with 0.01% formalin. Two milliliters of the oncolysate vaccine were injected i.m. at 2-week intervals. All patients, including those receiving autologous tumor cell vaccine, developed antibodies to both tumor and virus antigens. There was a direct correlation between the responses to viral and tumor cell antigens, but no correlation between the titer of antibody and the clinical course of disease either before or after immunization. In assessing the
3 24
FAYE C. AUSTIN A N D CHARLES W. BOONE
cellular immune response of 6 patients with extensive disease it was found that an immune response was generated only in patients who received autologous tumor cell vaccine, leading to the initial conclusion that autologous vaccine was necessary to stimulate cell-mediated immunity. However, further studies on 6 patients with less advanced disease indicated that 3 out of 4 patients who received allogeneic vaccine also developed cell-mediated immune responses and that the responses to autologous and allogenic vaccines were comparable in these patients with minimal disease. The 6 patients with minimal disease were receiving high-dose methotrexate chemotherapy. However, there was no evidence of suppression of the immune response in these patients receiving concurrent chemotherapy and viral oncolysate vaccine. The response to immunization correlated with disease status during the immunization. It was postulated that although the drug may have suppressed what would have been an even better response in the absence of chemotherapy, the drug may also have controlled the tumor to some extent and permitted patients to respond better to the vaccine. On the basis of these results, the authors are planning prospective imrnunotherapy trials. The use of recently isolated influenza virus strains in the preparation of viral oncolysate raises the question of the effect of viral antibodies in the patients' serum on the stimulation of immunity by viral oncolysates. Several patients did not have preexisting antiinfluenza antibody and responded anamnestically to the viral antigen present in the vaccine. The overall effect of this response on the effectiveness of the viral oncolysate in stimulating tumor immunity was not assessed. Antibody to the virus is known to inhibit the release of virus from infected cells (Dowdle et ul., 1974) and also, by neutralizing virus infectivity, would protect the patient from any spread of infectious virus. The effect of preexisting virus antibody on the ability of viral oncolysates to stimulate protective tumor immunity in animal model systems varied with the system studied (Lindenmann and Klein, 1967a; Boone rt al., 1974). Cassel et al. (1977) and Murray et ul. (1977) have studied the use of NDV-augmented melanoma cell extracts in the treatment of malignant melanoma. Because of the known oncolytic capability and minimal pathogenicity of this strain of NDV (Cassel and Garrett, 1965, 1966), viral oncolysates were administered without inactivation of the virus. The viral infectivity titers of the oncolysates ranged from lo3.' to 105.55070 egg-infectious doses per 0.05 ml. Viral oncolysate was prepared from primary explants of autochthonous melanoma and administered as a 10fold concentrate prepared in an ultrafiltration cell. The average protein concentration was 39 mg/ml. One milliliter of the 10-fold concentrate was the product of approximately 8 x lo6 cells. In a typical schedule, the
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
325
patients received 2 ml of the concentrated oncolysate S.C. every week. Distinct changes were observed in cell-mediated immunity. Lymphocyte cytotoxicity against cultured allogeneic melanoma cells increased in all patients within 1 to 2 weeks. Most patients showed a tendency for sustained increase both in relative and absolute levels of circulating T lymphocytes. T-lymphocyte levels in melanoma patients not receiving immunotherapy were low and decreased progressively until the patient died. Of the 13 patients with metastatic melanoma who were treated with viral oncolysate, 6 patients showed a decrease in the size of skin nodules or diseased lymph nodes. However, visceral lesions were not favorably influenced to any marked degree. This reduction in size in skin nodules and a noticeable lack of increase in size of existing superficial lesions in other cases were considered to be uncommon occurrences by the investigators. They noted that although the course of disseminated melanoma is often unpredictable, there is an inexorable fatal termination. The one surviving patient had the least advanced disease of all those studied, being the only one with no evidence of visceral disease. The patient had been unsuccessfully treated by surgery and chemotherapy with recurrent disease for over 4 years. At the time of the published report, the patient had no evidence of disease. The primary consideration in this pilot study was the assessment of the safety of viral oncolysates prepared from NDV-infected melanoma cells, without subsequent virus inactivation. No adverse responses were encountered, including evidence of any adverse effect on the central nervous system or any shedding of virus into the environment. The overall findings of this study suggested that the use of viral oncolysates may augment the immune response to malignant melanoma and produce benefit to the patient. As with all other studies, it appears that the patients could have been better helped if oncolysate treatment had been instituted earlier in the disease, particularly after removal of major tumor burden. The investigators in this project plan to extend these studies to high-risk cases with limited regional lymph node involvement. Wallack et a / . (1977), after demonstrating the effectiveness of vaccinia oncolysates in the immunoprophylaxis and immunotherapy of SV40-induced mouse peritoneal macrophage tumors, extended these findings to the study of vaccinia virus-augmented human tumor cell vaccines for active specific immunotherapy of human cancer (Wallack, 1979a). Since the safety of the vaccine strain of vaccinia virus has been established, it was considered unnecessary to inactivate any live virus in the oncolysates. Vaccine was prepared from autochthonous tumor cells after surgical removal, followed by preparation of the cells as a single-cell suspension culture. The cells were infected with vaccinia virus in vitro and
326
FAYE C. AUSTIN A N D CHARLES W . BOONE
incubated for 96 hours while both tumor cell lysis and virus titer were monitored. The cells were harvested, homogenized, and suspended to effect a concentration equivalent to los cells per milliliter. Doses of 1 ml of the vaccinia oncolysate were injected intradermally (i.d.) every 2 weeks for 3 months. Thereafter, 1-ml booster doses were given every month until recurrence of the tumor or exhaustion of the vaccine supply. A total of 29 tumor patients were reported to receive vaccinia oncolysate immunotherapy, representing 10 cases of melanoma, 9 cases of colon carcinoma, 2 cases of gastric carcinoma, 1 case of fibrosarcoma, and 1 case each of cancer of the cervix, lung, liver, ovary, kidney, breast, and thyroid. Although these patients probably had preexisting antivaccinia antibodies (from smallpox vaccination), experiments with mice had shown that preimmunization with virus alone prior to administration of vaccinia oncolysate did not abrogate the immunogenicity of the oncolysate. Side effects of the therapy were minimal, except for pain and inflammation at the injection sites. The patients' responses to vaccinia oncolysate immunotherapy were summarized as follows: (1) 0 out of 29 patients had generalized vaccinia, allergy, anaphylaxis, or tumor growth at injection sites: (2) 20 out of 29 patients died; (3) 15 out of 29 patients had delayed hypersensitivity reactions at vaccine injection sites, and of those, 9 of the 15 remained alive with controlled tumor growth at the time of this report: (4) 9 out of 9 patients with controlled tumor growth had elevated vaccinia antibody titers-a measure of nonspecific immunostimulation. Although all patients involved in this study had advanced metastatic disease, they did show DH reactions to one or more recall antigens and had tumors that could be safely excised in the operating room. The initial Phase 1 trials demonstrated that the vaccinia oncolysate was safe, nontoxic, and may even have been effective in stimulating the patients' tumor-immune response (9 out of 29 patients with evidence of controlled tumor growth). Wallack (1979b) has begun randomized prospective trials to determine the true efficacy of this approach in the treatment of carcinoma of the colon and rectum, Studies are currently underway at M. D. Anderson Hospital to assess the efficacy of influenza virus-augmented tumor cell extracts in the treatment of carcinoma of the vulva and cervix (R. s. Freedman and J . Bowen, in preparation). Allogeneic cultured tumor cells derived from both a primary squamous carcinoma of the vulva and a primary squamous carcinoma of the uterine cervix were infected with influenza A (PR8) virus. Although the virus was not adapted to grow in these cells, both human tumor lines readily supported replication of the virus. As monitored by membrane immunofluorescence, maximal detectable surface antigenicity preceded maximal release of progeny virus by several hours.
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
327
The cells remained intact during this period, but plating studies showed that infected cells were unable to proliferate. Influenza virus-augmented tumor cell extract (1.5 mg/0.3 ml) was administered every 2 weeks. Seven patients received extracts of virus-augmented vulva1 carcinoma cells. All patients had been treated with a radical vulvectomy and lymphadenectomy, and had two or more positive regional lymph nodes. In 3 out of 7 cases, deep nodes were also involved. Five patients received X-ray therapy at 3 to 4 months after surgery. Historically, at the M. D. Anderson Hospital, patients with two or more positive lymph nodes have an 80% risk of recurrence, and 60% of recurrences will occur within 12 months of primary therapy. Preliminary evaluation of the patients who received virus-augmented immunotherapy indicates that 5 of the 7 have had no recurrence after more than 34 weeks. No patients have developed recurrence while on therapy, and local or systemic toxicity has not been observed. Tumor-enhancing antibodies were not detected in a lymphocyte-mediated cytotoxicity assay. A randomized prospective study is now in progress in which the effectiveness of radiation therapy plus virusaugmented extracts will be compared with that of radiation therapy alone in the treatment of squamous cervical carcinoma.
C. IMMUNODIAGNOSIS OF HUMAN CANCER WITH VIRUS-AUGMENTED SKINTESTANTIGENS: CLINICAL TRIALS The radioisotopic footpad assay in tumor-immune mice has proved to be a useful animal model for DH reactions to tumor cells and tumor cell extracts (Paranjpe and Boone, 1972; Boone er al., 1974, 1979b; Austin et al., 1977, and submitted for publication). In this assay, tumor-immune mice inoculated in a rear footpad with tumor extracts developed a DH response which was quantitated by determining the increase in leakage of radioiodine-labeled albumin from the blood vessels in the test foot as compared to the contralateral control foot. It appears that a significant proportion of cancer patients give positive DH skin test reactions to CM or soluble extracts from their own tumors. Some of these reports show that CM from allogeneic tumor cells also give positive cancer-specific skin test reactions (see review by Burdick et al., 1975). In studies of virus augmentation of tumor rejection antigens (TATA) in mice, we found that the DH response of tumor-immune mice to tumor cell CM was augmented if the CM were prepared from cells that had been infected with influenza virus (Boone et al., 1974, 1979b; Austin et al., 1977). Although CM extracts of uninfected E4 tumor cells did not induce protective tumor immunity, they did elicit a DH response in tumor-immune
328
FAYE C. AUSTIN A N D CHARLES W . BOONE
mice. However, as stated above, extracts of influenza virus-infected E4 tumor cells elicited a greater footpad DH response. This result was not obtained with egg-grown influenza virus alone or with CM from uninfected cells mixed with egg-grown virus (Austin et al., unpublished observation). We therefore turned to the evaluation of virus-augmented allogeneic tumor cell CM as cancer-specific skin test antigens in humans. We chose melanoma patients because the existence of a tumor tissue-specific antigen has been documented in this disease (Fass et al., 1970; Morton et a / . , 1970; Shiku et a / . , 1976; Liao et a / . , 1978). VSV was chosen as the augmenting virus because of its relatively low pathogenicity in humans (mild flulike syndrome; Fields and Hawkins, 1967). To evaluate crossreactivity of melanoma antigens, we used three different established cell culture lines of human melanoma. Since UV-inactivation of the virus did not abrogate virus-augmented antigenicity in the E4 tumor system in mice (Austin et a / . , in preparation), we used UV-irradiation to inactivate the VSV in these preparations. Unadapted virus sometimes took 4 to 5 days to produce a recognizable CPE. After one passage through the respective melanoma line, the virus always produced marked CPE within 24 hours. Such facile adaption may be due to the acquired membrane compatibility between virus and host cell membranes after tumor cell passage (Young and Ash, 1974). We found that the presence of bovine serum proteins in tumor cell extracts frequently caused a nonspecific reaction when present in skin test materials. The fetal bovine serum was therefore removed from the cell cultures and replaced with 5% human serum several days prior to virus infection. Irie rt a / . (1974) have shown that bovine antigens can be incorporated into the plasma cell membrane during growth of the cells in the presence of bovine serum. Virus-augmented human melanoma cell extracts produced positive skin tests in 41 out of 58 (71%) trials in 20 melanoma patients. Identical CM extracts from the same melanoma lines that had not been infected with VSV gave positive skin tests in 8 out of 58 (14%) trials in the same 20 melanoma patients. In 18 control patients with other cancers or normal volunteers, the virus-augmented extracts were positive in only 3 out of 58 (6%) trials. The virus-augmented CM extracts thus exhibited markedly greater sensitivity without significant loss of specificity, as compared to the unaugmented extracts, when used as tumor-specific melanoma skin test antigens (Boone et nl., 1978, 1979b). Several cultured cell lines of human lung and breast cancer are now being evaluated for their VSV-augmented antigenicity in DH responses of patients with the respective tumor type. The testing of various established cell lines of the same histological type will indicate any cross-reactivity between individual tumors of the same
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
329
type. Methods that have been developed for the isolation of SV40 TATA of mouse tumor cells (Austin et af., submitted for publication; see Section 111,A,2) are now being applied t o the preparation of soluble virus-augmented human tumor skin test antigens.
IV. Mechanisms of Virus Augmentation of TATA Activity
A. RELATIONSHIP O F VIRAL ANTIGENS TO HOST CELL ANTIGENS In contrast to earlier impressions, it is now generally assumed that the envelope of surface-budding viruses contains little or no host protein (Holland and Kiehn, 1970), but that the carbohydrates and lipids of the virus envelope are host-derived (Laver and Webster, 1966; Klenk and Choppin, 1969). Many virus infections result in the appearance of new virus-induced proteins in the cell surface membrane which can serve as foreign antigens in the host without further modification or after they have been glycosylated by host cell enzymes. Much work has been done to determine the effect of virus replication on cell surface membranes. Extensive reports characterizing influenza virus-induced proteins within the cell surface membrane have appeared (Lazarowitz et al., 1971; Hay, 1974; Meier-Ewert and Compans, 1974), and a vast literature describing influenza virus biosynthesis has been dealt with in several reviews (White, 1974: Compans and Caliguiri, 1975; Klenk et a/., 1975; Ehrnst and Sundqvist, 1976). An excellent summary of the surface antigens of virus-infected cells and methods for their detection are presented in a recent review by Burns and Allison (1977). During influenza virus infection, two nonstructural and seven structural proteins are synthesized. Of the latter, three are glycoproteins that occur on or in the plasma membranes of the infected cell and the envelope of intact virions: the hemagglutinin (HA), neuraminidase (NA), and membrane (M) proteins. The other four are internal proteins (White, 1974). HA appears to be synthesized on the rough endoplasmic reticulum. Glycosylation with glucosamine ensues, followed by migration of the molecule to the smooth endoplasmic membranes where glucose and other sugars are added (Lazarowitz ef nl., 1971: Compans, 1973). Posttranslational cleavage of the HA differs for various virus strains, is host dependent, and occurs in the plasma membrane (Compans, 1973: Lazarowitz et al., 1973). Influenza virus-induced antigens have been demonstrated to move in the plane of the plasma membrane and are present within the entire plasma membrane of the infected cells (Rutter and Mannweiler, 19731, a finding which is consistent with the currently accepted fluid
330
FAYE C. AUSTIN A N D CHARLES W . BOONE
mosaic model of cell membranes (Singer and Nicholson, 1972). It has been suggested that the nucleoprotein (NP) antigen, formerly known as S antigen, may also interact with the plasma membrane, since it is known that large quantities of NP antigen accumulate in the extracellular medium or allantoic fluid and must, therefore, in some way pass through the plasma membrane. N P can be detected by immunofluorescence on the plasma membrane of cultured cells several hours after infection (Virelizier et a / . , 1977). NDV and other paramyxoviruses have many similarities to the orthomyxoviruses (influenza viruses). These viruses have been shown to alter the host cell plasma membranes in several ways, including exposing receptor sites for lectins and inducing rearrangement of receptor sites (Rott et af., 1975: Reeve et a / . , 1975) and producing changes in the phospholipid content of the plasma membrane (Semmel et al., 1975). Semliki Forest virus is a member of the alphavirus group of togaviruses (formerly group A arboviruses), which contain only three or four structural proteins and are the simplest of the enveloped viruses. Viral protein is synthesized on membrane-associated ribosomes, and two proteins are found in the plasma membrane as glyoproteins. Budding of alphaviruses occurs primarily, although not exclusively, at the plasma membrane (reviewed in Fenner e f al., 1974: and in Burns and Allison, 1977). VSV, a well-studied rhabdovirus, codes for two viral proteins which are associated with the plasma membrane shortly after their synthesis. One is not glycosylated and probably makes little contribution to surface antigenicity of infected cells, while the other is a glycoprotein with HA activity (Wagner et a / . , 1971; David, 1973; Morrison and Lidish, 1975). All of the above-described viruses are known to mature by budding through the plasma membrane of the infected cell. By contrast, pox viruses, which are the largest animal viruses, acquire a lipid membrane coat not by budding through cell membranes but by a poorly understood process of membrane biogenesis in cytoplasmic virus factories (Dales and Mosbach, 1968; Ichibashi et al., 1971). Vaccinia virus enters cells by a process of direct fusion between the virus envelope and the plasma membrane, followed by rapid dispersion of viral antigens in the plasma membrane (Chang and Metz, 1976). During maturation, the virus migrates to the Golgi area and is internalized by Golgi cisternae. It then migrates to the cell surface where the plasma membrane and cisternal membrane fuse and eject the virus particle. The virus envelope and cell membranes are probably separated at all times (Dales and Mosbach, 1968). HA is produced on the surface of cells infected with pox virus, but unlike the situation for all other viruses that produce HA, it is found in areas away from where virus is being released (Ichibashi and Dales, 1971).
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
33 1
According to White (1974), there are at least four distinct mechanisms whereby virus-coded protein could be inserted into the plasma membrane of the infected cell. (1) Viral proteins could be inserted only into newly synthesized membranes. (2) They could be substituted in place of the membrane proteins which are being turned over. (3) Viral proteins could be initially incorporated into areas of cytoplasmic membranes which are noted for their vigorous synthetic activity, namely Golgi vesicles and adjacent areas of smooth membranes. HA and M proteins of influenza virus are found not only in plasma membranes but in other cytoplasmic smooth membranes as well (Holland and Kiehn, 1970; Compans, 1973; Klenk et al., 1975). (4) Viral proteins could be inserted into the plasma membrane by displacing existing protein. Much work has been done to study this latter mechanism, with varying results. Holland and Kiehn (1970) reported that influenza virus envelope proteins completely replaced host membrane proteins only in those spots where the virus buds out, while most of the original host protein was retained. Hay (1974) studied the incorporation of influenza (fowl plague) virus-specific polypeptides into the plasma membrane of infected chick cells and found no apparent displacement of any host cell membrane protein as the result of virus infection. Hecht and Summers (1976) have studied the interaction of VSV with histocompatibility antigens of mouse cells and found a loss of 70% of H-2 activity after VSV infection. A large part (75%) of this activity was recovered in the purified VSV. In experiments described above demonstrating augmented TATA activity after infection with VSV, Wise (1977) found a 50% decrease in H-2 activity of tumor cells after virus infection. By attempting to block the absorption of cytotoxic anti-H-2 antibodies to VSV-infected tumor cells by adding excess antiserum to VSV, no steric blocking of absorption of anti-H-2 antibodies was observed, indicating no spatial association between H-2 and VSV surface antigens. On the other hand, Schrader et al. (1975) studied the functional interaction of viral and H-2 antigens on the surface of infected tumor cells and found that these antigens copatch and cocap as hybrid antigens when exposed to specific antiserum. Since the insertion of virus proteins has been reported to cause a decrease in host cell histocompatibility antigens, it is reasonable to consider that virus infection may also decrease the concentration of TATA on the surface of tumor cells. The relationship between histocompatibility antigens and tumor antigens has been studied and an inverse relationship has been demonstrated between the two antigen concentrations (Haywood and McKhann, 1971; Ting and Herberman, 1971; Cikes et al., 1973; Fujimoto ef al., 1973; Tsakraklides et al., 1974). It has been suggested that tumor-specific antigens may be modified histocompatibility antigens
332
FAYE C. AUSTIN A N D CHARLES W . BOONE
(Germain et al., 1975; Invernizzi and Parmiani, 1975: Thomson et al., 1976). However, these interpretations are uncertain since several investigators have reported the separation of tumor antigens from histocompatibility antigens by affinity chromatography (Appella et ul., 1976; Henriksen et a f . , 1977) or by gel filtration (Siegert et al., 1977). Kon et ul. (1976) were able to distinguish between DH responses to tumor antigens and to H-2 antigens by a footpad swelling assay and found that only tumor antigens were detected in extracts of cultured tumor cells. They could not specify, however, if the tumor antigen inducing DH responses was the TATA since the tumor-rejection activity of the antigen was not assayed. In studies of melanoma patients, Pellegrino et ul. (1977) found no inverse relationship between HLA antigens and melanoma antigens. Therefore, it remains to be determined whether the appearance of viral antigens in the tumor cell surface membrane causes a decrease in the actual number of TATA molecules. In fact, Wise and Acton (1978) recently reported that infection of T-lymphoblastoid cells of AKR mice with VSV had no effect on the expression of Thy-1, H-2K, or gp70 surface antigens, but produced an increase in Gross cell surface antigens (GCSA) and p30 antigens. This selective increase in GCSA followed VSV maturation at the cell surface but preceded the cytopathogenic effect of this virus. It is interesting that enteroviruses, which have been studied for their oncolytic and oncotropic capabilities (described above), have not been shown to induce cell surface antigens in infected cells, even though cytoplastic membranes play an important role in their replication and assembly. Also no reports demonstrating virus-augmented TATA activity of enterovirus oncolysates have appeared. If virus-induced surface antigens are not formed in any detectable amount, then immune responses against the tumor cells in which the virus has replicated are not likely to be important. Neutralizing serum antibody appears to play the major role in defense against enterovirus infections (Allison, 1974). With the above background in mind, several general questions regarding the mechanism of virus augmentation of TATA may now be addressed. 1. Is fully adapted virus necessary to augment TATA activity? If the insertion of viral antigens in the tumor cell surface membrane, juxtaposed to the TATA of the tumor cell, were a sufficient condition to augment TATA activity, then the answer to this question would be “no.” In fact, a virus that could replicate just to the point of inducing a full complement of virus-induced surface antigens, but with a defect in late gene functions related to virus maturation, would be preferable to a fully adapted virus because it would eliminate the problem of infectious virus in the resultant
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
333
oncolysate. Several studies have demonstrated augmented TATA activity after incomplete virus replication (Hailer and Lindenmann, 1975: Green et a l . , 1976). However, it is not certain that the mere presence of virusinduced antigens is the only factor required for maximum augmentation of tumor antigens. For example, early work by Lindenmann (1970) demonstrated that although VSV-infection of tumor cells was complete at 24 hours after infection, maximum augmentation of TATA immunogenicity did not occur until 48 hours after infection. To ascertain the need for virus adaptation, it would be necessary to perform carefully controlled studies to compare augmentation of TATA immunogenicity at various time points after virus infection. That point in the virus replicative cycle which contributes most to augmenting TATA activity may vary with the type of virus studied. 2. Which is the augmented immrrnogen: The tumor-grown virus or thc virus-infected host-cell membrane;? Which virus-infected cell fraction is most immunogenic? Most studies have shown that the major portion of immunogenic activity resides in the plasma membrane fraction of virusinfected cells, with little activity in soluble cytoplasmic fractions or in highly purified virus grown in tumor cells. Original studies which demonstrated TATA activity in tumor-grown viruses did not claim to have purified the virus to any great extent (Lindenmann, 1971, 1973). However, in studies of an SV40-transformed tumor cell line, Ansel(l974) and colleagues (Ansel et al., 1977: Huet and Ansel, 1977) demonstrated immunogenicity of purified tumor-grown VSV and attributed this immunogenicity to the incorporation of host cell SV40 TATA into the lipid core of the virus. The role of lipid may prove important since it has been recently reported that lipophilic agents can increase the immunogenicity of chemically modified tumor cells and soluble antigens (Prager and Gordon, 1978). The role of intracellular membranes has not been well studied, but it would not be surprising if these membranes also possessed augmented TATA activity. The recent finding of TATA activity in nuclear fractions of SV40-transformed tumor cells (Anderson et al., 1977: Rogers et a l . , 1977) raises the question of whether this phenomenon is characteristic of SV40-transformed tumors only, of virus-induced tumors in general, or whether it may also have relevance to chemically induced or spontaneous tumors. The role of the host cell nucleus in the replicative cycle of influenza virus is still unclear. 3 . What is the nature of the association between iiiriis antigen and tumor antigen? Is there a special spatial relationship.?Are they coi~ulently linked? Can they be copurified and retain augmented TATA activity:) There have been conflicting results on the nature of the association of these antigens. Work in progress indicates that antigens solubilized from
334
FAYE C . AUSTIN A N D CHARLES W . BOONE
virus-infected tumor cells can be further fractionated while still retaining augmented TATA activity (Austin and Boone, unpublished results). Methods successful for purification and characterization of histocompatibility antigens have not been as useful for the isolation of tumor antigens. Most attempts to purify tumor antigens have been frustrated by low yields or loss of antigenic activity (see review by Law and Appella, 1975). It is important to keep in mind that soluble antigens may induce an altered immune response solely because of differences in the form in which they are presented to the host’s immune system. 4. What is the effect qf inrrctivntioii of infectious virus in viral oncolvsates? Most published studies claim little decrease of virus-augmented immunogenicity after inactivation of virus infectivity either by UV-irradiation or inactivation with formalin. However, further controlled studies would be helpful to establish more exactly what, if any, contribution to augmented activity is due to the infectious nature of the viru.s. In summary, many of the conclusions concerning the interaction of virus-induced antigens with host cell antigens have been reached by inductive reasoning from results produced in similar systems. These conclusions will remain conjectural until more definitive experiments have been performed that elucidate the interaction between the cell surface antigen induced by the augmenting virus and the TATA of the host tumor cell.
B. AUGMENTED INDUCTION RESPONSE
OF THE
PRIMARY ANTITUMOR IMMUNE
1 . Biological Mechanisms Several possible mechanisms have been considered to explain the augmented TATA activity of extracts of virus-infected tumor cells, including ( 1) simple adjuvant action, (2) chemical stabilization, (3) neuraminidase action, and (4) helper antigen activity. The role of the virus acting as a simple adjuvant was ruled out in early experiments which demonstrated that merely mixing egg-grown virus with homogenates of uninfected tumor cells did not augment the immunogenicity of the tumor cell extracts (Lindenmann and Klein, 1967a; Boone et a l . , 1974). The requirement for virus replication in the tumor cells to augment TATA immunogenicity has been repeatedly demonstrated. Boone ez al. (1974) considered the possibility that a structural alteration of the tumor cell plasma membrane might occur during viral replication that could result in stabilization of the TATA against chemical degradation following cell disruption. How-
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
335
ever, this mechanism was ruled out by the demonstration that mice made immunologically tolerant to influenza virus (by pretreatment with Cytoxan plus virus) could no longer be immunized against the tumor by injection with influenza virus-infected tumor homogenates. The role of the enzymic activity of the viral neuraminidase was also considered, since neuraminidase has been shown to unmask antigenic components of tumor cells (Simmons and Rios, 1972; Rios and Simmons, 1976). However, treatment of tumor cells with purified neuraminidase preparations did not result in augmented TATA activity (Boone et al., 1974). Many subsequent studies have involved viruses which do not possess neuraminidase activity, e.g., VSV (Lindenmann, 1970; Hakkinen and Halonen, 1971; Boone et al., 1974; Ansel, 1974; Wise, 1977). This is not to say, however, that the neuraminidase glycoprotein spike of the influenza viruses may not interact in some other antigenic fashion to augment TATA activity. It is interesting to note that Gillette and Boone (1976) found that influenza virus-infected tumor cell homogenates were more immunogenic than those prepared from VSV-infected tumor cells. Most of the evidence to date points to a helper antigen role for virus antigens in stimulating an increased response to TATA in extracts of infected tumor cells, as discussed by Mitchison (1970). In addition to the above-mentioned experiment demonstrating that mice made tolerant to influenza virus could no longer be immunized with virus-infected tumor cell homogenates, Boone et al. (1974) showed that priming mice with egg-grown virus reduced the degree of tumor immunity that could be induced with virus-infected tumor cell homogenates. Similar results were reported in a study of the immune response to vaccinia virus-augmented xenoantigens (Bandlow and Koszinowski, 1974). Although these findings are opposite to those of Lindenmann and Klein (1967a) who demonstrated increased induction of immunity in virus-primed mice, both sets of results demonstrate involvement of the host’s immune system in the virus augmentation phenomenon. The variable effects of priming with virus may be due to differences in dose, route, and timing of the priming and immunizing inoculations. The early finding that the addition of antivirus antibody abrogated the augmented immunogenicity of viral oncolysates (Lindenmann and Klein, 1967a) has been interpreted as evidence for juxtaposed virus antigens and tumor antigens functioning through the host’s immune system. Interestingly, Crum and McGregor (1977) recently demonstrated that soluble tumor-associated antigens were rendered immunogenic when incubated briefly with living BCG. Imrnunization with either BCG or tumor protein alone or with mixtures made under circumstances designed to impede association of tumor protein with BCG failed to stimulate an immune response. Their results imply
336
FAYE C. AUSTIN A N D CHARLES W . BOONE
that the immunopotentiating power of BCG was related in some way to the proximity of the BCG antigens and the tumor-associated antigens. The helper antigen effect of virus-induced cell surface antigens has been shown to augment the immune response to alloantigens and xenoantigens, as well as to tumor antigens. Infection with Sendai virus, vaccinia virus, and herpesvirus simplex type I enhanced the primary humoral immune response to Thy-1 surface antigens in virus-primed congenic mice. Both antiviral priming of the host and virus infection of the immunizing cell were necessary to obtain the full response (Lake and Bromberg, 1979; Bromberg et al., 1979). In addition, guinea pigs demonstrated an increased cell-mediated immunity against xenoantigens after active immunization with vaccinia virus-infected xenogeneic cells (Bandlow and Koszinowski, 1974). The reproduction of the virus in the cell against which the increased immune response was directed was a prerequisite for its augmenting effect. 2 . Intertiction of Vir-ris and Tiitnor- Antigens with the Host's Immirne
Systcm The details of the mechanism by which virus infection of tumor cells augments the immune response to their TATA remains to be worked out. We know for certain that during virus replication, virus-specific antigens are inserted into the host cell plasma membrane in close proximity to, or possibly chemically linked with, the TATA, and that the afferent limb of the host immune system, during the process of responding to the virus antigen, also responds more strongly to the adjacent TATA than it would ordinarily have done in the absence of the vicinal virus antigen. Enough facts are known to construct the general hypothesis that the TATA and virus antigens share one or more soluble or cellular mediators, both specific and nonspecific, of the afferent immune response to the virus antigens. For example, lymphocyte chemotactic factors, macrophage migration inhibition factor, and other lymphokines released by T lymphocytes responding specifically to the virus antigen could serve to locally concentrate lymphoid cells which would also be able to participate in the afferent response to the nearby TATA. Factors similar to "allogeneic effect factor" could also be involved in enhancing the response. As another example, macrophages activated by the virus antigen could phagocytose, process, and present to helper T cells both the virus antigen and TATA. The idea is tenable that virus antigens on the cell surface membrane, especially the hemagglutinins, are "sticky" glycoproteins which could
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
337
facilitate the binding of adjacent TATA to lymphocytes and/or macrophages. As pointed out previously (Section IV,B,l), in virus augmentation of TATA, the virus antigens appear to operate through the host’s immune system. Lindenmann (1979b) has proposed that virus infection could alter the “antigen presentation” of the TATA either by increasing their number, conformation, or topological association on the tumor cell or on the macrophage membrane after processing. However, if this were the case, the following facts would not have been determined. ( 1 ) Priming mice with virus alone enhances (Lindenmann, 1974; Rukavishnikova and Alekseyeva, 1976) or diminishes (Beverley et al., 1973; Boone et ul., 1974) the augmented immunogenicity of virus-infected tumor cell homogenates. (2) Homogenates of virus-infected cells to which antiviral antibody has been added do not show augmented TATA immunogenicity (Lindenmann, 1974: Rukavishnikova and Alekseyeva, 1976). (3) Mice made tolerant to influenza virus could no longer be immunized with virus-infected tumor cell homogenates (Boone rt ul., 1974). The mechanisms of interaction of the various helper and effector cells in an immune response are popular subjects in current immunology research, and new theories are constantly being presented and evaluated. (For a concise and clear review of today’s knowledge of the cellular basis of the immune response, see Golub, 1977.) We know from studies of humoral immunity that helper T cells can operate via close proximity to B cells. For example, T-cell interactions with the carrier portion of an antigen facilitate B-cell responses to the hapten portion. Close association of helper T cells with effector T cells on the same stimulator cell surface has also been shown to occur in the cell-mediated immune response. It seems reasonable to postulate that the presence of a virus-induced neoantigen on the cell surface could stimulate helper T cells to assist effector T-cell precursors in recognizing adjacent antigenic determinants such as the TATA. This attractive hypothesis is supported by reports that viruses have been shown to generate helper T-cell activity for antibody production in primary immunization against certain alloantigens which would otherwise fail to induce an immune response in congenic hosts (Lake and Bromberg, 1978; Bromberg et al., 1979). To induce this response, it was shown that the helper determinant must be carried on the same cell to be effective, but not necessarily on the same molecule. Also, priming with virus increased this effect. We would expect that the net result of immunizing a tumor-bearing host with virus-augmented TATA would be the generation of a larger, or
338
F A Y E C. AUSTIN A N D CHARLES W . BOONE
a more intensely reactive, population of effector lymphocytes and humoral factors that would be active against the TATA as it appears on tumor cells within the main tumor mass away from the site of immunization.
c. VIRUS-AUGMENTED DELAYEDHYPERSENSITIVITY S K I N TESTS The proposition that the presence of a strong helper antigen within the plasma membrane, closely associated with the tumor antigen, could stimulate the generation of cellular or soluble mediators that would synergistically intensify the DH response to the tumor antigen, as previously described for virus augmentation of a primary immune response, suffers from the fact that the 48-hour time period in eliciting a DH response may be too short a time to see any helpful effect from an intense primary response to the virus antigen. In the absence of further data, we favor the working hypothesis that the virus antigen may produce a helper effect by generating a mild acute inflammatory response, producing soluble and cellular mediators that intensify the DH response to the tumor antigen. This effect apparently still requires a close interaction of virus and tumor antigens, since only a slight augmentation effect is obtained if the virus is simply mixed with tumor cell extracts before injection (Austin et NI., unpublished observation).
V. Prospects for the Application of Virus-Augmented Tumor Antigens in lmmunodiagnosis and lmmunotherapy
A. IMMUNODIAGNOSIS
It is clear that without the existence of a TATA in a given cancer patient, both nonspecific immunotherapy ,such as with BCG, and specific immunotherapy with a TATA vaccine would be impossible. Yet the presence, strength, and cross-reactivity of TATA in the different tissue types and grades of human cancer is far from sufficiently documented. In our view, skin testing with virus-augmented tumor membrane extracts is a clinically simple way to carry out this documentation. We therefore feel that the initial goal of skin test antigen development should be to screen patients with different types of cancer for the presence, strength, and cross-reactivity of their TATA, with secondary emphasis on a concurrent exploration of the use of skin test antigens to diagnose cancer and monitor its clinical course. Obviously, virus augmentation of TATA
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
339
can be effective only when a TATA exists. High priority should be given to the development of stable, soluble, or homogeneously dispersed, purified cancer skin test antigens.
B. IMMUNOTHERAPY A small number of clinical immunotherapy trials using virus-augmented tumor cell extracts are currently in progress (see Table IV), and there are indications that more may follow. At present, results are too preliminary to draw any definite conclusions. Therapeutic effects are often assessed by various in vitro assays of the cell-mediated immune response. However, tumor-specific DH skin test reactions have been demonstrated to correlate better with the patients’ clinical status (Leventhal et al., 1972). We feel that skin testing with autologous or, somewhat less desirably, allogeneic non-virus-augmented tumor cell extracts should be performed prior to and concurrently with virus-augmented TATA immunotherapy. Conversion of a negative tumor-specific skin test to a positive one during the course of a therapeutic regimen would be highly valuable proof of the efficacy of virus-augmented TATA therapy. In the meantime, we await further clinical results with cautious enthusiasm.
VI. Summary
Although many animal tumor cells possess TATA on their surface membrane, their immunogenicity is lost when the tumor cells are disrupted. Stemming from studies of the virus therapy of cancer, it was found that infection of tumor cells with a lytic virus that induces viral cell-surface antigens will augment the TATA activity of the tumor cell homogenate, or oncolysate. Substantial data have indicated that the mechanism of virus augmentation operates through the host’s immune system. Positive results in immunoprophylaxis and immunotherapy with virus-augmented TATA in animal model systems have led to the initiation of several attempts at immunotherapy with virus-augmented TATA in human cancer patients. In addition, virus-augmented crude membrane extracts of cultured human tumor cells have been shown to elicit enhanced delayed hypersensitivity skin tests without any loss of skin test specificity. The results of immunotherapy clinical trials are still too preliminary to draw any definite conclusions, but initial results are encouraging.
340
FAYE C. AUSTIN AND CHARLES W . BOONE
ACKNOWLEDGMENT The authors wish to especially acknowledge and thank their friend and colleague, Dr. Noritoshi Takeichi, Cancer Institute, University of Hokkaido School of Medicine, Sapporo, Japan, for his critical experimentation and high level of productivity which did much to further the progress of research on virus augmentation of TATA.
REFERENCES Al-Ghazzouli, I. K., Donohoe, R. M., Huang, K. Y., Sass, B., Peters, R. L., and Kelloff, G. J . (1976). J. Inimrtnol. 117, 2239-2248. Allison, A. C. (1974). Transplmt. Rev. 19, 3-55. Anderson, J . L.. Martin, R. G., Chang, C., Mora, P. T., and Livingston, D. M. (1977). Virology 76, 420-425. Ansel, S. (1974). In/. J. Cancer 13, 773-784. Ansel, S., Huet C., and Tournier, P. (1977). I n / . J. Cancer 20, 51-60. Appella, E., Law, L . W., and Henriksen, 0. (1976). Cancer Res. 36, 3539-3544. Asada, T. (1974). Cancer 34, 1907-1928. Am. Austin, F. C., Takeichi, N., Robbins, M. L., and Boone, C. W. (1977). Annrr. MPC/. Soc. Microhiol. (Abstr.), 83. Axler, 1).A., and Girardi, A. J. (1970). Proc. Am. Assoc. Ccrncer Res. 11, 4. Bandlow, G., and Koszinowski, U. (1974). Arch. Cescrmre Virrr.~for.scIi.45, 122-127. Belisario, J. C., and Milton, G. W. (1961). Airst. J. Dawza/ol. 6, 113-118. Beverley, P. C. L., Lowenthal, R. M., and Tyrrel, P. A. J. (1973). I n / . J. Cancer 2, 212216. Bluming, A. Z., and Ziegler, J. L. (1971). Loncer 2, 105-106. Boone, C. W. (1972). Nail. Cancer Ins/. Monogr. 35, 301-307. Boone, C. W. (1974). Recent Re.s/tl/s Cancer Res. 47, 394-400. Boone, C. W., and Blackman, K. (1972). Cancer Res. 32, 1018-1022. Boone, C. W., and Gillette. R. W. (1977). In “Modulation of Host Immune Resistance in the Prevention or Treatment of Induced Neoplasias” ( M . A. Chirigos, ed.), GPO Fogarty lnternational Center Proceedings No. 28, pp. 191-194, Washington, D.C. Boone, C., Blackman, K., and Brandchaft, P. (1971). Natrrre (London) 231, 265-266. Boone, C. W., Orme, T . W., Blackman, K., and Gillette, R. (1973). J. Natl. Cancer Inst. 51, 1141-1144. Boone, C. W., Paranjpe, M., Orme, T., and Gillette, R. (1974). I n / . J. Cancer 13,543-551. Boone, C. W., Austin, F. C., Gail, M., Case, R., and Klein, E. (1978). Cancer 41, 17811787. Boone, C. W., Takeichi, N., Austin, F. C., Gotohda, E., Oikawa, T., and Gillette, R. (1979a). Cann Monogr. Cancer Res. (in press). Boone, C. W., Austin, F. C., Gail, M., Paranjpe, M., and Klein, E. (1979b). Gann Monogr. Cancer Res. (in press). Bromberg, J., Brenan, M., Clark, E. A., Lake, P.,Mitchison, N. A., Nakashima, I.. and Sainis, K. B. (1979). Gann Monogr. Cancer Res. (in press). Burdick, J . F., Wells, S. A., Jr., and Herberman, R. B. (1975). Surg. Cynecol. Ohsrtt. 141, 779-794. Burdick, K. H. (1960). Arch. Dermutol. 82, 438-439. Burdick, K. H., and Hawk, W. A. (1964). Cancer 17, 708-712.
VIRUS AUGMENTATION OF TUMOR C E L L EXTRACTS
34 1
Burns, W. H., and Allison, A. C. (1977). I n ”Virus Infection and the Cell Surface” (G. Poste and G. L. Nicholson, eds.), pp. 213-247. Elsevier North-Holland, Amsterdam. Cassel, W. A , , and Garrett, R. E. (1965). Cancer 18, 863-868. Cassel, W. A., and Garrett, R. E. ( 1966). J . Brtcteriol. 92, 792. Cassel, W. A,, and Garrett, R. E. (1967a). Crtncer 20, 433-439. Cassel, W. A., and Garrett, R. E. (1967b). Cancer 20,440-444. Cassel, W. A.. Murray, D. R., l’orbin, A. H., Olkowski, Z. L., and Moore, M. E. (1977). Cancer 40, 672-679. Chang, A., and Metz, D. H. (1976). J . G m . Virol. 32, 275-282. Cikes, M., Friberg, S., Jr., and Klein, G. (1973). J. Nut/. Crtncer I n s t . 50, 347-362. Compans, R. W. (1973). Virology 51, 56-70. Compans, R. W., and Caliguiri, L. A. (1975). In “Negative Strand Viruses” (B. W. J . Mahy and R. D. Barry, eds.), Vol. 2, pp. 573-594. Academic Press, New York. Crum, E. D., and McGregor, D. D. (1977). Int. J. Cancer 20, 805-812. Csatary, L. K. (1971). Lance1 2, 825. Dales, S., and Mosbach, E. H. (1968). Virology 35, 564-583. David, A . E. (1973). J. M o l Biol. 76, 135-148. DeMarchi, J. M., and Kaplan, A. S. (1977). 1. Virol. 23, 126-132. Dowdle, W. R., Downie, J . C., and Laver, W. G. (1974). J. Virol. 13, 269-275. Eaton, M. D., and Almquist, S. J. P. (1975). Proc. Soc. Exp. Biol. Med. 148, 1090-1094. . 1089Eaton, M. D., Levinthal, J. D., and Scala, A . R. (1967). J . Nntl. CNI?(’eI.I I I J I39, 1097. Eaton, M. D., Heller, J . A., and Scala, A. R. (1973). Cancer Res. 33, 3293-3298. Ehrnst, A., and Sundqvist, K.-G. (1976). Exp. Cell Biol. 44, 198-225. Eiselein, J., and Biggs, M. W. (1970). Cancer Res. 30, 1953-1957. Eiselein, J . E., Biggs, M. W.. and Walton, J . R. (1978). Cuncer Rcs. 38, 3817-3822. Evermann, J . F., and Burnstein, T. (1975). I n / . J . Cancer 16, 861-869. Fass, L., Herberman, R. B., Ziegler, J. L., and Kiryabuire, J . W. M. (1970). Lnncet I , 116-1 18. Fenner, F., McAuslan, B. R., Mims, C. A., Sambrook, J., and White, D. 0. (1974). “The Biology of Animal Viruses” (2nd ed.). Academic Press, New York. Fields, B. N., and Hawkins, K. (1967). N. Engl. J . M i d . 277, 989-994. Fujimoto, S., Chen, C. H., Sabbadini, E., and Sehon, A. H. (1973). J. In~rwoml.111, 10931100.
Gerber, A,, Sauter, C., and Lindenmann, J. (1973a). Arch. Geximte Viriisfor~ch.40, 137151.
Gerber, A., Sauter, C., and Lindenmann, J. (1973b). Arch. Giwtrntc, Vir/tsf0rsch. 40, 255264. Germain, R. N., Dorf, M. E., and Benacerraf, B. (1975). J. Exp. Med. 142, 1023-1028. Gillette, R. W., and Boone, C. W. (1976). Int. J. Cancer IS, 216-222. Golub, E. S. (1977). “The Cellular Basis of the Immune Response.” Sinauer, Sunderland, Massachusetts. Green, A. A., Webster, A. G., and Smith, K. (1975). Proc. Am. Assoc. Cancer Res. 16, 271. Green, A. A., Pratt, C., Webster, R. G., and Smith, K. (1976). Ann. N. Y. Acad. Sci. 277, 396-4 I I. Griffith, I. P., Crook, N . E., and White, D. 0. (1975). Br. J. Cancer 31, 603-613. Hakkinen, I., and Halonen, P. (1971). J . Natl. Cancer Inst. 46, 1161-1167. Haller, O., and Lindenmann, J. (1975). J . N u t / . Cancer Inst. 54, 459-464. Hay, A. J. (1974). Virology 60, 398-418.
342
FAYE C . AUSTIN A N D CHARLES W . BOONE
Haywood, G. R., and McKhann, C. F. (1971). J. Exp. M e d . 133, 1171-1187. Hecht. T. T., and Summers, D. F. (1976). J . Virol. 19, 833-845. Henriksen, 0.. Law, L. W., and Appella, E. (1977). J. N u t / . Crrncer I n s t . 58, 1785-1788. Holland, J . J.. and Kiehn, E. D. (1970). Scicvici. 167, 202-205. Huet, C., and Ansel, S. (1977). I n / . J . Cuncer 20, 61-66. Hunter-Craig, I., Newton, K. A., Westbury, G., and Lacey, B. W. (1970). Br. M i d . J . 2, 5 12-5 15. Ichibashi, Y., and Dales, S . (1971). Virology 46, 533-543. Ichibashi. Y . , Matsumoto, S., and Dales, S . (1971). Virology 46, 507-532. Illiger, H . J . , Sauter, C., and Lindenmann, J. (1975). Cancer Rc~s.35, 3623-3627. Invernizzi, G., and Parmiani, G. ( 1975). Nutrrrc, (London) 254, 713-714. Irie, R. F., Irie, K., and Morton, D. L. (1974).J. N o r / . L'otri.i>rI n s t . 53, 1545-1551. Klein, P. A. (1967). Pnthol. Microhiol. 30, 222-245. Klein, P. A. (19741. Arch. Gesumte Virrrsforsch. 45, 199-208. Klein, P. A., and Lindenmann, J . (1965). Pu//iol. Microhiol. 28, 698-703. Klenk, H.-D., and Choppin, P. W. (1969). Virology 38, 255-268. Klenk, H.-D., Wollert, W., Rott, R., and Scholtissek, C. (1975). In "Negative Strand Viruses" (B. W. J . Mahy and R. D. Barry, eds.). Vol. 2 , pp. 621-634. Academic Press, New York. Kobayashi, H. (1970). Actu Ptithol. J o p u n . 20, 441-450. Kobayashi, H. (1979). Ad,'. Cancer Res. 30, 279-299. Kobayashi, H., and Sendo, F. (1979). Gun/i Monogr. Cuncer Res. (in press). Kobayashi, H., Sendo, F. Shirai, T., Kaji, H., Kodama, T., and Saito, H. (1969). J . N u t / . Crrnc~~r I n s / . 42, 413-419. Kobayashi, H . , Sendo, F., Kaji, H., Shirai, T . , Saito, H., Takeichi, N., Hosokawa, M., and Kodama, T. (1970). J . Nltrl. Cuncer Inst. 44, I I- 19. Kobayashi. H., Kodama, T., and Gotohda, E. (1977). "Xenogenization of Tumor Cells" (Hokkaido Univ. Med. Libr. Ser.), Vol. 9. Hokkaido Univ. School of Medicine, Sapporo, Japan. Kon. N. D., Forbes, J . T., and Klein, P. A. (1976). I n / . J. Cancer 17, 613-618. Koprowski, H.. Love. R . , and Koprowska, I . (1957). Ti.n-. Rep, Biol. M d . 15, 559-576. Kuzumaki, N., and Kobayashi, H. (1976). 7rrrrisplontrrtion 22, 545-550. Kuzumaki, N., Fenyo, E. M., Giovanella, B. C., and Klein, G. (1978). I n / . J. Cuncw 21, 62-66. Lake, P., and Bromberg, J . (1979). Proc. I2th Lerrkocyto C r h r c , Conf. (in press). Laver, W. G., and Webster. R. G. (1966). Virology 30, 104-115. Law, L. W., and Appella, E. (1975). In "Cancer: A Comprehensive Treatise" (F. F. Becker, ed.), Vol. 4, pp. 135-154. Plenum Press, New York. Lazarowitz, S. G . , Cornpans, R. W., and Choppin, P. W. (1971). Virology 46, 830-843. Lazarowitz, S . G . , Cornpans, R. W., and Choppin, P. W. (1973). Virology 52, 199-212. Leventhal, B. G . , Halterman, R. H., Rosenberg, E. B., and Herberman, R. B. (1972). Currcvr Rcs. 32, 1820-1825. Liao, S.-K., Leong, S. P. L., Sutherland, C. M., Dent, P. B., Kwong, P. C., and Krementz, E. T. (19781. Cancer Rcs. 38, 4395-4400. Lindenmann, J . (1963). Pro(,. Soc. E.rp. Biol. M i d . 113, 85-91. Lindenmann, J . (1964). J . I n i n n r n ~ l 92, . 912-919. Lindenmann, J . (1970). Arch. G e s u n m Virrr.s:forsch.31, 61-70. Lindenmann, J . (1971). Proc. R. Soc. London, Ser. B 176, 419-423. Lindenmann, J . (19731. C'iho Swnp. 18, 197-215. Lindenmann, J . (1974). Biochini. Biophys. Acfo 355, 49-75.
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
343
Lindenmann, J. (1977). I n “Modulation of Host Immune Resistance in the Prevention or Treatment of Induced Neoplasias” (M. A. Chirigos, ed.), No. 28, pp. 187-190, GPO Fogarty International Center Proceedings, Washington, D.C. Lindenmann, J. (1979a). Gann Monogr. Cancer Res. (in press). Lindenrnann, J . (1979b). Garin Monogr. Cancer Res. (in press). Lindenrnann, J . , and Klein, P. A. (1967a). J . Exp. Med. 126, 93-108. Lindenmann, J . , and Klein, P. A. (1967b). Recent Resrtlts Cancer Res. 9. McMurtrey, M. J . , Campos, L. T., Sinkovics, J . G., Romero, J . J., Loh, K . K., and Romsdahl, M. M. (1976). I n “Neoplasms of t h e Skin and Malignant Melanoma,” pp. 471-484. 20th Clin. Conf. on Cancer, M. D. Anderson Hospital and Tumor Institute, Houston, Year Book Medical, Chicago. Meier-Ewert, H., and Compans, R. W. (1974).J. Virol. 14, 1083-1091. Milton, G. W., and Brown, M. M. L. (1966). Airst. N . Z . J . S ~ i r g .35, 286-290. Mitchison, N. A . (1970). Transplant. Proc. 2, 92-103. Moore, A. E. (1954).Annu. Re\*. Microbiol. 8, 393-410. Moore, A . E. (1960). Prog. ESP. Trimor Res. 1, 411-439. Morrison, T. G . , and Lodish, H. F. (1975). J . B i d . Chem. 250, 6955-6962. Morton, D. L., Eilber, F. R., Joseph, W. L., Wood, W. C., Trahan, E . , and Ketcham, A. S . (1970). Ann. Sirrg. 172, 740-749. Muceniece, A. J . (1972). “Oncotropism of Viruses and the Problems of Virotherapy of Malignant Tumors.” Riga Publishing House, Zinatne, Riga, Latvian SSR (In Russian, summary in English). Muceniece. A. J. (1978). Workshop on I~nmiinologiculX e n o g e ~ i i a t i o nof Trimor Cells. Sapporo, Japan, June, 1978, p. 33 (Abstr.). Murray, D. R., Cassel, W. A., Torbin, A. H., Olkowski, Z. L., and Moore, M. E. (1977). Cuiicer 40, 680-686. Paranjpe, M., and Boone, C. W. (1972). J. Nail Cutrccv Inst. 48, 563-566. Pasquinucci, G. (1971). Lancet 1, 136. Pellegrino, M. A,, Ferrone, S . , Reisfeld, R. A., Irie, R. F., and Golub, S. H . (1977). C u ~ i c u 40, 36-4 1 . Prager, M. D., and Gordon, W. C . (1978). Cirncer Res. 38, 2052-2057. Reeve, P., Poste, G., and Alexander, D. (1975). [ ti “Negative Strand Viruses” (B. W. J . Mahy and R. D. Barry, eds.), Vol. 2, pp. 859-873. Academic Press, New York. Rios, A., and Simmons, R. L. (1976). Ann. N. Y . A c a d . Sci. 276, 45-60. Roenigk, H. H., Deodhar, S., St. Jacques, R., and Burdick, K. (1974). Arch. Dermutol. 109, 668-673. Rogers, M. J., Law, L. W., and Appella, E. (1977).J . Natl. Cancer Inst. 59, 1291-1295. Rott, R., Becht, H., Hammer, G., Klenk, H.-D., and Scholtissek, C. (1975). I n “Negative Strand Viruses” (B. W. J. Mahy and R. D. Barry, eds.), Vol. 2, pp. 843-857. Academic Press, New York. Rukavishnikova, G. E., and Alekseyeva, A. K. (1976). Acta Virol. 20, 387-394. Rutter, G., and Mannweiler, K. (1973). Arch. Gesamte Virri.sforsch. 43, 169-172. Sauter, C., and Lindenmann, J. (1976). I n “Erkrankungen der Myelopoese” (A. Stacher and P. Hocker, eds.), pp. 280-281. Urban & Schwarzenberg, Munich, Berlin and Vienna (cited in Lindenmann, 1979a). Sauter, C., Gerber, A., Lindenmann. J., and Martz, G. (1972). Schweiz. Med. Wochenschr. 102, 285-290 (In German, summary in English). Sauter, C., Baumberger, U., Ekenbark, S., and Lindenmann, J . (1973). Cancer Res. 33, 3002-3007. Sauter, C., Bachi, T., and Lindenmann, J . (1975). G t r . J . Cancer 11, 59-63.
344
FAYE C . AUSTIN A N D CHARLES W . BOONE
Sauter, C., Cavalli, F., Lindenmann, J . , Gmur, J. P., Berchtold, W., Alberto, P., Obrecht, P., and Senn, H . J . (1978). In "Immunotherapy of Cancer: Present Status of Trials in Man" (W. D. Terry and D. Windhorst, eds.), pp. 355-363. Raven Press, New York. Schrader, J . W., Cunningham, B. A., and Edelman, G. M. (1975). Proc. N o t / . Acud. Sci. U.S.A. 72, 5066-5070. Sedmak, G. V . , Taylor, M. W., Mealey, J . , Jr., and Chen, T . T. (1972). Nutrrri, (London) New B i d . 238, 7-9. Semmel, M., Israel, A., and Audubert, F. (1975). I n "Negative Strand Viruses" (B. W. J. Mahy and R. D. Barry, eds.), Vol. 2, pp. 875-883. Academic Press, New York. Sendo, F., Kaji, H., Saito, H., and Kobayashi, H. (1970). Gunn 61, 223-226. Sharpless, G . R., Davies, M. C., and Cox, H. R. (1950). Proc. Soc. Exp. B i d . M d . 73, 270-275. Shiku, H., Takahashi, T., Oettgen, H . F., and Old, L. J. (1976). J. Exp. Med. 144, 873881. Siegert, W., Fenyo, E. M., and Klein, G. (1977). I n t . J . Cancer 20, 75-82. Simmons, R. L., and Rios, A. (1972). Srrrgery 71, 556-564. Singer, S. J., and Nicholson, G. L. (1972). ScicJnce 175, 720-731. Sinkovics, J . G. (1977). J . Am. Met/. Assoc. 237, 869. Sinkovics, J . G., Thota, H., Loh, K. K., Gonzales, F., Campos, L. T . , Romero, J . J . , Kay, H. D.. and King, D. K. (1974a). I n "Cancer Chemotherapy-Fundamental Concepts and Recent Advances," pp. 417-443. 19th Clin. Conf. on Cancer. Year Book Medical, Chicago. Sinkovics, J. G., Williams, D. E., Campos, L . T., Kay, H. D., and Romero, J. J . (1974b). Semin. O n d . I , 351-365. Sinkovics, J . G., Plager, C., McMurtrey, M. J . , Romero, J. J . , and Romsdahl, M. M. (1977). Proc. An?.Assoc. Crrnccv R P S .18, 86. Smith, R. R., Heubner. R. J . , R o w , W. P., Schatten, W. E.. and Thomas, L. B. (1956). Cuncxv 9, 1211-1218. Southam, C. M . (1960). Truns. N . Y . Acud. Sci. Ser. 2 22, 657-673. Suskind, R. G., Heubner, R. J . , Rowe, W. P., and Lowe, R. (1957). P r ~ cSoc. . Exp. H i d . M1.d. 94, 309-318. Svet-Moldavsky, G. (1979). Gunn Monogr. Cancer Res. (in press). Svet-Moldavsky, G . J., and Hamburg, V. P. (1964). Natrtw (London), 202, 303-304. Svet-Moldavsky, G. J., Gibadullin, R. A., Frotzova, A. E., Hamburg, V. P., Perevodchikova, N., and Bychkov, M. B. (1974). Vosprosi Virrrsol. 6, 713-719 (In Russian, Summary in English). Takeichi, N., Austin, F. C., Oikawa, T., and Boone, C. W. (1978). Ccinctr Res. 38, 45804584. Taylor, M. W., Davidson, J . N., Land, C., and Wall, R. (1970). J . N u t / . C ' u n w I n s t . 44, 515-519. Taylor, M. W., Cordell. B.. Souhrada, M., and Prather, S. (1971). Proc. Nut/. Acnd. Sci. U.S.A. 68, 836-840. Thomson, D. M . P., Gold, P., Freedman, S. 0.. and Shuster, J. (1976). Canccv Rev. 36, 3518-3525. Ting, C.-C., and Herberman, R. B. (1971). Nutiire N t w B i d . 232, 118-120. Tsakraklides, E., Smith, C., Hersey, J. H., and Good, R. A. (1974). J . N u t / . Cancw Inst. 52, 1499-1504. Virelizier, J.-L., Allison, A. C., Oxford, J.. and Schild, G. (1977). Nntrrrc (London) 266, 52-54.
VIRUS AUGMENTATION OF TUMOR CELL EXTRACTS
345
Wagner, R. R., Heine, J. W., Goldstein, G., and Schnaitrnan, C. A. (1971). J . Virol. 7 , 274-277. Wallack, M. K. (1979a). Gann Mongr. Cancer Res. (in press). Wallack, M. K. (1979b). J . Surg. Oncol. (in press). Wallack, M. K., Steplewski, Z., Koprowski, H., Rosato, E . , George, J., Hulihan B., and Johnson, J . (1977). Cancer 39, 560-564. Webb. H. E., and Gordon Smith, C. E. (1970). Lancet 1, 1206-1209. Webb, H. E., Wetherley-Mein, G., Gordon Smith, C. E., and McMahon, D. (1966). Rr. Mecl. J . 1, 258-266. White, D. 0. (1974). Cirrr. Top. Microhiol. Irnrnrtnol. 63, 1-48. Wise, K . S. (1977). J. N u / / . Cuncer I n s t . 58, 83-90. Wise, K. S., and Acton, R. T. (1978). Proc. A m . Assoc. Cancer Rcs. 19, 137. Young, N . P., and Ash, R. J . (1974). Appl. Microhiol. 28, 26-30.
This Page Intentionally Left Blank
SUBJECT INDEX A
Acute phase reactant proteins (ACRP), 143 i n animal tumors systems, 26-27 biological effects of, 28-30 in bladder cancers, 15-17 in breast cancer, 8- 10 extrahepatic synthesis and concentration of. 25-26 in GI tract cancers, 13-15 haptoglobin phenotypic variation of, 2728 in leukemia, 20-23 in liver cancer, 19 in lung cancer, 17- 19 in lymphomas, 20-23 as markers. 3 mathematical analysis of levels of, 30-37 cluster analysis. 36-37 discriminant analysis, 33-35 multivariate methods. 33-37 principal component analysis, 35-36 in peritoneal and pleural effusions, 23-24 production and half-lives of, 4-8 in prostate cancer, 8- 10 Adenomatous polyps, intestinal carcinogenesis and, 183 Animals. tumor systems phase reactant proteins of, 26-27 Antigens of intestinal tumors, 21 1-216 from virally infected tumor cells, 280287 increase of antigenicity of, 287 A-Rad leukemia virus. characteristics of, 67-68 B
Bladder cancer, acute phase reactant proteins in. 15-17 347
Bone marrow. role in leukemia induction, 49-52 Breast cancer acute phase reactant proteins in, 8- 10 hormone dysfunction in, 132- 133
C Cachexia, in cancer patient, 93 Cell-mediated immunity, in intestinal carcinogenesis. 214-2 I6 Cancer ( w e trlso Tumors) acute phase reactant proteins in, 1-43 hormonal tumor syndrome in, 92 immunodiagnosis of, by virus-augmented tumor antigen, 327-329, 338-339 virus therapy of. 303-307 Cancer embryonal antigen (CEA), in colon tumors, 21 I Carbohydrate metabolism, in tumor-host relationship, 93- 102 Carcinogenesis, hydrazine role in, 151- I64 Colon, antigens in lesions of, 213 D
Delayed hypersensitivity skin test, virusaugmented, 338 Diet, in intestinal carcinogenesis, 190- 193 Dimethylh ydrazine interaction, with cell components, 922226 as intestinal carcinogen, 168-169, 216226 modification, 221-222 tumor types, 171 metabolic pathways of, 219 DNA of herpesviruses, 259-262 in transformed cells, 264-266 DNA synthesis, in tumor hosts, 109- 112
348
SUBJECT I N D E X
D-Rad leukemia virus. characteristics of, 66-67 E
Enzymes in intestinal carcinogenesis, 207-208 in tumor hosts, 1 15- 125 Enteroviruses, in cancer therapy, 306-307 Environment, hydrazine in, I53 Epithelial tumors, classification of, 170 Epstein-Barr virus antigens from, 258-266 in human tumors, 247 properties of, 243 tissue-culture studies of, 249-258 Estrogens, intestinal carcinoma and, 213
G Gastrointestinal tract cancers, 165-237 acute phase reactant proteins in. 13- IS Genetics in intestinal carcinogenesis, 184- 185 of leukemia induction, 75-78 Glucose, tumors as traps for, 93-102, 139I40 Glycogen reserves. tumor effects on, 96 Glycoproteins, ACRPs as, 1-43 Gynecological cancers, acute phase reactant proteins in, 15- 17 H
Herpesvirus(es) (lymphotropic), 239-278 antigens from, 258-266 DNA of, 259-262 in transformed cells, 264-266 experimental tumor studies on, 248-249 lytic infections of. 251 in natural hosts, 244-249 persistant and latent infections from, 246-248 proteins of, 258-259 survey of, 240-244 tissue culture studies of, 249-258 transformation by, 25 1-258 viral diseases from, 244-246 H c q w s v i r u s ofi4i>.s.properties of, 24 I , 243
H(vpc,sviru.s of chimpanzee. properties of, 243 HrrpcJsvirus p ~ i p i properties ~. of, 243 Hcrpesvirus pongo, properties of, 243 Hrrpe.sviru.s suiiniri (HVS). 239 antigens from, 258-266 in natural host, 244-248 properties of, 242 tissue-culture studies of, 249-258 H r r p t w i r u s of turkeys (HVT), properties of. 240-24 I , 242 Hcrprsvirrts .s.dvilngu.s. properties of, 240, 242 Hodgkin's disease, hydrazine therapy of, 161 Hormonal tumor syndrome, 92 Hormones disorders in tumor hosts, 127-133 ectopic, from cancer patients, 92 Hyperthermia, as tumor therapy, 138- 139 Hydrazine( s) i n carcinogenesis. 151- 164 experiments on, 155-157 in humans, 159-160 mechanism, 158- 159 degradation of, 156 effects on liver. 152-153 in environment, 153 in Hodgkin disease therapy, 161 intestinal polyps and tumors from, 158 toxic effect of, 152
I
Immunodepression, in tumors, 133- 137 Immunology, of intestinal carcinogenesis, 189-190, 211-216 Influenza virus tumor cell antigenicity from. 3 10, 3 12315 mechanisms, 329 therapeutic use. 321 Intestinal cancer, 165-237 age factors in, 187 biochemistry of, 206-21 I cells of, kinetics of. 201-204 dietary factors in, 190- 193 electron microscopy of, 174- 177 enzyme role in, 208-21 I
349
SUBJECT INDEX
experimental models for, 166- I69 transplants, 169 tumor induction, 166- 169 factors affecting, 184- 196 genetic factors in. 184- 185 histochemistry of. 177- 179 hydrazine-induced, 158 immunological aspects of, 189- 191, 21 1. 216 antigens, 212-213 cell-mediated immunity, 2 14-216 intestinal function factors, 193- 197 kinetics of cell division in 196-201 microbiological aspects of, 187- 189 morphogenesis of, 179-184 morphology of, 169- 184 mucosal injury and, 196 nuclear proteins function in, 207-208 nucleic acid function in, 206-207 polyp role in, 183 sex hormone role in, 185-187 stem epithelial cell role in. 204-206 Infections, with cancer, 89, 92 Isonicotinylhydrazide as carcinogen, 153- 155 in humans. 159-160
L
Lactic dehydrogenase (LDH), in tumor hosts, 124 Leukemia acute phase reactive proteins in, 20-23 genetic control of, 75-78 induction of, in mice, 45-87 age-related susceptibility. 52-54 by leukemia virus, 62-81 thymus and bone marrow role, 49-52, 54-57 virus radiation induction of, 57-59 Leukemia virus radiation induction of, 57-59 variants of, 64-69 A-Rad LV, 67-68 D-Rad LV, 66-67 Rad LV, 64-66 RS-Rad LV, 68-69 target cells for, 69-72
Lipid metabolism, in tumor hosts, 112- I 15 Liver. hydrazine effects on, 152-153 Liver cancer, acute phase reactive proteins in, 19 Lung cancer, acute phase reactant proteins in, 17-19 Lymphoid tumors, radiation-induced, 5961 Lymphomas. acute phase reactive proteins in, 20-23 Lymphotropic herpesviruses, 239-278 M
Marek's disease virus (MDV), 239 antigens from, 258-266 in natural host, 244-248 properties of, 242 tissue-culture studies of, 249-258 Mathematic analysis, of acute phase reactant protein levels, 30-37 Methylhydrazine(s) as cytotoxins, 160 natural occurrence of, 153 Mice. leukemia induction in. 45-87 Microbiology, of intestinal carcinogenesis, 187- I89 Mucosa, intestinal, injury to, in carcinogenesis, 196 Mumps virus, i n cancer therapy. 307 Mushrooms, hydrazine in. 153 N
Newcastle disease virus tumor cell antigenicity from, 3 I I mechanism, 330 therapeutic use, 321 Nitrogen metabolism, in tumor patients, 102- I12 Nuclear proteins, in intestinal carcinogenesis, 207-208 Nucleoside kinases, in tumor hosts. 121
P Peritoneal effusions, acute phase reactive proteins in, 23-24
350
SUBJECT INDEX
Pleural effusions. acute phase reactive proteins in, 23-24 Polyps, adenomatous, intestinal carcinogenesis and, 183, 193- 194 Prostate cancer, acute phase reactant proteins in, 8- 10 R
Rad leukemia virus characteristics of, 64-66 immunization by, 72-75 phases of, 78-81 Radiation, leukemia induction by, 45-87 RNAs, in tumor host tissue, 125- 127 RNP complexes, in tumor host tissue, 12s- 127 RS-Rad leukemia virus, target cells for, 69-72
S Semliki Forest virus tumor cell antigenicity from, 3 I I mechanisms, 330 Sendai virus, tumor cell antigenicity from, 311 Sex hormones, in intestinal carcinogenesis. 185- I87 Small intestine tumors, antigens, 212 Stem epithelial cells, in intestinal carcinoge ne si s, 204- 206 T Tumor-associated transplantation antigen (TATA) viral augmentation of. 301-345 mechanisms, 329-338 Tumor cells antigenicity of virus augmentation, 301-341 animal studies, 308-320 human studies, 305-309, 320-327 immune response, 334-338 viral xenogenization of, 279-299 antigens of. 280-287 definition of, 286-287 immunogenicity increase in, 290 therapeutic aspects, 291-292
Thymus role i n leukemia induction. 49-52 phenotypic characteristics, 54-57 Thyroid function, in cancer patients, 132 Tissue culture. in herpesvirus studies. 249258 Tobacco. hydrazine in, 153 Transformation. by herpesviruses, 251-258 Tumors and tumor hosts ( w e trlso Cancer) cells of. sec’ Tumor cells competitive relationships with host, 93I so clinical aspects. 137- 139 effect on host tissue biology, 1 IS- 137 enzymes. 1 15- I25 RNAs, 125-127 as glucose traps. 93- 102 hormone disorders in, 127- 133 host relationships of, 89- 150 hyperthermia therapy of, 138-139 immunodeppression in, 133- 137 lipid metabolism in, 112- I IS markers for, list, 5 nitrogen metabolism i n , 102- I I2 therapy of, using viral xenogenization. 29 1-292
V
Vaccinia virus, in cancer therapy, 307. 311. 321 Vesicular stomatitis virus tumor cell antigenicity from. 3 I I mechanisms, 330 Viruses cancer therapy by. 303-307 in animals, 303-309 in humans, 305-309, 320-327 nature of, 318-320 leukemia, radiation induction of, 57-59 xenogenization of tumor cells by, 279299 X
Xenogenization of tumor cells, by viruses. 279-299 immune response against. 292-295
CONTENTS OF PREVIOUS VOLUMES
Volume 1 Electronic Configuration and Carcinogenesis C . A . Coirlsoii
Epidermal Carcinogenesis E . V . Coitdry The Milk Agent in the Origin of Mammary Tumors in Mice L. D l l l O C ~ l l O l t ~ S ! i i Hormonal Aspects of Experimental Tumorigenesis T . U . Grirrinu Properties of the Agent of Rous No. I Sarcoma R. .IC .. Hrrrris Applications of Radioisotopes to Studies of Carcinogenesis and Tumor Metabolism Cho rlr s H c,itlolhcrgcr
The Carcinogenic Aminoazo Dyes J f r r i w s A . Millrr r i i i i l Eli~rihrrhC'. M i l l ~ r The Chemistry of Cytotoxic Alkylating Agents M. C . .I. Ross Nutrition in Relation to Cancer Alhcrr Toiirrc~rihairriiairrl Herhut Silwrstorlr
Plasma Proteins in Cancer Ric.liarcl .I. Wiiizler A U T H O R INDEX-SUBJECT INDEX
Volume 2 The Reactions of Carcinogens with Macromolecules
Carcinogenesis and Tumor Pathogenesis I . Br~rerrblur?~ Ionizing Radiations and Cancer Austin M . Brrws Survival and Preservation of Tumors in the Frozen State Jrinir.s Craigic,
Energy and Nitrogen Metabolism in Cancer Lroirarcl D. FcirilingPr aiid G. Burroirghs Mitier
Some Aspects of the Clinical Use of Nitrogen Mustards C'rrlviii T . K l o p p cirri1 Jcurlir(, C . Bcrf(,ilraii Genetic Studies in Experimental Cancer L . w. La113 The Role of Viruses in the Production of Cancer C . Ohorlirig uiid M . Giwriii Experimental Cancer Chemotherapy C . Ch(,stor Stock A U T H O R INDEX-SUBJECT INDEX
Volume 3 Etiology of Lung Cancer Richirril Doll The Experimental Development and Metabolism of Thyroid Gland Tumors Harold P . Morris Electronic Structure and Carcinogenic Activity and Aromatic Molecules: New Developments A . Pulliiiuii o r i d B . P ~ I / / I ~ I N I I Some Aspects of Carcinogenesis
P. Rollclorli
P r t w Al~xrrriilcr
Pulmonary Tumors in Experimental Animals Mic~haolB. Shimkiii
Chemical Constitution and Carcinogenic Activity G . M . Brrclgrv
35 I
352
CONTENTS OF PREVIOUS VOLUMES
Oxidative Metabolism of Neoplastic Tissues Sidm,y W~,irlhousc, AUTHOR INDEX-SUBJECT INDEX
Volume 4 Advances in Chemotherapy of Cancer in Man Siclnc,y Fnrhrr. RiiiIolf Toch. Edicwd Muririirig Siwrs. urrd Doririld PiriAc4
The Use of Myleran and Similar Agents in Chronic Leukemias D. A . G . ~ n / t f J l l The Employment of Methods of Inhibition Analysis in the Normal and TumorBearing Mammalian Organism Ahrahnrii Goldiri Some Recent Work on Tumor Immunity P. A. Gowr
Inductive Tissue Interaction in Development Clij/orcl Grohs/cirr
Lipids in Cancer Frciricc,.~L . Hui’cii nrril W . R . Bloor The Relation between Carcinogenic Activity and the Physical and Chemical Properties of Angular Benzacridines A . L U C U S S UN~. IP.~ ~ BUN . H o i . R . Dciiit l c ~ l .nnd F. Zujclc,lii The Hormonal Genesis of Mammary Cancer 0.Mithlhoc~X AUTHOR INDEX- SU BJ ECT INDEX
Volume 5
The Newer Concept of Cancer Toxin Wnro Ntiknhurn a r i d FiiriiiXo Fuhiiokri
Chemically Induced Tumors of Fowls P. R . Pcci COCA Anemia in Cancer V i t i ~ t ~Ei .r P r i c , arid ~ Rohrrt E . Gri~c~rrjicltl Specific Tumor Antigens L . A . Zilher Chemistry, Carcinogenicity. and Metabolism of 2-Fluorenamine and Related Compounds Eliz(rhrth K . Weishurgi,r arid Johii H . Wc,i.shrrrgc,r AUTHOR INDEX-SUBJECT INDEX
Volume 6 Blood Enzymes in Cancer and Other Diseases Oscur Bodansky The Plant Tumor Problem Arriiiri C. Bruurr u i i c l Hidrrry N . Wood Cancer Chemotherapy by Perfusion Osc.iir C’rc,cc,h. Jr. tirrtl Et1liw-d T . KrcIill’lltZ
Viral Etiology of Mouse Leukemia Liit1nYi.k
Cross
Radiation Chimeras P . C. KolliJr. A . J . S. Dniics. urrtl S h c 4 r r M . A . Docik
Etiology and Pathogenesis of Mouse Leukemia J . F . A . P. Millcr Antagonists of Purine and Pyrimidine Metabolites and of Folic Acid C . M . Tii?irrris Behavior of Liver Enzymes in Hepatocarcinogenesis Grorgi) W r h r r
Tumor-Host Relations R . W . Bcgg
AUTHOR INDEX-SUBJECT INDEX
Primary Carcinoma of the Liver Chur1e.s B ~ r ~ t ~ i i i i
Protein Synthesis with Special Reference to Growth Processes both Normal and Abnormal P . N . C‘cirirphcll
Volume 7 Avian Virus Growths and Their Etiologic Agents .I. W . Beard
CONTENTS O F PREVIOUS VOLUMES
353
Mechanisms of Resistance to Anticancer Agents R . W . Brocknian Cross Resistance and Collateral Sensitivity Studies in Cancer Chemotherapy Dorris J . Hutchison Cytogenic Studies in Chronic Myeloid Leukemia W. M. Court Brotiln a d Ishbel M . T o u g h Ethionine Carcinogenesis Etiiiiiuiiud Furber Atmospheric Factors in Pathogenesis of Lung Cancer Paul Kotin and H a n s L. Falh Progress with Some Tumor Viruses of Chickens and Mammals: The Problem of Passenger Viruses G. NPgroIii
The Relation of the lmmune Reaction to Cancer Louis v. coso Amino Acid Transport in Tumor Cells R . M . J o h n s t o n c and P . G . S c h o l i $ d d Studies on the Development, Biochemistry. and Biology of Experimental Hepatomas Harold P . Morris Biochemistry of Normal and Leukemic Leucocytes, Thrombocytes. and Bone Marrow Cells 1. F . Seitz
AUTHOR INDEX-SUBJECT INDEX
Carcinogens. Enzyme Induction. and Gene Action H . V . Crlhoiti I i i Vitro Studies on Protein Synthesis by Malignant Cells A . Clark Gr(fli’ti The Enzymatic Pattern of Neoplastic Tissue w. Eigc,no Knox Carcinogenic Nitroso Compounds P. N. Mngc’c, crtrtl .I. M. Brrrnos The Sulfhydryl Group and Carcinogenesis J . S . Harriiigtoii The Treatment of Plasma Cell Myeloma Daniel E . Bcrgsccgel, K . M . Gri’jji’th, A . Haul. arid W . .I. StricLlcy. J r .
Volume 8 The Structure of Tumor Viruses and Its Bearing on Their Relation to Viruses i n General A . F . Hoii~atsoti Nuclear Proteins of Neoplastic Cells Harris Birsch and Willirrtii J . Stc’i40 Nucleolar Chromosomes: Structures. Interactions, and Perspectives M. J . K o ~ c Jatid ~ , Glaclys M . Mw,cyXo Carcinogenesis Related to Foods Contaminated by Processing and Fungal Metabolites H . F . Kraybill crntl M . B . SIiin~Xin Experimental Tobacco Carci noge ne si s Enic,st L. Wyntlrr r i n d Dic,tric,h H o j $ i i ~ i i AUTHOR INDEX-SUBJECT INDEX
Volume 9 Urinary Enzymes and Their Diagnostic Value in Human Cancer Ric~k~irrl Stn/llba/rgh ( / , I d SidIiCy WOiIi/1ou.sr
AUTHOR INDEX- SU BJ ECT INDEX
Volume 10
AUTHOR INDEX-SUBJECT INDEX
Volume 11 The Carcinogenic Action and Metabolism of Urethran and N-Hydroxyurethan Sitltic~yS . M i n i s / / Runting Syndromes. Autoimmunity, and Neoplasia
D. K m s t Viral-Induced Enzymes and the Problem of Viral Oncogenesis Saul Kit
354
CONTENTS O F PREVIOUS VOLUMES
The Growth-Regulating Activity of Polyanions: A Theoretical Discussion of Their Place in the intercellular Environment and Their Role in Cell Physiology Williuttl Rcgrlsoil Molecular Geometry and Carcinogenic Activity of Aromatic Compounds. New Perspectives Josc~plrC'. Arc.os otrtl Mury F.Argus AUTHOR INDEX-SUBJECT
INDEX
CUMULATIVE INDEX
Volume 12 Antigens Induced by the Mouse Leukemia Viruses G. P a s t r r i d Immunological Aspects of Carcinogenesis by Deoxyribonucleic Acid Tumor Viruses G . 1. Dc~ii~hititrtt Replication of Oncogenic Viruses in VirusInduced Tumor Cells-Their Persistence and Interaction with Other Viruses H . Hutiajirsa Cellular Immunity against Tumor Antigens Karl Erik Hc,llstrtiin atrrl Iiigiyyril Hcllstram
Perspectives in the Epidemiology of Leukemia Irving L . Kessler uritl Ahraliarn M . Lilienfeld AUTHOR INDEX-SUBJECT INDEX
Volume 13 The Role of lmmunoblasts in Host Resistance and lmmunotherapy of Primary Sarcomata P . Alexander and J . G . Hall Evidence for the Viral Etiology of Leukemia in the Domestic Mammals Osu~alilJarrett
The Function of the Delayed Sensitivity Reaction as Revealed in the Graft Reaction Culture Huii~iGiiishurg Epigenetic Processes and Their Relevance to the Study of Neoplasia Gqjutlaii V . Sherbet The Characteristics of Animal Cells Transformed i t ~Vitro ~ ( 1 1 7 Moc.plii,r.soii Role of Cell Association i n Virus Infection and Virus Rescue J . Sl~ohoilaN I I 1. ~ Hlolrinc~k Cancer of the Urinary Tract D . B . C'luysori arid E . H . Coop1.r Aspects of the EB Virus M. A . Epstritc AUTHOR INDEX-SUBJECT INDEX
Volume 14 Active lmmunotherapy GiwrgcJs Math4 The Investigation of Oncogenic Viral Genomes in Transformed Cells by Nucleic Acid Hybridization E n i t s t Winocorrr Viral Genome and Oncogenic Transformation: Nuclear and Plasma Membrane Events Gcnrgc Meycr Passive lmmunotherapy of Leukemia and Other Cancer Rolund Motta Humoral Regulators in the Development and Progression of Leukemia Donald Metcalf' Complement and Tumor Immunology Kusuya Nishioka Alpha-Fetoprotein in Ontogenesis and Its Association with Malignant Tumors G . 1. Abidev Low Dose Radiation Cancers in Man Alice Stritwrr AUTHOR INDEX-SUBJECT INDEX
CONTENTS O F PREVIOUS VOLUMES
Volume 15 Oncogenicity and Cell Transformation by Papovavirus SV40: The Role of the Viral Genome J . S . Bittc,l. S . S . T(,rvthia. anil J . L . Mc.1nick
Nasopharyngeal Carcinoma (NPC) J . H. C. H o Transcriptional Regulation in Eukaryotic Cells A . J . MucGillit,ra.v, J . Pairl. arid G. Thrclfill1
Atypical Transfer RNA's and Their Origin in Neoplastic Cells Erric,.st Borpk anil Sylvia J . Kerr Use of Genetic Markers to Study Cellular Origin and Development of Tumors in Human Females Philip .I. Fialkoit, Electron Spin Resonance Studies of Carcinogenesis Harold M . Sttwrtz Some Biochemical Aspects of the Relationship between the Tumor and the Host V . S . Shapot Nuclear Proteins and the Cell Cycle Gary Stein and Renaro Baserga AUTHOR INDEX-SUBJECT INDEX
355
1,3-Bis(2-Chloroethyl)- I-Nitrosourea (BCNU) and Other Nitrosoureas i n Cancer Treatment: A Review St(,phcn K. Cartcv. Frank M . Schahcl. Jr.. Laii,rc,nc,c, E. Brodcr, and Thorvus P. Johnston AUTHOR INDEX-SUBJECT INDEX
Volume 17 Polysaccharides in Cancer: Glycoproteins and Glycolipids Vijai N . Nigani and Antonio Cantc,ro Some Aspects of the Epidemiology and Etiology of Esophageal Cancer with Particular Emphasis on the Transkei, South Africa Gi,rald P. Warli-ick and John S . Harington
Genetic Control of Murine Viral Leukemogenesis Frank Lilly and Th~wclorc~ Pincus Marek's Disease: A Neoplastic Disease of Chickens Caused by a Herpesvirus K. Nazmriun Mutation and Human Cancer Alfred G . Knudson. Jr. Mammary Neoplasia in Mice S . Nandi and Charles M. McGruth AUTHOR INDEX-SUBJECT INDEX
Volume 16 Polysaccharides in Cancer Vijai N . Nigam arid Antonio Cantero Antitumor Effects of Interferon Ion Gresscu Transformation by Polyoma Virus and Simian Virus 40 Joc, Sanibronk Molecular Repair, Wound Healing, and Carcinogenesis: Tumor Production a Possible Overheding? Sir Alexander Haddobiz The Expression of Normal Histocompatibility Antigens in Tumor Cells Alena LengiJrovd
Volume 18 Immunological Aspects of Chemical Carcinogenesis R . W . Baldtisin lsozymes and Cancer Fanny Schapira Physiological and Biochemical Reviews of Sex Differences and Carcinogenesis with Particular Reference to the Liver Yec Chu Toh Immunodeficiency and Cancer John H . Kersey, Beatrice D. Spector. and Robert A . Good
356
CONTENTS OF PREVIOUS VOLUMES
Recent Observations Related to the Chemotherapy and Immunology of Gestational Chonocarcinoma K . D . Bogsliaw Glycolipids of Tumor Cell Membrane Stti-itiroh Hokomori Chemical Oncogenesis in Culture C‘harlcs Hi,idelhivger AUTHOR INDEX-SUBJECT INDEX
Principles of Immunological Tolerance and lmmunocyte Receptor Blockade G . .I. V . N o s s a l The Role of Macrophages in Defense against Neoplastic Disease Mic hac4 H. Lei:\’ untl E . Fretlorich WhiJeloch
Epoxides in Polycyclic Aromatic Hydrocarbon Metabolism and Carcinogenesis P . Sims and P. L . Grovc,r Virion and Tumor Cell Antigens of C-Type RNA Tumor Viruses Hviiiz Bairn
Volume 19 Comparative Aspects of Mammary Tumors J . M. Hamilrori The Cellular and Molecular Biology of RNA Tumor Viruses, Especially Avian Leukosis-Sarcoma Viruses, and Their Relatives Howwrcl M . Ttviiin Cancer, Differentiation, and Embryonic Antigens: Some Central Problems J . H . Coggin. Jr. and N . G . Aiitlrrsoii Simian Herpesviruses and Neoplasia Frcdrich W . Deinharilt. Laitwiice A . Falk. and L a u r ~ nG . Wo@ Cell-Mediated Immunity to Tumor Cells Ronald B . Herbermmi Herpesviruses and Cancer F r d Rapp Cyclic AMP and the Transformation of Fibroblasts Ira Pastan arid Gtorge S . Johtison Tumor Angiogenesis Jutlah Folkman SUBJECT INDEX
Volume 20 Tumor Cell Surfaces: General Alterations Detected by Agglutinins Annette M. C . Rapin and Max M . Burgcv
Addendum to “Molecular Repair, Wound Healing, and Carcinogenesis: Tumor Production a Possible Overhealing?” Sir Ali,xatrdcr Hadcloi~~ SUBJECT INDEX
Volume 21 Lung Tumors in Mice: Application to Carcinogenesis Bioassay Micliad 8 . Shimhin arid Gary D. Storriv Cell Deazh in Normal and Malignant Tissues E. H . CoopcJr. A . J . Bi~lfi,rd,and T . E . Keiiiiy
The Histocompatibility-Linked Immune Response Genes Baruj Bc~naccrraf’mil David H . Karz Horizontally and Vertically Transmitted Oncornaviruses of Cats M . Essc,x Epithelial Cells: Growth in Culture of Normal and Neoplastic Forms K t e f A . Rafferty. Jr. Selection of Biochemically Variant, in Some Cases Mutant, Mammalian Cells in Culture G . B. Clemmis The Role of DNA Repair and Somatic Mutation in Carcinogenesis James E . Trosko and Ernest H . Y . Chu SUBJECT INDEX
CONTENTS OF PREVIOUS VOLUMES
357
Volume 22
Volume 24
Renal Carcinogenesis J . M . Hatnilton Toxicity of Antineoplastic Agents in Man: Chromosomal Aberrations, Antifertility Effects. Congenital Malformations, and Carcinogenic Potential Susan M . Sii+iv ant1 Richard H . Atlutnson Interrelationships among RNA Tumor Viruses and Host Cells Ragrnond V . GiIdi>n Proteolytic Enzymes, Cell Surface Changes, and Viral Transformation Richard Rohlin. lih-Nan Chorr. atiil Ptritl H . Black lmmunodepression and Malignancy
The Murine Sarcoma Virus-Induced Tumor: Exception or General Model in Tumor Immunology'? J . P . Lc,iy und .I. C . Lrclarc, Organization of the Genomes of Polyoma Virus and SV40 Mike Fried arid Bcivrl! E . G r i j j / n @,-Microglobulin and the Major Histocompatibility Complex
Osins Stutrllnn SUBJECT INDEX
Volume 23 The Genetic Aspects of Human Cancer W . E . Heston The Structure and Function of Intercellular Junctions in Cancer Ronald S . Weinstein. Frcdv-ick B . Mcrk. and Joscph Alroy Genetics of Adenoviruses Harold S . Gitistwrg and C . S . H . Young Molecular Biology of the Carcinogen, 4-Nitroquinoline I-Oxide Minuko Nagao and Takashi Sugimuru Epstein-Barr Virus and Nonhuman Pnmates: Natural and Experimental Infection A . Frank, W . A . Anditnun. arid G. Miller Tumor Progression and Homeostasis Richmond T . Prehn Genetic Transformation of Animal Cells with Viral DNA or RNA Tumor Viruses Miroslav Hill and Jana Hillova SUBJECT INDEX
P u A . Pctorson. Lars Rusk. untl Lurs osth(,rg
Chromosomal Abnormalities and Their Specificity in Human Neoplasms: An Assessment of Recent Observations by Banding Techniques Joachirn Mark Temperature-Sensitive Mutations in Animal Cells Cluuilio Basilicw Current Concepts of the Biology of Human Cutaneous Malignant Melanoma WullacP H . Clark. J r . . Mic~hnrlJ . Mosrrangdo. Ann M. Ainsii.orth. Dai*icl B u d , Rohivt E . B~ll1,t.und Ei~rliriu A . Bcrtiurtlirio SUBJECT INDEX
Volume 25 Biological Activity of Tumor Virus DNA F. L . Graham Malignancy and Transformation: Expression i n Somatic Cell Hybrids and Variants H a r v q ~L . O z t r and Krishna K . Jha Tumor-Bound Immunoglobulins: 111 Sitrr Expressions of Humoral Immunity Isaac P . Witz TheAh Locus and the Metabolism of Chemical Carcinogens and Other Foreign Compounds Snorri S . Thorgc.irssori and Danirl W . Nebert
358
CONTENTS OF PREVIOUS VOLUMES
Formation and Metabolism of Alkylated Nucleosides: Possible Role in Carcinogenesis by Nitroso Compounds a n d Alkylating Agents Aritlioriy E . Pt'gg Immunosuppression and the Role of Suppressive Factors in Cancer Isno k'niri(~~ i i t Hcrrtiii/i l Fricilrtrcrri Passive lmmunotherapy of Cancer in Animals and Man Sti,wrr A . Ro.sijirhi.rg n r i t l Willicrrii I). Tiwy
The Choice of Animal Tumors for Experimental Studies of Cancer Therapy Hirroltl B . H r n , i t t Mass Spectrometry in Cancer Research Johrr Robor Marrow Transplantation in the Treatment of Acute Leukemia E . Doriiitill Tlioirius. C . D~arr Buc,!,tii,r. Alc.wriclc,r Fqfi.r. Purr1 E . Nc,iriirrrl. crriil Rciiiior Storh Susceptibility of Human Population Groups to Colon Cancer Mrirtiri Lipkiii
SUBJECT INDEX
Natural Cell-Mediated Immunity Roiicrltl B . Hrrh(,rrirciii o r i d Hoiiuril T . H(11tlc~ri
Volume 26
SUBJECT INDEX
The Epidemiology of Large-Bowel Cancer P e I ~ y oCorroo trritl Willicriir Htierr.sz,cl Interaction between Viral and Genetic Factors in Murine Mammary Cancer . I . HilRors urid P. B i , i i t w I w i i Inhibitors of Chemical Carcinogenesis L w W . W(rito/ihi,rg Latent Characteristics of Selected Herpesviruses Jl71~AG. s/(,lY,rr.s Antitumor Activity of Coryrirhric,/c,riirr?r puri,rr/ir mil(/.^ t / / i t l Mortiir
Li/l\(i
T. S ~ o t t
SUBJECT INDEX
Volume 28 Cancer: Somatic-Genetic Considerations F . M . Brrrrict Tumors Arising in Organ Transplant Recipients Isrriol Peirii Structure and Morphogenesis of Type-C Retroviruses Roriciltl C . Moiitclnro rrrril Doni P . Bologr10.si
BCG in Tumor lmmunotherapy Rohrrt W . B d d i ~ , i ~uriil i Mokolin
Volume 27 Translational Products of Type-C RNA Tumor Viruses Johir R . S t r p h e n s o i i , Srr.sliilkuiiior G. Devnrc. nritl Fri4 H . R ~ y i i ( ~ l J~r1. . ~ . Quantitative Theories of Oncogenesis Alii,i, S. Whittomore Gestational Trophoblastic Disease: Origin of Choriocarcinoma. lnvasive Mole and Choriocarcinoma Associated with Hydatidiform Mole. and Some Immunologic Aspects J . 1. Brcwrr, E . E . Toro!,. B . D. Kaliciii, C . R . Sititiliopo. irriil B . HolpiJrri
V.
Pillllll
The Biology of Cancer Invasion and Metastasis lsaitrh .I. Fitllc,r. DoriR1u.s M . G r r s t ( ~ t i r. ~ r t l luii R . Hnrt Bovine Leukemia Virus Involvement in Enzootic Bovine Leukosis A . Burtry. F . Bcx. H . Chcriitreirtie, Y . Clcrrfrr. D. D r h c ~ g c ~.I.l . Ghysdiic.1. R . K C ~ / I I I ~ IM I I.I ILivli~rcq, . J . Leiirr(,ri. M . Mnrnriiorickx. criril D . Portotc,llc Molecular Mechanisms of Steroid Hormone Action Stopheii .I. Higgiris atid Ulriclr Gc,lrririg SUBJECT INDEX
CONTENTS OF PREVIOUS VOLUMES
Volume 29
Influence of the Major Histocompatibility Complex on T-cell Activation J . F . A . P . Millrr Suppressor Cells: Permitters and Promoters of Malignancy'! Dmid Nuor Retrodifferentiation and the Fetal Patterns of Gene Expression in Cancer Josc' Urii4 The Role of Glutathione and Glutathione STransferases in the Metabolism of
359
Chemical Carcinogens and Other Electrophilic Agents L . F . Chus.sc~uut1 a-Fetoprotein in Cancer and Fetal Development Erl,ki Roo,sluhti uric1 Morkkrr Srppiilii Mammary Tumor Viruses Dutr H . Moore. Curolr A . Long. AXhil B . Viiiclw. Joc4 B . Shejjii4tl. Ar/iold S . Diou. n r i t l E/ioitic, Y . L u s f u r g u ~ s Role of Selenium i n the Chemoprevention of Cancer A . C ' l d Grqjirr SUBJECT INDEX
This Page Intentionally Left Blank